1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Changing options at runtime with a command
318 Some options can be changed during the operation of the filter using
319 a command. These options are marked 'T' on the output of
320 @command{ffmpeg} @option{-h filter=<name of filter>}.
321 The name of the command is the name of the option and the argument is
325 @chapter Options for filters with several inputs (framesync)
326 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
328 Some filters with several inputs support a common set of options.
329 These options can only be set by name, not with the short notation.
333 The action to take when EOF is encountered on the secondary input; it accepts
334 one of the following values:
338 Repeat the last frame (the default).
342 Pass the main input through.
346 If set to 1, force the output to terminate when the shortest input
347 terminates. Default value is 0.
350 If set to 1, force the filter to extend the last frame of secondary streams
351 until the end of the primary stream. A value of 0 disables this behavior.
355 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
357 @chapter Audio Filters
358 @c man begin AUDIO FILTERS
360 When you configure your FFmpeg build, you can disable any of the
361 existing filters using @code{--disable-filters}.
362 The configure output will show the audio filters included in your
365 Below is a description of the currently available audio filters.
369 A compressor is mainly used to reduce the dynamic range of a signal.
370 Especially modern music is mostly compressed at a high ratio to
371 improve the overall loudness. It's done to get the highest attention
372 of a listener, "fatten" the sound and bring more "power" to the track.
373 If a signal is compressed too much it may sound dull or "dead"
374 afterwards or it may start to "pump" (which could be a powerful effect
375 but can also destroy a track completely).
376 The right compression is the key to reach a professional sound and is
377 the high art of mixing and mastering. Because of its complex settings
378 it may take a long time to get the right feeling for this kind of effect.
380 Compression is done by detecting the volume above a chosen level
381 @code{threshold} and dividing it by the factor set with @code{ratio}.
382 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
383 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
384 the signal would cause distortion of the waveform the reduction can be
385 levelled over the time. This is done by setting "Attack" and "Release".
386 @code{attack} determines how long the signal has to rise above the threshold
387 before any reduction will occur and @code{release} sets the time the signal
388 has to fall below the threshold to reduce the reduction again. Shorter signals
389 than the chosen attack time will be left untouched.
390 The overall reduction of the signal can be made up afterwards with the
391 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
392 raising the makeup to this level results in a signal twice as loud than the
393 source. To gain a softer entry in the compression the @code{knee} flattens the
394 hard edge at the threshold in the range of the chosen decibels.
396 The filter accepts the following options:
400 Set input gain. Default is 1. Range is between 0.015625 and 64.
403 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
404 Default is @code{downward}.
407 If a signal of stream rises above this level it will affect the gain
409 By default it is 0.125. Range is between 0.00097563 and 1.
412 Set a ratio by which the signal is reduced. 1:2 means that if the level
413 rose 4dB above the threshold, it will be only 2dB above after the reduction.
414 Default is 2. Range is between 1 and 20.
417 Amount of milliseconds the signal has to rise above the threshold before gain
418 reduction starts. Default is 20. Range is between 0.01 and 2000.
421 Amount of milliseconds the signal has to fall below the threshold before
422 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
425 Set the amount by how much signal will be amplified after processing.
426 Default is 1. Range is from 1 to 64.
429 Curve the sharp knee around the threshold to enter gain reduction more softly.
430 Default is 2.82843. Range is between 1 and 8.
433 Choose if the @code{average} level between all channels of input stream
434 or the louder(@code{maximum}) channel of input stream affects the
435 reduction. Default is @code{average}.
438 Should the exact signal be taken in case of @code{peak} or an RMS one in case
439 of @code{rms}. Default is @code{rms} which is mostly smoother.
442 How much to use compressed signal in output. Default is 1.
443 Range is between 0 and 1.
448 This filter supports the all above options as @ref{commands}.
451 Simple audio dynamic range compression/expansion filter.
453 The filter accepts the following options:
457 Set contrast. Default is 33. Allowed range is between 0 and 100.
462 Copy the input audio source unchanged to the output. This is mainly useful for
467 Apply cross fade from one input audio stream to another input audio stream.
468 The cross fade is applied for specified duration near the end of first stream.
470 The filter accepts the following options:
474 Specify the number of samples for which the cross fade effect has to last.
475 At the end of the cross fade effect the first input audio will be completely
476 silent. Default is 44100.
479 Specify the duration of the cross fade effect. See
480 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
481 for the accepted syntax.
482 By default the duration is determined by @var{nb_samples}.
483 If set this option is used instead of @var{nb_samples}.
486 Should first stream end overlap with second stream start. Default is enabled.
489 Set curve for cross fade transition for first stream.
492 Set curve for cross fade transition for second stream.
494 For description of available curve types see @ref{afade} filter description.
501 Cross fade from one input to another:
503 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
507 Cross fade from one input to another but without overlapping:
509 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
514 Split audio stream into several bands.
516 This filter splits audio stream into two or more frequency ranges.
517 Summing all streams back will give flat output.
519 The filter accepts the following options:
523 Set split frequencies. Those must be positive and increasing.
526 Set filter order for each band split. This controls filter roll-off or steepness
527 of filter transfer function.
528 Available values are:
553 Default is @var{4th}.
556 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
559 Set output gain for each band. Default value is 1 for all bands.
566 Split input audio stream into two bands (low and high) with split frequency of 1500 Hz,
567 each band will be in separate stream:
569 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
573 Same as above, but with higher filter order:
575 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500:order=8th[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
579 Same as above, but also with additional middle band (frequencies between 1500 and 8000):
581 ffmpeg -i in.flac -filter_complex 'acrossover=split=1500 8000:order=8th[LOW][MID][HIGH]' -map '[LOW]' low.wav -map '[MID]' mid.wav -map '[HIGH]' high.wav
587 Reduce audio bit resolution.
589 This filter is bit crusher with enhanced functionality. A bit crusher
590 is used to audibly reduce number of bits an audio signal is sampled
591 with. This doesn't change the bit depth at all, it just produces the
592 effect. Material reduced in bit depth sounds more harsh and "digital".
593 This filter is able to even round to continuous values instead of discrete
595 Additionally it has a D/C offset which results in different crushing of
596 the lower and the upper half of the signal.
597 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
599 Another feature of this filter is the logarithmic mode.
600 This setting switches from linear distances between bits to logarithmic ones.
601 The result is a much more "natural" sounding crusher which doesn't gate low
602 signals for example. The human ear has a logarithmic perception,
603 so this kind of crushing is much more pleasant.
604 Logarithmic crushing is also able to get anti-aliased.
606 The filter accepts the following options:
622 Can be linear: @code{lin} or logarithmic: @code{log}.
631 Set sample reduction.
634 Enable LFO. By default disabled.
645 This filter supports the all above options as @ref{commands}.
649 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
653 Remove impulsive noise from input audio.
655 Samples detected as impulsive noise are replaced by interpolated samples using
656 autoregressive modelling.
660 Set window size, in milliseconds. Allowed range is from @code{10} to
661 @code{100}. Default value is @code{55} milliseconds.
662 This sets size of window which will be processed at once.
665 Set window overlap, in percentage of window size. Allowed range is from
666 @code{50} to @code{95}. Default value is @code{75} percent.
667 Setting this to a very high value increases impulsive noise removal but makes
668 whole process much slower.
671 Set autoregression order, in percentage of window size. Allowed range is from
672 @code{0} to @code{25}. Default value is @code{2} percent. This option also
673 controls quality of interpolated samples using neighbour good samples.
676 Set threshold value. Allowed range is from @code{1} to @code{100}.
677 Default value is @code{2}.
678 This controls the strength of impulsive noise which is going to be removed.
679 The lower value, the more samples will be detected as impulsive noise.
682 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
683 @code{10}. Default value is @code{2}.
684 If any two samples detected as noise are spaced less than this value then any
685 sample between those two samples will be also detected as noise.
690 It accepts the following values:
693 Select overlap-add method. Even not interpolated samples are slightly
694 changed with this method.
697 Select overlap-save method. Not interpolated samples remain unchanged.
700 Default value is @code{a}.
704 Remove clipped samples from input audio.
706 Samples detected as clipped are replaced by interpolated samples using
707 autoregressive modelling.
711 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
712 Default value is @code{55} milliseconds.
713 This sets size of window which will be processed at once.
716 Set window overlap, in percentage of window size. Allowed range is from @code{50}
717 to @code{95}. Default value is @code{75} percent.
720 Set autoregression order, in percentage of window size. Allowed range is from
721 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
722 quality of interpolated samples using neighbour good samples.
725 Set threshold value. Allowed range is from @code{1} to @code{100}.
726 Default value is @code{10}. Higher values make clip detection less aggressive.
729 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
730 Default value is @code{1000}. Higher values make clip detection less aggressive.
735 It accepts the following values:
738 Select overlap-add method. Even not interpolated samples are slightly changed
742 Select overlap-save method. Not interpolated samples remain unchanged.
745 Default value is @code{a}.
750 Delay one or more audio channels.
752 Samples in delayed channel are filled with silence.
754 The filter accepts the following option:
758 Set list of delays in milliseconds for each channel separated by '|'.
759 Unused delays will be silently ignored. If number of given delays is
760 smaller than number of channels all remaining channels will not be delayed.
761 If you want to delay exact number of samples, append 'S' to number.
762 If you want instead to delay in seconds, append 's' to number.
765 Use last set delay for all remaining channels. By default is disabled.
766 This option if enabled changes how option @code{delays} is interpreted.
773 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
774 the second channel (and any other channels that may be present) unchanged.
780 Delay second channel by 500 samples, the third channel by 700 samples and leave
781 the first channel (and any other channels that may be present) unchanged.
787 Delay all channels by same number of samples:
789 adelay=delays=64S:all=1
794 Remedy denormals in audio by adding extremely low-level noise.
796 This filter shall be placed before any filter that can produce denormals.
798 A description of the accepted parameters follows.
802 Set level of added noise in dB. Default is @code{-351}.
803 Allowed range is from -451 to -90.
806 Set type of added noise.
819 Default is @code{dc}.
824 This filter supports the all above options as @ref{commands}.
826 @section aderivative, aintegral
828 Compute derivative/integral of audio stream.
830 Applying both filters one after another produces original audio.
834 Apply echoing to the input audio.
836 Echoes are reflected sound and can occur naturally amongst mountains
837 (and sometimes large buildings) when talking or shouting; digital echo
838 effects emulate this behaviour and are often used to help fill out the
839 sound of a single instrument or vocal. The time difference between the
840 original signal and the reflection is the @code{delay}, and the
841 loudness of the reflected signal is the @code{decay}.
842 Multiple echoes can have different delays and decays.
844 A description of the accepted parameters follows.
848 Set input gain of reflected signal. Default is @code{0.6}.
851 Set output gain of reflected signal. Default is @code{0.3}.
854 Set list of time intervals in milliseconds between original signal and reflections
855 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
856 Default is @code{1000}.
859 Set list of loudness of reflected signals separated by '|'.
860 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
861 Default is @code{0.5}.
868 Make it sound as if there are twice as many instruments as are actually playing:
870 aecho=0.8:0.88:60:0.4
874 If delay is very short, then it sounds like a (metallic) robot playing music:
880 A longer delay will sound like an open air concert in the mountains:
882 aecho=0.8:0.9:1000:0.3
886 Same as above but with one more mountain:
888 aecho=0.8:0.9:1000|1800:0.3|0.25
893 Audio emphasis filter creates or restores material directly taken from LPs or
894 emphased CDs with different filter curves. E.g. to store music on vinyl the
895 signal has to be altered by a filter first to even out the disadvantages of
896 this recording medium.
897 Once the material is played back the inverse filter has to be applied to
898 restore the distortion of the frequency response.
900 The filter accepts the following options:
910 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
911 use @code{production} mode. Default is @code{reproduction} mode.
914 Set filter type. Selects medium. Can be one of the following:
926 select Compact Disc (CD).
932 select 50µs (FM-KF).
934 select 75µs (FM-KF).
940 This filter supports the all above options as @ref{commands}.
944 Modify an audio signal according to the specified expressions.
946 This filter accepts one or more expressions (one for each channel),
947 which are evaluated and used to modify a corresponding audio signal.
949 It accepts the following parameters:
953 Set the '|'-separated expressions list for each separate channel. If
954 the number of input channels is greater than the number of
955 expressions, the last specified expression is used for the remaining
958 @item channel_layout, c
959 Set output channel layout. If not specified, the channel layout is
960 specified by the number of expressions. If set to @samp{same}, it will
961 use by default the same input channel layout.
964 Each expression in @var{exprs} can contain the following constants and functions:
968 channel number of the current expression
971 number of the evaluated sample, starting from 0
977 time of the evaluated sample expressed in seconds
980 @item nb_out_channels
981 input and output number of channels
984 the value of input channel with number @var{CH}
987 Note: this filter is slow. For faster processing you should use a
996 aeval=val(ch)/2:c=same
1000 Invert phase of the second channel:
1002 aeval=val(0)|-val(1)
1009 Apply fade-in/out effect to input audio.
1011 A description of the accepted parameters follows.
1015 Specify the effect type, can be either @code{in} for fade-in, or
1016 @code{out} for a fade-out effect. Default is @code{in}.
1018 @item start_sample, ss
1019 Specify the number of the start sample for starting to apply the fade
1020 effect. Default is 0.
1022 @item nb_samples, ns
1023 Specify the number of samples for which the fade effect has to last. At
1024 the end of the fade-in effect the output audio will have the same
1025 volume as the input audio, at the end of the fade-out transition
1026 the output audio will be silence. Default is 44100.
1028 @item start_time, st
1029 Specify the start time of the fade effect. Default is 0.
1030 The value must be specified as a time duration; see
1031 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1032 for the accepted syntax.
1033 If set this option is used instead of @var{start_sample}.
1036 Specify the duration of the fade effect. See
1037 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1038 for the accepted syntax.
1039 At the end of the fade-in effect the output audio will have the same
1040 volume as the input audio, at the end of the fade-out transition
1041 the output audio will be silence.
1042 By default the duration is determined by @var{nb_samples}.
1043 If set this option is used instead of @var{nb_samples}.
1046 Set curve for fade transition.
1048 It accepts the following values:
1051 select triangular, linear slope (default)
1053 select quarter of sine wave
1055 select half of sine wave
1057 select exponential sine wave
1061 select inverted parabola
1075 select inverted quarter of sine wave
1077 select inverted half of sine wave
1079 select double-exponential seat
1081 select double-exponential sigmoid
1083 select logistic sigmoid
1085 select sine cardinal function
1087 select inverted sine cardinal function
1093 @subsection Commands
1095 This filter supports the all above options as @ref{commands}.
1097 @subsection Examples
1101 Fade in first 15 seconds of audio:
1103 afade=t=in:ss=0:d=15
1107 Fade out last 25 seconds of a 900 seconds audio:
1109 afade=t=out:st=875:d=25
1114 Denoise audio samples with FFT.
1116 A description of the accepted parameters follows.
1120 Set the noise reduction in dB, allowed range is 0.01 to 97.
1121 Default value is 12 dB.
1124 Set the noise floor in dB, allowed range is -80 to -20.
1125 Default value is -50 dB.
1130 It accepts the following values:
1139 Select shellac noise.
1142 Select custom noise, defined in @code{bn} option.
1144 Default value is white noise.
1148 Set custom band noise for every one of 15 bands.
1149 Bands are separated by ' ' or '|'.
1152 Set the residual floor in dB, allowed range is -80 to -20.
1153 Default value is -38 dB.
1156 Enable noise tracking. By default is disabled.
1157 With this enabled, noise floor is automatically adjusted.
1160 Enable residual tracking. By default is disabled.
1163 Set the output mode.
1165 It accepts the following values:
1168 Pass input unchanged.
1171 Pass noise filtered out.
1176 Default value is @var{o}.
1180 @subsection Commands
1182 This filter supports the following commands:
1184 @item sample_noise, sn
1185 Start or stop measuring noise profile.
1186 Syntax for the command is : "start" or "stop" string.
1187 After measuring noise profile is stopped it will be
1188 automatically applied in filtering.
1190 @item noise_reduction, nr
1191 Change noise reduction. Argument is single float number.
1192 Syntax for the command is : "@var{noise_reduction}"
1194 @item noise_floor, nf
1195 Change noise floor. Argument is single float number.
1196 Syntax for the command is : "@var{noise_floor}"
1198 @item output_mode, om
1199 Change output mode operation.
1200 Syntax for the command is : "i", "o" or "n" string.
1204 Apply arbitrary expressions to samples in frequency domain.
1208 Set frequency domain real expression for each separate channel separated
1209 by '|'. Default is "re".
1210 If the number of input channels is greater than the number of
1211 expressions, the last specified expression is used for the remaining
1215 Set frequency domain imaginary expression for each separate channel
1216 separated by '|'. Default is "im".
1218 Each expression in @var{real} and @var{imag} can contain the following
1219 constants and functions:
1226 current frequency bin number
1229 number of available bins
1232 channel number of the current expression
1241 current real part of frequency bin of current channel
1244 current imaginary part of frequency bin of current channel
1247 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1250 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1254 Set window size. Allowed range is from 16 to 131072.
1255 Default is @code{4096}
1258 Set window function. Default is @code{hann}.
1261 Set window overlap. If set to 1, the recommended overlap for selected
1262 window function will be picked. Default is @code{0.75}.
1265 @subsection Examples
1269 Leave almost only low frequencies in audio:
1271 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1275 Apply robotize effect:
1277 afftfilt="real='hypot(re,im)*sin(0)':imag='hypot(re,im)*cos(0)':win_size=512:overlap=0.75"
1281 Apply whisper effect:
1283 afftfilt="real='hypot(re,im)*cos((random(0)*2-1)*2*3.14)':imag='hypot(re,im)*sin((random(1)*2-1)*2*3.14)':win_size=128:overlap=0.8"
1290 Apply an arbitrary Finite Impulse Response filter.
1292 This filter is designed for applying long FIR filters,
1293 up to 60 seconds long.
1295 It can be used as component for digital crossover filters,
1296 room equalization, cross talk cancellation, wavefield synthesis,
1297 auralization, ambiophonics, ambisonics and spatialization.
1299 This filter uses the streams higher than first one as FIR coefficients.
1300 If the non-first stream holds a single channel, it will be used
1301 for all input channels in the first stream, otherwise
1302 the number of channels in the non-first stream must be same as
1303 the number of channels in the first stream.
1305 It accepts the following parameters:
1309 Set dry gain. This sets input gain.
1312 Set wet gain. This sets final output gain.
1315 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1318 Enable applying gain measured from power of IR.
1320 Set which approach to use for auto gain measurement.
1324 Do not apply any gain.
1327 select peak gain, very conservative approach. This is default value.
1330 select DC gain, limited application.
1333 select gain to noise approach, this is most popular one.
1337 Set gain to be applied to IR coefficients before filtering.
1338 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1341 Set format of IR stream. Can be @code{mono} or @code{input}.
1342 Default is @code{input}.
1345 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1346 Allowed range is 0.1 to 60 seconds.
1349 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1350 By default it is disabled.
1353 Set for which IR channel to display frequency response. By default is first channel
1354 displayed. This option is used only when @var{response} is enabled.
1357 Set video stream size. This option is used only when @var{response} is enabled.
1360 Set video stream frame rate. This option is used only when @var{response} is enabled.
1363 Set minimal partition size used for convolution. Default is @var{8192}.
1364 Allowed range is from @var{1} to @var{32768}.
1365 Lower values decreases latency at cost of higher CPU usage.
1368 Set maximal partition size used for convolution. Default is @var{8192}.
1369 Allowed range is from @var{8} to @var{32768}.
1370 Lower values may increase CPU usage.
1373 Set number of input impulse responses streams which will be switchable at runtime.
1374 Allowed range is from @var{1} to @var{32}. Default is @var{1}.
1377 Set IR stream which will be used for convolution, starting from @var{0}, should always be
1378 lower than supplied value by @code{nbirs} option. Default is @var{0}.
1379 This option can be changed at runtime via @ref{commands}.
1382 @subsection Examples
1386 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1388 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1395 Set output format constraints for the input audio. The framework will
1396 negotiate the most appropriate format to minimize conversions.
1398 It accepts the following parameters:
1401 @item sample_fmts, f
1402 A '|'-separated list of requested sample formats.
1404 @item sample_rates, r
1405 A '|'-separated list of requested sample rates.
1407 @item channel_layouts, cl
1408 A '|'-separated list of requested channel layouts.
1410 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1411 for the required syntax.
1414 If a parameter is omitted, all values are allowed.
1416 Force the output to either unsigned 8-bit or signed 16-bit stereo
1418 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1422 Apply frequency shift to input audio samples.
1424 The filter accepts the following options:
1428 Specify frequency shift. Allowed range is -INT_MAX to INT_MAX.
1429 Default value is 0.0.
1432 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
1433 Default value is 1.0.
1436 @subsection Commands
1438 This filter supports the all above options as @ref{commands}.
1442 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1443 processing reduces disturbing noise between useful signals.
1445 Gating is done by detecting the volume below a chosen level @var{threshold}
1446 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1447 floor is set via @var{range}. Because an exact manipulation of the signal
1448 would cause distortion of the waveform the reduction can be levelled over
1449 time. This is done by setting @var{attack} and @var{release}.
1451 @var{attack} determines how long the signal has to fall below the threshold
1452 before any reduction will occur and @var{release} sets the time the signal
1453 has to rise above the threshold to reduce the reduction again.
1454 Shorter signals than the chosen attack time will be left untouched.
1458 Set input level before filtering.
1459 Default is 1. Allowed range is from 0.015625 to 64.
1462 Set the mode of operation. Can be @code{upward} or @code{downward}.
1463 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1464 will be amplified, expanding dynamic range in upward direction.
1465 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1468 Set the level of gain reduction when the signal is below the threshold.
1469 Default is 0.06125. Allowed range is from 0 to 1.
1470 Setting this to 0 disables reduction and then filter behaves like expander.
1473 If a signal rises above this level the gain reduction is released.
1474 Default is 0.125. Allowed range is from 0 to 1.
1477 Set a ratio by which the signal is reduced.
1478 Default is 2. Allowed range is from 1 to 9000.
1481 Amount of milliseconds the signal has to rise above the threshold before gain
1483 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1486 Amount of milliseconds the signal has to fall below the threshold before the
1487 reduction is increased again. Default is 250 milliseconds.
1488 Allowed range is from 0.01 to 9000.
1491 Set amount of amplification of signal after processing.
1492 Default is 1. Allowed range is from 1 to 64.
1495 Curve the sharp knee around the threshold to enter gain reduction more softly.
1496 Default is 2.828427125. Allowed range is from 1 to 8.
1499 Choose if exact signal should be taken for detection or an RMS like one.
1500 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1503 Choose if the average level between all channels or the louder channel affects
1505 Default is @code{average}. Can be @code{average} or @code{maximum}.
1508 @subsection Commands
1510 This filter supports the all above options as @ref{commands}.
1514 Apply an arbitrary Infinite Impulse Response filter.
1516 It accepts the following parameters:
1520 Set B/numerator/zeros/reflection coefficients.
1523 Set A/denominator/poles/ladder coefficients.
1535 Set coefficients format.
1539 lattice-ladder function
1541 analog transfer function
1543 digital transfer function
1545 Z-plane zeros/poles, cartesian (default)
1547 Z-plane zeros/poles, polar radians
1549 Z-plane zeros/poles, polar degrees
1555 Set type of processing.
1567 Set filtering precision.
1571 double-precision floating-point (default)
1573 single-precision floating-point
1581 Normalize filter coefficients, by default is enabled.
1582 Enabling it will normalize magnitude response at DC to 0dB.
1585 How much to use filtered signal in output. Default is 1.
1586 Range is between 0 and 1.
1589 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1590 By default it is disabled.
1593 Set for which IR channel to display frequency response. By default is first channel
1594 displayed. This option is used only when @var{response} is enabled.
1597 Set video stream size. This option is used only when @var{response} is enabled.
1600 Coefficients in @code{tf} and @code{sf} format are separated by spaces and are in ascending
1603 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1604 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1607 Different coefficients and gains can be provided for every channel, in such case
1608 use '|' to separate coefficients or gains. Last provided coefficients will be
1609 used for all remaining channels.
1611 @subsection Examples
1615 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1617 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1621 Same as above but in @code{zp} format:
1623 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1627 Apply 3-rd order analog normalized Butterworth low-pass filter, using analog transfer function format:
1629 aiir=z=1.3057 0 0 0:p=1.3057 2.3892 2.1860 1:f=sf:r=d
1635 The limiter prevents an input signal from rising over a desired threshold.
1636 This limiter uses lookahead technology to prevent your signal from distorting.
1637 It means that there is a small delay after the signal is processed. Keep in mind
1638 that the delay it produces is the attack time you set.
1640 The filter accepts the following options:
1644 Set input gain. Default is 1.
1647 Set output gain. Default is 1.
1650 Don't let signals above this level pass the limiter. Default is 1.
1653 The limiter will reach its attenuation level in this amount of time in
1654 milliseconds. Default is 5 milliseconds.
1657 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1658 Default is 50 milliseconds.
1661 When gain reduction is always needed ASC takes care of releasing to an
1662 average reduction level rather than reaching a reduction of 0 in the release
1666 Select how much the release time is affected by ASC, 0 means nearly no changes
1667 in release time while 1 produces higher release times.
1670 Auto level output signal. Default is enabled.
1671 This normalizes audio back to 0dB if enabled.
1674 Depending on picked setting it is recommended to upsample input 2x or 4x times
1675 with @ref{aresample} before applying this filter.
1679 Apply a two-pole all-pass filter with central frequency (in Hz)
1680 @var{frequency}, and filter-width @var{width}.
1681 An all-pass filter changes the audio's frequency to phase relationship
1682 without changing its frequency to amplitude relationship.
1684 The filter accepts the following options:
1688 Set frequency in Hz.
1691 Set method to specify band-width of filter.
1706 Specify the band-width of a filter in width_type units.
1709 How much to use filtered signal in output. Default is 1.
1710 Range is between 0 and 1.
1713 Specify which channels to filter, by default all available are filtered.
1716 Normalize biquad coefficients, by default is disabled.
1717 Enabling it will normalize magnitude response at DC to 0dB.
1720 Set the filter order, can be 1 or 2. Default is 2.
1723 Set transform type of IIR filter.
1732 Set precison of filtering.
1735 Pick automatic sample format depending on surround filters.
1737 Always use signed 16-bit.
1739 Always use signed 32-bit.
1741 Always use float 32-bit.
1743 Always use float 64-bit.
1747 @subsection Commands
1749 This filter supports the following commands:
1752 Change allpass frequency.
1753 Syntax for the command is : "@var{frequency}"
1756 Change allpass width_type.
1757 Syntax for the command is : "@var{width_type}"
1760 Change allpass width.
1761 Syntax for the command is : "@var{width}"
1765 Syntax for the command is : "@var{mix}"
1772 The filter accepts the following options:
1776 Set the number of loops. Setting this value to -1 will result in infinite loops.
1780 Set maximal number of samples. Default is 0.
1783 Set first sample of loop. Default is 0.
1789 Merge two or more audio streams into a single multi-channel stream.
1791 The filter accepts the following options:
1796 Set the number of inputs. Default is 2.
1800 If the channel layouts of the inputs are disjoint, and therefore compatible,
1801 the channel layout of the output will be set accordingly and the channels
1802 will be reordered as necessary. If the channel layouts of the inputs are not
1803 disjoint, the output will have all the channels of the first input then all
1804 the channels of the second input, in that order, and the channel layout of
1805 the output will be the default value corresponding to the total number of
1808 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1809 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1810 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1811 first input, b1 is the first channel of the second input).
1813 On the other hand, if both input are in stereo, the output channels will be
1814 in the default order: a1, a2, b1, b2, and the channel layout will be
1815 arbitrarily set to 4.0, which may or may not be the expected value.
1817 All inputs must have the same sample rate, and format.
1819 If inputs do not have the same duration, the output will stop with the
1822 @subsection Examples
1826 Merge two mono files into a stereo stream:
1828 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1832 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1834 ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
1840 Mixes multiple audio inputs into a single output.
1842 Note that this filter only supports float samples (the @var{amerge}
1843 and @var{pan} audio filters support many formats). If the @var{amix}
1844 input has integer samples then @ref{aresample} will be automatically
1845 inserted to perform the conversion to float samples.
1849 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1851 will mix 3 input audio streams to a single output with the same duration as the
1852 first input and a dropout transition time of 3 seconds.
1854 It accepts the following parameters:
1858 The number of inputs. If unspecified, it defaults to 2.
1861 How to determine the end-of-stream.
1865 The duration of the longest input. (default)
1868 The duration of the shortest input.
1871 The duration of the first input.
1875 @item dropout_transition
1876 The transition time, in seconds, for volume renormalization when an input
1877 stream ends. The default value is 2 seconds.
1880 Specify weight of each input audio stream as sequence.
1881 Each weight is separated by space. By default all inputs have same weight.
1884 Do not scale inputs but instead do only summation of samples.
1885 Beware of heavy clipping if inputs are not normalized prior of filtering
1886 or output from @var{amix} normalized after filtering. By default is disabled.
1889 @subsection Commands
1891 This filter supports the following commands:
1895 Syntax is same as option with same name.
1900 Multiply first audio stream with second audio stream and store result
1901 in output audio stream. Multiplication is done by multiplying each
1902 sample from first stream with sample at same position from second stream.
1904 With this element-wise multiplication one can create amplitude fades and
1905 amplitude modulations.
1907 @section anequalizer
1909 High-order parametric multiband equalizer for each channel.
1911 It accepts the following parameters:
1915 This option string is in format:
1916 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1917 Each equalizer band is separated by '|'.
1921 Set channel number to which equalization will be applied.
1922 If input doesn't have that channel the entry is ignored.
1925 Set central frequency for band.
1926 If input doesn't have that frequency the entry is ignored.
1929 Set band width in Hertz.
1932 Set band gain in dB.
1935 Set filter type for band, optional, can be:
1939 Butterworth, this is default.
1950 With this option activated frequency response of anequalizer is displayed
1954 Set video stream size. Only useful if curves option is activated.
1957 Set max gain that will be displayed. Only useful if curves option is activated.
1958 Setting this to a reasonable value makes it possible to display gain which is derived from
1959 neighbour bands which are too close to each other and thus produce higher gain
1960 when both are activated.
1963 Set frequency scale used to draw frequency response in video output.
1964 Can be linear or logarithmic. Default is logarithmic.
1967 Set color for each channel curve which is going to be displayed in video stream.
1968 This is list of color names separated by space or by '|'.
1969 Unrecognised or missing colors will be replaced by white color.
1972 @subsection Examples
1976 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1977 for first 2 channels using Chebyshev type 1 filter:
1979 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1983 @subsection Commands
1985 This filter supports the following commands:
1988 Alter existing filter parameters.
1989 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1991 @var{fN} is existing filter number, starting from 0, if no such filter is available
1993 @var{freq} set new frequency parameter.
1994 @var{width} set new width parameter in Hertz.
1995 @var{gain} set new gain parameter in dB.
1997 Full filter invocation with asendcmd may look like this:
1998 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
2003 Reduce broadband noise in audio samples using Non-Local Means algorithm.
2005 Each sample is adjusted by looking for other samples with similar contexts. This
2006 context similarity is defined by comparing their surrounding patches of size
2007 @option{p}. Patches are searched in an area of @option{r} around the sample.
2009 The filter accepts the following options:
2013 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
2016 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
2017 Default value is 2 milliseconds.
2020 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
2021 Default value is 6 milliseconds.
2024 Set the output mode.
2026 It accepts the following values:
2029 Pass input unchanged.
2032 Pass noise filtered out.
2037 Default value is @var{o}.
2041 Set smooth factor. Default value is @var{11}. Allowed range is from @var{1} to @var{15}.
2044 @subsection Commands
2046 This filter supports the all above options as @ref{commands}.
2049 Apply Normalized Least-Mean-Squares algorithm to the first audio stream using the second audio stream.
2051 This adaptive filter is used to mimic a desired filter by finding the filter coefficients that
2052 relate to producing the least mean square of the error signal (difference between the desired,
2053 2nd input audio stream and the actual signal, the 1st input audio stream).
2055 A description of the accepted options follows.
2068 Set the filter leakage.
2071 It accepts the following values:
2080 Pass filtered samples.
2083 Pass difference between desired and filtered samples.
2085 Default value is @var{o}.
2089 @subsection Examples
2093 One of many usages of this filter is noise reduction, input audio is filtered
2094 with same samples that are delayed by fixed amount, one such example for stereo audio is:
2096 asplit[a][b],[a]adelay=32S|32S[a],[b][a]anlms=order=128:leakage=0.0005:mu=.5:out_mode=o
2100 @subsection Commands
2102 This filter supports the same commands as options, excluding option @code{order}.
2106 Pass the audio source unchanged to the output.
2110 Pad the end of an audio stream with silence.
2112 This can be used together with @command{ffmpeg} @option{-shortest} to
2113 extend audio streams to the same length as the video stream.
2115 A description of the accepted options follows.
2119 Set silence packet size. Default value is 4096.
2122 Set the number of samples of silence to add to the end. After the
2123 value is reached, the stream is terminated. This option is mutually
2124 exclusive with @option{whole_len}.
2127 Set the minimum total number of samples in the output audio stream. If
2128 the value is longer than the input audio length, silence is added to
2129 the end, until the value is reached. This option is mutually exclusive
2130 with @option{pad_len}.
2133 Specify the duration of samples of silence to add. See
2134 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2135 for the accepted syntax. Used only if set to non-zero value.
2138 Specify the minimum total duration in the output audio stream. See
2139 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2140 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
2141 the input audio length, silence is added to the end, until the value is reached.
2142 This option is mutually exclusive with @option{pad_dur}
2145 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
2146 nor @option{whole_dur} option is set, the filter will add silence to the end of
2147 the input stream indefinitely.
2149 @subsection Examples
2153 Add 1024 samples of silence to the end of the input:
2159 Make sure the audio output will contain at least 10000 samples, pad
2160 the input with silence if required:
2162 apad=whole_len=10000
2166 Use @command{ffmpeg} to pad the audio input with silence, so that the
2167 video stream will always result the shortest and will be converted
2168 until the end in the output file when using the @option{shortest}
2171 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
2176 Add a phasing effect to the input audio.
2178 A phaser filter creates series of peaks and troughs in the frequency spectrum.
2179 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
2181 A description of the accepted parameters follows.
2185 Set input gain. Default is 0.4.
2188 Set output gain. Default is 0.74
2191 Set delay in milliseconds. Default is 3.0.
2194 Set decay. Default is 0.4.
2197 Set modulation speed in Hz. Default is 0.5.
2200 Set modulation type. Default is triangular.
2202 It accepts the following values:
2209 @section aphaseshift
2210 Apply phase shift to input audio samples.
2212 The filter accepts the following options:
2216 Specify phase shift. Allowed range is from -1.0 to 1.0.
2217 Default value is 0.0.
2220 Set output gain applied to final output. Allowed range is from 0.0 to 1.0.
2221 Default value is 1.0.
2224 @subsection Commands
2226 This filter supports the all above options as @ref{commands}.
2230 Audio pulsator is something between an autopanner and a tremolo.
2231 But it can produce funny stereo effects as well. Pulsator changes the volume
2232 of the left and right channel based on a LFO (low frequency oscillator) with
2233 different waveforms and shifted phases.
2234 This filter have the ability to define an offset between left and right
2235 channel. An offset of 0 means that both LFO shapes match each other.
2236 The left and right channel are altered equally - a conventional tremolo.
2237 An offset of 50% means that the shape of the right channel is exactly shifted
2238 in phase (or moved backwards about half of the frequency) - pulsator acts as
2239 an autopanner. At 1 both curves match again. Every setting in between moves the
2240 phase shift gapless between all stages and produces some "bypassing" sounds with
2241 sine and triangle waveforms. The more you set the offset near 1 (starting from
2242 the 0.5) the faster the signal passes from the left to the right speaker.
2244 The filter accepts the following options:
2248 Set input gain. By default it is 1. Range is [0.015625 - 64].
2251 Set output gain. By default it is 1. Range is [0.015625 - 64].
2254 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
2255 sawup or sawdown. Default is sine.
2258 Set modulation. Define how much of original signal is affected by the LFO.
2261 Set left channel offset. Default is 0. Allowed range is [0 - 1].
2264 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
2267 Set pulse width. Default is 1. Allowed range is [0 - 2].
2270 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
2273 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
2277 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
2281 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
2282 if timing is set to hz.
2288 Resample the input audio to the specified parameters, using the
2289 libswresample library. If none are specified then the filter will
2290 automatically convert between its input and output.
2292 This filter is also able to stretch/squeeze the audio data to make it match
2293 the timestamps or to inject silence / cut out audio to make it match the
2294 timestamps, do a combination of both or do neither.
2296 The filter accepts the syntax
2297 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
2298 expresses a sample rate and @var{resampler_options} is a list of
2299 @var{key}=@var{value} pairs, separated by ":". See the
2300 @ref{Resampler Options,,"Resampler Options" section in the
2301 ffmpeg-resampler(1) manual,ffmpeg-resampler}
2302 for the complete list of supported options.
2304 @subsection Examples
2308 Resample the input audio to 44100Hz:
2314 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
2315 samples per second compensation:
2317 aresample=async=1000
2323 Reverse an audio clip.
2325 Warning: This filter requires memory to buffer the entire clip, so trimming
2328 @subsection Examples
2332 Take the first 5 seconds of a clip, and reverse it.
2334 atrim=end=5,areverse
2340 Reduce noise from speech using Recurrent Neural Networks.
2342 This filter accepts the following options:
2346 Set train model file to load. This option is always required.
2349 Set how much to mix filtered samples into final output.
2350 Allowed range is from -1 to 1. Default value is 1.
2351 Negative values are special, they set how much to keep filtered noise
2352 in the final filter output. Set this option to -1 to hear actual
2353 noise removed from input signal.
2356 @subsection Commands
2358 This filter supports the all above options as @ref{commands}.
2360 @section asetnsamples
2362 Set the number of samples per each output audio frame.
2364 The last output packet may contain a different number of samples, as
2365 the filter will flush all the remaining samples when the input audio
2368 The filter accepts the following options:
2372 @item nb_out_samples, n
2373 Set the number of frames per each output audio frame. The number is
2374 intended as the number of samples @emph{per each channel}.
2375 Default value is 1024.
2378 If set to 1, the filter will pad the last audio frame with zeroes, so
2379 that the last frame will contain the same number of samples as the
2380 previous ones. Default value is 1.
2383 For example, to set the number of per-frame samples to 1234 and
2384 disable padding for the last frame, use:
2386 asetnsamples=n=1234:p=0
2391 Set the sample rate without altering the PCM data.
2392 This will result in a change of speed and pitch.
2394 The filter accepts the following options:
2397 @item sample_rate, r
2398 Set the output sample rate. Default is 44100 Hz.
2403 Show a line containing various information for each input audio frame.
2404 The input audio is not modified.
2406 The shown line contains a sequence of key/value pairs of the form
2407 @var{key}:@var{value}.
2409 The following values are shown in the output:
2413 The (sequential) number of the input frame, starting from 0.
2416 The presentation timestamp of the input frame, in time base units; the time base
2417 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2420 The presentation timestamp of the input frame in seconds.
2423 position of the frame in the input stream, -1 if this information in
2424 unavailable and/or meaningless (for example in case of synthetic audio)
2433 The sample rate for the audio frame.
2436 The number of samples (per channel) in the frame.
2439 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2440 audio, the data is treated as if all the planes were concatenated.
2442 @item plane_checksums
2443 A list of Adler-32 checksums for each data plane.
2447 Apply audio soft clipping.
2449 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2450 along a smooth curve, rather than the abrupt shape of hard-clipping.
2452 This filter accepts the following options:
2456 Set type of soft-clipping.
2458 It accepts the following values:
2472 Set threshold from where to start clipping. Default value is 0dB or 1.
2475 Set gain applied to output. Default value is 0dB or 1.
2478 Set additional parameter which controls sigmoid function.
2481 Set oversampling factor.
2484 @subsection Commands
2486 This filter supports the all above options as @ref{commands}.
2489 Automatic Speech Recognition
2491 This filter uses PocketSphinx for speech recognition. To enable
2492 compilation of this filter, you need to configure FFmpeg with
2493 @code{--enable-pocketsphinx}.
2495 It accepts the following options:
2499 Set sampling rate of input audio. Defaults is @code{16000}.
2500 This need to match speech models, otherwise one will get poor results.
2503 Set dictionary containing acoustic model files.
2506 Set pronunciation dictionary.
2509 Set language model file.
2512 Set language model set.
2515 Set which language model to use.
2518 Set output for log messages.
2521 The filter exports recognized speech as the frame metadata @code{lavfi.asr.text}.
2526 Display time domain statistical information about the audio channels.
2527 Statistics are calculated and displayed for each audio channel and,
2528 where applicable, an overall figure is also given.
2530 It accepts the following option:
2533 Short window length in seconds, used for peak and trough RMS measurement.
2534 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2538 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2539 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2542 Available keys for each channel are:
2588 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2589 this @code{lavfi.astats.Overall.Peak_count}.
2591 For description what each key means read below.
2594 Set number of frame after which stats are going to be recalculated.
2595 Default is disabled.
2597 @item measure_perchannel
2598 Select the entries which need to be measured per channel. The metadata keys can
2599 be used as flags, default is @option{all} which measures everything.
2600 @option{none} disables all per channel measurement.
2602 @item measure_overall
2603 Select the entries which need to be measured overall. The metadata keys can
2604 be used as flags, default is @option{all} which measures everything.
2605 @option{none} disables all overall measurement.
2609 A description of each shown parameter follows:
2613 Mean amplitude displacement from zero.
2616 Minimal sample level.
2619 Maximal sample level.
2621 @item Min difference
2622 Minimal difference between two consecutive samples.
2624 @item Max difference
2625 Maximal difference between two consecutive samples.
2627 @item Mean difference
2628 Mean difference between two consecutive samples.
2629 The average of each difference between two consecutive samples.
2631 @item RMS difference
2632 Root Mean Square difference between two consecutive samples.
2636 Standard peak and RMS level measured in dBFS.
2640 Peak and trough values for RMS level measured over a short window.
2643 Standard ratio of peak to RMS level (note: not in dB).
2646 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2647 (i.e. either @var{Min level} or @var{Max level}).
2650 Number of occasions (not the number of samples) that the signal attained either
2651 @var{Min level} or @var{Max level}.
2653 @item Noise floor dB
2654 Minimum local peak measured in dBFS over a short window.
2656 @item Noise floor count
2657 Number of occasions (not the number of samples) that the signal attained
2661 Overall bit depth of audio. Number of bits used for each sample.
2664 Measured dynamic range of audio in dB.
2666 @item Zero crossings
2667 Number of points where the waveform crosses the zero level axis.
2669 @item Zero crossings rate
2670 Rate of Zero crossings and number of audio samples.
2674 Boost subwoofer frequencies.
2676 The filter accepts the following options:
2680 Set dry gain, how much of original signal is kept. Allowed range is from 0 to 1.
2681 Default value is 0.7.
2684 Set wet gain, how much of filtered signal is kept. Allowed range is from 0 to 1.
2685 Default value is 0.7.
2688 Set delay line decay gain value. Allowed range is from 0 to 1.
2689 Default value is 0.7.
2692 Set delay line feedback gain value. Allowed range is from 0 to 1.
2693 Default value is 0.9.
2696 Set cutoff frequency in Hertz. Allowed range is 50 to 900.
2697 Default value is 100.
2700 Set slope amount for cutoff frequency. Allowed range is 0.0001 to 1.
2701 Default value is 0.5.
2704 Set delay. Allowed range is from 1 to 100.
2705 Default value is 20.
2708 @subsection Commands
2710 This filter supports the all above options as @ref{commands}.
2713 Cut subwoofer frequencies.
2715 This filter allows to set custom, steeper
2716 roll off than highpass filter, and thus is able to more attenuate
2717 frequency content in stop-band.
2719 The filter accepts the following options:
2723 Set cutoff frequency in Hertz. Allowed range is 2 to 200.
2724 Default value is 20.
2727 Set filter order. Available values are from 3 to 20.
2728 Default value is 10.
2731 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2734 @subsection Commands
2736 This filter supports the all above options as @ref{commands}.
2739 Cut super frequencies.
2741 The filter accepts the following options:
2745 Set cutoff frequency in Hertz. Allowed range is 20000 to 192000.
2746 Default value is 20000.
2749 Set filter order. Available values are from 3 to 20.
2750 Default value is 10.
2753 Set input gain level. Allowed range is from 0 to 1. Default value is 1.
2756 @subsection Commands
2758 This filter supports the all above options as @ref{commands}.
2761 Apply high order Butterworth band-pass filter.
2763 The filter accepts the following options:
2767 Set center frequency in Hertz. Allowed range is 2 to 999999.
2768 Default value is 1000.
2771 Set filter order. Available values are from 4 to 20.
2775 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2778 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2781 @subsection Commands
2783 This filter supports the all above options as @ref{commands}.
2786 Apply high order Butterworth band-stop filter.
2788 The filter accepts the following options:
2792 Set center frequency in Hertz. Allowed range is 2 to 999999.
2793 Default value is 1000.
2796 Set filter order. Available values are from 4 to 20.
2800 Set Q-factor. Allowed range is from 0.01 to 100. Default value is 1.
2803 Set input gain level. Allowed range is from 0 to 2. Default value is 1.
2806 @subsection Commands
2808 This filter supports the all above options as @ref{commands}.
2814 The filter accepts exactly one parameter, the audio tempo. If not
2815 specified then the filter will assume nominal 1.0 tempo. Tempo must
2816 be in the [0.5, 100.0] range.
2818 Note that tempo greater than 2 will skip some samples rather than
2819 blend them in. If for any reason this is a concern it is always
2820 possible to daisy-chain several instances of atempo to achieve the
2821 desired product tempo.
2823 @subsection Examples
2827 Slow down audio to 80% tempo:
2833 To speed up audio to 300% tempo:
2839 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2841 atempo=sqrt(3),atempo=sqrt(3)
2845 @subsection Commands
2847 This filter supports the following commands:
2850 Change filter tempo scale factor.
2851 Syntax for the command is : "@var{tempo}"
2856 Trim the input so that the output contains one continuous subpart of the input.
2858 It accepts the following parameters:
2861 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2862 sample with the timestamp @var{start} will be the first sample in the output.
2865 Specify time of the first audio sample that will be dropped, i.e. the
2866 audio sample immediately preceding the one with the timestamp @var{end} will be
2867 the last sample in the output.
2870 Same as @var{start}, except this option sets the start timestamp in samples
2874 Same as @var{end}, except this option sets the end timestamp in samples instead
2878 The maximum duration of the output in seconds.
2881 The number of the first sample that should be output.
2884 The number of the first sample that should be dropped.
2887 @option{start}, @option{end}, and @option{duration} are expressed as time
2888 duration specifications; see
2889 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2891 Note that the first two sets of the start/end options and the @option{duration}
2892 option look at the frame timestamp, while the _sample options simply count the
2893 samples that pass through the filter. So start/end_pts and start/end_sample will
2894 give different results when the timestamps are wrong, inexact or do not start at
2895 zero. Also note that this filter does not modify the timestamps. If you wish
2896 to have the output timestamps start at zero, insert the asetpts filter after the
2899 If multiple start or end options are set, this filter tries to be greedy and
2900 keep all samples that match at least one of the specified constraints. To keep
2901 only the part that matches all the constraints at once, chain multiple atrim
2904 The defaults are such that all the input is kept. So it is possible to set e.g.
2905 just the end values to keep everything before the specified time.
2910 Drop everything except the second minute of input:
2912 ffmpeg -i INPUT -af atrim=60:120
2916 Keep only the first 1000 samples:
2918 ffmpeg -i INPUT -af atrim=end_sample=1000
2923 @section axcorrelate
2924 Calculate normalized cross-correlation between two input audio streams.
2926 Resulted samples are always between -1 and 1 inclusive.
2927 If result is 1 it means two input samples are highly correlated in that selected segment.
2928 Result 0 means they are not correlated at all.
2929 If result is -1 it means two input samples are out of phase, which means they cancel each
2932 The filter accepts the following options:
2936 Set size of segment over which cross-correlation is calculated.
2937 Default is 256. Allowed range is from 2 to 131072.
2940 Set algorithm for cross-correlation. Can be @code{slow} or @code{fast}.
2941 Default is @code{slow}. Fast algorithm assumes mean values over any given segment
2942 are always zero and thus need much less calculations to make.
2943 This is generally not true, but is valid for typical audio streams.
2946 @subsection Examples
2950 Calculate correlation between channels in stereo audio stream:
2952 ffmpeg -i stereo.wav -af channelsplit,axcorrelate=size=1024:algo=fast correlation.wav
2958 Apply a two-pole Butterworth band-pass filter with central
2959 frequency @var{frequency}, and (3dB-point) band-width width.
2960 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2961 instead of the default: constant 0dB peak gain.
2962 The filter roll off at 6dB per octave (20dB per decade).
2964 The filter accepts the following options:
2968 Set the filter's central frequency. Default is @code{3000}.
2971 Constant skirt gain if set to 1. Defaults to 0.
2974 Set method to specify band-width of filter.
2989 Specify the band-width of a filter in width_type units.
2992 How much to use filtered signal in output. Default is 1.
2993 Range is between 0 and 1.
2996 Specify which channels to filter, by default all available are filtered.
2999 Normalize biquad coefficients, by default is disabled.
3000 Enabling it will normalize magnitude response at DC to 0dB.
3003 Set transform type of IIR filter.
3012 Set precison of filtering.
3015 Pick automatic sample format depending on surround filters.
3017 Always use signed 16-bit.
3019 Always use signed 32-bit.
3021 Always use float 32-bit.
3023 Always use float 64-bit.
3027 @subsection Commands
3029 This filter supports the following commands:
3032 Change bandpass frequency.
3033 Syntax for the command is : "@var{frequency}"
3036 Change bandpass width_type.
3037 Syntax for the command is : "@var{width_type}"
3040 Change bandpass width.
3041 Syntax for the command is : "@var{width}"
3044 Change bandpass mix.
3045 Syntax for the command is : "@var{mix}"
3050 Apply a two-pole Butterworth band-reject filter with central
3051 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
3052 The filter roll off at 6dB per octave (20dB per decade).
3054 The filter accepts the following options:
3058 Set the filter's central frequency. Default is @code{3000}.
3061 Set method to specify band-width of filter.
3076 Specify the band-width of a filter in width_type units.
3079 How much to use filtered signal in output. Default is 1.
3080 Range is between 0 and 1.
3083 Specify which channels to filter, by default all available are filtered.
3086 Normalize biquad coefficients, by default is disabled.
3087 Enabling it will normalize magnitude response at DC to 0dB.
3090 Set transform type of IIR filter.
3099 Set precison of filtering.
3102 Pick automatic sample format depending on surround filters.
3104 Always use signed 16-bit.
3106 Always use signed 32-bit.
3108 Always use float 32-bit.
3110 Always use float 64-bit.
3114 @subsection Commands
3116 This filter supports the following commands:
3119 Change bandreject frequency.
3120 Syntax for the command is : "@var{frequency}"
3123 Change bandreject width_type.
3124 Syntax for the command is : "@var{width_type}"
3127 Change bandreject width.
3128 Syntax for the command is : "@var{width}"
3131 Change bandreject mix.
3132 Syntax for the command is : "@var{mix}"
3135 @section bass, lowshelf
3137 Boost or cut the bass (lower) frequencies of the audio using a two-pole
3138 shelving filter with a response similar to that of a standard
3139 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3141 The filter accepts the following options:
3145 Give the gain at 0 Hz. Its useful range is about -20
3146 (for a large cut) to +20 (for a large boost).
3147 Beware of clipping when using a positive gain.
3150 Set the filter's central frequency and so can be used
3151 to extend or reduce the frequency range to be boosted or cut.
3152 The default value is @code{100} Hz.
3155 Set method to specify band-width of filter.
3170 Determine how steep is the filter's shelf transition.
3173 Set number of poles. Default is 2.
3176 How much to use filtered signal in output. Default is 1.
3177 Range is between 0 and 1.
3180 Specify which channels to filter, by default all available are filtered.
3183 Normalize biquad coefficients, by default is disabled.
3184 Enabling it will normalize magnitude response at DC to 0dB.
3187 Set transform type of IIR filter.
3196 Set precison of filtering.
3199 Pick automatic sample format depending on surround filters.
3201 Always use signed 16-bit.
3203 Always use signed 32-bit.
3205 Always use float 32-bit.
3207 Always use float 64-bit.
3211 @subsection Commands
3213 This filter supports the following commands:
3216 Change bass frequency.
3217 Syntax for the command is : "@var{frequency}"
3220 Change bass width_type.
3221 Syntax for the command is : "@var{width_type}"
3225 Syntax for the command is : "@var{width}"
3229 Syntax for the command is : "@var{gain}"
3233 Syntax for the command is : "@var{mix}"
3238 Apply a biquad IIR filter with the given coefficients.
3239 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
3240 are the numerator and denominator coefficients respectively.
3241 and @var{channels}, @var{c} specify which channels to filter, by default all
3242 available are filtered.
3244 @subsection Commands
3246 This filter supports the following commands:
3254 Change biquad parameter.
3255 Syntax for the command is : "@var{value}"
3258 How much to use filtered signal in output. Default is 1.
3259 Range is between 0 and 1.
3262 Specify which channels to filter, by default all available are filtered.
3265 Normalize biquad coefficients, by default is disabled.
3266 Enabling it will normalize magnitude response at DC to 0dB.
3269 Set transform type of IIR filter.
3278 Set precison of filtering.
3281 Pick automatic sample format depending on surround filters.
3283 Always use signed 16-bit.
3285 Always use signed 32-bit.
3287 Always use float 32-bit.
3289 Always use float 64-bit.
3294 Bauer stereo to binaural transformation, which improves headphone listening of
3295 stereo audio records.
3297 To enable compilation of this filter you need to configure FFmpeg with
3298 @code{--enable-libbs2b}.
3300 It accepts the following parameters:
3304 Pre-defined crossfeed level.
3308 Default level (fcut=700, feed=50).
3311 Chu Moy circuit (fcut=700, feed=60).
3314 Jan Meier circuit (fcut=650, feed=95).
3319 Cut frequency (in Hz).
3328 Remap input channels to new locations.
3330 It accepts the following parameters:
3333 Map channels from input to output. The argument is a '|'-separated list of
3334 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
3335 @var{in_channel} form. @var{in_channel} can be either the name of the input
3336 channel (e.g. FL for front left) or its index in the input channel layout.
3337 @var{out_channel} is the name of the output channel or its index in the output
3338 channel layout. If @var{out_channel} is not given then it is implicitly an
3339 index, starting with zero and increasing by one for each mapping.
3341 @item channel_layout
3342 The channel layout of the output stream.
3345 If no mapping is present, the filter will implicitly map input channels to
3346 output channels, preserving indices.
3348 @subsection Examples
3352 For example, assuming a 5.1+downmix input MOV file,
3354 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
3356 will create an output WAV file tagged as stereo from the downmix channels of
3360 To fix a 5.1 WAV improperly encoded in AAC's native channel order
3362 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
3366 @section channelsplit
3368 Split each channel from an input audio stream into a separate output stream.
3370 It accepts the following parameters:
3372 @item channel_layout
3373 The channel layout of the input stream. The default is "stereo".
3375 A channel layout describing the channels to be extracted as separate output streams
3376 or "all" to extract each input channel as a separate stream. The default is "all".
3378 Choosing channels not present in channel layout in the input will result in an error.
3381 @subsection Examples
3385 For example, assuming a stereo input MP3 file,
3387 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
3389 will create an output Matroska file with two audio streams, one containing only
3390 the left channel and the other the right channel.
3393 Split a 5.1 WAV file into per-channel files:
3395 ffmpeg -i in.wav -filter_complex
3396 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
3397 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
3398 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
3403 Extract only LFE from a 5.1 WAV file:
3405 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
3406 -map '[LFE]' lfe.wav
3411 Add a chorus effect to the audio.
3413 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
3415 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
3416 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
3417 The modulation depth defines the range the modulated delay is played before or after
3418 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
3419 sound tuned around the original one, like in a chorus where some vocals are slightly
3422 It accepts the following parameters:
3425 Set input gain. Default is 0.4.
3428 Set output gain. Default is 0.4.
3431 Set delays. A typical delay is around 40ms to 60ms.
3443 @subsection Examples
3449 chorus=0.7:0.9:55:0.4:0.25:2
3455 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
3459 Fuller sounding chorus with three delays:
3461 chorus=0.5:0.9:50|60|40:0.4|0.32|0.3:0.25|0.4|0.3:2|2.3|1.3
3466 Compress or expand the audio's dynamic range.
3468 It accepts the following parameters:
3474 A list of times in seconds for each channel over which the instantaneous level
3475 of the input signal is averaged to determine its volume. @var{attacks} refers to
3476 increase of volume and @var{decays} refers to decrease of volume. For most
3477 situations, the attack time (response to the audio getting louder) should be
3478 shorter than the decay time, because the human ear is more sensitive to sudden
3479 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
3480 a typical value for decay is 0.8 seconds.
3481 If specified number of attacks & decays is lower than number of channels, the last
3482 set attack/decay will be used for all remaining channels.
3485 A list of points for the transfer function, specified in dB relative to the
3486 maximum possible signal amplitude. Each key points list must be defined using
3487 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
3488 @code{x0/y0 x1/y1 x2/y2 ....}
3490 The input values must be in strictly increasing order but the transfer function
3491 does not have to be monotonically rising. The point @code{0/0} is assumed but
3492 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
3493 function are @code{-70/-70|-60/-20|1/0}.
3496 Set the curve radius in dB for all joints. It defaults to 0.01.
3499 Set the additional gain in dB to be applied at all points on the transfer
3500 function. This allows for easy adjustment of the overall gain.
3504 Set an initial volume, in dB, to be assumed for each channel when filtering
3505 starts. This permits the user to supply a nominal level initially, so that, for
3506 example, a very large gain is not applied to initial signal levels before the
3507 companding has begun to operate. A typical value for audio which is initially
3508 quiet is -90 dB. It defaults to 0.
3511 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
3512 delayed before being fed to the volume adjuster. Specifying a delay
3513 approximately equal to the attack/decay times allows the filter to effectively
3514 operate in predictive rather than reactive mode. It defaults to 0.
3518 @subsection Examples
3522 Make music with both quiet and loud passages suitable for listening to in a
3525 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
3528 Another example for audio with whisper and explosion parts:
3530 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
3534 A noise gate for when the noise is at a lower level than the signal:
3536 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
3540 Here is another noise gate, this time for when the noise is at a higher level
3541 than the signal (making it, in some ways, similar to squelch):
3543 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
3547 2:1 compression starting at -6dB:
3549 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
3553 2:1 compression starting at -9dB:
3555 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
3559 2:1 compression starting at -12dB:
3561 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
3565 2:1 compression starting at -18dB:
3567 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
3571 3:1 compression starting at -15dB:
3573 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
3579 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
3585 compand=attacks=0:points=-80/-169|-54/-80|-49.5/-64.6|-41.1/-41.1|-25.8/-15|-10.8/-4.5|0/0|20/8.3
3589 Hard limiter at -6dB:
3591 compand=attacks=0:points=-80/-80|-6/-6|20/-6
3595 Hard limiter at -12dB:
3597 compand=attacks=0:points=-80/-80|-12/-12|20/-12
3601 Hard noise gate at -35 dB:
3603 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
3609 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
3613 @section compensationdelay
3615 Compensation Delay Line is a metric based delay to compensate differing
3616 positions of microphones or speakers.
3618 For example, you have recorded guitar with two microphones placed in
3619 different locations. Because the front of sound wave has fixed speed in
3620 normal conditions, the phasing of microphones can vary and depends on
3621 their location and interposition. The best sound mix can be achieved when
3622 these microphones are in phase (synchronized). Note that a distance of
3623 ~30 cm between microphones makes one microphone capture the signal in
3624 antiphase to the other microphone. That makes the final mix sound moody.
3625 This filter helps to solve phasing problems by adding different delays
3626 to each microphone track and make them synchronized.
3628 The best result can be reached when you take one track as base and
3629 synchronize other tracks one by one with it.
3630 Remember that synchronization/delay tolerance depends on sample rate, too.
3631 Higher sample rates will give more tolerance.
3633 The filter accepts the following parameters:
3637 Set millimeters distance. This is compensation distance for fine tuning.
3641 Set cm distance. This is compensation distance for tightening distance setup.
3645 Set meters distance. This is compensation distance for hard distance setup.
3649 Set dry amount. Amount of unprocessed (dry) signal.
3653 Set wet amount. Amount of processed (wet) signal.
3657 Set temperature in degrees Celsius. This is the temperature of the environment.
3662 Apply headphone crossfeed filter.
3664 Crossfeed is the process of blending the left and right channels of stereo
3666 It is mainly used to reduce extreme stereo separation of low frequencies.
3668 The intent is to produce more speaker like sound to the listener.
3670 The filter accepts the following options:
3674 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
3675 This sets gain of low shelf filter for side part of stereo image.
3676 Default is -6dB. Max allowed is -30db when strength is set to 1.
3679 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
3680 This sets cut off frequency of low shelf filter. Default is cut off near
3681 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
3684 Set curve slope of low shelf filter. Default is 0.5.
3685 Allowed range is from 0.01 to 1.
3688 Set input gain. Default is 0.9.
3691 Set output gain. Default is 1.
3694 @subsection Commands
3696 This filter supports the all above options as @ref{commands}.
3698 @section crystalizer
3699 Simple algorithm for audio noise sharpening.
3701 This filter linearly increases differences betweeen each audio sample.
3703 The filter accepts the following options:
3707 Sets the intensity of effect (default: 2.0). Must be in range between -10.0 to 0
3708 (unchanged sound) to 10.0 (maximum effect).
3709 To inverse filtering use negative value.
3712 Enable clipping. By default is enabled.
3715 @subsection Commands
3717 This filter supports the all above options as @ref{commands}.
3720 Apply a DC shift to the audio.
3722 This can be useful to remove a DC offset (caused perhaps by a hardware problem
3723 in the recording chain) from the audio. The effect of a DC offset is reduced
3724 headroom and hence volume. The @ref{astats} filter can be used to determine if
3725 a signal has a DC offset.
3729 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
3733 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
3734 used to prevent clipping.
3739 Apply de-essing to the audio samples.
3743 Set intensity for triggering de-essing. Allowed range is from 0 to 1.
3747 Set amount of ducking on treble part of sound. Allowed range is from 0 to 1.
3751 How much of original frequency content to keep when de-essing. Allowed range is from 0 to 1.
3755 Set the output mode.
3757 It accepts the following values:
3760 Pass input unchanged.
3763 Pass ess filtered out.
3768 Default value is @var{o}.
3774 Measure audio dynamic range.
3776 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3777 is found in transition material. And anything less that 8 have very poor dynamics
3778 and is very compressed.
3780 The filter accepts the following options:
3784 Set window length in seconds used to split audio into segments of equal length.
3785 Default is 3 seconds.
3789 Dynamic Audio Normalizer.
3791 This filter applies a certain amount of gain to the input audio in order
3792 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3793 contrast to more "simple" normalization algorithms, the Dynamic Audio
3794 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3795 This allows for applying extra gain to the "quiet" sections of the audio
3796 while avoiding distortions or clipping the "loud" sections. In other words:
3797 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3798 sections, in the sense that the volume of each section is brought to the
3799 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3800 this goal *without* applying "dynamic range compressing". It will retain 100%
3801 of the dynamic range *within* each section of the audio file.
3805 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3806 Default is 500 milliseconds.
3807 The Dynamic Audio Normalizer processes the input audio in small chunks,
3808 referred to as frames. This is required, because a peak magnitude has no
3809 meaning for just a single sample value. Instead, we need to determine the
3810 peak magnitude for a contiguous sequence of sample values. While a "standard"
3811 normalizer would simply use the peak magnitude of the complete file, the
3812 Dynamic Audio Normalizer determines the peak magnitude individually for each
3813 frame. The length of a frame is specified in milliseconds. By default, the
3814 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3815 been found to give good results with most files.
3816 Note that the exact frame length, in number of samples, will be determined
3817 automatically, based on the sampling rate of the individual input audio file.
3820 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3821 number. Default is 31.
3822 Probably the most important parameter of the Dynamic Audio Normalizer is the
3823 @code{window size} of the Gaussian smoothing filter. The filter's window size
3824 is specified in frames, centered around the current frame. For the sake of
3825 simplicity, this must be an odd number. Consequently, the default value of 31
3826 takes into account the current frame, as well as the 15 preceding frames and
3827 the 15 subsequent frames. Using a larger window results in a stronger
3828 smoothing effect and thus in less gain variation, i.e. slower gain
3829 adaptation. Conversely, using a smaller window results in a weaker smoothing
3830 effect and thus in more gain variation, i.e. faster gain adaptation.
3831 In other words, the more you increase this value, the more the Dynamic Audio
3832 Normalizer will behave like a "traditional" normalization filter. On the
3833 contrary, the more you decrease this value, the more the Dynamic Audio
3834 Normalizer will behave like a dynamic range compressor.
3837 Set the target peak value. This specifies the highest permissible magnitude
3838 level for the normalized audio input. This filter will try to approach the
3839 target peak magnitude as closely as possible, but at the same time it also
3840 makes sure that the normalized signal will never exceed the peak magnitude.
3841 A frame's maximum local gain factor is imposed directly by the target peak
3842 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3843 It is not recommended to go above this value.
3846 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3847 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3848 factor for each input frame, i.e. the maximum gain factor that does not
3849 result in clipping or distortion. The maximum gain factor is determined by
3850 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3851 additionally bounds the frame's maximum gain factor by a predetermined
3852 (global) maximum gain factor. This is done in order to avoid excessive gain
3853 factors in "silent" or almost silent frames. By default, the maximum gain
3854 factor is 10.0, For most inputs the default value should be sufficient and
3855 it usually is not recommended to increase this value. Though, for input
3856 with an extremely low overall volume level, it may be necessary to allow even
3857 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3858 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3859 Instead, a "sigmoid" threshold function will be applied. This way, the
3860 gain factors will smoothly approach the threshold value, but never exceed that
3864 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3865 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3866 This means that the maximum local gain factor for each frame is defined
3867 (only) by the frame's highest magnitude sample. This way, the samples can
3868 be amplified as much as possible without exceeding the maximum signal
3869 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3870 Normalizer can also take into account the frame's root mean square,
3871 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3872 determine the power of a time-varying signal. It is therefore considered
3873 that the RMS is a better approximation of the "perceived loudness" than
3874 just looking at the signal's peak magnitude. Consequently, by adjusting all
3875 frames to a constant RMS value, a uniform "perceived loudness" can be
3876 established. If a target RMS value has been specified, a frame's local gain
3877 factor is defined as the factor that would result in exactly that RMS value.
3878 Note, however, that the maximum local gain factor is still restricted by the
3879 frame's highest magnitude sample, in order to prevent clipping.
3882 Enable channels coupling. By default is enabled.
3883 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3884 amount. This means the same gain factor will be applied to all channels, i.e.
3885 the maximum possible gain factor is determined by the "loudest" channel.
3886 However, in some recordings, it may happen that the volume of the different
3887 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3888 In this case, this option can be used to disable the channel coupling. This way,
3889 the gain factor will be determined independently for each channel, depending
3890 only on the individual channel's highest magnitude sample. This allows for
3891 harmonizing the volume of the different channels.
3894 Enable DC bias correction. By default is disabled.
3895 An audio signal (in the time domain) is a sequence of sample values.
3896 In the Dynamic Audio Normalizer these sample values are represented in the
3897 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3898 audio signal, or "waveform", should be centered around the zero point.
3899 That means if we calculate the mean value of all samples in a file, or in a
3900 single frame, then the result should be 0.0 or at least very close to that
3901 value. If, however, there is a significant deviation of the mean value from
3902 0.0, in either positive or negative direction, this is referred to as a
3903 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3904 Audio Normalizer provides optional DC bias correction.
3905 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3906 the mean value, or "DC correction" offset, of each input frame and subtract
3907 that value from all of the frame's sample values which ensures those samples
3908 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3909 boundaries, the DC correction offset values will be interpolated smoothly
3910 between neighbouring frames.
3912 @item altboundary, b
3913 Enable alternative boundary mode. By default is disabled.
3914 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3915 around each frame. This includes the preceding frames as well as the
3916 subsequent frames. However, for the "boundary" frames, located at the very
3917 beginning and at the very end of the audio file, not all neighbouring
3918 frames are available. In particular, for the first few frames in the audio
3919 file, the preceding frames are not known. And, similarly, for the last few
3920 frames in the audio file, the subsequent frames are not known. Thus, the
3921 question arises which gain factors should be assumed for the missing frames
3922 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3923 to deal with this situation. The default boundary mode assumes a gain factor
3924 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3925 "fade out" at the beginning and at the end of the input, respectively.
3928 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3929 By default, the Dynamic Audio Normalizer does not apply "traditional"
3930 compression. This means that signal peaks will not be pruned and thus the
3931 full dynamic range will be retained within each local neighbourhood. However,
3932 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3933 normalization algorithm with a more "traditional" compression.
3934 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3935 (thresholding) function. If (and only if) the compression feature is enabled,
3936 all input frames will be processed by a soft knee thresholding function prior
3937 to the actual normalization process. Put simply, the thresholding function is
3938 going to prune all samples whose magnitude exceeds a certain threshold value.
3939 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3940 value. Instead, the threshold value will be adjusted for each individual
3942 In general, smaller parameters result in stronger compression, and vice versa.
3943 Values below 3.0 are not recommended, because audible distortion may appear.
3946 Set the target threshold value. This specifies the lowest permissible
3947 magnitude level for the audio input which will be normalized.
3948 If input frame volume is above this value frame will be normalized.
3949 Otherwise frame may not be normalized at all. The default value is set
3950 to 0, which means all input frames will be normalized.
3951 This option is mostly useful if digital noise is not wanted to be amplified.
3954 @subsection Commands
3956 This filter supports the all above options as @ref{commands}.
3960 Make audio easier to listen to on headphones.
3962 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3963 so that when listened to on headphones the stereo image is moved from
3964 inside your head (standard for headphones) to outside and in front of
3965 the listener (standard for speakers).
3971 Apply a two-pole peaking equalisation (EQ) filter. With this
3972 filter, the signal-level at and around a selected frequency can
3973 be increased or decreased, whilst (unlike bandpass and bandreject
3974 filters) that at all other frequencies is unchanged.
3976 In order to produce complex equalisation curves, this filter can
3977 be given several times, each with a different central frequency.
3979 The filter accepts the following options:
3983 Set the filter's central frequency in Hz.
3986 Set method to specify band-width of filter.
4001 Specify the band-width of a filter in width_type units.
4004 Set the required gain or attenuation in dB.
4005 Beware of clipping when using a positive gain.
4008 How much to use filtered signal in output. Default is 1.
4009 Range is between 0 and 1.
4012 Specify which channels to filter, by default all available are filtered.
4015 Normalize biquad coefficients, by default is disabled.
4016 Enabling it will normalize magnitude response at DC to 0dB.
4019 Set transform type of IIR filter.
4028 Set precison of filtering.
4031 Pick automatic sample format depending on surround filters.
4033 Always use signed 16-bit.
4035 Always use signed 32-bit.
4037 Always use float 32-bit.
4039 Always use float 64-bit.
4043 @subsection Examples
4046 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
4048 equalizer=f=1000:t=h:width=200:g=-10
4052 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
4054 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
4058 @subsection Commands
4060 This filter supports the following commands:
4063 Change equalizer frequency.
4064 Syntax for the command is : "@var{frequency}"
4067 Change equalizer width_type.
4068 Syntax for the command is : "@var{width_type}"
4071 Change equalizer width.
4072 Syntax for the command is : "@var{width}"
4075 Change equalizer gain.
4076 Syntax for the command is : "@var{gain}"
4079 Change equalizer mix.
4080 Syntax for the command is : "@var{mix}"
4083 @section extrastereo
4085 Linearly increases the difference between left and right channels which
4086 adds some sort of "live" effect to playback.
4088 The filter accepts the following options:
4092 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
4093 (average of both channels), with 1.0 sound will be unchanged, with
4094 -1.0 left and right channels will be swapped.
4097 Enable clipping. By default is enabled.
4100 @subsection Commands
4102 This filter supports the all above options as @ref{commands}.
4104 @section firequalizer
4105 Apply FIR Equalization using arbitrary frequency response.
4107 The filter accepts the following option:
4111 Set gain curve equation (in dB). The expression can contain variables:
4114 the evaluated frequency
4118 channel number, set to 0 when multichannels evaluation is disabled
4120 channel id, see libavutil/channel_layout.h, set to the first channel id when
4121 multichannels evaluation is disabled
4125 channel_layout, see libavutil/channel_layout.h
4130 @item gain_interpolate(f)
4131 interpolate gain on frequency f based on gain_entry
4132 @item cubic_interpolate(f)
4133 same as gain_interpolate, but smoother
4135 This option is also available as command. Default is @code{gain_interpolate(f)}.
4138 Set gain entry for gain_interpolate function. The expression can
4142 store gain entry at frequency f with value g
4144 This option is also available as command.
4147 Set filter delay in seconds. Higher value means more accurate.
4148 Default is @code{0.01}.
4151 Set filter accuracy in Hz. Lower value means more accurate.
4152 Default is @code{5}.
4155 Set window function. Acceptable values are:
4158 rectangular window, useful when gain curve is already smooth
4160 hann window (default)
4166 3-terms continuous 1st derivative nuttall window
4168 minimum 3-terms discontinuous nuttall window
4170 4-terms continuous 1st derivative nuttall window
4172 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
4174 blackman-harris window
4180 If enabled, use fixed number of audio samples. This improves speed when
4181 filtering with large delay. Default is disabled.
4184 Enable multichannels evaluation on gain. Default is disabled.
4187 Enable zero phase mode by subtracting timestamp to compensate delay.
4188 Default is disabled.
4191 Set scale used by gain. Acceptable values are:
4194 linear frequency, linear gain
4196 linear frequency, logarithmic (in dB) gain (default)
4198 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
4200 logarithmic frequency, logarithmic gain
4204 Set file for dumping, suitable for gnuplot.
4207 Set scale for dumpfile. Acceptable values are same with scale option.
4211 Enable 2-channel convolution using complex FFT. This improves speed significantly.
4212 Default is disabled.
4215 Enable minimum phase impulse response. Default is disabled.
4218 @subsection Examples
4223 firequalizer=gain='if(lt(f,1000), 0, -INF)'
4226 lowpass at 1000 Hz with gain_entry:
4228 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
4231 custom equalization:
4233 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
4236 higher delay with zero phase to compensate delay:
4238 firequalizer=delay=0.1:fixed=on:zero_phase=on
4241 lowpass on left channel, highpass on right channel:
4243 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
4244 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
4249 Apply a flanging effect to the audio.
4251 The filter accepts the following options:
4255 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
4258 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
4261 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
4265 Set percentage of delayed signal mixed with original. Range from 0 to 100.
4266 Default value is 71.
4269 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
4272 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
4273 Default value is @var{sinusoidal}.
4276 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
4277 Default value is 25.
4280 Set delay-line interpolation, @var{linear} or @var{quadratic}.
4281 Default is @var{linear}.
4285 Apply Haas effect to audio.
4287 Note that this makes most sense to apply on mono signals.
4288 With this filter applied to mono signals it give some directionality and
4289 stretches its stereo image.
4291 The filter accepts the following options:
4295 Set input level. By default is @var{1}, or 0dB
4298 Set output level. By default is @var{1}, or 0dB.
4301 Set gain applied to side part of signal. By default is @var{1}.
4304 Set kind of middle source. Can be one of the following:
4314 Pick middle part signal of stereo image.
4317 Pick side part signal of stereo image.
4321 Change middle phase. By default is disabled.
4324 Set left channel delay. By default is @var{2.05} milliseconds.
4327 Set left channel balance. By default is @var{-1}.
4330 Set left channel gain. By default is @var{1}.
4333 Change left phase. By default is disabled.
4336 Set right channel delay. By defaults is @var{2.12} milliseconds.
4339 Set right channel balance. By default is @var{1}.
4342 Set right channel gain. By default is @var{1}.
4345 Change right phase. By default is enabled.
4350 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
4351 embedded HDCD codes is expanded into a 20-bit PCM stream.
4353 The filter supports the Peak Extend and Low-level Gain Adjustment features
4354 of HDCD, and detects the Transient Filter flag.
4357 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
4360 When using the filter with wav, note the default encoding for wav is 16-bit,
4361 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
4362 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
4364 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
4365 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
4368 The filter accepts the following options:
4371 @item disable_autoconvert
4372 Disable any automatic format conversion or resampling in the filter graph.
4374 @item process_stereo
4375 Process the stereo channels together. If target_gain does not match between
4376 channels, consider it invalid and use the last valid target_gain.
4379 Set the code detect timer period in ms.
4382 Always extend peaks above -3dBFS even if PE isn't signaled.
4385 Replace audio with a solid tone and adjust the amplitude to signal some
4386 specific aspect of the decoding process. The output file can be loaded in
4387 an audio editor alongside the original to aid analysis.
4389 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
4396 Gain adjustment level at each sample
4398 Samples where peak extend occurs
4400 Samples where the code detect timer is active
4402 Samples where the target gain does not match between channels
4408 Apply head-related transfer functions (HRTFs) to create virtual
4409 loudspeakers around the user for binaural listening via headphones.
4410 The HRIRs are provided via additional streams, for each channel
4411 one stereo input stream is needed.
4413 The filter accepts the following options:
4417 Set mapping of input streams for convolution.
4418 The argument is a '|'-separated list of channel names in order as they
4419 are given as additional stream inputs for filter.
4420 This also specify number of input streams. Number of input streams
4421 must be not less than number of channels in first stream plus one.
4424 Set gain applied to audio. Value is in dB. Default is 0.
4427 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4428 processing audio in time domain which is slow.
4429 @var{freq} is processing audio in frequency domain which is fast.
4430 Default is @var{freq}.
4433 Set custom gain for LFE channels. Value is in dB. Default is 0.
4436 Set size of frame in number of samples which will be processed at once.
4437 Default value is @var{1024}. Allowed range is from 1024 to 96000.
4440 Set format of hrir stream.
4441 Default value is @var{stereo}. Alternative value is @var{multich}.
4442 If value is set to @var{stereo}, number of additional streams should
4443 be greater or equal to number of input channels in first input stream.
4444 Also each additional stream should have stereo number of channels.
4445 If value is set to @var{multich}, number of additional streams should
4446 be exactly one. Also number of input channels of additional stream
4447 should be equal or greater than twice number of channels of first input
4451 @subsection Examples
4455 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4456 each amovie filter use stereo file with IR coefficients as input.
4457 The files give coefficients for each position of virtual loudspeaker:
4460 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
4465 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
4466 but now in @var{multich} @var{hrir} format.
4468 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
4475 Apply a high-pass filter with 3dB point frequency.
4476 The filter can be either single-pole, or double-pole (the default).
4477 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4479 The filter accepts the following options:
4483 Set frequency in Hz. Default is 3000.
4486 Set number of poles. Default is 2.
4489 Set method to specify band-width of filter.
4504 Specify the band-width of a filter in width_type units.
4505 Applies only to double-pole filter.
4506 The default is 0.707q and gives a Butterworth response.
4509 How much to use filtered signal in output. Default is 1.
4510 Range is between 0 and 1.
4513 Specify which channels to filter, by default all available are filtered.
4516 Normalize biquad coefficients, by default is disabled.
4517 Enabling it will normalize magnitude response at DC to 0dB.
4520 Set transform type of IIR filter.
4529 Set precison of filtering.
4532 Pick automatic sample format depending on surround filters.
4534 Always use signed 16-bit.
4536 Always use signed 32-bit.
4538 Always use float 32-bit.
4540 Always use float 64-bit.
4544 @subsection Commands
4546 This filter supports the following commands:
4549 Change highpass frequency.
4550 Syntax for the command is : "@var{frequency}"
4553 Change highpass width_type.
4554 Syntax for the command is : "@var{width_type}"
4557 Change highpass width.
4558 Syntax for the command is : "@var{width}"
4561 Change highpass mix.
4562 Syntax for the command is : "@var{mix}"
4567 Join multiple input streams into one multi-channel stream.
4569 It accepts the following parameters:
4573 The number of input streams. It defaults to 2.
4575 @item channel_layout
4576 The desired output channel layout. It defaults to stereo.
4579 Map channels from inputs to output. The argument is a '|'-separated list of
4580 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
4581 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
4582 can be either the name of the input channel (e.g. FL for front left) or its
4583 index in the specified input stream. @var{out_channel} is the name of the output
4587 The filter will attempt to guess the mappings when they are not specified
4588 explicitly. It does so by first trying to find an unused matching input channel
4589 and if that fails it picks the first unused input channel.
4591 Join 3 inputs (with properly set channel layouts):
4593 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
4596 Build a 5.1 output from 6 single-channel streams:
4598 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
4599 'join=inputs=6:channel_layout=5.1:map=0.0-FL|1.0-FR|2.0-FC|3.0-SL|4.0-SR|5.0-LFE'
4605 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
4607 To enable compilation of this filter you need to configure FFmpeg with
4608 @code{--enable-ladspa}.
4612 Specifies the name of LADSPA plugin library to load. If the environment
4613 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
4614 each one of the directories specified by the colon separated list in
4615 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
4616 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
4617 @file{/usr/lib/ladspa/}.
4620 Specifies the plugin within the library. Some libraries contain only
4621 one plugin, but others contain many of them. If this is not set filter
4622 will list all available plugins within the specified library.
4625 Set the '|' separated list of controls which are zero or more floating point
4626 values that determine the behavior of the loaded plugin (for example delay,
4628 Controls need to be defined using the following syntax:
4629 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
4630 @var{valuei} is the value set on the @var{i}-th control.
4631 Alternatively they can be also defined using the following syntax:
4632 @var{value0}|@var{value1}|@var{value2}|..., where
4633 @var{valuei} is the value set on the @var{i}-th control.
4634 If @option{controls} is set to @code{help}, all available controls and
4635 their valid ranges are printed.
4637 @item sample_rate, s
4638 Specify the sample rate, default to 44100. Only used if plugin have
4642 Set the number of samples per channel per each output frame, default
4643 is 1024. Only used if plugin have zero inputs.
4646 Set the minimum duration of the sourced audio. See
4647 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4648 for the accepted syntax.
4649 Note that the resulting duration may be greater than the specified duration,
4650 as the generated audio is always cut at the end of a complete frame.
4651 If not specified, or the expressed duration is negative, the audio is
4652 supposed to be generated forever.
4653 Only used if plugin have zero inputs.
4656 Enable latency compensation, by default is disabled.
4657 Only used if plugin have inputs.
4660 @subsection Examples
4664 List all available plugins within amp (LADSPA example plugin) library:
4670 List all available controls and their valid ranges for @code{vcf_notch}
4671 plugin from @code{VCF} library:
4673 ladspa=f=vcf:p=vcf_notch:c=help
4677 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
4680 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
4684 Add reverberation to the audio using TAP-plugins
4685 (Tom's Audio Processing plugins):
4687 ladspa=file=tap_reverb:tap_reverb
4691 Generate white noise, with 0.2 amplitude:
4693 ladspa=file=cmt:noise_source_white:c=c0=.2
4697 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
4698 @code{C* Audio Plugin Suite} (CAPS) library:
4700 ladspa=file=caps:Click:c=c1=20'
4704 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
4706 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
4710 Increase volume by 20dB using fast lookahead limiter from Steve Harris
4711 @code{SWH Plugins} collection:
4713 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
4717 Attenuate low frequencies using Multiband EQ from Steve Harris
4718 @code{SWH Plugins} collection:
4720 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
4724 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
4727 ladspa=caps:Narrower
4731 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
4733 ladspa=caps:White:.2
4737 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
4739 ladspa=caps:Fractal:c=c1=1
4743 Dynamic volume normalization using @code{VLevel} plugin:
4745 ladspa=vlevel-ladspa:vlevel_mono
4749 @subsection Commands
4751 This filter supports the following commands:
4754 Modify the @var{N}-th control value.
4756 If the specified value is not valid, it is ignored and prior one is kept.
4761 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
4762 Support for both single pass (livestreams, files) and double pass (files) modes.
4763 This algorithm can target IL, LRA, and maximum true peak. In dynamic mode, to accurately
4764 detect true peaks, the audio stream will be upsampled to 192 kHz.
4765 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
4767 The filter accepts the following options:
4771 Set integrated loudness target.
4772 Range is -70.0 - -5.0. Default value is -24.0.
4775 Set loudness range target.
4776 Range is 1.0 - 20.0. Default value is 7.0.
4779 Set maximum true peak.
4780 Range is -9.0 - +0.0. Default value is -2.0.
4782 @item measured_I, measured_i
4783 Measured IL of input file.
4784 Range is -99.0 - +0.0.
4786 @item measured_LRA, measured_lra
4787 Measured LRA of input file.
4788 Range is 0.0 - 99.0.
4790 @item measured_TP, measured_tp
4791 Measured true peak of input file.
4792 Range is -99.0 - +99.0.
4794 @item measured_thresh
4795 Measured threshold of input file.
4796 Range is -99.0 - +0.0.
4799 Set offset gain. Gain is applied before the true-peak limiter.
4800 Range is -99.0 - +99.0. Default is +0.0.
4803 Normalize by linearly scaling the source audio.
4804 @code{measured_I}, @code{measured_LRA}, @code{measured_TP},
4805 and @code{measured_thresh} must all be specified. Target LRA shouldn't
4806 be lower than source LRA and the change in integrated loudness shouldn't
4807 result in a true peak which exceeds the target TP. If any of these
4808 conditions aren't met, normalization mode will revert to @var{dynamic}.
4809 Options are @code{true} or @code{false}. Default is @code{true}.
4812 Treat mono input files as "dual-mono". If a mono file is intended for playback
4813 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
4814 If set to @code{true}, this option will compensate for this effect.
4815 Multi-channel input files are not affected by this option.
4816 Options are true or false. Default is false.
4819 Set print format for stats. Options are summary, json, or none.
4820 Default value is none.
4825 Apply a low-pass filter with 3dB point frequency.
4826 The filter can be either single-pole or double-pole (the default).
4827 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4829 The filter accepts the following options:
4833 Set frequency in Hz. Default is 500.
4836 Set number of poles. Default is 2.
4839 Set method to specify band-width of filter.
4854 Specify the band-width of a filter in width_type units.
4855 Applies only to double-pole filter.
4856 The default is 0.707q and gives a Butterworth response.
4859 How much to use filtered signal in output. Default is 1.
4860 Range is between 0 and 1.
4863 Specify which channels to filter, by default all available are filtered.
4866 Normalize biquad coefficients, by default is disabled.
4867 Enabling it will normalize magnitude response at DC to 0dB.
4870 Set transform type of IIR filter.
4879 Set precison of filtering.
4882 Pick automatic sample format depending on surround filters.
4884 Always use signed 16-bit.
4886 Always use signed 32-bit.
4888 Always use float 32-bit.
4890 Always use float 64-bit.
4894 @subsection Examples
4897 Lowpass only LFE channel, it LFE is not present it does nothing:
4903 @subsection Commands
4905 This filter supports the following commands:
4908 Change lowpass frequency.
4909 Syntax for the command is : "@var{frequency}"
4912 Change lowpass width_type.
4913 Syntax for the command is : "@var{width_type}"
4916 Change lowpass width.
4917 Syntax for the command is : "@var{width}"
4921 Syntax for the command is : "@var{mix}"
4926 Load a LV2 (LADSPA Version 2) plugin.
4928 To enable compilation of this filter you need to configure FFmpeg with
4929 @code{--enable-lv2}.
4933 Specifies the plugin URI. You may need to escape ':'.
4936 Set the '|' separated list of controls which are zero or more floating point
4937 values that determine the behavior of the loaded plugin (for example delay,
4939 If @option{controls} is set to @code{help}, all available controls and
4940 their valid ranges are printed.
4942 @item sample_rate, s
4943 Specify the sample rate, default to 44100. Only used if plugin have
4947 Set the number of samples per channel per each output frame, default
4948 is 1024. Only used if plugin have zero inputs.
4951 Set the minimum duration of the sourced audio. See
4952 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4953 for the accepted syntax.
4954 Note that the resulting duration may be greater than the specified duration,
4955 as the generated audio is always cut at the end of a complete frame.
4956 If not specified, or the expressed duration is negative, the audio is
4957 supposed to be generated forever.
4958 Only used if plugin have zero inputs.
4961 @subsection Examples
4965 Apply bass enhancer plugin from Calf:
4967 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4971 Apply vinyl plugin from Calf:
4973 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4977 Apply bit crusher plugin from ArtyFX:
4979 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4984 Multiband Compress or expand the audio's dynamic range.
4986 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4987 This is akin to the crossover of a loudspeaker, and results in flat frequency
4988 response when absent compander action.
4990 It accepts the following parameters:
4994 This option syntax is:
4995 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4996 For explanation of each item refer to compand filter documentation.
5002 Mix channels with specific gain levels. The filter accepts the output
5003 channel layout followed by a set of channels definitions.
5005 This filter is also designed to efficiently remap the channels of an audio
5008 The filter accepts parameters of the form:
5009 "@var{l}|@var{outdef}|@var{outdef}|..."
5013 output channel layout or number of channels
5016 output channel specification, of the form:
5017 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
5020 output channel to define, either a channel name (FL, FR, etc.) or a channel
5021 number (c0, c1, etc.)
5024 multiplicative coefficient for the channel, 1 leaving the volume unchanged
5027 input channel to use, see out_name for details; it is not possible to mix
5028 named and numbered input channels
5031 If the `=' in a channel specification is replaced by `<', then the gains for
5032 that specification will be renormalized so that the total is 1, thus
5033 avoiding clipping noise.
5035 @subsection Mixing examples
5037 For example, if you want to down-mix from stereo to mono, but with a bigger
5038 factor for the left channel:
5040 pan=1c|c0=0.9*c0+0.1*c1
5043 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
5044 7-channels surround:
5046 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
5049 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
5050 that should be preferred (see "-ac" option) unless you have very specific
5053 @subsection Remapping examples
5055 The channel remapping will be effective if, and only if:
5058 @item gain coefficients are zeroes or ones,
5059 @item only one input per channel output,
5062 If all these conditions are satisfied, the filter will notify the user ("Pure
5063 channel mapping detected"), and use an optimized and lossless method to do the
5066 For example, if you have a 5.1 source and want a stereo audio stream by
5067 dropping the extra channels:
5069 pan="stereo| c0=FL | c1=FR"
5072 Given the same source, you can also switch front left and front right channels
5073 and keep the input channel layout:
5075 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
5078 If the input is a stereo audio stream, you can mute the front left channel (and
5079 still keep the stereo channel layout) with:
5084 Still with a stereo audio stream input, you can copy the right channel in both
5085 front left and right:
5087 pan="stereo| c0=FR | c1=FR"
5092 ReplayGain scanner filter. This filter takes an audio stream as an input and
5093 outputs it unchanged.
5094 At end of filtering it displays @code{track_gain} and @code{track_peak}.
5098 Convert the audio sample format, sample rate and channel layout. It is
5099 not meant to be used directly.
5102 Apply time-stretching and pitch-shifting with librubberband.
5104 To enable compilation of this filter, you need to configure FFmpeg with
5105 @code{--enable-librubberband}.
5107 The filter accepts the following options:
5111 Set tempo scale factor.
5114 Set pitch scale factor.
5117 Set transients detector.
5118 Possible values are:
5127 Possible values are:
5136 Possible values are:
5143 Set processing window size.
5144 Possible values are:
5153 Possible values are:
5160 Enable formant preservation when shift pitching.
5161 Possible values are:
5169 Possible values are:
5178 Possible values are:
5185 @subsection Commands
5187 This filter supports the following commands:
5190 Change filter tempo scale factor.
5191 Syntax for the command is : "@var{tempo}"
5194 Change filter pitch scale factor.
5195 Syntax for the command is : "@var{pitch}"
5198 @section sidechaincompress
5200 This filter acts like normal compressor but has the ability to compress
5201 detected signal using second input signal.
5202 It needs two input streams and returns one output stream.
5203 First input stream will be processed depending on second stream signal.
5204 The filtered signal then can be filtered with other filters in later stages of
5205 processing. See @ref{pan} and @ref{amerge} filter.
5207 The filter accepts the following options:
5211 Set input gain. Default is 1. Range is between 0.015625 and 64.
5214 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
5215 Default is @code{downward}.
5218 If a signal of second stream raises above this level it will affect the gain
5219 reduction of first stream.
5220 By default is 0.125. Range is between 0.00097563 and 1.
5223 Set a ratio about which the signal is reduced. 1:2 means that if the level
5224 raised 4dB above the threshold, it will be only 2dB above after the reduction.
5225 Default is 2. Range is between 1 and 20.
5228 Amount of milliseconds the signal has to rise above the threshold before gain
5229 reduction starts. Default is 20. Range is between 0.01 and 2000.
5232 Amount of milliseconds the signal has to fall below the threshold before
5233 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
5236 Set the amount by how much signal will be amplified after processing.
5237 Default is 1. Range is from 1 to 64.
5240 Curve the sharp knee around the threshold to enter gain reduction more softly.
5241 Default is 2.82843. Range is between 1 and 8.
5244 Choose if the @code{average} level between all channels of side-chain stream
5245 or the louder(@code{maximum}) channel of side-chain stream affects the
5246 reduction. Default is @code{average}.
5249 Should the exact signal be taken in case of @code{peak} or an RMS one in case
5250 of @code{rms}. Default is @code{rms} which is mainly smoother.
5253 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
5256 How much to use compressed signal in output. Default is 1.
5257 Range is between 0 and 1.
5260 @subsection Commands
5262 This filter supports the all above options as @ref{commands}.
5264 @subsection Examples
5268 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
5269 depending on the signal of 2nd input and later compressed signal to be
5270 merged with 2nd input:
5272 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
5276 @section sidechaingate
5278 A sidechain gate acts like a normal (wideband) gate but has the ability to
5279 filter the detected signal before sending it to the gain reduction stage.
5280 Normally a gate uses the full range signal to detect a level above the
5282 For example: If you cut all lower frequencies from your sidechain signal
5283 the gate will decrease the volume of your track only if not enough highs
5284 appear. With this technique you are able to reduce the resonation of a
5285 natural drum or remove "rumbling" of muted strokes from a heavily distorted
5287 It needs two input streams and returns one output stream.
5288 First input stream will be processed depending on second stream signal.
5290 The filter accepts the following options:
5294 Set input level before filtering.
5295 Default is 1. Allowed range is from 0.015625 to 64.
5298 Set the mode of operation. Can be @code{upward} or @code{downward}.
5299 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
5300 will be amplified, expanding dynamic range in upward direction.
5301 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
5304 Set the level of gain reduction when the signal is below the threshold.
5305 Default is 0.06125. Allowed range is from 0 to 1.
5306 Setting this to 0 disables reduction and then filter behaves like expander.
5309 If a signal rises above this level the gain reduction is released.
5310 Default is 0.125. Allowed range is from 0 to 1.
5313 Set a ratio about which the signal is reduced.
5314 Default is 2. Allowed range is from 1 to 9000.
5317 Amount of milliseconds the signal has to rise above the threshold before gain
5319 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
5322 Amount of milliseconds the signal has to fall below the threshold before the
5323 reduction is increased again. Default is 250 milliseconds.
5324 Allowed range is from 0.01 to 9000.
5327 Set amount of amplification of signal after processing.
5328 Default is 1. Allowed range is from 1 to 64.
5331 Curve the sharp knee around the threshold to enter gain reduction more softly.
5332 Default is 2.828427125. Allowed range is from 1 to 8.
5335 Choose if exact signal should be taken for detection or an RMS like one.
5336 Default is rms. Can be peak or rms.
5339 Choose if the average level between all channels or the louder channel affects
5341 Default is average. Can be average or maximum.
5344 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
5347 @subsection Commands
5349 This filter supports the all above options as @ref{commands}.
5351 @section silencedetect
5353 Detect silence in an audio stream.
5355 This filter logs a message when it detects that the input audio volume is less
5356 or equal to a noise tolerance value for a duration greater or equal to the
5357 minimum detected noise duration.
5359 The printed times and duration are expressed in seconds. The
5360 @code{lavfi.silence_start} or @code{lavfi.silence_start.X} metadata key
5361 is set on the first frame whose timestamp equals or exceeds the detection
5362 duration and it contains the timestamp of the first frame of the silence.
5364 The @code{lavfi.silence_duration} or @code{lavfi.silence_duration.X}
5365 and @code{lavfi.silence_end} or @code{lavfi.silence_end.X} metadata
5366 keys are set on the first frame after the silence. If @option{mono} is
5367 enabled, and each channel is evaluated separately, the @code{.X}
5368 suffixed keys are used, and @code{X} corresponds to the channel number.
5370 The filter accepts the following options:
5374 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
5375 specified value) or amplitude ratio. Default is -60dB, or 0.001.
5378 Set silence duration until notification (default is 2 seconds). See
5379 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5380 for the accepted syntax.
5383 Process each channel separately, instead of combined. By default is disabled.
5386 @subsection Examples
5390 Detect 5 seconds of silence with -50dB noise tolerance:
5392 silencedetect=n=-50dB:d=5
5396 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
5397 tolerance in @file{silence.mp3}:
5399 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
5403 @section silenceremove
5405 Remove silence from the beginning, middle or end of the audio.
5407 The filter accepts the following options:
5411 This value is used to indicate if audio should be trimmed at beginning of
5412 the audio. A value of zero indicates no silence should be trimmed from the
5413 beginning. When specifying a non-zero value, it trims audio up until it
5414 finds non-silence. Normally, when trimming silence from beginning of audio
5415 the @var{start_periods} will be @code{1} but it can be increased to higher
5416 values to trim all audio up to specific count of non-silence periods.
5417 Default value is @code{0}.
5419 @item start_duration
5420 Specify the amount of time that non-silence must be detected before it stops
5421 trimming audio. By increasing the duration, bursts of noises can be treated
5422 as silence and trimmed off. Default value is @code{0}.
5424 @item start_threshold
5425 This indicates what sample value should be treated as silence. For digital
5426 audio, a value of @code{0} may be fine but for audio recorded from analog,
5427 you may wish to increase the value to account for background noise.
5428 Can be specified in dB (in case "dB" is appended to the specified value)
5429 or amplitude ratio. Default value is @code{0}.
5432 Specify max duration of silence at beginning that will be kept after
5433 trimming. Default is 0, which is equal to trimming all samples detected
5437 Specify mode of detection of silence end in start of multi-channel audio.
5438 Can be @var{any} or @var{all}. Default is @var{any}.
5439 With @var{any}, any sample that is detected as non-silence will cause
5440 stopped trimming of silence.
5441 With @var{all}, only if all channels are detected as non-silence will cause
5442 stopped trimming of silence.
5445 Set the count for trimming silence from the end of audio.
5446 To remove silence from the middle of a file, specify a @var{stop_periods}
5447 that is negative. This value is then treated as a positive value and is
5448 used to indicate the effect should restart processing as specified by
5449 @var{start_periods}, making it suitable for removing periods of silence
5450 in the middle of the audio.
5451 Default value is @code{0}.
5454 Specify a duration of silence that must exist before audio is not copied any
5455 more. By specifying a higher duration, silence that is wanted can be left in
5457 Default value is @code{0}.
5459 @item stop_threshold
5460 This is the same as @option{start_threshold} but for trimming silence from
5462 Can be specified in dB (in case "dB" is appended to the specified value)
5463 or amplitude ratio. Default value is @code{0}.
5466 Specify max duration of silence at end that will be kept after
5467 trimming. Default is 0, which is equal to trimming all samples detected
5471 Specify mode of detection of silence start in end of multi-channel audio.
5472 Can be @var{any} or @var{all}. Default is @var{any}.
5473 With @var{any}, any sample that is detected as non-silence will cause
5474 stopped trimming of silence.
5475 With @var{all}, only if all channels are detected as non-silence will cause
5476 stopped trimming of silence.
5479 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
5480 and works better with digital silence which is exactly 0.
5481 Default value is @code{rms}.
5484 Set duration in number of seconds used to calculate size of window in number
5485 of samples for detecting silence.
5486 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
5489 @subsection Examples
5493 The following example shows how this filter can be used to start a recording
5494 that does not contain the delay at the start which usually occurs between
5495 pressing the record button and the start of the performance:
5497 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
5501 Trim all silence encountered from beginning to end where there is more than 1
5502 second of silence in audio:
5504 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
5508 Trim all digital silence samples, using peak detection, from beginning to end
5509 where there is more than 0 samples of digital silence in audio and digital
5510 silence is detected in all channels at same positions in stream:
5512 silenceremove=window=0:detection=peak:stop_mode=all:start_mode=all:stop_periods=-1:stop_threshold=0
5518 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
5519 loudspeakers around the user for binaural listening via headphones (audio
5520 formats up to 9 channels supported).
5521 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
5522 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
5523 Austrian Academy of Sciences.
5525 To enable compilation of this filter you need to configure FFmpeg with
5526 @code{--enable-libmysofa}.
5528 The filter accepts the following options:
5532 Set the SOFA file used for rendering.
5535 Set gain applied to audio. Value is in dB. Default is 0.
5538 Set rotation of virtual loudspeakers in deg. Default is 0.
5541 Set elevation of virtual speakers in deg. Default is 0.
5544 Set distance in meters between loudspeakers and the listener with near-field
5545 HRTFs. Default is 1.
5548 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
5549 processing audio in time domain which is slow.
5550 @var{freq} is processing audio in frequency domain which is fast.
5551 Default is @var{freq}.
5554 Set custom positions of virtual loudspeakers. Syntax for this option is:
5555 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
5556 Each virtual loudspeaker is described with short channel name following with
5557 azimuth and elevation in degrees.
5558 Each virtual loudspeaker description is separated by '|'.
5559 For example to override front left and front right channel positions use:
5560 'speakers=FL 45 15|FR 345 15'.
5561 Descriptions with unrecognised channel names are ignored.
5564 Set custom gain for LFE channels. Value is in dB. Default is 0.
5567 Set custom frame size in number of samples. Default is 1024.
5568 Allowed range is from 1024 to 96000. Only used if option @samp{type}
5569 is set to @var{freq}.
5572 Should all IRs be normalized upon importing SOFA file.
5573 By default is enabled.
5576 Should nearest IRs be interpolated with neighbor IRs if exact position
5577 does not match. By default is disabled.
5580 Minphase all IRs upon loading of SOFA file. By default is disabled.
5583 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
5586 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
5589 @subsection Examples
5593 Using ClubFritz6 sofa file:
5595 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
5599 Using ClubFritz12 sofa file and bigger radius with small rotation:
5601 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
5605 Similar as above but with custom speaker positions for front left, front right, back left and back right
5606 and also with custom gain:
5608 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
5615 This filter expands or compresses each half-cycle of audio samples
5616 (local set of samples all above or all below zero and between two nearest zero crossings) depending
5617 on threshold value, so audio reaches target peak value under conditions controlled by below options.
5619 The filter accepts the following options:
5623 Set the expansion target peak value. This specifies the highest allowed absolute amplitude
5624 level for the normalized audio input. Default value is 0.95. Allowed range is from 0.0 to 1.0.
5627 Set the maximum expansion factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5628 This option controls maximum local half-cycle of samples expansion. The maximum expansion
5629 would be such that local peak value reaches target peak value but never to surpass it and that
5630 ratio between new and previous peak value does not surpass this option value.
5632 @item compression, c
5633 Set the maximum compression factor. Allowed range is from 1.0 to 50.0. Default value is 2.0.
5634 This option controls maximum local half-cycle of samples compression. This option is used
5635 only if @option{threshold} option is set to value greater than 0.0, then in such cases
5636 when local peak is lower or same as value set by @option{threshold} all samples belonging to
5637 that peak's half-cycle will be compressed by current compression factor.
5640 Set the threshold value. Default value is 0.0. Allowed range is from 0.0 to 1.0.
5641 This option specifies which half-cycles of samples will be compressed and which will be expanded.
5642 Any half-cycle samples with their local peak value below or same as this option value will be
5643 compressed by current compression factor, otherwise, if greater than threshold value they will be
5644 expanded with expansion factor so that it could reach peak target value but never surpass it.
5647 Set the expansion raising amount per each half-cycle of samples. Default value is 0.001.
5648 Allowed range is from 0.0 to 1.0. This controls how fast expansion factor is raised per
5649 each new half-cycle until it reaches @option{expansion} value.
5650 Setting this options too high may lead to distortions.
5653 Set the compression raising amount per each half-cycle of samples. Default value is 0.001.
5654 Allowed range is from 0.0 to 1.0. This controls how fast compression factor is raised per
5655 each new half-cycle until it reaches @option{compression} value.
5658 Specify which channels to filter, by default all available channels are filtered.
5661 Enable inverted filtering, by default is disabled. This inverts interpretation of @option{threshold}
5662 option. When enabled any half-cycle of samples with their local peak value below or same as
5663 @option{threshold} option will be expanded otherwise it will be compressed.
5666 Link channels when calculating gain applied to each filtered channel sample, by default is disabled.
5667 When disabled each filtered channel gain calculation is independent, otherwise when this option
5668 is enabled the minimum of all possible gains for each filtered channel is used.
5671 @subsection Commands
5673 This filter supports the all above options as @ref{commands}.
5675 @section stereotools
5677 This filter has some handy utilities to manage stereo signals, for converting
5678 M/S stereo recordings to L/R signal while having control over the parameters
5679 or spreading the stereo image of master track.
5681 The filter accepts the following options:
5685 Set input level before filtering for both channels. Defaults is 1.
5686 Allowed range is from 0.015625 to 64.
5689 Set output level after filtering for both channels. Defaults is 1.
5690 Allowed range is from 0.015625 to 64.
5693 Set input balance between both channels. Default is 0.
5694 Allowed range is from -1 to 1.
5697 Set output balance between both channels. Default is 0.
5698 Allowed range is from -1 to 1.
5701 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
5702 clipping. Disabled by default.
5705 Mute the left channel. Disabled by default.
5708 Mute the right channel. Disabled by default.
5711 Change the phase of the left channel. Disabled by default.
5714 Change the phase of the right channel. Disabled by default.
5717 Set stereo mode. Available values are:
5721 Left/Right to Left/Right, this is default.
5724 Left/Right to Mid/Side.
5727 Mid/Side to Left/Right.
5730 Left/Right to Left/Left.
5733 Left/Right to Right/Right.
5736 Left/Right to Left + Right.
5739 Left/Right to Right/Left.
5742 Mid/Side to Left/Left.
5745 Mid/Side to Right/Right.
5748 Mid/Side to Right/Left.
5751 Left/Right to Left - Right.
5755 Set level of side signal. Default is 1.
5756 Allowed range is from 0.015625 to 64.
5759 Set balance of side signal. Default is 0.
5760 Allowed range is from -1 to 1.
5763 Set level of the middle signal. Default is 1.
5764 Allowed range is from 0.015625 to 64.
5767 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
5770 Set stereo base between mono and inversed channels. Default is 0.
5771 Allowed range is from -1 to 1.
5774 Set delay in milliseconds how much to delay left from right channel and
5775 vice versa. Default is 0. Allowed range is from -20 to 20.
5778 Set S/C level. Default is 1. Allowed range is from 1 to 100.
5781 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
5783 @item bmode_in, bmode_out
5784 Set balance mode for balance_in/balance_out option.
5786 Can be one of the following:
5790 Classic balance mode. Attenuate one channel at time.
5791 Gain is raised up to 1.
5794 Similar as classic mode above but gain is raised up to 2.
5797 Equal power distribution, from -6dB to +6dB range.
5801 @subsection Commands
5803 This filter supports the all above options as @ref{commands}.
5805 @subsection Examples
5809 Apply karaoke like effect:
5811 stereotools=mlev=0.015625
5815 Convert M/S signal to L/R:
5817 "stereotools=mode=ms>lr"
5821 @section stereowiden
5823 This filter enhance the stereo effect by suppressing signal common to both
5824 channels and by delaying the signal of left into right and vice versa,
5825 thereby widening the stereo effect.
5827 The filter accepts the following options:
5831 Time in milliseconds of the delay of left signal into right and vice versa.
5832 Default is 20 milliseconds.
5835 Amount of gain in delayed signal into right and vice versa. Gives a delay
5836 effect of left signal in right output and vice versa which gives widening
5837 effect. Default is 0.3.
5840 Cross feed of left into right with inverted phase. This helps in suppressing
5841 the mono. If the value is 1 it will cancel all the signal common to both
5842 channels. Default is 0.3.
5845 Set level of input signal of original channel. Default is 0.8.
5848 @subsection Commands
5850 This filter supports the all above options except @code{delay} as @ref{commands}.
5852 @section superequalizer
5853 Apply 18 band equalizer.
5855 The filter accepts the following options:
5862 Set 131Hz band gain.
5864 Set 185Hz band gain.
5866 Set 262Hz band gain.
5868 Set 370Hz band gain.
5870 Set 523Hz band gain.
5872 Set 740Hz band gain.
5874 Set 1047Hz band gain.
5876 Set 1480Hz band gain.
5878 Set 2093Hz band gain.
5880 Set 2960Hz band gain.
5882 Set 4186Hz band gain.
5884 Set 5920Hz band gain.
5886 Set 8372Hz band gain.
5888 Set 11840Hz band gain.
5890 Set 16744Hz band gain.
5892 Set 20000Hz band gain.
5896 Apply audio surround upmix filter.
5898 This filter allows to produce multichannel output from audio stream.
5900 The filter accepts the following options:
5904 Set output channel layout. By default, this is @var{5.1}.
5906 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5907 for the required syntax.
5910 Set input channel layout. By default, this is @var{stereo}.
5912 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5913 for the required syntax.
5916 Set input volume level. By default, this is @var{1}.
5919 Set output volume level. By default, this is @var{1}.
5922 Enable LFE channel output if output channel layout has it. By default, this is enabled.
5925 Set LFE low cut off frequency. By default, this is @var{128} Hz.
5928 Set LFE high cut off frequency. By default, this is @var{256} Hz.
5931 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
5932 In @var{add} mode, LFE channel is created from input audio and added to output.
5933 In @var{sub} mode, LFE channel is created from input audio and added to output but
5934 also all non-LFE output channels are subtracted with output LFE channel.
5937 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
5938 Default is @var{90}.
5941 Set front center input volume. By default, this is @var{1}.
5944 Set front center output volume. By default, this is @var{1}.
5947 Set front left input volume. By default, this is @var{1}.
5950 Set front left output volume. By default, this is @var{1}.
5953 Set front right input volume. By default, this is @var{1}.
5956 Set front right output volume. By default, this is @var{1}.
5959 Set side left input volume. By default, this is @var{1}.
5962 Set side left output volume. By default, this is @var{1}.
5965 Set side right input volume. By default, this is @var{1}.
5968 Set side right output volume. By default, this is @var{1}.
5971 Set back left input volume. By default, this is @var{1}.
5974 Set back left output volume. By default, this is @var{1}.
5977 Set back right input volume. By default, this is @var{1}.
5980 Set back right output volume. By default, this is @var{1}.
5983 Set back center input volume. By default, this is @var{1}.
5986 Set back center output volume. By default, this is @var{1}.
5989 Set LFE input volume. By default, this is @var{1}.
5992 Set LFE output volume. By default, this is @var{1}.
5995 Set spread usage of stereo image across X axis for all channels.
5998 Set spread usage of stereo image across Y axis for all channels.
6000 @item fcx, flx, frx, blx, brx, slx, srx, bcx
6001 Set spread usage of stereo image across X axis for each channel.
6003 @item fcy, fly, fry, bly, bry, sly, sry, bcy
6004 Set spread usage of stereo image across Y axis for each channel.
6007 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
6010 Set window function.
6012 It accepts the following values:
6035 Default is @code{hann}.
6038 Set window overlap. If set to 1, the recommended overlap for selected
6039 window function will be picked. Default is @code{0.5}.
6042 @section treble, highshelf
6044 Boost or cut treble (upper) frequencies of the audio using a two-pole
6045 shelving filter with a response similar to that of a standard
6046 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
6048 The filter accepts the following options:
6052 Give the gain at whichever is the lower of ~22 kHz and the
6053 Nyquist frequency. Its useful range is about -20 (for a large cut)
6054 to +20 (for a large boost). Beware of clipping when using a positive gain.
6057 Set the filter's central frequency and so can be used
6058 to extend or reduce the frequency range to be boosted or cut.
6059 The default value is @code{3000} Hz.
6062 Set method to specify band-width of filter.
6077 Determine how steep is the filter's shelf transition.
6080 Set number of poles. Default is 2.
6083 How much to use filtered signal in output. Default is 1.
6084 Range is between 0 and 1.
6087 Specify which channels to filter, by default all available are filtered.
6090 Normalize biquad coefficients, by default is disabled.
6091 Enabling it will normalize magnitude response at DC to 0dB.
6094 Set transform type of IIR filter.
6103 Set precison of filtering.
6106 Pick automatic sample format depending on surround filters.
6108 Always use signed 16-bit.
6110 Always use signed 32-bit.
6112 Always use float 32-bit.
6114 Always use float 64-bit.
6118 @subsection Commands
6120 This filter supports the following commands:
6123 Change treble frequency.
6124 Syntax for the command is : "@var{frequency}"
6127 Change treble width_type.
6128 Syntax for the command is : "@var{width_type}"
6131 Change treble width.
6132 Syntax for the command is : "@var{width}"
6136 Syntax for the command is : "@var{gain}"
6140 Syntax for the command is : "@var{mix}"
6145 Sinusoidal amplitude modulation.
6147 The filter accepts the following options:
6151 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
6152 (20 Hz or lower) will result in a tremolo effect.
6153 This filter may also be used as a ring modulator by specifying
6154 a modulation frequency higher than 20 Hz.
6155 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6158 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6159 Default value is 0.5.
6164 Sinusoidal phase modulation.
6166 The filter accepts the following options:
6170 Modulation frequency in Hertz.
6171 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
6174 Depth of modulation as a percentage. Range is 0.0 - 1.0.
6175 Default value is 0.5.
6180 Adjust the input audio volume.
6182 It accepts the following parameters:
6186 Set audio volume expression.
6188 Output values are clipped to the maximum value.
6190 The output audio volume is given by the relation:
6192 @var{output_volume} = @var{volume} * @var{input_volume}
6195 The default value for @var{volume} is "1.0".
6198 This parameter represents the mathematical precision.
6200 It determines which input sample formats will be allowed, which affects the
6201 precision of the volume scaling.
6205 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
6207 32-bit floating-point; this limits input sample format to FLT. (default)
6209 64-bit floating-point; this limits input sample format to DBL.
6213 Choose the behaviour on encountering ReplayGain side data in input frames.
6217 Remove ReplayGain side data, ignoring its contents (the default).
6220 Ignore ReplayGain side data, but leave it in the frame.
6223 Prefer the track gain, if present.
6226 Prefer the album gain, if present.
6229 @item replaygain_preamp
6230 Pre-amplification gain in dB to apply to the selected replaygain gain.
6232 Default value for @var{replaygain_preamp} is 0.0.
6234 @item replaygain_noclip
6235 Prevent clipping by limiting the gain applied.
6237 Default value for @var{replaygain_noclip} is 1.
6240 Set when the volume expression is evaluated.
6242 It accepts the following values:
6245 only evaluate expression once during the filter initialization, or
6246 when the @samp{volume} command is sent
6249 evaluate expression for each incoming frame
6252 Default value is @samp{once}.
6255 The volume expression can contain the following parameters.
6259 frame number (starting at zero)
6262 @item nb_consumed_samples
6263 number of samples consumed by the filter
6265 number of samples in the current frame
6267 original frame position in the file
6273 PTS at start of stream
6275 time at start of stream
6281 last set volume value
6284 Note that when @option{eval} is set to @samp{once} only the
6285 @var{sample_rate} and @var{tb} variables are available, all other
6286 variables will evaluate to NAN.
6288 @subsection Commands
6290 This filter supports the following commands:
6293 Modify the volume expression.
6294 The command accepts the same syntax of the corresponding option.
6296 If the specified expression is not valid, it is kept at its current
6300 @subsection Examples
6304 Halve the input audio volume:
6308 volume=volume=-6.0206dB
6311 In all the above example the named key for @option{volume} can be
6312 omitted, for example like in:
6318 Increase input audio power by 6 decibels using fixed-point precision:
6320 volume=volume=6dB:precision=fixed
6324 Fade volume after time 10 with an annihilation period of 5 seconds:
6326 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
6330 @section volumedetect
6332 Detect the volume of the input video.
6334 The filter has no parameters. The input is not modified. Statistics about
6335 the volume will be printed in the log when the input stream end is reached.
6337 In particular it will show the mean volume (root mean square), maximum
6338 volume (on a per-sample basis), and the beginning of a histogram of the
6339 registered volume values (from the maximum value to a cumulated 1/1000 of
6342 All volumes are in decibels relative to the maximum PCM value.
6344 @subsection Examples
6346 Here is an excerpt of the output:
6348 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
6349 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
6350 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
6351 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
6352 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
6353 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
6354 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
6355 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
6356 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
6362 The mean square energy is approximately -27 dB, or 10^-2.7.
6364 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
6366 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
6369 In other words, raising the volume by +4 dB does not cause any clipping,
6370 raising it by +5 dB causes clipping for 6 samples, etc.
6372 @c man end AUDIO FILTERS
6374 @chapter Audio Sources
6375 @c man begin AUDIO SOURCES
6377 Below is a description of the currently available audio sources.
6381 Buffer audio frames, and make them available to the filter chain.
6383 This source is mainly intended for a programmatic use, in particular
6384 through the interface defined in @file{libavfilter/buffersrc.h}.
6386 It accepts the following parameters:
6390 The timebase which will be used for timestamps of submitted frames. It must be
6391 either a floating-point number or in @var{numerator}/@var{denominator} form.
6394 The sample rate of the incoming audio buffers.
6397 The sample format of the incoming audio buffers.
6398 Either a sample format name or its corresponding integer representation from
6399 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
6401 @item channel_layout
6402 The channel layout of the incoming audio buffers.
6403 Either a channel layout name from channel_layout_map in
6404 @file{libavutil/channel_layout.c} or its corresponding integer representation
6405 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
6408 The number of channels of the incoming audio buffers.
6409 If both @var{channels} and @var{channel_layout} are specified, then they
6414 @subsection Examples
6417 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
6420 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
6421 Since the sample format with name "s16p" corresponds to the number
6422 6 and the "stereo" channel layout corresponds to the value 0x3, this is
6425 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
6430 Generate an audio signal specified by an expression.
6432 This source accepts in input one or more expressions (one for each
6433 channel), which are evaluated and used to generate a corresponding
6436 This source accepts the following options:
6440 Set the '|'-separated expressions list for each separate channel. In case the
6441 @option{channel_layout} option is not specified, the selected channel layout
6442 depends on the number of provided expressions. Otherwise the last
6443 specified expression is applied to the remaining output channels.
6445 @item channel_layout, c
6446 Set the channel layout. The number of channels in the specified layout
6447 must be equal to the number of specified expressions.
6450 Set the minimum duration of the sourced audio. See
6451 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6452 for the accepted syntax.
6453 Note that the resulting duration may be greater than the specified
6454 duration, as the generated audio is always cut at the end of a
6457 If not specified, or the expressed duration is negative, the audio is
6458 supposed to be generated forever.
6461 Set the number of samples per channel per each output frame,
6464 @item sample_rate, s
6465 Specify the sample rate, default to 44100.
6468 Each expression in @var{exprs} can contain the following constants:
6472 number of the evaluated sample, starting from 0
6475 time of the evaluated sample expressed in seconds, starting from 0
6482 @subsection Examples
6492 Generate a sin signal with frequency of 440 Hz, set sample rate to
6495 aevalsrc="sin(440*2*PI*t):s=8000"
6499 Generate a two channels signal, specify the channel layout (Front
6500 Center + Back Center) explicitly:
6502 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
6506 Generate white noise:
6508 aevalsrc="-2+random(0)"
6512 Generate an amplitude modulated signal:
6514 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
6518 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
6520 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
6527 Generate a FIR coefficients using frequency sampling method.
6529 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6531 The filter accepts the following options:
6535 Set number of filter coefficents in output audio stream.
6536 Default value is 1025.
6539 Set frequency points from where magnitude and phase are set.
6540 This must be in non decreasing order, and first element must be 0, while last element
6541 must be 1. Elements are separated by white spaces.
6544 Set magnitude value for every frequency point set by @option{frequency}.
6545 Number of values must be same as number of frequency points.
6546 Values are separated by white spaces.
6549 Set phase value for every frequency point set by @option{frequency}.
6550 Number of values must be same as number of frequency points.
6551 Values are separated by white spaces.
6553 @item sample_rate, r
6554 Set sample rate, default is 44100.
6557 Set number of samples per each frame. Default is 1024.
6560 Set window function. Default is blackman.
6565 The null audio source, return unprocessed audio frames. It is mainly useful
6566 as a template and to be employed in analysis / debugging tools, or as
6567 the source for filters which ignore the input data (for example the sox
6570 This source accepts the following options:
6574 @item channel_layout, cl
6576 Specifies the channel layout, and can be either an integer or a string
6577 representing a channel layout. The default value of @var{channel_layout}
6580 Check the channel_layout_map definition in
6581 @file{libavutil/channel_layout.c} for the mapping between strings and
6582 channel layout values.
6584 @item sample_rate, r
6585 Specifies the sample rate, and defaults to 44100.
6588 Set the number of samples per requested frames.
6591 Set the duration of the sourced audio. See
6592 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
6593 for the accepted syntax.
6595 If not specified, or the expressed duration is negative, the audio is
6596 supposed to be generated forever.
6599 @subsection Examples
6603 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
6605 anullsrc=r=48000:cl=4
6609 Do the same operation with a more obvious syntax:
6611 anullsrc=r=48000:cl=mono
6615 All the parameters need to be explicitly defined.
6619 Synthesize a voice utterance using the libflite library.
6621 To enable compilation of this filter you need to configure FFmpeg with
6622 @code{--enable-libflite}.
6624 Note that versions of the flite library prior to 2.0 are not thread-safe.
6626 The filter accepts the following options:
6631 If set to 1, list the names of the available voices and exit
6632 immediately. Default value is 0.
6635 Set the maximum number of samples per frame. Default value is 512.
6638 Set the filename containing the text to speak.
6641 Set the text to speak.
6644 Set the voice to use for the speech synthesis. Default value is
6645 @code{kal}. See also the @var{list_voices} option.
6648 @subsection Examples
6652 Read from file @file{speech.txt}, and synthesize the text using the
6653 standard flite voice:
6655 flite=textfile=speech.txt
6659 Read the specified text selecting the @code{slt} voice:
6661 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6665 Input text to ffmpeg:
6667 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
6671 Make @file{ffplay} speak the specified text, using @code{flite} and
6672 the @code{lavfi} device:
6674 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
6678 For more information about libflite, check:
6679 @url{http://www.festvox.org/flite/}
6683 Generate a noise audio signal.
6685 The filter accepts the following options:
6688 @item sample_rate, r
6689 Specify the sample rate. Default value is 48000 Hz.
6692 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
6696 Specify the duration of the generated audio stream. Not specifying this option
6697 results in noise with an infinite length.
6699 @item color, colour, c
6700 Specify the color of noise. Available noise colors are white, pink, brown,
6701 blue, violet and velvet. Default color is white.
6704 Specify a value used to seed the PRNG.
6707 Set the number of samples per each output frame, default is 1024.
6710 @subsection Examples
6715 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
6717 anoisesrc=d=60:c=pink:r=44100:a=0.5
6723 Generate odd-tap Hilbert transform FIR coefficients.
6725 The resulting stream can be used with @ref{afir} filter for phase-shifting
6726 the signal by 90 degrees.
6728 This is used in many matrix coding schemes and for analytic signal generation.
6729 The process is often written as a multiplication by i (or j), the imaginary unit.
6731 The filter accepts the following options:
6735 @item sample_rate, s
6736 Set sample rate, default is 44100.
6739 Set length of FIR filter, default is 22051.
6742 Set number of samples per each frame.
6745 Set window function to be used when generating FIR coefficients.
6750 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
6752 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
6754 The filter accepts the following options:
6757 @item sample_rate, r
6758 Set sample rate, default is 44100.
6761 Set number of samples per each frame. Default is 1024.
6764 Set high-pass frequency. Default is 0.
6767 Set low-pass frequency. Default is 0.
6768 If high-pass frequency is lower than low-pass frequency and low-pass frequency
6769 is higher than 0 then filter will create band-pass filter coefficients,
6770 otherwise band-reject filter coefficients.
6773 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
6776 Set Kaiser window beta.
6779 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
6782 Enable rounding, by default is disabled.
6785 Set number of taps for high-pass filter.
6788 Set number of taps for low-pass filter.
6793 Generate an audio signal made of a sine wave with amplitude 1/8.
6795 The audio signal is bit-exact.
6797 The filter accepts the following options:
6802 Set the carrier frequency. Default is 440 Hz.
6804 @item beep_factor, b
6805 Enable a periodic beep every second with frequency @var{beep_factor} times
6806 the carrier frequency. Default is 0, meaning the beep is disabled.
6808 @item sample_rate, r
6809 Specify the sample rate, default is 44100.
6812 Specify the duration of the generated audio stream.
6814 @item samples_per_frame
6815 Set the number of samples per output frame.
6817 The expression can contain the following constants:
6821 The (sequential) number of the output audio frame, starting from 0.
6824 The PTS (Presentation TimeStamp) of the output audio frame,
6825 expressed in @var{TB} units.
6828 The PTS of the output audio frame, expressed in seconds.
6831 The timebase of the output audio frames.
6834 Default is @code{1024}.
6837 @subsection Examples
6842 Generate a simple 440 Hz sine wave:
6848 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
6852 sine=frequency=220:beep_factor=4:duration=5
6856 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
6859 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
6863 @c man end AUDIO SOURCES
6865 @chapter Audio Sinks
6866 @c man begin AUDIO SINKS
6868 Below is a description of the currently available audio sinks.
6870 @section abuffersink
6872 Buffer audio frames, and make them available to the end of filter chain.
6874 This sink is mainly intended for programmatic use, in particular
6875 through the interface defined in @file{libavfilter/buffersink.h}
6876 or the options system.
6878 It accepts a pointer to an AVABufferSinkContext structure, which
6879 defines the incoming buffers' formats, to be passed as the opaque
6880 parameter to @code{avfilter_init_filter} for initialization.
6883 Null audio sink; do absolutely nothing with the input audio. It is
6884 mainly useful as a template and for use in analysis / debugging
6887 @c man end AUDIO SINKS
6889 @chapter Video Filters
6890 @c man begin VIDEO FILTERS
6892 When you configure your FFmpeg build, you can disable any of the
6893 existing filters using @code{--disable-filters}.
6894 The configure output will show the video filters included in your
6897 Below is a description of the currently available video filters.
6901 Mark a region of interest in a video frame.
6903 The frame data is passed through unchanged, but metadata is attached
6904 to the frame indicating regions of interest which can affect the
6905 behaviour of later encoding. Multiple regions can be marked by
6906 applying the filter multiple times.
6910 Region distance in pixels from the left edge of the frame.
6912 Region distance in pixels from the top edge of the frame.
6914 Region width in pixels.
6916 Region height in pixels.
6918 The parameters @var{x}, @var{y}, @var{w} and @var{h} are expressions,
6919 and may contain the following variables:
6922 Width of the input frame.
6924 Height of the input frame.
6928 Quantisation offset to apply within the region.
6930 This must be a real value in the range -1 to +1. A value of zero
6931 indicates no quality change. A negative value asks for better quality
6932 (less quantisation), while a positive value asks for worse quality
6933 (greater quantisation).
6935 The range is calibrated so that the extreme values indicate the
6936 largest possible offset - if the rest of the frame is encoded with the
6937 worst possible quality, an offset of -1 indicates that this region
6938 should be encoded with the best possible quality anyway. Intermediate
6939 values are then interpolated in some codec-dependent way.
6941 For example, in 10-bit H.264 the quantisation parameter varies between
6942 -12 and 51. A typical qoffset value of -1/10 therefore indicates that
6943 this region should be encoded with a QP around one-tenth of the full
6944 range better than the rest of the frame. So, if most of the frame
6945 were to be encoded with a QP of around 30, this region would get a QP
6946 of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3).
6947 An extreme value of -1 would indicate that this region should be
6948 encoded with the best possible quality regardless of the treatment of
6949 the rest of the frame - that is, should be encoded at a QP of -12.
6951 If set to true, remove any existing regions of interest marked on the
6952 frame before adding the new one.
6955 @subsection Examples
6959 Mark the centre quarter of the frame as interesting.
6961 addroi=iw/4:ih/4:iw/2:ih/2:-1/10
6964 Mark the 100-pixel-wide region on the left edge of the frame as very
6965 uninteresting (to be encoded at much lower quality than the rest of
6968 addroi=0:0:100:ih:+1/5
6972 @section alphaextract
6974 Extract the alpha component from the input as a grayscale video. This
6975 is especially useful with the @var{alphamerge} filter.
6979 Add or replace the alpha component of the primary input with the
6980 grayscale value of a second input. This is intended for use with
6981 @var{alphaextract} to allow the transmission or storage of frame
6982 sequences that have alpha in a format that doesn't support an alpha
6985 For example, to reconstruct full frames from a normal YUV-encoded video
6986 and a separate video created with @var{alphaextract}, you might use:
6988 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
6993 Amplify differences between current pixel and pixels of adjacent frames in
6994 same pixel location.
6996 This filter accepts the following options:
7000 Set frame radius. Default is 2. Allowed range is from 1 to 63.
7001 For example radius of 3 will instruct filter to calculate average of 7 frames.
7004 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
7007 Set threshold for difference amplification. Any difference greater or equal to
7008 this value will not alter source pixel. Default is 10.
7009 Allowed range is from 0 to 65535.
7012 Set tolerance for difference amplification. Any difference lower to
7013 this value will not alter source pixel. Default is 0.
7014 Allowed range is from 0 to 65535.
7017 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7018 This option controls maximum possible value that will decrease source pixel value.
7021 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
7022 This option controls maximum possible value that will increase source pixel value.
7025 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
7028 @subsection Commands
7030 This filter supports the following @ref{commands} that corresponds to option of same name:
7042 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
7043 and libavformat to work. On the other hand, it is limited to ASS (Advanced
7044 Substation Alpha) subtitles files.
7046 This filter accepts the following option in addition to the common options from
7047 the @ref{subtitles} filter:
7051 Set the shaping engine
7053 Available values are:
7056 The default libass shaping engine, which is the best available.
7058 Fast, font-agnostic shaper that can do only substitutions
7060 Slower shaper using OpenType for substitutions and positioning
7063 The default is @code{auto}.
7067 Apply an Adaptive Temporal Averaging Denoiser to the video input.
7069 The filter accepts the following options:
7073 Set threshold A for 1st plane. Default is 0.02.
7074 Valid range is 0 to 0.3.
7077 Set threshold B for 1st plane. Default is 0.04.
7078 Valid range is 0 to 5.
7081 Set threshold A for 2nd plane. Default is 0.02.
7082 Valid range is 0 to 0.3.
7085 Set threshold B for 2nd plane. Default is 0.04.
7086 Valid range is 0 to 5.
7089 Set threshold A for 3rd plane. Default is 0.02.
7090 Valid range is 0 to 0.3.
7093 Set threshold B for 3rd plane. Default is 0.04.
7094 Valid range is 0 to 5.
7096 Threshold A is designed to react on abrupt changes in the input signal and
7097 threshold B is designed to react on continuous changes in the input signal.
7100 Set number of frames filter will use for averaging. Default is 9. Must be odd
7101 number in range [5, 129].
7104 Set what planes of frame filter will use for averaging. Default is all.
7107 Set what variant of algorithm filter will use for averaging. Default is @code{p} parallel.
7108 Alternatively can be set to @code{s} serial.
7110 Parallel can be faster then serial, while other way around is never true.
7111 Parallel will abort early on first change being greater then thresholds, while serial
7112 will continue processing other side of frames if they are equal or below thresholds.
7117 Set sigma for 1st plane, 2nd plane or 3rd plane. Default is 32767.
7118 Valid range is from 0 to 32767.
7119 This options controls weight for each pixel in radius defined by size.
7120 Default value means every pixel have same weight.
7121 Setting this option to 0 effectively disables filtering.
7124 @subsection Commands
7125 This filter supports same @ref{commands} as options except option @code{s}.
7126 The command accepts the same syntax of the corresponding option.
7130 Apply average blur filter.
7132 The filter accepts the following options:
7136 Set horizontal radius size.
7139 Set which planes to filter. By default all planes are filtered.
7142 Set vertical radius size, if zero it will be same as @code{sizeX}.
7143 Default is @code{0}.
7146 @subsection Commands
7147 This filter supports same commands as options.
7148 The command accepts the same syntax of the corresponding option.
7150 If the specified expression is not valid, it is kept at its current
7155 Compute the bounding box for the non-black pixels in the input frame
7158 This filter computes the bounding box containing all the pixels with a
7159 luminance value greater than the minimum allowed value.
7160 The parameters describing the bounding box are printed on the filter
7163 The filter accepts the following option:
7167 Set the minimal luminance value. Default is @code{16}.
7170 @subsection Commands
7172 This filter supports the all above options as @ref{commands}.
7175 Apply bilateral filter, spatial smoothing while preserving edges.
7177 The filter accepts the following options:
7180 Set sigma of gaussian function to calculate spatial weight.
7181 Allowed range is 0 to 512. Default is 0.1.
7184 Set sigma of gaussian function to calculate range weight.
7185 Allowed range is 0 to 1. Default is 0.1.
7188 Set planes to filter. Default is first only.
7191 @subsection Commands
7193 This filter supports the all above options as @ref{commands}.
7195 @section bitplanenoise
7197 Show and measure bit plane noise.
7199 The filter accepts the following options:
7203 Set which plane to analyze. Default is @code{1}.
7206 Filter out noisy pixels from @code{bitplane} set above.
7207 Default is disabled.
7210 @section blackdetect
7212 Detect video intervals that are (almost) completely black. Can be
7213 useful to detect chapter transitions, commercials, or invalid
7216 The filter outputs its detection analysis to both the log as well as
7217 frame metadata. If a black segment of at least the specified minimum
7218 duration is found, a line with the start and end timestamps as well
7219 as duration is printed to the log with level @code{info}. In addition,
7220 a log line with level @code{debug} is printed per frame showing the
7221 black amount detected for that frame.
7223 The filter also attaches metadata to the first frame of a black
7224 segment with key @code{lavfi.black_start} and to the first frame
7225 after the black segment ends with key @code{lavfi.black_end}. The
7226 value is the frame's timestamp. This metadata is added regardless
7227 of the minimum duration specified.
7229 The filter accepts the following options:
7232 @item black_min_duration, d
7233 Set the minimum detected black duration expressed in seconds. It must
7234 be a non-negative floating point number.
7236 Default value is 2.0.
7238 @item picture_black_ratio_th, pic_th
7239 Set the threshold for considering a picture "black".
7240 Express the minimum value for the ratio:
7242 @var{nb_black_pixels} / @var{nb_pixels}
7245 for which a picture is considered black.
7246 Default value is 0.98.
7248 @item pixel_black_th, pix_th
7249 Set the threshold for considering a pixel "black".
7251 The threshold expresses the maximum pixel luminance value for which a
7252 pixel is considered "black". The provided value is scaled according to
7253 the following equation:
7255 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
7258 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
7259 the input video format, the range is [0-255] for YUV full-range
7260 formats and [16-235] for YUV non full-range formats.
7262 Default value is 0.10.
7265 The following example sets the maximum pixel threshold to the minimum
7266 value, and detects only black intervals of 2 or more seconds:
7268 blackdetect=d=2:pix_th=0.00
7273 Detect frames that are (almost) completely black. Can be useful to
7274 detect chapter transitions or commercials. Output lines consist of
7275 the frame number of the detected frame, the percentage of blackness,
7276 the position in the file if known or -1 and the timestamp in seconds.
7278 In order to display the output lines, you need to set the loglevel at
7279 least to the AV_LOG_INFO value.
7281 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
7282 The value represents the percentage of pixels in the picture that
7283 are below the threshold value.
7285 It accepts the following parameters:
7290 The percentage of the pixels that have to be below the threshold; it defaults to
7293 @item threshold, thresh
7294 The threshold below which a pixel value is considered black; it defaults to
7302 Blend two video frames into each other.
7304 The @code{blend} filter takes two input streams and outputs one
7305 stream, the first input is the "top" layer and second input is
7306 "bottom" layer. By default, the output terminates when the longest input terminates.
7308 The @code{tblend} (time blend) filter takes two consecutive frames
7309 from one single stream, and outputs the result obtained by blending
7310 the new frame on top of the old frame.
7312 A description of the accepted options follows.
7320 Set blend mode for specific pixel component or all pixel components in case
7321 of @var{all_mode}. Default value is @code{normal}.
7323 Available values for component modes are:
7365 Set blend opacity for specific pixel component or all pixel components in case
7366 of @var{all_opacity}. Only used in combination with pixel component blend modes.
7373 Set blend expression for specific pixel component or all pixel components in case
7374 of @var{all_expr}. Note that related mode options will be ignored if those are set.
7376 The expressions can use the following variables:
7380 The sequential number of the filtered frame, starting from @code{0}.
7384 the coordinates of the current sample
7388 the width and height of currently filtered plane
7392 Width and height scale for the plane being filtered. It is the
7393 ratio between the dimensions of the current plane to the luma plane,
7394 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
7395 the luma plane and @code{0.5,0.5} for the chroma planes.
7398 Time of the current frame, expressed in seconds.
7401 Value of pixel component at current location for first video frame (top layer).
7404 Value of pixel component at current location for second video frame (bottom layer).
7408 The @code{blend} filter also supports the @ref{framesync} options.
7410 @subsection Examples
7414 Apply transition from bottom layer to top layer in first 10 seconds:
7416 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
7420 Apply linear horizontal transition from top layer to bottom layer:
7422 blend=all_expr='A*(X/W)+B*(1-X/W)'
7426 Apply 1x1 checkerboard effect:
7428 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
7432 Apply uncover left effect:
7434 blend=all_expr='if(gte(N*SW+X,W),A,B)'
7438 Apply uncover down effect:
7440 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
7444 Apply uncover up-left effect:
7446 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
7450 Split diagonally video and shows top and bottom layer on each side:
7452 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
7456 Display differences between the current and the previous frame:
7458 tblend=all_mode=grainextract
7462 @subsection Commands
7463 This filter supports same @ref{commands} as options.
7467 Denoise frames using Block-Matching 3D algorithm.
7469 The filter accepts the following options.
7473 Set denoising strength. Default value is 1.
7474 Allowed range is from 0 to 999.9.
7475 The denoising algorithm is very sensitive to sigma, so adjust it
7476 according to the source.
7479 Set local patch size. This sets dimensions in 2D.
7482 Set sliding step for processing blocks. Default value is 4.
7483 Allowed range is from 1 to 64.
7484 Smaller values allows processing more reference blocks and is slower.
7487 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
7488 When set to 1, no block matching is done. Larger values allows more blocks
7490 Allowed range is from 1 to 256.
7493 Set radius for search block matching. Default is 9.
7494 Allowed range is from 1 to INT32_MAX.
7497 Set step between two search locations for block matching. Default is 1.
7498 Allowed range is from 1 to 64. Smaller is slower.
7501 Set threshold of mean square error for block matching. Valid range is 0 to
7505 Set thresholding parameter for hard thresholding in 3D transformed domain.
7506 Larger values results in stronger hard-thresholding filtering in frequency
7510 Set filtering estimation mode. Can be @code{basic} or @code{final}.
7511 Default is @code{basic}.
7514 If enabled, filter will use 2nd stream for block matching.
7515 Default is disabled for @code{basic} value of @var{estim} option,
7516 and always enabled if value of @var{estim} is @code{final}.
7519 Set planes to filter. Default is all available except alpha.
7522 @subsection Examples
7526 Basic filtering with bm3d:
7528 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
7532 Same as above, but filtering only luma:
7534 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
7538 Same as above, but with both estimation modes:
7540 split[a][b],[a]bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
7544 Same as above, but prefilter with @ref{nlmeans} filter instead:
7546 split[a][b],[a]nlmeans=s=3:r=7:p=3[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
7552 Apply a boxblur algorithm to the input video.
7554 It accepts the following parameters:
7558 @item luma_radius, lr
7559 @item luma_power, lp
7560 @item chroma_radius, cr
7561 @item chroma_power, cp
7562 @item alpha_radius, ar
7563 @item alpha_power, ap
7567 A description of the accepted options follows.
7570 @item luma_radius, lr
7571 @item chroma_radius, cr
7572 @item alpha_radius, ar
7573 Set an expression for the box radius in pixels used for blurring the
7574 corresponding input plane.
7576 The radius value must be a non-negative number, and must not be
7577 greater than the value of the expression @code{min(w,h)/2} for the
7578 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
7581 Default value for @option{luma_radius} is "2". If not specified,
7582 @option{chroma_radius} and @option{alpha_radius} default to the
7583 corresponding value set for @option{luma_radius}.
7585 The expressions can contain the following constants:
7589 The input width and height in pixels.
7593 The input chroma image width and height in pixels.
7597 The horizontal and vertical chroma subsample values. For example, for the
7598 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
7601 @item luma_power, lp
7602 @item chroma_power, cp
7603 @item alpha_power, ap
7604 Specify how many times the boxblur filter is applied to the
7605 corresponding plane.
7607 Default value for @option{luma_power} is 2. If not specified,
7608 @option{chroma_power} and @option{alpha_power} default to the
7609 corresponding value set for @option{luma_power}.
7611 A value of 0 will disable the effect.
7614 @subsection Examples
7618 Apply a boxblur filter with the luma, chroma, and alpha radii
7621 boxblur=luma_radius=2:luma_power=1
7626 Set the luma radius to 2, and alpha and chroma radius to 0:
7628 boxblur=2:1:cr=0:ar=0
7632 Set the luma and chroma radii to a fraction of the video dimension:
7634 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
7640 Deinterlace the input video ("bwdif" stands for "Bob Weaver
7641 Deinterlacing Filter").
7643 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
7644 interpolation algorithms.
7645 It accepts the following parameters:
7649 The interlacing mode to adopt. It accepts one of the following values:
7653 Output one frame for each frame.
7655 Output one frame for each field.
7658 The default value is @code{send_field}.
7661 The picture field parity assumed for the input interlaced video. It accepts one
7662 of the following values:
7666 Assume the top field is first.
7668 Assume the bottom field is first.
7670 Enable automatic detection of field parity.
7673 The default value is @code{auto}.
7674 If the interlacing is unknown or the decoder does not export this information,
7675 top field first will be assumed.
7678 Specify which frames to deinterlace. Accepts one of the following
7683 Deinterlace all frames.
7685 Only deinterlace frames marked as interlaced.
7688 The default value is @code{all}.
7693 Apply Contrast Adaptive Sharpen filter to video stream.
7695 The filter accepts the following options:
7699 Set the sharpening strength. Default value is 0.
7702 Set planes to filter. Default value is to filter all
7703 planes except alpha plane.
7706 @subsection Commands
7707 This filter supports same @ref{commands} as options.
7710 Remove all color information for all colors except for certain one.
7712 The filter accepts the following options:
7716 The color which will not be replaced with neutral chroma.
7719 Similarity percentage with the above color.
7720 0.01 matches only the exact key color, while 1.0 matches everything.
7724 0.0 makes pixels either fully gray, or not gray at all.
7725 Higher values result in more preserved color.
7728 Signals that the color passed is already in YUV instead of RGB.
7730 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7731 This can be used to pass exact YUV values as hexadecimal numbers.
7734 @subsection Commands
7735 This filter supports same @ref{commands} as options.
7736 The command accepts the same syntax of the corresponding option.
7738 If the specified expression is not valid, it is kept at its current
7742 YUV colorspace color/chroma keying.
7744 The filter accepts the following options:
7748 The color which will be replaced with transparency.
7751 Similarity percentage with the key color.
7753 0.01 matches only the exact key color, while 1.0 matches everything.
7758 0.0 makes pixels either fully transparent, or not transparent at all.
7760 Higher values result in semi-transparent pixels, with a higher transparency
7761 the more similar the pixels color is to the key color.
7764 Signals that the color passed is already in YUV instead of RGB.
7766 Literal colors like "green" or "red" don't make sense with this enabled anymore.
7767 This can be used to pass exact YUV values as hexadecimal numbers.
7770 @subsection Commands
7771 This filter supports same @ref{commands} as options.
7772 The command accepts the same syntax of the corresponding option.
7774 If the specified expression is not valid, it is kept at its current
7777 @subsection Examples
7781 Make every green pixel in the input image transparent:
7783 ffmpeg -i input.png -vf chromakey=green out.png
7787 Overlay a greenscreen-video on top of a static black background.
7789 ffmpeg -f lavfi -i color=c=black:s=1280x720 -i video.mp4 -shortest -filter_complex "[1:v]chromakey=0x70de77:0.1:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.mkv
7794 Reduce chrominance noise.
7796 The filter accepts the following options:
7800 Set threshold for averaging chrominance values.
7801 Sum of absolute difference of Y, U and V pixel components of current
7802 pixel and neighbour pixels lower than this threshold will be used in
7803 averaging. Luma component is left unchanged and is copied to output.
7804 Default value is 30. Allowed range is from 1 to 200.
7807 Set horizontal radius of rectangle used for averaging.
7808 Allowed range is from 1 to 100. Default value is 5.
7811 Set vertical radius of rectangle used for averaging.
7812 Allowed range is from 1 to 100. Default value is 5.
7815 Set horizontal step when averaging. Default value is 1.
7816 Allowed range is from 1 to 50.
7817 Mostly useful to speed-up filtering.
7820 Set vertical step when averaging. Default value is 1.
7821 Allowed range is from 1 to 50.
7822 Mostly useful to speed-up filtering.
7825 Set Y threshold for averaging chrominance values.
7826 Set finer control for max allowed difference between Y components
7827 of current pixel and neigbour pixels.
7828 Default value is 200. Allowed range is from 1 to 200.
7831 Set U threshold for averaging chrominance values.
7832 Set finer control for max allowed difference between U components
7833 of current pixel and neigbour pixels.
7834 Default value is 200. Allowed range is from 1 to 200.
7837 Set V threshold for averaging chrominance values.
7838 Set finer control for max allowed difference between V components
7839 of current pixel and neigbour pixels.
7840 Default value is 200. Allowed range is from 1 to 200.
7843 @subsection Commands
7844 This filter supports same @ref{commands} as options.
7845 The command accepts the same syntax of the corresponding option.
7847 @section chromashift
7848 Shift chroma pixels horizontally and/or vertically.
7850 The filter accepts the following options:
7853 Set amount to shift chroma-blue horizontally.
7855 Set amount to shift chroma-blue vertically.
7857 Set amount to shift chroma-red horizontally.
7859 Set amount to shift chroma-red vertically.
7861 Set edge mode, can be @var{smear}, default, or @var{warp}.
7864 @subsection Commands
7866 This filter supports the all above options as @ref{commands}.
7870 Display CIE color diagram with pixels overlaid onto it.
7872 The filter accepts the following options:
7887 @item uhdtv, rec2020
7901 Set what gamuts to draw.
7903 See @code{system} option for available values.
7906 Set ciescope size, by default set to 512.
7909 Set intensity used to map input pixel values to CIE diagram.
7912 Set contrast used to draw tongue colors that are out of active color system gamut.
7915 Correct gamma displayed on scope, by default enabled.
7918 Show white point on CIE diagram, by default disabled.
7921 Set input gamma. Used only with XYZ input color space.
7926 Visualize information exported by some codecs.
7928 Some codecs can export information through frames using side-data or other
7929 means. For example, some MPEG based codecs export motion vectors through the
7930 @var{export_mvs} flag in the codec @option{flags2} option.
7932 The filter accepts the following option:
7936 Set motion vectors to visualize.
7938 Available flags for @var{mv} are:
7942 forward predicted MVs of P-frames
7944 forward predicted MVs of B-frames
7946 backward predicted MVs of B-frames
7950 Display quantization parameters using the chroma planes.
7953 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
7955 Available flags for @var{mv_type} are:
7959 forward predicted MVs
7961 backward predicted MVs
7964 @item frame_type, ft
7965 Set frame type to visualize motion vectors of.
7967 Available flags for @var{frame_type} are:
7971 intra-coded frames (I-frames)
7973 predicted frames (P-frames)
7975 bi-directionally predicted frames (B-frames)
7979 @subsection Examples
7983 Visualize forward predicted MVs of all frames using @command{ffplay}:
7985 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
7989 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
7991 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
7995 @section colorbalance
7996 Modify intensity of primary colors (red, green and blue) of input frames.
7998 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
7999 regions for the red-cyan, green-magenta or blue-yellow balance.
8001 A positive adjustment value shifts the balance towards the primary color, a negative
8002 value towards the complementary color.
8004 The filter accepts the following options:
8010 Adjust red, green and blue shadows (darkest pixels).
8015 Adjust red, green and blue midtones (medium pixels).
8020 Adjust red, green and blue highlights (brightest pixels).
8022 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8025 Preserve lightness when changing color balance. Default is disabled.
8028 @subsection Examples
8032 Add red color cast to shadows:
8038 @subsection Commands
8040 This filter supports the all above options as @ref{commands}.
8042 @section colorcontrast
8044 Adjust color contrast between RGB components.
8046 The filter accepts the following options:
8050 Set the red-cyan contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8053 Set the green-magenta contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8056 Set the blue-yellow contrast. Defaults is 0.0. Allowed range is from -1.0 to 1.0.
8061 Set the weight of each @code{rc}, @code{gm}, @code{by} option value. Default value is 0.0.
8062 Allowed range is from 0.0 to 1.0. If all weights are 0.0 filtering is disabled.
8065 Set the amount of preserving lightness. Default value is 0.0. Allowed range is from 0.0 to 1.0.
8068 @subsection Commands
8070 This filter supports the all above options as @ref{commands}.
8072 @section colorcorrect
8074 Adjust color white balance selectively for blacks and whites.
8075 This filter operates in YUV colorspace.
8077 The filter accepts the following options:
8081 Set the red shadow spot. Allowed range is from -1.0 to 1.0.
8085 Set the blue shadow spot. Allowed range is from -1.0 to 1.0.
8089 Set the red highlight spot. Allowed range is from -1.0 to 1.0.
8093 Set the red highlight spot. Allowed range is from -1.0 to 1.0.
8097 Set the amount of saturation. Allowed range is from -3.0 to 3.0.
8101 @subsection Commands
8103 This filter supports the all above options as @ref{commands}.
8105 @section colorchannelmixer
8107 Adjust video input frames by re-mixing color channels.
8109 This filter modifies a color channel by adding the values associated to
8110 the other channels of the same pixels. For example if the value to
8111 modify is red, the output value will be:
8113 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
8116 The filter accepts the following options:
8123 Adjust contribution of input red, green, blue and alpha channels for output red channel.
8124 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
8130 Adjust contribution of input red, green, blue and alpha channels for output green channel.
8131 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
8137 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
8138 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
8144 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
8145 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
8147 Allowed ranges for options are @code{[-2.0, 2.0]}.
8150 Preserve lightness when changing colors. Allowed range is from @code{[0.0, 1.0]}.
8151 Default is @code{0.0}, thus disabled.
8154 @subsection Examples
8158 Convert source to grayscale:
8160 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
8163 Simulate sepia tones:
8165 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
8169 @subsection Commands
8171 This filter supports the all above options as @ref{commands}.
8174 Overlay a solid color on the video stream.
8176 The filter accepts the following options:
8180 Set the color hue. Allowed range is from 0 to 360.
8184 Set the color saturation. Allowed range is from 0 to 1.
8185 Default value is 0.5.
8188 Set the color lightness. Allowed range is from 0 to 1.
8189 Default value is 0.5.
8192 Set the mix of source lightness. By default is set to 1.0.
8193 Allowed range is from 0.0 to 1.0.
8196 @subsection Commands
8198 This filter supports the all above options as @ref{commands}.
8201 RGB colorspace color keying.
8203 The filter accepts the following options:
8207 The color which will be replaced with transparency.
8210 Similarity percentage with the key color.
8212 0.01 matches only the exact key color, while 1.0 matches everything.
8217 0.0 makes pixels either fully transparent, or not transparent at all.
8219 Higher values result in semi-transparent pixels, with a higher transparency
8220 the more similar the pixels color is to the key color.
8223 @subsection Examples
8227 Make every green pixel in the input image transparent:
8229 ffmpeg -i input.png -vf colorkey=green out.png
8233 Overlay a greenscreen-video on top of a static background image.
8235 ffmpeg -i background.png -i video.mp4 -filter_complex "[1:v]colorkey=0x3BBD1E:0.3:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.flv
8239 @subsection Commands
8240 This filter supports same @ref{commands} as options.
8241 The command accepts the same syntax of the corresponding option.
8243 If the specified expression is not valid, it is kept at its current
8247 Remove all color information for all RGB colors except for certain one.
8249 The filter accepts the following options:
8253 The color which will not be replaced with neutral gray.
8256 Similarity percentage with the above color.
8257 0.01 matches only the exact key color, while 1.0 matches everything.
8260 Blend percentage. 0.0 makes pixels fully gray.
8261 Higher values result in more preserved color.
8264 @subsection Commands
8265 This filter supports same @ref{commands} as options.
8266 The command accepts the same syntax of the corresponding option.
8268 If the specified expression is not valid, it is kept at its current
8271 @section colorlevels
8273 Adjust video input frames using levels.
8275 The filter accepts the following options:
8282 Adjust red, green, blue and alpha input black point.
8283 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8289 Adjust red, green, blue and alpha input white point.
8290 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
8292 Input levels are used to lighten highlights (bright tones), darken shadows
8293 (dark tones), change the balance of bright and dark tones.
8299 Adjust red, green, blue and alpha output black point.
8300 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
8306 Adjust red, green, blue and alpha output white point.
8307 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
8309 Output levels allows manual selection of a constrained output level range.
8312 @subsection Examples
8316 Make video output darker:
8318 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
8324 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
8328 Make video output lighter:
8330 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
8334 Increase brightness:
8336 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
8340 @subsection Commands
8342 This filter supports the all above options as @ref{commands}.
8344 @section colormatrix
8346 Convert color matrix.
8348 The filter accepts the following options:
8353 Specify the source and destination color matrix. Both values must be
8356 The accepted values are:
8384 For example to convert from BT.601 to SMPTE-240M, use the command:
8386 colormatrix=bt601:smpte240m
8391 Convert colorspace, transfer characteristics or color primaries.
8392 Input video needs to have an even size.
8394 The filter accepts the following options:
8399 Specify all color properties at once.
8401 The accepted values are:
8431 Specify output colorspace.
8433 The accepted values are:
8442 BT.470BG or BT.601-6 625
8445 SMPTE-170M or BT.601-6 525
8454 BT.2020 with non-constant luminance
8460 Specify output transfer characteristics.
8462 The accepted values are:
8474 Constant gamma of 2.2
8477 Constant gamma of 2.8
8480 SMPTE-170M, BT.601-6 625 or BT.601-6 525
8498 BT.2020 for 10-bits content
8501 BT.2020 for 12-bits content
8507 Specify output color primaries.
8509 The accepted values are:
8518 BT.470BG or BT.601-6 625
8521 SMPTE-170M or BT.601-6 525
8545 Specify output color range.
8547 The accepted values are:
8550 TV (restricted) range
8553 MPEG (restricted) range
8564 Specify output color format.
8566 The accepted values are:
8569 YUV 4:2:0 planar 8-bits
8572 YUV 4:2:0 planar 10-bits
8575 YUV 4:2:0 planar 12-bits
8578 YUV 4:2:2 planar 8-bits
8581 YUV 4:2:2 planar 10-bits
8584 YUV 4:2:2 planar 12-bits
8587 YUV 4:4:4 planar 8-bits
8590 YUV 4:4:4 planar 10-bits
8593 YUV 4:4:4 planar 12-bits
8598 Do a fast conversion, which skips gamma/primary correction. This will take
8599 significantly less CPU, but will be mathematically incorrect. To get output
8600 compatible with that produced by the colormatrix filter, use fast=1.
8603 Specify dithering mode.
8605 The accepted values are:
8611 Floyd-Steinberg dithering
8615 Whitepoint adaptation mode.
8617 The accepted values are:
8620 Bradford whitepoint adaptation
8623 von Kries whitepoint adaptation
8626 identity whitepoint adaptation (i.e. no whitepoint adaptation)
8630 Override all input properties at once. Same accepted values as @ref{all}.
8633 Override input colorspace. Same accepted values as @ref{space}.
8636 Override input color primaries. Same accepted values as @ref{primaries}.
8639 Override input transfer characteristics. Same accepted values as @ref{trc}.
8642 Override input color range. Same accepted values as @ref{range}.
8646 The filter converts the transfer characteristics, color space and color
8647 primaries to the specified user values. The output value, if not specified,
8648 is set to a default value based on the "all" property. If that property is
8649 also not specified, the filter will log an error. The output color range and
8650 format default to the same value as the input color range and format. The
8651 input transfer characteristics, color space, color primaries and color range
8652 should be set on the input data. If any of these are missing, the filter will
8653 log an error and no conversion will take place.
8655 For example to convert the input to SMPTE-240M, use the command:
8657 colorspace=smpte240m
8660 @section colortemperature
8661 Adjust color temperature in video to simulate variations in ambient color temperature.
8663 The filter accepts the following options:
8667 Set the temperature in Kelvin. Allowed range is from 1000 to 40000.
8668 Default value is 6500 K.
8671 Set mixing with filtered output. Allowed range is from 0 to 1.
8675 Set the amount of preserving lightness. Allowed range is from 0 to 1.
8679 @subsection Commands
8680 This filter supports same @ref{commands} as options.
8682 @section convolution
8684 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
8686 The filter accepts the following options:
8693 Set matrix for each plane.
8694 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
8695 and from 1 to 49 odd number of signed integers in @var{row} mode.
8701 Set multiplier for calculated value for each plane.
8702 If unset or 0, it will be sum of all matrix elements.
8708 Set bias for each plane. This value is added to the result of the multiplication.
8709 Useful for making the overall image brighter or darker. Default is 0.0.
8715 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
8716 Default is @var{square}.
8719 @subsection Commands
8721 This filter supports the all above options as @ref{commands}.
8723 @subsection Examples
8729 convolution="0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0"
8735 convolution="1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9"
8741 convolution="0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128"
8747 convolution="0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128"
8751 Apply laplacian edge detector which includes diagonals:
8753 convolution="1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0"
8759 convolution="-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2"
8765 Apply 2D convolution of video stream in frequency domain using second stream
8768 The filter accepts the following options:
8772 Set which planes to process.
8775 Set which impulse video frames will be processed, can be @var{first}
8776 or @var{all}. Default is @var{all}.
8779 The @code{convolve} filter also supports the @ref{framesync} options.
8783 Copy the input video source unchanged to the output. This is mainly useful for
8788 Video filtering on GPU using Apple's CoreImage API on OSX.
8790 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8791 processed by video hardware. However, software-based OpenGL implementations
8792 exist which means there is no guarantee for hardware processing. It depends on
8795 There are many filters and image generators provided by Apple that come with a
8796 large variety of options. The filter has to be referenced by its name along
8799 The coreimage filter accepts the following options:
8802 List all available filters and generators along with all their respective
8803 options as well as possible minimum and maximum values along with the default
8810 Specify all filters by their respective name and options.
8811 Use @var{list_filters} to determine all valid filter names and options.
8812 Numerical options are specified by a float value and are automatically clamped
8813 to their respective value range. Vector and color options have to be specified
8814 by a list of space separated float values. Character escaping has to be done.
8815 A special option name @code{default} is available to use default options for a
8818 It is required to specify either @code{default} or at least one of the filter options.
8819 All omitted options are used with their default values.
8820 The syntax of the filter string is as follows:
8822 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8826 Specify a rectangle where the output of the filter chain is copied into the
8827 input image. It is given by a list of space separated float values:
8829 output_rect=x\ y\ width\ height
8831 If not given, the output rectangle equals the dimensions of the input image.
8832 The output rectangle is automatically cropped at the borders of the input
8833 image. Negative values are valid for each component.
8835 output_rect=25\ 25\ 100\ 100
8839 Several filters can be chained for successive processing without GPU-HOST
8840 transfers allowing for fast processing of complex filter chains.
8841 Currently, only filters with zero (generators) or exactly one (filters) input
8842 image and one output image are supported. Also, transition filters are not yet
8845 Some filters generate output images with additional padding depending on the
8846 respective filter kernel. The padding is automatically removed to ensure the
8847 filter output has the same size as the input image.
8849 For image generators, the size of the output image is determined by the
8850 previous output image of the filter chain or the input image of the whole
8851 filterchain, respectively. The generators do not use the pixel information of
8852 this image to generate their output. However, the generated output is
8853 blended onto this image, resulting in partial or complete coverage of the
8856 The @ref{coreimagesrc} video source can be used for generating input images
8857 which are directly fed into the filter chain. By using it, providing input
8858 images by another video source or an input video is not required.
8860 @subsection Examples
8865 List all filters available:
8867 coreimage=list_filters=true
8871 Use the CIBoxBlur filter with default options to blur an image:
8873 coreimage=filter=CIBoxBlur@@default
8877 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8878 its center at 100x100 and a radius of 50 pixels:
8880 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8884 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8885 given as complete and escaped command-line for Apple's standard bash shell:
8887 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8893 Cover a rectangular object
8895 It accepts the following options:
8899 Filepath of the optional cover image, needs to be in yuv420.
8904 It accepts the following values:
8907 cover it by the supplied image
8909 cover it by interpolating the surrounding pixels
8912 Default value is @var{blur}.
8915 @subsection Examples
8919 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8921 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8927 Crop the input video to given dimensions.
8929 It accepts the following parameters:
8933 The width of the output video. It defaults to @code{iw}.
8934 This expression is evaluated only once during the filter
8935 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8938 The height of the output video. It defaults to @code{ih}.
8939 This expression is evaluated only once during the filter
8940 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8943 The horizontal position, in the input video, of the left edge of the output
8944 video. It defaults to @code{(in_w-out_w)/2}.
8945 This expression is evaluated per-frame.
8948 The vertical position, in the input video, of the top edge of the output video.
8949 It defaults to @code{(in_h-out_h)/2}.
8950 This expression is evaluated per-frame.
8953 If set to 1 will force the output display aspect ratio
8954 to be the same of the input, by changing the output sample aspect
8955 ratio. It defaults to 0.
8958 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
8959 width/height/x/y as specified and will not be rounded to nearest smaller value.
8963 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
8964 expressions containing the following constants:
8969 The computed values for @var{x} and @var{y}. They are evaluated for
8974 The input width and height.
8978 These are the same as @var{in_w} and @var{in_h}.
8982 The output (cropped) width and height.
8986 These are the same as @var{out_w} and @var{out_h}.
8989 same as @var{iw} / @var{ih}
8992 input sample aspect ratio
8995 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
8999 horizontal and vertical chroma subsample values. For example for the
9000 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9003 The number of the input frame, starting from 0.
9006 the position in the file of the input frame, NAN if unknown
9009 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
9013 The expression for @var{out_w} may depend on the value of @var{out_h},
9014 and the expression for @var{out_h} may depend on @var{out_w}, but they
9015 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
9016 evaluated after @var{out_w} and @var{out_h}.
9018 The @var{x} and @var{y} parameters specify the expressions for the
9019 position of the top-left corner of the output (non-cropped) area. They
9020 are evaluated for each frame. If the evaluated value is not valid, it
9021 is approximated to the nearest valid value.
9023 The expression for @var{x} may depend on @var{y}, and the expression
9024 for @var{y} may depend on @var{x}.
9026 @subsection Examples
9030 Crop area with size 100x100 at position (12,34).
9035 Using named options, the example above becomes:
9037 crop=w=100:h=100:x=12:y=34
9041 Crop the central input area with size 100x100:
9047 Crop the central input area with size 2/3 of the input video:
9049 crop=2/3*in_w:2/3*in_h
9053 Crop the input video central square:
9060 Delimit the rectangle with the top-left corner placed at position
9061 100:100 and the right-bottom corner corresponding to the right-bottom
9062 corner of the input image.
9064 crop=in_w-100:in_h-100:100:100
9068 Crop 10 pixels from the left and right borders, and 20 pixels from
9069 the top and bottom borders
9071 crop=in_w-2*10:in_h-2*20
9075 Keep only the bottom right quarter of the input image:
9077 crop=in_w/2:in_h/2:in_w/2:in_h/2
9081 Crop height for getting Greek harmony:
9083 crop=in_w:1/PHI*in_w
9087 Apply trembling effect:
9089 crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(n/10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(n/7)
9093 Apply erratic camera effect depending on timestamp:
9095 crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(t*10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(t*13)"
9099 Set x depending on the value of y:
9101 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
9105 @subsection Commands
9107 This filter supports the following commands:
9113 Set width/height of the output video and the horizontal/vertical position
9115 The command accepts the same syntax of the corresponding option.
9117 If the specified expression is not valid, it is kept at its current
9123 Auto-detect the crop size.
9125 It calculates the necessary cropping parameters and prints the
9126 recommended parameters via the logging system. The detected dimensions
9127 correspond to the non-black area of the input video.
9129 It accepts the following parameters:
9134 Set higher black value threshold, which can be optionally specified
9135 from nothing (0) to everything (255 for 8-bit based formats). An intensity
9136 value greater to the set value is considered non-black. It defaults to 24.
9137 You can also specify a value between 0.0 and 1.0 which will be scaled depending
9138 on the bitdepth of the pixel format.
9141 The value which the width/height should be divisible by. It defaults to
9142 16. The offset is automatically adjusted to center the video. Use 2 to
9143 get only even dimensions (needed for 4:2:2 video). 16 is best when
9144 encoding to most video codecs.
9147 Set the number of initial frames for which evaluation is skipped.
9148 Default is 2. Range is 0 to INT_MAX.
9150 @item reset_count, reset
9151 Set the counter that determines after how many frames cropdetect will
9152 reset the previously detected largest video area and start over to
9153 detect the current optimal crop area. Default value is 0.
9155 This can be useful when channel logos distort the video area. 0
9156 indicates 'never reset', and returns the largest area encountered during
9163 Delay video filtering until a given wallclock timestamp. The filter first
9164 passes on @option{preroll} amount of frames, then it buffers at most
9165 @option{buffer} amount of frames and waits for the cue. After reaching the cue
9166 it forwards the buffered frames and also any subsequent frames coming in its
9169 The filter can be used synchronize the output of multiple ffmpeg processes for
9170 realtime output devices like decklink. By putting the delay in the filtering
9171 chain and pre-buffering frames the process can pass on data to output almost
9172 immediately after the target wallclock timestamp is reached.
9174 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
9180 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
9183 The duration of content to pass on as preroll expressed in seconds. Default is 0.
9186 The maximum duration of content to buffer before waiting for the cue expressed
9187 in seconds. Default is 0.
9194 Apply color adjustments using curves.
9196 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
9197 component (red, green and blue) has its values defined by @var{N} key points
9198 tied from each other using a smooth curve. The x-axis represents the pixel
9199 values from the input frame, and the y-axis the new pixel values to be set for
9202 By default, a component curve is defined by the two points @var{(0;0)} and
9203 @var{(1;1)}. This creates a straight line where each original pixel value is
9204 "adjusted" to its own value, which means no change to the image.
9206 The filter allows you to redefine these two points and add some more. A new
9207 curve (using a natural cubic spline interpolation) will be define to pass
9208 smoothly through all these new coordinates. The new defined points needs to be
9209 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
9210 be in the @var{[0;1]} interval. If the computed curves happened to go outside
9211 the vector spaces, the values will be clipped accordingly.
9213 The filter accepts the following options:
9217 Select one of the available color presets. This option can be used in addition
9218 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
9219 options takes priority on the preset values.
9220 Available presets are:
9223 @item color_negative
9226 @item increase_contrast
9228 @item linear_contrast
9229 @item medium_contrast
9231 @item strong_contrast
9234 Default is @code{none}.
9236 Set the master key points. These points will define a second pass mapping. It
9237 is sometimes called a "luminance" or "value" mapping. It can be used with
9238 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
9239 post-processing LUT.
9241 Set the key points for the red component.
9243 Set the key points for the green component.
9245 Set the key points for the blue component.
9247 Set the key points for all components (not including master).
9248 Can be used in addition to the other key points component
9249 options. In this case, the unset component(s) will fallback on this
9250 @option{all} setting.
9252 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
9254 Save Gnuplot script of the curves in specified file.
9257 To avoid some filtergraph syntax conflicts, each key points list need to be
9258 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
9260 @subsection Commands
9262 This filter supports same @ref{commands} as options.
9264 @subsection Examples
9268 Increase slightly the middle level of blue:
9270 curves=blue='0/0 0.5/0.58 1/1'
9276 curves=r='0/0.11 .42/.51 1/0.95':g='0/0 0.50/0.48 1/1':b='0/0.22 .49/.44 1/0.8'
9278 Here we obtain the following coordinates for each components:
9281 @code{(0;0.11) (0.42;0.51) (1;0.95)}
9283 @code{(0;0) (0.50;0.48) (1;1)}
9285 @code{(0;0.22) (0.49;0.44) (1;0.80)}
9289 The previous example can also be achieved with the associated built-in preset:
9291 curves=preset=vintage
9301 Use a Photoshop preset and redefine the points of the green component:
9303 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
9307 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
9308 and @command{gnuplot}:
9310 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
9311 gnuplot -p /tmp/curves.plt
9317 Video data analysis filter.
9319 This filter shows hexadecimal pixel values of part of video.
9321 The filter accepts the following options:
9325 Set output video size.
9328 Set x offset from where to pick pixels.
9331 Set y offset from where to pick pixels.
9334 Set scope mode, can be one of the following:
9337 Draw hexadecimal pixel values with white color on black background.
9340 Draw hexadecimal pixel values with input video pixel color on black
9344 Draw hexadecimal pixel values on color background picked from input video,
9345 the text color is picked in such way so its always visible.
9349 Draw rows and columns numbers on left and top of video.
9352 Set background opacity.
9355 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
9358 Set pixel components to display. By default all pixel components are displayed.
9362 Apply Directional blur filter.
9364 The filter accepts the following options:
9368 Set angle of directional blur. Default is @code{45}.
9371 Set radius of directional blur. Default is @code{5}.
9374 Set which planes to filter. By default all planes are filtered.
9377 @subsection Commands
9378 This filter supports same @ref{commands} as options.
9379 The command accepts the same syntax of the corresponding option.
9381 If the specified expression is not valid, it is kept at its current
9386 Denoise frames using 2D DCT (frequency domain filtering).
9388 This filter is not designed for real time.
9390 The filter accepts the following options:
9394 Set the noise sigma constant.
9396 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
9397 coefficient (absolute value) below this threshold with be dropped.
9399 If you need a more advanced filtering, see @option{expr}.
9401 Default is @code{0}.
9404 Set number overlapping pixels for each block. Since the filter can be slow, you
9405 may want to reduce this value, at the cost of a less effective filter and the
9406 risk of various artefacts.
9408 If the overlapping value doesn't permit processing the whole input width or
9409 height, a warning will be displayed and according borders won't be denoised.
9411 Default value is @var{blocksize}-1, which is the best possible setting.
9414 Set the coefficient factor expression.
9416 For each coefficient of a DCT block, this expression will be evaluated as a
9417 multiplier value for the coefficient.
9419 If this is option is set, the @option{sigma} option will be ignored.
9421 The absolute value of the coefficient can be accessed through the @var{c}
9425 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
9426 @var{blocksize}, which is the width and height of the processed blocks.
9428 The default value is @var{3} (8x8) and can be raised to @var{4} for a
9429 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
9430 on the speed processing. Also, a larger block size does not necessarily means a
9434 @subsection Examples
9436 Apply a denoise with a @option{sigma} of @code{4.5}:
9441 The same operation can be achieved using the expression system:
9443 dctdnoiz=e='gte(c, 4.5*3)'
9446 Violent denoise using a block size of @code{16x16}:
9453 Remove banding artifacts from input video.
9454 It works by replacing banded pixels with average value of referenced pixels.
9456 The filter accepts the following options:
9463 Set banding detection threshold for each plane. Default is 0.02.
9464 Valid range is 0.00003 to 0.5.
9465 If difference between current pixel and reference pixel is less than threshold,
9466 it will be considered as banded.
9469 Banding detection range in pixels. Default is 16. If positive, random number
9470 in range 0 to set value will be used. If negative, exact absolute value
9472 The range defines square of four pixels around current pixel.
9475 Set direction in radians from which four pixel will be compared. If positive,
9476 random direction from 0 to set direction will be picked. If negative, exact of
9477 absolute value will be picked. For example direction 0, -PI or -2*PI radians
9478 will pick only pixels on same row and -PI/2 will pick only pixels on same
9482 If enabled, current pixel is compared with average value of all four
9483 surrounding pixels. The default is enabled. If disabled current pixel is
9484 compared with all four surrounding pixels. The pixel is considered banded
9485 if only all four differences with surrounding pixels are less than threshold.
9488 If enabled, current pixel is changed if and only if all pixel components are banded,
9489 e.g. banding detection threshold is triggered for all color components.
9490 The default is disabled.
9493 @subsection Commands
9495 This filter supports the all above options as @ref{commands}.
9499 Remove blocking artifacts from input video.
9501 The filter accepts the following options:
9505 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
9506 This controls what kind of deblocking is applied.
9509 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
9515 Set blocking detection thresholds. Allowed range is 0 to 1.
9516 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
9517 Using higher threshold gives more deblocking strength.
9518 Setting @var{alpha} controls threshold detection at exact edge of block.
9519 Remaining options controls threshold detection near the edge. Each one for
9520 below/above or left/right. Setting any of those to @var{0} disables
9524 Set planes to filter. Default is to filter all available planes.
9527 @subsection Examples
9531 Deblock using weak filter and block size of 4 pixels.
9533 deblock=filter=weak:block=4
9537 Deblock using strong filter, block size of 4 pixels and custom thresholds for
9538 deblocking more edges.
9540 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
9544 Similar as above, but filter only first plane.
9546 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
9550 Similar as above, but filter only second and third plane.
9552 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
9556 @subsection Commands
9558 This filter supports the all above options as @ref{commands}.
9563 Drop duplicated frames at regular intervals.
9565 The filter accepts the following options:
9569 Set the number of frames from which one will be dropped. Setting this to
9570 @var{N} means one frame in every batch of @var{N} frames will be dropped.
9571 Default is @code{5}.
9574 Set the threshold for duplicate detection. If the difference metric for a frame
9575 is less than or equal to this value, then it is declared as duplicate. Default
9579 Set scene change threshold. Default is @code{15}.
9583 Set the size of the x and y-axis blocks used during metric calculations.
9584 Larger blocks give better noise suppression, but also give worse detection of
9585 small movements. Must be a power of two. Default is @code{32}.
9588 Mark main input as a pre-processed input and activate clean source input
9589 stream. This allows the input to be pre-processed with various filters to help
9590 the metrics calculation while keeping the frame selection lossless. When set to
9591 @code{1}, the first stream is for the pre-processed input, and the second
9592 stream is the clean source from where the kept frames are chosen. Default is
9596 Set whether or not chroma is considered in the metric calculations. Default is
9602 Apply 2D deconvolution of video stream in frequency domain using second stream
9605 The filter accepts the following options:
9609 Set which planes to process.
9612 Set which impulse video frames will be processed, can be @var{first}
9613 or @var{all}. Default is @var{all}.
9616 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
9617 and height are not same and not power of 2 or if stream prior to convolving
9621 The @code{deconvolve} filter also supports the @ref{framesync} options.
9625 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
9627 It accepts the following options:
9631 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
9632 @var{rainbows} for cross-color reduction.
9635 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
9638 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
9641 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
9644 Set temporal chroma threshold. Lower values increases reduction of cross-color.
9649 Apply deflate effect to the video.
9651 This filter replaces the pixel by the local(3x3) average by taking into account
9652 only values lower than the pixel.
9654 It accepts the following options:
9661 Limit the maximum change for each plane, default is 65535.
9662 If 0, plane will remain unchanged.
9665 @subsection Commands
9667 This filter supports the all above options as @ref{commands}.
9671 Remove temporal frame luminance variations.
9673 It accepts the following options:
9677 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
9680 Set averaging mode to smooth temporal luminance variations.
9682 Available values are:
9707 Do not actually modify frame. Useful when one only wants metadata.
9712 Remove judder produced by partially interlaced telecined content.
9714 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
9715 source was partially telecined content then the output of @code{pullup,dejudder}
9716 will have a variable frame rate. May change the recorded frame rate of the
9717 container. Aside from that change, this filter will not affect constant frame
9720 The option available in this filter is:
9724 Specify the length of the window over which the judder repeats.
9726 Accepts any integer greater than 1. Useful values are:
9730 If the original was telecined from 24 to 30 fps (Film to NTSC).
9733 If the original was telecined from 25 to 30 fps (PAL to NTSC).
9736 If a mixture of the two.
9739 The default is @samp{4}.
9744 Suppress a TV station logo by a simple interpolation of the surrounding
9745 pixels. Just set a rectangle covering the logo and watch it disappear
9746 (and sometimes something even uglier appear - your mileage may vary).
9748 It accepts the following parameters:
9753 Specify the top left corner coordinates of the logo. They must be
9758 Specify the width and height of the logo to clear. They must be
9762 Specify the thickness of the fuzzy edge of the rectangle (added to
9763 @var{w} and @var{h}). The default value is 1. This option is
9764 deprecated, setting higher values should no longer be necessary and
9768 When set to 1, a green rectangle is drawn on the screen to simplify
9769 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9770 The default value is 0.
9772 The rectangle is drawn on the outermost pixels which will be (partly)
9773 replaced with interpolated values. The values of the next pixels
9774 immediately outside this rectangle in each direction will be used to
9775 compute the interpolated pixel values inside the rectangle.
9779 @subsection Examples
9783 Set a rectangle covering the area with top left corner coordinates 0,0
9784 and size 100x77, and a band of size 10:
9786 delogo=x=0:y=0:w=100:h=77:band=10
9794 Remove the rain in the input image/video by applying the derain methods based on
9795 convolutional neural networks. Supported models:
9799 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9800 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9803 Training as well as model generation scripts are provided in
9804 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9806 Native model files (.model) can be generated from TensorFlow model
9807 files (.pb) by using tools/python/convert.py
9809 The filter accepts the following options:
9813 Specify which filter to use. This option accepts the following values:
9817 Derain filter. To conduct derain filter, you need to use a derain model.
9820 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9822 Default value is @samp{derain}.
9825 Specify which DNN backend to use for model loading and execution. This option accepts
9826 the following values:
9830 Native implementation of DNN loading and execution.
9833 TensorFlow backend. To enable this backend you
9834 need to install the TensorFlow for C library (see
9835 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9836 @code{--enable-libtensorflow}
9838 Default value is @samp{native}.
9841 Set path to model file specifying network architecture and its parameters.
9842 Note that different backends use different file formats. TensorFlow and native
9843 backend can load files for only its format.
9846 It can also be finished with @ref{dnn_processing} filter.
9850 Attempt to fix small changes in horizontal and/or vertical shift. This
9851 filter helps remove camera shake from hand-holding a camera, bumping a
9852 tripod, moving on a vehicle, etc.
9854 The filter accepts the following options:
9862 Specify a rectangular area where to limit the search for motion
9864 If desired the search for motion vectors can be limited to a
9865 rectangular area of the frame defined by its top left corner, width
9866 and height. These parameters have the same meaning as the drawbox
9867 filter which can be used to visualise the position of the bounding
9870 This is useful when simultaneous movement of subjects within the frame
9871 might be confused for camera motion by the motion vector search.
9873 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9874 then the full frame is used. This allows later options to be set
9875 without specifying the bounding box for the motion vector search.
9877 Default - search the whole frame.
9881 Specify the maximum extent of movement in x and y directions in the
9882 range 0-64 pixels. Default 16.
9885 Specify how to generate pixels to fill blanks at the edge of the
9886 frame. Available values are:
9889 Fill zeroes at blank locations
9891 Original image at blank locations
9893 Extruded edge value at blank locations
9895 Mirrored edge at blank locations
9897 Default value is @samp{mirror}.
9900 Specify the blocksize to use for motion search. Range 4-128 pixels,
9904 Specify the contrast threshold for blocks. Only blocks with more than
9905 the specified contrast (difference between darkest and lightest
9906 pixels) will be considered. Range 1-255, default 125.
9909 Specify the search strategy. Available values are:
9912 Set exhaustive search
9914 Set less exhaustive search.
9916 Default value is @samp{exhaustive}.
9919 If set then a detailed log of the motion search is written to the
9926 Remove unwanted contamination of foreground colors, caused by reflected color of
9927 greenscreen or bluescreen.
9929 This filter accepts the following options:
9933 Set what type of despill to use.
9936 Set how spillmap will be generated.
9939 Set how much to get rid of still remaining spill.
9942 Controls amount of red in spill area.
9945 Controls amount of green in spill area.
9946 Should be -1 for greenscreen.
9949 Controls amount of blue in spill area.
9950 Should be -1 for bluescreen.
9953 Controls brightness of spill area, preserving colors.
9956 Modify alpha from generated spillmap.
9959 @subsection Commands
9961 This filter supports the all above options as @ref{commands}.
9965 Apply an exact inverse of the telecine operation. It requires a predefined
9966 pattern specified using the pattern option which must be the same as that passed
9967 to the telecine filter.
9969 This filter accepts the following options:
9978 The default value is @code{top}.
9982 A string of numbers representing the pulldown pattern you wish to apply.
9983 The default value is @code{23}.
9986 A number representing position of the first frame with respect to the telecine
9987 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9992 Apply dilation effect to the video.
9994 This filter replaces the pixel by the local(3x3) maximum.
9996 It accepts the following options:
10003 Limit the maximum change for each plane, default is 65535.
10004 If 0, plane will remain unchanged.
10007 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
10010 Flags to local 3x3 coordinates maps like this:
10017 @subsection Commands
10019 This filter supports the all above options as @ref{commands}.
10023 Displace pixels as indicated by second and third input stream.
10025 It takes three input streams and outputs one stream, the first input is the
10026 source, and second and third input are displacement maps.
10028 The second input specifies how much to displace pixels along the
10029 x-axis, while the third input specifies how much to displace pixels
10031 If one of displacement map streams terminates, last frame from that
10032 displacement map will be used.
10034 Note that once generated, displacements maps can be reused over and over again.
10036 A description of the accepted options follows.
10040 Set displace behavior for pixels that are out of range.
10042 Available values are:
10045 Missing pixels are replaced by black pixels.
10048 Adjacent pixels will spread out to replace missing pixels.
10051 Out of range pixels are wrapped so they point to pixels of other side.
10054 Out of range pixels will be replaced with mirrored pixels.
10056 Default is @samp{smear}.
10060 @subsection Examples
10064 Add ripple effect to rgb input of video size hd720:
10066 ffmpeg -i INPUT -f lavfi -i nullsrc=s=hd720,lutrgb=128:128:128 -f lavfi -i nullsrc=s=hd720,geq='r=128+30*sin(2*PI*X/400+T):g=128+30*sin(2*PI*X/400+T):b=128+30*sin(2*PI*X/400+T)' -lavfi '[0][1][2]displace' OUTPUT
10070 Add wave effect to rgb input of video size hd720:
10072 ffmpeg -i INPUT -f lavfi -i nullsrc=hd720,geq='r=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):g=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):b=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T))' -lavfi '[1]split[x][y],[0][x][y]displace' OUTPUT
10076 @anchor{dnn_processing}
10077 @section dnn_processing
10079 Do image processing with deep neural networks. It works together with another filter
10080 which converts the pixel format of the Frame to what the dnn network requires.
10082 The filter accepts the following options:
10086 Specify which DNN backend to use for model loading and execution. This option accepts
10087 the following values:
10091 Native implementation of DNN loading and execution.
10094 TensorFlow backend. To enable this backend you
10095 need to install the TensorFlow for C library (see
10096 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
10097 @code{--enable-libtensorflow}
10100 OpenVINO backend. To enable this backend you
10101 need to build and install the OpenVINO for C library (see
10102 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
10103 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
10104 be needed if the header files and libraries are not installed into system path)
10108 Default value is @samp{native}.
10111 Set path to model file specifying network architecture and its parameters.
10112 Note that different backends use different file formats. TensorFlow, OpenVINO and native
10113 backend can load files for only its format.
10115 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
10118 Set the input name of the dnn network.
10121 Set the output name of the dnn network.
10124 use DNN async execution if set (default: set),
10125 roll back to sync execution if the backend does not support async.
10129 @subsection Examples
10133 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
10135 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
10139 Halve the pixel value of the frame with format gray32f:
10141 ffmpeg -i input.jpg -vf format=grayf32,dnn_processing=model=halve_gray_float.model:input=dnn_in:output=dnn_out:dnn_backend=native -y out.native.png
10145 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
10147 ./ffmpeg -i 480p.jpg -vf format=yuv420p,scale=w=iw*2:h=ih*2,dnn_processing=dnn_backend=tensorflow:model=srcnn.pb:input=x:output=y -y srcnn.jpg
10151 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
10153 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
10160 Draw a colored box on the input image.
10162 It accepts the following parameters:
10167 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
10171 The expressions which specify the width and height of the box; if 0 they are interpreted as
10172 the input width and height. It defaults to 0.
10175 Specify the color of the box to write. For the general syntax of this option,
10176 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10177 value @code{invert} is used, the box edge color is the same as the
10178 video with inverted luma.
10181 The expression which sets the thickness of the box edge.
10182 A value of @code{fill} will create a filled box. Default value is @code{3}.
10184 See below for the list of accepted constants.
10187 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
10188 will overwrite the video's color and alpha pixels.
10189 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
10192 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10193 following constants:
10197 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10201 horizontal and vertical chroma subsample values. For example for the
10202 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10206 The input width and height.
10209 The input sample aspect ratio.
10213 The x and y offset coordinates where the box is drawn.
10217 The width and height of the drawn box.
10220 The thickness of the drawn box.
10222 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10223 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10227 @subsection Examples
10231 Draw a black box around the edge of the input image:
10237 Draw a box with color red and an opacity of 50%:
10239 drawbox=10:20:200:60:red@@0.5
10242 The previous example can be specified as:
10244 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
10248 Fill the box with pink color:
10250 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
10254 Draw a 2-pixel red 2.40:1 mask:
10256 drawbox=x=-t:y=0.5*(ih-iw/2.4)-t:w=iw+t*2:h=iw/2.4+t*2:t=2:c=red
10260 @subsection Commands
10261 This filter supports same commands as options.
10262 The command accepts the same syntax of the corresponding option.
10264 If the specified expression is not valid, it is kept at its current
10269 Draw a graph using input video metadata.
10271 It accepts the following parameters:
10275 Set 1st frame metadata key from which metadata values will be used to draw a graph.
10278 Set 1st foreground color expression.
10281 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
10284 Set 2nd foreground color expression.
10287 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
10290 Set 3rd foreground color expression.
10293 Set 4th frame metadata key from which metadata values will be used to draw a graph.
10296 Set 4th foreground color expression.
10299 Set minimal value of metadata value.
10302 Set maximal value of metadata value.
10305 Set graph background color. Default is white.
10310 Available values for mode is:
10317 Default is @code{line}.
10322 Available values for slide is:
10325 Draw new frame when right border is reached.
10328 Replace old columns with new ones.
10331 Scroll from right to left.
10334 Scroll from left to right.
10337 Draw single picture.
10340 Default is @code{frame}.
10343 Set size of graph video. For the syntax of this option, check the
10344 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
10345 The default value is @code{900x256}.
10348 Set the output frame rate. Default value is @code{25}.
10350 The foreground color expressions can use the following variables:
10353 Minimal value of metadata value.
10356 Maximal value of metadata value.
10359 Current metadata key value.
10362 The color is defined as 0xAABBGGRR.
10365 Example using metadata from @ref{signalstats} filter:
10367 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
10370 Example using metadata from @ref{ebur128} filter:
10372 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
10377 Draw a grid on the input image.
10379 It accepts the following parameters:
10384 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
10388 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
10389 input width and height, respectively, minus @code{thickness}, so image gets
10390 framed. Default to 0.
10393 Specify the color of the grid. For the general syntax of this option,
10394 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10395 value @code{invert} is used, the grid color is the same as the
10396 video with inverted luma.
10399 The expression which sets the thickness of the grid line. Default value is @code{1}.
10401 See below for the list of accepted constants.
10404 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
10405 will overwrite the video's color and alpha pixels.
10406 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
10409 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10410 following constants:
10414 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10418 horizontal and vertical chroma subsample values. For example for the
10419 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10423 The input grid cell width and height.
10426 The input sample aspect ratio.
10430 The x and y coordinates of some point of grid intersection (meant to configure offset).
10434 The width and height of the drawn cell.
10437 The thickness of the drawn cell.
10439 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10440 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10444 @subsection Examples
10448 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
10450 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
10454 Draw a white 3x3 grid with an opacity of 50%:
10456 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
10460 @subsection Commands
10461 This filter supports same commands as options.
10462 The command accepts the same syntax of the corresponding option.
10464 If the specified expression is not valid, it is kept at its current
10470 Draw a text string or text from a specified file on top of a video, using the
10471 libfreetype library.
10473 To enable compilation of this filter, you need to configure FFmpeg with
10474 @code{--enable-libfreetype}.
10475 To enable default font fallback and the @var{font} option you need to
10476 configure FFmpeg with @code{--enable-libfontconfig}.
10477 To enable the @var{text_shaping} option, you need to configure FFmpeg with
10478 @code{--enable-libfribidi}.
10482 It accepts the following parameters:
10487 Used to draw a box around text using the background color.
10488 The value must be either 1 (enable) or 0 (disable).
10489 The default value of @var{box} is 0.
10492 Set the width of the border to be drawn around the box using @var{boxcolor}.
10493 The default value of @var{boxborderw} is 0.
10496 The color to be used for drawing box around text. For the syntax of this
10497 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10499 The default value of @var{boxcolor} is "white".
10502 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
10503 The default value of @var{line_spacing} is 0.
10506 Set the width of the border to be drawn around the text using @var{bordercolor}.
10507 The default value of @var{borderw} is 0.
10510 Set the color to be used for drawing border around text. For the syntax of this
10511 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10513 The default value of @var{bordercolor} is "black".
10516 Select how the @var{text} is expanded. Can be either @code{none},
10517 @code{strftime} (deprecated) or
10518 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
10522 Set a start time for the count. Value is in microseconds. Only applied
10523 in the deprecated strftime expansion mode. To emulate in normal expansion
10524 mode use the @code{pts} function, supplying the start time (in seconds)
10525 as the second argument.
10528 If true, check and fix text coords to avoid clipping.
10531 The color to be used for drawing fonts. For the syntax of this option, check
10532 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10534 The default value of @var{fontcolor} is "black".
10536 @item fontcolor_expr
10537 String which is expanded the same way as @var{text} to obtain dynamic
10538 @var{fontcolor} value. By default this option has empty value and is not
10539 processed. When this option is set, it overrides @var{fontcolor} option.
10542 The font family to be used for drawing text. By default Sans.
10545 The font file to be used for drawing text. The path must be included.
10546 This parameter is mandatory if the fontconfig support is disabled.
10549 Draw the text applying alpha blending. The value can
10550 be a number between 0.0 and 1.0.
10551 The expression accepts the same variables @var{x, y} as well.
10552 The default value is 1.
10553 Please see @var{fontcolor_expr}.
10556 The font size to be used for drawing text.
10557 The default value of @var{fontsize} is 16.
10560 If set to 1, attempt to shape the text (for example, reverse the order of
10561 right-to-left text and join Arabic characters) before drawing it.
10562 Otherwise, just draw the text exactly as given.
10563 By default 1 (if supported).
10565 @item ft_load_flags
10566 The flags to be used for loading the fonts.
10568 The flags map the corresponding flags supported by libfreetype, and are
10569 a combination of the following values:
10576 @item vertical_layout
10577 @item force_autohint
10580 @item ignore_global_advance_width
10582 @item ignore_transform
10584 @item linear_design
10588 Default value is "default".
10590 For more information consult the documentation for the FT_LOAD_*
10594 The color to be used for drawing a shadow behind the drawn text. For the
10595 syntax of this option, check the @ref{color syntax,,"Color" section in the
10596 ffmpeg-utils manual,ffmpeg-utils}.
10598 The default value of @var{shadowcolor} is "black".
10602 The x and y offsets for the text shadow position with respect to the
10603 position of the text. They can be either positive or negative
10604 values. The default value for both is "0".
10607 The starting frame number for the n/frame_num variable. The default value
10611 The size in number of spaces to use for rendering the tab.
10612 Default value is 4.
10615 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
10616 format. It can be used with or without text parameter. @var{timecode_rate}
10617 option must be specified.
10619 @item timecode_rate, rate, r
10620 Set the timecode frame rate (timecode only). Value will be rounded to nearest
10621 integer. Minimum value is "1".
10622 Drop-frame timecode is supported for frame rates 30 & 60.
10625 If set to 1, the output of the timecode option will wrap around at 24 hours.
10626 Default is 0 (disabled).
10629 The text string to be drawn. The text must be a sequence of UTF-8
10630 encoded characters.
10631 This parameter is mandatory if no file is specified with the parameter
10635 A text file containing text to be drawn. The text must be a sequence
10636 of UTF-8 encoded characters.
10638 This parameter is mandatory if no text string is specified with the
10639 parameter @var{text}.
10641 If both @var{text} and @var{textfile} are specified, an error is thrown.
10644 If set to 1, the @var{textfile} will be reloaded before each frame.
10645 Be sure to update it atomically, or it may be read partially, or even fail.
10649 The expressions which specify the offsets where text will be drawn
10650 within the video frame. They are relative to the top/left border of the
10653 The default value of @var{x} and @var{y} is "0".
10655 See below for the list of accepted constants and functions.
10658 The parameters for @var{x} and @var{y} are expressions containing the
10659 following constants and functions:
10663 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
10667 horizontal and vertical chroma subsample values. For example for the
10668 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10671 the height of each text line
10679 @item max_glyph_a, ascent
10680 the maximum distance from the baseline to the highest/upper grid
10681 coordinate used to place a glyph outline point, for all the rendered
10683 It is a positive value, due to the grid's orientation with the Y axis
10686 @item max_glyph_d, descent
10687 the maximum distance from the baseline to the lowest grid coordinate
10688 used to place a glyph outline point, for all the rendered glyphs.
10689 This is a negative value, due to the grid's orientation, with the Y axis
10693 maximum glyph height, that is the maximum height for all the glyphs
10694 contained in the rendered text, it is equivalent to @var{ascent} -
10698 maximum glyph width, that is the maximum width for all the glyphs
10699 contained in the rendered text
10702 the number of input frame, starting from 0
10704 @item rand(min, max)
10705 return a random number included between @var{min} and @var{max}
10708 The input sample aspect ratio.
10711 timestamp expressed in seconds, NAN if the input timestamp is unknown
10714 the height of the rendered text
10717 the width of the rendered text
10721 the x and y offset coordinates where the text is drawn.
10723 These parameters allow the @var{x} and @var{y} expressions to refer
10724 to each other, so you can for example specify @code{y=x/dar}.
10727 A one character description of the current frame's picture type.
10730 The current packet's position in the input file or stream
10731 (in bytes, from the start of the input). A value of -1 indicates
10732 this info is not available.
10735 The current packet's duration, in seconds.
10738 The current packet's size (in bytes).
10741 @anchor{drawtext_expansion}
10742 @subsection Text expansion
10744 If @option{expansion} is set to @code{strftime},
10745 the filter recognizes strftime() sequences in the provided text and
10746 expands them accordingly. Check the documentation of strftime(). This
10747 feature is deprecated.
10749 If @option{expansion} is set to @code{none}, the text is printed verbatim.
10751 If @option{expansion} is set to @code{normal} (which is the default),
10752 the following expansion mechanism is used.
10754 The backslash character @samp{\}, followed by any character, always expands to
10755 the second character.
10757 Sequences of the form @code{%@{...@}} are expanded. The text between the
10758 braces is a function name, possibly followed by arguments separated by ':'.
10759 If the arguments contain special characters or delimiters (':' or '@}'),
10760 they should be escaped.
10762 Note that they probably must also be escaped as the value for the
10763 @option{text} option in the filter argument string and as the filter
10764 argument in the filtergraph description, and possibly also for the shell,
10765 that makes up to four levels of escaping; using a text file avoids these
10768 The following functions are available:
10773 The expression evaluation result.
10775 It must take one argument specifying the expression to be evaluated,
10776 which accepts the same constants and functions as the @var{x} and
10777 @var{y} values. Note that not all constants should be used, for
10778 example the text size is not known when evaluating the expression, so
10779 the constants @var{text_w} and @var{text_h} will have an undefined
10782 @item expr_int_format, eif
10783 Evaluate the expression's value and output as formatted integer.
10785 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10786 The second argument specifies the output format. Allowed values are @samp{x},
10787 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10788 @code{printf} function.
10789 The third parameter is optional and sets the number of positions taken by the output.
10790 It can be used to add padding with zeros from the left.
10793 The time at which the filter is running, expressed in UTC.
10794 It can accept an argument: a strftime() format string.
10797 The time at which the filter is running, expressed in the local time zone.
10798 It can accept an argument: a strftime() format string.
10801 Frame metadata. Takes one or two arguments.
10803 The first argument is mandatory and specifies the metadata key.
10805 The second argument is optional and specifies a default value, used when the
10806 metadata key is not found or empty.
10808 Available metadata can be identified by inspecting entries
10809 starting with TAG included within each frame section
10810 printed by running @code{ffprobe -show_frames}.
10812 String metadata generated in filters leading to
10813 the drawtext filter are also available.
10816 The frame number, starting from 0.
10819 A one character description of the current picture type.
10822 The timestamp of the current frame.
10823 It can take up to three arguments.
10825 The first argument is the format of the timestamp; it defaults to @code{flt}
10826 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10827 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10828 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10829 @code{localtime} stands for the timestamp of the frame formatted as
10830 local time zone time.
10832 The second argument is an offset added to the timestamp.
10834 If the format is set to @code{hms}, a third argument @code{24HH} may be
10835 supplied to present the hour part of the formatted timestamp in 24h format
10838 If the format is set to @code{localtime} or @code{gmtime},
10839 a third argument may be supplied: a strftime() format string.
10840 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10843 @subsection Commands
10845 This filter supports altering parameters via commands:
10848 Alter existing filter parameters.
10850 Syntax for the argument is the same as for filter invocation, e.g.
10853 fontsize=56:fontcolor=green:text='Hello World'
10856 Full filter invocation with sendcmd would look like this:
10859 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10863 If the entire argument can't be parsed or applied as valid values then the filter will
10864 continue with its existing parameters.
10866 @subsection Examples
10870 Draw "Test Text" with font FreeSerif, using the default values for the
10871 optional parameters.
10874 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10878 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10879 and y=50 (counting from the top-left corner of the screen), text is
10880 yellow with a red box around it. Both the text and the box have an
10884 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10885 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10888 Note that the double quotes are not necessary if spaces are not used
10889 within the parameter list.
10892 Show the text at the center of the video frame:
10894 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10898 Show the text at a random position, switching to a new position every 30 seconds:
10900 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=if(eq(mod(t\,30)\,0)\,rand(0\,(w-text_w))\,x):y=if(eq(mod(t\,30)\,0)\,rand(0\,(h-text_h))\,y)"
10904 Show a text line sliding from right to left in the last row of the video
10905 frame. The file @file{LONG_LINE} is assumed to contain a single line
10908 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10912 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10914 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10918 Draw a single green letter "g", at the center of the input video.
10919 The glyph baseline is placed at half screen height.
10921 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10925 Show text for 1 second every 3 seconds:
10927 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10931 Use fontconfig to set the font. Note that the colons need to be escaped.
10933 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10937 Draw "Test Text" with font size dependent on height of the video.
10939 drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
10943 Print the date of a real-time encoding (see strftime(3)):
10945 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10949 Show text fading in and out (appearing/disappearing):
10952 DS=1.0 # display start
10953 DE=10.0 # display end
10954 FID=1.5 # fade in duration
10955 FOD=5 # fade out duration
10956 ffplay -f lavfi "color,drawtext=text=TEST:fontsize=50:fontfile=FreeSerif.ttf:fontcolor_expr=ff0000%@{eif\\\\: clip(255*(1*between(t\\, $DS + $FID\\, $DE - $FOD) + ((t - $DS)/$FID)*between(t\\, $DS\\, $DS + $FID) + (-(t - $DE)/$FOD)*between(t\\, $DE - $FOD\\, $DE) )\\, 0\\, 255) \\\\: x\\\\: 2 @}"
10960 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10961 and the @option{fontsize} value are included in the @option{y} offset.
10963 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10964 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10968 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10969 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10970 must have option @option{-export_path_metadata 1} for the special metadata fields
10971 to be available for filters.
10973 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10978 For more information about libfreetype, check:
10979 @url{http://www.freetype.org/}.
10981 For more information about fontconfig, check:
10982 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10984 For more information about libfribidi, check:
10985 @url{http://fribidi.org/}.
10987 @section edgedetect
10989 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10991 The filter accepts the following options:
10996 Set low and high threshold values used by the Canny thresholding
10999 The high threshold selects the "strong" edge pixels, which are then
11000 connected through 8-connectivity with the "weak" edge pixels selected
11001 by the low threshold.
11003 @var{low} and @var{high} threshold values must be chosen in the range
11004 [0,1], and @var{low} should be lesser or equal to @var{high}.
11006 Default value for @var{low} is @code{20/255}, and default value for @var{high}
11010 Define the drawing mode.
11014 Draw white/gray wires on black background.
11017 Mix the colors to create a paint/cartoon effect.
11020 Apply Canny edge detector on all selected planes.
11022 Default value is @var{wires}.
11025 Select planes for filtering. By default all available planes are filtered.
11028 @subsection Examples
11032 Standard edge detection with custom values for the hysteresis thresholding:
11034 edgedetect=low=0.1:high=0.4
11038 Painting effect without thresholding:
11040 edgedetect=mode=colormix:high=0
11046 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
11048 For each input image, the filter will compute the optimal mapping from
11049 the input to the output given the codebook length, that is the number
11050 of distinct output colors.
11052 This filter accepts the following options.
11055 @item codebook_length, l
11056 Set codebook length. The value must be a positive integer, and
11057 represents the number of distinct output colors. Default value is 256.
11060 Set the maximum number of iterations to apply for computing the optimal
11061 mapping. The higher the value the better the result and the higher the
11062 computation time. Default value is 1.
11065 Set a random seed, must be an integer included between 0 and
11066 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
11067 will try to use a good random seed on a best effort basis.
11070 Set pal8 output pixel format. This option does not work with codebook
11071 length greater than 256.
11076 Measure graylevel entropy in histogram of color channels of video frames.
11078 It accepts the following parameters:
11082 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
11084 @var{diff} mode measures entropy of histogram delta values, absolute differences
11085 between neighbour histogram values.
11089 Apply the EPX magnification filter which is designed for pixel art.
11091 It accepts the following option:
11095 Set the scaling dimension: @code{2} for @code{2xEPX}, @code{3} for
11097 Default is @code{3}.
11101 Set brightness, contrast, saturation and approximate gamma adjustment.
11103 The filter accepts the following options:
11107 Set the contrast expression. The value must be a float value in range
11108 @code{-1000.0} to @code{1000.0}. The default value is "1".
11111 Set the brightness expression. The value must be a float value in
11112 range @code{-1.0} to @code{1.0}. The default value is "0".
11115 Set the saturation expression. The value must be a float in
11116 range @code{0.0} to @code{3.0}. The default value is "1".
11119 Set the gamma expression. The value must be a float in range
11120 @code{0.1} to @code{10.0}. The default value is "1".
11123 Set the gamma expression for red. The value must be a float in
11124 range @code{0.1} to @code{10.0}. The default value is "1".
11127 Set the gamma expression for green. The value must be a float in range
11128 @code{0.1} to @code{10.0}. The default value is "1".
11131 Set the gamma expression for blue. The value must be a float in range
11132 @code{0.1} to @code{10.0}. The default value is "1".
11135 Set the gamma weight expression. It can be used to reduce the effect
11136 of a high gamma value on bright image areas, e.g. keep them from
11137 getting overamplified and just plain white. The value must be a float
11138 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
11139 gamma correction all the way down while @code{1.0} leaves it at its
11140 full strength. Default is "1".
11143 Set when the expressions for brightness, contrast, saturation and
11144 gamma expressions are evaluated.
11146 It accepts the following values:
11149 only evaluate expressions once during the filter initialization or
11150 when a command is processed
11153 evaluate expressions for each incoming frame
11156 Default value is @samp{init}.
11159 The expressions accept the following parameters:
11162 frame count of the input frame starting from 0
11165 byte position of the corresponding packet in the input file, NAN if
11169 frame rate of the input video, NAN if the input frame rate is unknown
11172 timestamp expressed in seconds, NAN if the input timestamp is unknown
11175 @subsection Commands
11176 The filter supports the following commands:
11180 Set the contrast expression.
11183 Set the brightness expression.
11186 Set the saturation expression.
11189 Set the gamma expression.
11192 Set the gamma_r expression.
11195 Set gamma_g expression.
11198 Set gamma_b expression.
11201 Set gamma_weight expression.
11203 The command accepts the same syntax of the corresponding option.
11205 If the specified expression is not valid, it is kept at its current
11212 Apply erosion effect to the video.
11214 This filter replaces the pixel by the local(3x3) minimum.
11216 It accepts the following options:
11223 Limit the maximum change for each plane, default is 65535.
11224 If 0, plane will remain unchanged.
11227 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
11230 Flags to local 3x3 coordinates maps like this:
11237 @subsection Commands
11239 This filter supports the all above options as @ref{commands}.
11243 Deinterlace the input video ("estdif" stands for "Edge Slope
11244 Tracing Deinterlacing Filter").
11246 Spatial only filter that uses edge slope tracing algorithm
11247 to interpolate missing lines.
11248 It accepts the following parameters:
11252 The interlacing mode to adopt. It accepts one of the following values:
11256 Output one frame for each frame.
11258 Output one frame for each field.
11261 The default value is @code{field}.
11264 The picture field parity assumed for the input interlaced video. It accepts one
11265 of the following values:
11269 Assume the top field is first.
11271 Assume the bottom field is first.
11273 Enable automatic detection of field parity.
11276 The default value is @code{auto}.
11277 If the interlacing is unknown or the decoder does not export this information,
11278 top field first will be assumed.
11281 Specify which frames to deinterlace. Accepts one of the following
11286 Deinterlace all frames.
11288 Only deinterlace frames marked as interlaced.
11291 The default value is @code{all}.
11294 Specify the search radius for edge slope tracing. Default value is 1.
11295 Allowed range is from 1 to 15.
11298 Specify the search radius for best edge matching. Default value is 2.
11299 Allowed range is from 0 to 15.
11302 Specify the interpolation used. Default is 4-point interpolation. It accepts one
11303 of the following values:
11307 Two-point interpolation.
11309 Four-point interpolation.
11311 Six-point interpolation.
11315 @subsection Commands
11316 This filter supports same @ref{commands} as options.
11318 @section extractplanes
11320 Extract color channel components from input video stream into
11321 separate grayscale video streams.
11323 The filter accepts the following option:
11327 Set plane(s) to extract.
11329 Available values for planes are:
11340 Choosing planes not available in the input will result in an error.
11341 That means you cannot select @code{r}, @code{g}, @code{b} planes
11342 with @code{y}, @code{u}, @code{v} planes at same time.
11345 @subsection Examples
11349 Extract luma, u and v color channel component from input video frame
11350 into 3 grayscale outputs:
11352 ffmpeg -i video.avi -filter_complex 'extractplanes=y+u+v[y][u][v]' -map '[y]' y.avi -map '[u]' u.avi -map '[v]' v.avi
11358 Apply a fade-in/out effect to the input video.
11360 It accepts the following parameters:
11364 The effect type can be either "in" for a fade-in, or "out" for a fade-out
11366 Default is @code{in}.
11368 @item start_frame, s
11369 Specify the number of the frame to start applying the fade
11370 effect at. Default is 0.
11373 The number of frames that the fade effect lasts. At the end of the
11374 fade-in effect, the output video will have the same intensity as the input video.
11375 At the end of the fade-out transition, the output video will be filled with the
11376 selected @option{color}.
11380 If set to 1, fade only alpha channel, if one exists on the input.
11381 Default value is 0.
11383 @item start_time, st
11384 Specify the timestamp (in seconds) of the frame to start to apply the fade
11385 effect. If both start_frame and start_time are specified, the fade will start at
11386 whichever comes last. Default is 0.
11389 The number of seconds for which the fade effect has to last. At the end of the
11390 fade-in effect the output video will have the same intensity as the input video,
11391 at the end of the fade-out transition the output video will be filled with the
11392 selected @option{color}.
11393 If both duration and nb_frames are specified, duration is used. Default is 0
11394 (nb_frames is used by default).
11397 Specify the color of the fade. Default is "black".
11400 @subsection Examples
11404 Fade in the first 30 frames of video:
11409 The command above is equivalent to:
11415 Fade out the last 45 frames of a 200-frame video:
11418 fade=type=out:start_frame=155:nb_frames=45
11422 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
11424 fade=in:0:25, fade=out:975:25
11428 Make the first 5 frames yellow, then fade in from frame 5-24:
11430 fade=in:5:20:color=yellow
11434 Fade in alpha over first 25 frames of video:
11436 fade=in:0:25:alpha=1
11440 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
11442 fade=t=in:st=5.5:d=0.5
11448 Denoise frames using 3D FFT (frequency domain filtering).
11450 The filter accepts the following options:
11454 Set the noise sigma constant. This sets denoising strength.
11455 Default value is 1. Allowed range is from 0 to 30.
11456 Using very high sigma with low overlap may give blocking artifacts.
11459 Set amount of denoising. By default all detected noise is reduced.
11460 Default value is 1. Allowed range is from 0 to 1.
11463 Set size of block, Default is 4, can be 3, 4, 5 or 6.
11464 Actual size of block in pixels is 2 to power of @var{block}, so by default
11465 block size in pixels is 2^4 which is 16.
11468 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
11471 Set number of previous frames to use for denoising. By default is set to 0.
11474 Set number of next frames to to use for denoising. By default is set to 0.
11477 Set planes which will be filtered, by default are all available filtered
11482 Apply arbitrary expressions to samples in frequency domain
11486 Adjust the dc value (gain) of the luma plane of the image. The filter
11487 accepts an integer value in range @code{0} to @code{1000}. The default
11488 value is set to @code{0}.
11491 Adjust the dc value (gain) of the 1st chroma plane of the image. The
11492 filter accepts an integer value in range @code{0} to @code{1000}. The
11493 default value is set to @code{0}.
11496 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
11497 filter accepts an integer value in range @code{0} to @code{1000}. The
11498 default value is set to @code{0}.
11501 Set the frequency domain weight expression for the luma plane.
11504 Set the frequency domain weight expression for the 1st chroma plane.
11507 Set the frequency domain weight expression for the 2nd chroma plane.
11510 Set when the expressions are evaluated.
11512 It accepts the following values:
11515 Only evaluate expressions once during the filter initialization.
11518 Evaluate expressions for each incoming frame.
11521 Default value is @samp{init}.
11523 The filter accepts the following variables:
11526 The coordinates of the current sample.
11530 The width and height of the image.
11533 The number of input frame, starting from 0.
11536 @subsection Examples
11542 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
11548 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
11554 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
11560 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
11567 Extract a single field from an interlaced image using stride
11568 arithmetic to avoid wasting CPU time. The output frames are marked as
11571 The filter accepts the following options:
11575 Specify whether to extract the top (if the value is @code{0} or
11576 @code{top}) or the bottom field (if the value is @code{1} or
11582 Create new frames by copying the top and bottom fields from surrounding frames
11583 supplied as numbers by the hint file.
11587 Set file containing hints: absolute/relative frame numbers.
11589 There must be one line for each frame in a clip. Each line must contain two
11590 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
11591 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
11592 is current frame number for @code{absolute} mode or out of [-1, 1] range
11593 for @code{relative} mode. First number tells from which frame to pick up top
11594 field and second number tells from which frame to pick up bottom field.
11596 If optionally followed by @code{+} output frame will be marked as interlaced,
11597 else if followed by @code{-} output frame will be marked as progressive, else
11598 it will be marked same as input frame.
11599 If optionally followed by @code{t} output frame will use only top field, or in
11600 case of @code{b} it will use only bottom field.
11601 If line starts with @code{#} or @code{;} that line is skipped.
11604 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
11607 Example of first several lines of @code{hint} file for @code{relative} mode:
11609 0,0 - # first frame
11610 1,0 - # second frame, use third's frame top field and second's frame bottom field
11611 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
11626 @section fieldmatch
11628 Field matching filter for inverse telecine. It is meant to reconstruct the
11629 progressive frames from a telecined stream. The filter does not drop duplicated
11630 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
11631 followed by a decimation filter such as @ref{decimate} in the filtergraph.
11633 The separation of the field matching and the decimation is notably motivated by
11634 the possibility of inserting a de-interlacing filter fallback between the two.
11635 If the source has mixed telecined and real interlaced content,
11636 @code{fieldmatch} will not be able to match fields for the interlaced parts.
11637 But these remaining combed frames will be marked as interlaced, and thus can be
11638 de-interlaced by a later filter such as @ref{yadif} before decimation.
11640 In addition to the various configuration options, @code{fieldmatch} can take an
11641 optional second stream, activated through the @option{ppsrc} option. If
11642 enabled, the frames reconstruction will be based on the fields and frames from
11643 this second stream. This allows the first input to be pre-processed in order to
11644 help the various algorithms of the filter, while keeping the output lossless
11645 (assuming the fields are matched properly). Typically, a field-aware denoiser,
11646 or brightness/contrast adjustments can help.
11648 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
11649 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
11650 which @code{fieldmatch} is based on. While the semantic and usage are very
11651 close, some behaviour and options names can differ.
11653 The @ref{decimate} filter currently only works for constant frame rate input.
11654 If your input has mixed telecined (30fps) and progressive content with a lower
11655 framerate like 24fps use the following filterchain to produce the necessary cfr
11656 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
11658 The filter accepts the following options:
11662 Specify the assumed field order of the input stream. Available values are:
11666 Auto detect parity (use FFmpeg's internal parity value).
11668 Assume bottom field first.
11670 Assume top field first.
11673 Note that it is sometimes recommended not to trust the parity announced by the
11676 Default value is @var{auto}.
11679 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
11680 sense that it won't risk creating jerkiness due to duplicate frames when
11681 possible, but if there are bad edits or blended fields it will end up
11682 outputting combed frames when a good match might actually exist. On the other
11683 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
11684 but will almost always find a good frame if there is one. The other values are
11685 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
11686 jerkiness and creating duplicate frames versus finding good matches in sections
11687 with bad edits, orphaned fields, blended fields, etc.
11689 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
11691 Available values are:
11695 2-way matching (p/c)
11697 2-way matching, and trying 3rd match if still combed (p/c + n)
11699 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
11701 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
11702 still combed (p/c + n + u/b)
11704 3-way matching (p/c/n)
11706 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
11707 detected as combed (p/c/n + u/b)
11710 The parenthesis at the end indicate the matches that would be used for that
11711 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
11714 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
11717 Default value is @var{pc_n}.
11720 Mark the main input stream as a pre-processed input, and enable the secondary
11721 input stream as the clean source to pick the fields from. See the filter
11722 introduction for more details. It is similar to the @option{clip2} feature from
11725 Default value is @code{0} (disabled).
11728 Set the field to match from. It is recommended to set this to the same value as
11729 @option{order} unless you experience matching failures with that setting. In
11730 certain circumstances changing the field that is used to match from can have a
11731 large impact on matching performance. Available values are:
11735 Automatic (same value as @option{order}).
11737 Match from the bottom field.
11739 Match from the top field.
11742 Default value is @var{auto}.
11745 Set whether or not chroma is included during the match comparisons. In most
11746 cases it is recommended to leave this enabled. You should set this to @code{0}
11747 only if your clip has bad chroma problems such as heavy rainbowing or other
11748 artifacts. Setting this to @code{0} could also be used to speed things up at
11749 the cost of some accuracy.
11751 Default value is @code{1}.
11755 These define an exclusion band which excludes the lines between @option{y0} and
11756 @option{y1} from being included in the field matching decision. An exclusion
11757 band can be used to ignore subtitles, a logo, or other things that may
11758 interfere with the matching. @option{y0} sets the starting scan line and
11759 @option{y1} sets the ending line; all lines in between @option{y0} and
11760 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
11761 @option{y0} and @option{y1} to the same value will disable the feature.
11762 @option{y0} and @option{y1} defaults to @code{0}.
11765 Set the scene change detection threshold as a percentage of maximum change on
11766 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
11767 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
11768 @option{scthresh} is @code{[0.0, 100.0]}.
11770 Default value is @code{12.0}.
11773 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
11774 account the combed scores of matches when deciding what match to use as the
11775 final match. Available values are:
11779 No final matching based on combed scores.
11781 Combed scores are only used when a scene change is detected.
11783 Use combed scores all the time.
11786 Default is @var{sc}.
11789 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
11790 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
11791 Available values are:
11795 No forced calculation.
11797 Force p/c/n calculations.
11799 Force p/c/n/u/b calculations.
11802 Default value is @var{none}.
11805 This is the area combing threshold used for combed frame detection. This
11806 essentially controls how "strong" or "visible" combing must be to be detected.
11807 Larger values mean combing must be more visible and smaller values mean combing
11808 can be less visible or strong and still be detected. Valid settings are from
11809 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
11810 be detected as combed). This is basically a pixel difference value. A good
11811 range is @code{[8, 12]}.
11813 Default value is @code{9}.
11816 Sets whether or not chroma is considered in the combed frame decision. Only
11817 disable this if your source has chroma problems (rainbowing, etc.) that are
11818 causing problems for the combed frame detection with chroma enabled. Actually,
11819 using @option{chroma}=@var{0} is usually more reliable, except for the case
11820 where there is chroma only combing in the source.
11822 Default value is @code{0}.
11826 Respectively set the x-axis and y-axis size of the window used during combed
11827 frame detection. This has to do with the size of the area in which
11828 @option{combpel} pixels are required to be detected as combed for a frame to be
11829 declared combed. See the @option{combpel} parameter description for more info.
11830 Possible values are any number that is a power of 2 starting at 4 and going up
11833 Default value is @code{16}.
11836 The number of combed pixels inside any of the @option{blocky} by
11837 @option{blockx} size blocks on the frame for the frame to be detected as
11838 combed. While @option{cthresh} controls how "visible" the combing must be, this
11839 setting controls "how much" combing there must be in any localized area (a
11840 window defined by the @option{blockx} and @option{blocky} settings) on the
11841 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
11842 which point no frames will ever be detected as combed). This setting is known
11843 as @option{MI} in TFM/VFM vocabulary.
11845 Default value is @code{80}.
11848 @anchor{p/c/n/u/b meaning}
11849 @subsection p/c/n/u/b meaning
11851 @subsubsection p/c/n
11853 We assume the following telecined stream:
11856 Top fields: 1 2 2 3 4
11857 Bottom fields: 1 2 3 4 4
11860 The numbers correspond to the progressive frame the fields relate to. Here, the
11861 first two frames are progressive, the 3rd and 4th are combed, and so on.
11863 When @code{fieldmatch} is configured to run a matching from bottom
11864 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11869 B 1 2 3 4 4 <-- matching reference
11878 As a result of the field matching, we can see that some frames get duplicated.
11879 To perform a complete inverse telecine, you need to rely on a decimation filter
11880 after this operation. See for instance the @ref{decimate} filter.
11882 The same operation now matching from top fields (@option{field}=@var{top})
11887 T 1 2 2 3 4 <-- matching reference
11897 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11898 basically, they refer to the frame and field of the opposite parity:
11901 @item @var{p} matches the field of the opposite parity in the previous frame
11902 @item @var{c} matches the field of the opposite parity in the current frame
11903 @item @var{n} matches the field of the opposite parity in the next frame
11908 The @var{u} and @var{b} matching are a bit special in the sense that they match
11909 from the opposite parity flag. In the following examples, we assume that we are
11910 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11911 'x' is placed above and below each matched fields.
11913 With bottom matching (@option{field}=@var{bottom}):
11918 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11919 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11927 With top matching (@option{field}=@var{top}):
11932 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11933 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11941 @subsection Examples
11943 Simple IVTC of a top field first telecined stream:
11945 fieldmatch=order=tff:combmatch=none, decimate
11948 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11950 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11953 @section fieldorder
11955 Transform the field order of the input video.
11957 It accepts the following parameters:
11962 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11963 for bottom field first.
11966 The default value is @samp{tff}.
11968 The transformation is done by shifting the picture content up or down
11969 by one line, and filling the remaining line with appropriate picture content.
11970 This method is consistent with most broadcast field order converters.
11972 If the input video is not flagged as being interlaced, or it is already
11973 flagged as being of the required output field order, then this filter does
11974 not alter the incoming video.
11976 It is very useful when converting to or from PAL DV material,
11977 which is bottom field first.
11981 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11984 @section fifo, afifo
11986 Buffer input images and send them when they are requested.
11988 It is mainly useful when auto-inserted by the libavfilter
11991 It does not take parameters.
11993 @section fillborders
11995 Fill borders of the input video, without changing video stream dimensions.
11996 Sometimes video can have garbage at the four edges and you may not want to
11997 crop video input to keep size multiple of some number.
11999 This filter accepts the following options:
12003 Number of pixels to fill from left border.
12006 Number of pixels to fill from right border.
12009 Number of pixels to fill from top border.
12012 Number of pixels to fill from bottom border.
12017 It accepts the following values:
12020 fill pixels using outermost pixels
12023 fill pixels using mirroring (half sample symmetric)
12026 fill pixels with constant value
12029 fill pixels using reflecting (whole sample symmetric)
12032 fill pixels using wrapping
12035 fade pixels to constant value
12038 Default is @var{smear}.
12041 Set color for pixels in fixed or fade mode. Default is @var{black}.
12044 @subsection Commands
12045 This filter supports same @ref{commands} as options.
12046 The command accepts the same syntax of the corresponding option.
12048 If the specified expression is not valid, it is kept at its current
12053 Find a rectangular object
12055 It accepts the following options:
12059 Filepath of the object image, needs to be in gray8.
12062 Detection threshold, default is 0.5.
12065 Number of mipmaps, default is 3.
12067 @item xmin, ymin, xmax, ymax
12068 Specifies the rectangle in which to search.
12071 @subsection Examples
12075 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
12077 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
12083 Flood area with values of same pixel components with another values.
12085 It accepts the following options:
12088 Set pixel x coordinate.
12091 Set pixel y coordinate.
12094 Set source #0 component value.
12097 Set source #1 component value.
12100 Set source #2 component value.
12103 Set source #3 component value.
12106 Set destination #0 component value.
12109 Set destination #1 component value.
12112 Set destination #2 component value.
12115 Set destination #3 component value.
12121 Convert the input video to one of the specified pixel formats.
12122 Libavfilter will try to pick one that is suitable as input to
12125 It accepts the following parameters:
12129 A '|'-separated list of pixel format names, such as
12130 "pix_fmts=yuv420p|monow|rgb24".
12134 @subsection Examples
12138 Convert the input video to the @var{yuv420p} format
12140 format=pix_fmts=yuv420p
12143 Convert the input video to any of the formats in the list
12145 format=pix_fmts=yuv420p|yuv444p|yuv410p
12152 Convert the video to specified constant frame rate by duplicating or dropping
12153 frames as necessary.
12155 It accepts the following parameters:
12159 The desired output frame rate. The default is @code{25}.
12162 Assume the first PTS should be the given value, in seconds. This allows for
12163 padding/trimming at the start of stream. By default, no assumption is made
12164 about the first frame's expected PTS, so no padding or trimming is done.
12165 For example, this could be set to 0 to pad the beginning with duplicates of
12166 the first frame if a video stream starts after the audio stream or to trim any
12167 frames with a negative PTS.
12170 Timestamp (PTS) rounding method.
12172 Possible values are:
12179 round towards -infinity
12181 round towards +infinity
12185 The default is @code{near}.
12188 Action performed when reading the last frame.
12190 Possible values are:
12193 Use same timestamp rounding method as used for other frames.
12195 Pass through last frame if input duration has not been reached yet.
12197 The default is @code{round}.
12201 Alternatively, the options can be specified as a flat string:
12202 @var{fps}[:@var{start_time}[:@var{round}]].
12204 See also the @ref{setpts} filter.
12206 @subsection Examples
12210 A typical usage in order to set the fps to 25:
12216 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
12218 fps=fps=film:round=near
12224 Pack two different video streams into a stereoscopic video, setting proper
12225 metadata on supported codecs. The two views should have the same size and
12226 framerate and processing will stop when the shorter video ends. Please note
12227 that you may conveniently adjust view properties with the @ref{scale} and
12230 It accepts the following parameters:
12234 The desired packing format. Supported values are:
12239 The views are next to each other (default).
12242 The views are on top of each other.
12245 The views are packed by line.
12248 The views are packed by column.
12251 The views are temporally interleaved.
12260 # Convert left and right views into a frame-sequential video
12261 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
12263 # Convert views into a side-by-side video with the same output resolution as the input
12264 ffmpeg -i LEFT -i RIGHT -filter_complex [0:v]scale=w=iw/2[left],[1:v]scale=w=iw/2[right],[left][right]framepack=sbs OUTPUT
12269 Change the frame rate by interpolating new video output frames from the source
12272 This filter is not designed to function correctly with interlaced media. If
12273 you wish to change the frame rate of interlaced media then you are required
12274 to deinterlace before this filter and re-interlace after this filter.
12276 A description of the accepted options follows.
12280 Specify the output frames per second. This option can also be specified
12281 as a value alone. The default is @code{50}.
12284 Specify the start of a range where the output frame will be created as a
12285 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12286 the default is @code{15}.
12289 Specify the end of a range where the output frame will be created as a
12290 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12291 the default is @code{240}.
12294 Specify the level at which a scene change is detected as a value between
12295 0 and 100 to indicate a new scene; a low value reflects a low
12296 probability for the current frame to introduce a new scene, while a higher
12297 value means the current frame is more likely to be one.
12298 The default is @code{8.2}.
12301 Specify flags influencing the filter process.
12303 Available value for @var{flags} is:
12306 @item scene_change_detect, scd
12307 Enable scene change detection using the value of the option @var{scene}.
12308 This flag is enabled by default.
12314 Select one frame every N-th frame.
12316 This filter accepts the following option:
12319 Select frame after every @code{step} frames.
12320 Allowed values are positive integers higher than 0. Default value is @code{1}.
12323 @section freezedetect
12325 Detect frozen video.
12327 This filter logs a message and sets frame metadata when it detects that the
12328 input video has no significant change in content during a specified duration.
12329 Video freeze detection calculates the mean average absolute difference of all
12330 the components of video frames and compares it to a noise floor.
12332 The printed times and duration are expressed in seconds. The
12333 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
12334 whose timestamp equals or exceeds the detection duration and it contains the
12335 timestamp of the first frame of the freeze. The
12336 @code{lavfi.freezedetect.freeze_duration} and
12337 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
12340 The filter accepts the following options:
12344 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
12345 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
12349 Set freeze duration until notification (default is 2 seconds).
12352 @section freezeframes
12354 Freeze video frames.
12356 This filter freezes video frames using frame from 2nd input.
12358 The filter accepts the following options:
12362 Set number of first frame from which to start freeze.
12365 Set number of last frame from which to end freeze.
12368 Set number of frame from 2nd input which will be used instead of replaced frames.
12374 Apply a frei0r effect to the input video.
12376 To enable the compilation of this filter, you need to install the frei0r
12377 header and configure FFmpeg with @code{--enable-frei0r}.
12379 It accepts the following parameters:
12384 The name of the frei0r effect to load. If the environment variable
12385 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
12386 directories specified by the colon-separated list in @env{FREI0R_PATH}.
12387 Otherwise, the standard frei0r paths are searched, in this order:
12388 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
12389 @file{/usr/lib/frei0r-1/}.
12391 @item filter_params
12392 A '|'-separated list of parameters to pass to the frei0r effect.
12396 A frei0r effect parameter can be a boolean (its value is either
12397 "y" or "n"), a double, a color (specified as
12398 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
12399 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
12400 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
12401 a position (specified as @var{X}/@var{Y}, where
12402 @var{X} and @var{Y} are floating point numbers) and/or a string.
12404 The number and types of parameters depend on the loaded effect. If an
12405 effect parameter is not specified, the default value is set.
12407 @subsection Examples
12411 Apply the distort0r effect, setting the first two double parameters:
12413 frei0r=filter_name=distort0r:filter_params=0.5|0.01
12417 Apply the colordistance effect, taking a color as the first parameter:
12419 frei0r=colordistance:0.2/0.3/0.4
12420 frei0r=colordistance:violet
12421 frei0r=colordistance:0x112233
12425 Apply the perspective effect, specifying the top left and top right image
12428 frei0r=perspective:0.2/0.2|0.8/0.2
12432 For more information, see
12433 @url{http://frei0r.dyne.org}
12435 @subsection Commands
12437 This filter supports the @option{filter_params} option as @ref{commands}.
12441 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
12443 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
12444 processing filter, one of them is performed once per block, not per pixel.
12445 This allows for much higher speed.
12447 The filter accepts the following options:
12451 Set quality. This option defines the number of levels for averaging. It accepts
12452 an integer in the range 4-5. Default value is @code{4}.
12455 Force a constant quantization parameter. It accepts an integer in range 0-63.
12456 If not set, the filter will use the QP from the video stream (if available).
12459 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
12460 more details but also more artifacts, while higher values make the image smoother
12461 but also blurrier. Default value is @code{0} − PSNR optimal.
12463 @item use_bframe_qp
12464 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12465 option may cause flicker since the B-Frames have often larger QP. Default is
12466 @code{0} (not enabled).
12472 Apply Gaussian blur filter.
12474 The filter accepts the following options:
12478 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
12481 Set number of steps for Gaussian approximation. Default is @code{1}.
12484 Set which planes to filter. By default all planes are filtered.
12487 Set vertical sigma, if negative it will be same as @code{sigma}.
12488 Default is @code{-1}.
12491 @subsection Commands
12492 This filter supports same commands as options.
12493 The command accepts the same syntax of the corresponding option.
12495 If the specified expression is not valid, it is kept at its current
12500 Apply generic equation to each pixel.
12502 The filter accepts the following options:
12505 @item lum_expr, lum
12506 Set the luminance expression.
12508 Set the chrominance blue expression.
12510 Set the chrominance red expression.
12511 @item alpha_expr, a
12512 Set the alpha expression.
12514 Set the red expression.
12515 @item green_expr, g
12516 Set the green expression.
12518 Set the blue expression.
12521 The colorspace is selected according to the specified options. If one
12522 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
12523 options is specified, the filter will automatically select a YCbCr
12524 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
12525 @option{blue_expr} options is specified, it will select an RGB
12528 If one of the chrominance expression is not defined, it falls back on the other
12529 one. If no alpha expression is specified it will evaluate to opaque value.
12530 If none of chrominance expressions are specified, they will evaluate
12531 to the luminance expression.
12533 The expressions can use the following variables and functions:
12537 The sequential number of the filtered frame, starting from @code{0}.
12541 The coordinates of the current sample.
12545 The width and height of the image.
12549 Width and height scale depending on the currently filtered plane. It is the
12550 ratio between the corresponding luma plane number of pixels and the current
12551 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
12552 @code{0.5,0.5} for chroma planes.
12555 Time of the current frame, expressed in seconds.
12558 Return the value of the pixel at location (@var{x},@var{y}) of the current
12562 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
12566 Return the value of the pixel at location (@var{x},@var{y}) of the
12567 blue-difference chroma plane. Return 0 if there is no such plane.
12570 Return the value of the pixel at location (@var{x},@var{y}) of the
12571 red-difference chroma plane. Return 0 if there is no such plane.
12576 Return the value of the pixel at location (@var{x},@var{y}) of the
12577 red/green/blue component. Return 0 if there is no such component.
12580 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
12581 plane. Return 0 if there is no such plane.
12583 @item psum(x,y), lumsum(x, y), cbsum(x,y), crsum(x,y), rsum(x,y), gsum(x,y), bsum(x,y), alphasum(x,y)
12584 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
12585 sums of samples within a rectangle. See the functions without the sum postfix.
12587 @item interpolation
12588 Set one of interpolation methods:
12593 Default is bilinear.
12596 For functions, if @var{x} and @var{y} are outside the area, the value will be
12597 automatically clipped to the closer edge.
12599 Please note that this filter can use multiple threads in which case each slice
12600 will have its own expression state. If you want to use only a single expression
12601 state because your expressions depend on previous state then you should limit
12602 the number of filter threads to 1.
12604 @subsection Examples
12608 Flip the image horizontally:
12614 Generate a bidimensional sine wave, with angle @code{PI/3} and a
12615 wavelength of 100 pixels:
12617 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
12621 Generate a fancy enigmatic moving light:
12623 nullsrc=s=256x256,geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128
12627 Generate a quick emboss effect:
12629 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
12633 Modify RGB components depending on pixel position:
12635 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
12639 Create a radial gradient that is the same size as the input (also see
12640 the @ref{vignette} filter):
12642 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
12648 Fix the banding artifacts that are sometimes introduced into nearly flat
12649 regions by truncation to 8-bit color depth.
12650 Interpolate the gradients that should go where the bands are, and
12653 It is designed for playback only. Do not use it prior to
12654 lossy compression, because compression tends to lose the dither and
12655 bring back the bands.
12657 It accepts the following parameters:
12662 The maximum amount by which the filter will change any one pixel. This is also
12663 the threshold for detecting nearly flat regions. Acceptable values range from
12664 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
12668 The neighborhood to fit the gradient to. A larger radius makes for smoother
12669 gradients, but also prevents the filter from modifying the pixels near detailed
12670 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
12671 values will be clipped to the valid range.
12675 Alternatively, the options can be specified as a flat string:
12676 @var{strength}[:@var{radius}]
12678 @subsection Examples
12682 Apply the filter with a @code{3.5} strength and radius of @code{8}:
12688 Specify radius, omitting the strength (which will fall-back to the default
12696 @anchor{graphmonitor}
12697 @section graphmonitor
12698 Show various filtergraph stats.
12700 With this filter one can debug complete filtergraph.
12701 Especially issues with links filling with queued frames.
12703 The filter accepts the following options:
12707 Set video output size. Default is @var{hd720}.
12710 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
12713 Set output mode, can be @var{fulll} or @var{compact}.
12714 In @var{compact} mode only filters with some queued frames have displayed stats.
12717 Set flags which enable which stats are shown in video.
12719 Available values for flags are:
12722 Display number of queued frames in each link.
12724 @item frame_count_in
12725 Display number of frames taken from filter.
12727 @item frame_count_out
12728 Display number of frames given out from filter.
12731 Display current filtered frame pts.
12734 Display current filtered frame time.
12737 Display time base for filter link.
12740 Display used format for filter link.
12743 Display video size or number of audio channels in case of audio used by filter link.
12746 Display video frame rate or sample rate in case of audio used by filter link.
12749 Display link output status.
12753 Set upper limit for video rate of output stream, Default value is @var{25}.
12754 This guarantee that output video frame rate will not be higher than this value.
12758 A color constancy variation filter which estimates scene illumination via grey edge algorithm
12759 and corrects the scene colors accordingly.
12761 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
12763 The filter accepts the following options:
12767 The order of differentiation to be applied on the scene. Must be chosen in the range
12768 [0,2] and default value is 1.
12771 The Minkowski parameter to be used for calculating the Minkowski distance. Must
12772 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
12773 max value instead of calculating Minkowski distance.
12776 The standard deviation of Gaussian blur to be applied on the scene. Must be
12777 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
12778 can't be equal to 0 if @var{difford} is greater than 0.
12781 @subsection Examples
12787 greyedge=difford=1:minknorm=5:sigma=2
12793 greyedge=difford=1:minknorm=0:sigma=2
12801 Apply a Hald CLUT to a video stream.
12803 First input is the video stream to process, and second one is the Hald CLUT.
12804 The Hald CLUT input can be a simple picture or a complete video stream.
12806 The filter accepts the following options:
12810 Force termination when the shortest input terminates. Default is @code{0}.
12812 Continue applying the last CLUT after the end of the stream. A value of
12813 @code{0} disable the filter after the last frame of the CLUT is reached.
12814 Default is @code{1}.
12817 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
12818 filters share the same internals).
12820 This filter also supports the @ref{framesync} options.
12822 More information about the Hald CLUT can be found on Eskil Steenberg's website
12823 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
12825 @subsection Commands
12827 This filter supports the @code{interp} option as @ref{commands}.
12829 @subsection Workflow examples
12831 @subsubsection Hald CLUT video stream
12833 Generate an identity Hald CLUT stream altered with various effects:
12835 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "hue=H=2*PI*t:s=sin(2*PI*t)+1, curves=cross_process" -t 10 -c:v ffv1 clut.nut
12838 Note: make sure you use a lossless codec.
12840 Then use it with @code{haldclut} to apply it on some random stream:
12842 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
12845 The Hald CLUT will be applied to the 10 first seconds (duration of
12846 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
12847 to the remaining frames of the @code{mandelbrot} stream.
12849 @subsubsection Hald CLUT with preview
12851 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
12852 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
12853 biggest possible square starting at the top left of the picture. The remaining
12854 padding pixels (bottom or right) will be ignored. This area can be used to add
12855 a preview of the Hald CLUT.
12857 Typically, the following generated Hald CLUT will be supported by the
12858 @code{haldclut} filter:
12861 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
12862 pad=iw+320 [padded_clut];
12863 smptebars=s=320x256, split [a][b];
12864 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
12865 [main][b] overlay=W-320" -frames:v 1 clut.png
12868 It contains the original and a preview of the effect of the CLUT: SMPTE color
12869 bars are displayed on the right-top, and below the same color bars processed by
12872 Then, the effect of this Hald CLUT can be visualized with:
12874 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
12879 Flip the input video horizontally.
12881 For example, to horizontally flip the input video with @command{ffmpeg}:
12883 ffmpeg -i in.avi -vf "hflip" out.avi
12887 This filter applies a global color histogram equalization on a
12890 It can be used to correct video that has a compressed range of pixel
12891 intensities. The filter redistributes the pixel intensities to
12892 equalize their distribution across the intensity range. It may be
12893 viewed as an "automatically adjusting contrast filter". This filter is
12894 useful only for correcting degraded or poorly captured source
12897 The filter accepts the following options:
12901 Determine the amount of equalization to be applied. As the strength
12902 is reduced, the distribution of pixel intensities more-and-more
12903 approaches that of the input frame. The value must be a float number
12904 in the range [0,1] and defaults to 0.200.
12907 Set the maximum intensity that can generated and scale the output
12908 values appropriately. The strength should be set as desired and then
12909 the intensity can be limited if needed to avoid washing-out. The value
12910 must be a float number in the range [0,1] and defaults to 0.210.
12913 Set the antibanding level. If enabled the filter will randomly vary
12914 the luminance of output pixels by a small amount to avoid banding of
12915 the histogram. Possible values are @code{none}, @code{weak} or
12916 @code{strong}. It defaults to @code{none}.
12922 Compute and draw a color distribution histogram for the input video.
12924 The computed histogram is a representation of the color component
12925 distribution in an image.
12927 Standard histogram displays the color components distribution in an image.
12928 Displays color graph for each color component. Shows distribution of
12929 the Y, U, V, A or R, G, B components, depending on input format, in the
12930 current frame. Below each graph a color component scale meter is shown.
12932 The filter accepts the following options:
12936 Set height of level. Default value is @code{200}.
12937 Allowed range is [50, 2048].
12940 Set height of color scale. Default value is @code{12}.
12941 Allowed range is [0, 40].
12945 It accepts the following values:
12948 Per color component graphs are placed below each other.
12951 Per color component graphs are placed side by side.
12954 Presents information identical to that in the @code{parade}, except
12955 that the graphs representing color components are superimposed directly
12958 Default is @code{stack}.
12961 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12962 Default is @code{linear}.
12965 Set what color components to display.
12966 Default is @code{7}.
12969 Set foreground opacity. Default is @code{0.7}.
12972 Set background opacity. Default is @code{0.5}.
12975 @subsection Examples
12980 Calculate and draw histogram:
12982 ffplay -i input -vf histogram
12990 This is a high precision/quality 3d denoise filter. It aims to reduce
12991 image noise, producing smooth images and making still images really
12992 still. It should enhance compressibility.
12994 It accepts the following optional parameters:
12998 A non-negative floating point number which specifies spatial luma strength.
12999 It defaults to 4.0.
13001 @item chroma_spatial
13002 A non-negative floating point number which specifies spatial chroma strength.
13003 It defaults to 3.0*@var{luma_spatial}/4.0.
13006 A floating point number which specifies luma temporal strength. It defaults to
13007 6.0*@var{luma_spatial}/4.0.
13010 A floating point number which specifies chroma temporal strength. It defaults to
13011 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
13014 @subsection Commands
13015 This filter supports same @ref{commands} as options.
13016 The command accepts the same syntax of the corresponding option.
13018 If the specified expression is not valid, it is kept at its current
13021 @anchor{hwdownload}
13022 @section hwdownload
13024 Download hardware frames to system memory.
13026 The input must be in hardware frames, and the output a non-hardware format.
13027 Not all formats will be supported on the output - it may be necessary to insert
13028 an additional @option{format} filter immediately following in the graph to get
13029 the output in a supported format.
13033 Map hardware frames to system memory or to another device.
13035 This filter has several different modes of operation; which one is used depends
13036 on the input and output formats:
13039 Hardware frame input, normal frame output
13041 Map the input frames to system memory and pass them to the output. If the
13042 original hardware frame is later required (for example, after overlaying
13043 something else on part of it), the @option{hwmap} filter can be used again
13044 in the next mode to retrieve it.
13046 Normal frame input, hardware frame output
13048 If the input is actually a software-mapped hardware frame, then unmap it -
13049 that is, return the original hardware frame.
13051 Otherwise, a device must be provided. Create new hardware surfaces on that
13052 device for the output, then map them back to the software format at the input
13053 and give those frames to the preceding filter. This will then act like the
13054 @option{hwupload} filter, but may be able to avoid an additional copy when
13055 the input is already in a compatible format.
13057 Hardware frame input and output
13059 A device must be supplied for the output, either directly or with the
13060 @option{derive_device} option. The input and output devices must be of
13061 different types and compatible - the exact meaning of this is
13062 system-dependent, but typically it means that they must refer to the same
13063 underlying hardware context (for example, refer to the same graphics card).
13065 If the input frames were originally created on the output device, then unmap
13066 to retrieve the original frames.
13068 Otherwise, map the frames to the output device - create new hardware frames
13069 on the output corresponding to the frames on the input.
13072 The following additional parameters are accepted:
13076 Set the frame mapping mode. Some combination of:
13079 The mapped frame should be readable.
13081 The mapped frame should be writeable.
13083 The mapping will always overwrite the entire frame.
13085 This may improve performance in some cases, as the original contents of the
13086 frame need not be loaded.
13088 The mapping must not involve any copying.
13090 Indirect mappings to copies of frames are created in some cases where either
13091 direct mapping is not possible or it would have unexpected properties.
13092 Setting this flag ensures that the mapping is direct and will fail if that is
13095 Defaults to @var{read+write} if not specified.
13097 @item derive_device @var{type}
13098 Rather than using the device supplied at initialisation, instead derive a new
13099 device of type @var{type} from the device the input frames exist on.
13102 In a hardware to hardware mapping, map in reverse - create frames in the sink
13103 and map them back to the source. This may be necessary in some cases where
13104 a mapping in one direction is required but only the opposite direction is
13105 supported by the devices being used.
13107 This option is dangerous - it may break the preceding filter in undefined
13108 ways if there are any additional constraints on that filter's output.
13109 Do not use it without fully understanding the implications of its use.
13115 Upload system memory frames to hardware surfaces.
13117 The device to upload to must be supplied when the filter is initialised. If
13118 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
13119 option or with the @option{derive_device} option. The input and output devices
13120 must be of different types and compatible - the exact meaning of this is
13121 system-dependent, but typically it means that they must refer to the same
13122 underlying hardware context (for example, refer to the same graphics card).
13124 The following additional parameters are accepted:
13127 @item derive_device @var{type}
13128 Rather than using the device supplied at initialisation, instead derive a new
13129 device of type @var{type} from the device the input frames exist on.
13132 @anchor{hwupload_cuda}
13133 @section hwupload_cuda
13135 Upload system memory frames to a CUDA device.
13137 It accepts the following optional parameters:
13141 The number of the CUDA device to use
13146 Apply a high-quality magnification filter designed for pixel art. This filter
13147 was originally created by Maxim Stepin.
13149 It accepts the following option:
13153 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
13154 @code{hq3x} and @code{4} for @code{hq4x}.
13155 Default is @code{3}.
13159 Stack input videos horizontally.
13161 All streams must be of same pixel format and of same height.
13163 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13164 to create same output.
13166 The filter accepts the following option:
13170 Set number of input streams. Default is 2.
13173 If set to 1, force the output to terminate when the shortest input
13174 terminates. Default value is 0.
13179 Modify the hue and/or the saturation of the input.
13181 It accepts the following parameters:
13185 Specify the hue angle as a number of degrees. It accepts an expression,
13186 and defaults to "0".
13189 Specify the saturation in the [-10,10] range. It accepts an expression and
13193 Specify the hue angle as a number of radians. It accepts an
13194 expression, and defaults to "0".
13197 Specify the brightness in the [-10,10] range. It accepts an expression and
13201 @option{h} and @option{H} are mutually exclusive, and can't be
13202 specified at the same time.
13204 The @option{b}, @option{h}, @option{H} and @option{s} option values are
13205 expressions containing the following constants:
13209 frame count of the input frame starting from 0
13212 presentation timestamp of the input frame expressed in time base units
13215 frame rate of the input video, NAN if the input frame rate is unknown
13218 timestamp expressed in seconds, NAN if the input timestamp is unknown
13221 time base of the input video
13224 @subsection Examples
13228 Set the hue to 90 degrees and the saturation to 1.0:
13234 Same command but expressing the hue in radians:
13240 Rotate hue and make the saturation swing between 0
13241 and 2 over a period of 1 second:
13243 hue="H=2*PI*t: s=sin(2*PI*t)+1"
13247 Apply a 3 seconds saturation fade-in effect starting at 0:
13249 hue="s=min(t/3\,1)"
13252 The general fade-in expression can be written as:
13254 hue="s=min(0\, max((t-START)/DURATION\, 1))"
13258 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
13260 hue="s=max(0\, min(1\, (8-t)/3))"
13263 The general fade-out expression can be written as:
13265 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
13270 @subsection Commands
13272 This filter supports the following commands:
13278 Modify the hue and/or the saturation and/or brightness of the input video.
13279 The command accepts the same syntax of the corresponding option.
13281 If the specified expression is not valid, it is kept at its current
13285 @section hysteresis
13287 Grow first stream into second stream by connecting components.
13288 This makes it possible to build more robust edge masks.
13290 This filter accepts the following options:
13294 Set which planes will be processed as bitmap, unprocessed planes will be
13295 copied from first stream.
13296 By default value 0xf, all planes will be processed.
13299 Set threshold which is used in filtering. If pixel component value is higher than
13300 this value filter algorithm for connecting components is activated.
13301 By default value is 0.
13304 The @code{hysteresis} filter also supports the @ref{framesync} options.
13308 Detect video interlacing type.
13310 This filter tries to detect if the input frames are interlaced, progressive,
13311 top or bottom field first. It will also try to detect fields that are
13312 repeated between adjacent frames (a sign of telecine).
13314 Single frame detection considers only immediately adjacent frames when classifying each frame.
13315 Multiple frame detection incorporates the classification history of previous frames.
13317 The filter will log these metadata values:
13320 @item single.current_frame
13321 Detected type of current frame using single-frame detection. One of:
13322 ``tff'' (top field first), ``bff'' (bottom field first),
13323 ``progressive'', or ``undetermined''
13326 Cumulative number of frames detected as top field first using single-frame detection.
13329 Cumulative number of frames detected as top field first using multiple-frame detection.
13332 Cumulative number of frames detected as bottom field first using single-frame detection.
13334 @item multiple.current_frame
13335 Detected type of current frame using multiple-frame detection. One of:
13336 ``tff'' (top field first), ``bff'' (bottom field first),
13337 ``progressive'', or ``undetermined''
13340 Cumulative number of frames detected as bottom field first using multiple-frame detection.
13342 @item single.progressive
13343 Cumulative number of frames detected as progressive using single-frame detection.
13345 @item multiple.progressive
13346 Cumulative number of frames detected as progressive using multiple-frame detection.
13348 @item single.undetermined
13349 Cumulative number of frames that could not be classified using single-frame detection.
13351 @item multiple.undetermined
13352 Cumulative number of frames that could not be classified using multiple-frame detection.
13354 @item repeated.current_frame
13355 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
13357 @item repeated.neither
13358 Cumulative number of frames with no repeated field.
13361 Cumulative number of frames with the top field repeated from the previous frame's top field.
13363 @item repeated.bottom
13364 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
13367 The filter accepts the following options:
13371 Set interlacing threshold.
13373 Set progressive threshold.
13375 Threshold for repeated field detection.
13377 Number of frames after which a given frame's contribution to the
13378 statistics is halved (i.e., it contributes only 0.5 to its
13379 classification). The default of 0 means that all frames seen are given
13380 full weight of 1.0 forever.
13381 @item analyze_interlaced_flag
13382 When this is not 0 then idet will use the specified number of frames to determine
13383 if the interlaced flag is accurate, it will not count undetermined frames.
13384 If the flag is found to be accurate it will be used without any further
13385 computations, if it is found to be inaccurate it will be cleared without any
13386 further computations. This allows inserting the idet filter as a low computational
13387 method to clean up the interlaced flag
13392 Deinterleave or interleave fields.
13394 This filter allows one to process interlaced images fields without
13395 deinterlacing them. Deinterleaving splits the input frame into 2
13396 fields (so called half pictures). Odd lines are moved to the top
13397 half of the output image, even lines to the bottom half.
13398 You can process (filter) them independently and then re-interleave them.
13400 The filter accepts the following options:
13404 @item chroma_mode, c
13405 @item alpha_mode, a
13406 Available values for @var{luma_mode}, @var{chroma_mode} and
13407 @var{alpha_mode} are:
13413 @item deinterleave, d
13414 Deinterleave fields, placing one above the other.
13416 @item interleave, i
13417 Interleave fields. Reverse the effect of deinterleaving.
13419 Default value is @code{none}.
13421 @item luma_swap, ls
13422 @item chroma_swap, cs
13423 @item alpha_swap, as
13424 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
13427 @subsection Commands
13429 This filter supports the all above options as @ref{commands}.
13433 Apply inflate effect to the video.
13435 This filter replaces the pixel by the local(3x3) average by taking into account
13436 only values higher than the pixel.
13438 It accepts the following options:
13445 Limit the maximum change for each plane, default is 65535.
13446 If 0, plane will remain unchanged.
13449 @subsection Commands
13451 This filter supports the all above options as @ref{commands}.
13455 Simple interlacing filter from progressive contents. This interleaves upper (or
13456 lower) lines from odd frames with lower (or upper) lines from even frames,
13457 halving the frame rate and preserving image height.
13460 Original Original New Frame
13461 Frame 'j' Frame 'j+1' (tff)
13462 ========== =========== ==================
13463 Line 0 --------------------> Frame 'j' Line 0
13464 Line 1 Line 1 ----> Frame 'j+1' Line 1
13465 Line 2 ---------------------> Frame 'j' Line 2
13466 Line 3 Line 3 ----> Frame 'j+1' Line 3
13468 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
13471 It accepts the following optional parameters:
13475 This determines whether the interlaced frame is taken from the even
13476 (tff - default) or odd (bff) lines of the progressive frame.
13479 Vertical lowpass filter to avoid twitter interlacing and
13480 reduce moire patterns.
13484 Disable vertical lowpass filter
13487 Enable linear filter (default)
13490 Enable complex filter. This will slightly less reduce twitter and moire
13491 but better retain detail and subjective sharpness impression.
13498 Deinterlace input video by applying Donald Graft's adaptive kernel
13499 deinterling. Work on interlaced parts of a video to produce
13500 progressive frames.
13502 The description of the accepted parameters follows.
13506 Set the threshold which affects the filter's tolerance when
13507 determining if a pixel line must be processed. It must be an integer
13508 in the range [0,255] and defaults to 10. A value of 0 will result in
13509 applying the process on every pixels.
13512 Paint pixels exceeding the threshold value to white if set to 1.
13516 Set the fields order. Swap fields if set to 1, leave fields alone if
13520 Enable additional sharpening if set to 1. Default is 0.
13523 Enable twoway sharpening if set to 1. Default is 0.
13526 @subsection Examples
13530 Apply default values:
13532 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
13536 Enable additional sharpening:
13542 Paint processed pixels in white:
13549 Apply kirsch operator to input video stream.
13551 The filter accepts the following option:
13555 Set which planes will be processed, unprocessed planes will be copied.
13556 By default value 0xf, all planes will be processed.
13559 Set value which will be multiplied with filtered result.
13562 Set value which will be added to filtered result.
13565 @subsection Commands
13567 This filter supports the all above options as @ref{commands}.
13571 Slowly update darker pixels.
13573 This filter makes short flashes of light appear longer.
13574 This filter accepts the following options:
13578 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
13581 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
13584 @subsection Commands
13586 This filter supports the all above options as @ref{commands}.
13588 @section lenscorrection
13590 Correct radial lens distortion
13592 This filter can be used to correct for radial distortion as can result from the use
13593 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
13594 one can use tools available for example as part of opencv or simply trial-and-error.
13595 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
13596 and extract the k1 and k2 coefficients from the resulting matrix.
13598 Note that effectively the same filter is available in the open-source tools Krita and
13599 Digikam from the KDE project.
13601 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
13602 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
13603 brightness distribution, so you may want to use both filters together in certain
13604 cases, though you will have to take care of ordering, i.e. whether vignetting should
13605 be applied before or after lens correction.
13607 @subsection Options
13609 The filter accepts the following options:
13613 Relative x-coordinate of the focal point of the image, and thereby the center of the
13614 distortion. This value has a range [0,1] and is expressed as fractions of the image
13615 width. Default is 0.5.
13617 Relative y-coordinate of the focal point of the image, and thereby the center of the
13618 distortion. This value has a range [0,1] and is expressed as fractions of the image
13619 height. Default is 0.5.
13621 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
13622 no correction. Default is 0.
13624 Coefficient of the double quadratic correction term. This value has a range [-1,1].
13625 0 means no correction. Default is 0.
13627 Set interpolation type. Can be @code{nearest} or @code{bilinear}.
13628 Default is @code{nearest}.
13630 Specify the color of the unmapped pixels. For the syntax of this option,
13631 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13632 manual,ffmpeg-utils}. Default color is @code{black@@0}.
13635 The formula that generates the correction is:
13637 @var{r_src} = @var{r_tgt} * (1 + @var{k1} * (@var{r_tgt} / @var{r_0})^2 + @var{k2} * (@var{r_tgt} / @var{r_0})^4)
13639 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
13640 distances from the focal point in the source and target images, respectively.
13642 @subsection Commands
13644 This filter supports the all above options as @ref{commands}.
13648 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
13650 The @code{lensfun} filter requires the camera make, camera model, and lens model
13651 to apply the lens correction. The filter will load the lensfun database and
13652 query it to find the corresponding camera and lens entries in the database. As
13653 long as these entries can be found with the given options, the filter can
13654 perform corrections on frames. Note that incomplete strings will result in the
13655 filter choosing the best match with the given options, and the filter will
13656 output the chosen camera and lens models (logged with level "info"). You must
13657 provide the make, camera model, and lens model as they are required.
13659 The filter accepts the following options:
13663 The make of the camera (for example, "Canon"). This option is required.
13666 The model of the camera (for example, "Canon EOS 100D"). This option is
13670 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
13671 option is required.
13674 The type of correction to apply. The following values are valid options:
13678 Enables fixing lens vignetting.
13681 Enables fixing lens geometry. This is the default.
13684 Enables fixing chromatic aberrations.
13687 Enables fixing lens vignetting and lens geometry.
13690 Enables fixing lens vignetting and chromatic aberrations.
13693 Enables fixing both lens geometry and chromatic aberrations.
13696 Enables all possible corrections.
13700 The focal length of the image/video (zoom; expected constant for video). For
13701 example, a 18--55mm lens has focal length range of [18--55], so a value in that
13702 range should be chosen when using that lens. Default 18.
13705 The aperture of the image/video (expected constant for video). Note that
13706 aperture is only used for vignetting correction. Default 3.5.
13708 @item focus_distance
13709 The focus distance of the image/video (expected constant for video). Note that
13710 focus distance is only used for vignetting and only slightly affects the
13711 vignetting correction process. If unknown, leave it at the default value (which
13715 The scale factor which is applied after transformation. After correction the
13716 video is no longer necessarily rectangular. This parameter controls how much of
13717 the resulting image is visible. The value 0 means that a value will be chosen
13718 automatically such that there is little or no unmapped area in the output
13719 image. 1.0 means that no additional scaling is done. Lower values may result
13720 in more of the corrected image being visible, while higher values may avoid
13721 unmapped areas in the output.
13723 @item target_geometry
13724 The target geometry of the output image/video. The following values are valid
13728 @item rectilinear (default)
13731 @item equirectangular
13732 @item fisheye_orthographic
13733 @item fisheye_stereographic
13734 @item fisheye_equisolid
13735 @item fisheye_thoby
13738 Apply the reverse of image correction (instead of correcting distortion, apply
13741 @item interpolation
13742 The type of interpolation used when correcting distortion. The following values
13747 @item linear (default)
13752 @subsection Examples
13756 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
13757 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
13761 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8 -c:v h264 -b:v 8000k output.mov
13765 Apply the same as before, but only for the first 5 seconds of video.
13768 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8:enable='lte(t\,5)' -c:v h264 -b:v 8000k output.mov
13775 Obtain the VMAF (Video Multi-Method Assessment Fusion)
13776 score between two input videos.
13778 The obtained VMAF score is printed through the logging system.
13780 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
13781 After installing the library it can be enabled using:
13782 @code{./configure --enable-libvmaf}.
13783 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
13785 The filter has following options:
13789 Set the model path which is to be used for SVM.
13790 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
13793 Set the file path to be used to store logs.
13796 Set the format of the log file (csv, json or xml).
13798 @item enable_transform
13799 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
13800 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
13801 Default value: @code{false}
13804 Invokes the phone model which will generate VMAF scores higher than in the
13805 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
13806 Default value: @code{false}
13809 Enables computing psnr along with vmaf.
13810 Default value: @code{false}
13813 Enables computing ssim along with vmaf.
13814 Default value: @code{false}
13817 Enables computing ms_ssim along with vmaf.
13818 Default value: @code{false}
13821 Set the pool method to be used for computing vmaf.
13822 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
13825 Set number of threads to be used when computing vmaf.
13826 Default value: @code{0}, which makes use of all available logical processors.
13829 Set interval for frame subsampling used when computing vmaf.
13830 Default value: @code{1}
13832 @item enable_conf_interval
13833 Enables confidence interval.
13834 Default value: @code{false}
13837 This filter also supports the @ref{framesync} options.
13839 @subsection Examples
13842 On the below examples the input file @file{main.mpg} being processed is
13843 compared with the reference file @file{ref.mpg}.
13846 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
13850 Example with options:
13852 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
13856 Example with options and different containers:
13858 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]libvmaf=psnr=1:log_fmt=json" -f null -
13864 Limits the pixel components values to the specified range [min, max].
13866 The filter accepts the following options:
13870 Lower bound. Defaults to the lowest allowed value for the input.
13873 Upper bound. Defaults to the highest allowed value for the input.
13876 Specify which planes will be processed. Defaults to all available.
13879 @subsection Commands
13881 This filter supports the all above options as @ref{commands}.
13887 The filter accepts the following options:
13891 Set the number of loops. Setting this value to -1 will result in infinite loops.
13895 Set maximal size in number of frames. Default is 0.
13898 Set first frame of loop. Default is 0.
13901 @subsection Examples
13905 Loop single first frame infinitely:
13907 loop=loop=-1:size=1:start=0
13911 Loop single first frame 10 times:
13913 loop=loop=10:size=1:start=0
13917 Loop 10 first frames 5 times:
13919 loop=loop=5:size=10:start=0
13925 Apply a 1D LUT to an input video.
13927 The filter accepts the following options:
13931 Set the 1D LUT file name.
13933 Currently supported formats:
13942 Select interpolation mode.
13944 Available values are:
13948 Use values from the nearest defined point.
13950 Interpolate values using the linear interpolation.
13952 Interpolate values using the cosine interpolation.
13954 Interpolate values using the cubic interpolation.
13956 Interpolate values using the spline interpolation.
13960 @subsection Commands
13962 This filter supports the all above options as @ref{commands}.
13967 Apply a 3D LUT to an input video.
13969 The filter accepts the following options:
13973 Set the 3D LUT file name.
13975 Currently supported formats:
13989 Select interpolation mode.
13991 Available values are:
13995 Use values from the nearest defined point.
13997 Interpolate values using the 8 points defining a cube.
13999 Interpolate values using a tetrahedron.
14001 Interpolate values using a pyramid.
14003 Interpolate values using a prism.
14007 @subsection Commands
14009 This filter supports the @code{interp} option as @ref{commands}.
14013 Turn certain luma values into transparency.
14015 The filter accepts the following options:
14019 Set the luma which will be used as base for transparency.
14020 Default value is @code{0}.
14023 Set the range of luma values to be keyed out.
14024 Default value is @code{0.01}.
14027 Set the range of softness. Default value is @code{0}.
14028 Use this to control gradual transition from zero to full transparency.
14031 @subsection Commands
14032 This filter supports same @ref{commands} as options.
14033 The command accepts the same syntax of the corresponding option.
14035 If the specified expression is not valid, it is kept at its current
14038 @section lut, lutrgb, lutyuv
14040 Compute a look-up table for binding each pixel component input value
14041 to an output value, and apply it to the input video.
14043 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
14044 to an RGB input video.
14046 These filters accept the following parameters:
14049 set first pixel component expression
14051 set second pixel component expression
14053 set third pixel component expression
14055 set fourth pixel component expression, corresponds to the alpha component
14058 set red component expression
14060 set green component expression
14062 set blue component expression
14064 alpha component expression
14067 set Y/luminance component expression
14069 set U/Cb component expression
14071 set V/Cr component expression
14074 Each of them specifies the expression to use for computing the lookup table for
14075 the corresponding pixel component values.
14077 The exact component associated to each of the @var{c*} options depends on the
14080 The @var{lut} filter requires either YUV or RGB pixel formats in input,
14081 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
14083 The expressions can contain the following constants and functions:
14088 The input width and height.
14091 The input value for the pixel component.
14094 The input value, clipped to the @var{minval}-@var{maxval} range.
14097 The maximum value for the pixel component.
14100 The minimum value for the pixel component.
14103 The negated value for the pixel component value, clipped to the
14104 @var{minval}-@var{maxval} range; it corresponds to the expression
14105 "maxval-clipval+minval".
14108 The computed value in @var{val}, clipped to the
14109 @var{minval}-@var{maxval} range.
14111 @item gammaval(gamma)
14112 The computed gamma correction value of the pixel component value,
14113 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
14115 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
14119 All expressions default to "val".
14121 @subsection Commands
14123 This filter supports same @ref{commands} as options.
14125 @subsection Examples
14129 Negate input video:
14131 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
14132 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
14135 The above is the same as:
14137 lutrgb="r=negval:g=negval:b=negval"
14138 lutyuv="y=negval:u=negval:v=negval"
14148 Remove chroma components, turning the video into a graytone image:
14150 lutyuv="u=128:v=128"
14154 Apply a luma burning effect:
14160 Remove green and blue components:
14166 Set a constant alpha channel value on input:
14168 format=rgba,lutrgb=a="maxval-minval/2"
14172 Correct luminance gamma by a factor of 0.5:
14174 lutyuv=y=gammaval(0.5)
14178 Discard least significant bits of luma:
14180 lutyuv=y='bitand(val, 128+64+32)'
14184 Technicolor like effect:
14186 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
14190 @section lut2, tlut2
14192 The @code{lut2} filter takes two input streams and outputs one
14195 The @code{tlut2} (time lut2) filter takes two consecutive frames
14196 from one single stream.
14198 This filter accepts the following parameters:
14201 set first pixel component expression
14203 set second pixel component expression
14205 set third pixel component expression
14207 set fourth pixel component expression, corresponds to the alpha component
14210 set output bit depth, only available for @code{lut2} filter. By default is 0,
14211 which means bit depth is automatically picked from first input format.
14214 The @code{lut2} filter also supports the @ref{framesync} options.
14216 Each of them specifies the expression to use for computing the lookup table for
14217 the corresponding pixel component values.
14219 The exact component associated to each of the @var{c*} options depends on the
14222 The expressions can contain the following constants:
14227 The input width and height.
14230 The first input value for the pixel component.
14233 The second input value for the pixel component.
14236 The first input video bit depth.
14239 The second input video bit depth.
14242 All expressions default to "x".
14244 @subsection Commands
14246 This filter supports the all above options as @ref{commands} except option @code{d}.
14248 @subsection Examples
14252 Highlight differences between two RGB video streams:
14254 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1)'
14258 Highlight differences between two YUV video streams:
14260 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1)'
14264 Show max difference between two video streams:
14266 lut2='if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1)))'
14270 @section maskedclamp
14272 Clamp the first input stream with the second input and third input stream.
14274 Returns the value of first stream to be between second input
14275 stream - @code{undershoot} and third input stream + @code{overshoot}.
14277 This filter accepts the following options:
14280 Default value is @code{0}.
14283 Default value is @code{0}.
14286 Set which planes will be processed as bitmap, unprocessed planes will be
14287 copied from first stream.
14288 By default value 0xf, all planes will be processed.
14291 @subsection Commands
14293 This filter supports the all above options as @ref{commands}.
14297 Merge the second and third input stream into output stream using absolute differences
14298 between second input stream and first input stream and absolute difference between
14299 third input stream and first input stream. The picked value will be from second input
14300 stream if second absolute difference is greater than first one or from third input stream
14303 This filter accepts the following options:
14306 Set which planes will be processed as bitmap, unprocessed planes will be
14307 copied from first stream.
14308 By default value 0xf, all planes will be processed.
14311 @subsection Commands
14313 This filter supports the all above options as @ref{commands}.
14315 @section maskedmerge
14317 Merge the first input stream with the second input stream using per pixel
14318 weights in the third input stream.
14320 A value of 0 in the third stream pixel component means that pixel component
14321 from first stream is returned unchanged, while maximum value (eg. 255 for
14322 8-bit videos) means that pixel component from second stream is returned
14323 unchanged. Intermediate values define the amount of merging between both
14324 input stream's pixel components.
14326 This filter accepts the following options:
14329 Set which planes will be processed as bitmap, unprocessed planes will be
14330 copied from first stream.
14331 By default value 0xf, all planes will be processed.
14334 @subsection Commands
14336 This filter supports the all above options as @ref{commands}.
14340 Merge the second and third input stream into output stream using absolute differences
14341 between second input stream and first input stream and absolute difference between
14342 third input stream and first input stream. The picked value will be from second input
14343 stream if second absolute difference is less than first one or from third input stream
14346 This filter accepts the following options:
14349 Set which planes will be processed as bitmap, unprocessed planes will be
14350 copied from first stream.
14351 By default value 0xf, all planes will be processed.
14354 @subsection Commands
14356 This filter supports the all above options as @ref{commands}.
14358 @section maskedthreshold
14359 Pick pixels comparing absolute difference of two video streams with fixed
14362 If absolute difference between pixel component of first and second video
14363 stream is equal or lower than user supplied threshold than pixel component
14364 from first video stream is picked, otherwise pixel component from second
14365 video stream is picked.
14367 This filter accepts the following options:
14370 Set threshold used when picking pixels from absolute difference from two input
14374 Set which planes will be processed as bitmap, unprocessed planes will be
14375 copied from second stream.
14376 By default value 0xf, all planes will be processed.
14379 @subsection Commands
14381 This filter supports the all above options as @ref{commands}.
14384 Create mask from input video.
14386 For example it is useful to create motion masks after @code{tblend} filter.
14388 This filter accepts the following options:
14392 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
14395 Set high threshold. Any pixel component higher than this value will be set to max value
14396 allowed for current pixel format.
14399 Set planes to filter, by default all available planes are filtered.
14402 Fill all frame pixels with this value.
14405 Set max average pixel value for frame. If sum of all pixel components is higher that this
14406 average, output frame will be completely filled with value set by @var{fill} option.
14407 Typically useful for scene changes when used in combination with @code{tblend} filter.
14410 @subsection Commands
14412 This filter supports the all above options as @ref{commands}.
14416 Apply motion-compensation deinterlacing.
14418 It needs one field per frame as input and must thus be used together
14419 with yadif=1/3 or equivalent.
14421 This filter accepts the following options:
14424 Set the deinterlacing mode.
14426 It accepts one of the following values:
14431 use iterative motion estimation
14433 like @samp{slow}, but use multiple reference frames.
14435 Default value is @samp{fast}.
14438 Set the picture field parity assumed for the input video. It must be
14439 one of the following values:
14443 assume top field first
14445 assume bottom field first
14448 Default value is @samp{bff}.
14451 Set per-block quantization parameter (QP) used by the internal
14454 Higher values should result in a smoother motion vector field but less
14455 optimal individual vectors. Default value is 1.
14460 Pick median pixel from certain rectangle defined by radius.
14462 This filter accepts the following options:
14466 Set horizontal radius size. Default value is @code{1}.
14467 Allowed range is integer from 1 to 127.
14470 Set which planes to process. Default is @code{15}, which is all available planes.
14473 Set vertical radius size. Default value is @code{0}.
14474 Allowed range is integer from 0 to 127.
14475 If it is 0, value will be picked from horizontal @code{radius} option.
14478 Set median percentile. Default value is @code{0.5}.
14479 Default value of @code{0.5} will pick always median values, while @code{0} will pick
14480 minimum values, and @code{1} maximum values.
14483 @subsection Commands
14484 This filter supports same @ref{commands} as options.
14485 The command accepts the same syntax of the corresponding option.
14487 If the specified expression is not valid, it is kept at its current
14490 @section mergeplanes
14492 Merge color channel components from several video streams.
14494 The filter accepts up to 4 input streams, and merge selected input
14495 planes to the output video.
14497 This filter accepts the following options:
14500 Set input to output plane mapping. Default is @code{0}.
14502 The mappings is specified as a bitmap. It should be specified as a
14503 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
14504 mapping for the first plane of the output stream. 'A' sets the number of
14505 the input stream to use (from 0 to 3), and 'a' the plane number of the
14506 corresponding input to use (from 0 to 3). The rest of the mappings is
14507 similar, 'Bb' describes the mapping for the output stream second
14508 plane, 'Cc' describes the mapping for the output stream third plane and
14509 'Dd' describes the mapping for the output stream fourth plane.
14512 Set output pixel format. Default is @code{yuva444p}.
14515 @subsection Examples
14519 Merge three gray video streams of same width and height into single video stream:
14521 [a0][a1][a2]mergeplanes=0x001020:yuv444p
14525 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
14527 [a0][a1]mergeplanes=0x00010210:yuva444p
14531 Swap Y and A plane in yuva444p stream:
14533 format=yuva444p,mergeplanes=0x03010200:yuva444p
14537 Swap U and V plane in yuv420p stream:
14539 format=yuv420p,mergeplanes=0x000201:yuv420p
14543 Cast a rgb24 clip to yuv444p:
14545 format=rgb24,mergeplanes=0x000102:yuv444p
14551 Estimate and export motion vectors using block matching algorithms.
14552 Motion vectors are stored in frame side data to be used by other filters.
14554 This filter accepts the following options:
14557 Specify the motion estimation method. Accepts one of the following values:
14561 Exhaustive search algorithm.
14563 Three step search algorithm.
14565 Two dimensional logarithmic search algorithm.
14567 New three step search algorithm.
14569 Four step search algorithm.
14571 Diamond search algorithm.
14573 Hexagon-based search algorithm.
14575 Enhanced predictive zonal search algorithm.
14577 Uneven multi-hexagon search algorithm.
14579 Default value is @samp{esa}.
14582 Macroblock size. Default @code{16}.
14585 Search parameter. Default @code{7}.
14588 @section midequalizer
14590 Apply Midway Image Equalization effect using two video streams.
14592 Midway Image Equalization adjusts a pair of images to have the same
14593 histogram, while maintaining their dynamics as much as possible. It's
14594 useful for e.g. matching exposures from a pair of stereo cameras.
14596 This filter has two inputs and one output, which must be of same pixel format, but
14597 may be of different sizes. The output of filter is first input adjusted with
14598 midway histogram of both inputs.
14600 This filter accepts the following option:
14604 Set which planes to process. Default is @code{15}, which is all available planes.
14607 @section minterpolate
14609 Convert the video to specified frame rate using motion interpolation.
14611 This filter accepts the following options:
14614 Specify the output frame rate. This can be rational e.g. @code{60000/1001}. Frames are dropped if @var{fps} is lower than source fps. Default @code{60}.
14617 Motion interpolation mode. Following values are accepted:
14620 Duplicate previous or next frame for interpolating new ones.
14622 Blend source frames. Interpolated frame is mean of previous and next frames.
14624 Motion compensated interpolation. Following options are effective when this mode is selected:
14628 Motion compensation mode. Following values are accepted:
14631 Overlapped block motion compensation.
14633 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
14635 Default mode is @samp{obmc}.
14638 Motion estimation mode. Following values are accepted:
14641 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
14643 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
14645 Default mode is @samp{bilat}.
14648 The algorithm to be used for motion estimation. Following values are accepted:
14651 Exhaustive search algorithm.
14653 Three step search algorithm.
14655 Two dimensional logarithmic search algorithm.
14657 New three step search algorithm.
14659 Four step search algorithm.
14661 Diamond search algorithm.
14663 Hexagon-based search algorithm.
14665 Enhanced predictive zonal search algorithm.
14667 Uneven multi-hexagon search algorithm.
14669 Default algorithm is @samp{epzs}.
14672 Macroblock size. Default @code{16}.
14675 Motion estimation search parameter. Default @code{32}.
14678 Enable variable-size block motion compensation. Motion estimation is applied with smaller block sizes at object boundaries in order to make the them less blur. Default is @code{0} (disabled).
14683 Scene change detection method. Scene change leads motion vectors to be in random direction. Scene change detection replace interpolated frames by duplicate ones. May not be needed for other modes. Following values are accepted:
14686 Disable scene change detection.
14688 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
14690 Default method is @samp{fdiff}.
14692 @item scd_threshold
14693 Scene change detection threshold. Default is @code{10.}.
14698 Mix several video input streams into one video stream.
14700 A description of the accepted options follows.
14704 The number of inputs. If unspecified, it defaults to 2.
14707 Specify weight of each input video stream as sequence.
14708 Each weight is separated by space. If number of weights
14709 is smaller than number of @var{frames} last specified
14710 weight will be used for all remaining unset weights.
14713 Specify scale, if it is set it will be multiplied with sum
14714 of each weight multiplied with pixel values to give final destination
14715 pixel value. By default @var{scale} is auto scaled to sum of weights.
14718 Specify how end of stream is determined.
14721 The duration of the longest input. (default)
14724 The duration of the shortest input.
14727 The duration of the first input.
14731 @subsection Commands
14733 This filter supports the following commands:
14737 Syntax is same as option with same name.
14740 @section mpdecimate
14742 Drop frames that do not differ greatly from the previous frame in
14743 order to reduce frame rate.
14745 The main use of this filter is for very-low-bitrate encoding
14746 (e.g. streaming over dialup modem), but it could in theory be used for
14747 fixing movies that were inverse-telecined incorrectly.
14749 A description of the accepted options follows.
14753 Set the maximum number of consecutive frames which can be dropped (if
14754 positive), or the minimum interval between dropped frames (if
14755 negative). If the value is 0, the frame is dropped disregarding the
14756 number of previous sequentially dropped frames.
14758 Default value is 0.
14763 Set the dropping threshold values.
14765 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
14766 represent actual pixel value differences, so a threshold of 64
14767 corresponds to 1 unit of difference for each pixel, or the same spread
14768 out differently over the block.
14770 A frame is a candidate for dropping if no 8x8 blocks differ by more
14771 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
14772 meaning the whole image) differ by more than a threshold of @option{lo}.
14774 Default value for @option{hi} is 64*12, default value for @option{lo} is
14775 64*5, and default value for @option{frac} is 0.33.
14781 Negate (invert) the input video.
14783 It accepts the following option:
14788 With value 1, it negates the alpha component, if present. Default value is 0.
14791 @subsection Commands
14793 This filter supports same @ref{commands} as options.
14798 Denoise frames using Non-Local Means algorithm.
14800 Each pixel is adjusted by looking for other pixels with similar contexts. This
14801 context similarity is defined by comparing their surrounding patches of size
14802 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
14805 Note that the research area defines centers for patches, which means some
14806 patches will be made of pixels outside that research area.
14808 The filter accepts the following options.
14812 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
14815 Set patch size. Default is 7. Must be odd number in range [0, 99].
14818 Same as @option{p} but for chroma planes.
14820 The default value is @var{0} and means automatic.
14823 Set research size. Default is 15. Must be odd number in range [0, 99].
14826 Same as @option{r} but for chroma planes.
14828 The default value is @var{0} and means automatic.
14833 Deinterlace video using neural network edge directed interpolation.
14835 This filter accepts the following options:
14839 Mandatory option, without binary file filter can not work.
14840 Currently file can be found here:
14841 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
14844 Set which frames to deinterlace, by default it is @code{all}.
14845 Can be @code{all} or @code{interlaced}.
14848 Set mode of operation.
14850 Can be one of the following:
14854 Use frame flags, both fields.
14856 Use frame flags, single field.
14858 Use top field only.
14860 Use bottom field only.
14862 Use both fields, top first.
14864 Use both fields, bottom first.
14868 Set which planes to process, by default filter process all frames.
14871 Set size of local neighborhood around each pixel, used by the predictor neural
14874 Can be one of the following:
14887 Set the number of neurons in predictor neural network.
14888 Can be one of the following:
14899 Controls the number of different neural network predictions that are blended
14900 together to compute the final output value. Can be @code{fast}, default or
14904 Set which set of weights to use in the predictor.
14905 Can be one of the following:
14909 weights trained to minimize absolute error
14911 weights trained to minimize squared error
14915 Controls whether or not the prescreener neural network is used to decide
14916 which pixels should be processed by the predictor neural network and which
14917 can be handled by simple cubic interpolation.
14918 The prescreener is trained to know whether cubic interpolation will be
14919 sufficient for a pixel or whether it should be predicted by the predictor nn.
14920 The computational complexity of the prescreener nn is much less than that of
14921 the predictor nn. Since most pixels can be handled by cubic interpolation,
14922 using the prescreener generally results in much faster processing.
14923 The prescreener is pretty accurate, so the difference between using it and not
14924 using it is almost always unnoticeable.
14926 Can be one of the following:
14936 Default is @code{new}.
14939 @subsection Commands
14940 This filter supports same @ref{commands} as options, excluding @var{weights} option.
14944 Force libavfilter not to use any of the specified pixel formats for the
14945 input to the next filter.
14947 It accepts the following parameters:
14951 A '|'-separated list of pixel format names, such as
14952 pix_fmts=yuv420p|monow|rgb24".
14956 @subsection Examples
14960 Force libavfilter to use a format different from @var{yuv420p} for the
14961 input to the vflip filter:
14963 noformat=pix_fmts=yuv420p,vflip
14967 Convert the input video to any of the formats not contained in the list:
14969 noformat=yuv420p|yuv444p|yuv410p
14975 Add noise on video input frame.
14977 The filter accepts the following options:
14985 Set noise seed for specific pixel component or all pixel components in case
14986 of @var{all_seed}. Default value is @code{123457}.
14988 @item all_strength, alls
14989 @item c0_strength, c0s
14990 @item c1_strength, c1s
14991 @item c2_strength, c2s
14992 @item c3_strength, c3s
14993 Set noise strength for specific pixel component or all pixel components in case
14994 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
14996 @item all_flags, allf
14997 @item c0_flags, c0f
14998 @item c1_flags, c1f
14999 @item c2_flags, c2f
15000 @item c3_flags, c3f
15001 Set pixel component flags or set flags for all components if @var{all_flags}.
15002 Available values for component flags are:
15005 averaged temporal noise (smoother)
15007 mix random noise with a (semi)regular pattern
15009 temporal noise (noise pattern changes between frames)
15011 uniform noise (gaussian otherwise)
15015 @subsection Examples
15017 Add temporal and uniform noise to input video:
15019 noise=alls=20:allf=t+u
15024 Normalize RGB video (aka histogram stretching, contrast stretching).
15025 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
15027 For each channel of each frame, the filter computes the input range and maps
15028 it linearly to the user-specified output range. The output range defaults
15029 to the full dynamic range from pure black to pure white.
15031 Temporal smoothing can be used on the input range to reduce flickering (rapid
15032 changes in brightness) caused when small dark or bright objects enter or leave
15033 the scene. This is similar to the auto-exposure (automatic gain control) on a
15034 video camera, and, like a video camera, it may cause a period of over- or
15035 under-exposure of the video.
15037 The R,G,B channels can be normalized independently, which may cause some
15038 color shifting, or linked together as a single channel, which prevents
15039 color shifting. Linked normalization preserves hue. Independent normalization
15040 does not, so it can be used to remove some color casts. Independent and linked
15041 normalization can be combined in any ratio.
15043 The normalize filter accepts the following options:
15048 Colors which define the output range. The minimum input value is mapped to
15049 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
15050 The defaults are black and white respectively. Specifying white for
15051 @var{blackpt} and black for @var{whitept} will give color-inverted,
15052 normalized video. Shades of grey can be used to reduce the dynamic range
15053 (contrast). Specifying saturated colors here can create some interesting
15057 The number of previous frames to use for temporal smoothing. The input range
15058 of each channel is smoothed using a rolling average over the current frame
15059 and the @var{smoothing} previous frames. The default is 0 (no temporal
15063 Controls the ratio of independent (color shifting) channel normalization to
15064 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
15065 independent. Defaults to 1.0 (fully independent).
15068 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
15069 expensive no-op. Defaults to 1.0 (full strength).
15073 @subsection Commands
15074 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
15075 The command accepts the same syntax of the corresponding option.
15077 If the specified expression is not valid, it is kept at its current
15080 @subsection Examples
15082 Stretch video contrast to use the full dynamic range, with no temporal
15083 smoothing; may flicker depending on the source content:
15085 normalize=blackpt=black:whitept=white:smoothing=0
15088 As above, but with 50 frames of temporal smoothing; flicker should be
15089 reduced, depending on the source content:
15091 normalize=blackpt=black:whitept=white:smoothing=50
15094 As above, but with hue-preserving linked channel normalization:
15096 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
15099 As above, but with half strength:
15101 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
15104 Map the darkest input color to red, the brightest input color to cyan:
15106 normalize=blackpt=red:whitept=cyan
15111 Pass the video source unchanged to the output.
15114 Optical Character Recognition
15116 This filter uses Tesseract for optical character recognition. To enable
15117 compilation of this filter, you need to configure FFmpeg with
15118 @code{--enable-libtesseract}.
15120 It accepts the following options:
15124 Set datapath to tesseract data. Default is to use whatever was
15125 set at installation.
15128 Set language, default is "eng".
15131 Set character whitelist.
15134 Set character blacklist.
15137 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
15138 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
15142 Apply a video transform using libopencv.
15144 To enable this filter, install the libopencv library and headers and
15145 configure FFmpeg with @code{--enable-libopencv}.
15147 It accepts the following parameters:
15152 The name of the libopencv filter to apply.
15154 @item filter_params
15155 The parameters to pass to the libopencv filter. If not specified, the default
15156 values are assumed.
15160 Refer to the official libopencv documentation for more precise
15162 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
15164 Several libopencv filters are supported; see the following subsections.
15169 Dilate an image by using a specific structuring element.
15170 It corresponds to the libopencv function @code{cvDilate}.
15172 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
15174 @var{struct_el} represents a structuring element, and has the syntax:
15175 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
15177 @var{cols} and @var{rows} represent the number of columns and rows of
15178 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
15179 point, and @var{shape} the shape for the structuring element. @var{shape}
15180 must be "rect", "cross", "ellipse", or "custom".
15182 If the value for @var{shape} is "custom", it must be followed by a
15183 string of the form "=@var{filename}". The file with name
15184 @var{filename} is assumed to represent a binary image, with each
15185 printable character corresponding to a bright pixel. When a custom
15186 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
15187 or columns and rows of the read file are assumed instead.
15189 The default value for @var{struct_el} is "3x3+0x0/rect".
15191 @var{nb_iterations} specifies the number of times the transform is
15192 applied to the image, and defaults to 1.
15196 # Use the default values
15199 # Dilate using a structuring element with a 5x5 cross, iterating two times
15200 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
15202 # Read the shape from the file diamond.shape, iterating two times.
15203 # The file diamond.shape may contain a pattern of characters like this
15209 # The specified columns and rows are ignored
15210 # but the anchor point coordinates are not
15211 ocv=dilate:0x0+2x2/custom=diamond.shape|2
15216 Erode an image by using a specific structuring element.
15217 It corresponds to the libopencv function @code{cvErode}.
15219 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
15220 with the same syntax and semantics as the @ref{dilate} filter.
15224 Smooth the input video.
15226 The filter takes the following parameters:
15227 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
15229 @var{type} is the type of smooth filter to apply, and must be one of
15230 the following values: "blur", "blur_no_scale", "median", "gaussian",
15231 or "bilateral". The default value is "gaussian".
15233 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
15234 depends on the smooth type. @var{param1} and
15235 @var{param2} accept integer positive values or 0. @var{param3} and
15236 @var{param4} accept floating point values.
15238 The default value for @var{param1} is 3. The default value for the
15239 other parameters is 0.
15241 These parameters correspond to the parameters assigned to the
15242 libopencv function @code{cvSmooth}.
15244 @section oscilloscope
15246 2D Video Oscilloscope.
15248 Useful to measure spatial impulse, step responses, chroma delays, etc.
15250 It accepts the following parameters:
15254 Set scope center x position.
15257 Set scope center y position.
15260 Set scope size, relative to frame diagonal.
15263 Set scope tilt/rotation.
15269 Set trace center x position.
15272 Set trace center y position.
15275 Set trace width, relative to width of frame.
15278 Set trace height, relative to height of frame.
15281 Set which components to trace. By default it traces first three components.
15284 Draw trace grid. By default is enabled.
15287 Draw some statistics. By default is enabled.
15290 Draw scope. By default is enabled.
15293 @subsection Commands
15294 This filter supports same @ref{commands} as options.
15295 The command accepts the same syntax of the corresponding option.
15297 If the specified expression is not valid, it is kept at its current
15300 @subsection Examples
15304 Inspect full first row of video frame.
15306 oscilloscope=x=0.5:y=0:s=1
15310 Inspect full last row of video frame.
15312 oscilloscope=x=0.5:y=1:s=1
15316 Inspect full 5th line of video frame of height 1080.
15318 oscilloscope=x=0.5:y=5/1080:s=1
15322 Inspect full last column of video frame.
15324 oscilloscope=x=1:y=0.5:s=1:t=1
15332 Overlay one video on top of another.
15334 It takes two inputs and has one output. The first input is the "main"
15335 video on which the second input is overlaid.
15337 It accepts the following parameters:
15339 A description of the accepted options follows.
15344 Set the expression for the x and y coordinates of the overlaid video
15345 on the main video. Default value is "0" for both expressions. In case
15346 the expression is invalid, it is set to a huge value (meaning that the
15347 overlay will not be displayed within the output visible area).
15350 See @ref{framesync}.
15353 Set when the expressions for @option{x}, and @option{y} are evaluated.
15355 It accepts the following values:
15358 only evaluate expressions once during the filter initialization or
15359 when a command is processed
15362 evaluate expressions for each incoming frame
15365 Default value is @samp{frame}.
15368 See @ref{framesync}.
15371 Set the format for the output video.
15373 It accepts the following values:
15376 force YUV420 output
15379 force YUV420p10 output
15382 force YUV422 output
15385 force YUV422p10 output
15388 force YUV444 output
15391 force packed RGB output
15394 force planar RGB output
15397 automatically pick format
15400 Default value is @samp{yuv420}.
15403 See @ref{framesync}.
15406 Set format of alpha of the overlaid video, it can be @var{straight} or
15407 @var{premultiplied}. Default is @var{straight}.
15410 The @option{x}, and @option{y} expressions can contain the following
15416 The main input width and height.
15420 The overlay input width and height.
15424 The computed values for @var{x} and @var{y}. They are evaluated for
15429 horizontal and vertical chroma subsample values of the output
15430 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
15434 the number of input frame, starting from 0
15437 the position in the file of the input frame, NAN if unknown
15440 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
15444 This filter also supports the @ref{framesync} options.
15446 Note that the @var{n}, @var{pos}, @var{t} variables are available only
15447 when evaluation is done @emph{per frame}, and will evaluate to NAN
15448 when @option{eval} is set to @samp{init}.
15450 Be aware that frames are taken from each input video in timestamp
15451 order, hence, if their initial timestamps differ, it is a good idea
15452 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
15453 have them begin in the same zero timestamp, as the example for
15454 the @var{movie} filter does.
15456 You can chain together more overlays but you should test the
15457 efficiency of such approach.
15459 @subsection Commands
15461 This filter supports the following commands:
15465 Modify the x and y of the overlay input.
15466 The command accepts the same syntax of the corresponding option.
15468 If the specified expression is not valid, it is kept at its current
15472 @subsection Examples
15476 Draw the overlay at 10 pixels from the bottom right corner of the main
15479 overlay=main_w-overlay_w-10:main_h-overlay_h-10
15482 Using named options the example above becomes:
15484 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
15488 Insert a transparent PNG logo in the bottom left corner of the input,
15489 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
15491 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
15495 Insert 2 different transparent PNG logos (second logo on bottom
15496 right corner) using the @command{ffmpeg} tool:
15498 ffmpeg -i input -i logo1 -i logo2 -filter_complex 'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
15502 Add a transparent color layer on top of the main video; @code{WxH}
15503 must specify the size of the main input to the overlay filter:
15505 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
15509 Play an original video and a filtered version (here with the deshake
15510 filter) side by side using the @command{ffplay} tool:
15512 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
15515 The above command is the same as:
15517 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
15521 Make a sliding overlay appearing from the left to the right top part of the
15522 screen starting since time 2:
15524 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
15528 Compose output by putting two input videos side to side:
15530 ffmpeg -i left.avi -i right.avi -filter_complex "
15531 nullsrc=size=200x100 [background];
15532 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
15533 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
15534 [background][left] overlay=shortest=1 [background+left];
15535 [background+left][right] overlay=shortest=1:x=100 [left+right]
15540 Mask 10-20 seconds of a video by applying the delogo filter to a section
15542 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
15543 -vf '[in]split[split_main][split_delogo];[split_delogo]trim=start=360:end=371,delogo=0:0:640:480[delogoed];[split_main][delogoed]overlay=eof_action=pass[out]'
15548 Chain several overlays in cascade:
15550 nullsrc=s=200x200 [bg];
15551 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
15552 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
15553 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
15554 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
15555 [in3] null, [mid2] overlay=100:100 [out0]
15560 @anchor{overlay_cuda}
15561 @section overlay_cuda
15563 Overlay one video on top of another.
15565 This is the CUDA variant of the @ref{overlay} filter.
15566 It only accepts CUDA frames. The underlying input pixel formats have to match.
15568 It takes two inputs and has one output. The first input is the "main"
15569 video on which the second input is overlaid.
15571 It accepts the following parameters:
15576 Set the x and y coordinates of the overlaid video on the main video.
15577 Default value is "0" for both expressions.
15580 See @ref{framesync}.
15583 See @ref{framesync}.
15586 See @ref{framesync}.
15590 This filter also supports the @ref{framesync} options.
15594 Apply Overcomplete Wavelet denoiser.
15596 The filter accepts the following options:
15602 Larger depth values will denoise lower frequency components more, but
15603 slow down filtering.
15605 Must be an int in the range 8-16, default is @code{8}.
15607 @item luma_strength, ls
15610 Must be a double value in the range 0-1000, default is @code{1.0}.
15612 @item chroma_strength, cs
15613 Set chroma strength.
15615 Must be a double value in the range 0-1000, default is @code{1.0}.
15621 Add paddings to the input image, and place the original input at the
15622 provided @var{x}, @var{y} coordinates.
15624 It accepts the following parameters:
15629 Specify an expression for the size of the output image with the
15630 paddings added. If the value for @var{width} or @var{height} is 0, the
15631 corresponding input size is used for the output.
15633 The @var{width} expression can reference the value set by the
15634 @var{height} expression, and vice versa.
15636 The default value of @var{width} and @var{height} is 0.
15640 Specify the offsets to place the input image at within the padded area,
15641 with respect to the top/left border of the output image.
15643 The @var{x} expression can reference the value set by the @var{y}
15644 expression, and vice versa.
15646 The default value of @var{x} and @var{y} is 0.
15648 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
15649 so the input image is centered on the padded area.
15652 Specify the color of the padded area. For the syntax of this option,
15653 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15654 manual,ffmpeg-utils}.
15656 The default value of @var{color} is "black".
15659 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
15661 It accepts the following values:
15665 Only evaluate expressions once during the filter initialization or when
15666 a command is processed.
15669 Evaluate expressions for each incoming frame.
15673 Default value is @samp{init}.
15676 Pad to aspect instead to a resolution.
15680 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
15681 options are expressions containing the following constants:
15686 The input video width and height.
15690 These are the same as @var{in_w} and @var{in_h}.
15694 The output width and height (the size of the padded area), as
15695 specified by the @var{width} and @var{height} expressions.
15699 These are the same as @var{out_w} and @var{out_h}.
15703 The x and y offsets as specified by the @var{x} and @var{y}
15704 expressions, or NAN if not yet specified.
15707 same as @var{iw} / @var{ih}
15710 input sample aspect ratio
15713 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
15717 The horizontal and vertical chroma subsample values. For example for the
15718 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15721 @subsection Examples
15725 Add paddings with the color "violet" to the input video. The output video
15726 size is 640x480, and the top-left corner of the input video is placed at
15729 pad=640:480:0:40:violet
15732 The example above is equivalent to the following command:
15734 pad=width=640:height=480:x=0:y=40:color=violet
15738 Pad the input to get an output with dimensions increased by 3/2,
15739 and put the input video at the center of the padded area:
15741 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
15745 Pad the input to get a squared output with size equal to the maximum
15746 value between the input width and height, and put the input video at
15747 the center of the padded area:
15749 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
15753 Pad the input to get a final w/h ratio of 16:9:
15755 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
15759 In case of anamorphic video, in order to set the output display aspect
15760 correctly, it is necessary to use @var{sar} in the expression,
15761 according to the relation:
15763 (ih * X / ih) * sar = output_dar
15764 X = output_dar / sar
15767 Thus the previous example needs to be modified to:
15769 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
15773 Double the output size and put the input video in the bottom-right
15774 corner of the output padded area:
15776 pad="2*iw:2*ih:ow-iw:oh-ih"
15780 @anchor{palettegen}
15781 @section palettegen
15783 Generate one palette for a whole video stream.
15785 It accepts the following options:
15789 Set the maximum number of colors to quantize in the palette.
15790 Note: the palette will still contain 256 colors; the unused palette entries
15793 @item reserve_transparent
15794 Create a palette of 255 colors maximum and reserve the last one for
15795 transparency. Reserving the transparency color is useful for GIF optimization.
15796 If not set, the maximum of colors in the palette will be 256. You probably want
15797 to disable this option for a standalone image.
15800 @item transparency_color
15801 Set the color that will be used as background for transparency.
15804 Set statistics mode.
15806 It accepts the following values:
15809 Compute full frame histograms.
15811 Compute histograms only for the part that differs from previous frame. This
15812 might be relevant to give more importance to the moving part of your input if
15813 the background is static.
15815 Compute new histogram for each frame.
15818 Default value is @var{full}.
15821 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
15822 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
15823 color quantization of the palette. This information is also visible at
15824 @var{info} logging level.
15826 @subsection Examples
15830 Generate a representative palette of a given video using @command{ffmpeg}:
15832 ffmpeg -i input.mkv -vf palettegen palette.png
15836 @section paletteuse
15838 Use a palette to downsample an input video stream.
15840 The filter takes two inputs: one video stream and a palette. The palette must
15841 be a 256 pixels image.
15843 It accepts the following options:
15847 Select dithering mode. Available algorithms are:
15850 Ordered 8x8 bayer dithering (deterministic)
15852 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
15853 Note: this dithering is sometimes considered "wrong" and is included as a
15855 @item floyd_steinberg
15856 Floyd and Steingberg dithering (error diffusion)
15858 Frankie Sierra dithering v2 (error diffusion)
15860 Frankie Sierra dithering v2 "Lite" (error diffusion)
15863 Default is @var{sierra2_4a}.
15866 When @var{bayer} dithering is selected, this option defines the scale of the
15867 pattern (how much the crosshatch pattern is visible). A low value means more
15868 visible pattern for less banding, and higher value means less visible pattern
15869 at the cost of more banding.
15871 The option must be an integer value in the range [0,5]. Default is @var{2}.
15874 If set, define the zone to process
15878 Only the changing rectangle will be reprocessed. This is similar to GIF
15879 cropping/offsetting compression mechanism. This option can be useful for speed
15880 if only a part of the image is changing, and has use cases such as limiting the
15881 scope of the error diffusal @option{dither} to the rectangle that bounds the
15882 moving scene (it leads to more deterministic output if the scene doesn't change
15883 much, and as a result less moving noise and better GIF compression).
15886 Default is @var{none}.
15889 Take new palette for each output frame.
15891 @item alpha_threshold
15892 Sets the alpha threshold for transparency. Alpha values above this threshold
15893 will be treated as completely opaque, and values below this threshold will be
15894 treated as completely transparent.
15896 The option must be an integer value in the range [0,255]. Default is @var{128}.
15899 @subsection Examples
15903 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
15904 using @command{ffmpeg}:
15906 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
15910 @section perspective
15912 Correct perspective of video not recorded perpendicular to the screen.
15914 A description of the accepted parameters follows.
15925 Set coordinates expression for top left, top right, bottom left and bottom right corners.
15926 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
15927 If the @code{sense} option is set to @code{source}, then the specified points will be sent
15928 to the corners of the destination. If the @code{sense} option is set to @code{destination},
15929 then the corners of the source will be sent to the specified coordinates.
15931 The expressions can use the following variables:
15936 the width and height of video frame.
15940 Output frame count.
15943 @item interpolation
15944 Set interpolation for perspective correction.
15946 It accepts the following values:
15952 Default value is @samp{linear}.
15955 Set interpretation of coordinate options.
15957 It accepts the following values:
15961 Send point in the source specified by the given coordinates to
15962 the corners of the destination.
15964 @item 1, destination
15966 Send the corners of the source to the point in the destination specified
15967 by the given coordinates.
15969 Default value is @samp{source}.
15973 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
15975 It accepts the following values:
15978 only evaluate expressions once during the filter initialization or
15979 when a command is processed
15982 evaluate expressions for each incoming frame
15985 Default value is @samp{init}.
15990 Delay interlaced video by one field time so that the field order changes.
15992 The intended use is to fix PAL movies that have been captured with the
15993 opposite field order to the film-to-video transfer.
15995 A description of the accepted parameters follows.
16001 It accepts the following values:
16004 Capture field order top-first, transfer bottom-first.
16005 Filter will delay the bottom field.
16008 Capture field order bottom-first, transfer top-first.
16009 Filter will delay the top field.
16012 Capture and transfer with the same field order. This mode only exists
16013 for the documentation of the other options to refer to, but if you
16014 actually select it, the filter will faithfully do nothing.
16017 Capture field order determined automatically by field flags, transfer
16019 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
16020 basis using field flags. If no field information is available,
16021 then this works just like @samp{u}.
16024 Capture unknown or varying, transfer opposite.
16025 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
16026 analyzing the images and selecting the alternative that produces best
16027 match between the fields.
16030 Capture top-first, transfer unknown or varying.
16031 Filter selects among @samp{t} and @samp{p} using image analysis.
16034 Capture bottom-first, transfer unknown or varying.
16035 Filter selects among @samp{b} and @samp{p} using image analysis.
16038 Capture determined by field flags, transfer unknown or varying.
16039 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
16040 image analysis. If no field information is available, then this works just
16041 like @samp{U}. This is the default mode.
16044 Both capture and transfer unknown or varying.
16045 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
16049 @subsection Commands
16051 This filter supports the all above options as @ref{commands}.
16053 @section photosensitivity
16054 Reduce various flashes in video, so to help users with epilepsy.
16056 It accepts the following options:
16059 Set how many frames to use when filtering. Default is 30.
16062 Set detection threshold factor. Default is 1.
16066 Set how many pixels to skip when sampling frames. Default is 1.
16067 Allowed range is from 1 to 1024.
16070 Leave frames unchanged. Default is disabled.
16073 @section pixdesctest
16075 Pixel format descriptor test filter, mainly useful for internal
16076 testing. The output video should be equal to the input video.
16080 format=monow, pixdesctest
16083 can be used to test the monowhite pixel format descriptor definition.
16087 Display sample values of color channels. Mainly useful for checking color
16088 and levels. Minimum supported resolution is 640x480.
16090 The filters accept the following options:
16094 Set scope X position, relative offset on X axis.
16097 Set scope Y position, relative offset on Y axis.
16106 Set window opacity. This window also holds statistics about pixel area.
16109 Set window X position, relative offset on X axis.
16112 Set window Y position, relative offset on Y axis.
16117 Enable the specified chain of postprocessing subfilters using libpostproc. This
16118 library should be automatically selected with a GPL build (@code{--enable-gpl}).
16119 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
16120 Each subfilter and some options have a short and a long name that can be used
16121 interchangeably, i.e. dr/dering are the same.
16123 The filters accept the following options:
16127 Set postprocessing subfilters string.
16130 All subfilters share common options to determine their scope:
16134 Honor the quality commands for this subfilter.
16137 Do chrominance filtering, too (default).
16140 Do luminance filtering only (no chrominance).
16143 Do chrominance filtering only (no luminance).
16146 These options can be appended after the subfilter name, separated by a '|'.
16148 Available subfilters are:
16151 @item hb/hdeblock[|difference[|flatness]]
16152 Horizontal deblocking filter
16155 Difference factor where higher values mean more deblocking (default: @code{32}).
16157 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16160 @item vb/vdeblock[|difference[|flatness]]
16161 Vertical deblocking filter
16164 Difference factor where higher values mean more deblocking (default: @code{32}).
16166 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16169 @item ha/hadeblock[|difference[|flatness]]
16170 Accurate horizontal deblocking filter
16173 Difference factor where higher values mean more deblocking (default: @code{32}).
16175 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16178 @item va/vadeblock[|difference[|flatness]]
16179 Accurate vertical deblocking filter
16182 Difference factor where higher values mean more deblocking (default: @code{32}).
16184 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16188 The horizontal and vertical deblocking filters share the difference and
16189 flatness values so you cannot set different horizontal and vertical
16193 @item h1/x1hdeblock
16194 Experimental horizontal deblocking filter
16196 @item v1/x1vdeblock
16197 Experimental vertical deblocking filter
16202 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
16205 larger -> stronger filtering
16207 larger -> stronger filtering
16209 larger -> stronger filtering
16212 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
16215 Stretch luminance to @code{0-255}.
16218 @item lb/linblenddeint
16219 Linear blend deinterlacing filter that deinterlaces the given block by
16220 filtering all lines with a @code{(1 2 1)} filter.
16222 @item li/linipoldeint
16223 Linear interpolating deinterlacing filter that deinterlaces the given block by
16224 linearly interpolating every second line.
16226 @item ci/cubicipoldeint
16227 Cubic interpolating deinterlacing filter deinterlaces the given block by
16228 cubically interpolating every second line.
16230 @item md/mediandeint
16231 Median deinterlacing filter that deinterlaces the given block by applying a
16232 median filter to every second line.
16234 @item fd/ffmpegdeint
16235 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
16236 second line with a @code{(-1 4 2 4 -1)} filter.
16239 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
16240 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
16242 @item fq/forceQuant[|quantizer]
16243 Overrides the quantizer table from the input with the constant quantizer you
16251 Default pp filter combination (@code{hb|a,vb|a,dr|a})
16254 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
16257 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
16260 @subsection Examples
16264 Apply horizontal and vertical deblocking, deringing and automatic
16265 brightness/contrast:
16271 Apply default filters without brightness/contrast correction:
16277 Apply default filters and temporal denoiser:
16279 pp=default/tmpnoise|1|2|3
16283 Apply deblocking on luminance only, and switch vertical deblocking on or off
16284 automatically depending on available CPU time:
16291 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
16292 similar to spp = 6 with 7 point DCT, where only the center sample is
16295 The filter accepts the following options:
16299 Force a constant quantization parameter. It accepts an integer in range
16300 0 to 63. If not set, the filter will use the QP from the video stream
16304 Set thresholding mode. Available modes are:
16308 Set hard thresholding.
16310 Set soft thresholding (better de-ringing effect, but likely blurrier).
16312 Set medium thresholding (good results, default).
16316 @section premultiply
16317 Apply alpha premultiply effect to input video stream using first plane
16318 of second stream as alpha.
16320 Both streams must have same dimensions and same pixel format.
16322 The filter accepts the following option:
16326 Set which planes will be processed, unprocessed planes will be copied.
16327 By default value 0xf, all planes will be processed.
16330 Do not require 2nd input for processing, instead use alpha plane from input stream.
16334 Apply prewitt operator to input video stream.
16336 The filter accepts the following option:
16340 Set which planes will be processed, unprocessed planes will be copied.
16341 By default value 0xf, all planes will be processed.
16344 Set value which will be multiplied with filtered result.
16347 Set value which will be added to filtered result.
16350 @subsection Commands
16352 This filter supports the all above options as @ref{commands}.
16354 @section pseudocolor
16356 Alter frame colors in video with pseudocolors.
16358 This filter accepts the following options:
16362 set pixel first component expression
16365 set pixel second component expression
16368 set pixel third component expression
16371 set pixel fourth component expression, corresponds to the alpha component
16374 set component to use as base for altering colors
16377 Pick one of built-in LUTs. By default is set to none.
16393 Each of them specifies the expression to use for computing the lookup table for
16394 the corresponding pixel component values.
16396 The expressions can contain the following constants and functions:
16401 The input width and height.
16404 The input value for the pixel component.
16406 @item ymin, umin, vmin, amin
16407 The minimum allowed component value.
16409 @item ymax, umax, vmax, amax
16410 The maximum allowed component value.
16413 All expressions default to "val".
16415 @subsection Commands
16417 This filter supports the all above options as @ref{commands}.
16419 @subsection Examples
16423 Change too high luma values to gradient:
16425 pseudocolor="'if(between(val,ymax,amax),lerp(ymin,ymax,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(umax,umin,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(vmin,vmax,(val-ymax)/(amax-ymax)),-1):-1'"
16431 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
16432 Ratio) between two input videos.
16434 This filter takes in input two input videos, the first input is
16435 considered the "main" source and is passed unchanged to the
16436 output. The second input is used as a "reference" video for computing
16439 Both video inputs must have the same resolution and pixel format for
16440 this filter to work correctly. Also it assumes that both inputs
16441 have the same number of frames, which are compared one by one.
16443 The obtained average PSNR is printed through the logging system.
16445 The filter stores the accumulated MSE (mean squared error) of each
16446 frame, and at the end of the processing it is averaged across all frames
16447 equally, and the following formula is applied to obtain the PSNR:
16450 PSNR = 10*log10(MAX^2/MSE)
16453 Where MAX is the average of the maximum values of each component of the
16456 The description of the accepted parameters follows.
16459 @item stats_file, f
16460 If specified the filter will use the named file to save the PSNR of
16461 each individual frame. When filename equals "-" the data is sent to
16464 @item stats_version
16465 Specifies which version of the stats file format to use. Details of
16466 each format are written below.
16467 Default value is 1.
16469 @item stats_add_max
16470 Determines whether the max value is output to the stats log.
16471 Default value is 0.
16472 Requires stats_version >= 2. If this is set and stats_version < 2,
16473 the filter will return an error.
16476 This filter also supports the @ref{framesync} options.
16478 The file printed if @var{stats_file} is selected, contains a sequence of
16479 key/value pairs of the form @var{key}:@var{value} for each compared
16482 If a @var{stats_version} greater than 1 is specified, a header line precedes
16483 the list of per-frame-pair stats, with key value pairs following the frame
16484 format with the following parameters:
16487 @item psnr_log_version
16488 The version of the log file format. Will match @var{stats_version}.
16491 A comma separated list of the per-frame-pair parameters included in
16495 A description of each shown per-frame-pair parameter follows:
16499 sequential number of the input frame, starting from 1
16502 Mean Square Error pixel-by-pixel average difference of the compared
16503 frames, averaged over all the image components.
16505 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
16506 Mean Square Error pixel-by-pixel average difference of the compared
16507 frames for the component specified by the suffix.
16509 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
16510 Peak Signal to Noise ratio of the compared frames for the component
16511 specified by the suffix.
16513 @item max_avg, max_y, max_u, max_v
16514 Maximum allowed value for each channel, and average over all
16518 @subsection Examples
16523 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16524 [main][ref] psnr="stats_file=stats.log" [out]
16527 On this example the input file being processed is compared with the
16528 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
16529 is stored in @file{stats.log}.
16532 Another example with different containers:
16534 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]psnr" -f null -
16541 Pulldown reversal (inverse telecine) filter, capable of handling mixed
16542 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
16545 The pullup filter is designed to take advantage of future context in making
16546 its decisions. This filter is stateless in the sense that it does not lock
16547 onto a pattern to follow, but it instead looks forward to the following
16548 fields in order to identify matches and rebuild progressive frames.
16550 To produce content with an even framerate, insert the fps filter after
16551 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
16552 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
16554 The filter accepts the following options:
16561 These options set the amount of "junk" to ignore at the left, right, top, and
16562 bottom of the image, respectively. Left and right are in units of 8 pixels,
16563 while top and bottom are in units of 2 lines.
16564 The default is 8 pixels on each side.
16567 Set the strict breaks. Setting this option to 1 will reduce the chances of
16568 filter generating an occasional mismatched frame, but it may also cause an
16569 excessive number of frames to be dropped during high motion sequences.
16570 Conversely, setting it to -1 will make filter match fields more easily.
16571 This may help processing of video where there is slight blurring between
16572 the fields, but may also cause there to be interlaced frames in the output.
16573 Default value is @code{0}.
16576 Set the metric plane to use. It accepts the following values:
16582 Use chroma blue plane.
16585 Use chroma red plane.
16588 This option may be set to use chroma plane instead of the default luma plane
16589 for doing filter's computations. This may improve accuracy on very clean
16590 source material, but more likely will decrease accuracy, especially if there
16591 is chroma noise (rainbow effect) or any grayscale video.
16592 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
16593 load and make pullup usable in realtime on slow machines.
16596 For best results (without duplicated frames in the output file) it is
16597 necessary to change the output frame rate. For example, to inverse
16598 telecine NTSC input:
16600 ffmpeg -i input -vf pullup -r 24000/1001 ...
16605 Change video quantization parameters (QP).
16607 The filter accepts the following option:
16611 Set expression for quantization parameter.
16614 The expression is evaluated through the eval API and can contain, among others,
16615 the following constants:
16619 1 if index is not 129, 0 otherwise.
16622 Sequential index starting from -129 to 128.
16625 @subsection Examples
16629 Some equation like:
16637 Flush video frames from internal cache of frames into a random order.
16638 No frame is discarded.
16639 Inspired by @ref{frei0r} nervous filter.
16643 Set size in number of frames of internal cache, in range from @code{2} to
16644 @code{512}. Default is @code{30}.
16647 Set seed for random number generator, must be an integer included between
16648 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16649 less than @code{0}, the filter will try to use a good random seed on a
16653 @section readeia608
16655 Read closed captioning (EIA-608) information from the top lines of a video frame.
16657 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
16658 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
16659 with EIA-608 data (starting from 0). A description of each metadata value follows:
16662 @item lavfi.readeia608.X.cc
16663 The two bytes stored as EIA-608 data (printed in hexadecimal).
16665 @item lavfi.readeia608.X.line
16666 The number of the line on which the EIA-608 data was identified and read.
16669 This filter accepts the following options:
16673 Set the line to start scanning for EIA-608 data. Default is @code{0}.
16676 Set the line to end scanning for EIA-608 data. Default is @code{29}.
16679 Set the ratio of width reserved for sync code detection.
16680 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
16683 Enable checking the parity bit. In the event of a parity error, the filter will output
16684 @code{0x00} for that character. Default is false.
16687 Lowpass lines prior to further processing. Default is enabled.
16690 @subsection Commands
16692 This filter supports the all above options as @ref{commands}.
16694 @subsection Examples
16698 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
16700 ffprobe -f lavfi -i movie=captioned_video.mov,readeia608 -show_entries frame=pkt_pts_time:frame_tags=lavfi.readeia608.0.cc,lavfi.readeia608.1.cc -of csv
16706 Read vertical interval timecode (VITC) information from the top lines of a
16709 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
16710 timecode value, if a valid timecode has been detected. Further metadata key
16711 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
16712 timecode data has been found or not.
16714 This filter accepts the following options:
16718 Set the maximum number of lines to scan for VITC data. If the value is set to
16719 @code{-1} the full video frame is scanned. Default is @code{45}.
16722 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
16723 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
16726 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
16727 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
16730 @subsection Examples
16734 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
16735 draw @code{--:--:--:--} as a placeholder:
16737 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
16743 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
16745 Destination pixel at position (X, Y) will be picked from source (x, y) position
16746 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
16747 value for pixel will be used for destination pixel.
16749 Xmap and Ymap input video streams must be of same dimensions. Output video stream
16750 will have Xmap/Ymap video stream dimensions.
16751 Xmap and Ymap input video streams are 16bit depth, single channel.
16755 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
16756 Default is @code{color}.
16759 Specify the color of the unmapped pixels. For the syntax of this option,
16760 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16761 manual,ffmpeg-utils}. Default color is @code{black}.
16764 @section removegrain
16766 The removegrain filter is a spatial denoiser for progressive video.
16770 Set mode for the first plane.
16773 Set mode for the second plane.
16776 Set mode for the third plane.
16779 Set mode for the fourth plane.
16782 Range of mode is from 0 to 24. Description of each mode follows:
16786 Leave input plane unchanged. Default.
16789 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
16792 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
16795 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
16798 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
16799 This is equivalent to a median filter.
16802 Line-sensitive clipping giving the minimal change.
16805 Line-sensitive clipping, intermediate.
16808 Line-sensitive clipping, intermediate.
16811 Line-sensitive clipping, intermediate.
16814 Line-sensitive clipping on a line where the neighbours pixels are the closest.
16817 Replaces the target pixel with the closest neighbour.
16820 [1 2 1] horizontal and vertical kernel blur.
16826 Bob mode, interpolates top field from the line where the neighbours
16827 pixels are the closest.
16830 Bob mode, interpolates bottom field from the line where the neighbours
16831 pixels are the closest.
16834 Bob mode, interpolates top field. Same as 13 but with a more complicated
16835 interpolation formula.
16838 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
16839 interpolation formula.
16842 Clips the pixel with the minimum and maximum of respectively the maximum and
16843 minimum of each pair of opposite neighbour pixels.
16846 Line-sensitive clipping using opposite neighbours whose greatest distance from
16847 the current pixel is minimal.
16850 Replaces the pixel with the average of its 8 neighbours.
16853 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
16856 Clips pixels using the averages of opposite neighbour.
16859 Same as mode 21 but simpler and faster.
16862 Small edge and halo removal, but reputed useless.
16868 @section removelogo
16870 Suppress a TV station logo, using an image file to determine which
16871 pixels comprise the logo. It works by filling in the pixels that
16872 comprise the logo with neighboring pixels.
16874 The filter accepts the following options:
16878 Set the filter bitmap file, which can be any image format supported by
16879 libavformat. The width and height of the image file must match those of the
16880 video stream being processed.
16883 Pixels in the provided bitmap image with a value of zero are not
16884 considered part of the logo, non-zero pixels are considered part of
16885 the logo. If you use white (255) for the logo and black (0) for the
16886 rest, you will be safe. For making the filter bitmap, it is
16887 recommended to take a screen capture of a black frame with the logo
16888 visible, and then using a threshold filter followed by the erode
16889 filter once or twice.
16891 If needed, little splotches can be fixed manually. Remember that if
16892 logo pixels are not covered, the filter quality will be much
16893 reduced. Marking too many pixels as part of the logo does not hurt as
16894 much, but it will increase the amount of blurring needed to cover over
16895 the image and will destroy more information than necessary, and extra
16896 pixels will slow things down on a large logo.
16898 @section repeatfields
16900 This filter uses the repeat_field flag from the Video ES headers and hard repeats
16901 fields based on its value.
16905 Reverse a video clip.
16907 Warning: This filter requires memory to buffer the entire clip, so trimming
16910 @subsection Examples
16914 Take the first 5 seconds of a clip, and reverse it.
16921 Shift R/G/B/A pixels horizontally and/or vertically.
16923 The filter accepts the following options:
16926 Set amount to shift red horizontally.
16928 Set amount to shift red vertically.
16930 Set amount to shift green horizontally.
16932 Set amount to shift green vertically.
16934 Set amount to shift blue horizontally.
16936 Set amount to shift blue vertically.
16938 Set amount to shift alpha horizontally.
16940 Set amount to shift alpha vertically.
16942 Set edge mode, can be @var{smear}, default, or @var{warp}.
16945 @subsection Commands
16947 This filter supports the all above options as @ref{commands}.
16950 Apply roberts cross operator to input video stream.
16952 The filter accepts the following option:
16956 Set which planes will be processed, unprocessed planes will be copied.
16957 By default value 0xf, all planes will be processed.
16960 Set value which will be multiplied with filtered result.
16963 Set value which will be added to filtered result.
16966 @subsection Commands
16968 This filter supports the all above options as @ref{commands}.
16972 Rotate video by an arbitrary angle expressed in radians.
16974 The filter accepts the following options:
16976 A description of the optional parameters follows.
16979 Set an expression for the angle by which to rotate the input video
16980 clockwise, expressed as a number of radians. A negative value will
16981 result in a counter-clockwise rotation. By default it is set to "0".
16983 This expression is evaluated for each frame.
16986 Set the output width expression, default value is "iw".
16987 This expression is evaluated just once during configuration.
16990 Set the output height expression, default value is "ih".
16991 This expression is evaluated just once during configuration.
16994 Enable bilinear interpolation if set to 1, a value of 0 disables
16995 it. Default value is 1.
16998 Set the color used to fill the output area not covered by the rotated
16999 image. For the general syntax of this option, check the
17000 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17001 If the special value "none" is selected then no
17002 background is printed (useful for example if the background is never shown).
17004 Default value is "black".
17007 The expressions for the angle and the output size can contain the
17008 following constants and functions:
17012 sequential number of the input frame, starting from 0. It is always NAN
17013 before the first frame is filtered.
17016 time in seconds of the input frame, it is set to 0 when the filter is
17017 configured. It is always NAN before the first frame is filtered.
17021 horizontal and vertical chroma subsample values. For example for the
17022 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17026 the input video width and height
17030 the output width and height, that is the size of the padded area as
17031 specified by the @var{width} and @var{height} expressions
17035 the minimal width/height required for completely containing the input
17036 video rotated by @var{a} radians.
17038 These are only available when computing the @option{out_w} and
17039 @option{out_h} expressions.
17042 @subsection Examples
17046 Rotate the input by PI/6 radians clockwise:
17052 Rotate the input by PI/6 radians counter-clockwise:
17058 Rotate the input by 45 degrees clockwise:
17064 Apply a constant rotation with period T, starting from an angle of PI/3:
17066 rotate=PI/3+2*PI*t/T
17070 Make the input video rotation oscillating with a period of T
17071 seconds and an amplitude of A radians:
17073 rotate=A*sin(2*PI/T*t)
17077 Rotate the video, output size is chosen so that the whole rotating
17078 input video is always completely contained in the output:
17080 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
17084 Rotate the video, reduce the output size so that no background is ever
17087 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
17091 @subsection Commands
17093 The filter supports the following commands:
17097 Set the angle expression.
17098 The command accepts the same syntax of the corresponding option.
17100 If the specified expression is not valid, it is kept at its current
17106 Apply Shape Adaptive Blur.
17108 The filter accepts the following options:
17111 @item luma_radius, lr
17112 Set luma blur filter strength, must be a value in range 0.1-4.0, default
17113 value is 1.0. A greater value will result in a more blurred image, and
17114 in slower processing.
17116 @item luma_pre_filter_radius, lpfr
17117 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
17120 @item luma_strength, ls
17121 Set luma maximum difference between pixels to still be considered, must
17122 be a value in the 0.1-100.0 range, default value is 1.0.
17124 @item chroma_radius, cr
17125 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
17126 greater value will result in a more blurred image, and in slower
17129 @item chroma_pre_filter_radius, cpfr
17130 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
17132 @item chroma_strength, cs
17133 Set chroma maximum difference between pixels to still be considered,
17134 must be a value in the -0.9-100.0 range.
17137 Each chroma option value, if not explicitly specified, is set to the
17138 corresponding luma option value.
17143 Scale (resize) the input video, using the libswscale library.
17145 The scale filter forces the output display aspect ratio to be the same
17146 of the input, by changing the output sample aspect ratio.
17148 If the input image format is different from the format requested by
17149 the next filter, the scale filter will convert the input to the
17152 @subsection Options
17153 The filter accepts the following options, or any of the options
17154 supported by the libswscale scaler.
17156 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
17157 the complete list of scaler options.
17162 Set the output video dimension expression. Default value is the input
17165 If the @var{width} or @var{w} value is 0, the input width is used for
17166 the output. If the @var{height} or @var{h} value is 0, the input height
17167 is used for the output.
17169 If one and only one of the values is -n with n >= 1, the scale filter
17170 will use a value that maintains the aspect ratio of the input image,
17171 calculated from the other specified dimension. After that it will,
17172 however, make sure that the calculated dimension is divisible by n and
17173 adjust the value if necessary.
17175 If both values are -n with n >= 1, the behavior will be identical to
17176 both values being set to 0 as previously detailed.
17178 See below for the list of accepted constants for use in the dimension
17182 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
17186 Only evaluate expressions once during the filter initialization or when a command is processed.
17189 Evaluate expressions for each incoming frame.
17193 Default value is @samp{init}.
17197 Set the interlacing mode. It accepts the following values:
17201 Force interlaced aware scaling.
17204 Do not apply interlaced scaling.
17207 Select interlaced aware scaling depending on whether the source frames
17208 are flagged as interlaced or not.
17211 Default value is @samp{0}.
17214 Set libswscale scaling flags. See
17215 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17216 complete list of values. If not explicitly specified the filter applies
17220 @item param0, param1
17221 Set libswscale input parameters for scaling algorithms that need them. See
17222 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17223 complete documentation. If not explicitly specified the filter applies
17229 Set the video size. For the syntax of this option, check the
17230 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17232 @item in_color_matrix
17233 @item out_color_matrix
17234 Set in/output YCbCr color space type.
17236 This allows the autodetected value to be overridden as well as allows forcing
17237 a specific value used for the output and encoder.
17239 If not specified, the color space type depends on the pixel format.
17245 Choose automatically.
17248 Format conforming to International Telecommunication Union (ITU)
17249 Recommendation BT.709.
17252 Set color space conforming to the United States Federal Communications
17253 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
17258 Set color space conforming to:
17262 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
17265 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
17268 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
17273 Set color space conforming to SMPTE ST 240:1999.
17276 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
17281 Set in/output YCbCr sample range.
17283 This allows the autodetected value to be overridden as well as allows forcing
17284 a specific value used for the output and encoder. If not specified, the
17285 range depends on the pixel format. Possible values:
17289 Choose automatically.
17292 Set full range (0-255 in case of 8-bit luma).
17294 @item mpeg/limited/tv
17295 Set "MPEG" range (16-235 in case of 8-bit luma).
17298 @item force_original_aspect_ratio
17299 Enable decreasing or increasing output video width or height if necessary to
17300 keep the original aspect ratio. Possible values:
17304 Scale the video as specified and disable this feature.
17307 The output video dimensions will automatically be decreased if needed.
17310 The output video dimensions will automatically be increased if needed.
17314 One useful instance of this option is that when you know a specific device's
17315 maximum allowed resolution, you can use this to limit the output video to
17316 that, while retaining the aspect ratio. For example, device A allows
17317 1280x720 playback, and your video is 1920x800. Using this option (set it to
17318 decrease) and specifying 1280x720 to the command line makes the output
17321 Please note that this is a different thing than specifying -1 for @option{w}
17322 or @option{h}, you still need to specify the output resolution for this option
17325 @item force_divisible_by
17326 Ensures that both the output dimensions, width and height, are divisible by the
17327 given integer when used together with @option{force_original_aspect_ratio}. This
17328 works similar to using @code{-n} in the @option{w} and @option{h} options.
17330 This option respects the value set for @option{force_original_aspect_ratio},
17331 increasing or decreasing the resolution accordingly. The video's aspect ratio
17332 may be slightly modified.
17334 This option can be handy if you need to have a video fit within or exceed
17335 a defined resolution using @option{force_original_aspect_ratio} but also have
17336 encoder restrictions on width or height divisibility.
17340 The values of the @option{w} and @option{h} options are expressions
17341 containing the following constants:
17346 The input width and height
17350 These are the same as @var{in_w} and @var{in_h}.
17354 The output (scaled) width and height
17358 These are the same as @var{out_w} and @var{out_h}
17361 The same as @var{iw} / @var{ih}
17364 input sample aspect ratio
17367 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17371 horizontal and vertical input chroma subsample values. For example for the
17372 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17376 horizontal and vertical output chroma subsample values. For example for the
17377 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17380 The (sequential) number of the input frame, starting from 0.
17381 Only available with @code{eval=frame}.
17384 The presentation timestamp of the input frame, expressed as a number of
17385 seconds. Only available with @code{eval=frame}.
17388 The position (byte offset) of the frame in the input stream, or NaN if
17389 this information is unavailable and/or meaningless (for example in case of synthetic video).
17390 Only available with @code{eval=frame}.
17393 @subsection Examples
17397 Scale the input video to a size of 200x100
17402 This is equivalent to:
17413 Specify a size abbreviation for the output size:
17418 which can also be written as:
17424 Scale the input to 2x:
17426 scale=w=2*iw:h=2*ih
17430 The above is the same as:
17432 scale=2*in_w:2*in_h
17436 Scale the input to 2x with forced interlaced scaling:
17438 scale=2*iw:2*ih:interl=1
17442 Scale the input to half size:
17444 scale=w=iw/2:h=ih/2
17448 Increase the width, and set the height to the same size:
17454 Seek Greek harmony:
17461 Increase the height, and set the width to 3/2 of the height:
17463 scale=w=3/2*oh:h=3/5*ih
17467 Increase the size, making the size a multiple of the chroma
17470 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
17474 Increase the width to a maximum of 500 pixels,
17475 keeping the same aspect ratio as the input:
17477 scale=w='min(500\, iw*3/2):h=-1'
17481 Make pixels square by combining scale and setsar:
17483 scale='trunc(ih*dar):ih',setsar=1/1
17487 Make pixels square by combining scale and setsar,
17488 making sure the resulting resolution is even (required by some codecs):
17490 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
17494 @subsection Commands
17496 This filter supports the following commands:
17500 Set the output video dimension expression.
17501 The command accepts the same syntax of the corresponding option.
17503 If the specified expression is not valid, it is kept at its current
17509 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
17510 format conversion on CUDA video frames. Setting the output width and height
17511 works in the same way as for the @var{scale} filter.
17513 The following additional options are accepted:
17516 The pixel format of the output CUDA frames. If set to the string "same" (the
17517 default), the input format will be kept. Note that automatic format negotiation
17518 and conversion is not yet supported for hardware frames
17521 The interpolation algorithm used for resizing. One of the following:
17528 @item cubic2p_bspline
17529 2-parameter cubic (B=1, C=0)
17531 @item cubic2p_catmullrom
17532 2-parameter cubic (B=0, C=1/2)
17534 @item cubic2p_b05c03
17535 2-parameter cubic (B=1/2, C=3/10)
17543 @item force_original_aspect_ratio
17544 Enable decreasing or increasing output video width or height if necessary to
17545 keep the original aspect ratio. Possible values:
17549 Scale the video as specified and disable this feature.
17552 The output video dimensions will automatically be decreased if needed.
17555 The output video dimensions will automatically be increased if needed.
17559 One useful instance of this option is that when you know a specific device's
17560 maximum allowed resolution, you can use this to limit the output video to
17561 that, while retaining the aspect ratio. For example, device A allows
17562 1280x720 playback, and your video is 1920x800. Using this option (set it to
17563 decrease) and specifying 1280x720 to the command line makes the output
17566 Please note that this is a different thing than specifying -1 for @option{w}
17567 or @option{h}, you still need to specify the output resolution for this option
17570 @item force_divisible_by
17571 Ensures that both the output dimensions, width and height, are divisible by the
17572 given integer when used together with @option{force_original_aspect_ratio}. This
17573 works similar to using @code{-n} in the @option{w} and @option{h} options.
17575 This option respects the value set for @option{force_original_aspect_ratio},
17576 increasing or decreasing the resolution accordingly. The video's aspect ratio
17577 may be slightly modified.
17579 This option can be handy if you need to have a video fit within or exceed
17580 a defined resolution using @option{force_original_aspect_ratio} but also have
17581 encoder restrictions on width or height divisibility.
17587 Scale (resize) the input video, based on a reference video.
17589 See the scale filter for available options, scale2ref supports the same but
17590 uses the reference video instead of the main input as basis. scale2ref also
17591 supports the following additional constants for the @option{w} and
17592 @option{h} options:
17597 The main input video's width and height
17600 The same as @var{main_w} / @var{main_h}
17603 The main input video's sample aspect ratio
17605 @item main_dar, mdar
17606 The main input video's display aspect ratio. Calculated from
17607 @code{(main_w / main_h) * main_sar}.
17611 The main input video's horizontal and vertical chroma subsample values.
17612 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
17616 The (sequential) number of the main input frame, starting from 0.
17617 Only available with @code{eval=frame}.
17620 The presentation timestamp of the main input frame, expressed as a number of
17621 seconds. Only available with @code{eval=frame}.
17624 The position (byte offset) of the frame in the main input stream, or NaN if
17625 this information is unavailable and/or meaningless (for example in case of synthetic video).
17626 Only available with @code{eval=frame}.
17629 @subsection Examples
17633 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
17635 'scale2ref[b][a];[a][b]overlay'
17639 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
17641 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
17645 @subsection Commands
17647 This filter supports the following commands:
17651 Set the output video dimension expression.
17652 The command accepts the same syntax of the corresponding option.
17654 If the specified expression is not valid, it is kept at its current
17659 Scroll input video horizontally and/or vertically by constant speed.
17661 The filter accepts the following options:
17663 @item horizontal, h
17664 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
17665 Negative values changes scrolling direction.
17668 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
17669 Negative values changes scrolling direction.
17672 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
17675 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
17678 @subsection Commands
17680 This filter supports the following @ref{commands}:
17682 @item horizontal, h
17683 Set the horizontal scrolling speed.
17685 Set the vertical scrolling speed.
17691 Detect video scene change.
17693 This filter sets frame metadata with mafd between frame, the scene score, and
17694 forward the frame to the next filter, so they can use these metadata to detect
17695 scene change or others.
17697 In addition, this filter logs a message and sets frame metadata when it detects
17698 a scene change by @option{threshold}.
17700 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
17702 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
17703 to detect scene change.
17705 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
17706 detect scene change with @option{threshold}.
17708 The filter accepts the following options:
17712 Set the scene change detection threshold as a percentage of maximum change. Good
17713 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
17716 Default value is @code{10.}.
17719 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
17720 You can enable it if you want to get snapshot of scene change frames only.
17723 @anchor{selectivecolor}
17724 @section selectivecolor
17726 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
17727 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
17728 by the "purity" of the color (that is, how saturated it already is).
17730 This filter is similar to the Adobe Photoshop Selective Color tool.
17732 The filter accepts the following options:
17735 @item correction_method
17736 Select color correction method.
17738 Available values are:
17741 Specified adjustments are applied "as-is" (added/subtracted to original pixel
17744 Specified adjustments are relative to the original component value.
17746 Default is @code{absolute}.
17748 Adjustments for red pixels (pixels where the red component is the maximum)
17750 Adjustments for yellow pixels (pixels where the blue component is the minimum)
17752 Adjustments for green pixels (pixels where the green component is the maximum)
17754 Adjustments for cyan pixels (pixels where the red component is the minimum)
17756 Adjustments for blue pixels (pixels where the blue component is the maximum)
17758 Adjustments for magenta pixels (pixels where the green component is the minimum)
17760 Adjustments for white pixels (pixels where all components are greater than 128)
17762 Adjustments for all pixels except pure black and pure white
17764 Adjustments for black pixels (pixels where all components are lesser than 128)
17766 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
17769 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
17770 4 space separated floating point adjustment values in the [-1,1] range,
17771 respectively to adjust the amount of cyan, magenta, yellow and black for the
17772 pixels of its range.
17774 @subsection Examples
17778 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
17779 increase magenta by 27% in blue areas:
17781 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
17785 Use a Photoshop selective color preset:
17787 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
17791 @anchor{separatefields}
17792 @section separatefields
17794 The @code{separatefields} takes a frame-based video input and splits
17795 each frame into its components fields, producing a new half height clip
17796 with twice the frame rate and twice the frame count.
17798 This filter use field-dominance information in frame to decide which
17799 of each pair of fields to place first in the output.
17800 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
17802 @section setdar, setsar
17804 The @code{setdar} filter sets the Display Aspect Ratio for the filter
17807 This is done by changing the specified Sample (aka Pixel) Aspect
17808 Ratio, according to the following equation:
17810 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
17813 Keep in mind that the @code{setdar} filter does not modify the pixel
17814 dimensions of the video frame. Also, the display aspect ratio set by
17815 this filter may be changed by later filters in the filterchain,
17816 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
17819 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
17820 the filter output video.
17822 Note that as a consequence of the application of this filter, the
17823 output display aspect ratio will change according to the equation
17826 Keep in mind that the sample aspect ratio set by the @code{setsar}
17827 filter may be changed by later filters in the filterchain, e.g. if
17828 another "setsar" or a "setdar" filter is applied.
17830 It accepts the following parameters:
17833 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
17834 Set the aspect ratio used by the filter.
17836 The parameter can be a floating point number string, an expression, or
17837 a string of the form @var{num}:@var{den}, where @var{num} and
17838 @var{den} are the numerator and denominator of the aspect ratio. If
17839 the parameter is not specified, it is assumed the value "0".
17840 In case the form "@var{num}:@var{den}" is used, the @code{:} character
17844 Set the maximum integer value to use for expressing numerator and
17845 denominator when reducing the expressed aspect ratio to a rational.
17846 Default value is @code{100}.
17850 The parameter @var{sar} is an expression containing
17851 the following constants:
17855 These are approximated values for the mathematical constants e
17856 (Euler's number), pi (Greek pi), and phi (the golden ratio).
17859 The input width and height.
17862 These are the same as @var{w} / @var{h}.
17865 The input sample aspect ratio.
17868 The input display aspect ratio. It is the same as
17869 (@var{w} / @var{h}) * @var{sar}.
17872 Horizontal and vertical chroma subsample values. For example, for the
17873 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17876 @subsection Examples
17881 To change the display aspect ratio to 16:9, specify one of the following:
17888 To change the sample aspect ratio to 10:11, specify:
17894 To set a display aspect ratio of 16:9, and specify a maximum integer value of
17895 1000 in the aspect ratio reduction, use the command:
17897 setdar=ratio=16/9:max=1000
17905 Force field for the output video frame.
17907 The @code{setfield} filter marks the interlace type field for the
17908 output frames. It does not change the input frame, but only sets the
17909 corresponding property, which affects how the frame is treated by
17910 following filters (e.g. @code{fieldorder} or @code{yadif}).
17912 The filter accepts the following options:
17917 Available values are:
17921 Keep the same field property.
17924 Mark the frame as bottom-field-first.
17927 Mark the frame as top-field-first.
17930 Mark the frame as progressive.
17937 Force frame parameter for the output video frame.
17939 The @code{setparams} filter marks interlace and color range for the
17940 output frames. It does not change the input frame, but only sets the
17941 corresponding property, which affects how the frame is treated by
17946 Available values are:
17950 Keep the same field property (default).
17953 Mark the frame as bottom-field-first.
17956 Mark the frame as top-field-first.
17959 Mark the frame as progressive.
17963 Available values are:
17967 Keep the same color range property (default).
17969 @item unspecified, unknown
17970 Mark the frame as unspecified color range.
17972 @item limited, tv, mpeg
17973 Mark the frame as limited range.
17975 @item full, pc, jpeg
17976 Mark the frame as full range.
17979 @item color_primaries
17980 Set the color primaries.
17981 Available values are:
17985 Keep the same color primaries property (default).
18002 Set the color transfer.
18003 Available values are:
18007 Keep the same color trc property (default).
18029 Set the colorspace.
18030 Available values are:
18034 Keep the same colorspace property (default).
18047 @item chroma-derived-nc
18048 @item chroma-derived-c
18054 Apply shear transform to input video.
18056 This filter supports the following options:
18060 Shear factor in X-direction. Default value is 0.
18061 Allowed range is from -2 to 2.
18064 Shear factor in Y-direction. Default value is 0.
18065 Allowed range is from -2 to 2.
18068 Set the color used to fill the output area not covered by the transformed
18069 video. For the general syntax of this option, check the
18070 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18071 If the special value "none" is selected then no
18072 background is printed (useful for example if the background is never shown).
18074 Default value is "black".
18077 Set interpolation type. Can be @code{bilinear} or @code{nearest}. Default is @code{bilinear}.
18080 @subsection Commands
18082 This filter supports the all above options as @ref{commands}.
18086 Show a line containing various information for each input video frame.
18087 The input video is not modified.
18089 This filter supports the following options:
18093 Calculate checksums of each plane. By default enabled.
18096 The shown line contains a sequence of key/value pairs of the form
18097 @var{key}:@var{value}.
18099 The following values are shown in the output:
18103 The (sequential) number of the input frame, starting from 0.
18106 The Presentation TimeStamp of the input frame, expressed as a number of
18107 time base units. The time base unit depends on the filter input pad.
18110 The Presentation TimeStamp of the input frame, expressed as a number of
18114 The position of the frame in the input stream, or -1 if this information is
18115 unavailable and/or meaningless (for example in case of synthetic video).
18118 The pixel format name.
18121 The sample aspect ratio of the input frame, expressed in the form
18122 @var{num}/@var{den}.
18125 The size of the input frame. For the syntax of this option, check the
18126 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18129 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
18130 for bottom field first).
18133 This is 1 if the frame is a key frame, 0 otherwise.
18136 The picture type of the input frame ("I" for an I-frame, "P" for a
18137 P-frame, "B" for a B-frame, or "?" for an unknown type).
18138 Also refer to the documentation of the @code{AVPictureType} enum and of
18139 the @code{av_get_picture_type_char} function defined in
18140 @file{libavutil/avutil.h}.
18143 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
18145 @item plane_checksum
18146 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
18147 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
18150 The mean value of pixels in each plane of the input frame, expressed in the form
18151 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
18154 The standard deviation of pixel values in each plane of the input frame, expressed
18155 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
18159 @section showpalette
18161 Displays the 256 colors palette of each frame. This filter is only relevant for
18162 @var{pal8} pixel format frames.
18164 It accepts the following option:
18168 Set the size of the box used to represent one palette color entry. Default is
18169 @code{30} (for a @code{30x30} pixel box).
18172 @section shuffleframes
18174 Reorder and/or duplicate and/or drop video frames.
18176 It accepts the following parameters:
18180 Set the destination indexes of input frames.
18181 This is space or '|' separated list of indexes that maps input frames to output
18182 frames. Number of indexes also sets maximal value that each index may have.
18183 '-1' index have special meaning and that is to drop frame.
18186 The first frame has the index 0. The default is to keep the input unchanged.
18188 @subsection Examples
18192 Swap second and third frame of every three frames of the input:
18194 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
18198 Swap 10th and 1st frame of every ten frames of the input:
18200 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
18204 @section shufflepixels
18206 Reorder pixels in video frames.
18208 This filter accepts the following options:
18212 Set shuffle direction. Can be forward or inverse direction.
18213 Default direction is forward.
18216 Set shuffle mode. Can be horizontal, vertical or block mode.
18220 Set shuffle block_size. In case of horizontal shuffle mode only width
18221 part of size is used, and in case of vertical shuffle mode only height
18222 part of size is used.
18225 Set random seed used with shuffling pixels. Mainly useful to set to be able
18226 to reverse filtering process to get original input.
18227 For example, to reverse forward shuffle you need to use same parameters
18228 and exact same seed and to set direction to inverse.
18231 @section shuffleplanes
18233 Reorder and/or duplicate video planes.
18235 It accepts the following parameters:
18240 The index of the input plane to be used as the first output plane.
18243 The index of the input plane to be used as the second output plane.
18246 The index of the input plane to be used as the third output plane.
18249 The index of the input plane to be used as the fourth output plane.
18253 The first plane has the index 0. The default is to keep the input unchanged.
18255 @subsection Examples
18259 Swap the second and third planes of the input:
18261 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
18265 @anchor{signalstats}
18266 @section signalstats
18267 Evaluate various visual metrics that assist in determining issues associated
18268 with the digitization of analog video media.
18270 By default the filter will log these metadata values:
18274 Display the minimal Y value contained within the input frame. Expressed in
18278 Display the Y value at the 10% percentile within the input frame. Expressed in
18282 Display the average Y value within the input frame. Expressed in range of
18286 Display the Y value at the 90% percentile within the input frame. Expressed in
18290 Display the maximum Y value contained within the input frame. Expressed in
18294 Display the minimal U value contained within the input frame. Expressed in
18298 Display the U value at the 10% percentile within the input frame. Expressed in
18302 Display the average U value within the input frame. Expressed in range of
18306 Display the U value at the 90% percentile within the input frame. Expressed in
18310 Display the maximum U value contained within the input frame. Expressed in
18314 Display the minimal V value contained within the input frame. Expressed in
18318 Display the V value at the 10% percentile within the input frame. Expressed in
18322 Display the average V value within the input frame. Expressed in range of
18326 Display the V value at the 90% percentile within the input frame. Expressed in
18330 Display the maximum V value contained within the input frame. Expressed in
18334 Display the minimal saturation value contained within the input frame.
18335 Expressed in range of [0-~181.02].
18338 Display the saturation value at the 10% percentile within the input frame.
18339 Expressed in range of [0-~181.02].
18342 Display the average saturation value within the input frame. Expressed in range
18346 Display the saturation value at the 90% percentile within the input frame.
18347 Expressed in range of [0-~181.02].
18350 Display the maximum saturation value contained within the input frame.
18351 Expressed in range of [0-~181.02].
18354 Display the median value for hue within the input frame. Expressed in range of
18358 Display the average value for hue within the input frame. Expressed in range of
18362 Display the average of sample value difference between all values of the Y
18363 plane in the current frame and corresponding values of the previous input frame.
18364 Expressed in range of [0-255].
18367 Display the average of sample value difference between all values of the U
18368 plane in the current frame and corresponding values of the previous input frame.
18369 Expressed in range of [0-255].
18372 Display the average of sample value difference between all values of the V
18373 plane in the current frame and corresponding values of the previous input frame.
18374 Expressed in range of [0-255].
18377 Display bit depth of Y plane in current frame.
18378 Expressed in range of [0-16].
18381 Display bit depth of U plane in current frame.
18382 Expressed in range of [0-16].
18385 Display bit depth of V plane in current frame.
18386 Expressed in range of [0-16].
18389 The filter accepts the following options:
18395 @option{stat} specify an additional form of image analysis.
18396 @option{out} output video with the specified type of pixel highlighted.
18398 Both options accept the following values:
18402 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
18403 unlike the neighboring pixels of the same field. Examples of temporal outliers
18404 include the results of video dropouts, head clogs, or tape tracking issues.
18407 Identify @var{vertical line repetition}. Vertical line repetition includes
18408 similar rows of pixels within a frame. In born-digital video vertical line
18409 repetition is common, but this pattern is uncommon in video digitized from an
18410 analog source. When it occurs in video that results from the digitization of an
18411 analog source it can indicate concealment from a dropout compensator.
18414 Identify pixels that fall outside of legal broadcast range.
18418 Set the highlight color for the @option{out} option. The default color is
18422 @subsection Examples
18426 Output data of various video metrics:
18428 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
18432 Output specific data about the minimum and maximum values of the Y plane per frame:
18434 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
18438 Playback video while highlighting pixels that are outside of broadcast range in red.
18440 ffplay example.mov -vf signalstats="out=brng:color=red"
18444 Playback video with signalstats metadata drawn over the frame.
18446 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
18449 The contents of signalstat_drawtext.txt used in the command are:
18452 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
18453 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
18454 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
18455 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
18463 Calculates the MPEG-7 Video Signature. The filter can handle more than one
18464 input. In this case the matching between the inputs can be calculated additionally.
18465 The filter always passes through the first input. The signature of each stream can
18466 be written into a file.
18468 It accepts the following options:
18472 Enable or disable the matching process.
18474 Available values are:
18478 Disable the calculation of a matching (default).
18480 Calculate the matching for the whole video and output whether the whole video
18481 matches or only parts.
18483 Calculate only until a matching is found or the video ends. Should be faster in
18488 Set the number of inputs. The option value must be a non negative integer.
18489 Default value is 1.
18492 Set the path to which the output is written. If there is more than one input,
18493 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
18494 integer), that will be replaced with the input number. If no filename is
18495 specified, no output will be written. This is the default.
18498 Choose the output format.
18500 Available values are:
18504 Use the specified binary representation (default).
18506 Use the specified xml representation.
18510 Set threshold to detect one word as similar. The option value must be an integer
18511 greater than zero. The default value is 9000.
18514 Set threshold to detect all words as similar. The option value must be an integer
18515 greater than zero. The default value is 60000.
18518 Set threshold to detect frames as similar. The option value must be an integer
18519 greater than zero. The default value is 116.
18522 Set the minimum length of a sequence in frames to recognize it as matching
18523 sequence. The option value must be a non negative integer value.
18524 The default value is 0.
18527 Set the minimum relation, that matching frames to all frames must have.
18528 The option value must be a double value between 0 and 1. The default value is 0.5.
18531 @subsection Examples
18535 To calculate the signature of an input video and store it in signature.bin:
18537 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
18541 To detect whether two videos match and store the signatures in XML format in
18542 signature0.xml and signature1.xml:
18544 ffmpeg -i input1.mkv -i input2.mkv -filter_complex "[0:v][1:v] signature=nb_inputs=2:detectmode=full:format=xml:filename=signature%d.xml" -map :v -f null -
18552 Blur the input video without impacting the outlines.
18554 It accepts the following options:
18557 @item luma_radius, lr
18558 Set the luma radius. The option value must be a float number in
18559 the range [0.1,5.0] that specifies the variance of the gaussian filter
18560 used to blur the image (slower if larger). Default value is 1.0.
18562 @item luma_strength, ls
18563 Set the luma strength. The option value must be a float number
18564 in the range [-1.0,1.0] that configures the blurring. A value included
18565 in [0.0,1.0] will blur the image whereas a value included in
18566 [-1.0,0.0] will sharpen the image. Default value is 1.0.
18568 @item luma_threshold, lt
18569 Set the luma threshold used as a coefficient to determine
18570 whether a pixel should be blurred or not. The option value must be an
18571 integer in the range [-30,30]. A value of 0 will filter all the image,
18572 a value included in [0,30] will filter flat areas and a value included
18573 in [-30,0] will filter edges. Default value is 0.
18575 @item chroma_radius, cr
18576 Set the chroma radius. The option value must be a float number in
18577 the range [0.1,5.0] that specifies the variance of the gaussian filter
18578 used to blur the image (slower if larger). Default value is @option{luma_radius}.
18580 @item chroma_strength, cs
18581 Set the chroma strength. The option value must be a float number
18582 in the range [-1.0,1.0] that configures the blurring. A value included
18583 in [0.0,1.0] will blur the image whereas a value included in
18584 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
18586 @item chroma_threshold, ct
18587 Set the chroma threshold used as a coefficient to determine
18588 whether a pixel should be blurred or not. The option value must be an
18589 integer in the range [-30,30]. A value of 0 will filter all the image,
18590 a value included in [0,30] will filter flat areas and a value included
18591 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
18594 If a chroma option is not explicitly set, the corresponding luma value
18598 Apply sobel operator to input video stream.
18600 The filter accepts the following option:
18604 Set which planes will be processed, unprocessed planes will be copied.
18605 By default value 0xf, all planes will be processed.
18608 Set value which will be multiplied with filtered result.
18611 Set value which will be added to filtered result.
18614 @subsection Commands
18616 This filter supports the all above options as @ref{commands}.
18621 Apply a simple postprocessing filter that compresses and decompresses the image
18622 at several (or - in the case of @option{quality} level @code{6} - all) shifts
18623 and average the results.
18625 The filter accepts the following options:
18629 Set quality. This option defines the number of levels for averaging. It accepts
18630 an integer in the range 0-6. If set to @code{0}, the filter will have no
18631 effect. A value of @code{6} means the higher quality. For each increment of
18632 that value the speed drops by a factor of approximately 2. Default value is
18636 Force a constant quantization parameter. If not set, the filter will use the QP
18637 from the video stream (if available).
18640 Set thresholding mode. Available modes are:
18644 Set hard thresholding (default).
18646 Set soft thresholding (better de-ringing effect, but likely blurrier).
18649 @item use_bframe_qp
18650 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
18651 option may cause flicker since the B-Frames have often larger QP. Default is
18652 @code{0} (not enabled).
18655 @subsection Commands
18657 This filter supports the following commands:
18659 @item quality, level
18660 Set quality level. The value @code{max} can be used to set the maximum level,
18661 currently @code{6}.
18667 Scale the input by applying one of the super-resolution methods based on
18668 convolutional neural networks. Supported models:
18672 Super-Resolution Convolutional Neural Network model (SRCNN).
18673 See @url{https://arxiv.org/abs/1501.00092}.
18676 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
18677 See @url{https://arxiv.org/abs/1609.05158}.
18680 Training scripts as well as scripts for model file (.pb) saving can be found at
18681 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
18682 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
18684 Native model files (.model) can be generated from TensorFlow model
18685 files (.pb) by using tools/python/convert.py
18687 The filter accepts the following options:
18691 Specify which DNN backend to use for model loading and execution. This option accepts
18692 the following values:
18696 Native implementation of DNN loading and execution.
18699 TensorFlow backend. To enable this backend you
18700 need to install the TensorFlow for C library (see
18701 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
18702 @code{--enable-libtensorflow}
18705 Default value is @samp{native}.
18708 Set path to model file specifying network architecture and its parameters.
18709 Note that different backends use different file formats. TensorFlow backend
18710 can load files for both formats, while native backend can load files for only
18714 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
18715 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
18716 input upscaled using bicubic upscaling with proper scale factor.
18719 This feature can also be finished with @ref{dnn_processing} filter.
18723 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
18725 This filter takes in input two input videos, the first input is
18726 considered the "main" source and is passed unchanged to the
18727 output. The second input is used as a "reference" video for computing
18730 Both video inputs must have the same resolution and pixel format for
18731 this filter to work correctly. Also it assumes that both inputs
18732 have the same number of frames, which are compared one by one.
18734 The filter stores the calculated SSIM of each frame.
18736 The description of the accepted parameters follows.
18739 @item stats_file, f
18740 If specified the filter will use the named file to save the SSIM of
18741 each individual frame. When filename equals "-" the data is sent to
18745 The file printed if @var{stats_file} is selected, contains a sequence of
18746 key/value pairs of the form @var{key}:@var{value} for each compared
18749 A description of each shown parameter follows:
18753 sequential number of the input frame, starting from 1
18755 @item Y, U, V, R, G, B
18756 SSIM of the compared frames for the component specified by the suffix.
18759 SSIM of the compared frames for the whole frame.
18762 Same as above but in dB representation.
18765 This filter also supports the @ref{framesync} options.
18767 @subsection Examples
18772 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
18773 [main][ref] ssim="stats_file=stats.log" [out]
18776 On this example the input file being processed is compared with the
18777 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
18778 is stored in @file{stats.log}.
18781 Another example with both psnr and ssim at same time:
18783 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
18787 Another example with different containers:
18789 ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]ssim" -f null -
18795 Convert between different stereoscopic image formats.
18797 The filters accept the following options:
18801 Set stereoscopic image format of input.
18803 Available values for input image formats are:
18806 side by side parallel (left eye left, right eye right)
18809 side by side crosseye (right eye left, left eye right)
18812 side by side parallel with half width resolution
18813 (left eye left, right eye right)
18816 side by side crosseye with half width resolution
18817 (right eye left, left eye right)
18821 above-below (left eye above, right eye below)
18825 above-below (right eye above, left eye below)
18829 above-below with half height resolution
18830 (left eye above, right eye below)
18834 above-below with half height resolution
18835 (right eye above, left eye below)
18838 alternating frames (left eye first, right eye second)
18841 alternating frames (right eye first, left eye second)
18844 interleaved rows (left eye has top row, right eye starts on next row)
18847 interleaved rows (right eye has top row, left eye starts on next row)
18850 interleaved columns, left eye first
18853 interleaved columns, right eye first
18855 Default value is @samp{sbsl}.
18859 Set stereoscopic image format of output.
18863 side by side parallel (left eye left, right eye right)
18866 side by side crosseye (right eye left, left eye right)
18869 side by side parallel with half width resolution
18870 (left eye left, right eye right)
18873 side by side crosseye with half width resolution
18874 (right eye left, left eye right)
18878 above-below (left eye above, right eye below)
18882 above-below (right eye above, left eye below)
18886 above-below with half height resolution
18887 (left eye above, right eye below)
18891 above-below with half height resolution
18892 (right eye above, left eye below)
18895 alternating frames (left eye first, right eye second)
18898 alternating frames (right eye first, left eye second)
18901 interleaved rows (left eye has top row, right eye starts on next row)
18904 interleaved rows (right eye has top row, left eye starts on next row)
18907 anaglyph red/blue gray
18908 (red filter on left eye, blue filter on right eye)
18911 anaglyph red/green gray
18912 (red filter on left eye, green filter on right eye)
18915 anaglyph red/cyan gray
18916 (red filter on left eye, cyan filter on right eye)
18919 anaglyph red/cyan half colored
18920 (red filter on left eye, cyan filter on right eye)
18923 anaglyph red/cyan color
18924 (red filter on left eye, cyan filter on right eye)
18927 anaglyph red/cyan color optimized with the least squares projection of dubois
18928 (red filter on left eye, cyan filter on right eye)
18931 anaglyph green/magenta gray
18932 (green filter on left eye, magenta filter on right eye)
18935 anaglyph green/magenta half colored
18936 (green filter on left eye, magenta filter on right eye)
18939 anaglyph green/magenta colored
18940 (green filter on left eye, magenta filter on right eye)
18943 anaglyph green/magenta color optimized with the least squares projection of dubois
18944 (green filter on left eye, magenta filter on right eye)
18947 anaglyph yellow/blue gray
18948 (yellow filter on left eye, blue filter on right eye)
18951 anaglyph yellow/blue half colored
18952 (yellow filter on left eye, blue filter on right eye)
18955 anaglyph yellow/blue colored
18956 (yellow filter on left eye, blue filter on right eye)
18959 anaglyph yellow/blue color optimized with the least squares projection of dubois
18960 (yellow filter on left eye, blue filter on right eye)
18963 mono output (left eye only)
18966 mono output (right eye only)
18969 checkerboard, left eye first
18972 checkerboard, right eye first
18975 interleaved columns, left eye first
18978 interleaved columns, right eye first
18984 Default value is @samp{arcd}.
18987 @subsection Examples
18991 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
18997 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
19003 @section streamselect, astreamselect
19004 Select video or audio streams.
19006 The filter accepts the following options:
19010 Set number of inputs. Default is 2.
19013 Set input indexes to remap to outputs.
19016 @subsection Commands
19018 The @code{streamselect} and @code{astreamselect} filter supports the following
19023 Set input indexes to remap to outputs.
19026 @subsection Examples
19030 Select first 5 seconds 1st stream and rest of time 2nd stream:
19032 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
19036 Same as above, but for audio:
19038 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
19045 Draw subtitles on top of input video using the libass library.
19047 To enable compilation of this filter you need to configure FFmpeg with
19048 @code{--enable-libass}. This filter also requires a build with libavcodec and
19049 libavformat to convert the passed subtitles file to ASS (Advanced Substation
19050 Alpha) subtitles format.
19052 The filter accepts the following options:
19056 Set the filename of the subtitle file to read. It must be specified.
19058 @item original_size
19059 Specify the size of the original video, the video for which the ASS file
19060 was composed. For the syntax of this option, check the
19061 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19062 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
19063 correctly scale the fonts if the aspect ratio has been changed.
19066 Set a directory path containing fonts that can be used by the filter.
19067 These fonts will be used in addition to whatever the font provider uses.
19070 Process alpha channel, by default alpha channel is untouched.
19073 Set subtitles input character encoding. @code{subtitles} filter only. Only
19074 useful if not UTF-8.
19076 @item stream_index, si
19077 Set subtitles stream index. @code{subtitles} filter only.
19080 Override default style or script info parameters of the subtitles. It accepts a
19081 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
19084 If the first key is not specified, it is assumed that the first value
19085 specifies the @option{filename}.
19087 For example, to render the file @file{sub.srt} on top of the input
19088 video, use the command:
19093 which is equivalent to:
19095 subtitles=filename=sub.srt
19098 To render the default subtitles stream from file @file{video.mkv}, use:
19100 subtitles=video.mkv
19103 To render the second subtitles stream from that file, use:
19105 subtitles=video.mkv:si=1
19108 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
19109 @code{DejaVu Serif}, use:
19111 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
19114 @section super2xsai
19116 Scale the input by 2x and smooth using the Super2xSaI (Scale and
19117 Interpolate) pixel art scaling algorithm.
19119 Useful for enlarging pixel art images without reducing sharpness.
19123 Swap two rectangular objects in video.
19125 This filter accepts the following options:
19135 Set 1st rect x coordinate.
19138 Set 1st rect y coordinate.
19141 Set 2nd rect x coordinate.
19144 Set 2nd rect y coordinate.
19146 All expressions are evaluated once for each frame.
19149 The all options are expressions containing the following constants:
19154 The input width and height.
19157 same as @var{w} / @var{h}
19160 input sample aspect ratio
19163 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
19166 The number of the input frame, starting from 0.
19169 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
19172 the position in the file of the input frame, NAN if unknown
19175 @subsection Commands
19177 This filter supports the all above options as @ref{commands}.
19183 Blend successive video frames.
19189 Apply telecine process to the video.
19191 This filter accepts the following options:
19200 The default value is @code{top}.
19204 A string of numbers representing the pulldown pattern you wish to apply.
19205 The default value is @code{23}.
19209 Some typical patterns:
19214 24p: 2332 (preferred)
19221 24p: 222222222223 ("Euro pulldown")
19226 @section thistogram
19228 Compute and draw a color distribution histogram for the input video across time.
19230 Unlike @ref{histogram} video filter which only shows histogram of single input frame
19231 at certain time, this filter shows also past histograms of number of frames defined
19232 by @code{width} option.
19234 The computed histogram is a representation of the color component
19235 distribution in an image.
19237 The filter accepts the following options:
19241 Set width of single color component output. Default value is @code{0}.
19242 Value of @code{0} means width will be picked from input video.
19243 This also set number of passed histograms to keep.
19244 Allowed range is [0, 8192].
19246 @item display_mode, d
19248 It accepts the following values:
19251 Per color component graphs are placed below each other.
19254 Per color component graphs are placed side by side.
19257 Presents information identical to that in the @code{parade}, except
19258 that the graphs representing color components are superimposed directly
19261 Default is @code{stack}.
19263 @item levels_mode, m
19264 Set mode. Can be either @code{linear}, or @code{logarithmic}.
19265 Default is @code{linear}.
19267 @item components, c
19268 Set what color components to display.
19269 Default is @code{7}.
19272 Set background opacity. Default is @code{0.9}.
19275 Show envelope. Default is disabled.
19278 Set envelope color. Default is @code{gold}.
19283 Available values for slide is:
19286 Draw new frame when right border is reached.
19289 Replace old columns with new ones.
19292 Scroll from right to left.
19295 Scroll from left to right.
19298 Draw single picture.
19301 Default is @code{replace}.
19306 Apply threshold effect to video stream.
19308 This filter needs four video streams to perform thresholding.
19309 First stream is stream we are filtering.
19310 Second stream is holding threshold values, third stream is holding min values,
19311 and last, fourth stream is holding max values.
19313 The filter accepts the following option:
19317 Set which planes will be processed, unprocessed planes will be copied.
19318 By default value 0xf, all planes will be processed.
19321 For example if first stream pixel's component value is less then threshold value
19322 of pixel component from 2nd threshold stream, third stream value will picked,
19323 otherwise fourth stream pixel component value will be picked.
19325 Using color source filter one can perform various types of thresholding:
19327 @subsection Examples
19331 Binary threshold, using gray color as threshold:
19333 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
19337 Inverted binary threshold, using gray color as threshold:
19339 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
19343 Truncate binary threshold, using gray color as threshold:
19345 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
19349 Threshold to zero, using gray color as threshold:
19351 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
19355 Inverted threshold to zero, using gray color as threshold:
19357 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
19362 Select the most representative frame in a given sequence of consecutive frames.
19364 The filter accepts the following options:
19368 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
19369 will pick one of them, and then handle the next batch of @var{n} frames until
19370 the end. Default is @code{100}.
19373 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
19374 value will result in a higher memory usage, so a high value is not recommended.
19376 @subsection Examples
19380 Extract one picture each 50 frames:
19386 Complete example of a thumbnail creation with @command{ffmpeg}:
19388 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
19395 Tile several successive frames together.
19397 The @ref{untile} filter can do the reverse.
19399 The filter accepts the following options:
19404 Set the grid size (i.e. the number of lines and columns). For the syntax of
19405 this option, check the
19406 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19409 Set the maximum number of frames to render in the given area. It must be less
19410 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
19411 the area will be used.
19414 Set the outer border margin in pixels.
19417 Set the inner border thickness (i.e. the number of pixels between frames). For
19418 more advanced padding options (such as having different values for the edges),
19419 refer to the pad video filter.
19422 Specify the color of the unused area. For the syntax of this option, check the
19423 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19424 The default value of @var{color} is "black".
19427 Set the number of frames to overlap when tiling several successive frames together.
19428 The value must be between @code{0} and @var{nb_frames - 1}.
19431 Set the number of frames to initially be empty before displaying first output frame.
19432 This controls how soon will one get first output frame.
19433 The value must be between @code{0} and @var{nb_frames - 1}.
19436 @subsection Examples
19440 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
19442 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
19444 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
19445 duplicating each output frame to accommodate the originally detected frame
19449 Display @code{5} pictures in an area of @code{3x2} frames,
19450 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
19451 mixed flat and named options:
19453 tile=3x2:nb_frames=5:padding=7:margin=2
19457 @section tinterlace
19459 Perform various types of temporal field interlacing.
19461 Frames are counted starting from 1, so the first input frame is
19464 The filter accepts the following options:
19469 Specify the mode of the interlacing. This option can also be specified
19470 as a value alone. See below for a list of values for this option.
19472 Available values are:
19476 Move odd frames into the upper field, even into the lower field,
19477 generating a double height frame at half frame rate.
19481 Frame 1 Frame 2 Frame 3 Frame 4
19483 11111 22222 33333 44444
19484 11111 22222 33333 44444
19485 11111 22222 33333 44444
19486 11111 22222 33333 44444
19500 Only output odd frames, even frames are dropped, generating a frame with
19501 unchanged height at half frame rate.
19506 Frame 1 Frame 2 Frame 3 Frame 4
19508 11111 22222 33333 44444
19509 11111 22222 33333 44444
19510 11111 22222 33333 44444
19511 11111 22222 33333 44444
19521 Only output even frames, odd frames are dropped, generating a frame with
19522 unchanged height at half frame rate.
19527 Frame 1 Frame 2 Frame 3 Frame 4
19529 11111 22222 33333 44444
19530 11111 22222 33333 44444
19531 11111 22222 33333 44444
19532 11111 22222 33333 44444
19542 Expand each frame to full height, but pad alternate lines with black,
19543 generating a frame with double height at the same input frame rate.
19548 Frame 1 Frame 2 Frame 3 Frame 4
19550 11111 22222 33333 44444
19551 11111 22222 33333 44444
19552 11111 22222 33333 44444
19553 11111 22222 33333 44444
19556 11111 ..... 33333 .....
19557 ..... 22222 ..... 44444
19558 11111 ..... 33333 .....
19559 ..... 22222 ..... 44444
19560 11111 ..... 33333 .....
19561 ..... 22222 ..... 44444
19562 11111 ..... 33333 .....
19563 ..... 22222 ..... 44444
19567 @item interleave_top, 4
19568 Interleave the upper field from odd frames with the lower field from
19569 even frames, generating a frame with unchanged height at half frame rate.
19574 Frame 1 Frame 2 Frame 3 Frame 4
19576 11111<- 22222 33333<- 44444
19577 11111 22222<- 33333 44444<-
19578 11111<- 22222 33333<- 44444
19579 11111 22222<- 33333 44444<-
19589 @item interleave_bottom, 5
19590 Interleave the lower field from odd frames with the upper field from
19591 even frames, generating a frame with unchanged height at half frame rate.
19596 Frame 1 Frame 2 Frame 3 Frame 4
19598 11111 22222<- 33333 44444<-
19599 11111<- 22222 33333<- 44444
19600 11111 22222<- 33333 44444<-
19601 11111<- 22222 33333<- 44444
19611 @item interlacex2, 6
19612 Double frame rate with unchanged height. Frames are inserted each
19613 containing the second temporal field from the previous input frame and
19614 the first temporal field from the next input frame. This mode relies on
19615 the top_field_first flag. Useful for interlaced video displays with no
19616 field synchronisation.
19621 Frame 1 Frame 2 Frame 3 Frame 4
19623 11111 22222 33333 44444
19624 11111 22222 33333 44444
19625 11111 22222 33333 44444
19626 11111 22222 33333 44444
19629 11111 22222 22222 33333 33333 44444 44444
19630 11111 11111 22222 22222 33333 33333 44444
19631 11111 22222 22222 33333 33333 44444 44444
19632 11111 11111 22222 22222 33333 33333 44444
19637 Move odd frames into the upper field, even into the lower field,
19638 generating a double height frame at same frame rate.
19643 Frame 1 Frame 2 Frame 3 Frame 4
19645 11111 22222 33333 44444
19646 11111 22222 33333 44444
19647 11111 22222 33333 44444
19648 11111 22222 33333 44444
19651 11111 33333 33333 55555
19652 22222 22222 44444 44444
19653 11111 33333 33333 55555
19654 22222 22222 44444 44444
19655 11111 33333 33333 55555
19656 22222 22222 44444 44444
19657 11111 33333 33333 55555
19658 22222 22222 44444 44444
19663 Numeric values are deprecated but are accepted for backward
19664 compatibility reasons.
19666 Default mode is @code{merge}.
19669 Specify flags influencing the filter process.
19671 Available value for @var{flags} is:
19674 @item low_pass_filter, vlpf
19675 Enable linear vertical low-pass filtering in the filter.
19676 Vertical low-pass filtering is required when creating an interlaced
19677 destination from a progressive source which contains high-frequency
19678 vertical detail. Filtering will reduce interlace 'twitter' and Moire
19681 @item complex_filter, cvlpf
19682 Enable complex vertical low-pass filtering.
19683 This will slightly less reduce interlace 'twitter' and Moire
19684 patterning but better retain detail and subjective sharpness impression.
19687 Bypass already interlaced frames, only adjust the frame rate.
19690 Vertical low-pass filtering and bypassing already interlaced frames can only be
19691 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
19696 Pick median pixels from several successive input video frames.
19698 The filter accepts the following options:
19702 Set radius of median filter.
19703 Default is 1. Allowed range is from 1 to 127.
19706 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
19709 Set median percentile. Default value is @code{0.5}.
19710 Default value of @code{0.5} will pick always median values, while @code{0} will pick
19711 minimum values, and @code{1} maximum values.
19714 @subsection Commands
19716 This filter supports all above options as @ref{commands}, excluding option @code{radius}.
19718 @section tmidequalizer
19720 Apply Temporal Midway Video Equalization effect.
19722 Midway Video Equalization adjusts a sequence of video frames to have the same
19723 histograms, while maintaining their dynamics as much as possible. It's
19724 useful for e.g. matching exposures from a video frames sequence.
19726 This filter accepts the following option:
19730 Set filtering radius. Default is @code{5}. Allowed range is from 1 to 127.
19733 Set filtering sigma. Default is @code{0.5}. This controls strength of filtering.
19734 Setting this option to 0 effectively does nothing.
19737 Set which planes to process. Default is @code{15}, which is all available planes.
19742 Mix successive video frames.
19744 A description of the accepted options follows.
19748 The number of successive frames to mix. If unspecified, it defaults to 3.
19751 Specify weight of each input video frame.
19752 Each weight is separated by space. If number of weights is smaller than
19753 number of @var{frames} last specified weight will be used for all remaining
19757 Specify scale, if it is set it will be multiplied with sum
19758 of each weight multiplied with pixel values to give final destination
19759 pixel value. By default @var{scale} is auto scaled to sum of weights.
19762 @subsection Examples
19766 Average 7 successive frames:
19768 tmix=frames=7:weights="1 1 1 1 1 1 1"
19772 Apply simple temporal convolution:
19774 tmix=frames=3:weights="-1 3 -1"
19778 Similar as above but only showing temporal differences:
19780 tmix=frames=3:weights="-1 2 -1":scale=1
19784 @subsection Commands
19786 This filter supports the following commands:
19790 Syntax is same as option with same name.
19795 Tone map colors from different dynamic ranges.
19797 This filter expects data in single precision floating point, as it needs to
19798 operate on (and can output) out-of-range values. Another filter, such as
19799 @ref{zscale}, is needed to convert the resulting frame to a usable format.
19801 The tonemapping algorithms implemented only work on linear light, so input
19802 data should be linearized beforehand (and possibly correctly tagged).
19805 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
19808 @subsection Options
19809 The filter accepts the following options.
19813 Set the tone map algorithm to use.
19815 Possible values are:
19818 Do not apply any tone map, only desaturate overbright pixels.
19821 Hard-clip any out-of-range values. Use it for perfect color accuracy for
19822 in-range values, while distorting out-of-range values.
19825 Stretch the entire reference gamut to a linear multiple of the display.
19828 Fit a logarithmic transfer between the tone curves.
19831 Preserve overall image brightness with a simple curve, using nonlinear
19832 contrast, which results in flattening details and degrading color accuracy.
19835 Preserve both dark and bright details better than @var{reinhard}, at the cost
19836 of slightly darkening everything. Use it when detail preservation is more
19837 important than color and brightness accuracy.
19840 Smoothly map out-of-range values, while retaining contrast and colors for
19841 in-range material as much as possible. Use it when color accuracy is more
19842 important than detail preservation.
19848 Tune the tone mapping algorithm.
19850 This affects the following algorithms:
19856 Specifies the scale factor to use while stretching.
19860 Specifies the exponent of the function.
19864 Specify an extra linear coefficient to multiply into the signal before clipping.
19868 Specify the local contrast coefficient at the display peak.
19869 Default to 0.5, which means that in-gamut values will be about half as bright
19876 Specify the transition point from linear to mobius transform. Every value
19877 below this point is guaranteed to be mapped 1:1. The higher the value, the
19878 more accurate the result will be, at the cost of losing bright details.
19879 Default to 0.3, which due to the steep initial slope still preserves in-range
19880 colors fairly accurately.
19884 Apply desaturation for highlights that exceed this level of brightness. The
19885 higher the parameter, the more color information will be preserved. This
19886 setting helps prevent unnaturally blown-out colors for super-highlights, by
19887 (smoothly) turning into white instead. This makes images feel more natural,
19888 at the cost of reducing information about out-of-range colors.
19890 The default of 2.0 is somewhat conservative and will mostly just apply to
19891 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
19893 This option works only if the input frame has a supported color tag.
19896 Override signal/nominal/reference peak with this value. Useful when the
19897 embedded peak information in display metadata is not reliable or when tone
19898 mapping from a lower range to a higher range.
19903 Temporarily pad video frames.
19905 The filter accepts the following options:
19909 Specify number of delay frames before input video stream. Default is 0.
19912 Specify number of padding frames after input video stream.
19913 Set to -1 to pad indefinitely. Default is 0.
19916 Set kind of frames added to beginning of stream.
19917 Can be either @var{add} or @var{clone}.
19918 With @var{add} frames of solid-color are added.
19919 With @var{clone} frames are clones of first frame.
19920 Default is @var{add}.
19923 Set kind of frames added to end of stream.
19924 Can be either @var{add} or @var{clone}.
19925 With @var{add} frames of solid-color are added.
19926 With @var{clone} frames are clones of last frame.
19927 Default is @var{add}.
19929 @item start_duration, stop_duration
19930 Specify the duration of the start/stop delay. See
19931 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19932 for the accepted syntax.
19933 These options override @var{start} and @var{stop}. Default is 0.
19936 Specify the color of the padded area. For the syntax of this option,
19937 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
19938 manual,ffmpeg-utils}.
19940 The default value of @var{color} is "black".
19946 Transpose rows with columns in the input video and optionally flip it.
19948 It accepts the following parameters:
19953 Specify the transposition direction.
19955 Can assume the following values:
19957 @item 0, 4, cclock_flip
19958 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
19966 Rotate by 90 degrees clockwise, that is:
19974 Rotate by 90 degrees counterclockwise, that is:
19981 @item 3, 7, clock_flip
19982 Rotate by 90 degrees clockwise and vertically flip, that is:
19990 For values between 4-7, the transposition is only done if the input
19991 video geometry is portrait and not landscape. These values are
19992 deprecated, the @code{passthrough} option should be used instead.
19994 Numerical values are deprecated, and should be dropped in favor of
19995 symbolic constants.
19998 Do not apply the transposition if the input geometry matches the one
19999 specified by the specified value. It accepts the following values:
20002 Always apply transposition.
20004 Preserve portrait geometry (when @var{height} >= @var{width}).
20006 Preserve landscape geometry (when @var{width} >= @var{height}).
20009 Default value is @code{none}.
20012 For example to rotate by 90 degrees clockwise and preserve portrait
20015 transpose=dir=1:passthrough=portrait
20018 The command above can also be specified as:
20020 transpose=1:portrait
20023 @section transpose_npp
20025 Transpose rows with columns in the input video and optionally flip it.
20026 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
20028 It accepts the following parameters:
20033 Specify the transposition direction.
20035 Can assume the following values:
20038 Rotate by 90 degrees counterclockwise and vertically flip. (default)
20041 Rotate by 90 degrees clockwise.
20044 Rotate by 90 degrees counterclockwise.
20047 Rotate by 90 degrees clockwise and vertically flip.
20051 Do not apply the transposition if the input geometry matches the one
20052 specified by the specified value. It accepts the following values:
20055 Always apply transposition. (default)
20057 Preserve portrait geometry (when @var{height} >= @var{width}).
20059 Preserve landscape geometry (when @var{width} >= @var{height}).
20065 Trim the input so that the output contains one continuous subpart of the input.
20067 It accepts the following parameters:
20070 Specify the time of the start of the kept section, i.e. the frame with the
20071 timestamp @var{start} will be the first frame in the output.
20074 Specify the time of the first frame that will be dropped, i.e. the frame
20075 immediately preceding the one with the timestamp @var{end} will be the last
20076 frame in the output.
20079 This is the same as @var{start}, except this option sets the start timestamp
20080 in timebase units instead of seconds.
20083 This is the same as @var{end}, except this option sets the end timestamp
20084 in timebase units instead of seconds.
20087 The maximum duration of the output in seconds.
20090 The number of the first frame that should be passed to the output.
20093 The number of the first frame that should be dropped.
20096 @option{start}, @option{end}, and @option{duration} are expressed as time
20097 duration specifications; see
20098 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20099 for the accepted syntax.
20101 Note that the first two sets of the start/end options and the @option{duration}
20102 option look at the frame timestamp, while the _frame variants simply count the
20103 frames that pass through the filter. Also note that this filter does not modify
20104 the timestamps. If you wish for the output timestamps to start at zero, insert a
20105 setpts filter after the trim filter.
20107 If multiple start or end options are set, this filter tries to be greedy and
20108 keep all the frames that match at least one of the specified constraints. To keep
20109 only the part that matches all the constraints at once, chain multiple trim
20112 The defaults are such that all the input is kept. So it is possible to set e.g.
20113 just the end values to keep everything before the specified time.
20118 Drop everything except the second minute of input:
20120 ffmpeg -i INPUT -vf trim=60:120
20124 Keep only the first second:
20126 ffmpeg -i INPUT -vf trim=duration=1
20131 @section unpremultiply
20132 Apply alpha unpremultiply effect to input video stream using first plane
20133 of second stream as alpha.
20135 Both streams must have same dimensions and same pixel format.
20137 The filter accepts the following option:
20141 Set which planes will be processed, unprocessed planes will be copied.
20142 By default value 0xf, all planes will be processed.
20144 If the format has 1 or 2 components, then luma is bit 0.
20145 If the format has 3 or 4 components:
20146 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
20147 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
20148 If present, the alpha channel is always the last bit.
20151 Do not require 2nd input for processing, instead use alpha plane from input stream.
20157 Sharpen or blur the input video.
20159 It accepts the following parameters:
20162 @item luma_msize_x, lx
20163 Set the luma matrix horizontal size. It must be an odd integer between
20164 3 and 23. The default value is 5.
20166 @item luma_msize_y, ly
20167 Set the luma matrix vertical size. It must be an odd integer between 3
20168 and 23. The default value is 5.
20170 @item luma_amount, la
20171 Set the luma effect strength. It must be a floating point number, reasonable
20172 values lay between -1.5 and 1.5.
20174 Negative values will blur the input video, while positive values will
20175 sharpen it, a value of zero will disable the effect.
20177 Default value is 1.0.
20179 @item chroma_msize_x, cx
20180 Set the chroma matrix horizontal size. It must be an odd integer
20181 between 3 and 23. The default value is 5.
20183 @item chroma_msize_y, cy
20184 Set the chroma matrix vertical size. It must be an odd integer
20185 between 3 and 23. The default value is 5.
20187 @item chroma_amount, ca
20188 Set the chroma effect strength. It must be a floating point number, reasonable
20189 values lay between -1.5 and 1.5.
20191 Negative values will blur the input video, while positive values will
20192 sharpen it, a value of zero will disable the effect.
20194 Default value is 0.0.
20198 All parameters are optional and default to the equivalent of the
20199 string '5:5:1.0:5:5:0.0'.
20201 @subsection Examples
20205 Apply strong luma sharpen effect:
20207 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
20211 Apply a strong blur of both luma and chroma parameters:
20213 unsharp=7:7:-2:7:7:-2
20220 Decompose a video made of tiled images into the individual images.
20222 The frame rate of the output video is the frame rate of the input video
20223 multiplied by the number of tiles.
20225 This filter does the reverse of @ref{tile}.
20227 The filter accepts the following options:
20232 Set the grid size (i.e. the number of lines and columns). For the syntax of
20233 this option, check the
20234 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20237 @subsection Examples
20241 Produce a 1-second video from a still image file made of 25 frames stacked
20242 vertically, like an analogic film reel:
20244 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
20250 Apply ultra slow/simple postprocessing filter that compresses and decompresses
20251 the image at several (or - in the case of @option{quality} level @code{8} - all)
20252 shifts and average the results.
20254 The way this differs from the behavior of spp is that uspp actually encodes &
20255 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
20256 DCT similar to MJPEG.
20258 The filter accepts the following options:
20262 Set quality. This option defines the number of levels for averaging. It accepts
20263 an integer in the range 0-8. If set to @code{0}, the filter will have no
20264 effect. A value of @code{8} means the higher quality. For each increment of
20265 that value the speed drops by a factor of approximately 2. Default value is
20269 Force a constant quantization parameter. If not set, the filter will use the QP
20270 from the video stream (if available).
20275 Convert 360 videos between various formats.
20277 The filter accepts the following options:
20283 Set format of the input/output video.
20291 Equirectangular projection.
20296 Cubemap with 3x2/6x1/1x6 layout.
20298 Format specific options:
20303 Set padding proportion for the input/output cubemap. Values in decimals.
20310 1% of face is padding. For example, with 1920x1280 resolution face size would be 640x640 and padding would be 3 pixels from each side. (640 * 0.01 = 6 pixels)
20313 Default value is @b{@samp{0}}.
20314 Maximum value is @b{@samp{0.1}}.
20318 Set fixed padding for the input/output cubemap. Values in pixels.
20320 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
20324 Set order of faces for the input/output cubemap. Choose one direction for each position.
20326 Designation of directions:
20342 Default value is @b{@samp{rludfb}}.
20346 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
20348 Designation of angles:
20351 0 degrees clockwise
20353 90 degrees clockwise
20355 180 degrees clockwise
20357 270 degrees clockwise
20360 Default value is @b{@samp{000000}}.
20364 Equi-Angular Cubemap.
20371 Format specific options:
20376 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20378 If diagonal field of view is set it overrides horizontal and vertical field of view.
20383 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20385 If diagonal field of view is set it overrides horizontal and vertical field of view.
20391 Format specific options:
20396 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20398 If diagonal field of view is set it overrides horizontal and vertical field of view.
20403 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20405 If diagonal field of view is set it overrides horizontal and vertical field of view.
20411 Facebook's 360 formats.
20414 Stereographic format.
20416 Format specific options:
20421 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20423 If diagonal field of view is set it overrides horizontal and vertical field of view.
20428 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20430 If diagonal field of view is set it overrides horizontal and vertical field of view.
20437 Ball format, gives significant distortion toward the back.
20440 Hammer-Aitoff map projection format.
20443 Sinusoidal map projection format.
20446 Fisheye projection.
20448 Format specific options:
20453 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20455 If diagonal field of view is set it overrides horizontal and vertical field of view.
20460 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20462 If diagonal field of view is set it overrides horizontal and vertical field of view.
20466 Pannini projection.
20468 Format specific options:
20471 Set output pannini parameter.
20474 Set input pannini parameter.
20478 Cylindrical projection.
20480 Format specific options:
20485 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20487 If diagonal field of view is set it overrides horizontal and vertical field of view.
20492 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20494 If diagonal field of view is set it overrides horizontal and vertical field of view.
20498 Perspective projection. @i{(output only)}
20500 Format specific options:
20503 Set perspective parameter.
20507 Tetrahedron projection.
20510 Truncated square pyramid projection.
20514 Half equirectangular projection.
20519 Format specific options:
20524 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20526 If diagonal field of view is set it overrides horizontal and vertical field of view.
20531 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20533 If diagonal field of view is set it overrides horizontal and vertical field of view.
20537 Orthographic format.
20539 Format specific options:
20544 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20546 If diagonal field of view is set it overrides horizontal and vertical field of view.
20551 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20553 If diagonal field of view is set it overrides horizontal and vertical field of view.
20557 Octahedron projection.
20561 Set interpolation method.@*
20562 @i{Note: more complex interpolation methods require much more memory to run.}
20572 Bilinear interpolation.
20574 Lagrange9 interpolation.
20577 Bicubic interpolation.
20580 Lanczos interpolation.
20583 Spline16 interpolation.
20586 Gaussian interpolation.
20588 Mitchell interpolation.
20591 Default value is @b{@samp{line}}.
20595 Set the output video resolution.
20597 Default resolution depends on formats.
20601 Set the input/output stereo format.
20612 Default value is @b{@samp{2d}} for input and output format.
20617 Set rotation for the output video. Values in degrees.
20620 Set rotation order for the output video. Choose one item for each position.
20631 Default value is @b{@samp{ypr}}.
20636 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
20640 Set if input video is flipped horizontally/vertically. Boolean values.
20643 Set if input video is transposed. Boolean value, by default disabled.
20646 Set if output video needs to be transposed. Boolean value, by default disabled.
20649 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
20652 @subsection Examples
20656 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
20658 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
20661 Extract back view of Equi-Angular Cubemap:
20663 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
20666 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
20668 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
20672 @subsection Commands
20674 This filter supports subset of above options as @ref{commands}.
20676 @section vaguedenoiser
20678 Apply a wavelet based denoiser.
20680 It transforms each frame from the video input into the wavelet domain,
20681 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
20682 the obtained coefficients. It does an inverse wavelet transform after.
20683 Due to wavelet properties, it should give a nice smoothed result, and
20684 reduced noise, without blurring picture features.
20686 This filter accepts the following options:
20690 The filtering strength. The higher, the more filtered the video will be.
20691 Hard thresholding can use a higher threshold than soft thresholding
20692 before the video looks overfiltered. Default value is 2.
20695 The filtering method the filter will use.
20697 It accepts the following values:
20700 All values under the threshold will be zeroed.
20703 All values under the threshold will be zeroed. All values above will be
20704 reduced by the threshold.
20707 Scales or nullifies coefficients - intermediary between (more) soft and
20708 (less) hard thresholding.
20711 Default is garrote.
20714 Number of times, the wavelet will decompose the picture. Picture can't
20715 be decomposed beyond a particular point (typically, 8 for a 640x480
20716 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
20719 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
20722 A list of the planes to process. By default all planes are processed.
20725 The threshold type the filter will use.
20727 It accepts the following values:
20730 Threshold used is same for all decompositions.
20733 Threshold used depends also on each decomposition coefficients.
20736 Default is universal.
20739 @section vectorscope
20741 Display 2 color component values in the two dimensional graph (which is called
20744 This filter accepts the following options:
20748 Set vectorscope mode.
20750 It accepts the following values:
20754 Gray values are displayed on graph, higher brightness means more pixels have
20755 same component color value on location in graph. This is the default mode.
20758 Gray values are displayed on graph. Surrounding pixels values which are not
20759 present in video frame are drawn in gradient of 2 color components which are
20760 set by option @code{x} and @code{y}. The 3rd color component is static.
20763 Actual color components values present in video frame are displayed on graph.
20766 Similar as color2 but higher frequency of same values @code{x} and @code{y}
20767 on graph increases value of another color component, which is luminance by
20768 default values of @code{x} and @code{y}.
20771 Actual colors present in video frame are displayed on graph. If two different
20772 colors map to same position on graph then color with higher value of component
20773 not present in graph is picked.
20776 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
20777 component picked from radial gradient.
20781 Set which color component will be represented on X-axis. Default is @code{1}.
20784 Set which color component will be represented on Y-axis. Default is @code{2}.
20787 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
20788 of color component which represents frequency of (X, Y) location in graph.
20793 No envelope, this is default.
20796 Instant envelope, even darkest single pixel will be clearly highlighted.
20799 Hold maximum and minimum values presented in graph over time. This way you
20800 can still spot out of range values without constantly looking at vectorscope.
20803 Peak and instant envelope combined together.
20807 Set what kind of graticule to draw.
20816 Set graticule opacity.
20819 Set graticule flags.
20823 Draw graticule for white point.
20826 Draw graticule for black point.
20829 Draw color points short names.
20833 Set background opacity.
20835 @item lthreshold, l
20836 Set low threshold for color component not represented on X or Y axis.
20837 Values lower than this value will be ignored. Default is 0.
20838 Note this value is multiplied with actual max possible value one pixel component
20839 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
20842 @item hthreshold, h
20843 Set high threshold for color component not represented on X or Y axis.
20844 Values higher than this value will be ignored. Default is 1.
20845 Note this value is multiplied with actual max possible value one pixel component
20846 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
20847 is 0.9 * 255 = 230.
20849 @item colorspace, c
20850 Set what kind of colorspace to use when drawing graticule.
20860 Set color tint for gray/tint vectorscope mode. By default both options are zero.
20861 This means no tint, and output will remain gray.
20864 @anchor{vidstabdetect}
20865 @section vidstabdetect
20867 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
20868 @ref{vidstabtransform} for pass 2.
20870 This filter generates a file with relative translation and rotation
20871 transform information about subsequent frames, which is then used by
20872 the @ref{vidstabtransform} filter.
20874 To enable compilation of this filter you need to configure FFmpeg with
20875 @code{--enable-libvidstab}.
20877 This filter accepts the following options:
20881 Set the path to the file used to write the transforms information.
20882 Default value is @file{transforms.trf}.
20885 Set how shaky the video is and how quick the camera is. It accepts an
20886 integer in the range 1-10, a value of 1 means little shakiness, a
20887 value of 10 means strong shakiness. Default value is 5.
20890 Set the accuracy of the detection process. It must be a value in the
20891 range 1-15. A value of 1 means low accuracy, a value of 15 means high
20892 accuracy. Default value is 15.
20895 Set stepsize of the search process. The region around minimum is
20896 scanned with 1 pixel resolution. Default value is 6.
20899 Set minimum contrast. Below this value a local measurement field is
20900 discarded. Must be a floating point value in the range 0-1. Default
20904 Set reference frame number for tripod mode.
20906 If enabled, the motion of the frames is compared to a reference frame
20907 in the filtered stream, identified by the specified number. The idea
20908 is to compensate all movements in a more-or-less static scene and keep
20909 the camera view absolutely still.
20911 If set to 0, it is disabled. The frames are counted starting from 1.
20914 Show fields and transforms in the resulting frames. It accepts an
20915 integer in the range 0-2. Default value is 0, which disables any
20919 @subsection Examples
20923 Use default values:
20929 Analyze strongly shaky movie and put the results in file
20930 @file{mytransforms.trf}:
20932 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
20936 Visualize the result of internal transformations in the resulting
20939 vidstabdetect=show=1
20943 Analyze a video with medium shakiness using @command{ffmpeg}:
20945 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
20949 @anchor{vidstabtransform}
20950 @section vidstabtransform
20952 Video stabilization/deshaking: pass 2 of 2,
20953 see @ref{vidstabdetect} for pass 1.
20955 Read a file with transform information for each frame and
20956 apply/compensate them. Together with the @ref{vidstabdetect}
20957 filter this can be used to deshake videos. See also
20958 @url{http://public.hronopik.de/vid.stab}. It is important to also use
20959 the @ref{unsharp} filter, see below.
20961 To enable compilation of this filter you need to configure FFmpeg with
20962 @code{--enable-libvidstab}.
20964 @subsection Options
20968 Set path to the file used to read the transforms. Default value is
20969 @file{transforms.trf}.
20972 Set the number of frames (value*2 + 1) used for lowpass filtering the
20973 camera movements. Default value is 10.
20975 For example a number of 10 means that 21 frames are used (10 in the
20976 past and 10 in the future) to smoothen the motion in the video. A
20977 larger value leads to a smoother video, but limits the acceleration of
20978 the camera (pan/tilt movements). 0 is a special case where a static
20979 camera is simulated.
20982 Set the camera path optimization algorithm.
20984 Accepted values are:
20987 gaussian kernel low-pass filter on camera motion (default)
20989 averaging on transformations
20993 Set maximal number of pixels to translate frames. Default value is -1,
20997 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
20998 value is -1, meaning no limit.
21001 Specify how to deal with borders that may be visible due to movement
21004 Available values are:
21007 keep image information from previous frame (default)
21009 fill the border black
21013 Invert transforms if set to 1. Default value is 0.
21016 Consider transforms as relative to previous frame if set to 1,
21017 absolute if set to 0. Default value is 0.
21020 Set percentage to zoom. A positive value will result in a zoom-in
21021 effect, a negative value in a zoom-out effect. Default value is 0 (no
21025 Set optimal zooming to avoid borders.
21027 Accepted values are:
21032 optimal static zoom value is determined (only very strong movements
21033 will lead to visible borders) (default)
21035 optimal adaptive zoom value is determined (no borders will be
21036 visible), see @option{zoomspeed}
21039 Note that the value given at zoom is added to the one calculated here.
21042 Set percent to zoom maximally each frame (enabled when
21043 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
21047 Specify type of interpolation.
21049 Available values are:
21054 linear only horizontal
21056 linear in both directions (default)
21058 cubic in both directions (slow)
21062 Enable virtual tripod mode if set to 1, which is equivalent to
21063 @code{relative=0:smoothing=0}. Default value is 0.
21065 Use also @code{tripod} option of @ref{vidstabdetect}.
21068 Increase log verbosity if set to 1. Also the detected global motions
21069 are written to the temporary file @file{global_motions.trf}. Default
21073 @subsection Examples
21077 Use @command{ffmpeg} for a typical stabilization with default values:
21079 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
21082 Note the use of the @ref{unsharp} filter which is always recommended.
21085 Zoom in a bit more and load transform data from a given file:
21087 vidstabtransform=zoom=5:input="mytransforms.trf"
21091 Smoothen the video even more:
21093 vidstabtransform=smoothing=30
21099 Flip the input video vertically.
21101 For example, to vertically flip a video with @command{ffmpeg}:
21103 ffmpeg -i in.avi -vf "vflip" out.avi
21108 Detect variable frame rate video.
21110 This filter tries to detect if the input is variable or constant frame rate.
21112 At end it will output number of frames detected as having variable delta pts,
21113 and ones with constant delta pts.
21114 If there was frames with variable delta, than it will also show min, max and
21115 average delta encountered.
21119 Boost or alter saturation.
21121 The filter accepts the following options:
21124 Set strength of boost if positive value or strength of alter if negative value.
21125 Default is 0. Allowed range is from -2 to 2.
21128 Set the red balance. Default is 1. Allowed range is from -10 to 10.
21131 Set the green balance. Default is 1. Allowed range is from -10 to 10.
21134 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
21137 Set the red luma coefficient.
21140 Set the green luma coefficient.
21143 Set the blue luma coefficient.
21146 If @code{intensity} is negative and this is set to 1, colors will change,
21147 otherwise colors will be less saturated, more towards gray.
21150 @subsection Commands
21152 This filter supports the all above options as @ref{commands}.
21157 Make or reverse a natural vignetting effect.
21159 The filter accepts the following options:
21163 Set lens angle expression as a number of radians.
21165 The value is clipped in the @code{[0,PI/2]} range.
21167 Default value: @code{"PI/5"}
21171 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
21175 Set forward/backward mode.
21177 Available modes are:
21180 The larger the distance from the central point, the darker the image becomes.
21183 The larger the distance from the central point, the brighter the image becomes.
21184 This can be used to reverse a vignette effect, though there is no automatic
21185 detection to extract the lens @option{angle} and other settings (yet). It can
21186 also be used to create a burning effect.
21189 Default value is @samp{forward}.
21192 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
21194 It accepts the following values:
21197 Evaluate expressions only once during the filter initialization.
21200 Evaluate expressions for each incoming frame. This is way slower than the
21201 @samp{init} mode since it requires all the scalers to be re-computed, but it
21202 allows advanced dynamic expressions.
21205 Default value is @samp{init}.
21208 Set dithering to reduce the circular banding effects. Default is @code{1}
21212 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
21213 Setting this value to the SAR of the input will make a rectangular vignetting
21214 following the dimensions of the video.
21216 Default is @code{1/1}.
21219 @subsection Expressions
21221 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
21222 following parameters.
21227 input width and height
21230 the number of input frame, starting from 0
21233 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
21234 @var{TB} units, NAN if undefined
21237 frame rate of the input video, NAN if the input frame rate is unknown
21240 the PTS (Presentation TimeStamp) of the filtered video frame,
21241 expressed in seconds, NAN if undefined
21244 time base of the input video
21248 @subsection Examples
21252 Apply simple strong vignetting effect:
21258 Make a flickering vignetting:
21260 vignette='PI/4+random(1)*PI/50':eval=frame
21265 @section vmafmotion
21267 Obtain the average VMAF motion score of a video.
21268 It is one of the component metrics of VMAF.
21270 The obtained average motion score is printed through the logging system.
21272 The filter accepts the following options:
21276 If specified, the filter will use the named file to save the motion score of
21277 each frame with respect to the previous frame.
21278 When filename equals "-" the data is sent to standard output.
21283 ffmpeg -i ref.mpg -vf vmafmotion -f null -
21287 Stack input videos vertically.
21289 All streams must be of same pixel format and of same width.
21291 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
21292 to create same output.
21294 The filter accepts the following options:
21298 Set number of input streams. Default is 2.
21301 If set to 1, force the output to terminate when the shortest input
21302 terminates. Default value is 0.
21307 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
21308 Deinterlacing Filter").
21310 Based on the process described by Martin Weston for BBC R&D, and
21311 implemented based on the de-interlace algorithm written by Jim
21312 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
21313 uses filter coefficients calculated by BBC R&D.
21315 This filter uses field-dominance information in frame to decide which
21316 of each pair of fields to place first in the output.
21317 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
21319 There are two sets of filter coefficients, so called "simple"
21320 and "complex". Which set of filter coefficients is used can
21321 be set by passing an optional parameter:
21325 Set the interlacing filter coefficients. Accepts one of the following values:
21329 Simple filter coefficient set.
21331 More-complex filter coefficient set.
21333 Default value is @samp{complex}.
21336 The interlacing mode to adopt. It accepts one of the following values:
21340 Output one frame for each frame.
21342 Output one frame for each field.
21345 The default value is @code{field}.
21348 The picture field parity assumed for the input interlaced video. It accepts one
21349 of the following values:
21353 Assume the top field is first.
21355 Assume the bottom field is first.
21357 Enable automatic detection of field parity.
21360 The default value is @code{auto}.
21361 If the interlacing is unknown or the decoder does not export this information,
21362 top field first will be assumed.
21365 Specify which frames to deinterlace. Accepts one of the following values:
21369 Deinterlace all frames,
21371 Only deinterlace frames marked as interlaced.
21374 Default value is @samp{all}.
21377 @subsection Commands
21378 This filter supports same @ref{commands} as options.
21381 Video waveform monitor.
21383 The waveform monitor plots color component intensity. By default luminance
21384 only. Each column of the waveform corresponds to a column of pixels in the
21387 It accepts the following options:
21391 Can be either @code{row}, or @code{column}. Default is @code{column}.
21392 In row mode, the graph on the left side represents color component value 0 and
21393 the right side represents value = 255. In column mode, the top side represents
21394 color component value = 0 and bottom side represents value = 255.
21397 Set intensity. Smaller values are useful to find out how many values of the same
21398 luminance are distributed across input rows/columns.
21399 Default value is @code{0.04}. Allowed range is [0, 1].
21402 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
21403 In mirrored mode, higher values will be represented on the left
21404 side for @code{row} mode and at the top for @code{column} mode. Default is
21405 @code{1} (mirrored).
21409 It accepts the following values:
21412 Presents information identical to that in the @code{parade}, except
21413 that the graphs representing color components are superimposed directly
21416 This display mode makes it easier to spot relative differences or similarities
21417 in overlapping areas of the color components that are supposed to be identical,
21418 such as neutral whites, grays, or blacks.
21421 Display separate graph for the color components side by side in
21422 @code{row} mode or one below the other in @code{column} mode.
21425 Display separate graph for the color components side by side in
21426 @code{column} mode or one below the other in @code{row} mode.
21428 Using this display mode makes it easy to spot color casts in the highlights
21429 and shadows of an image, by comparing the contours of the top and the bottom
21430 graphs of each waveform. Since whites, grays, and blacks are characterized
21431 by exactly equal amounts of red, green, and blue, neutral areas of the picture
21432 should display three waveforms of roughly equal width/height. If not, the
21433 correction is easy to perform by making level adjustments the three waveforms.
21435 Default is @code{stack}.
21437 @item components, c
21438 Set which color components to display. Default is 1, which means only luminance
21439 or red color component if input is in RGB colorspace. If is set for example to
21440 7 it will display all 3 (if) available color components.
21445 No envelope, this is default.
21448 Instant envelope, minimum and maximum values presented in graph will be easily
21449 visible even with small @code{step} value.
21452 Hold minimum and maximum values presented in graph across time. This way you
21453 can still spot out of range values without constantly looking at waveforms.
21456 Peak and instant envelope combined together.
21462 No filtering, this is default.
21465 Luma and chroma combined together.
21468 Similar as above, but shows difference between blue and red chroma.
21471 Similar as above, but use different colors.
21474 Similar as above, but again with different colors.
21477 Displays only chroma.
21480 Displays actual color value on waveform.
21483 Similar as above, but with luma showing frequency of chroma values.
21487 Set which graticule to display.
21491 Do not display graticule.
21494 Display green graticule showing legal broadcast ranges.
21497 Display orange graticule showing legal broadcast ranges.
21500 Display invert graticule showing legal broadcast ranges.
21504 Set graticule opacity.
21507 Set graticule flags.
21511 Draw numbers above lines. By default enabled.
21514 Draw dots instead of lines.
21518 Set scale used for displaying graticule.
21525 Default is digital.
21528 Set background opacity.
21532 Set tint for output.
21533 Only used with lowpass filter and when display is not overlay and input
21534 pixel formats are not RGB.
21537 @section weave, doubleweave
21539 The @code{weave} takes a field-based video input and join
21540 each two sequential fields into single frame, producing a new double
21541 height clip with half the frame rate and half the frame count.
21543 The @code{doubleweave} works same as @code{weave} but without
21544 halving frame rate and frame count.
21546 It accepts the following option:
21550 Set first field. Available values are:
21554 Set the frame as top-field-first.
21557 Set the frame as bottom-field-first.
21561 @subsection Examples
21565 Interlace video using @ref{select} and @ref{separatefields} filter:
21567 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
21572 Apply the xBR high-quality magnification filter which is designed for pixel
21573 art. It follows a set of edge-detection rules, see
21574 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
21576 It accepts the following option:
21580 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
21581 @code{3xBR} and @code{4} for @code{4xBR}.
21582 Default is @code{3}.
21587 Apply cross fade from one input video stream to another input video stream.
21588 The cross fade is applied for specified duration.
21590 The filter accepts the following options:
21594 Set one of available transition effects:
21642 Default transition effect is fade.
21645 Set cross fade duration in seconds.
21646 Default duration is 1 second.
21649 Set cross fade start relative to first input stream in seconds.
21650 Default offset is 0.
21653 Set expression for custom transition effect.
21655 The expressions can use the following variables and functions:
21660 The coordinates of the current sample.
21664 The width and height of the image.
21667 Progress of transition effect.
21670 Currently processed plane.
21673 Return value of first input at current location and plane.
21676 Return value of second input at current location and plane.
21682 Return the value of the pixel at location (@var{x},@var{y}) of the
21683 first/second/third/fourth component of first input.
21689 Return the value of the pixel at location (@var{x},@var{y}) of the
21690 first/second/third/fourth component of second input.
21694 @subsection Examples
21698 Cross fade from one input video to another input video, with fade transition and duration of transition
21699 of 2 seconds starting at offset of 5 seconds:
21701 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
21706 Pick median pixels from several input videos.
21708 The filter accepts the following options:
21712 Set number of inputs.
21713 Default is 3. Allowed range is from 3 to 255.
21714 If number of inputs is even number, than result will be mean value between two median values.
21717 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
21720 Set median percentile. Default value is @code{0.5}.
21721 Default value of @code{0.5} will pick always median values, while @code{0} will pick
21722 minimum values, and @code{1} maximum values.
21725 @subsection Commands
21727 This filter supports all above options as @ref{commands}, excluding option @code{inputs}.
21730 Stack video inputs into custom layout.
21732 All streams must be of same pixel format.
21734 The filter accepts the following options:
21738 Set number of input streams. Default is 2.
21741 Specify layout of inputs.
21742 This option requires the desired layout configuration to be explicitly set by the user.
21743 This sets position of each video input in output. Each input
21744 is separated by '|'.
21745 The first number represents the column, and the second number represents the row.
21746 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
21747 where X is video input from which to take width or height.
21748 Multiple values can be used when separated by '+'. In such
21749 case values are summed together.
21751 Note that if inputs are of different sizes gaps may appear, as not all of
21752 the output video frame will be filled. Similarly, videos can overlap each
21753 other if their position doesn't leave enough space for the full frame of
21756 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
21757 a layout must be set by the user.
21760 If set to 1, force the output to terminate when the shortest input
21761 terminates. Default value is 0.
21764 If set to valid color, all unused pixels will be filled with that color.
21765 By default fill is set to none, so it is disabled.
21768 @subsection Examples
21772 Display 4 inputs into 2x2 grid.
21776 input1(0, 0) | input3(w0, 0)
21777 input2(0, h0) | input4(w0, h0)
21781 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
21784 Note that if inputs are of different sizes, gaps or overlaps may occur.
21787 Display 4 inputs into 1x4 grid.
21794 input4(0, h0+h1+h2)
21798 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
21801 Note that if inputs are of different widths, unused space will appear.
21804 Display 9 inputs into 3x3 grid.
21808 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
21809 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
21810 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
21814 xstack=inputs=9:layout=0_0|0_h0|0_h0+h1|w0_0|w0_h0|w0_h0+h1|w0+w3_0|w0+w3_h0|w0+w3_h0+h1
21817 Note that if inputs are of different sizes, gaps or overlaps may occur.
21820 Display 16 inputs into 4x4 grid.
21824 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
21825 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
21826 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
21827 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
21831 xstack=inputs=16:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2|w0_0|w0_h0|w0_h0+h1|w0_h0+h1+h2|w0+w4_0|
21832 w0+w4_h0|w0+w4_h0+h1|w0+w4_h0+h1+h2|w0+w4+w8_0|w0+w4+w8_h0|w0+w4+w8_h0+h1|w0+w4+w8_h0+h1+h2
21835 Note that if inputs are of different sizes, gaps or overlaps may occur.
21842 Deinterlace the input video ("yadif" means "yet another deinterlacing
21845 It accepts the following parameters:
21851 The interlacing mode to adopt. It accepts one of the following values:
21854 @item 0, send_frame
21855 Output one frame for each frame.
21856 @item 1, send_field
21857 Output one frame for each field.
21858 @item 2, send_frame_nospatial
21859 Like @code{send_frame}, but it skips the spatial interlacing check.
21860 @item 3, send_field_nospatial
21861 Like @code{send_field}, but it skips the spatial interlacing check.
21864 The default value is @code{send_frame}.
21867 The picture field parity assumed for the input interlaced video. It accepts one
21868 of the following values:
21872 Assume the top field is first.
21874 Assume the bottom field is first.
21876 Enable automatic detection of field parity.
21879 The default value is @code{auto}.
21880 If the interlacing is unknown or the decoder does not export this information,
21881 top field first will be assumed.
21884 Specify which frames to deinterlace. Accepts one of the following
21889 Deinterlace all frames.
21890 @item 1, interlaced
21891 Only deinterlace frames marked as interlaced.
21894 The default value is @code{all}.
21897 @section yadif_cuda
21899 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
21900 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
21903 It accepts the following parameters:
21909 The interlacing mode to adopt. It accepts one of the following values:
21912 @item 0, send_frame
21913 Output one frame for each frame.
21914 @item 1, send_field
21915 Output one frame for each field.
21916 @item 2, send_frame_nospatial
21917 Like @code{send_frame}, but it skips the spatial interlacing check.
21918 @item 3, send_field_nospatial
21919 Like @code{send_field}, but it skips the spatial interlacing check.
21922 The default value is @code{send_frame}.
21925 The picture field parity assumed for the input interlaced video. It accepts one
21926 of the following values:
21930 Assume the top field is first.
21932 Assume the bottom field is first.
21934 Enable automatic detection of field parity.
21937 The default value is @code{auto}.
21938 If the interlacing is unknown or the decoder does not export this information,
21939 top field first will be assumed.
21942 Specify which frames to deinterlace. Accepts one of the following
21947 Deinterlace all frames.
21948 @item 1, interlaced
21949 Only deinterlace frames marked as interlaced.
21952 The default value is @code{all}.
21957 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
21958 The algorithm is described in
21959 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
21961 It accepts the following parameters:
21965 Set the window radius. Default value is 3.
21968 Set which planes to filter. Default is only the first plane.
21971 Set blur strength. Default value is 128.
21974 @subsection Commands
21975 This filter supports same @ref{commands} as options.
21979 Apply Zoom & Pan effect.
21981 This filter accepts the following options:
21985 Set the zoom expression. Range is 1-10. Default is 1.
21989 Set the x and y expression. Default is 0.
21992 Set the duration expression in number of frames.
21993 This sets for how many number of frames effect will last for
21994 single input image.
21997 Set the output image size, default is 'hd720'.
22000 Set the output frame rate, default is '25'.
22003 Each expression can contain the following constants:
22022 Output frame count.
22025 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
22027 @item out_time, time, ot
22028 The output timestamp expressed in seconds.
22032 Last calculated 'x' and 'y' position from 'x' and 'y' expression
22033 for current input frame.
22037 'x' and 'y' of last output frame of previous input frame or 0 when there was
22038 not yet such frame (first input frame).
22041 Last calculated zoom from 'z' expression for current input frame.
22044 Last calculated zoom of last output frame of previous input frame.
22047 Number of output frames for current input frame. Calculated from 'd' expression
22048 for each input frame.
22051 number of output frames created for previous input frame
22054 Rational number: input width / input height
22057 sample aspect ratio
22060 display aspect ratio
22064 @subsection Examples
22068 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
22070 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='if(gte(zoom,1.5),x,x+1/a)':y='if(gte(zoom,1.5),y,y+1)':s=640x360
22074 Zoom in up to 1.5x and pan always at center of picture:
22076 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22080 Same as above but without pausing:
22082 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22086 Zoom in 2x into center of picture only for the first second of the input video:
22088 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22095 Scale (resize) the input video, using the z.lib library:
22096 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
22097 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
22099 The zscale filter forces the output display aspect ratio to be the same
22100 as the input, by changing the output sample aspect ratio.
22102 If the input image format is different from the format requested by
22103 the next filter, the zscale filter will convert the input to the
22106 @subsection Options
22107 The filter accepts the following options.
22112 Set the output video dimension expression. Default value is the input
22115 If the @var{width} or @var{w} value is 0, the input width is used for
22116 the output. If the @var{height} or @var{h} value is 0, the input height
22117 is used for the output.
22119 If one and only one of the values is -n with n >= 1, the zscale filter
22120 will use a value that maintains the aspect ratio of the input image,
22121 calculated from the other specified dimension. After that it will,
22122 however, make sure that the calculated dimension is divisible by n and
22123 adjust the value if necessary.
22125 If both values are -n with n >= 1, the behavior will be identical to
22126 both values being set to 0 as previously detailed.
22128 See below for the list of accepted constants for use in the dimension
22132 Set the video size. For the syntax of this option, check the
22133 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22136 Set the dither type.
22138 Possible values are:
22143 @item error_diffusion
22149 Set the resize filter type.
22151 Possible values are:
22161 Default is bilinear.
22164 Set the color range.
22166 Possible values are:
22173 Default is same as input.
22176 Set the color primaries.
22178 Possible values are:
22188 Default is same as input.
22191 Set the transfer characteristics.
22193 Possible values are:
22207 Default is same as input.
22210 Set the colorspace matrix.
22212 Possible value are:
22223 Default is same as input.
22226 Set the input color range.
22228 Possible values are:
22235 Default is same as input.
22237 @item primariesin, pin
22238 Set the input color primaries.
22240 Possible values are:
22250 Default is same as input.
22252 @item transferin, tin
22253 Set the input transfer characteristics.
22255 Possible values are:
22266 Default is same as input.
22268 @item matrixin, min
22269 Set the input colorspace matrix.
22271 Possible value are:
22283 Set the output chroma location.
22285 Possible values are:
22296 @item chromalin, cin
22297 Set the input chroma location.
22299 Possible values are:
22311 Set the nominal peak luminance.
22314 The values of the @option{w} and @option{h} options are expressions
22315 containing the following constants:
22320 The input width and height
22324 These are the same as @var{in_w} and @var{in_h}.
22328 The output (scaled) width and height
22332 These are the same as @var{out_w} and @var{out_h}
22335 The same as @var{iw} / @var{ih}
22338 input sample aspect ratio
22341 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
22345 horizontal and vertical input chroma subsample values. For example for the
22346 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22350 horizontal and vertical output chroma subsample values. For example for the
22351 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22354 @subsection Commands
22356 This filter supports the following commands:
22360 Set the output video dimension expression.
22361 The command accepts the same syntax of the corresponding option.
22363 If the specified expression is not valid, it is kept at its current
22367 @c man end VIDEO FILTERS
22369 @chapter OpenCL Video Filters
22370 @c man begin OPENCL VIDEO FILTERS
22372 Below is a description of the currently available OpenCL video filters.
22374 To enable compilation of these filters you need to configure FFmpeg with
22375 @code{--enable-opencl}.
22377 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
22380 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
22381 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
22382 given device parameters.
22384 @item -filter_hw_device @var{name}
22385 Pass the hardware device called @var{name} to all filters in any filter graph.
22389 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
22393 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
22395 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
22399 Since OpenCL filters are not able to access frame data in normal memory, all frame data needs to be uploaded(@ref{hwupload}) to hardware surfaces connected to the appropriate device before being used and then downloaded(@ref{hwdownload}) back to normal memory. Note that @ref{hwupload} will upload to a surface with the same layout as the software frame, so it may be necessary to add a @ref{format} filter immediately before to get the input into the right format and @ref{hwdownload} does not support all formats on the output - it may be necessary to insert an additional @ref{format} filter immediately following in the graph to get the output in a supported format.
22401 @section avgblur_opencl
22403 Apply average blur filter.
22405 The filter accepts the following options:
22409 Set horizontal radius size.
22410 Range is @code{[1, 1024]} and default value is @code{1}.
22413 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22416 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
22419 @subsection Example
22423 Apply average blur filter with horizontal and vertical size of 3, setting each pixel of the output to the average value of the 7x7 region centered on it in the input. For pixels on the edges of the image, the region does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
22425 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
22429 @section boxblur_opencl
22431 Apply a boxblur algorithm to the input video.
22433 It accepts the following parameters:
22437 @item luma_radius, lr
22438 @item luma_power, lp
22439 @item chroma_radius, cr
22440 @item chroma_power, cp
22441 @item alpha_radius, ar
22442 @item alpha_power, ap
22446 A description of the accepted options follows.
22449 @item luma_radius, lr
22450 @item chroma_radius, cr
22451 @item alpha_radius, ar
22452 Set an expression for the box radius in pixels used for blurring the
22453 corresponding input plane.
22455 The radius value must be a non-negative number, and must not be
22456 greater than the value of the expression @code{min(w,h)/2} for the
22457 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
22460 Default value for @option{luma_radius} is "2". If not specified,
22461 @option{chroma_radius} and @option{alpha_radius} default to the
22462 corresponding value set for @option{luma_radius}.
22464 The expressions can contain the following constants:
22468 The input width and height in pixels.
22472 The input chroma image width and height in pixels.
22476 The horizontal and vertical chroma subsample values. For example, for the
22477 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
22480 @item luma_power, lp
22481 @item chroma_power, cp
22482 @item alpha_power, ap
22483 Specify how many times the boxblur filter is applied to the
22484 corresponding plane.
22486 Default value for @option{luma_power} is 2. If not specified,
22487 @option{chroma_power} and @option{alpha_power} default to the
22488 corresponding value set for @option{luma_power}.
22490 A value of 0 will disable the effect.
22493 @subsection Examples
22495 Apply boxblur filter, setting each pixel of the output to the average value of box-radiuses @var{luma_radius}, @var{chroma_radius}, @var{alpha_radius} for each plane respectively. The filter will apply @var{luma_power}, @var{chroma_power}, @var{alpha_power} times onto the corresponding plane. For pixels on the edges of the image, the radius does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
22499 Apply a boxblur filter with the luma, chroma, and alpha radius
22500 set to 2 and luma, chroma, and alpha power set to 3. The filter will run 3 times with box-radius set to 2 for every plane of the image.
22502 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
22503 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
22507 Apply a boxblur filter with luma radius set to 2, luma_power to 1, chroma_radius to 4, chroma_power to 5, alpha_radius to 3 and alpha_power to 7.
22509 For the luma plane, a 2x2 box radius will be run once.
22511 For the chroma plane, a 4x4 box radius will be run 5 times.
22513 For the alpha plane, a 3x3 box radius will be run 7 times.
22515 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
22519 @section colorkey_opencl
22520 RGB colorspace color keying.
22522 The filter accepts the following options:
22526 The color which will be replaced with transparency.
22529 Similarity percentage with the key color.
22531 0.01 matches only the exact key color, while 1.0 matches everything.
22536 0.0 makes pixels either fully transparent, or not transparent at all.
22538 Higher values result in semi-transparent pixels, with a higher transparency
22539 the more similar the pixels color is to the key color.
22542 @subsection Examples
22546 Make every semi-green pixel in the input transparent with some slight blending:
22548 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
22552 @section convolution_opencl
22554 Apply convolution of 3x3, 5x5, 7x7 matrix.
22556 The filter accepts the following options:
22563 Set matrix for each plane.
22564 Matrix is sequence of 9, 25 or 49 signed numbers.
22565 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
22571 Set multiplier for calculated value for each plane.
22572 If unset or 0, it will be sum of all matrix elements.
22573 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
22579 Set bias for each plane. This value is added to the result of the multiplication.
22580 Useful for making the overall image brighter or darker.
22581 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
22585 @subsection Examples
22591 -i INPUT -vf "hwupload, convolution_opencl=0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0, hwdownload" OUTPUT
22597 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9, hwdownload" OUTPUT
22601 Apply edge enhance:
22603 -i INPUT -vf "hwupload, convolution_opencl=0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128, hwdownload" OUTPUT
22609 -i INPUT -vf "hwupload, convolution_opencl=0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128, hwdownload" OUTPUT
22613 Apply laplacian edge detector which includes diagonals:
22615 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0, hwdownload" OUTPUT
22621 -i INPUT -vf "hwupload, convolution_opencl=-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2, hwdownload" OUTPUT
22625 @section erosion_opencl
22627 Apply erosion effect to the video.
22629 This filter replaces the pixel by the local(3x3) minimum.
22631 It accepts the following options:
22638 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22639 If @code{0}, plane will remain unchanged.
22642 Flag which specifies the pixel to refer to.
22643 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22645 Flags to local 3x3 coordinates region centered on @code{x}:
22654 @subsection Example
22658 Apply erosion filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local minimum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local minimum is more then threshold of the corresponding plane, output pixel will be set to input pixel - threshold of corresponding plane.
22660 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22664 @section deshake_opencl
22665 Feature-point based video stabilization filter.
22667 The filter accepts the following options:
22671 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
22674 Whether or not additional debug info should be displayed, both in the processed output and in the console.
22676 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
22678 Viewing point matches in the output video is only supported for RGB input.
22680 Defaults to @code{0}.
22682 @item adaptive_crop
22683 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
22685 Defaults to @code{1}.
22687 @item refine_features
22688 Whether or not feature points should be refined at a sub-pixel level.
22690 This can be turned off for a slight performance gain at the cost of precision.
22692 Defaults to @code{1}.
22694 @item smooth_strength
22695 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
22697 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
22699 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
22701 Defaults to @code{0.0}.
22703 @item smooth_window_multiplier
22704 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
22706 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
22708 Acceptable values range from @code{0.1} to @code{10.0}.
22710 Larger values increase the amount of motion data available for determining how to smooth the camera path,
22711 potentially improving smoothness, but also increase latency and memory usage.
22713 Defaults to @code{2.0}.
22717 @subsection Examples
22721 Stabilize a video with a fixed, medium smoothing strength:
22723 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
22727 Stabilize a video with debugging (both in console and in rendered video):
22729 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
22733 @section dilation_opencl
22735 Apply dilation effect to the video.
22737 This filter replaces the pixel by the local(3x3) maximum.
22739 It accepts the following options:
22746 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22747 If @code{0}, plane will remain unchanged.
22750 Flag which specifies the pixel to refer to.
22751 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22753 Flags to local 3x3 coordinates region centered on @code{x}:
22762 @subsection Example
22766 Apply dilation filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local maximum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local maximum is more then threshold of the corresponding plane, output pixel will be set to input pixel + threshold of corresponding plane.
22768 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22772 @section nlmeans_opencl
22774 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
22776 @section overlay_opencl
22778 Overlay one video on top of another.
22780 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
22781 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
22783 The filter accepts the following options:
22788 Set the x coordinate of the overlaid video on the main video.
22789 Default value is @code{0}.
22792 Set the y coordinate of the overlaid video on the main video.
22793 Default value is @code{0}.
22797 @subsection Examples
22801 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
22803 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22806 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
22808 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22813 @section pad_opencl
22815 Add paddings to the input image, and place the original input at the
22816 provided @var{x}, @var{y} coordinates.
22818 It accepts the following options:
22823 Specify an expression for the size of the output image with the
22824 paddings added. If the value for @var{width} or @var{height} is 0, the
22825 corresponding input size is used for the output.
22827 The @var{width} expression can reference the value set by the
22828 @var{height} expression, and vice versa.
22830 The default value of @var{width} and @var{height} is 0.
22834 Specify the offsets to place the input image at within the padded area,
22835 with respect to the top/left border of the output image.
22837 The @var{x} expression can reference the value set by the @var{y}
22838 expression, and vice versa.
22840 The default value of @var{x} and @var{y} is 0.
22842 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
22843 so the input image is centered on the padded area.
22846 Specify the color of the padded area. For the syntax of this option,
22847 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
22848 manual,ffmpeg-utils}.
22851 Pad to an aspect instead to a resolution.
22854 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
22855 options are expressions containing the following constants:
22860 The input video width and height.
22864 These are the same as @var{in_w} and @var{in_h}.
22868 The output width and height (the size of the padded area), as
22869 specified by the @var{width} and @var{height} expressions.
22873 These are the same as @var{out_w} and @var{out_h}.
22877 The x and y offsets as specified by the @var{x} and @var{y}
22878 expressions, or NAN if not yet specified.
22881 same as @var{iw} / @var{ih}
22884 input sample aspect ratio
22887 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
22890 @section prewitt_opencl
22892 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
22894 The filter accepts the following option:
22898 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22901 Set value which will be multiplied with filtered result.
22902 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22905 Set value which will be added to filtered result.
22906 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22909 @subsection Example
22913 Apply the Prewitt operator with scale set to 2 and delta set to 10.
22915 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
22919 @anchor{program_opencl}
22920 @section program_opencl
22922 Filter video using an OpenCL program.
22927 OpenCL program source file.
22930 Kernel name in program.
22933 Number of inputs to the filter. Defaults to 1.
22936 Size of output frames. Defaults to the same as the first input.
22940 The @code{program_opencl} filter also supports the @ref{framesync} options.
22942 The program source file must contain a kernel function with the given name,
22943 which will be run once for each plane of the output. Each run on a plane
22944 gets enqueued as a separate 2D global NDRange with one work-item for each
22945 pixel to be generated. The global ID offset for each work-item is therefore
22946 the coordinates of a pixel in the destination image.
22948 The kernel function needs to take the following arguments:
22951 Destination image, @var{__write_only image2d_t}.
22953 This image will become the output; the kernel should write all of it.
22955 Frame index, @var{unsigned int}.
22957 This is a counter starting from zero and increasing by one for each frame.
22959 Source images, @var{__read_only image2d_t}.
22961 These are the most recent images on each input. The kernel may read from
22962 them to generate the output, but they can't be written to.
22969 Copy the input to the output (output must be the same size as the input).
22971 __kernel void copy(__write_only image2d_t destination,
22972 unsigned int index,
22973 __read_only image2d_t source)
22975 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
22977 int2 location = (int2)(get_global_id(0), get_global_id(1));
22979 float4 value = read_imagef(source, sampler, location);
22981 write_imagef(destination, location, value);
22986 Apply a simple transformation, rotating the input by an amount increasing
22987 with the index counter. Pixel values are linearly interpolated by the
22988 sampler, and the output need not have the same dimensions as the input.
22990 __kernel void rotate_image(__write_only image2d_t dst,
22991 unsigned int index,
22992 __read_only image2d_t src)
22994 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22995 CLK_FILTER_LINEAR);
22997 float angle = (float)index / 100.0f;
22999 float2 dst_dim = convert_float2(get_image_dim(dst));
23000 float2 src_dim = convert_float2(get_image_dim(src));
23002 float2 dst_cen = dst_dim / 2.0f;
23003 float2 src_cen = src_dim / 2.0f;
23005 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23007 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
23009 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
23010 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
23012 src_pos = src_pos * src_dim / dst_dim;
23014 float2 src_loc = src_pos + src_cen;
23016 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
23017 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
23018 write_imagef(dst, dst_loc, 0.5f);
23020 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
23025 Blend two inputs together, with the amount of each input used varying
23026 with the index counter.
23028 __kernel void blend_images(__write_only image2d_t dst,
23029 unsigned int index,
23030 __read_only image2d_t src1,
23031 __read_only image2d_t src2)
23033 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23034 CLK_FILTER_LINEAR);
23036 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
23038 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23039 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
23040 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
23042 float4 val1 = read_imagef(src1, sampler, src1_loc);
23043 float4 val2 = read_imagef(src2, sampler, src2_loc);
23045 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
23051 @section roberts_opencl
23052 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
23054 The filter accepts the following option:
23058 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23061 Set value which will be multiplied with filtered result.
23062 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23065 Set value which will be added to filtered result.
23066 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23069 @subsection Example
23073 Apply the Roberts cross operator with scale set to 2 and delta set to 10
23075 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
23079 @section sobel_opencl
23081 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
23083 The filter accepts the following option:
23087 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23090 Set value which will be multiplied with filtered result.
23091 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23094 Set value which will be added to filtered result.
23095 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23098 @subsection Example
23102 Apply sobel operator with scale set to 2 and delta set to 10
23104 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
23108 @section tonemap_opencl
23110 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
23112 It accepts the following parameters:
23116 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
23119 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
23122 Apply desaturation for highlights that exceed this level of brightness. The
23123 higher the parameter, the more color information will be preserved. This
23124 setting helps prevent unnaturally blown-out colors for super-highlights, by
23125 (smoothly) turning into white instead. This makes images feel more natural,
23126 at the cost of reducing information about out-of-range colors.
23128 The default value is 0.5, and the algorithm here is a little different from
23129 the cpu version tonemap currently. A setting of 0.0 disables this option.
23132 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
23133 is used to detect whether the scene has changed or not. If the distance between
23134 the current frame average brightness and the current running average exceeds
23135 a threshold value, we would re-calculate scene average and peak brightness.
23136 The default value is 0.2.
23139 Specify the output pixel format.
23141 Currently supported formats are:
23148 Set the output color range.
23150 Possible values are:
23156 Default is same as input.
23159 Set the output color primaries.
23161 Possible values are:
23167 Default is same as input.
23170 Set the output transfer characteristics.
23172 Possible values are:
23181 Set the output colorspace matrix.
23183 Possible value are:
23189 Default is same as input.
23193 @subsection Example
23197 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
23199 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
23203 @section unsharp_opencl
23205 Sharpen or blur the input video.
23207 It accepts the following parameters:
23210 @item luma_msize_x, lx
23211 Set the luma matrix horizontal size.
23212 Range is @code{[1, 23]} and default value is @code{5}.
23214 @item luma_msize_y, ly
23215 Set the luma matrix vertical size.
23216 Range is @code{[1, 23]} and default value is @code{5}.
23218 @item luma_amount, la
23219 Set the luma effect strength.
23220 Range is @code{[-10, 10]} and default value is @code{1.0}.
23222 Negative values will blur the input video, while positive values will
23223 sharpen it, a value of zero will disable the effect.
23225 @item chroma_msize_x, cx
23226 Set the chroma matrix horizontal size.
23227 Range is @code{[1, 23]} and default value is @code{5}.
23229 @item chroma_msize_y, cy
23230 Set the chroma matrix vertical size.
23231 Range is @code{[1, 23]} and default value is @code{5}.
23233 @item chroma_amount, ca
23234 Set the chroma effect strength.
23235 Range is @code{[-10, 10]} and default value is @code{0.0}.
23237 Negative values will blur the input video, while positive values will
23238 sharpen it, a value of zero will disable the effect.
23242 All parameters are optional and default to the equivalent of the
23243 string '5:5:1.0:5:5:0.0'.
23245 @subsection Examples
23249 Apply strong luma sharpen effect:
23251 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
23255 Apply a strong blur of both luma and chroma parameters:
23257 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
23261 @section xfade_opencl
23263 Cross fade two videos with custom transition effect by using OpenCL.
23265 It accepts the following options:
23269 Set one of possible transition effects.
23273 Select custom transition effect, the actual transition description
23274 will be picked from source and kernel options.
23286 Default transition is fade.
23290 OpenCL program source file for custom transition.
23293 Set name of kernel to use for custom transition from program source file.
23296 Set duration of video transition.
23299 Set time of start of transition relative to first video.
23302 The program source file must contain a kernel function with the given name,
23303 which will be run once for each plane of the output. Each run on a plane
23304 gets enqueued as a separate 2D global NDRange with one work-item for each
23305 pixel to be generated. The global ID offset for each work-item is therefore
23306 the coordinates of a pixel in the destination image.
23308 The kernel function needs to take the following arguments:
23311 Destination image, @var{__write_only image2d_t}.
23313 This image will become the output; the kernel should write all of it.
23316 First Source image, @var{__read_only image2d_t}.
23317 Second Source image, @var{__read_only image2d_t}.
23319 These are the most recent images on each input. The kernel may read from
23320 them to generate the output, but they can't be written to.
23323 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
23330 Apply dots curtain transition effect:
23332 __kernel void blend_images(__write_only image2d_t dst,
23333 __read_only image2d_t src1,
23334 __read_only image2d_t src2,
23337 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23338 CLK_FILTER_LINEAR);
23339 int2 p = (int2)(get_global_id(0), get_global_id(1));
23340 float2 rp = (float2)(get_global_id(0), get_global_id(1));
23341 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
23344 float2 dots = (float2)(20.0, 20.0);
23345 float2 center = (float2)(0,0);
23348 float4 val1 = read_imagef(src1, sampler, p);
23349 float4 val2 = read_imagef(src2, sampler, p);
23350 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
23352 write_imagef(dst, p, next ? val1 : val2);
23358 @c man end OPENCL VIDEO FILTERS
23360 @chapter VAAPI Video Filters
23361 @c man begin VAAPI VIDEO FILTERS
23363 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
23365 To enable compilation of these filters you need to configure FFmpeg with
23366 @code{--enable-vaapi}.
23368 To use vaapi filters, you need to setup the vaapi device correctly. For more information, please read @url{https://trac.ffmpeg.org/wiki/Hardware/VAAPI}
23370 @section tonemap_vaapi
23372 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
23373 It maps the dynamic range of HDR10 content to the SDR content.
23374 It currently only accepts HDR10 as input.
23376 It accepts the following parameters:
23380 Specify the output pixel format.
23382 Currently supported formats are:
23391 Set the output color primaries.
23393 Default is same as input.
23396 Set the output transfer characteristics.
23401 Set the output colorspace matrix.
23403 Default is same as input.
23407 @subsection Example
23411 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
23413 tonemap_vaapi=format=p010:t=bt2020-10
23417 @c man end VAAPI VIDEO FILTERS
23419 @chapter Video Sources
23420 @c man begin VIDEO SOURCES
23422 Below is a description of the currently available video sources.
23426 Buffer video frames, and make them available to the filter chain.
23428 This source is mainly intended for a programmatic use, in particular
23429 through the interface defined in @file{libavfilter/buffersrc.h}.
23431 It accepts the following parameters:
23436 Specify the size (width and height) of the buffered video frames. For the
23437 syntax of this option, check the
23438 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23441 The input video width.
23444 The input video height.
23447 A string representing the pixel format of the buffered video frames.
23448 It may be a number corresponding to a pixel format, or a pixel format
23452 Specify the timebase assumed by the timestamps of the buffered frames.
23455 Specify the frame rate expected for the video stream.
23457 @item pixel_aspect, sar
23458 The sample (pixel) aspect ratio of the input video.
23461 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
23462 to the filtergraph description to specify swscale flags for automatically
23463 inserted scalers. See @ref{Filtergraph syntax}.
23465 @item hw_frames_ctx
23466 When using a hardware pixel format, this should be a reference to an
23467 AVHWFramesContext describing input frames.
23472 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
23475 will instruct the source to accept video frames with size 320x240 and
23476 with format "yuv410p", assuming 1/24 as the timestamps timebase and
23477 square pixels (1:1 sample aspect ratio).
23478 Since the pixel format with name "yuv410p" corresponds to the number 6
23479 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
23480 this example corresponds to:
23482 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
23485 Alternatively, the options can be specified as a flat string, but this
23486 syntax is deprecated:
23488 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
23492 Create a pattern generated by an elementary cellular automaton.
23494 The initial state of the cellular automaton can be defined through the
23495 @option{filename} and @option{pattern} options. If such options are
23496 not specified an initial state is created randomly.
23498 At each new frame a new row in the video is filled with the result of
23499 the cellular automaton next generation. The behavior when the whole
23500 frame is filled is defined by the @option{scroll} option.
23502 This source accepts the following options:
23506 Read the initial cellular automaton state, i.e. the starting row, from
23507 the specified file.
23508 In the file, each non-whitespace character is considered an alive
23509 cell, a newline will terminate the row, and further characters in the
23510 file will be ignored.
23513 Read the initial cellular automaton state, i.e. the starting row, from
23514 the specified string.
23516 Each non-whitespace character in the string is considered an alive
23517 cell, a newline will terminate the row, and further characters in the
23518 string will be ignored.
23521 Set the video rate, that is the number of frames generated per second.
23524 @item random_fill_ratio, ratio
23525 Set the random fill ratio for the initial cellular automaton row. It
23526 is a floating point number value ranging from 0 to 1, defaults to
23529 This option is ignored when a file or a pattern is specified.
23531 @item random_seed, seed
23532 Set the seed for filling randomly the initial row, must be an integer
23533 included between 0 and UINT32_MAX. If not specified, or if explicitly
23534 set to -1, the filter will try to use a good random seed on a best
23538 Set the cellular automaton rule, it is a number ranging from 0 to 255.
23539 Default value is 110.
23542 Set the size of the output video. For the syntax of this option, check the
23543 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23545 If @option{filename} or @option{pattern} is specified, the size is set
23546 by default to the width of the specified initial state row, and the
23547 height is set to @var{width} * PHI.
23549 If @option{size} is set, it must contain the width of the specified
23550 pattern string, and the specified pattern will be centered in the
23553 If a filename or a pattern string is not specified, the size value
23554 defaults to "320x518" (used for a randomly generated initial state).
23557 If set to 1, scroll the output upward when all the rows in the output
23558 have been already filled. If set to 0, the new generated row will be
23559 written over the top row just after the bottom row is filled.
23562 @item start_full, full
23563 If set to 1, completely fill the output with generated rows before
23564 outputting the first frame.
23565 This is the default behavior, for disabling set the value to 0.
23568 If set to 1, stitch the left and right row edges together.
23569 This is the default behavior, for disabling set the value to 0.
23572 @subsection Examples
23576 Read the initial state from @file{pattern}, and specify an output of
23579 cellauto=f=pattern:s=200x400
23583 Generate a random initial row with a width of 200 cells, with a fill
23586 cellauto=ratio=2/3:s=200x200
23590 Create a pattern generated by rule 18 starting by a single alive cell
23591 centered on an initial row with width 100:
23593 cellauto=p=@@:s=100x400:full=0:rule=18
23597 Specify a more elaborated initial pattern:
23599 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
23604 @anchor{coreimagesrc}
23605 @section coreimagesrc
23606 Video source generated on GPU using Apple's CoreImage API on OSX.
23608 This video source is a specialized version of the @ref{coreimage} video filter.
23609 Use a core image generator at the beginning of the applied filterchain to
23610 generate the content.
23612 The coreimagesrc video source accepts the following options:
23614 @item list_generators
23615 List all available generators along with all their respective options as well as
23616 possible minimum and maximum values along with the default values.
23618 list_generators=true
23622 Specify the size of the sourced video. For the syntax of this option, check the
23623 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23624 The default value is @code{320x240}.
23627 Specify the frame rate of the sourced video, as the number of frames
23628 generated per second. It has to be a string in the format
23629 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23630 number or a valid video frame rate abbreviation. The default value is
23634 Set the sample aspect ratio of the sourced video.
23637 Set the duration of the sourced video. See
23638 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23639 for the accepted syntax.
23641 If not specified, or the expressed duration is negative, the video is
23642 supposed to be generated forever.
23645 Additionally, all options of the @ref{coreimage} video filter are accepted.
23646 A complete filterchain can be used for further processing of the
23647 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
23648 and examples for details.
23650 @subsection Examples
23655 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
23656 given as complete and escaped command-line for Apple's standard bash shell:
23658 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
23660 This example is equivalent to the QRCode example of @ref{coreimage} without the
23661 need for a nullsrc video source.
23666 Generate several gradients.
23670 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23671 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23674 Set frame rate, expressed as number of frames per second. Default
23677 @item c0, c1, c2, c3, c4, c5, c6, c7
23678 Set 8 colors. Default values for colors is to pick random one.
23680 @item x0, y0, y0, y1
23681 Set gradient line source and destination points. If negative or out of range, random ones
23685 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
23688 Set seed for picking gradient line points.
23691 Set the duration of the sourced video. See
23692 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23693 for the accepted syntax.
23695 If not specified, or the expressed duration is negative, the video is
23696 supposed to be generated forever.
23699 Set speed of gradients rotation.
23703 @section mandelbrot
23705 Generate a Mandelbrot set fractal, and progressively zoom towards the
23706 point specified with @var{start_x} and @var{start_y}.
23708 This source accepts the following options:
23713 Set the terminal pts value. Default value is 400.
23716 Set the terminal scale value.
23717 Must be a floating point value. Default value is 0.3.
23720 Set the inner coloring mode, that is the algorithm used to draw the
23721 Mandelbrot fractal internal region.
23723 It shall assume one of the following values:
23728 Show time until convergence.
23730 Set color based on point closest to the origin of the iterations.
23735 Default value is @var{mincol}.
23738 Set the bailout value. Default value is 10.0.
23741 Set the maximum of iterations performed by the rendering
23742 algorithm. Default value is 7189.
23745 Set outer coloring mode.
23746 It shall assume one of following values:
23748 @item iteration_count
23749 Set iteration count mode.
23750 @item normalized_iteration_count
23751 set normalized iteration count mode.
23753 Default value is @var{normalized_iteration_count}.
23756 Set frame rate, expressed as number of frames per second. Default
23760 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23761 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23764 Set the initial scale value. Default value is 3.0.
23767 Set the initial x position. Must be a floating point value between
23768 -100 and 100. Default value is -0.743643887037158704752191506114774.
23771 Set the initial y position. Must be a floating point value between
23772 -100 and 100. Default value is -0.131825904205311970493132056385139.
23777 Generate various test patterns, as generated by the MPlayer test filter.
23779 The size of the generated video is fixed, and is 256x256.
23780 This source is useful in particular for testing encoding features.
23782 This source accepts the following options:
23787 Specify the frame rate of the sourced video, as the number of frames
23788 generated per second. It has to be a string in the format
23789 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23790 number or a valid video frame rate abbreviation. The default value is
23794 Set the duration of the sourced video. See
23795 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23796 for the accepted syntax.
23798 If not specified, or the expressed duration is negative, the video is
23799 supposed to be generated forever.
23803 Set the number or the name of the test to perform. Supported tests are:
23817 @item max_frames, m
23818 Set the maximum number of frames generated for each test, default value is 30.
23822 Default value is "all", which will cycle through the list of all tests.
23827 mptestsrc=t=dc_luma
23830 will generate a "dc_luma" test pattern.
23832 @section frei0r_src
23834 Provide a frei0r source.
23836 To enable compilation of this filter you need to install the frei0r
23837 header and configure FFmpeg with @code{--enable-frei0r}.
23839 This source accepts the following parameters:
23844 The size of the video to generate. For the syntax of this option, check the
23845 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23848 The framerate of the generated video. It may be a string of the form
23849 @var{num}/@var{den} or a frame rate abbreviation.
23852 The name to the frei0r source to load. For more information regarding frei0r and
23853 how to set the parameters, read the @ref{frei0r} section in the video filters
23856 @item filter_params
23857 A '|'-separated list of parameters to pass to the frei0r source.
23861 For example, to generate a frei0r partik0l source with size 200x200
23862 and frame rate 10 which is overlaid on the overlay filter main input:
23864 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
23869 Generate a life pattern.
23871 This source is based on a generalization of John Conway's life game.
23873 The sourced input represents a life grid, each pixel represents a cell
23874 which can be in one of two possible states, alive or dead. Every cell
23875 interacts with its eight neighbours, which are the cells that are
23876 horizontally, vertically, or diagonally adjacent.
23878 At each interaction the grid evolves according to the adopted rule,
23879 which specifies the number of neighbor alive cells which will make a
23880 cell stay alive or born. The @option{rule} option allows one to specify
23883 This source accepts the following options:
23887 Set the file from which to read the initial grid state. In the file,
23888 each non-whitespace character is considered an alive cell, and newline
23889 is used to delimit the end of each row.
23891 If this option is not specified, the initial grid is generated
23895 Set the video rate, that is the number of frames generated per second.
23898 @item random_fill_ratio, ratio
23899 Set the random fill ratio for the initial random grid. It is a
23900 floating point number value ranging from 0 to 1, defaults to 1/PHI.
23901 It is ignored when a file is specified.
23903 @item random_seed, seed
23904 Set the seed for filling the initial random grid, must be an integer
23905 included between 0 and UINT32_MAX. If not specified, or if explicitly
23906 set to -1, the filter will try to use a good random seed on a best
23912 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
23913 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
23914 @var{NS} specifies the number of alive neighbor cells which make a
23915 live cell stay alive, and @var{NB} the number of alive neighbor cells
23916 which make a dead cell to become alive (i.e. to "born").
23917 "s" and "b" can be used in place of "S" and "B", respectively.
23919 Alternatively a rule can be specified by an 18-bits integer. The 9
23920 high order bits are used to encode the next cell state if it is alive
23921 for each number of neighbor alive cells, the low order bits specify
23922 the rule for "borning" new cells. Higher order bits encode for an
23923 higher number of neighbor cells.
23924 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
23925 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
23927 Default value is "S23/B3", which is the original Conway's game of life
23928 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
23929 cells, and will born a new cell if there are three alive cells around
23933 Set the size of the output video. For the syntax of this option, check the
23934 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23936 If @option{filename} is specified, the size is set by default to the
23937 same size of the input file. If @option{size} is set, it must contain
23938 the size specified in the input file, and the initial grid defined in
23939 that file is centered in the larger resulting area.
23941 If a filename is not specified, the size value defaults to "320x240"
23942 (used for a randomly generated initial grid).
23945 If set to 1, stitch the left and right grid edges together, and the
23946 top and bottom edges also. Defaults to 1.
23949 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
23950 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
23951 value from 0 to 255.
23954 Set the color of living (or new born) cells.
23957 Set the color of dead cells. If @option{mold} is set, this is the first color
23958 used to represent a dead cell.
23961 Set mold color, for definitely dead and moldy cells.
23963 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
23964 ffmpeg-utils manual,ffmpeg-utils}.
23967 @subsection Examples
23971 Read a grid from @file{pattern}, and center it on a grid of size
23974 life=f=pattern:s=300x300
23978 Generate a random grid of size 200x200, with a fill ratio of 2/3:
23980 life=ratio=2/3:s=200x200
23984 Specify a custom rule for evolving a randomly generated grid:
23990 Full example with slow death effect (mold) using @command{ffplay}:
23992 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
23999 @anchor{haldclutsrc}
24002 @anchor{pal100bars}
24003 @anchor{rgbtestsrc}
24005 @anchor{smptehdbars}
24008 @anchor{yuvtestsrc}
24009 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
24011 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
24013 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
24015 The @code{color} source provides an uniformly colored input.
24017 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
24018 @ref{haldclut} filter.
24020 The @code{nullsrc} source returns unprocessed video frames. It is
24021 mainly useful to be employed in analysis / debugging tools, or as the
24022 source for filters which ignore the input data.
24024 The @code{pal75bars} source generates a color bars pattern, based on
24025 EBU PAL recommendations with 75% color levels.
24027 The @code{pal100bars} source generates a color bars pattern, based on
24028 EBU PAL recommendations with 100% color levels.
24030 The @code{rgbtestsrc} source generates an RGB test pattern useful for
24031 detecting RGB vs BGR issues. You should see a red, green and blue
24032 stripe from top to bottom.
24034 The @code{smptebars} source generates a color bars pattern, based on
24035 the SMPTE Engineering Guideline EG 1-1990.
24037 The @code{smptehdbars} source generates a color bars pattern, based on
24038 the SMPTE RP 219-2002.
24040 The @code{testsrc} source generates a test video pattern, showing a
24041 color pattern, a scrolling gradient and a timestamp. This is mainly
24042 intended for testing purposes.
24044 The @code{testsrc2} source is similar to testsrc, but supports more
24045 pixel formats instead of just @code{rgb24}. This allows using it as an
24046 input for other tests without requiring a format conversion.
24048 The @code{yuvtestsrc} source generates an YUV test pattern. You should
24049 see a y, cb and cr stripe from top to bottom.
24051 The sources accept the following parameters:
24056 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
24057 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
24058 pixels to be used as identity matrix for 3D lookup tables. Each component is
24059 coded on a @code{1/(N*N)} scale.
24062 Specify the color of the source, only available in the @code{color}
24063 source. For the syntax of this option, check the
24064 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
24067 Specify the size of the sourced video. For the syntax of this option, check the
24068 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24069 The default value is @code{320x240}.
24071 This option is not available with the @code{allrgb}, @code{allyuv}, and
24072 @code{haldclutsrc} filters.
24075 Specify the frame rate of the sourced video, as the number of frames
24076 generated per second. It has to be a string in the format
24077 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
24078 number or a valid video frame rate abbreviation. The default value is
24082 Set the duration of the sourced video. See
24083 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
24084 for the accepted syntax.
24086 If not specified, or the expressed duration is negative, the video is
24087 supposed to be generated forever.
24089 Since the frame rate is used as time base, all frames including the last one
24090 will have their full duration. If the specified duration is not a multiple
24091 of the frame duration, it will be rounded up.
24094 Set the sample aspect ratio of the sourced video.
24097 Specify the alpha (opacity) of the background, only available in the
24098 @code{testsrc2} source. The value must be between 0 (fully transparent) and
24099 255 (fully opaque, the default).
24102 Set the number of decimals to show in the timestamp, only available in the
24103 @code{testsrc} source.
24105 The displayed timestamp value will correspond to the original
24106 timestamp value multiplied by the power of 10 of the specified
24107 value. Default value is 0.
24110 @subsection Examples
24114 Generate a video with a duration of 5.3 seconds, with size
24115 176x144 and a frame rate of 10 frames per second:
24117 testsrc=duration=5.3:size=qcif:rate=10
24121 The following graph description will generate a red source
24122 with an opacity of 0.2, with size "qcif" and a frame rate of 10
24125 color=c=red@@0.2:s=qcif:r=10
24129 If the input content is to be ignored, @code{nullsrc} can be used. The
24130 following command generates noise in the luminance plane by employing
24131 the @code{geq} filter:
24133 nullsrc=s=256x256, geq=random(1)*255:128:128
24137 @subsection Commands
24139 The @code{color} source supports the following commands:
24143 Set the color of the created image. Accepts the same syntax of the
24144 corresponding @option{color} option.
24149 Generate video using an OpenCL program.
24154 OpenCL program source file.
24157 Kernel name in program.
24160 Size of frames to generate. This must be set.
24163 Pixel format to use for the generated frames. This must be set.
24166 Number of frames generated every second. Default value is '25'.
24170 For details of how the program loading works, see the @ref{program_opencl}
24177 Generate a colour ramp by setting pixel values from the position of the pixel
24178 in the output image. (Note that this will work with all pixel formats, but
24179 the generated output will not be the same.)
24181 __kernel void ramp(__write_only image2d_t dst,
24182 unsigned int index)
24184 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24187 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
24189 write_imagef(dst, loc, val);
24194 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
24196 __kernel void sierpinski_carpet(__write_only image2d_t dst,
24197 unsigned int index)
24199 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24201 float4 value = 0.0f;
24202 int x = loc.x + index;
24203 int y = loc.y + index;
24204 while (x > 0 || y > 0) {
24205 if (x % 3 == 1 && y % 3 == 1) {
24213 write_imagef(dst, loc, value);
24219 @section sierpinski
24221 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
24223 This source accepts the following options:
24227 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
24228 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
24231 Set frame rate, expressed as number of frames per second. Default
24235 Set seed which is used for random panning.
24238 Set max jump for single pan destination. Allowed range is from 1 to 10000.
24241 Set fractal type, can be default @code{carpet} or @code{triangle}.
24244 @c man end VIDEO SOURCES
24246 @chapter Video Sinks
24247 @c man begin VIDEO SINKS
24249 Below is a description of the currently available video sinks.
24251 @section buffersink
24253 Buffer video frames, and make them available to the end of the filter
24256 This sink is mainly intended for programmatic use, in particular
24257 through the interface defined in @file{libavfilter/buffersink.h}
24258 or the options system.
24260 It accepts a pointer to an AVBufferSinkContext structure, which
24261 defines the incoming buffers' formats, to be passed as the opaque
24262 parameter to @code{avfilter_init_filter} for initialization.
24266 Null video sink: do absolutely nothing with the input video. It is
24267 mainly useful as a template and for use in analysis / debugging
24270 @c man end VIDEO SINKS
24272 @chapter Multimedia Filters
24273 @c man begin MULTIMEDIA FILTERS
24275 Below is a description of the currently available multimedia filters.
24279 Convert input audio to a video output, displaying the audio bit scope.
24281 The filter accepts the following options:
24285 Set frame rate, expressed as number of frames per second. Default
24289 Specify the video size for the output. For the syntax of this option, check the
24290 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24291 Default value is @code{1024x256}.
24294 Specify list of colors separated by space or by '|' which will be used to
24295 draw channels. Unrecognized or missing colors will be replaced
24299 @section adrawgraph
24300 Draw a graph using input audio metadata.
24302 See @ref{drawgraph}
24304 @section agraphmonitor
24306 See @ref{graphmonitor}.
24308 @section ahistogram
24310 Convert input audio to a video output, displaying the volume histogram.
24312 The filter accepts the following options:
24316 Specify how histogram is calculated.
24318 It accepts the following values:
24321 Use single histogram for all channels.
24323 Use separate histogram for each channel.
24325 Default is @code{single}.
24328 Set frame rate, expressed as number of frames per second. Default
24332 Specify the video size for the output. For the syntax of this option, check the
24333 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24334 Default value is @code{hd720}.
24339 It accepts the following values:
24350 reverse logarithmic
24352 Default is @code{log}.
24355 Set amplitude scale.
24357 It accepts the following values:
24364 Default is @code{log}.
24367 Set how much frames to accumulate in histogram.
24368 Default is 1. Setting this to -1 accumulates all frames.
24371 Set histogram ratio of window height.
24374 Set sonogram sliding.
24376 It accepts the following values:
24379 replace old rows with new ones.
24381 scroll from top to bottom.
24383 Default is @code{replace}.
24386 @section aphasemeter
24388 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
24389 representing mean phase of current audio frame. A video output can also be produced and is
24390 enabled by default. The audio is passed through as first output.
24392 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
24393 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
24394 and @code{1} means channels are in phase.
24396 The filter accepts the following options, all related to its video output:
24400 Set the output frame rate. Default value is @code{25}.
24403 Set the video size for the output. For the syntax of this option, check the
24404 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24405 Default value is @code{800x400}.
24410 Specify the red, green, blue contrast. Default values are @code{2},
24411 @code{7} and @code{1}.
24412 Allowed range is @code{[0, 255]}.
24415 Set color which will be used for drawing median phase. If color is
24416 @code{none} which is default, no median phase value will be drawn.
24419 Enable video output. Default is enabled.
24422 @subsection phasing detection
24424 The filter also detects out of phase and mono sequences in stereo streams.
24425 It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
24427 The filter accepts the following options for this detection:
24431 Enable mono and out of phase detection. Default is disabled.
24434 Set phase tolerance for mono detection, in amplitude ratio. Default is @code{0}.
24435 Allowed range is @code{[0, 1]}.
24438 Set angle threshold for out of phase detection, in degree. Default is @code{170}.
24439 Allowed range is @code{[90, 180]}.
24442 Set mono or out of phase duration until notification, expressed in seconds. Default is @code{2}.
24445 @subsection Examples
24449 Complete example with @command{ffmpeg} to detect 1 second of mono with 0.001 phase tolerance:
24451 ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
24455 @section avectorscope
24457 Convert input audio to a video output, representing the audio vector
24460 The filter is used to measure the difference between channels of stereo
24461 audio stream. A monaural signal, consisting of identical left and right
24462 signal, results in straight vertical line. Any stereo separation is visible
24463 as a deviation from this line, creating a Lissajous figure.
24464 If the straight (or deviation from it) but horizontal line appears this
24465 indicates that the left and right channels are out of phase.
24467 The filter accepts the following options:
24471 Set the vectorscope mode.
24473 Available values are:
24476 Lissajous rotated by 45 degrees.
24479 Same as above but not rotated.
24482 Shape resembling half of circle.
24485 Default value is @samp{lissajous}.
24488 Set the video size for the output. For the syntax of this option, check the
24489 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24490 Default value is @code{400x400}.
24493 Set the output frame rate. Default value is @code{25}.
24499 Specify the red, green, blue and alpha contrast. Default values are @code{40},
24500 @code{160}, @code{80} and @code{255}.
24501 Allowed range is @code{[0, 255]}.
24507 Specify the red, green, blue and alpha fade. Default values are @code{15},
24508 @code{10}, @code{5} and @code{5}.
24509 Allowed range is @code{[0, 255]}.
24512 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
24513 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
24516 Set the vectorscope drawing mode.
24518 Available values are:
24521 Draw dot for each sample.
24524 Draw line between previous and current sample.
24527 Default value is @samp{dot}.
24530 Specify amplitude scale of audio samples.
24532 Available values are:
24548 Swap left channel axis with right channel axis.
24558 Mirror only x axis.
24561 Mirror only y axis.
24569 @subsection Examples
24573 Complete example using @command{ffplay}:
24575 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
24576 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
24580 @section bench, abench
24582 Benchmark part of a filtergraph.
24584 The filter accepts the following options:
24588 Start or stop a timer.
24590 Available values are:
24593 Get the current time, set it as frame metadata (using the key
24594 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
24597 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
24598 the input frame metadata to get the time difference. Time difference, average,
24599 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
24600 @code{min}) are then printed. The timestamps are expressed in seconds.
24604 @subsection Examples
24608 Benchmark @ref{selectivecolor} filter:
24610 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
24616 Concatenate audio and video streams, joining them together one after the
24619 The filter works on segments of synchronized video and audio streams. All
24620 segments must have the same number of streams of each type, and that will
24621 also be the number of streams at output.
24623 The filter accepts the following options:
24628 Set the number of segments. Default is 2.
24631 Set the number of output video streams, that is also the number of video
24632 streams in each segment. Default is 1.
24635 Set the number of output audio streams, that is also the number of audio
24636 streams in each segment. Default is 0.
24639 Activate unsafe mode: do not fail if segments have a different format.
24643 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
24644 @var{a} audio outputs.
24646 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
24647 segment, in the same order as the outputs, then the inputs for the second
24650 Related streams do not always have exactly the same duration, for various
24651 reasons including codec frame size or sloppy authoring. For that reason,
24652 related synchronized streams (e.g. a video and its audio track) should be
24653 concatenated at once. The concat filter will use the duration of the longest
24654 stream in each segment (except the last one), and if necessary pad shorter
24655 audio streams with silence.
24657 For this filter to work correctly, all segments must start at timestamp 0.
24659 All corresponding streams must have the same parameters in all segments; the
24660 filtering system will automatically select a common pixel format for video
24661 streams, and a common sample format, sample rate and channel layout for
24662 audio streams, but other settings, such as resolution, must be converted
24663 explicitly by the user.
24665 Different frame rates are acceptable but will result in variable frame rate
24666 at output; be sure to configure the output file to handle it.
24668 @subsection Examples
24672 Concatenate an opening, an episode and an ending, all in bilingual version
24673 (video in stream 0, audio in streams 1 and 2):
24675 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
24676 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
24677 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
24678 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
24682 Concatenate two parts, handling audio and video separately, using the
24683 (a)movie sources, and adjusting the resolution:
24685 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
24686 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
24687 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
24689 Note that a desync will happen at the stitch if the audio and video streams
24690 do not have exactly the same duration in the first file.
24694 @subsection Commands
24696 This filter supports the following commands:
24699 Close the current segment and step to the next one
24705 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
24706 level. By default, it logs a message at a frequency of 10Hz with the
24707 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
24708 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
24710 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
24711 sample format is double-precision floating point. The input stream will be converted to
24712 this specification, if needed. Users may need to insert aformat and/or aresample filters
24713 after this filter to obtain the original parameters.
24715 The filter also has a video output (see the @var{video} option) with a real
24716 time graph to observe the loudness evolution. The graphic contains the logged
24717 message mentioned above, so it is not printed anymore when this option is set,
24718 unless the verbose logging is set. The main graphing area contains the
24719 short-term loudness (3 seconds of analysis), and the gauge on the right is for
24720 the momentary loudness (400 milliseconds), but can optionally be configured
24721 to instead display short-term loudness (see @var{gauge}).
24723 The green area marks a +/- 1LU target range around the target loudness
24724 (-23LUFS by default, unless modified through @var{target}).
24726 More information about the Loudness Recommendation EBU R128 on
24727 @url{http://tech.ebu.ch/loudness}.
24729 The filter accepts the following options:
24734 Activate the video output. The audio stream is passed unchanged whether this
24735 option is set or no. The video stream will be the first output stream if
24736 activated. Default is @code{0}.
24739 Set the video size. This option is for video only. For the syntax of this
24741 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24742 Default and minimum resolution is @code{640x480}.
24745 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
24746 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
24747 other integer value between this range is allowed.
24750 Set metadata injection. If set to @code{1}, the audio input will be segmented
24751 into 100ms output frames, each of them containing various loudness information
24752 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
24754 Default is @code{0}.
24757 Force the frame logging level.
24759 Available values are:
24762 information logging level
24764 verbose logging level
24767 By default, the logging level is set to @var{info}. If the @option{video} or
24768 the @option{metadata} options are set, it switches to @var{verbose}.
24773 Available modes can be cumulated (the option is a @code{flag} type). Possible
24777 Disable any peak mode (default).
24779 Enable sample-peak mode.
24781 Simple peak mode looking for the higher sample value. It logs a message
24782 for sample-peak (identified by @code{SPK}).
24784 Enable true-peak mode.
24786 If enabled, the peak lookup is done on an over-sampled version of the input
24787 stream for better peak accuracy. It logs a message for true-peak.
24788 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
24789 This mode requires a build with @code{libswresample}.
24793 Treat mono input files as "dual mono". If a mono file is intended for playback
24794 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
24795 If set to @code{true}, this option will compensate for this effect.
24796 Multi-channel input files are not affected by this option.
24799 Set a specific pan law to be used for the measurement of dual mono files.
24800 This parameter is optional, and has a default value of -3.01dB.
24803 Set a specific target level (in LUFS) used as relative zero in the visualization.
24804 This parameter is optional and has a default value of -23LUFS as specified
24805 by EBU R128. However, material published online may prefer a level of -16LUFS
24806 (e.g. for use with podcasts or video platforms).
24809 Set the value displayed by the gauge. Valid values are @code{momentary} and s
24810 @code{shortterm}. By default the momentary value will be used, but in certain
24811 scenarios it may be more useful to observe the short term value instead (e.g.
24815 Sets the display scale for the loudness. Valid parameters are @code{absolute}
24816 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
24817 video output, not the summary or continuous log output.
24820 @subsection Examples
24824 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
24826 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
24830 Run an analysis with @command{ffmpeg}:
24832 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
24836 @section interleave, ainterleave
24838 Temporally interleave frames from several inputs.
24840 @code{interleave} works with video inputs, @code{ainterleave} with audio.
24842 These filters read frames from several inputs and send the oldest
24843 queued frame to the output.
24845 Input streams must have well defined, monotonically increasing frame
24848 In order to submit one frame to output, these filters need to enqueue
24849 at least one frame for each input, so they cannot work in case one
24850 input is not yet terminated and will not receive incoming frames.
24852 For example consider the case when one input is a @code{select} filter
24853 which always drops input frames. The @code{interleave} filter will keep
24854 reading from that input, but it will never be able to send new frames
24855 to output until the input sends an end-of-stream signal.
24857 Also, depending on inputs synchronization, the filters will drop
24858 frames in case one input receives more frames than the other ones, and
24859 the queue is already filled.
24861 These filters accept the following options:
24865 Set the number of different inputs, it is 2 by default.
24868 How to determine the end-of-stream.
24872 The duration of the longest input. (default)
24875 The duration of the shortest input.
24878 The duration of the first input.
24883 @subsection Examples
24887 Interleave frames belonging to different streams using @command{ffmpeg}:
24889 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
24893 Add flickering blur effect:
24895 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
24899 @section metadata, ametadata
24901 Manipulate frame metadata.
24903 This filter accepts the following options:
24907 Set mode of operation of the filter.
24909 Can be one of the following:
24913 If both @code{value} and @code{key} is set, select frames
24914 which have such metadata. If only @code{key} is set, select
24915 every frame that has such key in metadata.
24918 Add new metadata @code{key} and @code{value}. If key is already available
24922 Modify value of already present key.
24925 If @code{value} is set, delete only keys that have such value.
24926 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
24930 Print key and its value if metadata was found. If @code{key} is not set print all
24931 metadata values available in frame.
24935 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
24938 Set metadata value which will be used. This option is mandatory for
24939 @code{modify} and @code{add} mode.
24942 Which function to use when comparing metadata value and @code{value}.
24944 Can be one of following:
24948 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
24951 Values are interpreted as strings, returns true if metadata value starts with
24952 the @code{value} option string.
24955 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
24958 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
24961 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
24964 Values are interpreted as floats, returns true if expression from option @code{expr}
24968 Values are interpreted as strings, returns true if metadata value ends with
24969 the @code{value} option string.
24973 Set expression which is used when @code{function} is set to @code{expr}.
24974 The expression is evaluated through the eval API and can contain the following
24979 Float representation of @code{value} from metadata key.
24982 Float representation of @code{value} as supplied by user in @code{value} option.
24986 If specified in @code{print} mode, output is written to the named file. Instead of
24987 plain filename any writable url can be specified. Filename ``-'' is a shorthand
24988 for standard output. If @code{file} option is not set, output is written to the log
24989 with AV_LOG_INFO loglevel.
24992 Reduces buffering in print mode when output is written to a URL set using @var{file}.
24996 @subsection Examples
25000 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
25003 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
25006 Print silencedetect output to file @file{metadata.txt}.
25008 silencedetect,ametadata=mode=print:file=metadata.txt
25011 Direct all metadata to a pipe with file descriptor 4.
25013 metadata=mode=print:file='pipe\:4'
25017 @section perms, aperms
25019 Set read/write permissions for the output frames.
25021 These filters are mainly aimed at developers to test direct path in the
25022 following filter in the filtergraph.
25024 The filters accept the following options:
25028 Select the permissions mode.
25030 It accepts the following values:
25033 Do nothing. This is the default.
25035 Set all the output frames read-only.
25037 Set all the output frames directly writable.
25039 Make the frame read-only if writable, and writable if read-only.
25041 Set each output frame read-only or writable randomly.
25045 Set the seed for the @var{random} mode, must be an integer included between
25046 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
25047 @code{-1}, the filter will try to use a good random seed on a best effort
25051 Note: in case of auto-inserted filter between the permission filter and the
25052 following one, the permission might not be received as expected in that
25053 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
25054 perms/aperms filter can avoid this problem.
25056 @section realtime, arealtime
25058 Slow down filtering to match real time approximately.
25060 These filters will pause the filtering for a variable amount of time to
25061 match the output rate with the input timestamps.
25062 They are similar to the @option{re} option to @code{ffmpeg}.
25064 They accept the following options:
25068 Time limit for the pauses. Any pause longer than that will be considered
25069 a timestamp discontinuity and reset the timer. Default is 2 seconds.
25071 Speed factor for processing. The value must be a float larger than zero.
25072 Values larger than 1.0 will result in faster than realtime processing,
25073 smaller will slow processing down. The @var{limit} is automatically adapted
25074 accordingly. Default is 1.0.
25076 A processing speed faster than what is possible without these filters cannot
25081 @section select, aselect
25083 Select frames to pass in output.
25085 This filter accepts the following options:
25090 Set expression, which is evaluated for each input frame.
25092 If the expression is evaluated to zero, the frame is discarded.
25094 If the evaluation result is negative or NaN, the frame is sent to the
25095 first output; otherwise it is sent to the output with index
25096 @code{ceil(val)-1}, assuming that the input index starts from 0.
25098 For example a value of @code{1.2} corresponds to the output with index
25099 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
25102 Set the number of outputs. The output to which to send the selected
25103 frame is based on the result of the evaluation. Default value is 1.
25106 The expression can contain the following constants:
25110 The (sequential) number of the filtered frame, starting from 0.
25113 The (sequential) number of the selected frame, starting from 0.
25115 @item prev_selected_n
25116 The sequential number of the last selected frame. It's NAN if undefined.
25119 The timebase of the input timestamps.
25122 The PTS (Presentation TimeStamp) of the filtered video frame,
25123 expressed in @var{TB} units. It's NAN if undefined.
25126 The PTS of the filtered video frame,
25127 expressed in seconds. It's NAN if undefined.
25130 The PTS of the previously filtered video frame. It's NAN if undefined.
25132 @item prev_selected_pts
25133 The PTS of the last previously filtered video frame. It's NAN if undefined.
25135 @item prev_selected_t
25136 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
25139 The PTS of the first video frame in the video. It's NAN if undefined.
25142 The time of the first video frame in the video. It's NAN if undefined.
25144 @item pict_type @emph{(video only)}
25145 The type of the filtered frame. It can assume one of the following
25157 @item interlace_type @emph{(video only)}
25158 The frame interlace type. It can assume one of the following values:
25161 The frame is progressive (not interlaced).
25163 The frame is top-field-first.
25165 The frame is bottom-field-first.
25168 @item consumed_sample_n @emph{(audio only)}
25169 the number of selected samples before the current frame
25171 @item samples_n @emph{(audio only)}
25172 the number of samples in the current frame
25174 @item sample_rate @emph{(audio only)}
25175 the input sample rate
25178 This is 1 if the filtered frame is a key-frame, 0 otherwise.
25181 the position in the file of the filtered frame, -1 if the information
25182 is not available (e.g. for synthetic video)
25184 @item scene @emph{(video only)}
25185 value between 0 and 1 to indicate a new scene; a low value reflects a low
25186 probability for the current frame to introduce a new scene, while a higher
25187 value means the current frame is more likely to be one (see the example below)
25189 @item concatdec_select
25190 The concat demuxer can select only part of a concat input file by setting an
25191 inpoint and an outpoint, but the output packets may not be entirely contained
25192 in the selected interval. By using this variable, it is possible to skip frames
25193 generated by the concat demuxer which are not exactly contained in the selected
25196 This works by comparing the frame pts against the @var{lavf.concat.start_time}
25197 and the @var{lavf.concat.duration} packet metadata values which are also
25198 present in the decoded frames.
25200 The @var{concatdec_select} variable is -1 if the frame pts is at least
25201 start_time and either the duration metadata is missing or the frame pts is less
25202 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
25205 That basically means that an input frame is selected if its pts is within the
25206 interval set by the concat demuxer.
25210 The default value of the select expression is "1".
25212 @subsection Examples
25216 Select all frames in input:
25221 The example above is the same as:
25233 Select only I-frames:
25235 select='eq(pict_type\,I)'
25239 Select one frame every 100:
25241 select='not(mod(n\,100))'
25245 Select only frames contained in the 10-20 time interval:
25247 select=between(t\,10\,20)
25251 Select only I-frames contained in the 10-20 time interval:
25253 select=between(t\,10\,20)*eq(pict_type\,I)
25257 Select frames with a minimum distance of 10 seconds:
25259 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
25263 Use aselect to select only audio frames with samples number > 100:
25265 aselect='gt(samples_n\,100)'
25269 Create a mosaic of the first scenes:
25271 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
25274 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
25278 Send even and odd frames to separate outputs, and compose them:
25280 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
25284 Select useful frames from an ffconcat file which is using inpoints and
25285 outpoints but where the source files are not intra frame only.
25287 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
25291 @section sendcmd, asendcmd
25293 Send commands to filters in the filtergraph.
25295 These filters read commands to be sent to other filters in the
25298 @code{sendcmd} must be inserted between two video filters,
25299 @code{asendcmd} must be inserted between two audio filters, but apart
25300 from that they act the same way.
25302 The specification of commands can be provided in the filter arguments
25303 with the @var{commands} option, or in a file specified by the
25304 @var{filename} option.
25306 These filters accept the following options:
25309 Set the commands to be read and sent to the other filters.
25311 Set the filename of the commands to be read and sent to the other
25315 @subsection Commands syntax
25317 A commands description consists of a sequence of interval
25318 specifications, comprising a list of commands to be executed when a
25319 particular event related to that interval occurs. The occurring event
25320 is typically the current frame time entering or leaving a given time
25323 An interval is specified by the following syntax:
25325 @var{START}[-@var{END}] @var{COMMANDS};
25328 The time interval is specified by the @var{START} and @var{END} times.
25329 @var{END} is optional and defaults to the maximum time.
25331 The current frame time is considered within the specified interval if
25332 it is included in the interval [@var{START}, @var{END}), that is when
25333 the time is greater or equal to @var{START} and is lesser than
25336 @var{COMMANDS} consists of a sequence of one or more command
25337 specifications, separated by ",", relating to that interval. The
25338 syntax of a command specification is given by:
25340 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
25343 @var{FLAGS} is optional and specifies the type of events relating to
25344 the time interval which enable sending the specified command, and must
25345 be a non-null sequence of identifier flags separated by "+" or "|" and
25346 enclosed between "[" and "]".
25348 The following flags are recognized:
25351 The command is sent when the current frame timestamp enters the
25352 specified interval. In other words, the command is sent when the
25353 previous frame timestamp was not in the given interval, and the
25357 The command is sent when the current frame timestamp leaves the
25358 specified interval. In other words, the command is sent when the
25359 previous frame timestamp was in the given interval, and the
25363 The command @var{ARG} is interpreted as expression and result of
25364 expression is passed as @var{ARG}.
25366 The expression is evaluated through the eval API and can contain the following
25371 Original position in the file of the frame, or undefined if undefined
25372 for the current frame.
25375 The presentation timestamp in input.
25378 The count of the input frame for video or audio, starting from 0.
25381 The time in seconds of the current frame.
25384 The start time in seconds of the current command interval.
25387 The end time in seconds of the current command interval.
25390 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
25395 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
25398 @var{TARGET} specifies the target of the command, usually the name of
25399 the filter class or a specific filter instance name.
25401 @var{COMMAND} specifies the name of the command for the target filter.
25403 @var{ARG} is optional and specifies the optional list of argument for
25404 the given @var{COMMAND}.
25406 Between one interval specification and another, whitespaces, or
25407 sequences of characters starting with @code{#} until the end of line,
25408 are ignored and can be used to annotate comments.
25410 A simplified BNF description of the commands specification syntax
25413 @var{COMMAND_FLAG} ::= "enter" | "leave"
25414 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
25415 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
25416 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
25417 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
25418 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
25421 @subsection Examples
25425 Specify audio tempo change at second 4:
25427 asendcmd=c='4.0 atempo tempo 1.5',atempo
25431 Target a specific filter instance:
25433 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
25437 Specify a list of drawtext and hue commands in a file.
25439 # show text in the interval 5-10
25440 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
25441 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
25443 # desaturate the image in the interval 15-20
25444 15.0-20.0 [enter] hue s 0,
25445 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
25447 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
25449 # apply an exponential saturation fade-out effect, starting from time 25
25450 25 [enter] hue s exp(25-t)
25453 A filtergraph allowing to read and process the above command list
25454 stored in a file @file{test.cmd}, can be specified with:
25456 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
25461 @section setpts, asetpts
25463 Change the PTS (presentation timestamp) of the input frames.
25465 @code{setpts} works on video frames, @code{asetpts} on audio frames.
25467 This filter accepts the following options:
25472 The expression which is evaluated for each frame to construct its timestamp.
25476 The expression is evaluated through the eval API and can contain the following
25480 @item FRAME_RATE, FR
25481 frame rate, only defined for constant frame-rate video
25484 The presentation timestamp in input
25487 The count of the input frame for video or the number of consumed samples,
25488 not including the current frame for audio, starting from 0.
25490 @item NB_CONSUMED_SAMPLES
25491 The number of consumed samples, not including the current frame (only
25494 @item NB_SAMPLES, S
25495 The number of samples in the current frame (only audio)
25497 @item SAMPLE_RATE, SR
25498 The audio sample rate.
25501 The PTS of the first frame.
25504 the time in seconds of the first frame
25507 State whether the current frame is interlaced.
25510 the time in seconds of the current frame
25513 original position in the file of the frame, or undefined if undefined
25514 for the current frame
25517 The previous input PTS.
25520 previous input time in seconds
25523 The previous output PTS.
25526 previous output time in seconds
25529 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
25533 The wallclock (RTC) time at the start of the movie in microseconds.
25536 The timebase of the input timestamps.
25540 @subsection Examples
25544 Start counting PTS from zero
25546 setpts=PTS-STARTPTS
25550 Apply fast motion effect:
25556 Apply slow motion effect:
25562 Set fixed rate of 25 frames per second:
25568 Set fixed rate 25 fps with some jitter:
25570 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
25574 Apply an offset of 10 seconds to the input PTS:
25580 Generate timestamps from a "live source" and rebase onto the current timebase:
25582 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
25586 Generate timestamps by counting samples:
25595 Force color range for the output video frame.
25597 The @code{setrange} filter marks the color range property for the
25598 output frames. It does not change the input frame, but only sets the
25599 corresponding property, which affects how the frame is treated by
25602 The filter accepts the following options:
25607 Available values are:
25611 Keep the same color range property.
25613 @item unspecified, unknown
25614 Set the color range as unspecified.
25616 @item limited, tv, mpeg
25617 Set the color range as limited.
25619 @item full, pc, jpeg
25620 Set the color range as full.
25624 @section settb, asettb
25626 Set the timebase to use for the output frames timestamps.
25627 It is mainly useful for testing timebase configuration.
25629 It accepts the following parameters:
25634 The expression which is evaluated into the output timebase.
25638 The value for @option{tb} is an arithmetic expression representing a
25639 rational. The expression can contain the constants "AVTB" (the default
25640 timebase), "intb" (the input timebase) and "sr" (the sample rate,
25641 audio only). Default value is "intb".
25643 @subsection Examples
25647 Set the timebase to 1/25:
25653 Set the timebase to 1/10:
25659 Set the timebase to 1001/1000:
25665 Set the timebase to 2*intb:
25671 Set the default timebase value:
25678 Convert input audio to a video output representing frequency spectrum
25679 logarithmically using Brown-Puckette constant Q transform algorithm with
25680 direct frequency domain coefficient calculation (but the transform itself
25681 is not really constant Q, instead the Q factor is actually variable/clamped),
25682 with musical tone scale, from E0 to D#10.
25684 The filter accepts the following options:
25688 Specify the video size for the output. It must be even. For the syntax of this option,
25689 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25690 Default value is @code{1920x1080}.
25693 Set the output frame rate. Default value is @code{25}.
25696 Set the bargraph height. It must be even. Default value is @code{-1} which
25697 computes the bargraph height automatically.
25700 Set the axis height. It must be even. Default value is @code{-1} which computes
25701 the axis height automatically.
25704 Set the sonogram height. It must be even. Default value is @code{-1} which
25705 computes the sonogram height automatically.
25708 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
25709 instead. Default value is @code{1}.
25711 @item sono_v, volume
25712 Specify the sonogram volume expression. It can contain variables:
25715 the @var{bar_v} evaluated expression
25716 @item frequency, freq, f
25717 the frequency where it is evaluated
25718 @item timeclamp, tc
25719 the value of @var{timeclamp} option
25723 @item a_weighting(f)
25724 A-weighting of equal loudness
25725 @item b_weighting(f)
25726 B-weighting of equal loudness
25727 @item c_weighting(f)
25728 C-weighting of equal loudness.
25730 Default value is @code{16}.
25732 @item bar_v, volume2
25733 Specify the bargraph volume expression. It can contain variables:
25736 the @var{sono_v} evaluated expression
25737 @item frequency, freq, f
25738 the frequency where it is evaluated
25739 @item timeclamp, tc
25740 the value of @var{timeclamp} option
25744 @item a_weighting(f)
25745 A-weighting of equal loudness
25746 @item b_weighting(f)
25747 B-weighting of equal loudness
25748 @item c_weighting(f)
25749 C-weighting of equal loudness.
25751 Default value is @code{sono_v}.
25753 @item sono_g, gamma
25754 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
25755 higher gamma makes the spectrum having more range. Default value is @code{3}.
25756 Acceptable range is @code{[1, 7]}.
25758 @item bar_g, gamma2
25759 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
25763 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
25764 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
25766 @item timeclamp, tc
25767 Specify the transform timeclamp. At low frequency, there is trade-off between
25768 accuracy in time domain and frequency domain. If timeclamp is lower,
25769 event in time domain is represented more accurately (such as fast bass drum),
25770 otherwise event in frequency domain is represented more accurately
25771 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
25774 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
25775 limits future samples by applying asymmetric windowing in time domain, useful
25776 when low latency is required. Accepted range is @code{[0, 1]}.
25779 Specify the transform base frequency. Default value is @code{20.01523126408007475},
25780 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
25783 Specify the transform end frequency. Default value is @code{20495.59681441799654},
25784 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
25787 This option is deprecated and ignored.
25790 Specify the transform length in time domain. Use this option to control accuracy
25791 trade-off between time domain and frequency domain at every frequency sample.
25792 It can contain variables:
25794 @item frequency, freq, f
25795 the frequency where it is evaluated
25796 @item timeclamp, tc
25797 the value of @var{timeclamp} option.
25799 Default value is @code{384*tc/(384+tc*f)}.
25802 Specify the transform count for every video frame. Default value is @code{6}.
25803 Acceptable range is @code{[1, 30]}.
25806 Specify the transform count for every single pixel. Default value is @code{0},
25807 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
25810 Specify font file for use with freetype to draw the axis. If not specified,
25811 use embedded font. Note that drawing with font file or embedded font is not
25812 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
25816 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
25817 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
25821 Specify font color expression. This is arithmetic expression that should return
25822 integer value 0xRRGGBB. It can contain variables:
25824 @item frequency, freq, f
25825 the frequency where it is evaluated
25826 @item timeclamp, tc
25827 the value of @var{timeclamp} option
25832 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
25833 @item r(x), g(x), b(x)
25834 red, green, and blue value of intensity x.
25836 Default value is @code{st(0, (midi(f)-59.5)/12);
25837 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
25838 r(1-ld(1)) + b(ld(1))}.
25841 Specify image file to draw the axis. This option override @var{fontfile} and
25842 @var{fontcolor} option.
25845 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
25846 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
25847 Default value is @code{1}.
25850 Set colorspace. The accepted values are:
25853 Unspecified (default)
25862 BT.470BG or BT.601-6 625
25865 SMPTE-170M or BT.601-6 525
25871 BT.2020 with non-constant luminance
25876 Set spectrogram color scheme. This is list of floating point values with format
25877 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
25878 The default is @code{1|0.5|0|0|0.5|1}.
25882 @subsection Examples
25886 Playing audio while showing the spectrum:
25888 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
25892 Same as above, but with frame rate 30 fps:
25894 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
25898 Playing at 1280x720:
25900 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
25904 Disable sonogram display:
25910 A1 and its harmonics: A1, A2, (near)E3, A3:
25912 ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
25913 asplit[a][out1]; [a] showcqt [out0]'
25917 Same as above, but with more accuracy in frequency domain:
25919 ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
25920 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
25926 bar_v=10:sono_v=bar_v*a_weighting(f)
25930 Custom gamma, now spectrum is linear to the amplitude.
25936 Custom tlength equation:
25938 tc=0.33:tlength='st(0,0.17); 384*tc / (384 / ld(0) + tc*f /(1-ld(0))) + 384*tc / (tc*f / ld(0) + 384 /(1-ld(0)))'
25942 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
25944 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
25948 Custom font using fontconfig:
25950 font='Courier New,Monospace,mono|bold'
25954 Custom frequency range with custom axis using image file:
25956 axisfile=myaxis.png:basefreq=40:endfreq=10000
25962 Convert input audio to video output representing the audio power spectrum.
25963 Audio amplitude is on Y-axis while frequency is on X-axis.
25965 The filter accepts the following options:
25969 Specify size of video. For the syntax of this option, check the
25970 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25971 Default is @code{1024x512}.
25975 This set how each frequency bin will be represented.
25977 It accepts the following values:
25983 Default is @code{bar}.
25986 Set amplitude scale.
25988 It accepts the following values:
26002 Default is @code{log}.
26005 Set frequency scale.
26007 It accepts the following values:
26016 Reverse logarithmic scale.
26018 Default is @code{lin}.
26021 Set window size. Allowed range is from 16 to 65536.
26023 Default is @code{2048}
26026 Set windowing function.
26028 It accepts the following values:
26051 Default is @code{hanning}.
26054 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26055 which means optimal overlap for selected window function will be picked.
26058 Set time averaging. Setting this to 0 will display current maximal peaks.
26059 Default is @code{1}, which means time averaging is disabled.
26062 Specify list of colors separated by space or by '|' which will be used to
26063 draw channel frequencies. Unrecognized or missing colors will be replaced
26067 Set channel display mode.
26069 It accepts the following values:
26074 Default is @code{combined}.
26077 Set minimum amplitude used in @code{log} amplitude scaler.
26080 Set data display mode.
26082 It accepts the following values:
26088 Default is @code{magnitude}.
26091 @section showspatial
26093 Convert stereo input audio to a video output, representing the spatial relationship
26094 between two channels.
26096 The filter accepts the following options:
26100 Specify the video size for the output. For the syntax of this option, check the
26101 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26102 Default value is @code{512x512}.
26105 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
26108 Set window function.
26110 It accepts the following values:
26135 Default value is @code{hann}.
26138 Set ratio of overlap window. Default value is @code{0.5}.
26139 When value is @code{1} overlap is set to recommended size for specific
26140 window function currently used.
26143 @anchor{showspectrum}
26144 @section showspectrum
26146 Convert input audio to a video output, representing the audio frequency
26149 The filter accepts the following options:
26153 Specify the video size for the output. For the syntax of this option, check the
26154 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26155 Default value is @code{640x512}.
26158 Specify how the spectrum should slide along the window.
26160 It accepts the following values:
26163 the samples start again on the left when they reach the right
26165 the samples scroll from right to left
26167 frames are only produced when the samples reach the right
26169 the samples scroll from left to right
26172 Default value is @code{replace}.
26175 Specify display mode.
26177 It accepts the following values:
26180 all channels are displayed in the same row
26182 all channels are displayed in separate rows
26185 Default value is @samp{combined}.
26188 Specify display color mode.
26190 It accepts the following values:
26193 each channel is displayed in a separate color
26195 each channel is displayed using the same color scheme
26197 each channel is displayed using the rainbow color scheme
26199 each channel is displayed using the moreland color scheme
26201 each channel is displayed using the nebulae color scheme
26203 each channel is displayed using the fire color scheme
26205 each channel is displayed using the fiery color scheme
26207 each channel is displayed using the fruit color scheme
26209 each channel is displayed using the cool color scheme
26211 each channel is displayed using the magma color scheme
26213 each channel is displayed using the green color scheme
26215 each channel is displayed using the viridis color scheme
26217 each channel is displayed using the plasma color scheme
26219 each channel is displayed using the cividis color scheme
26221 each channel is displayed using the terrain color scheme
26224 Default value is @samp{channel}.
26227 Specify scale used for calculating intensity color values.
26229 It accepts the following values:
26234 square root, default
26245 Default value is @samp{sqrt}.
26248 Specify frequency scale.
26250 It accepts the following values:
26258 Default value is @samp{lin}.
26261 Set saturation modifier for displayed colors. Negative values provide
26262 alternative color scheme. @code{0} is no saturation at all.
26263 Saturation must be in [-10.0, 10.0] range.
26264 Default value is @code{1}.
26267 Set window function.
26269 It accepts the following values:
26294 Default value is @code{hann}.
26297 Set orientation of time vs frequency axis. Can be @code{vertical} or
26298 @code{horizontal}. Default is @code{vertical}.
26301 Set ratio of overlap window. Default value is @code{0}.
26302 When value is @code{1} overlap is set to recommended size for specific
26303 window function currently used.
26306 Set scale gain for calculating intensity color values.
26307 Default value is @code{1}.
26310 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
26313 Set color rotation, must be in [-1.0, 1.0] range.
26314 Default value is @code{0}.
26317 Set start frequency from which to display spectrogram. Default is @code{0}.
26320 Set stop frequency to which to display spectrogram. Default is @code{0}.
26323 Set upper frame rate limit. Default is @code{auto}, unlimited.
26326 Draw time and frequency axes and legends. Default is disabled.
26329 The usage is very similar to the showwaves filter; see the examples in that
26332 @subsection Examples
26336 Large window with logarithmic color scaling:
26338 showspectrum=s=1280x480:scale=log
26342 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
26344 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
26345 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
26349 @section showspectrumpic
26351 Convert input audio to a single video frame, representing the audio frequency
26354 The filter accepts the following options:
26358 Specify the video size for the output. For the syntax of this option, check the
26359 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26360 Default value is @code{4096x2048}.
26363 Specify display mode.
26365 It accepts the following values:
26368 all channels are displayed in the same row
26370 all channels are displayed in separate rows
26372 Default value is @samp{combined}.
26375 Specify display color mode.
26377 It accepts the following values:
26380 each channel is displayed in a separate color
26382 each channel is displayed using the same color scheme
26384 each channel is displayed using the rainbow color scheme
26386 each channel is displayed using the moreland color scheme
26388 each channel is displayed using the nebulae color scheme
26390 each channel is displayed using the fire color scheme
26392 each channel is displayed using the fiery color scheme
26394 each channel is displayed using the fruit color scheme
26396 each channel is displayed using the cool color scheme
26398 each channel is displayed using the magma color scheme
26400 each channel is displayed using the green color scheme
26402 each channel is displayed using the viridis color scheme
26404 each channel is displayed using the plasma color scheme
26406 each channel is displayed using the cividis color scheme
26408 each channel is displayed using the terrain color scheme
26410 Default value is @samp{intensity}.
26413 Specify scale used for calculating intensity color values.
26415 It accepts the following values:
26420 square root, default
26430 Default value is @samp{log}.
26433 Specify frequency scale.
26435 It accepts the following values:
26443 Default value is @samp{lin}.
26446 Set saturation modifier for displayed colors. Negative values provide
26447 alternative color scheme. @code{0} is no saturation at all.
26448 Saturation must be in [-10.0, 10.0] range.
26449 Default value is @code{1}.
26452 Set window function.
26454 It accepts the following values:
26478 Default value is @code{hann}.
26481 Set orientation of time vs frequency axis. Can be @code{vertical} or
26482 @code{horizontal}. Default is @code{vertical}.
26485 Set scale gain for calculating intensity color values.
26486 Default value is @code{1}.
26489 Draw time and frequency axes and legends. Default is enabled.
26492 Set color rotation, must be in [-1.0, 1.0] range.
26493 Default value is @code{0}.
26496 Set start frequency from which to display spectrogram. Default is @code{0}.
26499 Set stop frequency to which to display spectrogram. Default is @code{0}.
26502 @subsection Examples
26506 Extract an audio spectrogram of a whole audio track
26507 in a 1024x1024 picture using @command{ffmpeg}:
26509 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
26513 @section showvolume
26515 Convert input audio volume to a video output.
26517 The filter accepts the following options:
26524 Set border width, allowed range is [0, 5]. Default is 1.
26527 Set channel width, allowed range is [80, 8192]. Default is 400.
26530 Set channel height, allowed range is [1, 900]. Default is 20.
26533 Set fade, allowed range is [0, 1]. Default is 0.95.
26536 Set volume color expression.
26538 The expression can use the following variables:
26542 Current max volume of channel in dB.
26548 Current channel number, starting from 0.
26552 If set, displays channel names. Default is enabled.
26555 If set, displays volume values. Default is enabled.
26558 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
26559 default is @code{h}.
26562 Set step size, allowed range is [0, 5]. Default is 0, which means
26566 Set background opacity, allowed range is [0, 1]. Default is 0.
26569 Set metering mode, can be peak: @code{p} or rms: @code{r},
26570 default is @code{p}.
26573 Set display scale, can be linear: @code{lin} or log: @code{log},
26574 default is @code{lin}.
26578 If set to > 0., display a line for the max level
26579 in the previous seconds.
26580 default is disabled: @code{0.}
26583 The color of the max line. Use when @code{dm} option is set to > 0.
26584 default is: @code{orange}
26589 Convert input audio to a video output, representing the samples waves.
26591 The filter accepts the following options:
26595 Specify the video size for the output. For the syntax of this option, check the
26596 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26597 Default value is @code{600x240}.
26602 Available values are:
26605 Draw a point for each sample.
26608 Draw a vertical line for each sample.
26611 Draw a point for each sample and a line between them.
26614 Draw a centered vertical line for each sample.
26617 Default value is @code{point}.
26620 Set the number of samples which are printed on the same column. A
26621 larger value will decrease the frame rate. Must be a positive
26622 integer. This option can be set only if the value for @var{rate}
26623 is not explicitly specified.
26626 Set the (approximate) output frame rate. This is done by setting the
26627 option @var{n}. Default value is "25".
26629 @item split_channels
26630 Set if channels should be drawn separately or overlap. Default value is 0.
26633 Set colors separated by '|' which are going to be used for drawing of each channel.
26636 Set amplitude scale.
26638 Available values are:
26656 Set the draw mode. This is mostly useful to set for high @var{n}.
26658 Available values are:
26661 Scale pixel values for each drawn sample.
26664 Draw every sample directly.
26667 Default value is @code{scale}.
26670 @subsection Examples
26674 Output the input file audio and the corresponding video representation
26677 amovie=a.mp3,asplit[out0],showwaves[out1]
26681 Create a synthetic signal and show it with showwaves, forcing a
26682 frame rate of 30 frames per second:
26684 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
26688 @section showwavespic
26690 Convert input audio to a single video frame, representing the samples waves.
26692 The filter accepts the following options:
26696 Specify the video size for the output. For the syntax of this option, check the
26697 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26698 Default value is @code{600x240}.
26700 @item split_channels
26701 Set if channels should be drawn separately or overlap. Default value is 0.
26704 Set colors separated by '|' which are going to be used for drawing of each channel.
26707 Set amplitude scale.
26709 Available values are:
26729 Available values are:
26732 Scale pixel values for each drawn sample.
26735 Draw every sample directly.
26738 Default value is @code{scale}.
26741 Set the filter mode.
26743 Available values are:
26746 Use average samples values for each drawn sample.
26749 Use peak samples values for each drawn sample.
26752 Default value is @code{average}.
26755 @subsection Examples
26759 Extract a channel split representation of the wave form of a whole audio track
26760 in a 1024x800 picture using @command{ffmpeg}:
26762 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
26766 @section sidedata, asidedata
26768 Delete frame side data, or select frames based on it.
26770 This filter accepts the following options:
26774 Set mode of operation of the filter.
26776 Can be one of the following:
26780 Select every frame with side data of @code{type}.
26783 Delete side data of @code{type}. If @code{type} is not set, delete all side
26789 Set side data type used with all modes. Must be set for @code{select} mode. For
26790 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
26791 in @file{libavutil/frame.h}. For example, to choose
26792 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
26796 @section spectrumsynth
26798 Synthesize audio from 2 input video spectrums, first input stream represents
26799 magnitude across time and second represents phase across time.
26800 The filter will transform from frequency domain as displayed in videos back
26801 to time domain as presented in audio output.
26803 This filter is primarily created for reversing processed @ref{showspectrum}
26804 filter outputs, but can synthesize sound from other spectrograms too.
26805 But in such case results are going to be poor if the phase data is not
26806 available, because in such cases phase data need to be recreated, usually
26807 it's just recreated from random noise.
26808 For best results use gray only output (@code{channel} color mode in
26809 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
26810 @code{lin} scale for phase video. To produce phase, for 2nd video, use
26811 @code{data} option. Inputs videos should generally use @code{fullframe}
26812 slide mode as that saves resources needed for decoding video.
26814 The filter accepts the following options:
26818 Specify sample rate of output audio, the sample rate of audio from which
26819 spectrum was generated may differ.
26822 Set number of channels represented in input video spectrums.
26825 Set scale which was used when generating magnitude input spectrum.
26826 Can be @code{lin} or @code{log}. Default is @code{log}.
26829 Set slide which was used when generating inputs spectrums.
26830 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
26831 Default is @code{fullframe}.
26834 Set window function used for resynthesis.
26837 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26838 which means optimal overlap for selected window function will be picked.
26841 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
26842 Default is @code{vertical}.
26845 @subsection Examples
26849 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
26850 then resynthesize videos back to audio with spectrumsynth:
26852 ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=log:overlap=0.875:color=channel:slide=fullframe:data=magnitude -an -c:v rawvideo magnitude.nut
26853 ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=lin:overlap=0.875:color=channel:slide=fullframe:data=phase -an -c:v rawvideo phase.nut
26854 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
26858 @section split, asplit
26860 Split input into several identical outputs.
26862 @code{asplit} works with audio input, @code{split} with video.
26864 The filter accepts a single parameter which specifies the number of outputs. If
26865 unspecified, it defaults to 2.
26867 @subsection Examples
26871 Create two separate outputs from the same input:
26873 [in] split [out0][out1]
26877 To create 3 or more outputs, you need to specify the number of
26880 [in] asplit=3 [out0][out1][out2]
26884 Create two separate outputs from the same input, one cropped and
26887 [in] split [splitout1][splitout2];
26888 [splitout1] crop=100:100:0:0 [cropout];
26889 [splitout2] pad=200:200:100:100 [padout];
26893 Create 5 copies of the input audio with @command{ffmpeg}:
26895 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
26901 Receive commands sent through a libzmq client, and forward them to
26902 filters in the filtergraph.
26904 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
26905 must be inserted between two video filters, @code{azmq} between two
26906 audio filters. Both are capable to send messages to any filter type.
26908 To enable these filters you need to install the libzmq library and
26909 headers and configure FFmpeg with @code{--enable-libzmq}.
26911 For more information about libzmq see:
26912 @url{http://www.zeromq.org/}
26914 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
26915 receives messages sent through a network interface defined by the
26916 @option{bind_address} (or the abbreviation "@option{b}") option.
26917 Default value of this option is @file{tcp://localhost:5555}. You may
26918 want to alter this value to your needs, but do not forget to escape any
26919 ':' signs (see @ref{filtergraph escaping}).
26921 The received message must be in the form:
26923 @var{TARGET} @var{COMMAND} [@var{ARG}]
26926 @var{TARGET} specifies the target of the command, usually the name of
26927 the filter class or a specific filter instance name. The default
26928 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
26929 but you can override this by using the @samp{filter_name@@id} syntax
26930 (see @ref{Filtergraph syntax}).
26932 @var{COMMAND} specifies the name of the command for the target filter.
26934 @var{ARG} is optional and specifies the optional argument list for the
26935 given @var{COMMAND}.
26937 Upon reception, the message is processed and the corresponding command
26938 is injected into the filtergraph. Depending on the result, the filter
26939 will send a reply to the client, adopting the format:
26941 @var{ERROR_CODE} @var{ERROR_REASON}
26945 @var{MESSAGE} is optional.
26947 @subsection Examples
26949 Look at @file{tools/zmqsend} for an example of a zmq client which can
26950 be used to send commands processed by these filters.
26952 Consider the following filtergraph generated by @command{ffplay}.
26953 In this example the last overlay filter has an instance name. All other
26954 filters will have default instance names.
26957 ffplay -dumpgraph 1 -f lavfi "
26958 color=s=100x100:c=red [l];
26959 color=s=100x100:c=blue [r];
26960 nullsrc=s=200x100, zmq [bg];
26961 [bg][l] overlay [bg+l];
26962 [bg+l][r] overlay@@my=x=100 "
26965 To change the color of the left side of the video, the following
26966 command can be used:
26968 echo Parsed_color_0 c yellow | tools/zmqsend
26971 To change the right side:
26973 echo Parsed_color_1 c pink | tools/zmqsend
26976 To change the position of the right side:
26978 echo overlay@@my x 150 | tools/zmqsend
26982 @c man end MULTIMEDIA FILTERS
26984 @chapter Multimedia Sources
26985 @c man begin MULTIMEDIA SOURCES
26987 Below is a description of the currently available multimedia sources.
26991 This is the same as @ref{movie} source, except it selects an audio
26997 Read audio and/or video stream(s) from a movie container.
26999 It accepts the following parameters:
27003 The name of the resource to read (not necessarily a file; it can also be a
27004 device or a stream accessed through some protocol).
27006 @item format_name, f
27007 Specifies the format assumed for the movie to read, and can be either
27008 the name of a container or an input device. If not specified, the
27009 format is guessed from @var{movie_name} or by probing.
27011 @item seek_point, sp
27012 Specifies the seek point in seconds. The frames will be output
27013 starting from this seek point. The parameter is evaluated with
27014 @code{av_strtod}, so the numerical value may be suffixed by an IS
27015 postfix. The default value is "0".
27018 Specifies the streams to read. Several streams can be specified,
27019 separated by "+". The source will then have as many outputs, in the
27020 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
27021 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
27022 respectively the default (best suited) video and audio stream. Default
27023 is "dv", or "da" if the filter is called as "amovie".
27025 @item stream_index, si
27026 Specifies the index of the video stream to read. If the value is -1,
27027 the most suitable video stream will be automatically selected. The default
27028 value is "-1". Deprecated. If the filter is called "amovie", it will select
27029 audio instead of video.
27032 Specifies how many times to read the stream in sequence.
27033 If the value is 0, the stream will be looped infinitely.
27034 Default value is "1".
27036 Note that when the movie is looped the source timestamps are not
27037 changed, so it will generate non monotonically increasing timestamps.
27039 @item discontinuity
27040 Specifies the time difference between frames above which the point is
27041 considered a timestamp discontinuity which is removed by adjusting the later
27045 It allows overlaying a second video on top of the main input of
27046 a filtergraph, as shown in this graph:
27048 input -----------> deltapts0 --> overlay --> output
27051 movie --> scale--> deltapts1 -------+
27053 @subsection Examples
27057 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
27058 on top of the input labelled "in":
27060 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
27061 [in] setpts=PTS-STARTPTS [main];
27062 [main][over] overlay=16:16 [out]
27066 Read from a video4linux2 device, and overlay it on top of the input
27069 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
27070 [in] setpts=PTS-STARTPTS [main];
27071 [main][over] overlay=16:16 [out]
27075 Read the first video stream and the audio stream with id 0x81 from
27076 dvd.vob; the video is connected to the pad named "video" and the audio is
27077 connected to the pad named "audio":
27079 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
27083 @subsection Commands
27085 Both movie and amovie support the following commands:
27088 Perform seek using "av_seek_frame".
27089 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
27092 @var{stream_index}: If stream_index is -1, a default
27093 stream is selected, and @var{timestamp} is automatically converted
27094 from AV_TIME_BASE units to the stream specific time_base.
27096 @var{timestamp}: Timestamp in AVStream.time_base units
27097 or, if no stream is specified, in AV_TIME_BASE units.
27099 @var{flags}: Flags which select direction and seeking mode.
27103 Get movie duration in AV_TIME_BASE units.
27107 @c man end MULTIMEDIA SOURCES