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 Examples
9264 Increase slightly the middle level of blue:
9266 curves=blue='0/0 0.5/0.58 1/1'
9272 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'
9274 Here we obtain the following coordinates for each components:
9277 @code{(0;0.11) (0.42;0.51) (1;0.95)}
9279 @code{(0;0) (0.50;0.48) (1;1)}
9281 @code{(0;0.22) (0.49;0.44) (1;0.80)}
9285 The previous example can also be achieved with the associated built-in preset:
9287 curves=preset=vintage
9297 Use a Photoshop preset and redefine the points of the green component:
9299 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
9303 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
9304 and @command{gnuplot}:
9306 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
9307 gnuplot -p /tmp/curves.plt
9313 Video data analysis filter.
9315 This filter shows hexadecimal pixel values of part of video.
9317 The filter accepts the following options:
9321 Set output video size.
9324 Set x offset from where to pick pixels.
9327 Set y offset from where to pick pixels.
9330 Set scope mode, can be one of the following:
9333 Draw hexadecimal pixel values with white color on black background.
9336 Draw hexadecimal pixel values with input video pixel color on black
9340 Draw hexadecimal pixel values on color background picked from input video,
9341 the text color is picked in such way so its always visible.
9345 Draw rows and columns numbers on left and top of video.
9348 Set background opacity.
9351 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
9354 Set pixel components to display. By default all pixel components are displayed.
9358 Apply Directional blur filter.
9360 The filter accepts the following options:
9364 Set angle of directional blur. Default is @code{45}.
9367 Set radius of directional blur. Default is @code{5}.
9370 Set which planes to filter. By default all planes are filtered.
9373 @subsection Commands
9374 This filter supports same @ref{commands} as options.
9375 The command accepts the same syntax of the corresponding option.
9377 If the specified expression is not valid, it is kept at its current
9382 Denoise frames using 2D DCT (frequency domain filtering).
9384 This filter is not designed for real time.
9386 The filter accepts the following options:
9390 Set the noise sigma constant.
9392 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
9393 coefficient (absolute value) below this threshold with be dropped.
9395 If you need a more advanced filtering, see @option{expr}.
9397 Default is @code{0}.
9400 Set number overlapping pixels for each block. Since the filter can be slow, you
9401 may want to reduce this value, at the cost of a less effective filter and the
9402 risk of various artefacts.
9404 If the overlapping value doesn't permit processing the whole input width or
9405 height, a warning will be displayed and according borders won't be denoised.
9407 Default value is @var{blocksize}-1, which is the best possible setting.
9410 Set the coefficient factor expression.
9412 For each coefficient of a DCT block, this expression will be evaluated as a
9413 multiplier value for the coefficient.
9415 If this is option is set, the @option{sigma} option will be ignored.
9417 The absolute value of the coefficient can be accessed through the @var{c}
9421 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
9422 @var{blocksize}, which is the width and height of the processed blocks.
9424 The default value is @var{3} (8x8) and can be raised to @var{4} for a
9425 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
9426 on the speed processing. Also, a larger block size does not necessarily means a
9430 @subsection Examples
9432 Apply a denoise with a @option{sigma} of @code{4.5}:
9437 The same operation can be achieved using the expression system:
9439 dctdnoiz=e='gte(c, 4.5*3)'
9442 Violent denoise using a block size of @code{16x16}:
9449 Remove banding artifacts from input video.
9450 It works by replacing banded pixels with average value of referenced pixels.
9452 The filter accepts the following options:
9459 Set banding detection threshold for each plane. Default is 0.02.
9460 Valid range is 0.00003 to 0.5.
9461 If difference between current pixel and reference pixel is less than threshold,
9462 it will be considered as banded.
9465 Banding detection range in pixels. Default is 16. If positive, random number
9466 in range 0 to set value will be used. If negative, exact absolute value
9468 The range defines square of four pixels around current pixel.
9471 Set direction in radians from which four pixel will be compared. If positive,
9472 random direction from 0 to set direction will be picked. If negative, exact of
9473 absolute value will be picked. For example direction 0, -PI or -2*PI radians
9474 will pick only pixels on same row and -PI/2 will pick only pixels on same
9478 If enabled, current pixel is compared with average value of all four
9479 surrounding pixels. The default is enabled. If disabled current pixel is
9480 compared with all four surrounding pixels. The pixel is considered banded
9481 if only all four differences with surrounding pixels are less than threshold.
9484 If enabled, current pixel is changed if and only if all pixel components are banded,
9485 e.g. banding detection threshold is triggered for all color components.
9486 The default is disabled.
9489 @subsection Commands
9491 This filter supports the all above options as @ref{commands}.
9495 Remove blocking artifacts from input video.
9497 The filter accepts the following options:
9501 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
9502 This controls what kind of deblocking is applied.
9505 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
9511 Set blocking detection thresholds. Allowed range is 0 to 1.
9512 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
9513 Using higher threshold gives more deblocking strength.
9514 Setting @var{alpha} controls threshold detection at exact edge of block.
9515 Remaining options controls threshold detection near the edge. Each one for
9516 below/above or left/right. Setting any of those to @var{0} disables
9520 Set planes to filter. Default is to filter all available planes.
9523 @subsection Examples
9527 Deblock using weak filter and block size of 4 pixels.
9529 deblock=filter=weak:block=4
9533 Deblock using strong filter, block size of 4 pixels and custom thresholds for
9534 deblocking more edges.
9536 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
9540 Similar as above, but filter only first plane.
9542 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
9546 Similar as above, but filter only second and third plane.
9548 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
9552 @subsection Commands
9554 This filter supports the all above options as @ref{commands}.
9559 Drop duplicated frames at regular intervals.
9561 The filter accepts the following options:
9565 Set the number of frames from which one will be dropped. Setting this to
9566 @var{N} means one frame in every batch of @var{N} frames will be dropped.
9567 Default is @code{5}.
9570 Set the threshold for duplicate detection. If the difference metric for a frame
9571 is less than or equal to this value, then it is declared as duplicate. Default
9575 Set scene change threshold. Default is @code{15}.
9579 Set the size of the x and y-axis blocks used during metric calculations.
9580 Larger blocks give better noise suppression, but also give worse detection of
9581 small movements. Must be a power of two. Default is @code{32}.
9584 Mark main input as a pre-processed input and activate clean source input
9585 stream. This allows the input to be pre-processed with various filters to help
9586 the metrics calculation while keeping the frame selection lossless. When set to
9587 @code{1}, the first stream is for the pre-processed input, and the second
9588 stream is the clean source from where the kept frames are chosen. Default is
9592 Set whether or not chroma is considered in the metric calculations. Default is
9598 Apply 2D deconvolution of video stream in frequency domain using second stream
9601 The filter accepts the following options:
9605 Set which planes to process.
9608 Set which impulse video frames will be processed, can be @var{first}
9609 or @var{all}. Default is @var{all}.
9612 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
9613 and height are not same and not power of 2 or if stream prior to convolving
9617 The @code{deconvolve} filter also supports the @ref{framesync} options.
9621 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
9623 It accepts the following options:
9627 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
9628 @var{rainbows} for cross-color reduction.
9631 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
9634 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
9637 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
9640 Set temporal chroma threshold. Lower values increases reduction of cross-color.
9645 Apply deflate effect to the video.
9647 This filter replaces the pixel by the local(3x3) average by taking into account
9648 only values lower than the pixel.
9650 It accepts the following options:
9657 Limit the maximum change for each plane, default is 65535.
9658 If 0, plane will remain unchanged.
9661 @subsection Commands
9663 This filter supports the all above options as @ref{commands}.
9667 Remove temporal frame luminance variations.
9669 It accepts the following options:
9673 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
9676 Set averaging mode to smooth temporal luminance variations.
9678 Available values are:
9703 Do not actually modify frame. Useful when one only wants metadata.
9708 Remove judder produced by partially interlaced telecined content.
9710 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
9711 source was partially telecined content then the output of @code{pullup,dejudder}
9712 will have a variable frame rate. May change the recorded frame rate of the
9713 container. Aside from that change, this filter will not affect constant frame
9716 The option available in this filter is:
9720 Specify the length of the window over which the judder repeats.
9722 Accepts any integer greater than 1. Useful values are:
9726 If the original was telecined from 24 to 30 fps (Film to NTSC).
9729 If the original was telecined from 25 to 30 fps (PAL to NTSC).
9732 If a mixture of the two.
9735 The default is @samp{4}.
9740 Suppress a TV station logo by a simple interpolation of the surrounding
9741 pixels. Just set a rectangle covering the logo and watch it disappear
9742 (and sometimes something even uglier appear - your mileage may vary).
9744 It accepts the following parameters:
9749 Specify the top left corner coordinates of the logo. They must be
9754 Specify the width and height of the logo to clear. They must be
9758 Specify the thickness of the fuzzy edge of the rectangle (added to
9759 @var{w} and @var{h}). The default value is 1. This option is
9760 deprecated, setting higher values should no longer be necessary and
9764 When set to 1, a green rectangle is drawn on the screen to simplify
9765 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9766 The default value is 0.
9768 The rectangle is drawn on the outermost pixels which will be (partly)
9769 replaced with interpolated values. The values of the next pixels
9770 immediately outside this rectangle in each direction will be used to
9771 compute the interpolated pixel values inside the rectangle.
9775 @subsection Examples
9779 Set a rectangle covering the area with top left corner coordinates 0,0
9780 and size 100x77, and a band of size 10:
9782 delogo=x=0:y=0:w=100:h=77:band=10
9790 Remove the rain in the input image/video by applying the derain methods based on
9791 convolutional neural networks. Supported models:
9795 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9796 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9799 Training as well as model generation scripts are provided in
9800 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9802 Native model files (.model) can be generated from TensorFlow model
9803 files (.pb) by using tools/python/convert.py
9805 The filter accepts the following options:
9809 Specify which filter to use. This option accepts the following values:
9813 Derain filter. To conduct derain filter, you need to use a derain model.
9816 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9818 Default value is @samp{derain}.
9821 Specify which DNN backend to use for model loading and execution. This option accepts
9822 the following values:
9826 Native implementation of DNN loading and execution.
9829 TensorFlow backend. To enable this backend you
9830 need to install the TensorFlow for C library (see
9831 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9832 @code{--enable-libtensorflow}
9834 Default value is @samp{native}.
9837 Set path to model file specifying network architecture and its parameters.
9838 Note that different backends use different file formats. TensorFlow and native
9839 backend can load files for only its format.
9842 It can also be finished with @ref{dnn_processing} filter.
9846 Attempt to fix small changes in horizontal and/or vertical shift. This
9847 filter helps remove camera shake from hand-holding a camera, bumping a
9848 tripod, moving on a vehicle, etc.
9850 The filter accepts the following options:
9858 Specify a rectangular area where to limit the search for motion
9860 If desired the search for motion vectors can be limited to a
9861 rectangular area of the frame defined by its top left corner, width
9862 and height. These parameters have the same meaning as the drawbox
9863 filter which can be used to visualise the position of the bounding
9866 This is useful when simultaneous movement of subjects within the frame
9867 might be confused for camera motion by the motion vector search.
9869 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9870 then the full frame is used. This allows later options to be set
9871 without specifying the bounding box for the motion vector search.
9873 Default - search the whole frame.
9877 Specify the maximum extent of movement in x and y directions in the
9878 range 0-64 pixels. Default 16.
9881 Specify how to generate pixels to fill blanks at the edge of the
9882 frame. Available values are:
9885 Fill zeroes at blank locations
9887 Original image at blank locations
9889 Extruded edge value at blank locations
9891 Mirrored edge at blank locations
9893 Default value is @samp{mirror}.
9896 Specify the blocksize to use for motion search. Range 4-128 pixels,
9900 Specify the contrast threshold for blocks. Only blocks with more than
9901 the specified contrast (difference between darkest and lightest
9902 pixels) will be considered. Range 1-255, default 125.
9905 Specify the search strategy. Available values are:
9908 Set exhaustive search
9910 Set less exhaustive search.
9912 Default value is @samp{exhaustive}.
9915 If set then a detailed log of the motion search is written to the
9922 Remove unwanted contamination of foreground colors, caused by reflected color of
9923 greenscreen or bluescreen.
9925 This filter accepts the following options:
9929 Set what type of despill to use.
9932 Set how spillmap will be generated.
9935 Set how much to get rid of still remaining spill.
9938 Controls amount of red in spill area.
9941 Controls amount of green in spill area.
9942 Should be -1 for greenscreen.
9945 Controls amount of blue in spill area.
9946 Should be -1 for bluescreen.
9949 Controls brightness of spill area, preserving colors.
9952 Modify alpha from generated spillmap.
9955 @subsection Commands
9957 This filter supports the all above options as @ref{commands}.
9961 Apply an exact inverse of the telecine operation. It requires a predefined
9962 pattern specified using the pattern option which must be the same as that passed
9963 to the telecine filter.
9965 This filter accepts the following options:
9974 The default value is @code{top}.
9978 A string of numbers representing the pulldown pattern you wish to apply.
9979 The default value is @code{23}.
9982 A number representing position of the first frame with respect to the telecine
9983 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9988 Apply dilation effect to the video.
9990 This filter replaces the pixel by the local(3x3) maximum.
9992 It accepts the following options:
9999 Limit the maximum change for each plane, default is 65535.
10000 If 0, plane will remain unchanged.
10003 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
10006 Flags to local 3x3 coordinates maps like this:
10013 @subsection Commands
10015 This filter supports the all above options as @ref{commands}.
10019 Displace pixels as indicated by second and third input stream.
10021 It takes three input streams and outputs one stream, the first input is the
10022 source, and second and third input are displacement maps.
10024 The second input specifies how much to displace pixels along the
10025 x-axis, while the third input specifies how much to displace pixels
10027 If one of displacement map streams terminates, last frame from that
10028 displacement map will be used.
10030 Note that once generated, displacements maps can be reused over and over again.
10032 A description of the accepted options follows.
10036 Set displace behavior for pixels that are out of range.
10038 Available values are:
10041 Missing pixels are replaced by black pixels.
10044 Adjacent pixels will spread out to replace missing pixels.
10047 Out of range pixels are wrapped so they point to pixels of other side.
10050 Out of range pixels will be replaced with mirrored pixels.
10052 Default is @samp{smear}.
10056 @subsection Examples
10060 Add ripple effect to rgb input of video size hd720:
10062 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
10066 Add wave effect to rgb input of video size hd720:
10068 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
10072 @anchor{dnn_processing}
10073 @section dnn_processing
10075 Do image processing with deep neural networks. It works together with another filter
10076 which converts the pixel format of the Frame to what the dnn network requires.
10078 The filter accepts the following options:
10082 Specify which DNN backend to use for model loading and execution. This option accepts
10083 the following values:
10087 Native implementation of DNN loading and execution.
10090 TensorFlow backend. To enable this backend you
10091 need to install the TensorFlow for C library (see
10092 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
10093 @code{--enable-libtensorflow}
10096 OpenVINO backend. To enable this backend you
10097 need to build and install the OpenVINO for C library (see
10098 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
10099 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
10100 be needed if the header files and libraries are not installed into system path)
10104 Default value is @samp{native}.
10107 Set path to model file specifying network architecture and its parameters.
10108 Note that different backends use different file formats. TensorFlow, OpenVINO and native
10109 backend can load files for only its format.
10111 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
10114 Set the input name of the dnn network.
10117 Set the output name of the dnn network.
10120 use DNN async execution if set (default: set),
10121 roll back to sync execution if the backend does not support async.
10125 @subsection Examples
10129 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
10131 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
10135 Halve the pixel value of the frame with format gray32f:
10137 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
10141 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
10143 ./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
10147 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
10149 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
10156 Draw a colored box on the input image.
10158 It accepts the following parameters:
10163 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
10167 The expressions which specify the width and height of the box; if 0 they are interpreted as
10168 the input width and height. It defaults to 0.
10171 Specify the color of the box to write. For the general syntax of this option,
10172 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10173 value @code{invert} is used, the box edge color is the same as the
10174 video with inverted luma.
10177 The expression which sets the thickness of the box edge.
10178 A value of @code{fill} will create a filled box. Default value is @code{3}.
10180 See below for the list of accepted constants.
10183 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
10184 will overwrite the video's color and alpha pixels.
10185 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
10188 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10189 following constants:
10193 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10197 horizontal and vertical chroma subsample values. For example for the
10198 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10202 The input width and height.
10205 The input sample aspect ratio.
10209 The x and y offset coordinates where the box is drawn.
10213 The width and height of the drawn box.
10216 The thickness of the drawn box.
10218 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10219 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10223 @subsection Examples
10227 Draw a black box around the edge of the input image:
10233 Draw a box with color red and an opacity of 50%:
10235 drawbox=10:20:200:60:red@@0.5
10238 The previous example can be specified as:
10240 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
10244 Fill the box with pink color:
10246 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
10250 Draw a 2-pixel red 2.40:1 mask:
10252 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
10256 @subsection Commands
10257 This filter supports same commands as options.
10258 The command accepts the same syntax of the corresponding option.
10260 If the specified expression is not valid, it is kept at its current
10265 Draw a graph using input video metadata.
10267 It accepts the following parameters:
10271 Set 1st frame metadata key from which metadata values will be used to draw a graph.
10274 Set 1st foreground color expression.
10277 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
10280 Set 2nd foreground color expression.
10283 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
10286 Set 3rd foreground color expression.
10289 Set 4th frame metadata key from which metadata values will be used to draw a graph.
10292 Set 4th foreground color expression.
10295 Set minimal value of metadata value.
10298 Set maximal value of metadata value.
10301 Set graph background color. Default is white.
10306 Available values for mode is:
10313 Default is @code{line}.
10318 Available values for slide is:
10321 Draw new frame when right border is reached.
10324 Replace old columns with new ones.
10327 Scroll from right to left.
10330 Scroll from left to right.
10333 Draw single picture.
10336 Default is @code{frame}.
10339 Set size of graph video. For the syntax of this option, check the
10340 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
10341 The default value is @code{900x256}.
10344 Set the output frame rate. Default value is @code{25}.
10346 The foreground color expressions can use the following variables:
10349 Minimal value of metadata value.
10352 Maximal value of metadata value.
10355 Current metadata key value.
10358 The color is defined as 0xAABBGGRR.
10361 Example using metadata from @ref{signalstats} filter:
10363 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
10366 Example using metadata from @ref{ebur128} filter:
10368 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
10373 Draw a grid on the input image.
10375 It accepts the following parameters:
10380 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
10384 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
10385 input width and height, respectively, minus @code{thickness}, so image gets
10386 framed. Default to 0.
10389 Specify the color of the grid. For the general syntax of this option,
10390 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10391 value @code{invert} is used, the grid color is the same as the
10392 video with inverted luma.
10395 The expression which sets the thickness of the grid line. Default value is @code{1}.
10397 See below for the list of accepted constants.
10400 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
10401 will overwrite the video's color and alpha pixels.
10402 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
10405 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10406 following constants:
10410 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10414 horizontal and vertical chroma subsample values. For example for the
10415 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10419 The input grid cell width and height.
10422 The input sample aspect ratio.
10426 The x and y coordinates of some point of grid intersection (meant to configure offset).
10430 The width and height of the drawn cell.
10433 The thickness of the drawn cell.
10435 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10436 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10440 @subsection Examples
10444 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
10446 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
10450 Draw a white 3x3 grid with an opacity of 50%:
10452 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
10456 @subsection Commands
10457 This filter supports same commands as options.
10458 The command accepts the same syntax of the corresponding option.
10460 If the specified expression is not valid, it is kept at its current
10466 Draw a text string or text from a specified file on top of a video, using the
10467 libfreetype library.
10469 To enable compilation of this filter, you need to configure FFmpeg with
10470 @code{--enable-libfreetype}.
10471 To enable default font fallback and the @var{font} option you need to
10472 configure FFmpeg with @code{--enable-libfontconfig}.
10473 To enable the @var{text_shaping} option, you need to configure FFmpeg with
10474 @code{--enable-libfribidi}.
10478 It accepts the following parameters:
10483 Used to draw a box around text using the background color.
10484 The value must be either 1 (enable) or 0 (disable).
10485 The default value of @var{box} is 0.
10488 Set the width of the border to be drawn around the box using @var{boxcolor}.
10489 The default value of @var{boxborderw} is 0.
10492 The color to be used for drawing box around text. For the syntax of this
10493 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10495 The default value of @var{boxcolor} is "white".
10498 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
10499 The default value of @var{line_spacing} is 0.
10502 Set the width of the border to be drawn around the text using @var{bordercolor}.
10503 The default value of @var{borderw} is 0.
10506 Set the color to be used for drawing border around text. For the syntax of this
10507 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10509 The default value of @var{bordercolor} is "black".
10512 Select how the @var{text} is expanded. Can be either @code{none},
10513 @code{strftime} (deprecated) or
10514 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
10518 Set a start time for the count. Value is in microseconds. Only applied
10519 in the deprecated strftime expansion mode. To emulate in normal expansion
10520 mode use the @code{pts} function, supplying the start time (in seconds)
10521 as the second argument.
10524 If true, check and fix text coords to avoid clipping.
10527 The color to be used for drawing fonts. For the syntax of this option, check
10528 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10530 The default value of @var{fontcolor} is "black".
10532 @item fontcolor_expr
10533 String which is expanded the same way as @var{text} to obtain dynamic
10534 @var{fontcolor} value. By default this option has empty value and is not
10535 processed. When this option is set, it overrides @var{fontcolor} option.
10538 The font family to be used for drawing text. By default Sans.
10541 The font file to be used for drawing text. The path must be included.
10542 This parameter is mandatory if the fontconfig support is disabled.
10545 Draw the text applying alpha blending. The value can
10546 be a number between 0.0 and 1.0.
10547 The expression accepts the same variables @var{x, y} as well.
10548 The default value is 1.
10549 Please see @var{fontcolor_expr}.
10552 The font size to be used for drawing text.
10553 The default value of @var{fontsize} is 16.
10556 If set to 1, attempt to shape the text (for example, reverse the order of
10557 right-to-left text and join Arabic characters) before drawing it.
10558 Otherwise, just draw the text exactly as given.
10559 By default 1 (if supported).
10561 @item ft_load_flags
10562 The flags to be used for loading the fonts.
10564 The flags map the corresponding flags supported by libfreetype, and are
10565 a combination of the following values:
10572 @item vertical_layout
10573 @item force_autohint
10576 @item ignore_global_advance_width
10578 @item ignore_transform
10580 @item linear_design
10584 Default value is "default".
10586 For more information consult the documentation for the FT_LOAD_*
10590 The color to be used for drawing a shadow behind the drawn text. For the
10591 syntax of this option, check the @ref{color syntax,,"Color" section in the
10592 ffmpeg-utils manual,ffmpeg-utils}.
10594 The default value of @var{shadowcolor} is "black".
10598 The x and y offsets for the text shadow position with respect to the
10599 position of the text. They can be either positive or negative
10600 values. The default value for both is "0".
10603 The starting frame number for the n/frame_num variable. The default value
10607 The size in number of spaces to use for rendering the tab.
10608 Default value is 4.
10611 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
10612 format. It can be used with or without text parameter. @var{timecode_rate}
10613 option must be specified.
10615 @item timecode_rate, rate, r
10616 Set the timecode frame rate (timecode only). Value will be rounded to nearest
10617 integer. Minimum value is "1".
10618 Drop-frame timecode is supported for frame rates 30 & 60.
10621 If set to 1, the output of the timecode option will wrap around at 24 hours.
10622 Default is 0 (disabled).
10625 The text string to be drawn. The text must be a sequence of UTF-8
10626 encoded characters.
10627 This parameter is mandatory if no file is specified with the parameter
10631 A text file containing text to be drawn. The text must be a sequence
10632 of UTF-8 encoded characters.
10634 This parameter is mandatory if no text string is specified with the
10635 parameter @var{text}.
10637 If both @var{text} and @var{textfile} are specified, an error is thrown.
10640 If set to 1, the @var{textfile} will be reloaded before each frame.
10641 Be sure to update it atomically, or it may be read partially, or even fail.
10645 The expressions which specify the offsets where text will be drawn
10646 within the video frame. They are relative to the top/left border of the
10649 The default value of @var{x} and @var{y} is "0".
10651 See below for the list of accepted constants and functions.
10654 The parameters for @var{x} and @var{y} are expressions containing the
10655 following constants and functions:
10659 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
10663 horizontal and vertical chroma subsample values. For example for the
10664 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10667 the height of each text line
10675 @item max_glyph_a, ascent
10676 the maximum distance from the baseline to the highest/upper grid
10677 coordinate used to place a glyph outline point, for all the rendered
10679 It is a positive value, due to the grid's orientation with the Y axis
10682 @item max_glyph_d, descent
10683 the maximum distance from the baseline to the lowest grid coordinate
10684 used to place a glyph outline point, for all the rendered glyphs.
10685 This is a negative value, due to the grid's orientation, with the Y axis
10689 maximum glyph height, that is the maximum height for all the glyphs
10690 contained in the rendered text, it is equivalent to @var{ascent} -
10694 maximum glyph width, that is the maximum width for all the glyphs
10695 contained in the rendered text
10698 the number of input frame, starting from 0
10700 @item rand(min, max)
10701 return a random number included between @var{min} and @var{max}
10704 The input sample aspect ratio.
10707 timestamp expressed in seconds, NAN if the input timestamp is unknown
10710 the height of the rendered text
10713 the width of the rendered text
10717 the x and y offset coordinates where the text is drawn.
10719 These parameters allow the @var{x} and @var{y} expressions to refer
10720 to each other, so you can for example specify @code{y=x/dar}.
10723 A one character description of the current frame's picture type.
10726 The current packet's position in the input file or stream
10727 (in bytes, from the start of the input). A value of -1 indicates
10728 this info is not available.
10731 The current packet's duration, in seconds.
10734 The current packet's size (in bytes).
10737 @anchor{drawtext_expansion}
10738 @subsection Text expansion
10740 If @option{expansion} is set to @code{strftime},
10741 the filter recognizes strftime() sequences in the provided text and
10742 expands them accordingly. Check the documentation of strftime(). This
10743 feature is deprecated.
10745 If @option{expansion} is set to @code{none}, the text is printed verbatim.
10747 If @option{expansion} is set to @code{normal} (which is the default),
10748 the following expansion mechanism is used.
10750 The backslash character @samp{\}, followed by any character, always expands to
10751 the second character.
10753 Sequences of the form @code{%@{...@}} are expanded. The text between the
10754 braces is a function name, possibly followed by arguments separated by ':'.
10755 If the arguments contain special characters or delimiters (':' or '@}'),
10756 they should be escaped.
10758 Note that they probably must also be escaped as the value for the
10759 @option{text} option in the filter argument string and as the filter
10760 argument in the filtergraph description, and possibly also for the shell,
10761 that makes up to four levels of escaping; using a text file avoids these
10764 The following functions are available:
10769 The expression evaluation result.
10771 It must take one argument specifying the expression to be evaluated,
10772 which accepts the same constants and functions as the @var{x} and
10773 @var{y} values. Note that not all constants should be used, for
10774 example the text size is not known when evaluating the expression, so
10775 the constants @var{text_w} and @var{text_h} will have an undefined
10778 @item expr_int_format, eif
10779 Evaluate the expression's value and output as formatted integer.
10781 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10782 The second argument specifies the output format. Allowed values are @samp{x},
10783 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10784 @code{printf} function.
10785 The third parameter is optional and sets the number of positions taken by the output.
10786 It can be used to add padding with zeros from the left.
10789 The time at which the filter is running, expressed in UTC.
10790 It can accept an argument: a strftime() format string.
10793 The time at which the filter is running, expressed in the local time zone.
10794 It can accept an argument: a strftime() format string.
10797 Frame metadata. Takes one or two arguments.
10799 The first argument is mandatory and specifies the metadata key.
10801 The second argument is optional and specifies a default value, used when the
10802 metadata key is not found or empty.
10804 Available metadata can be identified by inspecting entries
10805 starting with TAG included within each frame section
10806 printed by running @code{ffprobe -show_frames}.
10808 String metadata generated in filters leading to
10809 the drawtext filter are also available.
10812 The frame number, starting from 0.
10815 A one character description of the current picture type.
10818 The timestamp of the current frame.
10819 It can take up to three arguments.
10821 The first argument is the format of the timestamp; it defaults to @code{flt}
10822 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10823 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10824 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10825 @code{localtime} stands for the timestamp of the frame formatted as
10826 local time zone time.
10828 The second argument is an offset added to the timestamp.
10830 If the format is set to @code{hms}, a third argument @code{24HH} may be
10831 supplied to present the hour part of the formatted timestamp in 24h format
10834 If the format is set to @code{localtime} or @code{gmtime},
10835 a third argument may be supplied: a strftime() format string.
10836 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10839 @subsection Commands
10841 This filter supports altering parameters via commands:
10844 Alter existing filter parameters.
10846 Syntax for the argument is the same as for filter invocation, e.g.
10849 fontsize=56:fontcolor=green:text='Hello World'
10852 Full filter invocation with sendcmd would look like this:
10855 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10859 If the entire argument can't be parsed or applied as valid values then the filter will
10860 continue with its existing parameters.
10862 @subsection Examples
10866 Draw "Test Text" with font FreeSerif, using the default values for the
10867 optional parameters.
10870 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10874 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10875 and y=50 (counting from the top-left corner of the screen), text is
10876 yellow with a red box around it. Both the text and the box have an
10880 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10881 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10884 Note that the double quotes are not necessary if spaces are not used
10885 within the parameter list.
10888 Show the text at the center of the video frame:
10890 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10894 Show the text at a random position, switching to a new position every 30 seconds:
10896 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)"
10900 Show a text line sliding from right to left in the last row of the video
10901 frame. The file @file{LONG_LINE} is assumed to contain a single line
10904 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10908 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10910 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10914 Draw a single green letter "g", at the center of the input video.
10915 The glyph baseline is placed at half screen height.
10917 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10921 Show text for 1 second every 3 seconds:
10923 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10927 Use fontconfig to set the font. Note that the colons need to be escaped.
10929 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10933 Draw "Test Text" with font size dependent on height of the video.
10935 drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
10939 Print the date of a real-time encoding (see strftime(3)):
10941 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10945 Show text fading in and out (appearing/disappearing):
10948 DS=1.0 # display start
10949 DE=10.0 # display end
10950 FID=1.5 # fade in duration
10951 FOD=5 # fade out duration
10952 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 @}"
10956 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10957 and the @option{fontsize} value are included in the @option{y} offset.
10959 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10960 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10964 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10965 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10966 must have option @option{-export_path_metadata 1} for the special metadata fields
10967 to be available for filters.
10969 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10974 For more information about libfreetype, check:
10975 @url{http://www.freetype.org/}.
10977 For more information about fontconfig, check:
10978 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10980 For more information about libfribidi, check:
10981 @url{http://fribidi.org/}.
10983 @section edgedetect
10985 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10987 The filter accepts the following options:
10992 Set low and high threshold values used by the Canny thresholding
10995 The high threshold selects the "strong" edge pixels, which are then
10996 connected through 8-connectivity with the "weak" edge pixels selected
10997 by the low threshold.
10999 @var{low} and @var{high} threshold values must be chosen in the range
11000 [0,1], and @var{low} should be lesser or equal to @var{high}.
11002 Default value for @var{low} is @code{20/255}, and default value for @var{high}
11006 Define the drawing mode.
11010 Draw white/gray wires on black background.
11013 Mix the colors to create a paint/cartoon effect.
11016 Apply Canny edge detector on all selected planes.
11018 Default value is @var{wires}.
11021 Select planes for filtering. By default all available planes are filtered.
11024 @subsection Examples
11028 Standard edge detection with custom values for the hysteresis thresholding:
11030 edgedetect=low=0.1:high=0.4
11034 Painting effect without thresholding:
11036 edgedetect=mode=colormix:high=0
11042 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
11044 For each input image, the filter will compute the optimal mapping from
11045 the input to the output given the codebook length, that is the number
11046 of distinct output colors.
11048 This filter accepts the following options.
11051 @item codebook_length, l
11052 Set codebook length. The value must be a positive integer, and
11053 represents the number of distinct output colors. Default value is 256.
11056 Set the maximum number of iterations to apply for computing the optimal
11057 mapping. The higher the value the better the result and the higher the
11058 computation time. Default value is 1.
11061 Set a random seed, must be an integer included between 0 and
11062 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
11063 will try to use a good random seed on a best effort basis.
11066 Set pal8 output pixel format. This option does not work with codebook
11067 length greater than 256.
11072 Measure graylevel entropy in histogram of color channels of video frames.
11074 It accepts the following parameters:
11078 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
11080 @var{diff} mode measures entropy of histogram delta values, absolute differences
11081 between neighbour histogram values.
11085 Apply the EPX magnification filter which is designed for pixel art.
11087 It accepts the following option:
11091 Set the scaling dimension: @code{2} for @code{2xEPX}, @code{3} for
11093 Default is @code{3}.
11097 Set brightness, contrast, saturation and approximate gamma adjustment.
11099 The filter accepts the following options:
11103 Set the contrast expression. The value must be a float value in range
11104 @code{-1000.0} to @code{1000.0}. The default value is "1".
11107 Set the brightness expression. The value must be a float value in
11108 range @code{-1.0} to @code{1.0}. The default value is "0".
11111 Set the saturation expression. The value must be a float in
11112 range @code{0.0} to @code{3.0}. The default value is "1".
11115 Set the gamma expression. The value must be a float in range
11116 @code{0.1} to @code{10.0}. The default value is "1".
11119 Set the gamma expression for red. The value must be a float in
11120 range @code{0.1} to @code{10.0}. The default value is "1".
11123 Set the gamma expression for green. The value must be a float in range
11124 @code{0.1} to @code{10.0}. The default value is "1".
11127 Set the gamma expression for blue. 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 weight expression. It can be used to reduce the effect
11132 of a high gamma value on bright image areas, e.g. keep them from
11133 getting overamplified and just plain white. The value must be a float
11134 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
11135 gamma correction all the way down while @code{1.0} leaves it at its
11136 full strength. Default is "1".
11139 Set when the expressions for brightness, contrast, saturation and
11140 gamma expressions are evaluated.
11142 It accepts the following values:
11145 only evaluate expressions once during the filter initialization or
11146 when a command is processed
11149 evaluate expressions for each incoming frame
11152 Default value is @samp{init}.
11155 The expressions accept the following parameters:
11158 frame count of the input frame starting from 0
11161 byte position of the corresponding packet in the input file, NAN if
11165 frame rate of the input video, NAN if the input frame rate is unknown
11168 timestamp expressed in seconds, NAN if the input timestamp is unknown
11171 @subsection Commands
11172 The filter supports the following commands:
11176 Set the contrast expression.
11179 Set the brightness expression.
11182 Set the saturation expression.
11185 Set the gamma expression.
11188 Set the gamma_r expression.
11191 Set gamma_g expression.
11194 Set gamma_b expression.
11197 Set gamma_weight expression.
11199 The command accepts the same syntax of the corresponding option.
11201 If the specified expression is not valid, it is kept at its current
11208 Apply erosion effect to the video.
11210 This filter replaces the pixel by the local(3x3) minimum.
11212 It accepts the following options:
11219 Limit the maximum change for each plane, default is 65535.
11220 If 0, plane will remain unchanged.
11223 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
11226 Flags to local 3x3 coordinates maps like this:
11233 @subsection Commands
11235 This filter supports the all above options as @ref{commands}.
11239 Deinterlace the input video ("estdif" stands for "Edge Slope
11240 Tracing Deinterlacing Filter").
11242 Spatial only filter that uses edge slope tracing algorithm
11243 to interpolate missing lines.
11244 It accepts the following parameters:
11248 The interlacing mode to adopt. It accepts one of the following values:
11252 Output one frame for each frame.
11254 Output one frame for each field.
11257 The default value is @code{field}.
11260 The picture field parity assumed for the input interlaced video. It accepts one
11261 of the following values:
11265 Assume the top field is first.
11267 Assume the bottom field is first.
11269 Enable automatic detection of field parity.
11272 The default value is @code{auto}.
11273 If the interlacing is unknown or the decoder does not export this information,
11274 top field first will be assumed.
11277 Specify which frames to deinterlace. Accepts one of the following
11282 Deinterlace all frames.
11284 Only deinterlace frames marked as interlaced.
11287 The default value is @code{all}.
11290 Specify the search radius for edge slope tracing. Default value is 1.
11291 Allowed range is from 1 to 15.
11294 Specify the search radius for best edge matching. Default value is 2.
11295 Allowed range is from 0 to 15.
11298 Specify the interpolation used. Default is 4-point interpolation. It accepts one
11299 of the following values:
11303 Two-point interpolation.
11305 Four-point interpolation.
11307 Six-point interpolation.
11311 @subsection Commands
11312 This filter supports same @ref{commands} as options.
11314 @section extractplanes
11316 Extract color channel components from input video stream into
11317 separate grayscale video streams.
11319 The filter accepts the following option:
11323 Set plane(s) to extract.
11325 Available values for planes are:
11336 Choosing planes not available in the input will result in an error.
11337 That means you cannot select @code{r}, @code{g}, @code{b} planes
11338 with @code{y}, @code{u}, @code{v} planes at same time.
11341 @subsection Examples
11345 Extract luma, u and v color channel component from input video frame
11346 into 3 grayscale outputs:
11348 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
11354 Apply a fade-in/out effect to the input video.
11356 It accepts the following parameters:
11360 The effect type can be either "in" for a fade-in, or "out" for a fade-out
11362 Default is @code{in}.
11364 @item start_frame, s
11365 Specify the number of the frame to start applying the fade
11366 effect at. Default is 0.
11369 The number of frames that the fade effect lasts. At the end of the
11370 fade-in effect, the output video will have the same intensity as the input video.
11371 At the end of the fade-out transition, the output video will be filled with the
11372 selected @option{color}.
11376 If set to 1, fade only alpha channel, if one exists on the input.
11377 Default value is 0.
11379 @item start_time, st
11380 Specify the timestamp (in seconds) of the frame to start to apply the fade
11381 effect. If both start_frame and start_time are specified, the fade will start at
11382 whichever comes last. Default is 0.
11385 The number of seconds for which the fade effect has to last. At the end of the
11386 fade-in effect the output video will have the same intensity as the input video,
11387 at the end of the fade-out transition the output video will be filled with the
11388 selected @option{color}.
11389 If both duration and nb_frames are specified, duration is used. Default is 0
11390 (nb_frames is used by default).
11393 Specify the color of the fade. Default is "black".
11396 @subsection Examples
11400 Fade in the first 30 frames of video:
11405 The command above is equivalent to:
11411 Fade out the last 45 frames of a 200-frame video:
11414 fade=type=out:start_frame=155:nb_frames=45
11418 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
11420 fade=in:0:25, fade=out:975:25
11424 Make the first 5 frames yellow, then fade in from frame 5-24:
11426 fade=in:5:20:color=yellow
11430 Fade in alpha over first 25 frames of video:
11432 fade=in:0:25:alpha=1
11436 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
11438 fade=t=in:st=5.5:d=0.5
11444 Denoise frames using 3D FFT (frequency domain filtering).
11446 The filter accepts the following options:
11450 Set the noise sigma constant. This sets denoising strength.
11451 Default value is 1. Allowed range is from 0 to 30.
11452 Using very high sigma with low overlap may give blocking artifacts.
11455 Set amount of denoising. By default all detected noise is reduced.
11456 Default value is 1. Allowed range is from 0 to 1.
11459 Set size of block, Default is 4, can be 3, 4, 5 or 6.
11460 Actual size of block in pixels is 2 to power of @var{block}, so by default
11461 block size in pixels is 2^4 which is 16.
11464 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
11467 Set number of previous frames to use for denoising. By default is set to 0.
11470 Set number of next frames to to use for denoising. By default is set to 0.
11473 Set planes which will be filtered, by default are all available filtered
11478 Apply arbitrary expressions to samples in frequency domain
11482 Adjust the dc value (gain) of the luma plane of the image. The filter
11483 accepts an integer value in range @code{0} to @code{1000}. The default
11484 value is set to @code{0}.
11487 Adjust the dc value (gain) of the 1st chroma plane of the image. The
11488 filter accepts an integer value in range @code{0} to @code{1000}. The
11489 default value is set to @code{0}.
11492 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
11493 filter accepts an integer value in range @code{0} to @code{1000}. The
11494 default value is set to @code{0}.
11497 Set the frequency domain weight expression for the luma plane.
11500 Set the frequency domain weight expression for the 1st chroma plane.
11503 Set the frequency domain weight expression for the 2nd chroma plane.
11506 Set when the expressions are evaluated.
11508 It accepts the following values:
11511 Only evaluate expressions once during the filter initialization.
11514 Evaluate expressions for each incoming frame.
11517 Default value is @samp{init}.
11519 The filter accepts the following variables:
11522 The coordinates of the current sample.
11526 The width and height of the image.
11529 The number of input frame, starting from 0.
11532 @subsection Examples
11538 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
11544 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
11550 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
11556 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
11563 Extract a single field from an interlaced image using stride
11564 arithmetic to avoid wasting CPU time. The output frames are marked as
11567 The filter accepts the following options:
11571 Specify whether to extract the top (if the value is @code{0} or
11572 @code{top}) or the bottom field (if the value is @code{1} or
11578 Create new frames by copying the top and bottom fields from surrounding frames
11579 supplied as numbers by the hint file.
11583 Set file containing hints: absolute/relative frame numbers.
11585 There must be one line for each frame in a clip. Each line must contain two
11586 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
11587 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
11588 is current frame number for @code{absolute} mode or out of [-1, 1] range
11589 for @code{relative} mode. First number tells from which frame to pick up top
11590 field and second number tells from which frame to pick up bottom field.
11592 If optionally followed by @code{+} output frame will be marked as interlaced,
11593 else if followed by @code{-} output frame will be marked as progressive, else
11594 it will be marked same as input frame.
11595 If optionally followed by @code{t} output frame will use only top field, or in
11596 case of @code{b} it will use only bottom field.
11597 If line starts with @code{#} or @code{;} that line is skipped.
11600 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
11603 Example of first several lines of @code{hint} file for @code{relative} mode:
11605 0,0 - # first frame
11606 1,0 - # second frame, use third's frame top field and second's frame bottom field
11607 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
11622 @section fieldmatch
11624 Field matching filter for inverse telecine. It is meant to reconstruct the
11625 progressive frames from a telecined stream. The filter does not drop duplicated
11626 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
11627 followed by a decimation filter such as @ref{decimate} in the filtergraph.
11629 The separation of the field matching and the decimation is notably motivated by
11630 the possibility of inserting a de-interlacing filter fallback between the two.
11631 If the source has mixed telecined and real interlaced content,
11632 @code{fieldmatch} will not be able to match fields for the interlaced parts.
11633 But these remaining combed frames will be marked as interlaced, and thus can be
11634 de-interlaced by a later filter such as @ref{yadif} before decimation.
11636 In addition to the various configuration options, @code{fieldmatch} can take an
11637 optional second stream, activated through the @option{ppsrc} option. If
11638 enabled, the frames reconstruction will be based on the fields and frames from
11639 this second stream. This allows the first input to be pre-processed in order to
11640 help the various algorithms of the filter, while keeping the output lossless
11641 (assuming the fields are matched properly). Typically, a field-aware denoiser,
11642 or brightness/contrast adjustments can help.
11644 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
11645 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
11646 which @code{fieldmatch} is based on. While the semantic and usage are very
11647 close, some behaviour and options names can differ.
11649 The @ref{decimate} filter currently only works for constant frame rate input.
11650 If your input has mixed telecined (30fps) and progressive content with a lower
11651 framerate like 24fps use the following filterchain to produce the necessary cfr
11652 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
11654 The filter accepts the following options:
11658 Specify the assumed field order of the input stream. Available values are:
11662 Auto detect parity (use FFmpeg's internal parity value).
11664 Assume bottom field first.
11666 Assume top field first.
11669 Note that it is sometimes recommended not to trust the parity announced by the
11672 Default value is @var{auto}.
11675 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
11676 sense that it won't risk creating jerkiness due to duplicate frames when
11677 possible, but if there are bad edits or blended fields it will end up
11678 outputting combed frames when a good match might actually exist. On the other
11679 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
11680 but will almost always find a good frame if there is one. The other values are
11681 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
11682 jerkiness and creating duplicate frames versus finding good matches in sections
11683 with bad edits, orphaned fields, blended fields, etc.
11685 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
11687 Available values are:
11691 2-way matching (p/c)
11693 2-way matching, and trying 3rd match if still combed (p/c + n)
11695 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
11697 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
11698 still combed (p/c + n + u/b)
11700 3-way matching (p/c/n)
11702 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
11703 detected as combed (p/c/n + u/b)
11706 The parenthesis at the end indicate the matches that would be used for that
11707 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
11710 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
11713 Default value is @var{pc_n}.
11716 Mark the main input stream as a pre-processed input, and enable the secondary
11717 input stream as the clean source to pick the fields from. See the filter
11718 introduction for more details. It is similar to the @option{clip2} feature from
11721 Default value is @code{0} (disabled).
11724 Set the field to match from. It is recommended to set this to the same value as
11725 @option{order} unless you experience matching failures with that setting. In
11726 certain circumstances changing the field that is used to match from can have a
11727 large impact on matching performance. Available values are:
11731 Automatic (same value as @option{order}).
11733 Match from the bottom field.
11735 Match from the top field.
11738 Default value is @var{auto}.
11741 Set whether or not chroma is included during the match comparisons. In most
11742 cases it is recommended to leave this enabled. You should set this to @code{0}
11743 only if your clip has bad chroma problems such as heavy rainbowing or other
11744 artifacts. Setting this to @code{0} could also be used to speed things up at
11745 the cost of some accuracy.
11747 Default value is @code{1}.
11751 These define an exclusion band which excludes the lines between @option{y0} and
11752 @option{y1} from being included in the field matching decision. An exclusion
11753 band can be used to ignore subtitles, a logo, or other things that may
11754 interfere with the matching. @option{y0} sets the starting scan line and
11755 @option{y1} sets the ending line; all lines in between @option{y0} and
11756 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
11757 @option{y0} and @option{y1} to the same value will disable the feature.
11758 @option{y0} and @option{y1} defaults to @code{0}.
11761 Set the scene change detection threshold as a percentage of maximum change on
11762 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
11763 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
11764 @option{scthresh} is @code{[0.0, 100.0]}.
11766 Default value is @code{12.0}.
11769 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
11770 account the combed scores of matches when deciding what match to use as the
11771 final match. Available values are:
11775 No final matching based on combed scores.
11777 Combed scores are only used when a scene change is detected.
11779 Use combed scores all the time.
11782 Default is @var{sc}.
11785 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
11786 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
11787 Available values are:
11791 No forced calculation.
11793 Force p/c/n calculations.
11795 Force p/c/n/u/b calculations.
11798 Default value is @var{none}.
11801 This is the area combing threshold used for combed frame detection. This
11802 essentially controls how "strong" or "visible" combing must be to be detected.
11803 Larger values mean combing must be more visible and smaller values mean combing
11804 can be less visible or strong and still be detected. Valid settings are from
11805 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
11806 be detected as combed). This is basically a pixel difference value. A good
11807 range is @code{[8, 12]}.
11809 Default value is @code{9}.
11812 Sets whether or not chroma is considered in the combed frame decision. Only
11813 disable this if your source has chroma problems (rainbowing, etc.) that are
11814 causing problems for the combed frame detection with chroma enabled. Actually,
11815 using @option{chroma}=@var{0} is usually more reliable, except for the case
11816 where there is chroma only combing in the source.
11818 Default value is @code{0}.
11822 Respectively set the x-axis and y-axis size of the window used during combed
11823 frame detection. This has to do with the size of the area in which
11824 @option{combpel} pixels are required to be detected as combed for a frame to be
11825 declared combed. See the @option{combpel} parameter description for more info.
11826 Possible values are any number that is a power of 2 starting at 4 and going up
11829 Default value is @code{16}.
11832 The number of combed pixels inside any of the @option{blocky} by
11833 @option{blockx} size blocks on the frame for the frame to be detected as
11834 combed. While @option{cthresh} controls how "visible" the combing must be, this
11835 setting controls "how much" combing there must be in any localized area (a
11836 window defined by the @option{blockx} and @option{blocky} settings) on the
11837 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
11838 which point no frames will ever be detected as combed). This setting is known
11839 as @option{MI} in TFM/VFM vocabulary.
11841 Default value is @code{80}.
11844 @anchor{p/c/n/u/b meaning}
11845 @subsection p/c/n/u/b meaning
11847 @subsubsection p/c/n
11849 We assume the following telecined stream:
11852 Top fields: 1 2 2 3 4
11853 Bottom fields: 1 2 3 4 4
11856 The numbers correspond to the progressive frame the fields relate to. Here, the
11857 first two frames are progressive, the 3rd and 4th are combed, and so on.
11859 When @code{fieldmatch} is configured to run a matching from bottom
11860 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11865 B 1 2 3 4 4 <-- matching reference
11874 As a result of the field matching, we can see that some frames get duplicated.
11875 To perform a complete inverse telecine, you need to rely on a decimation filter
11876 after this operation. See for instance the @ref{decimate} filter.
11878 The same operation now matching from top fields (@option{field}=@var{top})
11883 T 1 2 2 3 4 <-- matching reference
11893 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11894 basically, they refer to the frame and field of the opposite parity:
11897 @item @var{p} matches the field of the opposite parity in the previous frame
11898 @item @var{c} matches the field of the opposite parity in the current frame
11899 @item @var{n} matches the field of the opposite parity in the next frame
11904 The @var{u} and @var{b} matching are a bit special in the sense that they match
11905 from the opposite parity flag. In the following examples, we assume that we are
11906 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11907 'x' is placed above and below each matched fields.
11909 With bottom matching (@option{field}=@var{bottom}):
11914 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11915 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11923 With top matching (@option{field}=@var{top}):
11928 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11929 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11937 @subsection Examples
11939 Simple IVTC of a top field first telecined stream:
11941 fieldmatch=order=tff:combmatch=none, decimate
11944 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11946 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11949 @section fieldorder
11951 Transform the field order of the input video.
11953 It accepts the following parameters:
11958 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11959 for bottom field first.
11962 The default value is @samp{tff}.
11964 The transformation is done by shifting the picture content up or down
11965 by one line, and filling the remaining line with appropriate picture content.
11966 This method is consistent with most broadcast field order converters.
11968 If the input video is not flagged as being interlaced, or it is already
11969 flagged as being of the required output field order, then this filter does
11970 not alter the incoming video.
11972 It is very useful when converting to or from PAL DV material,
11973 which is bottom field first.
11977 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11980 @section fifo, afifo
11982 Buffer input images and send them when they are requested.
11984 It is mainly useful when auto-inserted by the libavfilter
11987 It does not take parameters.
11989 @section fillborders
11991 Fill borders of the input video, without changing video stream dimensions.
11992 Sometimes video can have garbage at the four edges and you may not want to
11993 crop video input to keep size multiple of some number.
11995 This filter accepts the following options:
11999 Number of pixels to fill from left border.
12002 Number of pixels to fill from right border.
12005 Number of pixels to fill from top border.
12008 Number of pixels to fill from bottom border.
12013 It accepts the following values:
12016 fill pixels using outermost pixels
12019 fill pixels using mirroring (half sample symmetric)
12022 fill pixels with constant value
12025 fill pixels using reflecting (whole sample symmetric)
12028 fill pixels using wrapping
12031 fade pixels to constant value
12034 Default is @var{smear}.
12037 Set color for pixels in fixed or fade mode. Default is @var{black}.
12040 @subsection Commands
12041 This filter supports same @ref{commands} as options.
12042 The command accepts the same syntax of the corresponding option.
12044 If the specified expression is not valid, it is kept at its current
12049 Find a rectangular object
12051 It accepts the following options:
12055 Filepath of the object image, needs to be in gray8.
12058 Detection threshold, default is 0.5.
12061 Number of mipmaps, default is 3.
12063 @item xmin, ymin, xmax, ymax
12064 Specifies the rectangle in which to search.
12067 @subsection Examples
12071 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
12073 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
12079 Flood area with values of same pixel components with another values.
12081 It accepts the following options:
12084 Set pixel x coordinate.
12087 Set pixel y coordinate.
12090 Set source #0 component value.
12093 Set source #1 component value.
12096 Set source #2 component value.
12099 Set source #3 component value.
12102 Set destination #0 component value.
12105 Set destination #1 component value.
12108 Set destination #2 component value.
12111 Set destination #3 component value.
12117 Convert the input video to one of the specified pixel formats.
12118 Libavfilter will try to pick one that is suitable as input to
12121 It accepts the following parameters:
12125 A '|'-separated list of pixel format names, such as
12126 "pix_fmts=yuv420p|monow|rgb24".
12130 @subsection Examples
12134 Convert the input video to the @var{yuv420p} format
12136 format=pix_fmts=yuv420p
12139 Convert the input video to any of the formats in the list
12141 format=pix_fmts=yuv420p|yuv444p|yuv410p
12148 Convert the video to specified constant frame rate by duplicating or dropping
12149 frames as necessary.
12151 It accepts the following parameters:
12155 The desired output frame rate. The default is @code{25}.
12158 Assume the first PTS should be the given value, in seconds. This allows for
12159 padding/trimming at the start of stream. By default, no assumption is made
12160 about the first frame's expected PTS, so no padding or trimming is done.
12161 For example, this could be set to 0 to pad the beginning with duplicates of
12162 the first frame if a video stream starts after the audio stream or to trim any
12163 frames with a negative PTS.
12166 Timestamp (PTS) rounding method.
12168 Possible values are:
12175 round towards -infinity
12177 round towards +infinity
12181 The default is @code{near}.
12184 Action performed when reading the last frame.
12186 Possible values are:
12189 Use same timestamp rounding method as used for other frames.
12191 Pass through last frame if input duration has not been reached yet.
12193 The default is @code{round}.
12197 Alternatively, the options can be specified as a flat string:
12198 @var{fps}[:@var{start_time}[:@var{round}]].
12200 See also the @ref{setpts} filter.
12202 @subsection Examples
12206 A typical usage in order to set the fps to 25:
12212 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
12214 fps=fps=film:round=near
12220 Pack two different video streams into a stereoscopic video, setting proper
12221 metadata on supported codecs. The two views should have the same size and
12222 framerate and processing will stop when the shorter video ends. Please note
12223 that you may conveniently adjust view properties with the @ref{scale} and
12226 It accepts the following parameters:
12230 The desired packing format. Supported values are:
12235 The views are next to each other (default).
12238 The views are on top of each other.
12241 The views are packed by line.
12244 The views are packed by column.
12247 The views are temporally interleaved.
12256 # Convert left and right views into a frame-sequential video
12257 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
12259 # Convert views into a side-by-side video with the same output resolution as the input
12260 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
12265 Change the frame rate by interpolating new video output frames from the source
12268 This filter is not designed to function correctly with interlaced media. If
12269 you wish to change the frame rate of interlaced media then you are required
12270 to deinterlace before this filter and re-interlace after this filter.
12272 A description of the accepted options follows.
12276 Specify the output frames per second. This option can also be specified
12277 as a value alone. The default is @code{50}.
12280 Specify the start of a range where the output frame will be created as a
12281 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12282 the default is @code{15}.
12285 Specify the end of a range where the output frame will be created as a
12286 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12287 the default is @code{240}.
12290 Specify the level at which a scene change is detected as a value between
12291 0 and 100 to indicate a new scene; a low value reflects a low
12292 probability for the current frame to introduce a new scene, while a higher
12293 value means the current frame is more likely to be one.
12294 The default is @code{8.2}.
12297 Specify flags influencing the filter process.
12299 Available value for @var{flags} is:
12302 @item scene_change_detect, scd
12303 Enable scene change detection using the value of the option @var{scene}.
12304 This flag is enabled by default.
12310 Select one frame every N-th frame.
12312 This filter accepts the following option:
12315 Select frame after every @code{step} frames.
12316 Allowed values are positive integers higher than 0. Default value is @code{1}.
12319 @section freezedetect
12321 Detect frozen video.
12323 This filter logs a message and sets frame metadata when it detects that the
12324 input video has no significant change in content during a specified duration.
12325 Video freeze detection calculates the mean average absolute difference of all
12326 the components of video frames and compares it to a noise floor.
12328 The printed times and duration are expressed in seconds. The
12329 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
12330 whose timestamp equals or exceeds the detection duration and it contains the
12331 timestamp of the first frame of the freeze. The
12332 @code{lavfi.freezedetect.freeze_duration} and
12333 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
12336 The filter accepts the following options:
12340 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
12341 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
12345 Set freeze duration until notification (default is 2 seconds).
12348 @section freezeframes
12350 Freeze video frames.
12352 This filter freezes video frames using frame from 2nd input.
12354 The filter accepts the following options:
12358 Set number of first frame from which to start freeze.
12361 Set number of last frame from which to end freeze.
12364 Set number of frame from 2nd input which will be used instead of replaced frames.
12370 Apply a frei0r effect to the input video.
12372 To enable the compilation of this filter, you need to install the frei0r
12373 header and configure FFmpeg with @code{--enable-frei0r}.
12375 It accepts the following parameters:
12380 The name of the frei0r effect to load. If the environment variable
12381 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
12382 directories specified by the colon-separated list in @env{FREI0R_PATH}.
12383 Otherwise, the standard frei0r paths are searched, in this order:
12384 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
12385 @file{/usr/lib/frei0r-1/}.
12387 @item filter_params
12388 A '|'-separated list of parameters to pass to the frei0r effect.
12392 A frei0r effect parameter can be a boolean (its value is either
12393 "y" or "n"), a double, a color (specified as
12394 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
12395 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
12396 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
12397 a position (specified as @var{X}/@var{Y}, where
12398 @var{X} and @var{Y} are floating point numbers) and/or a string.
12400 The number and types of parameters depend on the loaded effect. If an
12401 effect parameter is not specified, the default value is set.
12403 @subsection Examples
12407 Apply the distort0r effect, setting the first two double parameters:
12409 frei0r=filter_name=distort0r:filter_params=0.5|0.01
12413 Apply the colordistance effect, taking a color as the first parameter:
12415 frei0r=colordistance:0.2/0.3/0.4
12416 frei0r=colordistance:violet
12417 frei0r=colordistance:0x112233
12421 Apply the perspective effect, specifying the top left and top right image
12424 frei0r=perspective:0.2/0.2|0.8/0.2
12428 For more information, see
12429 @url{http://frei0r.dyne.org}
12431 @subsection Commands
12433 This filter supports the @option{filter_params} option as @ref{commands}.
12437 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
12439 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
12440 processing filter, one of them is performed once per block, not per pixel.
12441 This allows for much higher speed.
12443 The filter accepts the following options:
12447 Set quality. This option defines the number of levels for averaging. It accepts
12448 an integer in the range 4-5. Default value is @code{4}.
12451 Force a constant quantization parameter. It accepts an integer in range 0-63.
12452 If not set, the filter will use the QP from the video stream (if available).
12455 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
12456 more details but also more artifacts, while higher values make the image smoother
12457 but also blurrier. Default value is @code{0} − PSNR optimal.
12459 @item use_bframe_qp
12460 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12461 option may cause flicker since the B-Frames have often larger QP. Default is
12462 @code{0} (not enabled).
12468 Apply Gaussian blur filter.
12470 The filter accepts the following options:
12474 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
12477 Set number of steps for Gaussian approximation. Default is @code{1}.
12480 Set which planes to filter. By default all planes are filtered.
12483 Set vertical sigma, if negative it will be same as @code{sigma}.
12484 Default is @code{-1}.
12487 @subsection Commands
12488 This filter supports same commands as options.
12489 The command accepts the same syntax of the corresponding option.
12491 If the specified expression is not valid, it is kept at its current
12496 Apply generic equation to each pixel.
12498 The filter accepts the following options:
12501 @item lum_expr, lum
12502 Set the luminance expression.
12504 Set the chrominance blue expression.
12506 Set the chrominance red expression.
12507 @item alpha_expr, a
12508 Set the alpha expression.
12510 Set the red expression.
12511 @item green_expr, g
12512 Set the green expression.
12514 Set the blue expression.
12517 The colorspace is selected according to the specified options. If one
12518 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
12519 options is specified, the filter will automatically select a YCbCr
12520 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
12521 @option{blue_expr} options is specified, it will select an RGB
12524 If one of the chrominance expression is not defined, it falls back on the other
12525 one. If no alpha expression is specified it will evaluate to opaque value.
12526 If none of chrominance expressions are specified, they will evaluate
12527 to the luminance expression.
12529 The expressions can use the following variables and functions:
12533 The sequential number of the filtered frame, starting from @code{0}.
12537 The coordinates of the current sample.
12541 The width and height of the image.
12545 Width and height scale depending on the currently filtered plane. It is the
12546 ratio between the corresponding luma plane number of pixels and the current
12547 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
12548 @code{0.5,0.5} for chroma planes.
12551 Time of the current frame, expressed in seconds.
12554 Return the value of the pixel at location (@var{x},@var{y}) of the current
12558 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
12562 Return the value of the pixel at location (@var{x},@var{y}) of the
12563 blue-difference chroma plane. Return 0 if there is no such plane.
12566 Return the value of the pixel at location (@var{x},@var{y}) of the
12567 red-difference chroma plane. Return 0 if there is no such plane.
12572 Return the value of the pixel at location (@var{x},@var{y}) of the
12573 red/green/blue component. Return 0 if there is no such component.
12576 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
12577 plane. Return 0 if there is no such plane.
12579 @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)
12580 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
12581 sums of samples within a rectangle. See the functions without the sum postfix.
12583 @item interpolation
12584 Set one of interpolation methods:
12589 Default is bilinear.
12592 For functions, if @var{x} and @var{y} are outside the area, the value will be
12593 automatically clipped to the closer edge.
12595 Please note that this filter can use multiple threads in which case each slice
12596 will have its own expression state. If you want to use only a single expression
12597 state because your expressions depend on previous state then you should limit
12598 the number of filter threads to 1.
12600 @subsection Examples
12604 Flip the image horizontally:
12610 Generate a bidimensional sine wave, with angle @code{PI/3} and a
12611 wavelength of 100 pixels:
12613 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
12617 Generate a fancy enigmatic moving light:
12619 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
12623 Generate a quick emboss effect:
12625 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
12629 Modify RGB components depending on pixel position:
12631 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
12635 Create a radial gradient that is the same size as the input (also see
12636 the @ref{vignette} filter):
12638 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
12644 Fix the banding artifacts that are sometimes introduced into nearly flat
12645 regions by truncation to 8-bit color depth.
12646 Interpolate the gradients that should go where the bands are, and
12649 It is designed for playback only. Do not use it prior to
12650 lossy compression, because compression tends to lose the dither and
12651 bring back the bands.
12653 It accepts the following parameters:
12658 The maximum amount by which the filter will change any one pixel. This is also
12659 the threshold for detecting nearly flat regions. Acceptable values range from
12660 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
12664 The neighborhood to fit the gradient to. A larger radius makes for smoother
12665 gradients, but also prevents the filter from modifying the pixels near detailed
12666 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
12667 values will be clipped to the valid range.
12671 Alternatively, the options can be specified as a flat string:
12672 @var{strength}[:@var{radius}]
12674 @subsection Examples
12678 Apply the filter with a @code{3.5} strength and radius of @code{8}:
12684 Specify radius, omitting the strength (which will fall-back to the default
12692 @anchor{graphmonitor}
12693 @section graphmonitor
12694 Show various filtergraph stats.
12696 With this filter one can debug complete filtergraph.
12697 Especially issues with links filling with queued frames.
12699 The filter accepts the following options:
12703 Set video output size. Default is @var{hd720}.
12706 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
12709 Set output mode, can be @var{fulll} or @var{compact}.
12710 In @var{compact} mode only filters with some queued frames have displayed stats.
12713 Set flags which enable which stats are shown in video.
12715 Available values for flags are:
12718 Display number of queued frames in each link.
12720 @item frame_count_in
12721 Display number of frames taken from filter.
12723 @item frame_count_out
12724 Display number of frames given out from filter.
12727 Display current filtered frame pts.
12730 Display current filtered frame time.
12733 Display time base for filter link.
12736 Display used format for filter link.
12739 Display video size or number of audio channels in case of audio used by filter link.
12742 Display video frame rate or sample rate in case of audio used by filter link.
12745 Display link output status.
12749 Set upper limit for video rate of output stream, Default value is @var{25}.
12750 This guarantee that output video frame rate will not be higher than this value.
12754 A color constancy variation filter which estimates scene illumination via grey edge algorithm
12755 and corrects the scene colors accordingly.
12757 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
12759 The filter accepts the following options:
12763 The order of differentiation to be applied on the scene. Must be chosen in the range
12764 [0,2] and default value is 1.
12767 The Minkowski parameter to be used for calculating the Minkowski distance. Must
12768 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
12769 max value instead of calculating Minkowski distance.
12772 The standard deviation of Gaussian blur to be applied on the scene. Must be
12773 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
12774 can't be equal to 0 if @var{difford} is greater than 0.
12777 @subsection Examples
12783 greyedge=difford=1:minknorm=5:sigma=2
12789 greyedge=difford=1:minknorm=0:sigma=2
12797 Apply a Hald CLUT to a video stream.
12799 First input is the video stream to process, and second one is the Hald CLUT.
12800 The Hald CLUT input can be a simple picture or a complete video stream.
12802 The filter accepts the following options:
12806 Force termination when the shortest input terminates. Default is @code{0}.
12808 Continue applying the last CLUT after the end of the stream. A value of
12809 @code{0} disable the filter after the last frame of the CLUT is reached.
12810 Default is @code{1}.
12813 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
12814 filters share the same internals).
12816 This filter also supports the @ref{framesync} options.
12818 More information about the Hald CLUT can be found on Eskil Steenberg's website
12819 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
12821 @subsection Commands
12823 This filter supports the @code{interp} option as @ref{commands}.
12825 @subsection Workflow examples
12827 @subsubsection Hald CLUT video stream
12829 Generate an identity Hald CLUT stream altered with various effects:
12831 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
12834 Note: make sure you use a lossless codec.
12836 Then use it with @code{haldclut} to apply it on some random stream:
12838 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
12841 The Hald CLUT will be applied to the 10 first seconds (duration of
12842 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
12843 to the remaining frames of the @code{mandelbrot} stream.
12845 @subsubsection Hald CLUT with preview
12847 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
12848 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
12849 biggest possible square starting at the top left of the picture. The remaining
12850 padding pixels (bottom or right) will be ignored. This area can be used to add
12851 a preview of the Hald CLUT.
12853 Typically, the following generated Hald CLUT will be supported by the
12854 @code{haldclut} filter:
12857 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
12858 pad=iw+320 [padded_clut];
12859 smptebars=s=320x256, split [a][b];
12860 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
12861 [main][b] overlay=W-320" -frames:v 1 clut.png
12864 It contains the original and a preview of the effect of the CLUT: SMPTE color
12865 bars are displayed on the right-top, and below the same color bars processed by
12868 Then, the effect of this Hald CLUT can be visualized with:
12870 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
12875 Flip the input video horizontally.
12877 For example, to horizontally flip the input video with @command{ffmpeg}:
12879 ffmpeg -i in.avi -vf "hflip" out.avi
12883 This filter applies a global color histogram equalization on a
12886 It can be used to correct video that has a compressed range of pixel
12887 intensities. The filter redistributes the pixel intensities to
12888 equalize their distribution across the intensity range. It may be
12889 viewed as an "automatically adjusting contrast filter". This filter is
12890 useful only for correcting degraded or poorly captured source
12893 The filter accepts the following options:
12897 Determine the amount of equalization to be applied. As the strength
12898 is reduced, the distribution of pixel intensities more-and-more
12899 approaches that of the input frame. The value must be a float number
12900 in the range [0,1] and defaults to 0.200.
12903 Set the maximum intensity that can generated and scale the output
12904 values appropriately. The strength should be set as desired and then
12905 the intensity can be limited if needed to avoid washing-out. The value
12906 must be a float number in the range [0,1] and defaults to 0.210.
12909 Set the antibanding level. If enabled the filter will randomly vary
12910 the luminance of output pixels by a small amount to avoid banding of
12911 the histogram. Possible values are @code{none}, @code{weak} or
12912 @code{strong}. It defaults to @code{none}.
12918 Compute and draw a color distribution histogram for the input video.
12920 The computed histogram is a representation of the color component
12921 distribution in an image.
12923 Standard histogram displays the color components distribution in an image.
12924 Displays color graph for each color component. Shows distribution of
12925 the Y, U, V, A or R, G, B components, depending on input format, in the
12926 current frame. Below each graph a color component scale meter is shown.
12928 The filter accepts the following options:
12932 Set height of level. Default value is @code{200}.
12933 Allowed range is [50, 2048].
12936 Set height of color scale. Default value is @code{12}.
12937 Allowed range is [0, 40].
12941 It accepts the following values:
12944 Per color component graphs are placed below each other.
12947 Per color component graphs are placed side by side.
12950 Presents information identical to that in the @code{parade}, except
12951 that the graphs representing color components are superimposed directly
12954 Default is @code{stack}.
12957 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12958 Default is @code{linear}.
12961 Set what color components to display.
12962 Default is @code{7}.
12965 Set foreground opacity. Default is @code{0.7}.
12968 Set background opacity. Default is @code{0.5}.
12971 @subsection Examples
12976 Calculate and draw histogram:
12978 ffplay -i input -vf histogram
12986 This is a high precision/quality 3d denoise filter. It aims to reduce
12987 image noise, producing smooth images and making still images really
12988 still. It should enhance compressibility.
12990 It accepts the following optional parameters:
12994 A non-negative floating point number which specifies spatial luma strength.
12995 It defaults to 4.0.
12997 @item chroma_spatial
12998 A non-negative floating point number which specifies spatial chroma strength.
12999 It defaults to 3.0*@var{luma_spatial}/4.0.
13002 A floating point number which specifies luma temporal strength. It defaults to
13003 6.0*@var{luma_spatial}/4.0.
13006 A floating point number which specifies chroma temporal strength. It defaults to
13007 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
13010 @subsection Commands
13011 This filter supports same @ref{commands} as options.
13012 The command accepts the same syntax of the corresponding option.
13014 If the specified expression is not valid, it is kept at its current
13017 @anchor{hwdownload}
13018 @section hwdownload
13020 Download hardware frames to system memory.
13022 The input must be in hardware frames, and the output a non-hardware format.
13023 Not all formats will be supported on the output - it may be necessary to insert
13024 an additional @option{format} filter immediately following in the graph to get
13025 the output in a supported format.
13029 Map hardware frames to system memory or to another device.
13031 This filter has several different modes of operation; which one is used depends
13032 on the input and output formats:
13035 Hardware frame input, normal frame output
13037 Map the input frames to system memory and pass them to the output. If the
13038 original hardware frame is later required (for example, after overlaying
13039 something else on part of it), the @option{hwmap} filter can be used again
13040 in the next mode to retrieve it.
13042 Normal frame input, hardware frame output
13044 If the input is actually a software-mapped hardware frame, then unmap it -
13045 that is, return the original hardware frame.
13047 Otherwise, a device must be provided. Create new hardware surfaces on that
13048 device for the output, then map them back to the software format at the input
13049 and give those frames to the preceding filter. This will then act like the
13050 @option{hwupload} filter, but may be able to avoid an additional copy when
13051 the input is already in a compatible format.
13053 Hardware frame input and output
13055 A device must be supplied for the output, either directly or with the
13056 @option{derive_device} option. The input and output devices must be of
13057 different types and compatible - the exact meaning of this is
13058 system-dependent, but typically it means that they must refer to the same
13059 underlying hardware context (for example, refer to the same graphics card).
13061 If the input frames were originally created on the output device, then unmap
13062 to retrieve the original frames.
13064 Otherwise, map the frames to the output device - create new hardware frames
13065 on the output corresponding to the frames on the input.
13068 The following additional parameters are accepted:
13072 Set the frame mapping mode. Some combination of:
13075 The mapped frame should be readable.
13077 The mapped frame should be writeable.
13079 The mapping will always overwrite the entire frame.
13081 This may improve performance in some cases, as the original contents of the
13082 frame need not be loaded.
13084 The mapping must not involve any copying.
13086 Indirect mappings to copies of frames are created in some cases where either
13087 direct mapping is not possible or it would have unexpected properties.
13088 Setting this flag ensures that the mapping is direct and will fail if that is
13091 Defaults to @var{read+write} if not specified.
13093 @item derive_device @var{type}
13094 Rather than using the device supplied at initialisation, instead derive a new
13095 device of type @var{type} from the device the input frames exist on.
13098 In a hardware to hardware mapping, map in reverse - create frames in the sink
13099 and map them back to the source. This may be necessary in some cases where
13100 a mapping in one direction is required but only the opposite direction is
13101 supported by the devices being used.
13103 This option is dangerous - it may break the preceding filter in undefined
13104 ways if there are any additional constraints on that filter's output.
13105 Do not use it without fully understanding the implications of its use.
13111 Upload system memory frames to hardware surfaces.
13113 The device to upload to must be supplied when the filter is initialised. If
13114 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
13115 option or with the @option{derive_device} option. The input and output devices
13116 must be of different types and compatible - the exact meaning of this is
13117 system-dependent, but typically it means that they must refer to the same
13118 underlying hardware context (for example, refer to the same graphics card).
13120 The following additional parameters are accepted:
13123 @item derive_device @var{type}
13124 Rather than using the device supplied at initialisation, instead derive a new
13125 device of type @var{type} from the device the input frames exist on.
13128 @anchor{hwupload_cuda}
13129 @section hwupload_cuda
13131 Upload system memory frames to a CUDA device.
13133 It accepts the following optional parameters:
13137 The number of the CUDA device to use
13142 Apply a high-quality magnification filter designed for pixel art. This filter
13143 was originally created by Maxim Stepin.
13145 It accepts the following option:
13149 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
13150 @code{hq3x} and @code{4} for @code{hq4x}.
13151 Default is @code{3}.
13155 Stack input videos horizontally.
13157 All streams must be of same pixel format and of same height.
13159 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13160 to create same output.
13162 The filter accepts the following option:
13166 Set number of input streams. Default is 2.
13169 If set to 1, force the output to terminate when the shortest input
13170 terminates. Default value is 0.
13175 Modify the hue and/or the saturation of the input.
13177 It accepts the following parameters:
13181 Specify the hue angle as a number of degrees. It accepts an expression,
13182 and defaults to "0".
13185 Specify the saturation in the [-10,10] range. It accepts an expression and
13189 Specify the hue angle as a number of radians. It accepts an
13190 expression, and defaults to "0".
13193 Specify the brightness in the [-10,10] range. It accepts an expression and
13197 @option{h} and @option{H} are mutually exclusive, and can't be
13198 specified at the same time.
13200 The @option{b}, @option{h}, @option{H} and @option{s} option values are
13201 expressions containing the following constants:
13205 frame count of the input frame starting from 0
13208 presentation timestamp of the input frame expressed in time base units
13211 frame rate of the input video, NAN if the input frame rate is unknown
13214 timestamp expressed in seconds, NAN if the input timestamp is unknown
13217 time base of the input video
13220 @subsection Examples
13224 Set the hue to 90 degrees and the saturation to 1.0:
13230 Same command but expressing the hue in radians:
13236 Rotate hue and make the saturation swing between 0
13237 and 2 over a period of 1 second:
13239 hue="H=2*PI*t: s=sin(2*PI*t)+1"
13243 Apply a 3 seconds saturation fade-in effect starting at 0:
13245 hue="s=min(t/3\,1)"
13248 The general fade-in expression can be written as:
13250 hue="s=min(0\, max((t-START)/DURATION\, 1))"
13254 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
13256 hue="s=max(0\, min(1\, (8-t)/3))"
13259 The general fade-out expression can be written as:
13261 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
13266 @subsection Commands
13268 This filter supports the following commands:
13274 Modify the hue and/or the saturation and/or brightness of the input video.
13275 The command accepts the same syntax of the corresponding option.
13277 If the specified expression is not valid, it is kept at its current
13281 @section hysteresis
13283 Grow first stream into second stream by connecting components.
13284 This makes it possible to build more robust edge masks.
13286 This filter accepts the following options:
13290 Set which planes will be processed as bitmap, unprocessed planes will be
13291 copied from first stream.
13292 By default value 0xf, all planes will be processed.
13295 Set threshold which is used in filtering. If pixel component value is higher than
13296 this value filter algorithm for connecting components is activated.
13297 By default value is 0.
13300 The @code{hysteresis} filter also supports the @ref{framesync} options.
13304 Detect video interlacing type.
13306 This filter tries to detect if the input frames are interlaced, progressive,
13307 top or bottom field first. It will also try to detect fields that are
13308 repeated between adjacent frames (a sign of telecine).
13310 Single frame detection considers only immediately adjacent frames when classifying each frame.
13311 Multiple frame detection incorporates the classification history of previous frames.
13313 The filter will log these metadata values:
13316 @item single.current_frame
13317 Detected type of current frame using single-frame detection. One of:
13318 ``tff'' (top field first), ``bff'' (bottom field first),
13319 ``progressive'', or ``undetermined''
13322 Cumulative number of frames detected as top field first using single-frame detection.
13325 Cumulative number of frames detected as top field first using multiple-frame detection.
13328 Cumulative number of frames detected as bottom field first using single-frame detection.
13330 @item multiple.current_frame
13331 Detected type of current frame using multiple-frame detection. One of:
13332 ``tff'' (top field first), ``bff'' (bottom field first),
13333 ``progressive'', or ``undetermined''
13336 Cumulative number of frames detected as bottom field first using multiple-frame detection.
13338 @item single.progressive
13339 Cumulative number of frames detected as progressive using single-frame detection.
13341 @item multiple.progressive
13342 Cumulative number of frames detected as progressive using multiple-frame detection.
13344 @item single.undetermined
13345 Cumulative number of frames that could not be classified using single-frame detection.
13347 @item multiple.undetermined
13348 Cumulative number of frames that could not be classified using multiple-frame detection.
13350 @item repeated.current_frame
13351 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
13353 @item repeated.neither
13354 Cumulative number of frames with no repeated field.
13357 Cumulative number of frames with the top field repeated from the previous frame's top field.
13359 @item repeated.bottom
13360 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
13363 The filter accepts the following options:
13367 Set interlacing threshold.
13369 Set progressive threshold.
13371 Threshold for repeated field detection.
13373 Number of frames after which a given frame's contribution to the
13374 statistics is halved (i.e., it contributes only 0.5 to its
13375 classification). The default of 0 means that all frames seen are given
13376 full weight of 1.0 forever.
13377 @item analyze_interlaced_flag
13378 When this is not 0 then idet will use the specified number of frames to determine
13379 if the interlaced flag is accurate, it will not count undetermined frames.
13380 If the flag is found to be accurate it will be used without any further
13381 computations, if it is found to be inaccurate it will be cleared without any
13382 further computations. This allows inserting the idet filter as a low computational
13383 method to clean up the interlaced flag
13388 Deinterleave or interleave fields.
13390 This filter allows one to process interlaced images fields without
13391 deinterlacing them. Deinterleaving splits the input frame into 2
13392 fields (so called half pictures). Odd lines are moved to the top
13393 half of the output image, even lines to the bottom half.
13394 You can process (filter) them independently and then re-interleave them.
13396 The filter accepts the following options:
13400 @item chroma_mode, c
13401 @item alpha_mode, a
13402 Available values for @var{luma_mode}, @var{chroma_mode} and
13403 @var{alpha_mode} are:
13409 @item deinterleave, d
13410 Deinterleave fields, placing one above the other.
13412 @item interleave, i
13413 Interleave fields. Reverse the effect of deinterleaving.
13415 Default value is @code{none}.
13417 @item luma_swap, ls
13418 @item chroma_swap, cs
13419 @item alpha_swap, as
13420 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
13423 @subsection Commands
13425 This filter supports the all above options as @ref{commands}.
13429 Apply inflate effect to the video.
13431 This filter replaces the pixel by the local(3x3) average by taking into account
13432 only values higher than the pixel.
13434 It accepts the following options:
13441 Limit the maximum change for each plane, default is 65535.
13442 If 0, plane will remain unchanged.
13445 @subsection Commands
13447 This filter supports the all above options as @ref{commands}.
13451 Simple interlacing filter from progressive contents. This interleaves upper (or
13452 lower) lines from odd frames with lower (or upper) lines from even frames,
13453 halving the frame rate and preserving image height.
13456 Original Original New Frame
13457 Frame 'j' Frame 'j+1' (tff)
13458 ========== =========== ==================
13459 Line 0 --------------------> Frame 'j' Line 0
13460 Line 1 Line 1 ----> Frame 'j+1' Line 1
13461 Line 2 ---------------------> Frame 'j' Line 2
13462 Line 3 Line 3 ----> Frame 'j+1' Line 3
13464 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
13467 It accepts the following optional parameters:
13471 This determines whether the interlaced frame is taken from the even
13472 (tff - default) or odd (bff) lines of the progressive frame.
13475 Vertical lowpass filter to avoid twitter interlacing and
13476 reduce moire patterns.
13480 Disable vertical lowpass filter
13483 Enable linear filter (default)
13486 Enable complex filter. This will slightly less reduce twitter and moire
13487 but better retain detail and subjective sharpness impression.
13494 Deinterlace input video by applying Donald Graft's adaptive kernel
13495 deinterling. Work on interlaced parts of a video to produce
13496 progressive frames.
13498 The description of the accepted parameters follows.
13502 Set the threshold which affects the filter's tolerance when
13503 determining if a pixel line must be processed. It must be an integer
13504 in the range [0,255] and defaults to 10. A value of 0 will result in
13505 applying the process on every pixels.
13508 Paint pixels exceeding the threshold value to white if set to 1.
13512 Set the fields order. Swap fields if set to 1, leave fields alone if
13516 Enable additional sharpening if set to 1. Default is 0.
13519 Enable twoway sharpening if set to 1. Default is 0.
13522 @subsection Examples
13526 Apply default values:
13528 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
13532 Enable additional sharpening:
13538 Paint processed pixels in white:
13545 Apply kirsch operator to input video stream.
13547 The filter accepts the following option:
13551 Set which planes will be processed, unprocessed planes will be copied.
13552 By default value 0xf, all planes will be processed.
13555 Set value which will be multiplied with filtered result.
13558 Set value which will be added to filtered result.
13561 @subsection Commands
13563 This filter supports the all above options as @ref{commands}.
13567 Slowly update darker pixels.
13569 This filter makes short flashes of light appear longer.
13570 This filter accepts the following options:
13574 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
13577 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
13580 @subsection Commands
13582 This filter supports the all above options as @ref{commands}.
13584 @section lenscorrection
13586 Correct radial lens distortion
13588 This filter can be used to correct for radial distortion as can result from the use
13589 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
13590 one can use tools available for example as part of opencv or simply trial-and-error.
13591 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
13592 and extract the k1 and k2 coefficients from the resulting matrix.
13594 Note that effectively the same filter is available in the open-source tools Krita and
13595 Digikam from the KDE project.
13597 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
13598 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
13599 brightness distribution, so you may want to use both filters together in certain
13600 cases, though you will have to take care of ordering, i.e. whether vignetting should
13601 be applied before or after lens correction.
13603 @subsection Options
13605 The filter accepts the following options:
13609 Relative x-coordinate of the focal point of the image, and thereby the center of the
13610 distortion. This value has a range [0,1] and is expressed as fractions of the image
13611 width. Default is 0.5.
13613 Relative y-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 height. Default is 0.5.
13617 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
13618 no correction. Default is 0.
13620 Coefficient of the double quadratic correction term. This value has a range [-1,1].
13621 0 means no correction. Default is 0.
13623 Set interpolation type. Can be @code{nearest} or @code{bilinear}.
13624 Default is @code{nearest}.
13626 Specify the color of the unmapped pixels. For the syntax of this option,
13627 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13628 manual,ffmpeg-utils}. Default color is @code{black@@0}.
13631 The formula that generates the correction is:
13633 @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)
13635 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
13636 distances from the focal point in the source and target images, respectively.
13638 @subsection Commands
13640 This filter supports the all above options as @ref{commands}.
13644 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
13646 The @code{lensfun} filter requires the camera make, camera model, and lens model
13647 to apply the lens correction. The filter will load the lensfun database and
13648 query it to find the corresponding camera and lens entries in the database. As
13649 long as these entries can be found with the given options, the filter can
13650 perform corrections on frames. Note that incomplete strings will result in the
13651 filter choosing the best match with the given options, and the filter will
13652 output the chosen camera and lens models (logged with level "info"). You must
13653 provide the make, camera model, and lens model as they are required.
13655 The filter accepts the following options:
13659 The make of the camera (for example, "Canon"). This option is required.
13662 The model of the camera (for example, "Canon EOS 100D"). This option is
13666 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
13667 option is required.
13670 The type of correction to apply. The following values are valid options:
13674 Enables fixing lens vignetting.
13677 Enables fixing lens geometry. This is the default.
13680 Enables fixing chromatic aberrations.
13683 Enables fixing lens vignetting and lens geometry.
13686 Enables fixing lens vignetting and chromatic aberrations.
13689 Enables fixing both lens geometry and chromatic aberrations.
13692 Enables all possible corrections.
13696 The focal length of the image/video (zoom; expected constant for video). For
13697 example, a 18--55mm lens has focal length range of [18--55], so a value in that
13698 range should be chosen when using that lens. Default 18.
13701 The aperture of the image/video (expected constant for video). Note that
13702 aperture is only used for vignetting correction. Default 3.5.
13704 @item focus_distance
13705 The focus distance of the image/video (expected constant for video). Note that
13706 focus distance is only used for vignetting and only slightly affects the
13707 vignetting correction process. If unknown, leave it at the default value (which
13711 The scale factor which is applied after transformation. After correction the
13712 video is no longer necessarily rectangular. This parameter controls how much of
13713 the resulting image is visible. The value 0 means that a value will be chosen
13714 automatically such that there is little or no unmapped area in the output
13715 image. 1.0 means that no additional scaling is done. Lower values may result
13716 in more of the corrected image being visible, while higher values may avoid
13717 unmapped areas in the output.
13719 @item target_geometry
13720 The target geometry of the output image/video. The following values are valid
13724 @item rectilinear (default)
13727 @item equirectangular
13728 @item fisheye_orthographic
13729 @item fisheye_stereographic
13730 @item fisheye_equisolid
13731 @item fisheye_thoby
13734 Apply the reverse of image correction (instead of correcting distortion, apply
13737 @item interpolation
13738 The type of interpolation used when correcting distortion. The following values
13743 @item linear (default)
13748 @subsection Examples
13752 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
13753 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
13757 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
13761 Apply the same as before, but only for the first 5 seconds of video.
13764 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
13771 Obtain the VMAF (Video Multi-Method Assessment Fusion)
13772 score between two input videos.
13774 The obtained VMAF score is printed through the logging system.
13776 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
13777 After installing the library it can be enabled using:
13778 @code{./configure --enable-libvmaf}.
13779 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
13781 The filter has following options:
13785 Set the model path which is to be used for SVM.
13786 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
13789 Set the file path to be used to store logs.
13792 Set the format of the log file (csv, json or xml).
13794 @item enable_transform
13795 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
13796 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
13797 Default value: @code{false}
13800 Invokes the phone model which will generate VMAF scores higher than in the
13801 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
13802 Default value: @code{false}
13805 Enables computing psnr along with vmaf.
13806 Default value: @code{false}
13809 Enables computing ssim along with vmaf.
13810 Default value: @code{false}
13813 Enables computing ms_ssim along with vmaf.
13814 Default value: @code{false}
13817 Set the pool method to be used for computing vmaf.
13818 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
13821 Set number of threads to be used when computing vmaf.
13822 Default value: @code{0}, which makes use of all available logical processors.
13825 Set interval for frame subsampling used when computing vmaf.
13826 Default value: @code{1}
13828 @item enable_conf_interval
13829 Enables confidence interval.
13830 Default value: @code{false}
13833 This filter also supports the @ref{framesync} options.
13835 @subsection Examples
13838 On the below examples the input file @file{main.mpg} being processed is
13839 compared with the reference file @file{ref.mpg}.
13842 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
13846 Example with options:
13848 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
13852 Example with options and different containers:
13854 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 -
13860 Limits the pixel components values to the specified range [min, max].
13862 The filter accepts the following options:
13866 Lower bound. Defaults to the lowest allowed value for the input.
13869 Upper bound. Defaults to the highest allowed value for the input.
13872 Specify which planes will be processed. Defaults to all available.
13875 @subsection Commands
13877 This filter supports the all above options as @ref{commands}.
13883 The filter accepts the following options:
13887 Set the number of loops. Setting this value to -1 will result in infinite loops.
13891 Set maximal size in number of frames. Default is 0.
13894 Set first frame of loop. Default is 0.
13897 @subsection Examples
13901 Loop single first frame infinitely:
13903 loop=loop=-1:size=1:start=0
13907 Loop single first frame 10 times:
13909 loop=loop=10:size=1:start=0
13913 Loop 10 first frames 5 times:
13915 loop=loop=5:size=10:start=0
13921 Apply a 1D LUT to an input video.
13923 The filter accepts the following options:
13927 Set the 1D LUT file name.
13929 Currently supported formats:
13938 Select interpolation mode.
13940 Available values are:
13944 Use values from the nearest defined point.
13946 Interpolate values using the linear interpolation.
13948 Interpolate values using the cosine interpolation.
13950 Interpolate values using the cubic interpolation.
13952 Interpolate values using the spline interpolation.
13956 @subsection Commands
13958 This filter supports the all above options as @ref{commands}.
13963 Apply a 3D LUT to an input video.
13965 The filter accepts the following options:
13969 Set the 3D LUT file name.
13971 Currently supported formats:
13985 Select interpolation mode.
13987 Available values are:
13991 Use values from the nearest defined point.
13993 Interpolate values using the 8 points defining a cube.
13995 Interpolate values using a tetrahedron.
13997 Interpolate values using a pyramid.
13999 Interpolate values using a prism.
14003 @subsection Commands
14005 This filter supports the @code{interp} option as @ref{commands}.
14009 Turn certain luma values into transparency.
14011 The filter accepts the following options:
14015 Set the luma which will be used as base for transparency.
14016 Default value is @code{0}.
14019 Set the range of luma values to be keyed out.
14020 Default value is @code{0.01}.
14023 Set the range of softness. Default value is @code{0}.
14024 Use this to control gradual transition from zero to full transparency.
14027 @subsection Commands
14028 This filter supports same @ref{commands} as options.
14029 The command accepts the same syntax of the corresponding option.
14031 If the specified expression is not valid, it is kept at its current
14034 @section lut, lutrgb, lutyuv
14036 Compute a look-up table for binding each pixel component input value
14037 to an output value, and apply it to the input video.
14039 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
14040 to an RGB input video.
14042 These filters accept the following parameters:
14045 set first pixel component expression
14047 set second pixel component expression
14049 set third pixel component expression
14051 set fourth pixel component expression, corresponds to the alpha component
14054 set red component expression
14056 set green component expression
14058 set blue component expression
14060 alpha component expression
14063 set Y/luminance component expression
14065 set U/Cb component expression
14067 set V/Cr component expression
14070 Each of them specifies the expression to use for computing the lookup table for
14071 the corresponding pixel component values.
14073 The exact component associated to each of the @var{c*} options depends on the
14076 The @var{lut} filter requires either YUV or RGB pixel formats in input,
14077 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
14079 The expressions can contain the following constants and functions:
14084 The input width and height.
14087 The input value for the pixel component.
14090 The input value, clipped to the @var{minval}-@var{maxval} range.
14093 The maximum value for the pixel component.
14096 The minimum value for the pixel component.
14099 The negated value for the pixel component value, clipped to the
14100 @var{minval}-@var{maxval} range; it corresponds to the expression
14101 "maxval-clipval+minval".
14104 The computed value in @var{val}, clipped to the
14105 @var{minval}-@var{maxval} range.
14107 @item gammaval(gamma)
14108 The computed gamma correction value of the pixel component value,
14109 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
14111 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
14115 All expressions default to "val".
14117 @subsection Commands
14119 This filter supports same @ref{commands} as options.
14121 @subsection Examples
14125 Negate input video:
14127 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
14128 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
14131 The above is the same as:
14133 lutrgb="r=negval:g=negval:b=negval"
14134 lutyuv="y=negval:u=negval:v=negval"
14144 Remove chroma components, turning the video into a graytone image:
14146 lutyuv="u=128:v=128"
14150 Apply a luma burning effect:
14156 Remove green and blue components:
14162 Set a constant alpha channel value on input:
14164 format=rgba,lutrgb=a="maxval-minval/2"
14168 Correct luminance gamma by a factor of 0.5:
14170 lutyuv=y=gammaval(0.5)
14174 Discard least significant bits of luma:
14176 lutyuv=y='bitand(val, 128+64+32)'
14180 Technicolor like effect:
14182 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
14186 @section lut2, tlut2
14188 The @code{lut2} filter takes two input streams and outputs one
14191 The @code{tlut2} (time lut2) filter takes two consecutive frames
14192 from one single stream.
14194 This filter accepts the following parameters:
14197 set first pixel component expression
14199 set second pixel component expression
14201 set third pixel component expression
14203 set fourth pixel component expression, corresponds to the alpha component
14206 set output bit depth, only available for @code{lut2} filter. By default is 0,
14207 which means bit depth is automatically picked from first input format.
14210 The @code{lut2} filter also supports the @ref{framesync} options.
14212 Each of them specifies the expression to use for computing the lookup table for
14213 the corresponding pixel component values.
14215 The exact component associated to each of the @var{c*} options depends on the
14218 The expressions can contain the following constants:
14223 The input width and height.
14226 The first input value for the pixel component.
14229 The second input value for the pixel component.
14232 The first input video bit depth.
14235 The second input video bit depth.
14238 All expressions default to "x".
14240 @subsection Commands
14242 This filter supports the all above options as @ref{commands} except option @code{d}.
14244 @subsection Examples
14248 Highlight differences between two RGB video streams:
14250 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)'
14254 Highlight differences between two YUV video streams:
14256 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)'
14260 Show max difference between two video streams:
14262 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)))'
14266 @section maskedclamp
14268 Clamp the first input stream with the second input and third input stream.
14270 Returns the value of first stream to be between second input
14271 stream - @code{undershoot} and third input stream + @code{overshoot}.
14273 This filter accepts the following options:
14276 Default value is @code{0}.
14279 Default value is @code{0}.
14282 Set which planes will be processed as bitmap, unprocessed planes will be
14283 copied from first stream.
14284 By default value 0xf, all planes will be processed.
14287 @subsection Commands
14289 This filter supports the all above options as @ref{commands}.
14293 Merge the second and third input stream into output stream using absolute differences
14294 between second input stream and first input stream and absolute difference between
14295 third input stream and first input stream. The picked value will be from second input
14296 stream if second absolute difference is greater than first one or from third input stream
14299 This filter accepts the following options:
14302 Set which planes will be processed as bitmap, unprocessed planes will be
14303 copied from first stream.
14304 By default value 0xf, all planes will be processed.
14307 @subsection Commands
14309 This filter supports the all above options as @ref{commands}.
14311 @section maskedmerge
14313 Merge the first input stream with the second input stream using per pixel
14314 weights in the third input stream.
14316 A value of 0 in the third stream pixel component means that pixel component
14317 from first stream is returned unchanged, while maximum value (eg. 255 for
14318 8-bit videos) means that pixel component from second stream is returned
14319 unchanged. Intermediate values define the amount of merging between both
14320 input stream's pixel components.
14322 This filter accepts the following options:
14325 Set which planes will be processed as bitmap, unprocessed planes will be
14326 copied from first stream.
14327 By default value 0xf, all planes will be processed.
14330 @subsection Commands
14332 This filter supports the all above options as @ref{commands}.
14336 Merge the second and third input stream into output stream using absolute differences
14337 between second input stream and first input stream and absolute difference between
14338 third input stream and first input stream. The picked value will be from second input
14339 stream if second absolute difference is less than first one or from third input stream
14342 This filter accepts the following options:
14345 Set which planes will be processed as bitmap, unprocessed planes will be
14346 copied from first stream.
14347 By default value 0xf, all planes will be processed.
14350 @subsection Commands
14352 This filter supports the all above options as @ref{commands}.
14354 @section maskedthreshold
14355 Pick pixels comparing absolute difference of two video streams with fixed
14358 If absolute difference between pixel component of first and second video
14359 stream is equal or lower than user supplied threshold than pixel component
14360 from first video stream is picked, otherwise pixel component from second
14361 video stream is picked.
14363 This filter accepts the following options:
14366 Set threshold used when picking pixels from absolute difference from two input
14370 Set which planes will be processed as bitmap, unprocessed planes will be
14371 copied from second stream.
14372 By default value 0xf, all planes will be processed.
14375 @subsection Commands
14377 This filter supports the all above options as @ref{commands}.
14380 Create mask from input video.
14382 For example it is useful to create motion masks after @code{tblend} filter.
14384 This filter accepts the following options:
14388 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
14391 Set high threshold. Any pixel component higher than this value will be set to max value
14392 allowed for current pixel format.
14395 Set planes to filter, by default all available planes are filtered.
14398 Fill all frame pixels with this value.
14401 Set max average pixel value for frame. If sum of all pixel components is higher that this
14402 average, output frame will be completely filled with value set by @var{fill} option.
14403 Typically useful for scene changes when used in combination with @code{tblend} filter.
14406 @subsection Commands
14408 This filter supports the all above options as @ref{commands}.
14412 Apply motion-compensation deinterlacing.
14414 It needs one field per frame as input and must thus be used together
14415 with yadif=1/3 or equivalent.
14417 This filter accepts the following options:
14420 Set the deinterlacing mode.
14422 It accepts one of the following values:
14427 use iterative motion estimation
14429 like @samp{slow}, but use multiple reference frames.
14431 Default value is @samp{fast}.
14434 Set the picture field parity assumed for the input video. It must be
14435 one of the following values:
14439 assume top field first
14441 assume bottom field first
14444 Default value is @samp{bff}.
14447 Set per-block quantization parameter (QP) used by the internal
14450 Higher values should result in a smoother motion vector field but less
14451 optimal individual vectors. Default value is 1.
14456 Pick median pixel from certain rectangle defined by radius.
14458 This filter accepts the following options:
14462 Set horizontal radius size. Default value is @code{1}.
14463 Allowed range is integer from 1 to 127.
14466 Set which planes to process. Default is @code{15}, which is all available planes.
14469 Set vertical radius size. Default value is @code{0}.
14470 Allowed range is integer from 0 to 127.
14471 If it is 0, value will be picked from horizontal @code{radius} option.
14474 Set median percentile. Default value is @code{0.5}.
14475 Default value of @code{0.5} will pick always median values, while @code{0} will pick
14476 minimum values, and @code{1} maximum values.
14479 @subsection Commands
14480 This filter supports same @ref{commands} as options.
14481 The command accepts the same syntax of the corresponding option.
14483 If the specified expression is not valid, it is kept at its current
14486 @section mergeplanes
14488 Merge color channel components from several video streams.
14490 The filter accepts up to 4 input streams, and merge selected input
14491 planes to the output video.
14493 This filter accepts the following options:
14496 Set input to output plane mapping. Default is @code{0}.
14498 The mappings is specified as a bitmap. It should be specified as a
14499 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
14500 mapping for the first plane of the output stream. 'A' sets the number of
14501 the input stream to use (from 0 to 3), and 'a' the plane number of the
14502 corresponding input to use (from 0 to 3). The rest of the mappings is
14503 similar, 'Bb' describes the mapping for the output stream second
14504 plane, 'Cc' describes the mapping for the output stream third plane and
14505 'Dd' describes the mapping for the output stream fourth plane.
14508 Set output pixel format. Default is @code{yuva444p}.
14511 @subsection Examples
14515 Merge three gray video streams of same width and height into single video stream:
14517 [a0][a1][a2]mergeplanes=0x001020:yuv444p
14521 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
14523 [a0][a1]mergeplanes=0x00010210:yuva444p
14527 Swap Y and A plane in yuva444p stream:
14529 format=yuva444p,mergeplanes=0x03010200:yuva444p
14533 Swap U and V plane in yuv420p stream:
14535 format=yuv420p,mergeplanes=0x000201:yuv420p
14539 Cast a rgb24 clip to yuv444p:
14541 format=rgb24,mergeplanes=0x000102:yuv444p
14547 Estimate and export motion vectors using block matching algorithms.
14548 Motion vectors are stored in frame side data to be used by other filters.
14550 This filter accepts the following options:
14553 Specify the motion estimation method. Accepts one of the following values:
14557 Exhaustive search algorithm.
14559 Three step search algorithm.
14561 Two dimensional logarithmic search algorithm.
14563 New three step search algorithm.
14565 Four step search algorithm.
14567 Diamond search algorithm.
14569 Hexagon-based search algorithm.
14571 Enhanced predictive zonal search algorithm.
14573 Uneven multi-hexagon search algorithm.
14575 Default value is @samp{esa}.
14578 Macroblock size. Default @code{16}.
14581 Search parameter. Default @code{7}.
14584 @section midequalizer
14586 Apply Midway Image Equalization effect using two video streams.
14588 Midway Image Equalization adjusts a pair of images to have the same
14589 histogram, while maintaining their dynamics as much as possible. It's
14590 useful for e.g. matching exposures from a pair of stereo cameras.
14592 This filter has two inputs and one output, which must be of same pixel format, but
14593 may be of different sizes. The output of filter is first input adjusted with
14594 midway histogram of both inputs.
14596 This filter accepts the following option:
14600 Set which planes to process. Default is @code{15}, which is all available planes.
14603 @section minterpolate
14605 Convert the video to specified frame rate using motion interpolation.
14607 This filter accepts the following options:
14610 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}.
14613 Motion interpolation mode. Following values are accepted:
14616 Duplicate previous or next frame for interpolating new ones.
14618 Blend source frames. Interpolated frame is mean of previous and next frames.
14620 Motion compensated interpolation. Following options are effective when this mode is selected:
14624 Motion compensation mode. Following values are accepted:
14627 Overlapped block motion compensation.
14629 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
14631 Default mode is @samp{obmc}.
14634 Motion estimation mode. Following values are accepted:
14637 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
14639 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
14641 Default mode is @samp{bilat}.
14644 The algorithm to be used for motion estimation. Following values are accepted:
14647 Exhaustive search algorithm.
14649 Three step search algorithm.
14651 Two dimensional logarithmic search algorithm.
14653 New three step search algorithm.
14655 Four step search algorithm.
14657 Diamond search algorithm.
14659 Hexagon-based search algorithm.
14661 Enhanced predictive zonal search algorithm.
14663 Uneven multi-hexagon search algorithm.
14665 Default algorithm is @samp{epzs}.
14668 Macroblock size. Default @code{16}.
14671 Motion estimation search parameter. Default @code{32}.
14674 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).
14679 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:
14682 Disable scene change detection.
14684 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
14686 Default method is @samp{fdiff}.
14688 @item scd_threshold
14689 Scene change detection threshold. Default is @code{10.}.
14694 Mix several video input streams into one video stream.
14696 A description of the accepted options follows.
14700 The number of inputs. If unspecified, it defaults to 2.
14703 Specify weight of each input video stream as sequence.
14704 Each weight is separated by space. If number of weights
14705 is smaller than number of @var{frames} last specified
14706 weight will be used for all remaining unset weights.
14709 Specify scale, if it is set it will be multiplied with sum
14710 of each weight multiplied with pixel values to give final destination
14711 pixel value. By default @var{scale} is auto scaled to sum of weights.
14714 Specify how end of stream is determined.
14717 The duration of the longest input. (default)
14720 The duration of the shortest input.
14723 The duration of the first input.
14727 @subsection Commands
14729 This filter supports the following commands:
14733 Syntax is same as option with same name.
14736 @section mpdecimate
14738 Drop frames that do not differ greatly from the previous frame in
14739 order to reduce frame rate.
14741 The main use of this filter is for very-low-bitrate encoding
14742 (e.g. streaming over dialup modem), but it could in theory be used for
14743 fixing movies that were inverse-telecined incorrectly.
14745 A description of the accepted options follows.
14749 Set the maximum number of consecutive frames which can be dropped (if
14750 positive), or the minimum interval between dropped frames (if
14751 negative). If the value is 0, the frame is dropped disregarding the
14752 number of previous sequentially dropped frames.
14754 Default value is 0.
14759 Set the dropping threshold values.
14761 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
14762 represent actual pixel value differences, so a threshold of 64
14763 corresponds to 1 unit of difference for each pixel, or the same spread
14764 out differently over the block.
14766 A frame is a candidate for dropping if no 8x8 blocks differ by more
14767 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
14768 meaning the whole image) differ by more than a threshold of @option{lo}.
14770 Default value for @option{hi} is 64*12, default value for @option{lo} is
14771 64*5, and default value for @option{frac} is 0.33.
14777 Negate (invert) the input video.
14779 It accepts the following option:
14784 With value 1, it negates the alpha component, if present. Default value is 0.
14787 @subsection Commands
14789 This filter supports same @ref{commands} as options.
14794 Denoise frames using Non-Local Means algorithm.
14796 Each pixel is adjusted by looking for other pixels with similar contexts. This
14797 context similarity is defined by comparing their surrounding patches of size
14798 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
14801 Note that the research area defines centers for patches, which means some
14802 patches will be made of pixels outside that research area.
14804 The filter accepts the following options.
14808 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
14811 Set patch size. Default is 7. Must be odd number in range [0, 99].
14814 Same as @option{p} but for chroma planes.
14816 The default value is @var{0} and means automatic.
14819 Set research size. Default is 15. Must be odd number in range [0, 99].
14822 Same as @option{r} but for chroma planes.
14824 The default value is @var{0} and means automatic.
14829 Deinterlace video using neural network edge directed interpolation.
14831 This filter accepts the following options:
14835 Mandatory option, without binary file filter can not work.
14836 Currently file can be found here:
14837 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
14840 Set which frames to deinterlace, by default it is @code{all}.
14841 Can be @code{all} or @code{interlaced}.
14844 Set mode of operation.
14846 Can be one of the following:
14850 Use frame flags, both fields.
14852 Use frame flags, single field.
14854 Use top field only.
14856 Use bottom field only.
14858 Use both fields, top first.
14860 Use both fields, bottom first.
14864 Set which planes to process, by default filter process all frames.
14867 Set size of local neighborhood around each pixel, used by the predictor neural
14870 Can be one of the following:
14883 Set the number of neurons in predictor neural network.
14884 Can be one of the following:
14895 Controls the number of different neural network predictions that are blended
14896 together to compute the final output value. Can be @code{fast}, default or
14900 Set which set of weights to use in the predictor.
14901 Can be one of the following:
14905 weights trained to minimize absolute error
14907 weights trained to minimize squared error
14911 Controls whether or not the prescreener neural network is used to decide
14912 which pixels should be processed by the predictor neural network and which
14913 can be handled by simple cubic interpolation.
14914 The prescreener is trained to know whether cubic interpolation will be
14915 sufficient for a pixel or whether it should be predicted by the predictor nn.
14916 The computational complexity of the prescreener nn is much less than that of
14917 the predictor nn. Since most pixels can be handled by cubic interpolation,
14918 using the prescreener generally results in much faster processing.
14919 The prescreener is pretty accurate, so the difference between using it and not
14920 using it is almost always unnoticeable.
14922 Can be one of the following:
14932 Default is @code{new}.
14935 @subsection Commands
14936 This filter supports same @ref{commands} as options, excluding @var{weights} option.
14940 Force libavfilter not to use any of the specified pixel formats for the
14941 input to the next filter.
14943 It accepts the following parameters:
14947 A '|'-separated list of pixel format names, such as
14948 pix_fmts=yuv420p|monow|rgb24".
14952 @subsection Examples
14956 Force libavfilter to use a format different from @var{yuv420p} for the
14957 input to the vflip filter:
14959 noformat=pix_fmts=yuv420p,vflip
14963 Convert the input video to any of the formats not contained in the list:
14965 noformat=yuv420p|yuv444p|yuv410p
14971 Add noise on video input frame.
14973 The filter accepts the following options:
14981 Set noise seed for specific pixel component or all pixel components in case
14982 of @var{all_seed}. Default value is @code{123457}.
14984 @item all_strength, alls
14985 @item c0_strength, c0s
14986 @item c1_strength, c1s
14987 @item c2_strength, c2s
14988 @item c3_strength, c3s
14989 Set noise strength for specific pixel component or all pixel components in case
14990 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
14992 @item all_flags, allf
14993 @item c0_flags, c0f
14994 @item c1_flags, c1f
14995 @item c2_flags, c2f
14996 @item c3_flags, c3f
14997 Set pixel component flags or set flags for all components if @var{all_flags}.
14998 Available values for component flags are:
15001 averaged temporal noise (smoother)
15003 mix random noise with a (semi)regular pattern
15005 temporal noise (noise pattern changes between frames)
15007 uniform noise (gaussian otherwise)
15011 @subsection Examples
15013 Add temporal and uniform noise to input video:
15015 noise=alls=20:allf=t+u
15020 Normalize RGB video (aka histogram stretching, contrast stretching).
15021 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
15023 For each channel of each frame, the filter computes the input range and maps
15024 it linearly to the user-specified output range. The output range defaults
15025 to the full dynamic range from pure black to pure white.
15027 Temporal smoothing can be used on the input range to reduce flickering (rapid
15028 changes in brightness) caused when small dark or bright objects enter or leave
15029 the scene. This is similar to the auto-exposure (automatic gain control) on a
15030 video camera, and, like a video camera, it may cause a period of over- or
15031 under-exposure of the video.
15033 The R,G,B channels can be normalized independently, which may cause some
15034 color shifting, or linked together as a single channel, which prevents
15035 color shifting. Linked normalization preserves hue. Independent normalization
15036 does not, so it can be used to remove some color casts. Independent and linked
15037 normalization can be combined in any ratio.
15039 The normalize filter accepts the following options:
15044 Colors which define the output range. The minimum input value is mapped to
15045 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
15046 The defaults are black and white respectively. Specifying white for
15047 @var{blackpt} and black for @var{whitept} will give color-inverted,
15048 normalized video. Shades of grey can be used to reduce the dynamic range
15049 (contrast). Specifying saturated colors here can create some interesting
15053 The number of previous frames to use for temporal smoothing. The input range
15054 of each channel is smoothed using a rolling average over the current frame
15055 and the @var{smoothing} previous frames. The default is 0 (no temporal
15059 Controls the ratio of independent (color shifting) channel normalization to
15060 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
15061 independent. Defaults to 1.0 (fully independent).
15064 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
15065 expensive no-op. Defaults to 1.0 (full strength).
15069 @subsection Commands
15070 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
15071 The command accepts the same syntax of the corresponding option.
15073 If the specified expression is not valid, it is kept at its current
15076 @subsection Examples
15078 Stretch video contrast to use the full dynamic range, with no temporal
15079 smoothing; may flicker depending on the source content:
15081 normalize=blackpt=black:whitept=white:smoothing=0
15084 As above, but with 50 frames of temporal smoothing; flicker should be
15085 reduced, depending on the source content:
15087 normalize=blackpt=black:whitept=white:smoothing=50
15090 As above, but with hue-preserving linked channel normalization:
15092 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
15095 As above, but with half strength:
15097 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
15100 Map the darkest input color to red, the brightest input color to cyan:
15102 normalize=blackpt=red:whitept=cyan
15107 Pass the video source unchanged to the output.
15110 Optical Character Recognition
15112 This filter uses Tesseract for optical character recognition. To enable
15113 compilation of this filter, you need to configure FFmpeg with
15114 @code{--enable-libtesseract}.
15116 It accepts the following options:
15120 Set datapath to tesseract data. Default is to use whatever was
15121 set at installation.
15124 Set language, default is "eng".
15127 Set character whitelist.
15130 Set character blacklist.
15133 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
15134 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
15138 Apply a video transform using libopencv.
15140 To enable this filter, install the libopencv library and headers and
15141 configure FFmpeg with @code{--enable-libopencv}.
15143 It accepts the following parameters:
15148 The name of the libopencv filter to apply.
15150 @item filter_params
15151 The parameters to pass to the libopencv filter. If not specified, the default
15152 values are assumed.
15156 Refer to the official libopencv documentation for more precise
15158 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
15160 Several libopencv filters are supported; see the following subsections.
15165 Dilate an image by using a specific structuring element.
15166 It corresponds to the libopencv function @code{cvDilate}.
15168 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
15170 @var{struct_el} represents a structuring element, and has the syntax:
15171 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
15173 @var{cols} and @var{rows} represent the number of columns and rows of
15174 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
15175 point, and @var{shape} the shape for the structuring element. @var{shape}
15176 must be "rect", "cross", "ellipse", or "custom".
15178 If the value for @var{shape} is "custom", it must be followed by a
15179 string of the form "=@var{filename}". The file with name
15180 @var{filename} is assumed to represent a binary image, with each
15181 printable character corresponding to a bright pixel. When a custom
15182 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
15183 or columns and rows of the read file are assumed instead.
15185 The default value for @var{struct_el} is "3x3+0x0/rect".
15187 @var{nb_iterations} specifies the number of times the transform is
15188 applied to the image, and defaults to 1.
15192 # Use the default values
15195 # Dilate using a structuring element with a 5x5 cross, iterating two times
15196 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
15198 # Read the shape from the file diamond.shape, iterating two times.
15199 # The file diamond.shape may contain a pattern of characters like this
15205 # The specified columns and rows are ignored
15206 # but the anchor point coordinates are not
15207 ocv=dilate:0x0+2x2/custom=diamond.shape|2
15212 Erode an image by using a specific structuring element.
15213 It corresponds to the libopencv function @code{cvErode}.
15215 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
15216 with the same syntax and semantics as the @ref{dilate} filter.
15220 Smooth the input video.
15222 The filter takes the following parameters:
15223 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
15225 @var{type} is the type of smooth filter to apply, and must be one of
15226 the following values: "blur", "blur_no_scale", "median", "gaussian",
15227 or "bilateral". The default value is "gaussian".
15229 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
15230 depends on the smooth type. @var{param1} and
15231 @var{param2} accept integer positive values or 0. @var{param3} and
15232 @var{param4} accept floating point values.
15234 The default value for @var{param1} is 3. The default value for the
15235 other parameters is 0.
15237 These parameters correspond to the parameters assigned to the
15238 libopencv function @code{cvSmooth}.
15240 @section oscilloscope
15242 2D Video Oscilloscope.
15244 Useful to measure spatial impulse, step responses, chroma delays, etc.
15246 It accepts the following parameters:
15250 Set scope center x position.
15253 Set scope center y position.
15256 Set scope size, relative to frame diagonal.
15259 Set scope tilt/rotation.
15265 Set trace center x position.
15268 Set trace center y position.
15271 Set trace width, relative to width of frame.
15274 Set trace height, relative to height of frame.
15277 Set which components to trace. By default it traces first three components.
15280 Draw trace grid. By default is enabled.
15283 Draw some statistics. By default is enabled.
15286 Draw scope. By default is enabled.
15289 @subsection Commands
15290 This filter supports same @ref{commands} as options.
15291 The command accepts the same syntax of the corresponding option.
15293 If the specified expression is not valid, it is kept at its current
15296 @subsection Examples
15300 Inspect full first row of video frame.
15302 oscilloscope=x=0.5:y=0:s=1
15306 Inspect full last row of video frame.
15308 oscilloscope=x=0.5:y=1:s=1
15312 Inspect full 5th line of video frame of height 1080.
15314 oscilloscope=x=0.5:y=5/1080:s=1
15318 Inspect full last column of video frame.
15320 oscilloscope=x=1:y=0.5:s=1:t=1
15328 Overlay one video on top of another.
15330 It takes two inputs and has one output. The first input is the "main"
15331 video on which the second input is overlaid.
15333 It accepts the following parameters:
15335 A description of the accepted options follows.
15340 Set the expression for the x and y coordinates of the overlaid video
15341 on the main video. Default value is "0" for both expressions. In case
15342 the expression is invalid, it is set to a huge value (meaning that the
15343 overlay will not be displayed within the output visible area).
15346 See @ref{framesync}.
15349 Set when the expressions for @option{x}, and @option{y} are evaluated.
15351 It accepts the following values:
15354 only evaluate expressions once during the filter initialization or
15355 when a command is processed
15358 evaluate expressions for each incoming frame
15361 Default value is @samp{frame}.
15364 See @ref{framesync}.
15367 Set the format for the output video.
15369 It accepts the following values:
15372 force YUV420 output
15375 force YUV420p10 output
15378 force YUV422 output
15381 force YUV422p10 output
15384 force YUV444 output
15387 force packed RGB output
15390 force planar RGB output
15393 automatically pick format
15396 Default value is @samp{yuv420}.
15399 See @ref{framesync}.
15402 Set format of alpha of the overlaid video, it can be @var{straight} or
15403 @var{premultiplied}. Default is @var{straight}.
15406 The @option{x}, and @option{y} expressions can contain the following
15412 The main input width and height.
15416 The overlay input width and height.
15420 The computed values for @var{x} and @var{y}. They are evaluated for
15425 horizontal and vertical chroma subsample values of the output
15426 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
15430 the number of input frame, starting from 0
15433 the position in the file of the input frame, NAN if unknown
15436 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
15440 This filter also supports the @ref{framesync} options.
15442 Note that the @var{n}, @var{pos}, @var{t} variables are available only
15443 when evaluation is done @emph{per frame}, and will evaluate to NAN
15444 when @option{eval} is set to @samp{init}.
15446 Be aware that frames are taken from each input video in timestamp
15447 order, hence, if their initial timestamps differ, it is a good idea
15448 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
15449 have them begin in the same zero timestamp, as the example for
15450 the @var{movie} filter does.
15452 You can chain together more overlays but you should test the
15453 efficiency of such approach.
15455 @subsection Commands
15457 This filter supports the following commands:
15461 Modify the x and y of the overlay input.
15462 The command accepts the same syntax of the corresponding option.
15464 If the specified expression is not valid, it is kept at its current
15468 @subsection Examples
15472 Draw the overlay at 10 pixels from the bottom right corner of the main
15475 overlay=main_w-overlay_w-10:main_h-overlay_h-10
15478 Using named options the example above becomes:
15480 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
15484 Insert a transparent PNG logo in the bottom left corner of the input,
15485 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
15487 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
15491 Insert 2 different transparent PNG logos (second logo on bottom
15492 right corner) using the @command{ffmpeg} tool:
15494 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
15498 Add a transparent color layer on top of the main video; @code{WxH}
15499 must specify the size of the main input to the overlay filter:
15501 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
15505 Play an original video and a filtered version (here with the deshake
15506 filter) side by side using the @command{ffplay} tool:
15508 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
15511 The above command is the same as:
15513 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
15517 Make a sliding overlay appearing from the left to the right top part of the
15518 screen starting since time 2:
15520 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
15524 Compose output by putting two input videos side to side:
15526 ffmpeg -i left.avi -i right.avi -filter_complex "
15527 nullsrc=size=200x100 [background];
15528 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
15529 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
15530 [background][left] overlay=shortest=1 [background+left];
15531 [background+left][right] overlay=shortest=1:x=100 [left+right]
15536 Mask 10-20 seconds of a video by applying the delogo filter to a section
15538 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
15539 -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]'
15544 Chain several overlays in cascade:
15546 nullsrc=s=200x200 [bg];
15547 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
15548 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
15549 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
15550 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
15551 [in3] null, [mid2] overlay=100:100 [out0]
15556 @anchor{overlay_cuda}
15557 @section overlay_cuda
15559 Overlay one video on top of another.
15561 This is the CUDA variant of the @ref{overlay} filter.
15562 It only accepts CUDA frames. The underlying input pixel formats have to match.
15564 It takes two inputs and has one output. The first input is the "main"
15565 video on which the second input is overlaid.
15567 It accepts the following parameters:
15572 Set the x and y coordinates of the overlaid video on the main video.
15573 Default value is "0" for both expressions.
15576 See @ref{framesync}.
15579 See @ref{framesync}.
15582 See @ref{framesync}.
15586 This filter also supports the @ref{framesync} options.
15590 Apply Overcomplete Wavelet denoiser.
15592 The filter accepts the following options:
15598 Larger depth values will denoise lower frequency components more, but
15599 slow down filtering.
15601 Must be an int in the range 8-16, default is @code{8}.
15603 @item luma_strength, ls
15606 Must be a double value in the range 0-1000, default is @code{1.0}.
15608 @item chroma_strength, cs
15609 Set chroma strength.
15611 Must be a double value in the range 0-1000, default is @code{1.0}.
15617 Add paddings to the input image, and place the original input at the
15618 provided @var{x}, @var{y} coordinates.
15620 It accepts the following parameters:
15625 Specify an expression for the size of the output image with the
15626 paddings added. If the value for @var{width} or @var{height} is 0, the
15627 corresponding input size is used for the output.
15629 The @var{width} expression can reference the value set by the
15630 @var{height} expression, and vice versa.
15632 The default value of @var{width} and @var{height} is 0.
15636 Specify the offsets to place the input image at within the padded area,
15637 with respect to the top/left border of the output image.
15639 The @var{x} expression can reference the value set by the @var{y}
15640 expression, and vice versa.
15642 The default value of @var{x} and @var{y} is 0.
15644 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
15645 so the input image is centered on the padded area.
15648 Specify the color of the padded area. For the syntax of this option,
15649 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15650 manual,ffmpeg-utils}.
15652 The default value of @var{color} is "black".
15655 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
15657 It accepts the following values:
15661 Only evaluate expressions once during the filter initialization or when
15662 a command is processed.
15665 Evaluate expressions for each incoming frame.
15669 Default value is @samp{init}.
15672 Pad to aspect instead to a resolution.
15676 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
15677 options are expressions containing the following constants:
15682 The input video width and height.
15686 These are the same as @var{in_w} and @var{in_h}.
15690 The output width and height (the size of the padded area), as
15691 specified by the @var{width} and @var{height} expressions.
15695 These are the same as @var{out_w} and @var{out_h}.
15699 The x and y offsets as specified by the @var{x} and @var{y}
15700 expressions, or NAN if not yet specified.
15703 same as @var{iw} / @var{ih}
15706 input sample aspect ratio
15709 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
15713 The horizontal and vertical chroma subsample values. For example for the
15714 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15717 @subsection Examples
15721 Add paddings with the color "violet" to the input video. The output video
15722 size is 640x480, and the top-left corner of the input video is placed at
15725 pad=640:480:0:40:violet
15728 The example above is equivalent to the following command:
15730 pad=width=640:height=480:x=0:y=40:color=violet
15734 Pad the input to get an output with dimensions increased by 3/2,
15735 and put the input video at the center of the padded area:
15737 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
15741 Pad the input to get a squared output with size equal to the maximum
15742 value between the input width and height, and put the input video at
15743 the center of the padded area:
15745 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
15749 Pad the input to get a final w/h ratio of 16:9:
15751 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
15755 In case of anamorphic video, in order to set the output display aspect
15756 correctly, it is necessary to use @var{sar} in the expression,
15757 according to the relation:
15759 (ih * X / ih) * sar = output_dar
15760 X = output_dar / sar
15763 Thus the previous example needs to be modified to:
15765 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
15769 Double the output size and put the input video in the bottom-right
15770 corner of the output padded area:
15772 pad="2*iw:2*ih:ow-iw:oh-ih"
15776 @anchor{palettegen}
15777 @section palettegen
15779 Generate one palette for a whole video stream.
15781 It accepts the following options:
15785 Set the maximum number of colors to quantize in the palette.
15786 Note: the palette will still contain 256 colors; the unused palette entries
15789 @item reserve_transparent
15790 Create a palette of 255 colors maximum and reserve the last one for
15791 transparency. Reserving the transparency color is useful for GIF optimization.
15792 If not set, the maximum of colors in the palette will be 256. You probably want
15793 to disable this option for a standalone image.
15796 @item transparency_color
15797 Set the color that will be used as background for transparency.
15800 Set statistics mode.
15802 It accepts the following values:
15805 Compute full frame histograms.
15807 Compute histograms only for the part that differs from previous frame. This
15808 might be relevant to give more importance to the moving part of your input if
15809 the background is static.
15811 Compute new histogram for each frame.
15814 Default value is @var{full}.
15817 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
15818 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
15819 color quantization of the palette. This information is also visible at
15820 @var{info} logging level.
15822 @subsection Examples
15826 Generate a representative palette of a given video using @command{ffmpeg}:
15828 ffmpeg -i input.mkv -vf palettegen palette.png
15832 @section paletteuse
15834 Use a palette to downsample an input video stream.
15836 The filter takes two inputs: one video stream and a palette. The palette must
15837 be a 256 pixels image.
15839 It accepts the following options:
15843 Select dithering mode. Available algorithms are:
15846 Ordered 8x8 bayer dithering (deterministic)
15848 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
15849 Note: this dithering is sometimes considered "wrong" and is included as a
15851 @item floyd_steinberg
15852 Floyd and Steingberg dithering (error diffusion)
15854 Frankie Sierra dithering v2 (error diffusion)
15856 Frankie Sierra dithering v2 "Lite" (error diffusion)
15859 Default is @var{sierra2_4a}.
15862 When @var{bayer} dithering is selected, this option defines the scale of the
15863 pattern (how much the crosshatch pattern is visible). A low value means more
15864 visible pattern for less banding, and higher value means less visible pattern
15865 at the cost of more banding.
15867 The option must be an integer value in the range [0,5]. Default is @var{2}.
15870 If set, define the zone to process
15874 Only the changing rectangle will be reprocessed. This is similar to GIF
15875 cropping/offsetting compression mechanism. This option can be useful for speed
15876 if only a part of the image is changing, and has use cases such as limiting the
15877 scope of the error diffusal @option{dither} to the rectangle that bounds the
15878 moving scene (it leads to more deterministic output if the scene doesn't change
15879 much, and as a result less moving noise and better GIF compression).
15882 Default is @var{none}.
15885 Take new palette for each output frame.
15887 @item alpha_threshold
15888 Sets the alpha threshold for transparency. Alpha values above this threshold
15889 will be treated as completely opaque, and values below this threshold will be
15890 treated as completely transparent.
15892 The option must be an integer value in the range [0,255]. Default is @var{128}.
15895 @subsection Examples
15899 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
15900 using @command{ffmpeg}:
15902 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
15906 @section perspective
15908 Correct perspective of video not recorded perpendicular to the screen.
15910 A description of the accepted parameters follows.
15921 Set coordinates expression for top left, top right, bottom left and bottom right corners.
15922 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
15923 If the @code{sense} option is set to @code{source}, then the specified points will be sent
15924 to the corners of the destination. If the @code{sense} option is set to @code{destination},
15925 then the corners of the source will be sent to the specified coordinates.
15927 The expressions can use the following variables:
15932 the width and height of video frame.
15936 Output frame count.
15939 @item interpolation
15940 Set interpolation for perspective correction.
15942 It accepts the following values:
15948 Default value is @samp{linear}.
15951 Set interpretation of coordinate options.
15953 It accepts the following values:
15957 Send point in the source specified by the given coordinates to
15958 the corners of the destination.
15960 @item 1, destination
15962 Send the corners of the source to the point in the destination specified
15963 by the given coordinates.
15965 Default value is @samp{source}.
15969 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
15971 It accepts the following values:
15974 only evaluate expressions once during the filter initialization or
15975 when a command is processed
15978 evaluate expressions for each incoming frame
15981 Default value is @samp{init}.
15986 Delay interlaced video by one field time so that the field order changes.
15988 The intended use is to fix PAL movies that have been captured with the
15989 opposite field order to the film-to-video transfer.
15991 A description of the accepted parameters follows.
15997 It accepts the following values:
16000 Capture field order top-first, transfer bottom-first.
16001 Filter will delay the bottom field.
16004 Capture field order bottom-first, transfer top-first.
16005 Filter will delay the top field.
16008 Capture and transfer with the same field order. This mode only exists
16009 for the documentation of the other options to refer to, but if you
16010 actually select it, the filter will faithfully do nothing.
16013 Capture field order determined automatically by field flags, transfer
16015 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
16016 basis using field flags. If no field information is available,
16017 then this works just like @samp{u}.
16020 Capture unknown or varying, transfer opposite.
16021 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
16022 analyzing the images and selecting the alternative that produces best
16023 match between the fields.
16026 Capture top-first, transfer unknown or varying.
16027 Filter selects among @samp{t} and @samp{p} using image analysis.
16030 Capture bottom-first, transfer unknown or varying.
16031 Filter selects among @samp{b} and @samp{p} using image analysis.
16034 Capture determined by field flags, transfer unknown or varying.
16035 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
16036 image analysis. If no field information is available, then this works just
16037 like @samp{U}. This is the default mode.
16040 Both capture and transfer unknown or varying.
16041 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
16045 @subsection Commands
16047 This filter supports the all above options as @ref{commands}.
16049 @section photosensitivity
16050 Reduce various flashes in video, so to help users with epilepsy.
16052 It accepts the following options:
16055 Set how many frames to use when filtering. Default is 30.
16058 Set detection threshold factor. Default is 1.
16062 Set how many pixels to skip when sampling frames. Default is 1.
16063 Allowed range is from 1 to 1024.
16066 Leave frames unchanged. Default is disabled.
16069 @section pixdesctest
16071 Pixel format descriptor test filter, mainly useful for internal
16072 testing. The output video should be equal to the input video.
16076 format=monow, pixdesctest
16079 can be used to test the monowhite pixel format descriptor definition.
16083 Display sample values of color channels. Mainly useful for checking color
16084 and levels. Minimum supported resolution is 640x480.
16086 The filters accept the following options:
16090 Set scope X position, relative offset on X axis.
16093 Set scope Y position, relative offset on Y axis.
16102 Set window opacity. This window also holds statistics about pixel area.
16105 Set window X position, relative offset on X axis.
16108 Set window Y position, relative offset on Y axis.
16113 Enable the specified chain of postprocessing subfilters using libpostproc. This
16114 library should be automatically selected with a GPL build (@code{--enable-gpl}).
16115 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
16116 Each subfilter and some options have a short and a long name that can be used
16117 interchangeably, i.e. dr/dering are the same.
16119 The filters accept the following options:
16123 Set postprocessing subfilters string.
16126 All subfilters share common options to determine their scope:
16130 Honor the quality commands for this subfilter.
16133 Do chrominance filtering, too (default).
16136 Do luminance filtering only (no chrominance).
16139 Do chrominance filtering only (no luminance).
16142 These options can be appended after the subfilter name, separated by a '|'.
16144 Available subfilters are:
16147 @item hb/hdeblock[|difference[|flatness]]
16148 Horizontal deblocking filter
16151 Difference factor where higher values mean more deblocking (default: @code{32}).
16153 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16156 @item vb/vdeblock[|difference[|flatness]]
16157 Vertical deblocking filter
16160 Difference factor where higher values mean more deblocking (default: @code{32}).
16162 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16165 @item ha/hadeblock[|difference[|flatness]]
16166 Accurate horizontal deblocking filter
16169 Difference factor where higher values mean more deblocking (default: @code{32}).
16171 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16174 @item va/vadeblock[|difference[|flatness]]
16175 Accurate vertical deblocking filter
16178 Difference factor where higher values mean more deblocking (default: @code{32}).
16180 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16184 The horizontal and vertical deblocking filters share the difference and
16185 flatness values so you cannot set different horizontal and vertical
16189 @item h1/x1hdeblock
16190 Experimental horizontal deblocking filter
16192 @item v1/x1vdeblock
16193 Experimental vertical deblocking filter
16198 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
16201 larger -> stronger filtering
16203 larger -> stronger filtering
16205 larger -> stronger filtering
16208 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
16211 Stretch luminance to @code{0-255}.
16214 @item lb/linblenddeint
16215 Linear blend deinterlacing filter that deinterlaces the given block by
16216 filtering all lines with a @code{(1 2 1)} filter.
16218 @item li/linipoldeint
16219 Linear interpolating deinterlacing filter that deinterlaces the given block by
16220 linearly interpolating every second line.
16222 @item ci/cubicipoldeint
16223 Cubic interpolating deinterlacing filter deinterlaces the given block by
16224 cubically interpolating every second line.
16226 @item md/mediandeint
16227 Median deinterlacing filter that deinterlaces the given block by applying a
16228 median filter to every second line.
16230 @item fd/ffmpegdeint
16231 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
16232 second line with a @code{(-1 4 2 4 -1)} filter.
16235 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
16236 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
16238 @item fq/forceQuant[|quantizer]
16239 Overrides the quantizer table from the input with the constant quantizer you
16247 Default pp filter combination (@code{hb|a,vb|a,dr|a})
16250 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
16253 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
16256 @subsection Examples
16260 Apply horizontal and vertical deblocking, deringing and automatic
16261 brightness/contrast:
16267 Apply default filters without brightness/contrast correction:
16273 Apply default filters and temporal denoiser:
16275 pp=default/tmpnoise|1|2|3
16279 Apply deblocking on luminance only, and switch vertical deblocking on or off
16280 automatically depending on available CPU time:
16287 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
16288 similar to spp = 6 with 7 point DCT, where only the center sample is
16291 The filter accepts the following options:
16295 Force a constant quantization parameter. It accepts an integer in range
16296 0 to 63. If not set, the filter will use the QP from the video stream
16300 Set thresholding mode. Available modes are:
16304 Set hard thresholding.
16306 Set soft thresholding (better de-ringing effect, but likely blurrier).
16308 Set medium thresholding (good results, default).
16312 @section premultiply
16313 Apply alpha premultiply effect to input video stream using first plane
16314 of second stream as alpha.
16316 Both streams must have same dimensions and same pixel format.
16318 The filter accepts the following option:
16322 Set which planes will be processed, unprocessed planes will be copied.
16323 By default value 0xf, all planes will be processed.
16326 Do not require 2nd input for processing, instead use alpha plane from input stream.
16330 Apply prewitt operator to input video stream.
16332 The filter accepts the following option:
16336 Set which planes will be processed, unprocessed planes will be copied.
16337 By default value 0xf, all planes will be processed.
16340 Set value which will be multiplied with filtered result.
16343 Set value which will be added to filtered result.
16346 @subsection Commands
16348 This filter supports the all above options as @ref{commands}.
16350 @section pseudocolor
16352 Alter frame colors in video with pseudocolors.
16354 This filter accepts the following options:
16358 set pixel first component expression
16361 set pixel second component expression
16364 set pixel third component expression
16367 set pixel fourth component expression, corresponds to the alpha component
16370 set component to use as base for altering colors
16373 Pick one of built-in LUTs. By default is set to none.
16389 Each of them specifies the expression to use for computing the lookup table for
16390 the corresponding pixel component values.
16392 The expressions can contain the following constants and functions:
16397 The input width and height.
16400 The input value for the pixel component.
16402 @item ymin, umin, vmin, amin
16403 The minimum allowed component value.
16405 @item ymax, umax, vmax, amax
16406 The maximum allowed component value.
16409 All expressions default to "val".
16411 @subsection Commands
16413 This filter supports the all above options as @ref{commands}.
16415 @subsection Examples
16419 Change too high luma values to gradient:
16421 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'"
16427 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
16428 Ratio) between two input videos.
16430 This filter takes in input two input videos, the first input is
16431 considered the "main" source and is passed unchanged to the
16432 output. The second input is used as a "reference" video for computing
16435 Both video inputs must have the same resolution and pixel format for
16436 this filter to work correctly. Also it assumes that both inputs
16437 have the same number of frames, which are compared one by one.
16439 The obtained average PSNR is printed through the logging system.
16441 The filter stores the accumulated MSE (mean squared error) of each
16442 frame, and at the end of the processing it is averaged across all frames
16443 equally, and the following formula is applied to obtain the PSNR:
16446 PSNR = 10*log10(MAX^2/MSE)
16449 Where MAX is the average of the maximum values of each component of the
16452 The description of the accepted parameters follows.
16455 @item stats_file, f
16456 If specified the filter will use the named file to save the PSNR of
16457 each individual frame. When filename equals "-" the data is sent to
16460 @item stats_version
16461 Specifies which version of the stats file format to use. Details of
16462 each format are written below.
16463 Default value is 1.
16465 @item stats_add_max
16466 Determines whether the max value is output to the stats log.
16467 Default value is 0.
16468 Requires stats_version >= 2. If this is set and stats_version < 2,
16469 the filter will return an error.
16472 This filter also supports the @ref{framesync} options.
16474 The file printed if @var{stats_file} is selected, contains a sequence of
16475 key/value pairs of the form @var{key}:@var{value} for each compared
16478 If a @var{stats_version} greater than 1 is specified, a header line precedes
16479 the list of per-frame-pair stats, with key value pairs following the frame
16480 format with the following parameters:
16483 @item psnr_log_version
16484 The version of the log file format. Will match @var{stats_version}.
16487 A comma separated list of the per-frame-pair parameters included in
16491 A description of each shown per-frame-pair parameter follows:
16495 sequential number of the input frame, starting from 1
16498 Mean Square Error pixel-by-pixel average difference of the compared
16499 frames, averaged over all the image components.
16501 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
16502 Mean Square Error pixel-by-pixel average difference of the compared
16503 frames for the component specified by the suffix.
16505 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
16506 Peak Signal to Noise ratio of the compared frames for the component
16507 specified by the suffix.
16509 @item max_avg, max_y, max_u, max_v
16510 Maximum allowed value for each channel, and average over all
16514 @subsection Examples
16519 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16520 [main][ref] psnr="stats_file=stats.log" [out]
16523 On this example the input file being processed is compared with the
16524 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
16525 is stored in @file{stats.log}.
16528 Another example with different containers:
16530 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 -
16537 Pulldown reversal (inverse telecine) filter, capable of handling mixed
16538 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
16541 The pullup filter is designed to take advantage of future context in making
16542 its decisions. This filter is stateless in the sense that it does not lock
16543 onto a pattern to follow, but it instead looks forward to the following
16544 fields in order to identify matches and rebuild progressive frames.
16546 To produce content with an even framerate, insert the fps filter after
16547 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
16548 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
16550 The filter accepts the following options:
16557 These options set the amount of "junk" to ignore at the left, right, top, and
16558 bottom of the image, respectively. Left and right are in units of 8 pixels,
16559 while top and bottom are in units of 2 lines.
16560 The default is 8 pixels on each side.
16563 Set the strict breaks. Setting this option to 1 will reduce the chances of
16564 filter generating an occasional mismatched frame, but it may also cause an
16565 excessive number of frames to be dropped during high motion sequences.
16566 Conversely, setting it to -1 will make filter match fields more easily.
16567 This may help processing of video where there is slight blurring between
16568 the fields, but may also cause there to be interlaced frames in the output.
16569 Default value is @code{0}.
16572 Set the metric plane to use. It accepts the following values:
16578 Use chroma blue plane.
16581 Use chroma red plane.
16584 This option may be set to use chroma plane instead of the default luma plane
16585 for doing filter's computations. This may improve accuracy on very clean
16586 source material, but more likely will decrease accuracy, especially if there
16587 is chroma noise (rainbow effect) or any grayscale video.
16588 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
16589 load and make pullup usable in realtime on slow machines.
16592 For best results (without duplicated frames in the output file) it is
16593 necessary to change the output frame rate. For example, to inverse
16594 telecine NTSC input:
16596 ffmpeg -i input -vf pullup -r 24000/1001 ...
16601 Change video quantization parameters (QP).
16603 The filter accepts the following option:
16607 Set expression for quantization parameter.
16610 The expression is evaluated through the eval API and can contain, among others,
16611 the following constants:
16615 1 if index is not 129, 0 otherwise.
16618 Sequential index starting from -129 to 128.
16621 @subsection Examples
16625 Some equation like:
16633 Flush video frames from internal cache of frames into a random order.
16634 No frame is discarded.
16635 Inspired by @ref{frei0r} nervous filter.
16639 Set size in number of frames of internal cache, in range from @code{2} to
16640 @code{512}. Default is @code{30}.
16643 Set seed for random number generator, must be an integer included between
16644 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16645 less than @code{0}, the filter will try to use a good random seed on a
16649 @section readeia608
16651 Read closed captioning (EIA-608) information from the top lines of a video frame.
16653 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
16654 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
16655 with EIA-608 data (starting from 0). A description of each metadata value follows:
16658 @item lavfi.readeia608.X.cc
16659 The two bytes stored as EIA-608 data (printed in hexadecimal).
16661 @item lavfi.readeia608.X.line
16662 The number of the line on which the EIA-608 data was identified and read.
16665 This filter accepts the following options:
16669 Set the line to start scanning for EIA-608 data. Default is @code{0}.
16672 Set the line to end scanning for EIA-608 data. Default is @code{29}.
16675 Set the ratio of width reserved for sync code detection.
16676 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
16679 Enable checking the parity bit. In the event of a parity error, the filter will output
16680 @code{0x00} for that character. Default is false.
16683 Lowpass lines prior to further processing. Default is enabled.
16686 @subsection Commands
16688 This filter supports the all above options as @ref{commands}.
16690 @subsection Examples
16694 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
16696 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
16702 Read vertical interval timecode (VITC) information from the top lines of a
16705 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
16706 timecode value, if a valid timecode has been detected. Further metadata key
16707 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
16708 timecode data has been found or not.
16710 This filter accepts the following options:
16714 Set the maximum number of lines to scan for VITC data. If the value is set to
16715 @code{-1} the full video frame is scanned. Default is @code{45}.
16718 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
16719 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
16722 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
16723 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
16726 @subsection Examples
16730 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
16731 draw @code{--:--:--:--} as a placeholder:
16733 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
16739 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
16741 Destination pixel at position (X, Y) will be picked from source (x, y) position
16742 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
16743 value for pixel will be used for destination pixel.
16745 Xmap and Ymap input video streams must be of same dimensions. Output video stream
16746 will have Xmap/Ymap video stream dimensions.
16747 Xmap and Ymap input video streams are 16bit depth, single channel.
16751 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
16752 Default is @code{color}.
16755 Specify the color of the unmapped pixels. For the syntax of this option,
16756 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16757 manual,ffmpeg-utils}. Default color is @code{black}.
16760 @section removegrain
16762 The removegrain filter is a spatial denoiser for progressive video.
16766 Set mode for the first plane.
16769 Set mode for the second plane.
16772 Set mode for the third plane.
16775 Set mode for the fourth plane.
16778 Range of mode is from 0 to 24. Description of each mode follows:
16782 Leave input plane unchanged. Default.
16785 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
16788 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
16791 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
16794 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
16795 This is equivalent to a median filter.
16798 Line-sensitive clipping giving the minimal change.
16801 Line-sensitive clipping, intermediate.
16804 Line-sensitive clipping, intermediate.
16807 Line-sensitive clipping, intermediate.
16810 Line-sensitive clipping on a line where the neighbours pixels are the closest.
16813 Replaces the target pixel with the closest neighbour.
16816 [1 2 1] horizontal and vertical kernel blur.
16822 Bob mode, interpolates top field from the line where the neighbours
16823 pixels are the closest.
16826 Bob mode, interpolates bottom field from the line where the neighbours
16827 pixels are the closest.
16830 Bob mode, interpolates top field. Same as 13 but with a more complicated
16831 interpolation formula.
16834 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
16835 interpolation formula.
16838 Clips the pixel with the minimum and maximum of respectively the maximum and
16839 minimum of each pair of opposite neighbour pixels.
16842 Line-sensitive clipping using opposite neighbours whose greatest distance from
16843 the current pixel is minimal.
16846 Replaces the pixel with the average of its 8 neighbours.
16849 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
16852 Clips pixels using the averages of opposite neighbour.
16855 Same as mode 21 but simpler and faster.
16858 Small edge and halo removal, but reputed useless.
16864 @section removelogo
16866 Suppress a TV station logo, using an image file to determine which
16867 pixels comprise the logo. It works by filling in the pixels that
16868 comprise the logo with neighboring pixels.
16870 The filter accepts the following options:
16874 Set the filter bitmap file, which can be any image format supported by
16875 libavformat. The width and height of the image file must match those of the
16876 video stream being processed.
16879 Pixels in the provided bitmap image with a value of zero are not
16880 considered part of the logo, non-zero pixels are considered part of
16881 the logo. If you use white (255) for the logo and black (0) for the
16882 rest, you will be safe. For making the filter bitmap, it is
16883 recommended to take a screen capture of a black frame with the logo
16884 visible, and then using a threshold filter followed by the erode
16885 filter once or twice.
16887 If needed, little splotches can be fixed manually. Remember that if
16888 logo pixels are not covered, the filter quality will be much
16889 reduced. Marking too many pixels as part of the logo does not hurt as
16890 much, but it will increase the amount of blurring needed to cover over
16891 the image and will destroy more information than necessary, and extra
16892 pixels will slow things down on a large logo.
16894 @section repeatfields
16896 This filter uses the repeat_field flag from the Video ES headers and hard repeats
16897 fields based on its value.
16901 Reverse a video clip.
16903 Warning: This filter requires memory to buffer the entire clip, so trimming
16906 @subsection Examples
16910 Take the first 5 seconds of a clip, and reverse it.
16917 Shift R/G/B/A pixels horizontally and/or vertically.
16919 The filter accepts the following options:
16922 Set amount to shift red horizontally.
16924 Set amount to shift red vertically.
16926 Set amount to shift green horizontally.
16928 Set amount to shift green vertically.
16930 Set amount to shift blue horizontally.
16932 Set amount to shift blue vertically.
16934 Set amount to shift alpha horizontally.
16936 Set amount to shift alpha vertically.
16938 Set edge mode, can be @var{smear}, default, or @var{warp}.
16941 @subsection Commands
16943 This filter supports the all above options as @ref{commands}.
16946 Apply roberts cross operator to input video stream.
16948 The filter accepts the following option:
16952 Set which planes will be processed, unprocessed planes will be copied.
16953 By default value 0xf, all planes will be processed.
16956 Set value which will be multiplied with filtered result.
16959 Set value which will be added to filtered result.
16962 @subsection Commands
16964 This filter supports the all above options as @ref{commands}.
16968 Rotate video by an arbitrary angle expressed in radians.
16970 The filter accepts the following options:
16972 A description of the optional parameters follows.
16975 Set an expression for the angle by which to rotate the input video
16976 clockwise, expressed as a number of radians. A negative value will
16977 result in a counter-clockwise rotation. By default it is set to "0".
16979 This expression is evaluated for each frame.
16982 Set the output width expression, default value is "iw".
16983 This expression is evaluated just once during configuration.
16986 Set the output height expression, default value is "ih".
16987 This expression is evaluated just once during configuration.
16990 Enable bilinear interpolation if set to 1, a value of 0 disables
16991 it. Default value is 1.
16994 Set the color used to fill the output area not covered by the rotated
16995 image. For the general syntax of this option, check the
16996 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16997 If the special value "none" is selected then no
16998 background is printed (useful for example if the background is never shown).
17000 Default value is "black".
17003 The expressions for the angle and the output size can contain the
17004 following constants and functions:
17008 sequential number of the input frame, starting from 0. It is always NAN
17009 before the first frame is filtered.
17012 time in seconds of the input frame, it is set to 0 when the filter is
17013 configured. It is always NAN before the first frame is filtered.
17017 horizontal and vertical chroma subsample values. For example for the
17018 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17022 the input video width and height
17026 the output width and height, that is the size of the padded area as
17027 specified by the @var{width} and @var{height} expressions
17031 the minimal width/height required for completely containing the input
17032 video rotated by @var{a} radians.
17034 These are only available when computing the @option{out_w} and
17035 @option{out_h} expressions.
17038 @subsection Examples
17042 Rotate the input by PI/6 radians clockwise:
17048 Rotate the input by PI/6 radians counter-clockwise:
17054 Rotate the input by 45 degrees clockwise:
17060 Apply a constant rotation with period T, starting from an angle of PI/3:
17062 rotate=PI/3+2*PI*t/T
17066 Make the input video rotation oscillating with a period of T
17067 seconds and an amplitude of A radians:
17069 rotate=A*sin(2*PI/T*t)
17073 Rotate the video, output size is chosen so that the whole rotating
17074 input video is always completely contained in the output:
17076 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
17080 Rotate the video, reduce the output size so that no background is ever
17083 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
17087 @subsection Commands
17089 The filter supports the following commands:
17093 Set the angle expression.
17094 The command accepts the same syntax of the corresponding option.
17096 If the specified expression is not valid, it is kept at its current
17102 Apply Shape Adaptive Blur.
17104 The filter accepts the following options:
17107 @item luma_radius, lr
17108 Set luma blur filter strength, must be a value in range 0.1-4.0, default
17109 value is 1.0. A greater value will result in a more blurred image, and
17110 in slower processing.
17112 @item luma_pre_filter_radius, lpfr
17113 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
17116 @item luma_strength, ls
17117 Set luma maximum difference between pixels to still be considered, must
17118 be a value in the 0.1-100.0 range, default value is 1.0.
17120 @item chroma_radius, cr
17121 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
17122 greater value will result in a more blurred image, and in slower
17125 @item chroma_pre_filter_radius, cpfr
17126 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
17128 @item chroma_strength, cs
17129 Set chroma maximum difference between pixels to still be considered,
17130 must be a value in the -0.9-100.0 range.
17133 Each chroma option value, if not explicitly specified, is set to the
17134 corresponding luma option value.
17139 Scale (resize) the input video, using the libswscale library.
17141 The scale filter forces the output display aspect ratio to be the same
17142 of the input, by changing the output sample aspect ratio.
17144 If the input image format is different from the format requested by
17145 the next filter, the scale filter will convert the input to the
17148 @subsection Options
17149 The filter accepts the following options, or any of the options
17150 supported by the libswscale scaler.
17152 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
17153 the complete list of scaler options.
17158 Set the output video dimension expression. Default value is the input
17161 If the @var{width} or @var{w} value is 0, the input width is used for
17162 the output. If the @var{height} or @var{h} value is 0, the input height
17163 is used for the output.
17165 If one and only one of the values is -n with n >= 1, the scale filter
17166 will use a value that maintains the aspect ratio of the input image,
17167 calculated from the other specified dimension. After that it will,
17168 however, make sure that the calculated dimension is divisible by n and
17169 adjust the value if necessary.
17171 If both values are -n with n >= 1, the behavior will be identical to
17172 both values being set to 0 as previously detailed.
17174 See below for the list of accepted constants for use in the dimension
17178 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
17182 Only evaluate expressions once during the filter initialization or when a command is processed.
17185 Evaluate expressions for each incoming frame.
17189 Default value is @samp{init}.
17193 Set the interlacing mode. It accepts the following values:
17197 Force interlaced aware scaling.
17200 Do not apply interlaced scaling.
17203 Select interlaced aware scaling depending on whether the source frames
17204 are flagged as interlaced or not.
17207 Default value is @samp{0}.
17210 Set libswscale scaling flags. See
17211 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17212 complete list of values. If not explicitly specified the filter applies
17216 @item param0, param1
17217 Set libswscale input parameters for scaling algorithms that need them. See
17218 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17219 complete documentation. If not explicitly specified the filter applies
17225 Set the video size. For the syntax of this option, check the
17226 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17228 @item in_color_matrix
17229 @item out_color_matrix
17230 Set in/output YCbCr color space type.
17232 This allows the autodetected value to be overridden as well as allows forcing
17233 a specific value used for the output and encoder.
17235 If not specified, the color space type depends on the pixel format.
17241 Choose automatically.
17244 Format conforming to International Telecommunication Union (ITU)
17245 Recommendation BT.709.
17248 Set color space conforming to the United States Federal Communications
17249 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
17254 Set color space conforming to:
17258 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
17261 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
17264 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
17269 Set color space conforming to SMPTE ST 240:1999.
17272 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
17277 Set in/output YCbCr sample range.
17279 This allows the autodetected value to be overridden as well as allows forcing
17280 a specific value used for the output and encoder. If not specified, the
17281 range depends on the pixel format. Possible values:
17285 Choose automatically.
17288 Set full range (0-255 in case of 8-bit luma).
17290 @item mpeg/limited/tv
17291 Set "MPEG" range (16-235 in case of 8-bit luma).
17294 @item force_original_aspect_ratio
17295 Enable decreasing or increasing output video width or height if necessary to
17296 keep the original aspect ratio. Possible values:
17300 Scale the video as specified and disable this feature.
17303 The output video dimensions will automatically be decreased if needed.
17306 The output video dimensions will automatically be increased if needed.
17310 One useful instance of this option is that when you know a specific device's
17311 maximum allowed resolution, you can use this to limit the output video to
17312 that, while retaining the aspect ratio. For example, device A allows
17313 1280x720 playback, and your video is 1920x800. Using this option (set it to
17314 decrease) and specifying 1280x720 to the command line makes the output
17317 Please note that this is a different thing than specifying -1 for @option{w}
17318 or @option{h}, you still need to specify the output resolution for this option
17321 @item force_divisible_by
17322 Ensures that both the output dimensions, width and height, are divisible by the
17323 given integer when used together with @option{force_original_aspect_ratio}. This
17324 works similar to using @code{-n} in the @option{w} and @option{h} options.
17326 This option respects the value set for @option{force_original_aspect_ratio},
17327 increasing or decreasing the resolution accordingly. The video's aspect ratio
17328 may be slightly modified.
17330 This option can be handy if you need to have a video fit within or exceed
17331 a defined resolution using @option{force_original_aspect_ratio} but also have
17332 encoder restrictions on width or height divisibility.
17336 The values of the @option{w} and @option{h} options are expressions
17337 containing the following constants:
17342 The input width and height
17346 These are the same as @var{in_w} and @var{in_h}.
17350 The output (scaled) width and height
17354 These are the same as @var{out_w} and @var{out_h}
17357 The same as @var{iw} / @var{ih}
17360 input sample aspect ratio
17363 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17367 horizontal and vertical input chroma subsample values. For example for the
17368 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17372 horizontal and vertical output chroma subsample values. For example for the
17373 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17376 The (sequential) number of the input frame, starting from 0.
17377 Only available with @code{eval=frame}.
17380 The presentation timestamp of the input frame, expressed as a number of
17381 seconds. Only available with @code{eval=frame}.
17384 The position (byte offset) of the frame in the input stream, or NaN if
17385 this information is unavailable and/or meaningless (for example in case of synthetic video).
17386 Only available with @code{eval=frame}.
17389 @subsection Examples
17393 Scale the input video to a size of 200x100
17398 This is equivalent to:
17409 Specify a size abbreviation for the output size:
17414 which can also be written as:
17420 Scale the input to 2x:
17422 scale=w=2*iw:h=2*ih
17426 The above is the same as:
17428 scale=2*in_w:2*in_h
17432 Scale the input to 2x with forced interlaced scaling:
17434 scale=2*iw:2*ih:interl=1
17438 Scale the input to half size:
17440 scale=w=iw/2:h=ih/2
17444 Increase the width, and set the height to the same size:
17450 Seek Greek harmony:
17457 Increase the height, and set the width to 3/2 of the height:
17459 scale=w=3/2*oh:h=3/5*ih
17463 Increase the size, making the size a multiple of the chroma
17466 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
17470 Increase the width to a maximum of 500 pixels,
17471 keeping the same aspect ratio as the input:
17473 scale=w='min(500\, iw*3/2):h=-1'
17477 Make pixels square by combining scale and setsar:
17479 scale='trunc(ih*dar):ih',setsar=1/1
17483 Make pixels square by combining scale and setsar,
17484 making sure the resulting resolution is even (required by some codecs):
17486 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
17490 @subsection Commands
17492 This filter supports the following commands:
17496 Set the output video dimension expression.
17497 The command accepts the same syntax of the corresponding option.
17499 If the specified expression is not valid, it is kept at its current
17505 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
17506 format conversion on CUDA video frames. Setting the output width and height
17507 works in the same way as for the @var{scale} filter.
17509 The following additional options are accepted:
17512 The pixel format of the output CUDA frames. If set to the string "same" (the
17513 default), the input format will be kept. Note that automatic format negotiation
17514 and conversion is not yet supported for hardware frames
17517 The interpolation algorithm used for resizing. One of the following:
17524 @item cubic2p_bspline
17525 2-parameter cubic (B=1, C=0)
17527 @item cubic2p_catmullrom
17528 2-parameter cubic (B=0, C=1/2)
17530 @item cubic2p_b05c03
17531 2-parameter cubic (B=1/2, C=3/10)
17539 @item force_original_aspect_ratio
17540 Enable decreasing or increasing output video width or height if necessary to
17541 keep the original aspect ratio. Possible values:
17545 Scale the video as specified and disable this feature.
17548 The output video dimensions will automatically be decreased if needed.
17551 The output video dimensions will automatically be increased if needed.
17555 One useful instance of this option is that when you know a specific device's
17556 maximum allowed resolution, you can use this to limit the output video to
17557 that, while retaining the aspect ratio. For example, device A allows
17558 1280x720 playback, and your video is 1920x800. Using this option (set it to
17559 decrease) and specifying 1280x720 to the command line makes the output
17562 Please note that this is a different thing than specifying -1 for @option{w}
17563 or @option{h}, you still need to specify the output resolution for this option
17566 @item force_divisible_by
17567 Ensures that both the output dimensions, width and height, are divisible by the
17568 given integer when used together with @option{force_original_aspect_ratio}. This
17569 works similar to using @code{-n} in the @option{w} and @option{h} options.
17571 This option respects the value set for @option{force_original_aspect_ratio},
17572 increasing or decreasing the resolution accordingly. The video's aspect ratio
17573 may be slightly modified.
17575 This option can be handy if you need to have a video fit within or exceed
17576 a defined resolution using @option{force_original_aspect_ratio} but also have
17577 encoder restrictions on width or height divisibility.
17583 Scale (resize) the input video, based on a reference video.
17585 See the scale filter for available options, scale2ref supports the same but
17586 uses the reference video instead of the main input as basis. scale2ref also
17587 supports the following additional constants for the @option{w} and
17588 @option{h} options:
17593 The main input video's width and height
17596 The same as @var{main_w} / @var{main_h}
17599 The main input video's sample aspect ratio
17601 @item main_dar, mdar
17602 The main input video's display aspect ratio. Calculated from
17603 @code{(main_w / main_h) * main_sar}.
17607 The main input video's horizontal and vertical chroma subsample values.
17608 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
17612 The (sequential) number of the main input frame, starting from 0.
17613 Only available with @code{eval=frame}.
17616 The presentation timestamp of the main input frame, expressed as a number of
17617 seconds. Only available with @code{eval=frame}.
17620 The position (byte offset) of the frame in the main input stream, or NaN if
17621 this information is unavailable and/or meaningless (for example in case of synthetic video).
17622 Only available with @code{eval=frame}.
17625 @subsection Examples
17629 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
17631 'scale2ref[b][a];[a][b]overlay'
17635 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
17637 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
17641 @subsection Commands
17643 This filter supports the following commands:
17647 Set the output video dimension expression.
17648 The command accepts the same syntax of the corresponding option.
17650 If the specified expression is not valid, it is kept at its current
17655 Scroll input video horizontally and/or vertically by constant speed.
17657 The filter accepts the following options:
17659 @item horizontal, h
17660 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
17661 Negative values changes scrolling direction.
17664 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
17665 Negative values changes scrolling direction.
17668 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
17671 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
17674 @subsection Commands
17676 This filter supports the following @ref{commands}:
17678 @item horizontal, h
17679 Set the horizontal scrolling speed.
17681 Set the vertical scrolling speed.
17687 Detect video scene change.
17689 This filter sets frame metadata with mafd between frame, the scene score, and
17690 forward the frame to the next filter, so they can use these metadata to detect
17691 scene change or others.
17693 In addition, this filter logs a message and sets frame metadata when it detects
17694 a scene change by @option{threshold}.
17696 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
17698 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
17699 to detect scene change.
17701 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
17702 detect scene change with @option{threshold}.
17704 The filter accepts the following options:
17708 Set the scene change detection threshold as a percentage of maximum change. Good
17709 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
17712 Default value is @code{10.}.
17715 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
17716 You can enable it if you want to get snapshot of scene change frames only.
17719 @anchor{selectivecolor}
17720 @section selectivecolor
17722 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
17723 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
17724 by the "purity" of the color (that is, how saturated it already is).
17726 This filter is similar to the Adobe Photoshop Selective Color tool.
17728 The filter accepts the following options:
17731 @item correction_method
17732 Select color correction method.
17734 Available values are:
17737 Specified adjustments are applied "as-is" (added/subtracted to original pixel
17740 Specified adjustments are relative to the original component value.
17742 Default is @code{absolute}.
17744 Adjustments for red pixels (pixels where the red component is the maximum)
17746 Adjustments for yellow pixels (pixels where the blue component is the minimum)
17748 Adjustments for green pixels (pixels where the green component is the maximum)
17750 Adjustments for cyan pixels (pixels where the red component is the minimum)
17752 Adjustments for blue pixels (pixels where the blue component is the maximum)
17754 Adjustments for magenta pixels (pixels where the green component is the minimum)
17756 Adjustments for white pixels (pixels where all components are greater than 128)
17758 Adjustments for all pixels except pure black and pure white
17760 Adjustments for black pixels (pixels where all components are lesser than 128)
17762 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
17765 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
17766 4 space separated floating point adjustment values in the [-1,1] range,
17767 respectively to adjust the amount of cyan, magenta, yellow and black for the
17768 pixels of its range.
17770 @subsection Examples
17774 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
17775 increase magenta by 27% in blue areas:
17777 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
17781 Use a Photoshop selective color preset:
17783 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
17787 @anchor{separatefields}
17788 @section separatefields
17790 The @code{separatefields} takes a frame-based video input and splits
17791 each frame into its components fields, producing a new half height clip
17792 with twice the frame rate and twice the frame count.
17794 This filter use field-dominance information in frame to decide which
17795 of each pair of fields to place first in the output.
17796 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
17798 @section setdar, setsar
17800 The @code{setdar} filter sets the Display Aspect Ratio for the filter
17803 This is done by changing the specified Sample (aka Pixel) Aspect
17804 Ratio, according to the following equation:
17806 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
17809 Keep in mind that the @code{setdar} filter does not modify the pixel
17810 dimensions of the video frame. Also, the display aspect ratio set by
17811 this filter may be changed by later filters in the filterchain,
17812 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
17815 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
17816 the filter output video.
17818 Note that as a consequence of the application of this filter, the
17819 output display aspect ratio will change according to the equation
17822 Keep in mind that the sample aspect ratio set by the @code{setsar}
17823 filter may be changed by later filters in the filterchain, e.g. if
17824 another "setsar" or a "setdar" filter is applied.
17826 It accepts the following parameters:
17829 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
17830 Set the aspect ratio used by the filter.
17832 The parameter can be a floating point number string, an expression, or
17833 a string of the form @var{num}:@var{den}, where @var{num} and
17834 @var{den} are the numerator and denominator of the aspect ratio. If
17835 the parameter is not specified, it is assumed the value "0".
17836 In case the form "@var{num}:@var{den}" is used, the @code{:} character
17840 Set the maximum integer value to use for expressing numerator and
17841 denominator when reducing the expressed aspect ratio to a rational.
17842 Default value is @code{100}.
17846 The parameter @var{sar} is an expression containing
17847 the following constants:
17851 These are approximated values for the mathematical constants e
17852 (Euler's number), pi (Greek pi), and phi (the golden ratio).
17855 The input width and height.
17858 These are the same as @var{w} / @var{h}.
17861 The input sample aspect ratio.
17864 The input display aspect ratio. It is the same as
17865 (@var{w} / @var{h}) * @var{sar}.
17868 Horizontal and vertical chroma subsample values. For example, for the
17869 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17872 @subsection Examples
17877 To change the display aspect ratio to 16:9, specify one of the following:
17884 To change the sample aspect ratio to 10:11, specify:
17890 To set a display aspect ratio of 16:9, and specify a maximum integer value of
17891 1000 in the aspect ratio reduction, use the command:
17893 setdar=ratio=16/9:max=1000
17901 Force field for the output video frame.
17903 The @code{setfield} filter marks the interlace type field for the
17904 output frames. It does not change the input frame, but only sets the
17905 corresponding property, which affects how the frame is treated by
17906 following filters (e.g. @code{fieldorder} or @code{yadif}).
17908 The filter accepts the following options:
17913 Available values are:
17917 Keep the same field property.
17920 Mark the frame as bottom-field-first.
17923 Mark the frame as top-field-first.
17926 Mark the frame as progressive.
17933 Force frame parameter for the output video frame.
17935 The @code{setparams} filter marks interlace and color range for the
17936 output frames. It does not change the input frame, but only sets the
17937 corresponding property, which affects how the frame is treated by
17942 Available values are:
17946 Keep the same field property (default).
17949 Mark the frame as bottom-field-first.
17952 Mark the frame as top-field-first.
17955 Mark the frame as progressive.
17959 Available values are:
17963 Keep the same color range property (default).
17965 @item unspecified, unknown
17966 Mark the frame as unspecified color range.
17968 @item limited, tv, mpeg
17969 Mark the frame as limited range.
17971 @item full, pc, jpeg
17972 Mark the frame as full range.
17975 @item color_primaries
17976 Set the color primaries.
17977 Available values are:
17981 Keep the same color primaries property (default).
17998 Set the color transfer.
17999 Available values are:
18003 Keep the same color trc property (default).
18025 Set the colorspace.
18026 Available values are:
18030 Keep the same colorspace property (default).
18043 @item chroma-derived-nc
18044 @item chroma-derived-c
18050 Apply shear transform to input video.
18052 This filter supports the following options:
18056 Shear factor in X-direction. Default value is 0.
18057 Allowed range is from -2 to 2.
18060 Shear factor in Y-direction. Default value is 0.
18061 Allowed range is from -2 to 2.
18064 Set the color used to fill the output area not covered by the transformed
18065 video. For the general syntax of this option, check the
18066 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18067 If the special value "none" is selected then no
18068 background is printed (useful for example if the background is never shown).
18070 Default value is "black".
18073 Set interpolation type. Can be @code{bilinear} or @code{nearest}. Default is @code{bilinear}.
18076 @subsection Commands
18078 This filter supports the all above options as @ref{commands}.
18082 Show a line containing various information for each input video frame.
18083 The input video is not modified.
18085 This filter supports the following options:
18089 Calculate checksums of each plane. By default enabled.
18092 The shown line contains a sequence of key/value pairs of the form
18093 @var{key}:@var{value}.
18095 The following values are shown in the output:
18099 The (sequential) number of the input frame, starting from 0.
18102 The Presentation TimeStamp of the input frame, expressed as a number of
18103 time base units. The time base unit depends on the filter input pad.
18106 The Presentation TimeStamp of the input frame, expressed as a number of
18110 The position of the frame in the input stream, or -1 if this information is
18111 unavailable and/or meaningless (for example in case of synthetic video).
18114 The pixel format name.
18117 The sample aspect ratio of the input frame, expressed in the form
18118 @var{num}/@var{den}.
18121 The size of the input frame. For the syntax of this option, check the
18122 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18125 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
18126 for bottom field first).
18129 This is 1 if the frame is a key frame, 0 otherwise.
18132 The picture type of the input frame ("I" for an I-frame, "P" for a
18133 P-frame, "B" for a B-frame, or "?" for an unknown type).
18134 Also refer to the documentation of the @code{AVPictureType} enum and of
18135 the @code{av_get_picture_type_char} function defined in
18136 @file{libavutil/avutil.h}.
18139 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
18141 @item plane_checksum
18142 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
18143 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
18146 The mean value of pixels in each plane of the input frame, expressed in the form
18147 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
18150 The standard deviation of pixel values in each plane of the input frame, expressed
18151 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
18155 @section showpalette
18157 Displays the 256 colors palette of each frame. This filter is only relevant for
18158 @var{pal8} pixel format frames.
18160 It accepts the following option:
18164 Set the size of the box used to represent one palette color entry. Default is
18165 @code{30} (for a @code{30x30} pixel box).
18168 @section shuffleframes
18170 Reorder and/or duplicate and/or drop video frames.
18172 It accepts the following parameters:
18176 Set the destination indexes of input frames.
18177 This is space or '|' separated list of indexes that maps input frames to output
18178 frames. Number of indexes also sets maximal value that each index may have.
18179 '-1' index have special meaning and that is to drop frame.
18182 The first frame has the index 0. The default is to keep the input unchanged.
18184 @subsection Examples
18188 Swap second and third frame of every three frames of the input:
18190 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
18194 Swap 10th and 1st frame of every ten frames of the input:
18196 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
18200 @section shufflepixels
18202 Reorder pixels in video frames.
18204 This filter accepts the following options:
18208 Set shuffle direction. Can be forward or inverse direction.
18209 Default direction is forward.
18212 Set shuffle mode. Can be horizontal, vertical or block mode.
18216 Set shuffle block_size. In case of horizontal shuffle mode only width
18217 part of size is used, and in case of vertical shuffle mode only height
18218 part of size is used.
18221 Set random seed used with shuffling pixels. Mainly useful to set to be able
18222 to reverse filtering process to get original input.
18223 For example, to reverse forward shuffle you need to use same parameters
18224 and exact same seed and to set direction to inverse.
18227 @section shuffleplanes
18229 Reorder and/or duplicate video planes.
18231 It accepts the following parameters:
18236 The index of the input plane to be used as the first output plane.
18239 The index of the input plane to be used as the second output plane.
18242 The index of the input plane to be used as the third output plane.
18245 The index of the input plane to be used as the fourth output plane.
18249 The first plane has the index 0. The default is to keep the input unchanged.
18251 @subsection Examples
18255 Swap the second and third planes of the input:
18257 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
18261 @anchor{signalstats}
18262 @section signalstats
18263 Evaluate various visual metrics that assist in determining issues associated
18264 with the digitization of analog video media.
18266 By default the filter will log these metadata values:
18270 Display the minimal Y value contained within the input frame. Expressed in
18274 Display the Y value at the 10% percentile within the input frame. Expressed in
18278 Display the average Y value within the input frame. Expressed in range of
18282 Display the Y value at the 90% percentile within the input frame. Expressed in
18286 Display the maximum Y value contained within the input frame. Expressed in
18290 Display the minimal U value contained within the input frame. Expressed in
18294 Display the U value at the 10% percentile within the input frame. Expressed in
18298 Display the average U value within the input frame. Expressed in range of
18302 Display the U value at the 90% percentile within the input frame. Expressed in
18306 Display the maximum U value contained within the input frame. Expressed in
18310 Display the minimal V value contained within the input frame. Expressed in
18314 Display the V value at the 10% percentile within the input frame. Expressed in
18318 Display the average V value within the input frame. Expressed in range of
18322 Display the V value at the 90% percentile within the input frame. Expressed in
18326 Display the maximum V value contained within the input frame. Expressed in
18330 Display the minimal saturation value contained within the input frame.
18331 Expressed in range of [0-~181.02].
18334 Display the saturation value at the 10% percentile within the input frame.
18335 Expressed in range of [0-~181.02].
18338 Display the average saturation value within the input frame. Expressed in range
18342 Display the saturation value at the 90% percentile within the input frame.
18343 Expressed in range of [0-~181.02].
18346 Display the maximum saturation value contained within the input frame.
18347 Expressed in range of [0-~181.02].
18350 Display the median value for hue within the input frame. Expressed in range of
18354 Display the average value for hue within the input frame. Expressed in range of
18358 Display the average of sample value difference between all values of the Y
18359 plane in the current frame and corresponding values of the previous input frame.
18360 Expressed in range of [0-255].
18363 Display the average of sample value difference between all values of the U
18364 plane in the current frame and corresponding values of the previous input frame.
18365 Expressed in range of [0-255].
18368 Display the average of sample value difference between all values of the V
18369 plane in the current frame and corresponding values of the previous input frame.
18370 Expressed in range of [0-255].
18373 Display bit depth of Y plane in current frame.
18374 Expressed in range of [0-16].
18377 Display bit depth of U plane in current frame.
18378 Expressed in range of [0-16].
18381 Display bit depth of V plane in current frame.
18382 Expressed in range of [0-16].
18385 The filter accepts the following options:
18391 @option{stat} specify an additional form of image analysis.
18392 @option{out} output video with the specified type of pixel highlighted.
18394 Both options accept the following values:
18398 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
18399 unlike the neighboring pixels of the same field. Examples of temporal outliers
18400 include the results of video dropouts, head clogs, or tape tracking issues.
18403 Identify @var{vertical line repetition}. Vertical line repetition includes
18404 similar rows of pixels within a frame. In born-digital video vertical line
18405 repetition is common, but this pattern is uncommon in video digitized from an
18406 analog source. When it occurs in video that results from the digitization of an
18407 analog source it can indicate concealment from a dropout compensator.
18410 Identify pixels that fall outside of legal broadcast range.
18414 Set the highlight color for the @option{out} option. The default color is
18418 @subsection Examples
18422 Output data of various video metrics:
18424 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
18428 Output specific data about the minimum and maximum values of the Y plane per frame:
18430 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
18434 Playback video while highlighting pixels that are outside of broadcast range in red.
18436 ffplay example.mov -vf signalstats="out=brng:color=red"
18440 Playback video with signalstats metadata drawn over the frame.
18442 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
18445 The contents of signalstat_drawtext.txt used in the command are:
18448 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
18449 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
18450 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
18451 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
18459 Calculates the MPEG-7 Video Signature. The filter can handle more than one
18460 input. In this case the matching between the inputs can be calculated additionally.
18461 The filter always passes through the first input. The signature of each stream can
18462 be written into a file.
18464 It accepts the following options:
18468 Enable or disable the matching process.
18470 Available values are:
18474 Disable the calculation of a matching (default).
18476 Calculate the matching for the whole video and output whether the whole video
18477 matches or only parts.
18479 Calculate only until a matching is found or the video ends. Should be faster in
18484 Set the number of inputs. The option value must be a non negative integer.
18485 Default value is 1.
18488 Set the path to which the output is written. If there is more than one input,
18489 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
18490 integer), that will be replaced with the input number. If no filename is
18491 specified, no output will be written. This is the default.
18494 Choose the output format.
18496 Available values are:
18500 Use the specified binary representation (default).
18502 Use the specified xml representation.
18506 Set threshold to detect one word as similar. The option value must be an integer
18507 greater than zero. The default value is 9000.
18510 Set threshold to detect all words as similar. The option value must be an integer
18511 greater than zero. The default value is 60000.
18514 Set threshold to detect frames as similar. The option value must be an integer
18515 greater than zero. The default value is 116.
18518 Set the minimum length of a sequence in frames to recognize it as matching
18519 sequence. The option value must be a non negative integer value.
18520 The default value is 0.
18523 Set the minimum relation, that matching frames to all frames must have.
18524 The option value must be a double value between 0 and 1. The default value is 0.5.
18527 @subsection Examples
18531 To calculate the signature of an input video and store it in signature.bin:
18533 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
18537 To detect whether two videos match and store the signatures in XML format in
18538 signature0.xml and signature1.xml:
18540 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 -
18548 Blur the input video without impacting the outlines.
18550 It accepts the following options:
18553 @item luma_radius, lr
18554 Set the luma radius. The option value must be a float number in
18555 the range [0.1,5.0] that specifies the variance of the gaussian filter
18556 used to blur the image (slower if larger). Default value is 1.0.
18558 @item luma_strength, ls
18559 Set the luma strength. The option value must be a float number
18560 in the range [-1.0,1.0] that configures the blurring. A value included
18561 in [0.0,1.0] will blur the image whereas a value included in
18562 [-1.0,0.0] will sharpen the image. Default value is 1.0.
18564 @item luma_threshold, lt
18565 Set the luma threshold used as a coefficient to determine
18566 whether a pixel should be blurred or not. The option value must be an
18567 integer in the range [-30,30]. A value of 0 will filter all the image,
18568 a value included in [0,30] will filter flat areas and a value included
18569 in [-30,0] will filter edges. Default value is 0.
18571 @item chroma_radius, cr
18572 Set the chroma radius. The option value must be a float number in
18573 the range [0.1,5.0] that specifies the variance of the gaussian filter
18574 used to blur the image (slower if larger). Default value is @option{luma_radius}.
18576 @item chroma_strength, cs
18577 Set the chroma strength. The option value must be a float number
18578 in the range [-1.0,1.0] that configures the blurring. A value included
18579 in [0.0,1.0] will blur the image whereas a value included in
18580 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
18582 @item chroma_threshold, ct
18583 Set the chroma threshold used as a coefficient to determine
18584 whether a pixel should be blurred or not. The option value must be an
18585 integer in the range [-30,30]. A value of 0 will filter all the image,
18586 a value included in [0,30] will filter flat areas and a value included
18587 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
18590 If a chroma option is not explicitly set, the corresponding luma value
18594 Apply sobel operator to input video stream.
18596 The filter accepts the following option:
18600 Set which planes will be processed, unprocessed planes will be copied.
18601 By default value 0xf, all planes will be processed.
18604 Set value which will be multiplied with filtered result.
18607 Set value which will be added to filtered result.
18610 @subsection Commands
18612 This filter supports the all above options as @ref{commands}.
18617 Apply a simple postprocessing filter that compresses and decompresses the image
18618 at several (or - in the case of @option{quality} level @code{6} - all) shifts
18619 and average the results.
18621 The filter accepts the following options:
18625 Set quality. This option defines the number of levels for averaging. It accepts
18626 an integer in the range 0-6. If set to @code{0}, the filter will have no
18627 effect. A value of @code{6} means the higher quality. For each increment of
18628 that value the speed drops by a factor of approximately 2. Default value is
18632 Force a constant quantization parameter. If not set, the filter will use the QP
18633 from the video stream (if available).
18636 Set thresholding mode. Available modes are:
18640 Set hard thresholding (default).
18642 Set soft thresholding (better de-ringing effect, but likely blurrier).
18645 @item use_bframe_qp
18646 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
18647 option may cause flicker since the B-Frames have often larger QP. Default is
18648 @code{0} (not enabled).
18651 @subsection Commands
18653 This filter supports the following commands:
18655 @item quality, level
18656 Set quality level. The value @code{max} can be used to set the maximum level,
18657 currently @code{6}.
18663 Scale the input by applying one of the super-resolution methods based on
18664 convolutional neural networks. Supported models:
18668 Super-Resolution Convolutional Neural Network model (SRCNN).
18669 See @url{https://arxiv.org/abs/1501.00092}.
18672 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
18673 See @url{https://arxiv.org/abs/1609.05158}.
18676 Training scripts as well as scripts for model file (.pb) saving can be found at
18677 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
18678 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
18680 Native model files (.model) can be generated from TensorFlow model
18681 files (.pb) by using tools/python/convert.py
18683 The filter accepts the following options:
18687 Specify which DNN backend to use for model loading and execution. This option accepts
18688 the following values:
18692 Native implementation of DNN loading and execution.
18695 TensorFlow backend. To enable this backend you
18696 need to install the TensorFlow for C library (see
18697 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
18698 @code{--enable-libtensorflow}
18701 Default value is @samp{native}.
18704 Set path to model file specifying network architecture and its parameters.
18705 Note that different backends use different file formats. TensorFlow backend
18706 can load files for both formats, while native backend can load files for only
18710 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
18711 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
18712 input upscaled using bicubic upscaling with proper scale factor.
18715 This feature can also be finished with @ref{dnn_processing} filter.
18719 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
18721 This filter takes in input two input videos, the first input is
18722 considered the "main" source and is passed unchanged to the
18723 output. The second input is used as a "reference" video for computing
18726 Both video inputs must have the same resolution and pixel format for
18727 this filter to work correctly. Also it assumes that both inputs
18728 have the same number of frames, which are compared one by one.
18730 The filter stores the calculated SSIM of each frame.
18732 The description of the accepted parameters follows.
18735 @item stats_file, f
18736 If specified the filter will use the named file to save the SSIM of
18737 each individual frame. When filename equals "-" the data is sent to
18741 The file printed if @var{stats_file} is selected, contains a sequence of
18742 key/value pairs of the form @var{key}:@var{value} for each compared
18745 A description of each shown parameter follows:
18749 sequential number of the input frame, starting from 1
18751 @item Y, U, V, R, G, B
18752 SSIM of the compared frames for the component specified by the suffix.
18755 SSIM of the compared frames for the whole frame.
18758 Same as above but in dB representation.
18761 This filter also supports the @ref{framesync} options.
18763 @subsection Examples
18768 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
18769 [main][ref] ssim="stats_file=stats.log" [out]
18772 On this example the input file being processed is compared with the
18773 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
18774 is stored in @file{stats.log}.
18777 Another example with both psnr and ssim at same time:
18779 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
18783 Another example with different containers:
18785 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 -
18791 Convert between different stereoscopic image formats.
18793 The filters accept the following options:
18797 Set stereoscopic image format of input.
18799 Available values for input image formats are:
18802 side by side parallel (left eye left, right eye right)
18805 side by side crosseye (right eye left, left eye right)
18808 side by side parallel with half width resolution
18809 (left eye left, right eye right)
18812 side by side crosseye with half width resolution
18813 (right eye left, left eye right)
18817 above-below (left eye above, right eye below)
18821 above-below (right eye above, left eye below)
18825 above-below with half height resolution
18826 (left eye above, right eye below)
18830 above-below with half height resolution
18831 (right eye above, left eye below)
18834 alternating frames (left eye first, right eye second)
18837 alternating frames (right eye first, left eye second)
18840 interleaved rows (left eye has top row, right eye starts on next row)
18843 interleaved rows (right eye has top row, left eye starts on next row)
18846 interleaved columns, left eye first
18849 interleaved columns, right eye first
18851 Default value is @samp{sbsl}.
18855 Set stereoscopic image format of output.
18859 side by side parallel (left eye left, right eye right)
18862 side by side crosseye (right eye left, left eye right)
18865 side by side parallel with half width resolution
18866 (left eye left, right eye right)
18869 side by side crosseye with half width resolution
18870 (right eye left, left eye right)
18874 above-below (left eye above, right eye below)
18878 above-below (right eye above, left eye below)
18882 above-below with half height resolution
18883 (left eye above, right eye below)
18887 above-below with half height resolution
18888 (right eye above, left eye below)
18891 alternating frames (left eye first, right eye second)
18894 alternating frames (right eye first, left eye second)
18897 interleaved rows (left eye has top row, right eye starts on next row)
18900 interleaved rows (right eye has top row, left eye starts on next row)
18903 anaglyph red/blue gray
18904 (red filter on left eye, blue filter on right eye)
18907 anaglyph red/green gray
18908 (red filter on left eye, green filter on right eye)
18911 anaglyph red/cyan gray
18912 (red filter on left eye, cyan filter on right eye)
18915 anaglyph red/cyan half colored
18916 (red filter on left eye, cyan filter on right eye)
18919 anaglyph red/cyan color
18920 (red filter on left eye, cyan filter on right eye)
18923 anaglyph red/cyan color optimized with the least squares projection of dubois
18924 (red filter on left eye, cyan filter on right eye)
18927 anaglyph green/magenta gray
18928 (green filter on left eye, magenta filter on right eye)
18931 anaglyph green/magenta half colored
18932 (green filter on left eye, magenta filter on right eye)
18935 anaglyph green/magenta colored
18936 (green filter on left eye, magenta filter on right eye)
18939 anaglyph green/magenta color optimized with the least squares projection of dubois
18940 (green filter on left eye, magenta filter on right eye)
18943 anaglyph yellow/blue gray
18944 (yellow filter on left eye, blue filter on right eye)
18947 anaglyph yellow/blue half colored
18948 (yellow filter on left eye, blue filter on right eye)
18951 anaglyph yellow/blue colored
18952 (yellow filter on left eye, blue filter on right eye)
18955 anaglyph yellow/blue color optimized with the least squares projection of dubois
18956 (yellow filter on left eye, blue filter on right eye)
18959 mono output (left eye only)
18962 mono output (right eye only)
18965 checkerboard, left eye first
18968 checkerboard, right eye first
18971 interleaved columns, left eye first
18974 interleaved columns, right eye first
18980 Default value is @samp{arcd}.
18983 @subsection Examples
18987 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
18993 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
18999 @section streamselect, astreamselect
19000 Select video or audio streams.
19002 The filter accepts the following options:
19006 Set number of inputs. Default is 2.
19009 Set input indexes to remap to outputs.
19012 @subsection Commands
19014 The @code{streamselect} and @code{astreamselect} filter supports the following
19019 Set input indexes to remap to outputs.
19022 @subsection Examples
19026 Select first 5 seconds 1st stream and rest of time 2nd stream:
19028 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
19032 Same as above, but for audio:
19034 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
19041 Draw subtitles on top of input video using the libass library.
19043 To enable compilation of this filter you need to configure FFmpeg with
19044 @code{--enable-libass}. This filter also requires a build with libavcodec and
19045 libavformat to convert the passed subtitles file to ASS (Advanced Substation
19046 Alpha) subtitles format.
19048 The filter accepts the following options:
19052 Set the filename of the subtitle file to read. It must be specified.
19054 @item original_size
19055 Specify the size of the original video, the video for which the ASS file
19056 was composed. For the syntax of this option, check the
19057 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19058 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
19059 correctly scale the fonts if the aspect ratio has been changed.
19062 Set a directory path containing fonts that can be used by the filter.
19063 These fonts will be used in addition to whatever the font provider uses.
19066 Process alpha channel, by default alpha channel is untouched.
19069 Set subtitles input character encoding. @code{subtitles} filter only. Only
19070 useful if not UTF-8.
19072 @item stream_index, si
19073 Set subtitles stream index. @code{subtitles} filter only.
19076 Override default style or script info parameters of the subtitles. It accepts a
19077 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
19080 If the first key is not specified, it is assumed that the first value
19081 specifies the @option{filename}.
19083 For example, to render the file @file{sub.srt} on top of the input
19084 video, use the command:
19089 which is equivalent to:
19091 subtitles=filename=sub.srt
19094 To render the default subtitles stream from file @file{video.mkv}, use:
19096 subtitles=video.mkv
19099 To render the second subtitles stream from that file, use:
19101 subtitles=video.mkv:si=1
19104 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
19105 @code{DejaVu Serif}, use:
19107 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
19110 @section super2xsai
19112 Scale the input by 2x and smooth using the Super2xSaI (Scale and
19113 Interpolate) pixel art scaling algorithm.
19115 Useful for enlarging pixel art images without reducing sharpness.
19119 Swap two rectangular objects in video.
19121 This filter accepts the following options:
19131 Set 1st rect x coordinate.
19134 Set 1st rect y coordinate.
19137 Set 2nd rect x coordinate.
19140 Set 2nd rect y coordinate.
19142 All expressions are evaluated once for each frame.
19145 The all options are expressions containing the following constants:
19150 The input width and height.
19153 same as @var{w} / @var{h}
19156 input sample aspect ratio
19159 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
19162 The number of the input frame, starting from 0.
19165 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
19168 the position in the file of the input frame, NAN if unknown
19171 @subsection Commands
19173 This filter supports the all above options as @ref{commands}.
19179 Blend successive video frames.
19185 Apply telecine process to the video.
19187 This filter accepts the following options:
19196 The default value is @code{top}.
19200 A string of numbers representing the pulldown pattern you wish to apply.
19201 The default value is @code{23}.
19205 Some typical patterns:
19210 24p: 2332 (preferred)
19217 24p: 222222222223 ("Euro pulldown")
19222 @section thistogram
19224 Compute and draw a color distribution histogram for the input video across time.
19226 Unlike @ref{histogram} video filter which only shows histogram of single input frame
19227 at certain time, this filter shows also past histograms of number of frames defined
19228 by @code{width} option.
19230 The computed histogram is a representation of the color component
19231 distribution in an image.
19233 The filter accepts the following options:
19237 Set width of single color component output. Default value is @code{0}.
19238 Value of @code{0} means width will be picked from input video.
19239 This also set number of passed histograms to keep.
19240 Allowed range is [0, 8192].
19242 @item display_mode, d
19244 It accepts the following values:
19247 Per color component graphs are placed below each other.
19250 Per color component graphs are placed side by side.
19253 Presents information identical to that in the @code{parade}, except
19254 that the graphs representing color components are superimposed directly
19257 Default is @code{stack}.
19259 @item levels_mode, m
19260 Set mode. Can be either @code{linear}, or @code{logarithmic}.
19261 Default is @code{linear}.
19263 @item components, c
19264 Set what color components to display.
19265 Default is @code{7}.
19268 Set background opacity. Default is @code{0.9}.
19271 Show envelope. Default is disabled.
19274 Set envelope color. Default is @code{gold}.
19279 Available values for slide is:
19282 Draw new frame when right border is reached.
19285 Replace old columns with new ones.
19288 Scroll from right to left.
19291 Scroll from left to right.
19294 Draw single picture.
19297 Default is @code{replace}.
19302 Apply threshold effect to video stream.
19304 This filter needs four video streams to perform thresholding.
19305 First stream is stream we are filtering.
19306 Second stream is holding threshold values, third stream is holding min values,
19307 and last, fourth stream is holding max values.
19309 The filter accepts the following option:
19313 Set which planes will be processed, unprocessed planes will be copied.
19314 By default value 0xf, all planes will be processed.
19317 For example if first stream pixel's component value is less then threshold value
19318 of pixel component from 2nd threshold stream, third stream value will picked,
19319 otherwise fourth stream pixel component value will be picked.
19321 Using color source filter one can perform various types of thresholding:
19323 @subsection Examples
19327 Binary threshold, using gray color as threshold:
19329 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
19333 Inverted binary threshold, using gray color as threshold:
19335 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
19339 Truncate binary threshold, using gray color as threshold:
19341 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
19345 Threshold to zero, using gray color as threshold:
19347 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
19351 Inverted threshold to zero, using gray color as threshold:
19353 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
19358 Select the most representative frame in a given sequence of consecutive frames.
19360 The filter accepts the following options:
19364 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
19365 will pick one of them, and then handle the next batch of @var{n} frames until
19366 the end. Default is @code{100}.
19369 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
19370 value will result in a higher memory usage, so a high value is not recommended.
19372 @subsection Examples
19376 Extract one picture each 50 frames:
19382 Complete example of a thumbnail creation with @command{ffmpeg}:
19384 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
19391 Tile several successive frames together.
19393 The @ref{untile} filter can do the reverse.
19395 The filter accepts the following options:
19400 Set the grid size (i.e. the number of lines and columns). For the syntax of
19401 this option, check the
19402 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19405 Set the maximum number of frames to render in the given area. It must be less
19406 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
19407 the area will be used.
19410 Set the outer border margin in pixels.
19413 Set the inner border thickness (i.e. the number of pixels between frames). For
19414 more advanced padding options (such as having different values for the edges),
19415 refer to the pad video filter.
19418 Specify the color of the unused area. For the syntax of this option, check the
19419 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19420 The default value of @var{color} is "black".
19423 Set the number of frames to overlap when tiling several successive frames together.
19424 The value must be between @code{0} and @var{nb_frames - 1}.
19427 Set the number of frames to initially be empty before displaying first output frame.
19428 This controls how soon will one get first output frame.
19429 The value must be between @code{0} and @var{nb_frames - 1}.
19432 @subsection Examples
19436 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
19438 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
19440 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
19441 duplicating each output frame to accommodate the originally detected frame
19445 Display @code{5} pictures in an area of @code{3x2} frames,
19446 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
19447 mixed flat and named options:
19449 tile=3x2:nb_frames=5:padding=7:margin=2
19453 @section tinterlace
19455 Perform various types of temporal field interlacing.
19457 Frames are counted starting from 1, so the first input frame is
19460 The filter accepts the following options:
19465 Specify the mode of the interlacing. This option can also be specified
19466 as a value alone. See below for a list of values for this option.
19468 Available values are:
19472 Move odd frames into the upper field, even into the lower field,
19473 generating a double height frame at half frame rate.
19477 Frame 1 Frame 2 Frame 3 Frame 4
19479 11111 22222 33333 44444
19480 11111 22222 33333 44444
19481 11111 22222 33333 44444
19482 11111 22222 33333 44444
19496 Only output odd frames, even frames are dropped, generating a frame with
19497 unchanged height at half frame rate.
19502 Frame 1 Frame 2 Frame 3 Frame 4
19504 11111 22222 33333 44444
19505 11111 22222 33333 44444
19506 11111 22222 33333 44444
19507 11111 22222 33333 44444
19517 Only output even frames, odd frames are dropped, generating a frame with
19518 unchanged height at half frame rate.
19523 Frame 1 Frame 2 Frame 3 Frame 4
19525 11111 22222 33333 44444
19526 11111 22222 33333 44444
19527 11111 22222 33333 44444
19528 11111 22222 33333 44444
19538 Expand each frame to full height, but pad alternate lines with black,
19539 generating a frame with double height at the same input frame rate.
19544 Frame 1 Frame 2 Frame 3 Frame 4
19546 11111 22222 33333 44444
19547 11111 22222 33333 44444
19548 11111 22222 33333 44444
19549 11111 22222 33333 44444
19552 11111 ..... 33333 .....
19553 ..... 22222 ..... 44444
19554 11111 ..... 33333 .....
19555 ..... 22222 ..... 44444
19556 11111 ..... 33333 .....
19557 ..... 22222 ..... 44444
19558 11111 ..... 33333 .....
19559 ..... 22222 ..... 44444
19563 @item interleave_top, 4
19564 Interleave the upper field from odd frames with the lower field from
19565 even frames, generating a frame with unchanged height at half frame rate.
19570 Frame 1 Frame 2 Frame 3 Frame 4
19572 11111<- 22222 33333<- 44444
19573 11111 22222<- 33333 44444<-
19574 11111<- 22222 33333<- 44444
19575 11111 22222<- 33333 44444<-
19585 @item interleave_bottom, 5
19586 Interleave the lower field from odd frames with the upper field from
19587 even frames, generating a frame with unchanged height at half frame rate.
19592 Frame 1 Frame 2 Frame 3 Frame 4
19594 11111 22222<- 33333 44444<-
19595 11111<- 22222 33333<- 44444
19596 11111 22222<- 33333 44444<-
19597 11111<- 22222 33333<- 44444
19607 @item interlacex2, 6
19608 Double frame rate with unchanged height. Frames are inserted each
19609 containing the second temporal field from the previous input frame and
19610 the first temporal field from the next input frame. This mode relies on
19611 the top_field_first flag. Useful for interlaced video displays with no
19612 field synchronisation.
19617 Frame 1 Frame 2 Frame 3 Frame 4
19619 11111 22222 33333 44444
19620 11111 22222 33333 44444
19621 11111 22222 33333 44444
19622 11111 22222 33333 44444
19625 11111 22222 22222 33333 33333 44444 44444
19626 11111 11111 22222 22222 33333 33333 44444
19627 11111 22222 22222 33333 33333 44444 44444
19628 11111 11111 22222 22222 33333 33333 44444
19633 Move odd frames into the upper field, even into the lower field,
19634 generating a double height frame at same frame rate.
19639 Frame 1 Frame 2 Frame 3 Frame 4
19641 11111 22222 33333 44444
19642 11111 22222 33333 44444
19643 11111 22222 33333 44444
19644 11111 22222 33333 44444
19647 11111 33333 33333 55555
19648 22222 22222 44444 44444
19649 11111 33333 33333 55555
19650 22222 22222 44444 44444
19651 11111 33333 33333 55555
19652 22222 22222 44444 44444
19653 11111 33333 33333 55555
19654 22222 22222 44444 44444
19659 Numeric values are deprecated but are accepted for backward
19660 compatibility reasons.
19662 Default mode is @code{merge}.
19665 Specify flags influencing the filter process.
19667 Available value for @var{flags} is:
19670 @item low_pass_filter, vlpf
19671 Enable linear vertical low-pass filtering in the filter.
19672 Vertical low-pass filtering is required when creating an interlaced
19673 destination from a progressive source which contains high-frequency
19674 vertical detail. Filtering will reduce interlace 'twitter' and Moire
19677 @item complex_filter, cvlpf
19678 Enable complex vertical low-pass filtering.
19679 This will slightly less reduce interlace 'twitter' and Moire
19680 patterning but better retain detail and subjective sharpness impression.
19683 Bypass already interlaced frames, only adjust the frame rate.
19686 Vertical low-pass filtering and bypassing already interlaced frames can only be
19687 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
19692 Pick median pixels from several successive input video frames.
19694 The filter accepts the following options:
19698 Set radius of median filter.
19699 Default is 1. Allowed range is from 1 to 127.
19702 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
19705 Set median percentile. Default value is @code{0.5}.
19706 Default value of @code{0.5} will pick always median values, while @code{0} will pick
19707 minimum values, and @code{1} maximum values.
19710 @subsection Commands
19712 This filter supports all above options as @ref{commands}, excluding option @code{radius}.
19714 @section tmidequalizer
19716 Apply Temporal Midway Video Equalization effect.
19718 Midway Video Equalization adjusts a sequence of video frames to have the same
19719 histograms, while maintaining their dynamics as much as possible. It's
19720 useful for e.g. matching exposures from a video frames sequence.
19722 This filter accepts the following option:
19726 Set filtering radius. Default is @code{5}. Allowed range is from 1 to 127.
19729 Set filtering sigma. Default is @code{0.5}. This controls strength of filtering.
19730 Setting this option to 0 effectively does nothing.
19733 Set which planes to process. Default is @code{15}, which is all available planes.
19738 Mix successive video frames.
19740 A description of the accepted options follows.
19744 The number of successive frames to mix. If unspecified, it defaults to 3.
19747 Specify weight of each input video frame.
19748 Each weight is separated by space. If number of weights is smaller than
19749 number of @var{frames} last specified weight will be used for all remaining
19753 Specify scale, if it is set it will be multiplied with sum
19754 of each weight multiplied with pixel values to give final destination
19755 pixel value. By default @var{scale} is auto scaled to sum of weights.
19758 @subsection Examples
19762 Average 7 successive frames:
19764 tmix=frames=7:weights="1 1 1 1 1 1 1"
19768 Apply simple temporal convolution:
19770 tmix=frames=3:weights="-1 3 -1"
19774 Similar as above but only showing temporal differences:
19776 tmix=frames=3:weights="-1 2 -1":scale=1
19780 @subsection Commands
19782 This filter supports the following commands:
19786 Syntax is same as option with same name.
19791 Tone map colors from different dynamic ranges.
19793 This filter expects data in single precision floating point, as it needs to
19794 operate on (and can output) out-of-range values. Another filter, such as
19795 @ref{zscale}, is needed to convert the resulting frame to a usable format.
19797 The tonemapping algorithms implemented only work on linear light, so input
19798 data should be linearized beforehand (and possibly correctly tagged).
19801 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
19804 @subsection Options
19805 The filter accepts the following options.
19809 Set the tone map algorithm to use.
19811 Possible values are:
19814 Do not apply any tone map, only desaturate overbright pixels.
19817 Hard-clip any out-of-range values. Use it for perfect color accuracy for
19818 in-range values, while distorting out-of-range values.
19821 Stretch the entire reference gamut to a linear multiple of the display.
19824 Fit a logarithmic transfer between the tone curves.
19827 Preserve overall image brightness with a simple curve, using nonlinear
19828 contrast, which results in flattening details and degrading color accuracy.
19831 Preserve both dark and bright details better than @var{reinhard}, at the cost
19832 of slightly darkening everything. Use it when detail preservation is more
19833 important than color and brightness accuracy.
19836 Smoothly map out-of-range values, while retaining contrast and colors for
19837 in-range material as much as possible. Use it when color accuracy is more
19838 important than detail preservation.
19844 Tune the tone mapping algorithm.
19846 This affects the following algorithms:
19852 Specifies the scale factor to use while stretching.
19856 Specifies the exponent of the function.
19860 Specify an extra linear coefficient to multiply into the signal before clipping.
19864 Specify the local contrast coefficient at the display peak.
19865 Default to 0.5, which means that in-gamut values will be about half as bright
19872 Specify the transition point from linear to mobius transform. Every value
19873 below this point is guaranteed to be mapped 1:1. The higher the value, the
19874 more accurate the result will be, at the cost of losing bright details.
19875 Default to 0.3, which due to the steep initial slope still preserves in-range
19876 colors fairly accurately.
19880 Apply desaturation for highlights that exceed this level of brightness. The
19881 higher the parameter, the more color information will be preserved. This
19882 setting helps prevent unnaturally blown-out colors for super-highlights, by
19883 (smoothly) turning into white instead. This makes images feel more natural,
19884 at the cost of reducing information about out-of-range colors.
19886 The default of 2.0 is somewhat conservative and will mostly just apply to
19887 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
19889 This option works only if the input frame has a supported color tag.
19892 Override signal/nominal/reference peak with this value. Useful when the
19893 embedded peak information in display metadata is not reliable or when tone
19894 mapping from a lower range to a higher range.
19899 Temporarily pad video frames.
19901 The filter accepts the following options:
19905 Specify number of delay frames before input video stream. Default is 0.
19908 Specify number of padding frames after input video stream.
19909 Set to -1 to pad indefinitely. Default is 0.
19912 Set kind of frames added to beginning of stream.
19913 Can be either @var{add} or @var{clone}.
19914 With @var{add} frames of solid-color are added.
19915 With @var{clone} frames are clones of first frame.
19916 Default is @var{add}.
19919 Set kind of frames added to end of stream.
19920 Can be either @var{add} or @var{clone}.
19921 With @var{add} frames of solid-color are added.
19922 With @var{clone} frames are clones of last frame.
19923 Default is @var{add}.
19925 @item start_duration, stop_duration
19926 Specify the duration of the start/stop delay. See
19927 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19928 for the accepted syntax.
19929 These options override @var{start} and @var{stop}. Default is 0.
19932 Specify the color of the padded area. For the syntax of this option,
19933 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
19934 manual,ffmpeg-utils}.
19936 The default value of @var{color} is "black".
19942 Transpose rows with columns in the input video and optionally flip it.
19944 It accepts the following parameters:
19949 Specify the transposition direction.
19951 Can assume the following values:
19953 @item 0, 4, cclock_flip
19954 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
19962 Rotate by 90 degrees clockwise, that is:
19970 Rotate by 90 degrees counterclockwise, that is:
19977 @item 3, 7, clock_flip
19978 Rotate by 90 degrees clockwise and vertically flip, that is:
19986 For values between 4-7, the transposition is only done if the input
19987 video geometry is portrait and not landscape. These values are
19988 deprecated, the @code{passthrough} option should be used instead.
19990 Numerical values are deprecated, and should be dropped in favor of
19991 symbolic constants.
19994 Do not apply the transposition if the input geometry matches the one
19995 specified by the specified value. It accepts the following values:
19998 Always apply transposition.
20000 Preserve portrait geometry (when @var{height} >= @var{width}).
20002 Preserve landscape geometry (when @var{width} >= @var{height}).
20005 Default value is @code{none}.
20008 For example to rotate by 90 degrees clockwise and preserve portrait
20011 transpose=dir=1:passthrough=portrait
20014 The command above can also be specified as:
20016 transpose=1:portrait
20019 @section transpose_npp
20021 Transpose rows with columns in the input video and optionally flip it.
20022 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
20024 It accepts the following parameters:
20029 Specify the transposition direction.
20031 Can assume the following values:
20034 Rotate by 90 degrees counterclockwise and vertically flip. (default)
20037 Rotate by 90 degrees clockwise.
20040 Rotate by 90 degrees counterclockwise.
20043 Rotate by 90 degrees clockwise and vertically flip.
20047 Do not apply the transposition if the input geometry matches the one
20048 specified by the specified value. It accepts the following values:
20051 Always apply transposition. (default)
20053 Preserve portrait geometry (when @var{height} >= @var{width}).
20055 Preserve landscape geometry (when @var{width} >= @var{height}).
20061 Trim the input so that the output contains one continuous subpart of the input.
20063 It accepts the following parameters:
20066 Specify the time of the start of the kept section, i.e. the frame with the
20067 timestamp @var{start} will be the first frame in the output.
20070 Specify the time of the first frame that will be dropped, i.e. the frame
20071 immediately preceding the one with the timestamp @var{end} will be the last
20072 frame in the output.
20075 This is the same as @var{start}, except this option sets the start timestamp
20076 in timebase units instead of seconds.
20079 This is the same as @var{end}, except this option sets the end timestamp
20080 in timebase units instead of seconds.
20083 The maximum duration of the output in seconds.
20086 The number of the first frame that should be passed to the output.
20089 The number of the first frame that should be dropped.
20092 @option{start}, @option{end}, and @option{duration} are expressed as time
20093 duration specifications; see
20094 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20095 for the accepted syntax.
20097 Note that the first two sets of the start/end options and the @option{duration}
20098 option look at the frame timestamp, while the _frame variants simply count the
20099 frames that pass through the filter. Also note that this filter does not modify
20100 the timestamps. If you wish for the output timestamps to start at zero, insert a
20101 setpts filter after the trim filter.
20103 If multiple start or end options are set, this filter tries to be greedy and
20104 keep all the frames that match at least one of the specified constraints. To keep
20105 only the part that matches all the constraints at once, chain multiple trim
20108 The defaults are such that all the input is kept. So it is possible to set e.g.
20109 just the end values to keep everything before the specified time.
20114 Drop everything except the second minute of input:
20116 ffmpeg -i INPUT -vf trim=60:120
20120 Keep only the first second:
20122 ffmpeg -i INPUT -vf trim=duration=1
20127 @section unpremultiply
20128 Apply alpha unpremultiply effect to input video stream using first plane
20129 of second stream as alpha.
20131 Both streams must have same dimensions and same pixel format.
20133 The filter accepts the following option:
20137 Set which planes will be processed, unprocessed planes will be copied.
20138 By default value 0xf, all planes will be processed.
20140 If the format has 1 or 2 components, then luma is bit 0.
20141 If the format has 3 or 4 components:
20142 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
20143 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
20144 If present, the alpha channel is always the last bit.
20147 Do not require 2nd input for processing, instead use alpha plane from input stream.
20153 Sharpen or blur the input video.
20155 It accepts the following parameters:
20158 @item luma_msize_x, lx
20159 Set the luma matrix horizontal size. It must be an odd integer between
20160 3 and 23. The default value is 5.
20162 @item luma_msize_y, ly
20163 Set the luma matrix vertical size. It must be an odd integer between 3
20164 and 23. The default value is 5.
20166 @item luma_amount, la
20167 Set the luma effect strength. It must be a floating point number, reasonable
20168 values lay between -1.5 and 1.5.
20170 Negative values will blur the input video, while positive values will
20171 sharpen it, a value of zero will disable the effect.
20173 Default value is 1.0.
20175 @item chroma_msize_x, cx
20176 Set the chroma matrix horizontal size. It must be an odd integer
20177 between 3 and 23. The default value is 5.
20179 @item chroma_msize_y, cy
20180 Set the chroma matrix vertical size. It must be an odd integer
20181 between 3 and 23. The default value is 5.
20183 @item chroma_amount, ca
20184 Set the chroma effect strength. It must be a floating point number, reasonable
20185 values lay between -1.5 and 1.5.
20187 Negative values will blur the input video, while positive values will
20188 sharpen it, a value of zero will disable the effect.
20190 Default value is 0.0.
20194 All parameters are optional and default to the equivalent of the
20195 string '5:5:1.0:5:5:0.0'.
20197 @subsection Examples
20201 Apply strong luma sharpen effect:
20203 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
20207 Apply a strong blur of both luma and chroma parameters:
20209 unsharp=7:7:-2:7:7:-2
20216 Decompose a video made of tiled images into the individual images.
20218 The frame rate of the output video is the frame rate of the input video
20219 multiplied by the number of tiles.
20221 This filter does the reverse of @ref{tile}.
20223 The filter accepts the following options:
20228 Set the grid size (i.e. the number of lines and columns). For the syntax of
20229 this option, check the
20230 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20233 @subsection Examples
20237 Produce a 1-second video from a still image file made of 25 frames stacked
20238 vertically, like an analogic film reel:
20240 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
20246 Apply ultra slow/simple postprocessing filter that compresses and decompresses
20247 the image at several (or - in the case of @option{quality} level @code{8} - all)
20248 shifts and average the results.
20250 The way this differs from the behavior of spp is that uspp actually encodes &
20251 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
20252 DCT similar to MJPEG.
20254 The filter accepts the following options:
20258 Set quality. This option defines the number of levels for averaging. It accepts
20259 an integer in the range 0-8. If set to @code{0}, the filter will have no
20260 effect. A value of @code{8} means the higher quality. For each increment of
20261 that value the speed drops by a factor of approximately 2. Default value is
20265 Force a constant quantization parameter. If not set, the filter will use the QP
20266 from the video stream (if available).
20271 Convert 360 videos between various formats.
20273 The filter accepts the following options:
20279 Set format of the input/output video.
20287 Equirectangular projection.
20292 Cubemap with 3x2/6x1/1x6 layout.
20294 Format specific options:
20299 Set padding proportion for the input/output cubemap. Values in decimals.
20306 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)
20309 Default value is @b{@samp{0}}.
20310 Maximum value is @b{@samp{0.1}}.
20314 Set fixed padding for the input/output cubemap. Values in pixels.
20316 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
20320 Set order of faces for the input/output cubemap. Choose one direction for each position.
20322 Designation of directions:
20338 Default value is @b{@samp{rludfb}}.
20342 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
20344 Designation of angles:
20347 0 degrees clockwise
20349 90 degrees clockwise
20351 180 degrees clockwise
20353 270 degrees clockwise
20356 Default value is @b{@samp{000000}}.
20360 Equi-Angular Cubemap.
20367 Format specific options:
20372 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20374 If diagonal field of view is set it overrides horizontal and vertical field of view.
20379 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20381 If diagonal field of view is set it overrides horizontal and vertical field of view.
20387 Format specific options:
20392 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20394 If diagonal field of view is set it overrides horizontal and vertical field of view.
20399 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20401 If diagonal field of view is set it overrides horizontal and vertical field of view.
20407 Facebook's 360 formats.
20410 Stereographic format.
20412 Format specific options:
20417 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20419 If diagonal field of view is set it overrides horizontal and vertical field of view.
20424 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20426 If diagonal field of view is set it overrides horizontal and vertical field of view.
20433 Ball format, gives significant distortion toward the back.
20436 Hammer-Aitoff map projection format.
20439 Sinusoidal map projection format.
20442 Fisheye projection.
20444 Format specific options:
20449 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20451 If diagonal field of view is set it overrides horizontal and vertical field of view.
20456 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20458 If diagonal field of view is set it overrides horizontal and vertical field of view.
20462 Pannini projection.
20464 Format specific options:
20467 Set output pannini parameter.
20470 Set input pannini parameter.
20474 Cylindrical projection.
20476 Format specific options:
20481 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20483 If diagonal field of view is set it overrides horizontal and vertical field of view.
20488 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20490 If diagonal field of view is set it overrides horizontal and vertical field of view.
20494 Perspective projection. @i{(output only)}
20496 Format specific options:
20499 Set perspective parameter.
20503 Tetrahedron projection.
20506 Truncated square pyramid projection.
20510 Half equirectangular projection.
20515 Format specific options:
20520 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20522 If diagonal field of view is set it overrides horizontal and vertical field of view.
20527 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20529 If diagonal field of view is set it overrides horizontal and vertical field of view.
20533 Orthographic format.
20535 Format specific options:
20540 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20542 If diagonal field of view is set it overrides horizontal and vertical field of view.
20547 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20549 If diagonal field of view is set it overrides horizontal and vertical field of view.
20553 Octahedron projection.
20557 Set interpolation method.@*
20558 @i{Note: more complex interpolation methods require much more memory to run.}
20568 Bilinear interpolation.
20570 Lagrange9 interpolation.
20573 Bicubic interpolation.
20576 Lanczos interpolation.
20579 Spline16 interpolation.
20582 Gaussian interpolation.
20584 Mitchell interpolation.
20587 Default value is @b{@samp{line}}.
20591 Set the output video resolution.
20593 Default resolution depends on formats.
20597 Set the input/output stereo format.
20608 Default value is @b{@samp{2d}} for input and output format.
20613 Set rotation for the output video. Values in degrees.
20616 Set rotation order for the output video. Choose one item for each position.
20627 Default value is @b{@samp{ypr}}.
20632 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
20636 Set if input video is flipped horizontally/vertically. Boolean values.
20639 Set if input video is transposed. Boolean value, by default disabled.
20642 Set if output video needs to be transposed. Boolean value, by default disabled.
20645 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
20648 @subsection Examples
20652 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
20654 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
20657 Extract back view of Equi-Angular Cubemap:
20659 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
20662 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
20664 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
20668 @subsection Commands
20670 This filter supports subset of above options as @ref{commands}.
20672 @section vaguedenoiser
20674 Apply a wavelet based denoiser.
20676 It transforms each frame from the video input into the wavelet domain,
20677 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
20678 the obtained coefficients. It does an inverse wavelet transform after.
20679 Due to wavelet properties, it should give a nice smoothed result, and
20680 reduced noise, without blurring picture features.
20682 This filter accepts the following options:
20686 The filtering strength. The higher, the more filtered the video will be.
20687 Hard thresholding can use a higher threshold than soft thresholding
20688 before the video looks overfiltered. Default value is 2.
20691 The filtering method the filter will use.
20693 It accepts the following values:
20696 All values under the threshold will be zeroed.
20699 All values under the threshold will be zeroed. All values above will be
20700 reduced by the threshold.
20703 Scales or nullifies coefficients - intermediary between (more) soft and
20704 (less) hard thresholding.
20707 Default is garrote.
20710 Number of times, the wavelet will decompose the picture. Picture can't
20711 be decomposed beyond a particular point (typically, 8 for a 640x480
20712 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
20715 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
20718 A list of the planes to process. By default all planes are processed.
20721 The threshold type the filter will use.
20723 It accepts the following values:
20726 Threshold used is same for all decompositions.
20729 Threshold used depends also on each decomposition coefficients.
20732 Default is universal.
20735 @section vectorscope
20737 Display 2 color component values in the two dimensional graph (which is called
20740 This filter accepts the following options:
20744 Set vectorscope mode.
20746 It accepts the following values:
20750 Gray values are displayed on graph, higher brightness means more pixels have
20751 same component color value on location in graph. This is the default mode.
20754 Gray values are displayed on graph. Surrounding pixels values which are not
20755 present in video frame are drawn in gradient of 2 color components which are
20756 set by option @code{x} and @code{y}. The 3rd color component is static.
20759 Actual color components values present in video frame are displayed on graph.
20762 Similar as color2 but higher frequency of same values @code{x} and @code{y}
20763 on graph increases value of another color component, which is luminance by
20764 default values of @code{x} and @code{y}.
20767 Actual colors present in video frame are displayed on graph. If two different
20768 colors map to same position on graph then color with higher value of component
20769 not present in graph is picked.
20772 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
20773 component picked from radial gradient.
20777 Set which color component will be represented on X-axis. Default is @code{1}.
20780 Set which color component will be represented on Y-axis. Default is @code{2}.
20783 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
20784 of color component which represents frequency of (X, Y) location in graph.
20789 No envelope, this is default.
20792 Instant envelope, even darkest single pixel will be clearly highlighted.
20795 Hold maximum and minimum values presented in graph over time. This way you
20796 can still spot out of range values without constantly looking at vectorscope.
20799 Peak and instant envelope combined together.
20803 Set what kind of graticule to draw.
20812 Set graticule opacity.
20815 Set graticule flags.
20819 Draw graticule for white point.
20822 Draw graticule for black point.
20825 Draw color points short names.
20829 Set background opacity.
20831 @item lthreshold, l
20832 Set low threshold for color component not represented on X or Y axis.
20833 Values lower than this value will be ignored. Default is 0.
20834 Note this value is multiplied with actual max possible value one pixel component
20835 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
20838 @item hthreshold, h
20839 Set high threshold for color component not represented on X or Y axis.
20840 Values higher than this value will be ignored. Default is 1.
20841 Note this value is multiplied with actual max possible value one pixel component
20842 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
20843 is 0.9 * 255 = 230.
20845 @item colorspace, c
20846 Set what kind of colorspace to use when drawing graticule.
20856 Set color tint for gray/tint vectorscope mode. By default both options are zero.
20857 This means no tint, and output will remain gray.
20860 @anchor{vidstabdetect}
20861 @section vidstabdetect
20863 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
20864 @ref{vidstabtransform} for pass 2.
20866 This filter generates a file with relative translation and rotation
20867 transform information about subsequent frames, which is then used by
20868 the @ref{vidstabtransform} filter.
20870 To enable compilation of this filter you need to configure FFmpeg with
20871 @code{--enable-libvidstab}.
20873 This filter accepts the following options:
20877 Set the path to the file used to write the transforms information.
20878 Default value is @file{transforms.trf}.
20881 Set how shaky the video is and how quick the camera is. It accepts an
20882 integer in the range 1-10, a value of 1 means little shakiness, a
20883 value of 10 means strong shakiness. Default value is 5.
20886 Set the accuracy of the detection process. It must be a value in the
20887 range 1-15. A value of 1 means low accuracy, a value of 15 means high
20888 accuracy. Default value is 15.
20891 Set stepsize of the search process. The region around minimum is
20892 scanned with 1 pixel resolution. Default value is 6.
20895 Set minimum contrast. Below this value a local measurement field is
20896 discarded. Must be a floating point value in the range 0-1. Default
20900 Set reference frame number for tripod mode.
20902 If enabled, the motion of the frames is compared to a reference frame
20903 in the filtered stream, identified by the specified number. The idea
20904 is to compensate all movements in a more-or-less static scene and keep
20905 the camera view absolutely still.
20907 If set to 0, it is disabled. The frames are counted starting from 1.
20910 Show fields and transforms in the resulting frames. It accepts an
20911 integer in the range 0-2. Default value is 0, which disables any
20915 @subsection Examples
20919 Use default values:
20925 Analyze strongly shaky movie and put the results in file
20926 @file{mytransforms.trf}:
20928 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
20932 Visualize the result of internal transformations in the resulting
20935 vidstabdetect=show=1
20939 Analyze a video with medium shakiness using @command{ffmpeg}:
20941 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
20945 @anchor{vidstabtransform}
20946 @section vidstabtransform
20948 Video stabilization/deshaking: pass 2 of 2,
20949 see @ref{vidstabdetect} for pass 1.
20951 Read a file with transform information for each frame and
20952 apply/compensate them. Together with the @ref{vidstabdetect}
20953 filter this can be used to deshake videos. See also
20954 @url{http://public.hronopik.de/vid.stab}. It is important to also use
20955 the @ref{unsharp} filter, see below.
20957 To enable compilation of this filter you need to configure FFmpeg with
20958 @code{--enable-libvidstab}.
20960 @subsection Options
20964 Set path to the file used to read the transforms. Default value is
20965 @file{transforms.trf}.
20968 Set the number of frames (value*2 + 1) used for lowpass filtering the
20969 camera movements. Default value is 10.
20971 For example a number of 10 means that 21 frames are used (10 in the
20972 past and 10 in the future) to smoothen the motion in the video. A
20973 larger value leads to a smoother video, but limits the acceleration of
20974 the camera (pan/tilt movements). 0 is a special case where a static
20975 camera is simulated.
20978 Set the camera path optimization algorithm.
20980 Accepted values are:
20983 gaussian kernel low-pass filter on camera motion (default)
20985 averaging on transformations
20989 Set maximal number of pixels to translate frames. Default value is -1,
20993 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
20994 value is -1, meaning no limit.
20997 Specify how to deal with borders that may be visible due to movement
21000 Available values are:
21003 keep image information from previous frame (default)
21005 fill the border black
21009 Invert transforms if set to 1. Default value is 0.
21012 Consider transforms as relative to previous frame if set to 1,
21013 absolute if set to 0. Default value is 0.
21016 Set percentage to zoom. A positive value will result in a zoom-in
21017 effect, a negative value in a zoom-out effect. Default value is 0 (no
21021 Set optimal zooming to avoid borders.
21023 Accepted values are:
21028 optimal static zoom value is determined (only very strong movements
21029 will lead to visible borders) (default)
21031 optimal adaptive zoom value is determined (no borders will be
21032 visible), see @option{zoomspeed}
21035 Note that the value given at zoom is added to the one calculated here.
21038 Set percent to zoom maximally each frame (enabled when
21039 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
21043 Specify type of interpolation.
21045 Available values are:
21050 linear only horizontal
21052 linear in both directions (default)
21054 cubic in both directions (slow)
21058 Enable virtual tripod mode if set to 1, which is equivalent to
21059 @code{relative=0:smoothing=0}. Default value is 0.
21061 Use also @code{tripod} option of @ref{vidstabdetect}.
21064 Increase log verbosity if set to 1. Also the detected global motions
21065 are written to the temporary file @file{global_motions.trf}. Default
21069 @subsection Examples
21073 Use @command{ffmpeg} for a typical stabilization with default values:
21075 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
21078 Note the use of the @ref{unsharp} filter which is always recommended.
21081 Zoom in a bit more and load transform data from a given file:
21083 vidstabtransform=zoom=5:input="mytransforms.trf"
21087 Smoothen the video even more:
21089 vidstabtransform=smoothing=30
21095 Flip the input video vertically.
21097 For example, to vertically flip a video with @command{ffmpeg}:
21099 ffmpeg -i in.avi -vf "vflip" out.avi
21104 Detect variable frame rate video.
21106 This filter tries to detect if the input is variable or constant frame rate.
21108 At end it will output number of frames detected as having variable delta pts,
21109 and ones with constant delta pts.
21110 If there was frames with variable delta, than it will also show min, max and
21111 average delta encountered.
21115 Boost or alter saturation.
21117 The filter accepts the following options:
21120 Set strength of boost if positive value or strength of alter if negative value.
21121 Default is 0. Allowed range is from -2 to 2.
21124 Set the red balance. Default is 1. Allowed range is from -10 to 10.
21127 Set the green balance. Default is 1. Allowed range is from -10 to 10.
21130 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
21133 Set the red luma coefficient.
21136 Set the green luma coefficient.
21139 Set the blue luma coefficient.
21142 If @code{intensity} is negative and this is set to 1, colors will change,
21143 otherwise colors will be less saturated, more towards gray.
21146 @subsection Commands
21148 This filter supports the all above options as @ref{commands}.
21153 Make or reverse a natural vignetting effect.
21155 The filter accepts the following options:
21159 Set lens angle expression as a number of radians.
21161 The value is clipped in the @code{[0,PI/2]} range.
21163 Default value: @code{"PI/5"}
21167 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
21171 Set forward/backward mode.
21173 Available modes are:
21176 The larger the distance from the central point, the darker the image becomes.
21179 The larger the distance from the central point, the brighter the image becomes.
21180 This can be used to reverse a vignette effect, though there is no automatic
21181 detection to extract the lens @option{angle} and other settings (yet). It can
21182 also be used to create a burning effect.
21185 Default value is @samp{forward}.
21188 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
21190 It accepts the following values:
21193 Evaluate expressions only once during the filter initialization.
21196 Evaluate expressions for each incoming frame. This is way slower than the
21197 @samp{init} mode since it requires all the scalers to be re-computed, but it
21198 allows advanced dynamic expressions.
21201 Default value is @samp{init}.
21204 Set dithering to reduce the circular banding effects. Default is @code{1}
21208 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
21209 Setting this value to the SAR of the input will make a rectangular vignetting
21210 following the dimensions of the video.
21212 Default is @code{1/1}.
21215 @subsection Expressions
21217 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
21218 following parameters.
21223 input width and height
21226 the number of input frame, starting from 0
21229 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
21230 @var{TB} units, NAN if undefined
21233 frame rate of the input video, NAN if the input frame rate is unknown
21236 the PTS (Presentation TimeStamp) of the filtered video frame,
21237 expressed in seconds, NAN if undefined
21240 time base of the input video
21244 @subsection Examples
21248 Apply simple strong vignetting effect:
21254 Make a flickering vignetting:
21256 vignette='PI/4+random(1)*PI/50':eval=frame
21261 @section vmafmotion
21263 Obtain the average VMAF motion score of a video.
21264 It is one of the component metrics of VMAF.
21266 The obtained average motion score is printed through the logging system.
21268 The filter accepts the following options:
21272 If specified, the filter will use the named file to save the motion score of
21273 each frame with respect to the previous frame.
21274 When filename equals "-" the data is sent to standard output.
21279 ffmpeg -i ref.mpg -vf vmafmotion -f null -
21283 Stack input videos vertically.
21285 All streams must be of same pixel format and of same width.
21287 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
21288 to create same output.
21290 The filter accepts the following options:
21294 Set number of input streams. Default is 2.
21297 If set to 1, force the output to terminate when the shortest input
21298 terminates. Default value is 0.
21303 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
21304 Deinterlacing Filter").
21306 Based on the process described by Martin Weston for BBC R&D, and
21307 implemented based on the de-interlace algorithm written by Jim
21308 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
21309 uses filter coefficients calculated by BBC R&D.
21311 This filter uses field-dominance information in frame to decide which
21312 of each pair of fields to place first in the output.
21313 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
21315 There are two sets of filter coefficients, so called "simple"
21316 and "complex". Which set of filter coefficients is used can
21317 be set by passing an optional parameter:
21321 Set the interlacing filter coefficients. Accepts one of the following values:
21325 Simple filter coefficient set.
21327 More-complex filter coefficient set.
21329 Default value is @samp{complex}.
21332 The interlacing mode to adopt. It accepts one of the following values:
21336 Output one frame for each frame.
21338 Output one frame for each field.
21341 The default value is @code{field}.
21344 The picture field parity assumed for the input interlaced video. It accepts one
21345 of the following values:
21349 Assume the top field is first.
21351 Assume the bottom field is first.
21353 Enable automatic detection of field parity.
21356 The default value is @code{auto}.
21357 If the interlacing is unknown or the decoder does not export this information,
21358 top field first will be assumed.
21361 Specify which frames to deinterlace. Accepts one of the following values:
21365 Deinterlace all frames,
21367 Only deinterlace frames marked as interlaced.
21370 Default value is @samp{all}.
21373 @subsection Commands
21374 This filter supports same @ref{commands} as options.
21377 Video waveform monitor.
21379 The waveform monitor plots color component intensity. By default luminance
21380 only. Each column of the waveform corresponds to a column of pixels in the
21383 It accepts the following options:
21387 Can be either @code{row}, or @code{column}. Default is @code{column}.
21388 In row mode, the graph on the left side represents color component value 0 and
21389 the right side represents value = 255. In column mode, the top side represents
21390 color component value = 0 and bottom side represents value = 255.
21393 Set intensity. Smaller values are useful to find out how many values of the same
21394 luminance are distributed across input rows/columns.
21395 Default value is @code{0.04}. Allowed range is [0, 1].
21398 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
21399 In mirrored mode, higher values will be represented on the left
21400 side for @code{row} mode and at the top for @code{column} mode. Default is
21401 @code{1} (mirrored).
21405 It accepts the following values:
21408 Presents information identical to that in the @code{parade}, except
21409 that the graphs representing color components are superimposed directly
21412 This display mode makes it easier to spot relative differences or similarities
21413 in overlapping areas of the color components that are supposed to be identical,
21414 such as neutral whites, grays, or blacks.
21417 Display separate graph for the color components side by side in
21418 @code{row} mode or one below the other in @code{column} mode.
21421 Display separate graph for the color components side by side in
21422 @code{column} mode or one below the other in @code{row} mode.
21424 Using this display mode makes it easy to spot color casts in the highlights
21425 and shadows of an image, by comparing the contours of the top and the bottom
21426 graphs of each waveform. Since whites, grays, and blacks are characterized
21427 by exactly equal amounts of red, green, and blue, neutral areas of the picture
21428 should display three waveforms of roughly equal width/height. If not, the
21429 correction is easy to perform by making level adjustments the three waveforms.
21431 Default is @code{stack}.
21433 @item components, c
21434 Set which color components to display. Default is 1, which means only luminance
21435 or red color component if input is in RGB colorspace. If is set for example to
21436 7 it will display all 3 (if) available color components.
21441 No envelope, this is default.
21444 Instant envelope, minimum and maximum values presented in graph will be easily
21445 visible even with small @code{step} value.
21448 Hold minimum and maximum values presented in graph across time. This way you
21449 can still spot out of range values without constantly looking at waveforms.
21452 Peak and instant envelope combined together.
21458 No filtering, this is default.
21461 Luma and chroma combined together.
21464 Similar as above, but shows difference between blue and red chroma.
21467 Similar as above, but use different colors.
21470 Similar as above, but again with different colors.
21473 Displays only chroma.
21476 Displays actual color value on waveform.
21479 Similar as above, but with luma showing frequency of chroma values.
21483 Set which graticule to display.
21487 Do not display graticule.
21490 Display green graticule showing legal broadcast ranges.
21493 Display orange graticule showing legal broadcast ranges.
21496 Display invert graticule showing legal broadcast ranges.
21500 Set graticule opacity.
21503 Set graticule flags.
21507 Draw numbers above lines. By default enabled.
21510 Draw dots instead of lines.
21514 Set scale used for displaying graticule.
21521 Default is digital.
21524 Set background opacity.
21528 Set tint for output.
21529 Only used with lowpass filter and when display is not overlay and input
21530 pixel formats are not RGB.
21533 @section weave, doubleweave
21535 The @code{weave} takes a field-based video input and join
21536 each two sequential fields into single frame, producing a new double
21537 height clip with half the frame rate and half the frame count.
21539 The @code{doubleweave} works same as @code{weave} but without
21540 halving frame rate and frame count.
21542 It accepts the following option:
21546 Set first field. Available values are:
21550 Set the frame as top-field-first.
21553 Set the frame as bottom-field-first.
21557 @subsection Examples
21561 Interlace video using @ref{select} and @ref{separatefields} filter:
21563 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
21568 Apply the xBR high-quality magnification filter which is designed for pixel
21569 art. It follows a set of edge-detection rules, see
21570 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
21572 It accepts the following option:
21576 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
21577 @code{3xBR} and @code{4} for @code{4xBR}.
21578 Default is @code{3}.
21583 Apply cross fade from one input video stream to another input video stream.
21584 The cross fade is applied for specified duration.
21586 The filter accepts the following options:
21590 Set one of available transition effects:
21638 Default transition effect is fade.
21641 Set cross fade duration in seconds.
21642 Default duration is 1 second.
21645 Set cross fade start relative to first input stream in seconds.
21646 Default offset is 0.
21649 Set expression for custom transition effect.
21651 The expressions can use the following variables and functions:
21656 The coordinates of the current sample.
21660 The width and height of the image.
21663 Progress of transition effect.
21666 Currently processed plane.
21669 Return value of first input at current location and plane.
21672 Return value of second input at current location and plane.
21678 Return the value of the pixel at location (@var{x},@var{y}) of the
21679 first/second/third/fourth component of first input.
21685 Return the value of the pixel at location (@var{x},@var{y}) of the
21686 first/second/third/fourth component of second input.
21690 @subsection Examples
21694 Cross fade from one input video to another input video, with fade transition and duration of transition
21695 of 2 seconds starting at offset of 5 seconds:
21697 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
21702 Pick median pixels from several input videos.
21704 The filter accepts the following options:
21708 Set number of inputs.
21709 Default is 3. Allowed range is from 3 to 255.
21710 If number of inputs is even number, than result will be mean value between two median values.
21713 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
21716 Set median percentile. Default value is @code{0.5}.
21717 Default value of @code{0.5} will pick always median values, while @code{0} will pick
21718 minimum values, and @code{1} maximum values.
21721 @subsection Commands
21723 This filter supports all above options as @ref{commands}, excluding option @code{inputs}.
21726 Stack video inputs into custom layout.
21728 All streams must be of same pixel format.
21730 The filter accepts the following options:
21734 Set number of input streams. Default is 2.
21737 Specify layout of inputs.
21738 This option requires the desired layout configuration to be explicitly set by the user.
21739 This sets position of each video input in output. Each input
21740 is separated by '|'.
21741 The first number represents the column, and the second number represents the row.
21742 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
21743 where X is video input from which to take width or height.
21744 Multiple values can be used when separated by '+'. In such
21745 case values are summed together.
21747 Note that if inputs are of different sizes gaps may appear, as not all of
21748 the output video frame will be filled. Similarly, videos can overlap each
21749 other if their position doesn't leave enough space for the full frame of
21752 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
21753 a layout must be set by the user.
21756 If set to 1, force the output to terminate when the shortest input
21757 terminates. Default value is 0.
21760 If set to valid color, all unused pixels will be filled with that color.
21761 By default fill is set to none, so it is disabled.
21764 @subsection Examples
21768 Display 4 inputs into 2x2 grid.
21772 input1(0, 0) | input3(w0, 0)
21773 input2(0, h0) | input4(w0, h0)
21777 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
21780 Note that if inputs are of different sizes, gaps or overlaps may occur.
21783 Display 4 inputs into 1x4 grid.
21790 input4(0, h0+h1+h2)
21794 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
21797 Note that if inputs are of different widths, unused space will appear.
21800 Display 9 inputs into 3x3 grid.
21804 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
21805 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
21806 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
21810 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
21813 Note that if inputs are of different sizes, gaps or overlaps may occur.
21816 Display 16 inputs into 4x4 grid.
21820 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
21821 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
21822 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
21823 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
21827 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|
21828 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
21831 Note that if inputs are of different sizes, gaps or overlaps may occur.
21838 Deinterlace the input video ("yadif" means "yet another deinterlacing
21841 It accepts the following parameters:
21847 The interlacing mode to adopt. It accepts one of the following values:
21850 @item 0, send_frame
21851 Output one frame for each frame.
21852 @item 1, send_field
21853 Output one frame for each field.
21854 @item 2, send_frame_nospatial
21855 Like @code{send_frame}, but it skips the spatial interlacing check.
21856 @item 3, send_field_nospatial
21857 Like @code{send_field}, but it skips the spatial interlacing check.
21860 The default value is @code{send_frame}.
21863 The picture field parity assumed for the input interlaced video. It accepts one
21864 of the following values:
21868 Assume the top field is first.
21870 Assume the bottom field is first.
21872 Enable automatic detection of field parity.
21875 The default value is @code{auto}.
21876 If the interlacing is unknown or the decoder does not export this information,
21877 top field first will be assumed.
21880 Specify which frames to deinterlace. Accepts one of the following
21885 Deinterlace all frames.
21886 @item 1, interlaced
21887 Only deinterlace frames marked as interlaced.
21890 The default value is @code{all}.
21893 @section yadif_cuda
21895 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
21896 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
21899 It accepts the following parameters:
21905 The interlacing mode to adopt. It accepts one of the following values:
21908 @item 0, send_frame
21909 Output one frame for each frame.
21910 @item 1, send_field
21911 Output one frame for each field.
21912 @item 2, send_frame_nospatial
21913 Like @code{send_frame}, but it skips the spatial interlacing check.
21914 @item 3, send_field_nospatial
21915 Like @code{send_field}, but it skips the spatial interlacing check.
21918 The default value is @code{send_frame}.
21921 The picture field parity assumed for the input interlaced video. It accepts one
21922 of the following values:
21926 Assume the top field is first.
21928 Assume the bottom field is first.
21930 Enable automatic detection of field parity.
21933 The default value is @code{auto}.
21934 If the interlacing is unknown or the decoder does not export this information,
21935 top field first will be assumed.
21938 Specify which frames to deinterlace. Accepts one of the following
21943 Deinterlace all frames.
21944 @item 1, interlaced
21945 Only deinterlace frames marked as interlaced.
21948 The default value is @code{all}.
21953 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
21954 The algorithm is described in
21955 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
21957 It accepts the following parameters:
21961 Set the window radius. Default value is 3.
21964 Set which planes to filter. Default is only the first plane.
21967 Set blur strength. Default value is 128.
21970 @subsection Commands
21971 This filter supports same @ref{commands} as options.
21975 Apply Zoom & Pan effect.
21977 This filter accepts the following options:
21981 Set the zoom expression. Range is 1-10. Default is 1.
21985 Set the x and y expression. Default is 0.
21988 Set the duration expression in number of frames.
21989 This sets for how many number of frames effect will last for
21990 single input image.
21993 Set the output image size, default is 'hd720'.
21996 Set the output frame rate, default is '25'.
21999 Each expression can contain the following constants:
22018 Output frame count.
22021 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
22023 @item out_time, time, ot
22024 The output timestamp expressed in seconds.
22028 Last calculated 'x' and 'y' position from 'x' and 'y' expression
22029 for current input frame.
22033 'x' and 'y' of last output frame of previous input frame or 0 when there was
22034 not yet such frame (first input frame).
22037 Last calculated zoom from 'z' expression for current input frame.
22040 Last calculated zoom of last output frame of previous input frame.
22043 Number of output frames for current input frame. Calculated from 'd' expression
22044 for each input frame.
22047 number of output frames created for previous input frame
22050 Rational number: input width / input height
22053 sample aspect ratio
22056 display aspect ratio
22060 @subsection Examples
22064 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
22066 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
22070 Zoom in up to 1.5x and pan always at center of picture:
22072 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22076 Same as above but without pausing:
22078 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22082 Zoom in 2x into center of picture only for the first second of the input video:
22084 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22091 Scale (resize) the input video, using the z.lib library:
22092 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
22093 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
22095 The zscale filter forces the output display aspect ratio to be the same
22096 as the input, by changing the output sample aspect ratio.
22098 If the input image format is different from the format requested by
22099 the next filter, the zscale filter will convert the input to the
22102 @subsection Options
22103 The filter accepts the following options.
22108 Set the output video dimension expression. Default value is the input
22111 If the @var{width} or @var{w} value is 0, the input width is used for
22112 the output. If the @var{height} or @var{h} value is 0, the input height
22113 is used for the output.
22115 If one and only one of the values is -n with n >= 1, the zscale filter
22116 will use a value that maintains the aspect ratio of the input image,
22117 calculated from the other specified dimension. After that it will,
22118 however, make sure that the calculated dimension is divisible by n and
22119 adjust the value if necessary.
22121 If both values are -n with n >= 1, the behavior will be identical to
22122 both values being set to 0 as previously detailed.
22124 See below for the list of accepted constants for use in the dimension
22128 Set the video size. For the syntax of this option, check the
22129 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22132 Set the dither type.
22134 Possible values are:
22139 @item error_diffusion
22145 Set the resize filter type.
22147 Possible values are:
22157 Default is bilinear.
22160 Set the color range.
22162 Possible values are:
22169 Default is same as input.
22172 Set the color primaries.
22174 Possible values are:
22184 Default is same as input.
22187 Set the transfer characteristics.
22189 Possible values are:
22203 Default is same as input.
22206 Set the colorspace matrix.
22208 Possible value are:
22219 Default is same as input.
22222 Set the input color range.
22224 Possible values are:
22231 Default is same as input.
22233 @item primariesin, pin
22234 Set the input color primaries.
22236 Possible values are:
22246 Default is same as input.
22248 @item transferin, tin
22249 Set the input transfer characteristics.
22251 Possible values are:
22262 Default is same as input.
22264 @item matrixin, min
22265 Set the input colorspace matrix.
22267 Possible value are:
22279 Set the output chroma location.
22281 Possible values are:
22292 @item chromalin, cin
22293 Set the input chroma location.
22295 Possible values are:
22307 Set the nominal peak luminance.
22310 The values of the @option{w} and @option{h} options are expressions
22311 containing the following constants:
22316 The input width and height
22320 These are the same as @var{in_w} and @var{in_h}.
22324 The output (scaled) width and height
22328 These are the same as @var{out_w} and @var{out_h}
22331 The same as @var{iw} / @var{ih}
22334 input sample aspect ratio
22337 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
22341 horizontal and vertical input chroma subsample values. For example for the
22342 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22346 horizontal and vertical output chroma subsample values. For example for the
22347 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22350 @subsection Commands
22352 This filter supports the following commands:
22356 Set the output video dimension expression.
22357 The command accepts the same syntax of the corresponding option.
22359 If the specified expression is not valid, it is kept at its current
22363 @c man end VIDEO FILTERS
22365 @chapter OpenCL Video Filters
22366 @c man begin OPENCL VIDEO FILTERS
22368 Below is a description of the currently available OpenCL video filters.
22370 To enable compilation of these filters you need to configure FFmpeg with
22371 @code{--enable-opencl}.
22373 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
22376 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
22377 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
22378 given device parameters.
22380 @item -filter_hw_device @var{name}
22381 Pass the hardware device called @var{name} to all filters in any filter graph.
22385 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
22389 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
22391 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
22395 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.
22397 @section avgblur_opencl
22399 Apply average blur filter.
22401 The filter accepts the following options:
22405 Set horizontal radius size.
22406 Range is @code{[1, 1024]} and default value is @code{1}.
22409 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22412 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
22415 @subsection Example
22419 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.
22421 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
22425 @section boxblur_opencl
22427 Apply a boxblur algorithm to the input video.
22429 It accepts the following parameters:
22433 @item luma_radius, lr
22434 @item luma_power, lp
22435 @item chroma_radius, cr
22436 @item chroma_power, cp
22437 @item alpha_radius, ar
22438 @item alpha_power, ap
22442 A description of the accepted options follows.
22445 @item luma_radius, lr
22446 @item chroma_radius, cr
22447 @item alpha_radius, ar
22448 Set an expression for the box radius in pixels used for blurring the
22449 corresponding input plane.
22451 The radius value must be a non-negative number, and must not be
22452 greater than the value of the expression @code{min(w,h)/2} for the
22453 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
22456 Default value for @option{luma_radius} is "2". If not specified,
22457 @option{chroma_radius} and @option{alpha_radius} default to the
22458 corresponding value set for @option{luma_radius}.
22460 The expressions can contain the following constants:
22464 The input width and height in pixels.
22468 The input chroma image width and height in pixels.
22472 The horizontal and vertical chroma subsample values. For example, for the
22473 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
22476 @item luma_power, lp
22477 @item chroma_power, cp
22478 @item alpha_power, ap
22479 Specify how many times the boxblur filter is applied to the
22480 corresponding plane.
22482 Default value for @option{luma_power} is 2. If not specified,
22483 @option{chroma_power} and @option{alpha_power} default to the
22484 corresponding value set for @option{luma_power}.
22486 A value of 0 will disable the effect.
22489 @subsection Examples
22491 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.
22495 Apply a boxblur filter with the luma, chroma, and alpha radius
22496 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.
22498 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
22499 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
22503 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.
22505 For the luma plane, a 2x2 box radius will be run once.
22507 For the chroma plane, a 4x4 box radius will be run 5 times.
22509 For the alpha plane, a 3x3 box radius will be run 7 times.
22511 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
22515 @section colorkey_opencl
22516 RGB colorspace color keying.
22518 The filter accepts the following options:
22522 The color which will be replaced with transparency.
22525 Similarity percentage with the key color.
22527 0.01 matches only the exact key color, while 1.0 matches everything.
22532 0.0 makes pixels either fully transparent, or not transparent at all.
22534 Higher values result in semi-transparent pixels, with a higher transparency
22535 the more similar the pixels color is to the key color.
22538 @subsection Examples
22542 Make every semi-green pixel in the input transparent with some slight blending:
22544 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
22548 @section convolution_opencl
22550 Apply convolution of 3x3, 5x5, 7x7 matrix.
22552 The filter accepts the following options:
22559 Set matrix for each plane.
22560 Matrix is sequence of 9, 25 or 49 signed numbers.
22561 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
22567 Set multiplier for calculated value for each plane.
22568 If unset or 0, it will be sum of all matrix elements.
22569 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
22575 Set bias for each plane. This value is added to the result of the multiplication.
22576 Useful for making the overall image brighter or darker.
22577 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
22581 @subsection Examples
22587 -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
22593 -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
22597 Apply edge enhance:
22599 -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
22605 -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
22609 Apply laplacian edge detector which includes diagonals:
22611 -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
22617 -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
22621 @section erosion_opencl
22623 Apply erosion effect to the video.
22625 This filter replaces the pixel by the local(3x3) minimum.
22627 It accepts the following options:
22634 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22635 If @code{0}, plane will remain unchanged.
22638 Flag which specifies the pixel to refer to.
22639 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22641 Flags to local 3x3 coordinates region centered on @code{x}:
22650 @subsection Example
22654 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.
22656 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22660 @section deshake_opencl
22661 Feature-point based video stabilization filter.
22663 The filter accepts the following options:
22667 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
22670 Whether or not additional debug info should be displayed, both in the processed output and in the console.
22672 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
22674 Viewing point matches in the output video is only supported for RGB input.
22676 Defaults to @code{0}.
22678 @item adaptive_crop
22679 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
22681 Defaults to @code{1}.
22683 @item refine_features
22684 Whether or not feature points should be refined at a sub-pixel level.
22686 This can be turned off for a slight performance gain at the cost of precision.
22688 Defaults to @code{1}.
22690 @item smooth_strength
22691 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
22693 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
22695 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
22697 Defaults to @code{0.0}.
22699 @item smooth_window_multiplier
22700 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
22702 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
22704 Acceptable values range from @code{0.1} to @code{10.0}.
22706 Larger values increase the amount of motion data available for determining how to smooth the camera path,
22707 potentially improving smoothness, but also increase latency and memory usage.
22709 Defaults to @code{2.0}.
22713 @subsection Examples
22717 Stabilize a video with a fixed, medium smoothing strength:
22719 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
22723 Stabilize a video with debugging (both in console and in rendered video):
22725 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
22729 @section dilation_opencl
22731 Apply dilation effect to the video.
22733 This filter replaces the pixel by the local(3x3) maximum.
22735 It accepts the following options:
22742 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22743 If @code{0}, plane will remain unchanged.
22746 Flag which specifies the pixel to refer to.
22747 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22749 Flags to local 3x3 coordinates region centered on @code{x}:
22758 @subsection Example
22762 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.
22764 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22768 @section nlmeans_opencl
22770 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
22772 @section overlay_opencl
22774 Overlay one video on top of another.
22776 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
22777 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
22779 The filter accepts the following options:
22784 Set the x coordinate of the overlaid video on the main video.
22785 Default value is @code{0}.
22788 Set the y coordinate of the overlaid video on the main video.
22789 Default value is @code{0}.
22793 @subsection Examples
22797 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
22799 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22802 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
22804 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22809 @section pad_opencl
22811 Add paddings to the input image, and place the original input at the
22812 provided @var{x}, @var{y} coordinates.
22814 It accepts the following options:
22819 Specify an expression for the size of the output image with the
22820 paddings added. If the value for @var{width} or @var{height} is 0, the
22821 corresponding input size is used for the output.
22823 The @var{width} expression can reference the value set by the
22824 @var{height} expression, and vice versa.
22826 The default value of @var{width} and @var{height} is 0.
22830 Specify the offsets to place the input image at within the padded area,
22831 with respect to the top/left border of the output image.
22833 The @var{x} expression can reference the value set by the @var{y}
22834 expression, and vice versa.
22836 The default value of @var{x} and @var{y} is 0.
22838 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
22839 so the input image is centered on the padded area.
22842 Specify the color of the padded area. For the syntax of this option,
22843 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
22844 manual,ffmpeg-utils}.
22847 Pad to an aspect instead to a resolution.
22850 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
22851 options are expressions containing the following constants:
22856 The input video width and height.
22860 These are the same as @var{in_w} and @var{in_h}.
22864 The output width and height (the size of the padded area), as
22865 specified by the @var{width} and @var{height} expressions.
22869 These are the same as @var{out_w} and @var{out_h}.
22873 The x and y offsets as specified by the @var{x} and @var{y}
22874 expressions, or NAN if not yet specified.
22877 same as @var{iw} / @var{ih}
22880 input sample aspect ratio
22883 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
22886 @section prewitt_opencl
22888 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
22890 The filter accepts the following option:
22894 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22897 Set value which will be multiplied with filtered result.
22898 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22901 Set value which will be added to filtered result.
22902 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22905 @subsection Example
22909 Apply the Prewitt operator with scale set to 2 and delta set to 10.
22911 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
22915 @anchor{program_opencl}
22916 @section program_opencl
22918 Filter video using an OpenCL program.
22923 OpenCL program source file.
22926 Kernel name in program.
22929 Number of inputs to the filter. Defaults to 1.
22932 Size of output frames. Defaults to the same as the first input.
22936 The @code{program_opencl} filter also supports the @ref{framesync} options.
22938 The program source file must contain a kernel function with the given name,
22939 which will be run once for each plane of the output. Each run on a plane
22940 gets enqueued as a separate 2D global NDRange with one work-item for each
22941 pixel to be generated. The global ID offset for each work-item is therefore
22942 the coordinates of a pixel in the destination image.
22944 The kernel function needs to take the following arguments:
22947 Destination image, @var{__write_only image2d_t}.
22949 This image will become the output; the kernel should write all of it.
22951 Frame index, @var{unsigned int}.
22953 This is a counter starting from zero and increasing by one for each frame.
22955 Source images, @var{__read_only image2d_t}.
22957 These are the most recent images on each input. The kernel may read from
22958 them to generate the output, but they can't be written to.
22965 Copy the input to the output (output must be the same size as the input).
22967 __kernel void copy(__write_only image2d_t destination,
22968 unsigned int index,
22969 __read_only image2d_t source)
22971 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
22973 int2 location = (int2)(get_global_id(0), get_global_id(1));
22975 float4 value = read_imagef(source, sampler, location);
22977 write_imagef(destination, location, value);
22982 Apply a simple transformation, rotating the input by an amount increasing
22983 with the index counter. Pixel values are linearly interpolated by the
22984 sampler, and the output need not have the same dimensions as the input.
22986 __kernel void rotate_image(__write_only image2d_t dst,
22987 unsigned int index,
22988 __read_only image2d_t src)
22990 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22991 CLK_FILTER_LINEAR);
22993 float angle = (float)index / 100.0f;
22995 float2 dst_dim = convert_float2(get_image_dim(dst));
22996 float2 src_dim = convert_float2(get_image_dim(src));
22998 float2 dst_cen = dst_dim / 2.0f;
22999 float2 src_cen = src_dim / 2.0f;
23001 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23003 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
23005 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
23006 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
23008 src_pos = src_pos * src_dim / dst_dim;
23010 float2 src_loc = src_pos + src_cen;
23012 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
23013 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
23014 write_imagef(dst, dst_loc, 0.5f);
23016 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
23021 Blend two inputs together, with the amount of each input used varying
23022 with the index counter.
23024 __kernel void blend_images(__write_only image2d_t dst,
23025 unsigned int index,
23026 __read_only image2d_t src1,
23027 __read_only image2d_t src2)
23029 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23030 CLK_FILTER_LINEAR);
23032 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
23034 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
23035 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
23036 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
23038 float4 val1 = read_imagef(src1, sampler, src1_loc);
23039 float4 val2 = read_imagef(src2, sampler, src2_loc);
23041 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
23047 @section roberts_opencl
23048 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
23050 The filter accepts the following option:
23054 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23057 Set value which will be multiplied with filtered result.
23058 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23061 Set value which will be added to filtered result.
23062 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23065 @subsection Example
23069 Apply the Roberts cross operator with scale set to 2 and delta set to 10
23071 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
23075 @section sobel_opencl
23077 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
23079 The filter accepts the following option:
23083 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23086 Set value which will be multiplied with filtered result.
23087 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23090 Set value which will be added to filtered result.
23091 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23094 @subsection Example
23098 Apply sobel operator with scale set to 2 and delta set to 10
23100 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
23104 @section tonemap_opencl
23106 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
23108 It accepts the following parameters:
23112 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
23115 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
23118 Apply desaturation for highlights that exceed this level of brightness. The
23119 higher the parameter, the more color information will be preserved. This
23120 setting helps prevent unnaturally blown-out colors for super-highlights, by
23121 (smoothly) turning into white instead. This makes images feel more natural,
23122 at the cost of reducing information about out-of-range colors.
23124 The default value is 0.5, and the algorithm here is a little different from
23125 the cpu version tonemap currently. A setting of 0.0 disables this option.
23128 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
23129 is used to detect whether the scene has changed or not. If the distance between
23130 the current frame average brightness and the current running average exceeds
23131 a threshold value, we would re-calculate scene average and peak brightness.
23132 The default value is 0.2.
23135 Specify the output pixel format.
23137 Currently supported formats are:
23144 Set the output color range.
23146 Possible values are:
23152 Default is same as input.
23155 Set the output color primaries.
23157 Possible values are:
23163 Default is same as input.
23166 Set the output transfer characteristics.
23168 Possible values are:
23177 Set the output colorspace matrix.
23179 Possible value are:
23185 Default is same as input.
23189 @subsection Example
23193 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
23195 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
23199 @section unsharp_opencl
23201 Sharpen or blur the input video.
23203 It accepts the following parameters:
23206 @item luma_msize_x, lx
23207 Set the luma matrix horizontal size.
23208 Range is @code{[1, 23]} and default value is @code{5}.
23210 @item luma_msize_y, ly
23211 Set the luma matrix vertical size.
23212 Range is @code{[1, 23]} and default value is @code{5}.
23214 @item luma_amount, la
23215 Set the luma effect strength.
23216 Range is @code{[-10, 10]} and default value is @code{1.0}.
23218 Negative values will blur the input video, while positive values will
23219 sharpen it, a value of zero will disable the effect.
23221 @item chroma_msize_x, cx
23222 Set the chroma matrix horizontal size.
23223 Range is @code{[1, 23]} and default value is @code{5}.
23225 @item chroma_msize_y, cy
23226 Set the chroma matrix vertical size.
23227 Range is @code{[1, 23]} and default value is @code{5}.
23229 @item chroma_amount, ca
23230 Set the chroma effect strength.
23231 Range is @code{[-10, 10]} and default value is @code{0.0}.
23233 Negative values will blur the input video, while positive values will
23234 sharpen it, a value of zero will disable the effect.
23238 All parameters are optional and default to the equivalent of the
23239 string '5:5:1.0:5:5:0.0'.
23241 @subsection Examples
23245 Apply strong luma sharpen effect:
23247 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
23251 Apply a strong blur of both luma and chroma parameters:
23253 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
23257 @section xfade_opencl
23259 Cross fade two videos with custom transition effect by using OpenCL.
23261 It accepts the following options:
23265 Set one of possible transition effects.
23269 Select custom transition effect, the actual transition description
23270 will be picked from source and kernel options.
23282 Default transition is fade.
23286 OpenCL program source file for custom transition.
23289 Set name of kernel to use for custom transition from program source file.
23292 Set duration of video transition.
23295 Set time of start of transition relative to first video.
23298 The program source file must contain a kernel function with the given name,
23299 which will be run once for each plane of the output. Each run on a plane
23300 gets enqueued as a separate 2D global NDRange with one work-item for each
23301 pixel to be generated. The global ID offset for each work-item is therefore
23302 the coordinates of a pixel in the destination image.
23304 The kernel function needs to take the following arguments:
23307 Destination image, @var{__write_only image2d_t}.
23309 This image will become the output; the kernel should write all of it.
23312 First Source image, @var{__read_only image2d_t}.
23313 Second Source image, @var{__read_only image2d_t}.
23315 These are the most recent images on each input. The kernel may read from
23316 them to generate the output, but they can't be written to.
23319 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
23326 Apply dots curtain transition effect:
23328 __kernel void blend_images(__write_only image2d_t dst,
23329 __read_only image2d_t src1,
23330 __read_only image2d_t src2,
23333 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23334 CLK_FILTER_LINEAR);
23335 int2 p = (int2)(get_global_id(0), get_global_id(1));
23336 float2 rp = (float2)(get_global_id(0), get_global_id(1));
23337 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
23340 float2 dots = (float2)(20.0, 20.0);
23341 float2 center = (float2)(0,0);
23344 float4 val1 = read_imagef(src1, sampler, p);
23345 float4 val2 = read_imagef(src2, sampler, p);
23346 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
23348 write_imagef(dst, p, next ? val1 : val2);
23354 @c man end OPENCL VIDEO FILTERS
23356 @chapter VAAPI Video Filters
23357 @c man begin VAAPI VIDEO FILTERS
23359 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
23361 To enable compilation of these filters you need to configure FFmpeg with
23362 @code{--enable-vaapi}.
23364 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}
23366 @section tonemap_vaapi
23368 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
23369 It maps the dynamic range of HDR10 content to the SDR content.
23370 It currently only accepts HDR10 as input.
23372 It accepts the following parameters:
23376 Specify the output pixel format.
23378 Currently supported formats are:
23387 Set the output color primaries.
23389 Default is same as input.
23392 Set the output transfer characteristics.
23397 Set the output colorspace matrix.
23399 Default is same as input.
23403 @subsection Example
23407 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
23409 tonemap_vaapi=format=p010:t=bt2020-10
23413 @c man end VAAPI VIDEO FILTERS
23415 @chapter Video Sources
23416 @c man begin VIDEO SOURCES
23418 Below is a description of the currently available video sources.
23422 Buffer video frames, and make them available to the filter chain.
23424 This source is mainly intended for a programmatic use, in particular
23425 through the interface defined in @file{libavfilter/buffersrc.h}.
23427 It accepts the following parameters:
23432 Specify the size (width and height) of the buffered video frames. For the
23433 syntax of this option, check the
23434 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23437 The input video width.
23440 The input video height.
23443 A string representing the pixel format of the buffered video frames.
23444 It may be a number corresponding to a pixel format, or a pixel format
23448 Specify the timebase assumed by the timestamps of the buffered frames.
23451 Specify the frame rate expected for the video stream.
23453 @item pixel_aspect, sar
23454 The sample (pixel) aspect ratio of the input video.
23457 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
23458 to the filtergraph description to specify swscale flags for automatically
23459 inserted scalers. See @ref{Filtergraph syntax}.
23461 @item hw_frames_ctx
23462 When using a hardware pixel format, this should be a reference to an
23463 AVHWFramesContext describing input frames.
23468 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
23471 will instruct the source to accept video frames with size 320x240 and
23472 with format "yuv410p", assuming 1/24 as the timestamps timebase and
23473 square pixels (1:1 sample aspect ratio).
23474 Since the pixel format with name "yuv410p" corresponds to the number 6
23475 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
23476 this example corresponds to:
23478 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
23481 Alternatively, the options can be specified as a flat string, but this
23482 syntax is deprecated:
23484 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
23488 Create a pattern generated by an elementary cellular automaton.
23490 The initial state of the cellular automaton can be defined through the
23491 @option{filename} and @option{pattern} options. If such options are
23492 not specified an initial state is created randomly.
23494 At each new frame a new row in the video is filled with the result of
23495 the cellular automaton next generation. The behavior when the whole
23496 frame is filled is defined by the @option{scroll} option.
23498 This source accepts the following options:
23502 Read the initial cellular automaton state, i.e. the starting row, from
23503 the specified file.
23504 In the file, each non-whitespace character is considered an alive
23505 cell, a newline will terminate the row, and further characters in the
23506 file will be ignored.
23509 Read the initial cellular automaton state, i.e. the starting row, from
23510 the specified string.
23512 Each non-whitespace character in the string is considered an alive
23513 cell, a newline will terminate the row, and further characters in the
23514 string will be ignored.
23517 Set the video rate, that is the number of frames generated per second.
23520 @item random_fill_ratio, ratio
23521 Set the random fill ratio for the initial cellular automaton row. It
23522 is a floating point number value ranging from 0 to 1, defaults to
23525 This option is ignored when a file or a pattern is specified.
23527 @item random_seed, seed
23528 Set the seed for filling randomly the initial row, must be an integer
23529 included between 0 and UINT32_MAX. If not specified, or if explicitly
23530 set to -1, the filter will try to use a good random seed on a best
23534 Set the cellular automaton rule, it is a number ranging from 0 to 255.
23535 Default value is 110.
23538 Set the size of the output video. For the syntax of this option, check the
23539 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23541 If @option{filename} or @option{pattern} is specified, the size is set
23542 by default to the width of the specified initial state row, and the
23543 height is set to @var{width} * PHI.
23545 If @option{size} is set, it must contain the width of the specified
23546 pattern string, and the specified pattern will be centered in the
23549 If a filename or a pattern string is not specified, the size value
23550 defaults to "320x518" (used for a randomly generated initial state).
23553 If set to 1, scroll the output upward when all the rows in the output
23554 have been already filled. If set to 0, the new generated row will be
23555 written over the top row just after the bottom row is filled.
23558 @item start_full, full
23559 If set to 1, completely fill the output with generated rows before
23560 outputting the first frame.
23561 This is the default behavior, for disabling set the value to 0.
23564 If set to 1, stitch the left and right row edges together.
23565 This is the default behavior, for disabling set the value to 0.
23568 @subsection Examples
23572 Read the initial state from @file{pattern}, and specify an output of
23575 cellauto=f=pattern:s=200x400
23579 Generate a random initial row with a width of 200 cells, with a fill
23582 cellauto=ratio=2/3:s=200x200
23586 Create a pattern generated by rule 18 starting by a single alive cell
23587 centered on an initial row with width 100:
23589 cellauto=p=@@:s=100x400:full=0:rule=18
23593 Specify a more elaborated initial pattern:
23595 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
23600 @anchor{coreimagesrc}
23601 @section coreimagesrc
23602 Video source generated on GPU using Apple's CoreImage API on OSX.
23604 This video source is a specialized version of the @ref{coreimage} video filter.
23605 Use a core image generator at the beginning of the applied filterchain to
23606 generate the content.
23608 The coreimagesrc video source accepts the following options:
23610 @item list_generators
23611 List all available generators along with all their respective options as well as
23612 possible minimum and maximum values along with the default values.
23614 list_generators=true
23618 Specify the size of the sourced video. For the syntax of this option, check the
23619 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23620 The default value is @code{320x240}.
23623 Specify the frame rate of the sourced video, as the number of frames
23624 generated per second. It has to be a string in the format
23625 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23626 number or a valid video frame rate abbreviation. The default value is
23630 Set the sample aspect ratio of the sourced video.
23633 Set the duration of the sourced video. See
23634 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23635 for the accepted syntax.
23637 If not specified, or the expressed duration is negative, the video is
23638 supposed to be generated forever.
23641 Additionally, all options of the @ref{coreimage} video filter are accepted.
23642 A complete filterchain can be used for further processing of the
23643 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
23644 and examples for details.
23646 @subsection Examples
23651 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
23652 given as complete and escaped command-line for Apple's standard bash shell:
23654 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
23656 This example is equivalent to the QRCode example of @ref{coreimage} without the
23657 need for a nullsrc video source.
23662 Generate several gradients.
23666 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23667 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23670 Set frame rate, expressed as number of frames per second. Default
23673 @item c0, c1, c2, c3, c4, c5, c6, c7
23674 Set 8 colors. Default values for colors is to pick random one.
23676 @item x0, y0, y0, y1
23677 Set gradient line source and destination points. If negative or out of range, random ones
23681 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
23684 Set seed for picking gradient line points.
23687 Set the duration of the sourced video. See
23688 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23689 for the accepted syntax.
23691 If not specified, or the expressed duration is negative, the video is
23692 supposed to be generated forever.
23695 Set speed of gradients rotation.
23699 @section mandelbrot
23701 Generate a Mandelbrot set fractal, and progressively zoom towards the
23702 point specified with @var{start_x} and @var{start_y}.
23704 This source accepts the following options:
23709 Set the terminal pts value. Default value is 400.
23712 Set the terminal scale value.
23713 Must be a floating point value. Default value is 0.3.
23716 Set the inner coloring mode, that is the algorithm used to draw the
23717 Mandelbrot fractal internal region.
23719 It shall assume one of the following values:
23724 Show time until convergence.
23726 Set color based on point closest to the origin of the iterations.
23731 Default value is @var{mincol}.
23734 Set the bailout value. Default value is 10.0.
23737 Set the maximum of iterations performed by the rendering
23738 algorithm. Default value is 7189.
23741 Set outer coloring mode.
23742 It shall assume one of following values:
23744 @item iteration_count
23745 Set iteration count mode.
23746 @item normalized_iteration_count
23747 set normalized iteration count mode.
23749 Default value is @var{normalized_iteration_count}.
23752 Set frame rate, expressed as number of frames per second. Default
23756 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23757 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23760 Set the initial scale value. Default value is 3.0.
23763 Set the initial x position. Must be a floating point value between
23764 -100 and 100. Default value is -0.743643887037158704752191506114774.
23767 Set the initial y position. Must be a floating point value between
23768 -100 and 100. Default value is -0.131825904205311970493132056385139.
23773 Generate various test patterns, as generated by the MPlayer test filter.
23775 The size of the generated video is fixed, and is 256x256.
23776 This source is useful in particular for testing encoding features.
23778 This source accepts the following options:
23783 Specify the frame rate of the sourced video, as the number of frames
23784 generated per second. It has to be a string in the format
23785 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23786 number or a valid video frame rate abbreviation. The default value is
23790 Set the duration of the sourced video. See
23791 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23792 for the accepted syntax.
23794 If not specified, or the expressed duration is negative, the video is
23795 supposed to be generated forever.
23799 Set the number or the name of the test to perform. Supported tests are:
23813 @item max_frames, m
23814 Set the maximum number of frames generated for each test, default value is 30.
23818 Default value is "all", which will cycle through the list of all tests.
23823 mptestsrc=t=dc_luma
23826 will generate a "dc_luma" test pattern.
23828 @section frei0r_src
23830 Provide a frei0r source.
23832 To enable compilation of this filter you need to install the frei0r
23833 header and configure FFmpeg with @code{--enable-frei0r}.
23835 This source accepts the following parameters:
23840 The size of the video to generate. For the syntax of this option, check the
23841 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23844 The framerate of the generated video. It may be a string of the form
23845 @var{num}/@var{den} or a frame rate abbreviation.
23848 The name to the frei0r source to load. For more information regarding frei0r and
23849 how to set the parameters, read the @ref{frei0r} section in the video filters
23852 @item filter_params
23853 A '|'-separated list of parameters to pass to the frei0r source.
23857 For example, to generate a frei0r partik0l source with size 200x200
23858 and frame rate 10 which is overlaid on the overlay filter main input:
23860 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
23865 Generate a life pattern.
23867 This source is based on a generalization of John Conway's life game.
23869 The sourced input represents a life grid, each pixel represents a cell
23870 which can be in one of two possible states, alive or dead. Every cell
23871 interacts with its eight neighbours, which are the cells that are
23872 horizontally, vertically, or diagonally adjacent.
23874 At each interaction the grid evolves according to the adopted rule,
23875 which specifies the number of neighbor alive cells which will make a
23876 cell stay alive or born. The @option{rule} option allows one to specify
23879 This source accepts the following options:
23883 Set the file from which to read the initial grid state. In the file,
23884 each non-whitespace character is considered an alive cell, and newline
23885 is used to delimit the end of each row.
23887 If this option is not specified, the initial grid is generated
23891 Set the video rate, that is the number of frames generated per second.
23894 @item random_fill_ratio, ratio
23895 Set the random fill ratio for the initial random grid. It is a
23896 floating point number value ranging from 0 to 1, defaults to 1/PHI.
23897 It is ignored when a file is specified.
23899 @item random_seed, seed
23900 Set the seed for filling the initial random grid, must be an integer
23901 included between 0 and UINT32_MAX. If not specified, or if explicitly
23902 set to -1, the filter will try to use a good random seed on a best
23908 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
23909 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
23910 @var{NS} specifies the number of alive neighbor cells which make a
23911 live cell stay alive, and @var{NB} the number of alive neighbor cells
23912 which make a dead cell to become alive (i.e. to "born").
23913 "s" and "b" can be used in place of "S" and "B", respectively.
23915 Alternatively a rule can be specified by an 18-bits integer. The 9
23916 high order bits are used to encode the next cell state if it is alive
23917 for each number of neighbor alive cells, the low order bits specify
23918 the rule for "borning" new cells. Higher order bits encode for an
23919 higher number of neighbor cells.
23920 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
23921 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
23923 Default value is "S23/B3", which is the original Conway's game of life
23924 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
23925 cells, and will born a new cell if there are three alive cells around
23929 Set the size of the output video. For the syntax of this option, check the
23930 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23932 If @option{filename} is specified, the size is set by default to the
23933 same size of the input file. If @option{size} is set, it must contain
23934 the size specified in the input file, and the initial grid defined in
23935 that file is centered in the larger resulting area.
23937 If a filename is not specified, the size value defaults to "320x240"
23938 (used for a randomly generated initial grid).
23941 If set to 1, stitch the left and right grid edges together, and the
23942 top and bottom edges also. Defaults to 1.
23945 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
23946 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
23947 value from 0 to 255.
23950 Set the color of living (or new born) cells.
23953 Set the color of dead cells. If @option{mold} is set, this is the first color
23954 used to represent a dead cell.
23957 Set mold color, for definitely dead and moldy cells.
23959 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
23960 ffmpeg-utils manual,ffmpeg-utils}.
23963 @subsection Examples
23967 Read a grid from @file{pattern}, and center it on a grid of size
23970 life=f=pattern:s=300x300
23974 Generate a random grid of size 200x200, with a fill ratio of 2/3:
23976 life=ratio=2/3:s=200x200
23980 Specify a custom rule for evolving a randomly generated grid:
23986 Full example with slow death effect (mold) using @command{ffplay}:
23988 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
23995 @anchor{haldclutsrc}
23998 @anchor{pal100bars}
23999 @anchor{rgbtestsrc}
24001 @anchor{smptehdbars}
24004 @anchor{yuvtestsrc}
24005 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
24007 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
24009 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
24011 The @code{color} source provides an uniformly colored input.
24013 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
24014 @ref{haldclut} filter.
24016 The @code{nullsrc} source returns unprocessed video frames. It is
24017 mainly useful to be employed in analysis / debugging tools, or as the
24018 source for filters which ignore the input data.
24020 The @code{pal75bars} source generates a color bars pattern, based on
24021 EBU PAL recommendations with 75% color levels.
24023 The @code{pal100bars} source generates a color bars pattern, based on
24024 EBU PAL recommendations with 100% color levels.
24026 The @code{rgbtestsrc} source generates an RGB test pattern useful for
24027 detecting RGB vs BGR issues. You should see a red, green and blue
24028 stripe from top to bottom.
24030 The @code{smptebars} source generates a color bars pattern, based on
24031 the SMPTE Engineering Guideline EG 1-1990.
24033 The @code{smptehdbars} source generates a color bars pattern, based on
24034 the SMPTE RP 219-2002.
24036 The @code{testsrc} source generates a test video pattern, showing a
24037 color pattern, a scrolling gradient and a timestamp. This is mainly
24038 intended for testing purposes.
24040 The @code{testsrc2} source is similar to testsrc, but supports more
24041 pixel formats instead of just @code{rgb24}. This allows using it as an
24042 input for other tests without requiring a format conversion.
24044 The @code{yuvtestsrc} source generates an YUV test pattern. You should
24045 see a y, cb and cr stripe from top to bottom.
24047 The sources accept the following parameters:
24052 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
24053 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
24054 pixels to be used as identity matrix for 3D lookup tables. Each component is
24055 coded on a @code{1/(N*N)} scale.
24058 Specify the color of the source, only available in the @code{color}
24059 source. For the syntax of this option, check the
24060 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
24063 Specify the size of the sourced video. For the syntax of this option, check the
24064 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24065 The default value is @code{320x240}.
24067 This option is not available with the @code{allrgb}, @code{allyuv}, and
24068 @code{haldclutsrc} filters.
24071 Specify the frame rate of the sourced video, as the number of frames
24072 generated per second. It has to be a string in the format
24073 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
24074 number or a valid video frame rate abbreviation. The default value is
24078 Set the duration of the sourced video. See
24079 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
24080 for the accepted syntax.
24082 If not specified, or the expressed duration is negative, the video is
24083 supposed to be generated forever.
24085 Since the frame rate is used as time base, all frames including the last one
24086 will have their full duration. If the specified duration is not a multiple
24087 of the frame duration, it will be rounded up.
24090 Set the sample aspect ratio of the sourced video.
24093 Specify the alpha (opacity) of the background, only available in the
24094 @code{testsrc2} source. The value must be between 0 (fully transparent) and
24095 255 (fully opaque, the default).
24098 Set the number of decimals to show in the timestamp, only available in the
24099 @code{testsrc} source.
24101 The displayed timestamp value will correspond to the original
24102 timestamp value multiplied by the power of 10 of the specified
24103 value. Default value is 0.
24106 @subsection Examples
24110 Generate a video with a duration of 5.3 seconds, with size
24111 176x144 and a frame rate of 10 frames per second:
24113 testsrc=duration=5.3:size=qcif:rate=10
24117 The following graph description will generate a red source
24118 with an opacity of 0.2, with size "qcif" and a frame rate of 10
24121 color=c=red@@0.2:s=qcif:r=10
24125 If the input content is to be ignored, @code{nullsrc} can be used. The
24126 following command generates noise in the luminance plane by employing
24127 the @code{geq} filter:
24129 nullsrc=s=256x256, geq=random(1)*255:128:128
24133 @subsection Commands
24135 The @code{color} source supports the following commands:
24139 Set the color of the created image. Accepts the same syntax of the
24140 corresponding @option{color} option.
24145 Generate video using an OpenCL program.
24150 OpenCL program source file.
24153 Kernel name in program.
24156 Size of frames to generate. This must be set.
24159 Pixel format to use for the generated frames. This must be set.
24162 Number of frames generated every second. Default value is '25'.
24166 For details of how the program loading works, see the @ref{program_opencl}
24173 Generate a colour ramp by setting pixel values from the position of the pixel
24174 in the output image. (Note that this will work with all pixel formats, but
24175 the generated output will not be the same.)
24177 __kernel void ramp(__write_only image2d_t dst,
24178 unsigned int index)
24180 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24183 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
24185 write_imagef(dst, loc, val);
24190 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
24192 __kernel void sierpinski_carpet(__write_only image2d_t dst,
24193 unsigned int index)
24195 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24197 float4 value = 0.0f;
24198 int x = loc.x + index;
24199 int y = loc.y + index;
24200 while (x > 0 || y > 0) {
24201 if (x % 3 == 1 && y % 3 == 1) {
24209 write_imagef(dst, loc, value);
24215 @section sierpinski
24217 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
24219 This source accepts the following options:
24223 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
24224 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
24227 Set frame rate, expressed as number of frames per second. Default
24231 Set seed which is used for random panning.
24234 Set max jump for single pan destination. Allowed range is from 1 to 10000.
24237 Set fractal type, can be default @code{carpet} or @code{triangle}.
24240 @c man end VIDEO SOURCES
24242 @chapter Video Sinks
24243 @c man begin VIDEO SINKS
24245 Below is a description of the currently available video sinks.
24247 @section buffersink
24249 Buffer video frames, and make them available to the end of the filter
24252 This sink is mainly intended for programmatic use, in particular
24253 through the interface defined in @file{libavfilter/buffersink.h}
24254 or the options system.
24256 It accepts a pointer to an AVBufferSinkContext structure, which
24257 defines the incoming buffers' formats, to be passed as the opaque
24258 parameter to @code{avfilter_init_filter} for initialization.
24262 Null video sink: do absolutely nothing with the input video. It is
24263 mainly useful as a template and for use in analysis / debugging
24266 @c man end VIDEO SINKS
24268 @chapter Multimedia Filters
24269 @c man begin MULTIMEDIA FILTERS
24271 Below is a description of the currently available multimedia filters.
24275 Convert input audio to a video output, displaying the audio bit scope.
24277 The filter accepts the following options:
24281 Set frame rate, expressed as number of frames per second. Default
24285 Specify the video size for the output. For the syntax of this option, check the
24286 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24287 Default value is @code{1024x256}.
24290 Specify list of colors separated by space or by '|' which will be used to
24291 draw channels. Unrecognized or missing colors will be replaced
24295 @section adrawgraph
24296 Draw a graph using input audio metadata.
24298 See @ref{drawgraph}
24300 @section agraphmonitor
24302 See @ref{graphmonitor}.
24304 @section ahistogram
24306 Convert input audio to a video output, displaying the volume histogram.
24308 The filter accepts the following options:
24312 Specify how histogram is calculated.
24314 It accepts the following values:
24317 Use single histogram for all channels.
24319 Use separate histogram for each channel.
24321 Default is @code{single}.
24324 Set frame rate, expressed as number of frames per second. Default
24328 Specify the video size for the output. For the syntax of this option, check the
24329 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24330 Default value is @code{hd720}.
24335 It accepts the following values:
24346 reverse logarithmic
24348 Default is @code{log}.
24351 Set amplitude scale.
24353 It accepts the following values:
24360 Default is @code{log}.
24363 Set how much frames to accumulate in histogram.
24364 Default is 1. Setting this to -1 accumulates all frames.
24367 Set histogram ratio of window height.
24370 Set sonogram sliding.
24372 It accepts the following values:
24375 replace old rows with new ones.
24377 scroll from top to bottom.
24379 Default is @code{replace}.
24382 @section aphasemeter
24384 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
24385 representing mean phase of current audio frame. A video output can also be produced and is
24386 enabled by default. The audio is passed through as first output.
24388 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
24389 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
24390 and @code{1} means channels are in phase.
24392 The filter accepts the following options, all related to its video output:
24396 Set the output frame rate. Default value is @code{25}.
24399 Set the video size for the output. For the syntax of this option, check the
24400 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24401 Default value is @code{800x400}.
24406 Specify the red, green, blue contrast. Default values are @code{2},
24407 @code{7} and @code{1}.
24408 Allowed range is @code{[0, 255]}.
24411 Set color which will be used for drawing median phase. If color is
24412 @code{none} which is default, no median phase value will be drawn.
24415 Enable video output. Default is enabled.
24418 @subsection phasing detection
24420 The filter also detects out of phase and mono sequences in stereo streams.
24421 It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
24423 The filter accepts the following options for this detection:
24427 Enable mono and out of phase detection. Default is disabled.
24430 Set phase tolerance for mono detection, in amplitude ratio. Default is @code{0}.
24431 Allowed range is @code{[0, 1]}.
24434 Set angle threshold for out of phase detection, in degree. Default is @code{170}.
24435 Allowed range is @code{[90, 180]}.
24438 Set mono or out of phase duration until notification, expressed in seconds. Default is @code{2}.
24441 @subsection Examples
24445 Complete example with @command{ffmpeg} to detect 1 second of mono with 0.001 phase tolerance:
24447 ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
24451 @section avectorscope
24453 Convert input audio to a video output, representing the audio vector
24456 The filter is used to measure the difference between channels of stereo
24457 audio stream. A monaural signal, consisting of identical left and right
24458 signal, results in straight vertical line. Any stereo separation is visible
24459 as a deviation from this line, creating a Lissajous figure.
24460 If the straight (or deviation from it) but horizontal line appears this
24461 indicates that the left and right channels are out of phase.
24463 The filter accepts the following options:
24467 Set the vectorscope mode.
24469 Available values are:
24472 Lissajous rotated by 45 degrees.
24475 Same as above but not rotated.
24478 Shape resembling half of circle.
24481 Default value is @samp{lissajous}.
24484 Set the video size for the output. For the syntax of this option, check the
24485 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24486 Default value is @code{400x400}.
24489 Set the output frame rate. Default value is @code{25}.
24495 Specify the red, green, blue and alpha contrast. Default values are @code{40},
24496 @code{160}, @code{80} and @code{255}.
24497 Allowed range is @code{[0, 255]}.
24503 Specify the red, green, blue and alpha fade. Default values are @code{15},
24504 @code{10}, @code{5} and @code{5}.
24505 Allowed range is @code{[0, 255]}.
24508 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
24509 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
24512 Set the vectorscope drawing mode.
24514 Available values are:
24517 Draw dot for each sample.
24520 Draw line between previous and current sample.
24523 Default value is @samp{dot}.
24526 Specify amplitude scale of audio samples.
24528 Available values are:
24544 Swap left channel axis with right channel axis.
24554 Mirror only x axis.
24557 Mirror only y axis.
24565 @subsection Examples
24569 Complete example using @command{ffplay}:
24571 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
24572 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
24576 @section bench, abench
24578 Benchmark part of a filtergraph.
24580 The filter accepts the following options:
24584 Start or stop a timer.
24586 Available values are:
24589 Get the current time, set it as frame metadata (using the key
24590 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
24593 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
24594 the input frame metadata to get the time difference. Time difference, average,
24595 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
24596 @code{min}) are then printed. The timestamps are expressed in seconds.
24600 @subsection Examples
24604 Benchmark @ref{selectivecolor} filter:
24606 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
24612 Concatenate audio and video streams, joining them together one after the
24615 The filter works on segments of synchronized video and audio streams. All
24616 segments must have the same number of streams of each type, and that will
24617 also be the number of streams at output.
24619 The filter accepts the following options:
24624 Set the number of segments. Default is 2.
24627 Set the number of output video streams, that is also the number of video
24628 streams in each segment. Default is 1.
24631 Set the number of output audio streams, that is also the number of audio
24632 streams in each segment. Default is 0.
24635 Activate unsafe mode: do not fail if segments have a different format.
24639 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
24640 @var{a} audio outputs.
24642 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
24643 segment, in the same order as the outputs, then the inputs for the second
24646 Related streams do not always have exactly the same duration, for various
24647 reasons including codec frame size or sloppy authoring. For that reason,
24648 related synchronized streams (e.g. a video and its audio track) should be
24649 concatenated at once. The concat filter will use the duration of the longest
24650 stream in each segment (except the last one), and if necessary pad shorter
24651 audio streams with silence.
24653 For this filter to work correctly, all segments must start at timestamp 0.
24655 All corresponding streams must have the same parameters in all segments; the
24656 filtering system will automatically select a common pixel format for video
24657 streams, and a common sample format, sample rate and channel layout for
24658 audio streams, but other settings, such as resolution, must be converted
24659 explicitly by the user.
24661 Different frame rates are acceptable but will result in variable frame rate
24662 at output; be sure to configure the output file to handle it.
24664 @subsection Examples
24668 Concatenate an opening, an episode and an ending, all in bilingual version
24669 (video in stream 0, audio in streams 1 and 2):
24671 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
24672 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
24673 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
24674 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
24678 Concatenate two parts, handling audio and video separately, using the
24679 (a)movie sources, and adjusting the resolution:
24681 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
24682 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
24683 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
24685 Note that a desync will happen at the stitch if the audio and video streams
24686 do not have exactly the same duration in the first file.
24690 @subsection Commands
24692 This filter supports the following commands:
24695 Close the current segment and step to the next one
24701 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
24702 level. By default, it logs a message at a frequency of 10Hz with the
24703 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
24704 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
24706 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
24707 sample format is double-precision floating point. The input stream will be converted to
24708 this specification, if needed. Users may need to insert aformat and/or aresample filters
24709 after this filter to obtain the original parameters.
24711 The filter also has a video output (see the @var{video} option) with a real
24712 time graph to observe the loudness evolution. The graphic contains the logged
24713 message mentioned above, so it is not printed anymore when this option is set,
24714 unless the verbose logging is set. The main graphing area contains the
24715 short-term loudness (3 seconds of analysis), and the gauge on the right is for
24716 the momentary loudness (400 milliseconds), but can optionally be configured
24717 to instead display short-term loudness (see @var{gauge}).
24719 The green area marks a +/- 1LU target range around the target loudness
24720 (-23LUFS by default, unless modified through @var{target}).
24722 More information about the Loudness Recommendation EBU R128 on
24723 @url{http://tech.ebu.ch/loudness}.
24725 The filter accepts the following options:
24730 Activate the video output. The audio stream is passed unchanged whether this
24731 option is set or no. The video stream will be the first output stream if
24732 activated. Default is @code{0}.
24735 Set the video size. This option is for video only. For the syntax of this
24737 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24738 Default and minimum resolution is @code{640x480}.
24741 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
24742 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
24743 other integer value between this range is allowed.
24746 Set metadata injection. If set to @code{1}, the audio input will be segmented
24747 into 100ms output frames, each of them containing various loudness information
24748 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
24750 Default is @code{0}.
24753 Force the frame logging level.
24755 Available values are:
24758 information logging level
24760 verbose logging level
24763 By default, the logging level is set to @var{info}. If the @option{video} or
24764 the @option{metadata} options are set, it switches to @var{verbose}.
24769 Available modes can be cumulated (the option is a @code{flag} type). Possible
24773 Disable any peak mode (default).
24775 Enable sample-peak mode.
24777 Simple peak mode looking for the higher sample value. It logs a message
24778 for sample-peak (identified by @code{SPK}).
24780 Enable true-peak mode.
24782 If enabled, the peak lookup is done on an over-sampled version of the input
24783 stream for better peak accuracy. It logs a message for true-peak.
24784 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
24785 This mode requires a build with @code{libswresample}.
24789 Treat mono input files as "dual mono". If a mono file is intended for playback
24790 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
24791 If set to @code{true}, this option will compensate for this effect.
24792 Multi-channel input files are not affected by this option.
24795 Set a specific pan law to be used for the measurement of dual mono files.
24796 This parameter is optional, and has a default value of -3.01dB.
24799 Set a specific target level (in LUFS) used as relative zero in the visualization.
24800 This parameter is optional and has a default value of -23LUFS as specified
24801 by EBU R128. However, material published online may prefer a level of -16LUFS
24802 (e.g. for use with podcasts or video platforms).
24805 Set the value displayed by the gauge. Valid values are @code{momentary} and s
24806 @code{shortterm}. By default the momentary value will be used, but in certain
24807 scenarios it may be more useful to observe the short term value instead (e.g.
24811 Sets the display scale for the loudness. Valid parameters are @code{absolute}
24812 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
24813 video output, not the summary or continuous log output.
24816 @subsection Examples
24820 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
24822 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
24826 Run an analysis with @command{ffmpeg}:
24828 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
24832 @section interleave, ainterleave
24834 Temporally interleave frames from several inputs.
24836 @code{interleave} works with video inputs, @code{ainterleave} with audio.
24838 These filters read frames from several inputs and send the oldest
24839 queued frame to the output.
24841 Input streams must have well defined, monotonically increasing frame
24844 In order to submit one frame to output, these filters need to enqueue
24845 at least one frame for each input, so they cannot work in case one
24846 input is not yet terminated and will not receive incoming frames.
24848 For example consider the case when one input is a @code{select} filter
24849 which always drops input frames. The @code{interleave} filter will keep
24850 reading from that input, but it will never be able to send new frames
24851 to output until the input sends an end-of-stream signal.
24853 Also, depending on inputs synchronization, the filters will drop
24854 frames in case one input receives more frames than the other ones, and
24855 the queue is already filled.
24857 These filters accept the following options:
24861 Set the number of different inputs, it is 2 by default.
24864 How to determine the end-of-stream.
24868 The duration of the longest input. (default)
24871 The duration of the shortest input.
24874 The duration of the first input.
24879 @subsection Examples
24883 Interleave frames belonging to different streams using @command{ffmpeg}:
24885 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
24889 Add flickering blur effect:
24891 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
24895 @section metadata, ametadata
24897 Manipulate frame metadata.
24899 This filter accepts the following options:
24903 Set mode of operation of the filter.
24905 Can be one of the following:
24909 If both @code{value} and @code{key} is set, select frames
24910 which have such metadata. If only @code{key} is set, select
24911 every frame that has such key in metadata.
24914 Add new metadata @code{key} and @code{value}. If key is already available
24918 Modify value of already present key.
24921 If @code{value} is set, delete only keys that have such value.
24922 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
24926 Print key and its value if metadata was found. If @code{key} is not set print all
24927 metadata values available in frame.
24931 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
24934 Set metadata value which will be used. This option is mandatory for
24935 @code{modify} and @code{add} mode.
24938 Which function to use when comparing metadata value and @code{value}.
24940 Can be one of following:
24944 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
24947 Values are interpreted as strings, returns true if metadata value starts with
24948 the @code{value} option string.
24951 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
24954 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
24957 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
24960 Values are interpreted as floats, returns true if expression from option @code{expr}
24964 Values are interpreted as strings, returns true if metadata value ends with
24965 the @code{value} option string.
24969 Set expression which is used when @code{function} is set to @code{expr}.
24970 The expression is evaluated through the eval API and can contain the following
24975 Float representation of @code{value} from metadata key.
24978 Float representation of @code{value} as supplied by user in @code{value} option.
24982 If specified in @code{print} mode, output is written to the named file. Instead of
24983 plain filename any writable url can be specified. Filename ``-'' is a shorthand
24984 for standard output. If @code{file} option is not set, output is written to the log
24985 with AV_LOG_INFO loglevel.
24988 Reduces buffering in print mode when output is written to a URL set using @var{file}.
24992 @subsection Examples
24996 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
24999 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
25002 Print silencedetect output to file @file{metadata.txt}.
25004 silencedetect,ametadata=mode=print:file=metadata.txt
25007 Direct all metadata to a pipe with file descriptor 4.
25009 metadata=mode=print:file='pipe\:4'
25013 @section perms, aperms
25015 Set read/write permissions for the output frames.
25017 These filters are mainly aimed at developers to test direct path in the
25018 following filter in the filtergraph.
25020 The filters accept the following options:
25024 Select the permissions mode.
25026 It accepts the following values:
25029 Do nothing. This is the default.
25031 Set all the output frames read-only.
25033 Set all the output frames directly writable.
25035 Make the frame read-only if writable, and writable if read-only.
25037 Set each output frame read-only or writable randomly.
25041 Set the seed for the @var{random} mode, must be an integer included between
25042 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
25043 @code{-1}, the filter will try to use a good random seed on a best effort
25047 Note: in case of auto-inserted filter between the permission filter and the
25048 following one, the permission might not be received as expected in that
25049 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
25050 perms/aperms filter can avoid this problem.
25052 @section realtime, arealtime
25054 Slow down filtering to match real time approximately.
25056 These filters will pause the filtering for a variable amount of time to
25057 match the output rate with the input timestamps.
25058 They are similar to the @option{re} option to @code{ffmpeg}.
25060 They accept the following options:
25064 Time limit for the pauses. Any pause longer than that will be considered
25065 a timestamp discontinuity and reset the timer. Default is 2 seconds.
25067 Speed factor for processing. The value must be a float larger than zero.
25068 Values larger than 1.0 will result in faster than realtime processing,
25069 smaller will slow processing down. The @var{limit} is automatically adapted
25070 accordingly. Default is 1.0.
25072 A processing speed faster than what is possible without these filters cannot
25077 @section select, aselect
25079 Select frames to pass in output.
25081 This filter accepts the following options:
25086 Set expression, which is evaluated for each input frame.
25088 If the expression is evaluated to zero, the frame is discarded.
25090 If the evaluation result is negative or NaN, the frame is sent to the
25091 first output; otherwise it is sent to the output with index
25092 @code{ceil(val)-1}, assuming that the input index starts from 0.
25094 For example a value of @code{1.2} corresponds to the output with index
25095 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
25098 Set the number of outputs. The output to which to send the selected
25099 frame is based on the result of the evaluation. Default value is 1.
25102 The expression can contain the following constants:
25106 The (sequential) number of the filtered frame, starting from 0.
25109 The (sequential) number of the selected frame, starting from 0.
25111 @item prev_selected_n
25112 The sequential number of the last selected frame. It's NAN if undefined.
25115 The timebase of the input timestamps.
25118 The PTS (Presentation TimeStamp) of the filtered video frame,
25119 expressed in @var{TB} units. It's NAN if undefined.
25122 The PTS of the filtered video frame,
25123 expressed in seconds. It's NAN if undefined.
25126 The PTS of the previously filtered video frame. It's NAN if undefined.
25128 @item prev_selected_pts
25129 The PTS of the last previously filtered video frame. It's NAN if undefined.
25131 @item prev_selected_t
25132 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
25135 The PTS of the first video frame in the video. It's NAN if undefined.
25138 The time of the first video frame in the video. It's NAN if undefined.
25140 @item pict_type @emph{(video only)}
25141 The type of the filtered frame. It can assume one of the following
25153 @item interlace_type @emph{(video only)}
25154 The frame interlace type. It can assume one of the following values:
25157 The frame is progressive (not interlaced).
25159 The frame is top-field-first.
25161 The frame is bottom-field-first.
25164 @item consumed_sample_n @emph{(audio only)}
25165 the number of selected samples before the current frame
25167 @item samples_n @emph{(audio only)}
25168 the number of samples in the current frame
25170 @item sample_rate @emph{(audio only)}
25171 the input sample rate
25174 This is 1 if the filtered frame is a key-frame, 0 otherwise.
25177 the position in the file of the filtered frame, -1 if the information
25178 is not available (e.g. for synthetic video)
25180 @item scene @emph{(video only)}
25181 value between 0 and 1 to indicate a new scene; a low value reflects a low
25182 probability for the current frame to introduce a new scene, while a higher
25183 value means the current frame is more likely to be one (see the example below)
25185 @item concatdec_select
25186 The concat demuxer can select only part of a concat input file by setting an
25187 inpoint and an outpoint, but the output packets may not be entirely contained
25188 in the selected interval. By using this variable, it is possible to skip frames
25189 generated by the concat demuxer which are not exactly contained in the selected
25192 This works by comparing the frame pts against the @var{lavf.concat.start_time}
25193 and the @var{lavf.concat.duration} packet metadata values which are also
25194 present in the decoded frames.
25196 The @var{concatdec_select} variable is -1 if the frame pts is at least
25197 start_time and either the duration metadata is missing or the frame pts is less
25198 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
25201 That basically means that an input frame is selected if its pts is within the
25202 interval set by the concat demuxer.
25206 The default value of the select expression is "1".
25208 @subsection Examples
25212 Select all frames in input:
25217 The example above is the same as:
25229 Select only I-frames:
25231 select='eq(pict_type\,I)'
25235 Select one frame every 100:
25237 select='not(mod(n\,100))'
25241 Select only frames contained in the 10-20 time interval:
25243 select=between(t\,10\,20)
25247 Select only I-frames contained in the 10-20 time interval:
25249 select=between(t\,10\,20)*eq(pict_type\,I)
25253 Select frames with a minimum distance of 10 seconds:
25255 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
25259 Use aselect to select only audio frames with samples number > 100:
25261 aselect='gt(samples_n\,100)'
25265 Create a mosaic of the first scenes:
25267 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
25270 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
25274 Send even and odd frames to separate outputs, and compose them:
25276 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
25280 Select useful frames from an ffconcat file which is using inpoints and
25281 outpoints but where the source files are not intra frame only.
25283 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
25287 @section sendcmd, asendcmd
25289 Send commands to filters in the filtergraph.
25291 These filters read commands to be sent to other filters in the
25294 @code{sendcmd} must be inserted between two video filters,
25295 @code{asendcmd} must be inserted between two audio filters, but apart
25296 from that they act the same way.
25298 The specification of commands can be provided in the filter arguments
25299 with the @var{commands} option, or in a file specified by the
25300 @var{filename} option.
25302 These filters accept the following options:
25305 Set the commands to be read and sent to the other filters.
25307 Set the filename of the commands to be read and sent to the other
25311 @subsection Commands syntax
25313 A commands description consists of a sequence of interval
25314 specifications, comprising a list of commands to be executed when a
25315 particular event related to that interval occurs. The occurring event
25316 is typically the current frame time entering or leaving a given time
25319 An interval is specified by the following syntax:
25321 @var{START}[-@var{END}] @var{COMMANDS};
25324 The time interval is specified by the @var{START} and @var{END} times.
25325 @var{END} is optional and defaults to the maximum time.
25327 The current frame time is considered within the specified interval if
25328 it is included in the interval [@var{START}, @var{END}), that is when
25329 the time is greater or equal to @var{START} and is lesser than
25332 @var{COMMANDS} consists of a sequence of one or more command
25333 specifications, separated by ",", relating to that interval. The
25334 syntax of a command specification is given by:
25336 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
25339 @var{FLAGS} is optional and specifies the type of events relating to
25340 the time interval which enable sending the specified command, and must
25341 be a non-null sequence of identifier flags separated by "+" or "|" and
25342 enclosed between "[" and "]".
25344 The following flags are recognized:
25347 The command is sent when the current frame timestamp enters the
25348 specified interval. In other words, the command is sent when the
25349 previous frame timestamp was not in the given interval, and the
25353 The command is sent when the current frame timestamp leaves the
25354 specified interval. In other words, the command is sent when the
25355 previous frame timestamp was in the given interval, and the
25359 The command @var{ARG} is interpreted as expression and result of
25360 expression is passed as @var{ARG}.
25362 The expression is evaluated through the eval API and can contain the following
25367 Original position in the file of the frame, or undefined if undefined
25368 for the current frame.
25371 The presentation timestamp in input.
25374 The count of the input frame for video or audio, starting from 0.
25377 The time in seconds of the current frame.
25380 The start time in seconds of the current command interval.
25383 The end time in seconds of the current command interval.
25386 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
25391 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
25394 @var{TARGET} specifies the target of the command, usually the name of
25395 the filter class or a specific filter instance name.
25397 @var{COMMAND} specifies the name of the command for the target filter.
25399 @var{ARG} is optional and specifies the optional list of argument for
25400 the given @var{COMMAND}.
25402 Between one interval specification and another, whitespaces, or
25403 sequences of characters starting with @code{#} until the end of line,
25404 are ignored and can be used to annotate comments.
25406 A simplified BNF description of the commands specification syntax
25409 @var{COMMAND_FLAG} ::= "enter" | "leave"
25410 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
25411 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
25412 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
25413 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
25414 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
25417 @subsection Examples
25421 Specify audio tempo change at second 4:
25423 asendcmd=c='4.0 atempo tempo 1.5',atempo
25427 Target a specific filter instance:
25429 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
25433 Specify a list of drawtext and hue commands in a file.
25435 # show text in the interval 5-10
25436 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
25437 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
25439 # desaturate the image in the interval 15-20
25440 15.0-20.0 [enter] hue s 0,
25441 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
25443 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
25445 # apply an exponential saturation fade-out effect, starting from time 25
25446 25 [enter] hue s exp(25-t)
25449 A filtergraph allowing to read and process the above command list
25450 stored in a file @file{test.cmd}, can be specified with:
25452 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
25457 @section setpts, asetpts
25459 Change the PTS (presentation timestamp) of the input frames.
25461 @code{setpts} works on video frames, @code{asetpts} on audio frames.
25463 This filter accepts the following options:
25468 The expression which is evaluated for each frame to construct its timestamp.
25472 The expression is evaluated through the eval API and can contain the following
25476 @item FRAME_RATE, FR
25477 frame rate, only defined for constant frame-rate video
25480 The presentation timestamp in input
25483 The count of the input frame for video or the number of consumed samples,
25484 not including the current frame for audio, starting from 0.
25486 @item NB_CONSUMED_SAMPLES
25487 The number of consumed samples, not including the current frame (only
25490 @item NB_SAMPLES, S
25491 The number of samples in the current frame (only audio)
25493 @item SAMPLE_RATE, SR
25494 The audio sample rate.
25497 The PTS of the first frame.
25500 the time in seconds of the first frame
25503 State whether the current frame is interlaced.
25506 the time in seconds of the current frame
25509 original position in the file of the frame, or undefined if undefined
25510 for the current frame
25513 The previous input PTS.
25516 previous input time in seconds
25519 The previous output PTS.
25522 previous output time in seconds
25525 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
25529 The wallclock (RTC) time at the start of the movie in microseconds.
25532 The timebase of the input timestamps.
25536 @subsection Examples
25540 Start counting PTS from zero
25542 setpts=PTS-STARTPTS
25546 Apply fast motion effect:
25552 Apply slow motion effect:
25558 Set fixed rate of 25 frames per second:
25564 Set fixed rate 25 fps with some jitter:
25566 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
25570 Apply an offset of 10 seconds to the input PTS:
25576 Generate timestamps from a "live source" and rebase onto the current timebase:
25578 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
25582 Generate timestamps by counting samples:
25591 Force color range for the output video frame.
25593 The @code{setrange} filter marks the color range property for the
25594 output frames. It does not change the input frame, but only sets the
25595 corresponding property, which affects how the frame is treated by
25598 The filter accepts the following options:
25603 Available values are:
25607 Keep the same color range property.
25609 @item unspecified, unknown
25610 Set the color range as unspecified.
25612 @item limited, tv, mpeg
25613 Set the color range as limited.
25615 @item full, pc, jpeg
25616 Set the color range as full.
25620 @section settb, asettb
25622 Set the timebase to use for the output frames timestamps.
25623 It is mainly useful for testing timebase configuration.
25625 It accepts the following parameters:
25630 The expression which is evaluated into the output timebase.
25634 The value for @option{tb} is an arithmetic expression representing a
25635 rational. The expression can contain the constants "AVTB" (the default
25636 timebase), "intb" (the input timebase) and "sr" (the sample rate,
25637 audio only). Default value is "intb".
25639 @subsection Examples
25643 Set the timebase to 1/25:
25649 Set the timebase to 1/10:
25655 Set the timebase to 1001/1000:
25661 Set the timebase to 2*intb:
25667 Set the default timebase value:
25674 Convert input audio to a video output representing frequency spectrum
25675 logarithmically using Brown-Puckette constant Q transform algorithm with
25676 direct frequency domain coefficient calculation (but the transform itself
25677 is not really constant Q, instead the Q factor is actually variable/clamped),
25678 with musical tone scale, from E0 to D#10.
25680 The filter accepts the following options:
25684 Specify the video size for the output. It must be even. For the syntax of this option,
25685 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25686 Default value is @code{1920x1080}.
25689 Set the output frame rate. Default value is @code{25}.
25692 Set the bargraph height. It must be even. Default value is @code{-1} which
25693 computes the bargraph height automatically.
25696 Set the axis height. It must be even. Default value is @code{-1} which computes
25697 the axis height automatically.
25700 Set the sonogram height. It must be even. Default value is @code{-1} which
25701 computes the sonogram height automatically.
25704 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
25705 instead. Default value is @code{1}.
25707 @item sono_v, volume
25708 Specify the sonogram volume expression. It can contain variables:
25711 the @var{bar_v} evaluated expression
25712 @item frequency, freq, f
25713 the frequency where it is evaluated
25714 @item timeclamp, tc
25715 the value of @var{timeclamp} option
25719 @item a_weighting(f)
25720 A-weighting of equal loudness
25721 @item b_weighting(f)
25722 B-weighting of equal loudness
25723 @item c_weighting(f)
25724 C-weighting of equal loudness.
25726 Default value is @code{16}.
25728 @item bar_v, volume2
25729 Specify the bargraph volume expression. It can contain variables:
25732 the @var{sono_v} evaluated expression
25733 @item frequency, freq, f
25734 the frequency where it is evaluated
25735 @item timeclamp, tc
25736 the value of @var{timeclamp} option
25740 @item a_weighting(f)
25741 A-weighting of equal loudness
25742 @item b_weighting(f)
25743 B-weighting of equal loudness
25744 @item c_weighting(f)
25745 C-weighting of equal loudness.
25747 Default value is @code{sono_v}.
25749 @item sono_g, gamma
25750 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
25751 higher gamma makes the spectrum having more range. Default value is @code{3}.
25752 Acceptable range is @code{[1, 7]}.
25754 @item bar_g, gamma2
25755 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
25759 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
25760 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
25762 @item timeclamp, tc
25763 Specify the transform timeclamp. At low frequency, there is trade-off between
25764 accuracy in time domain and frequency domain. If timeclamp is lower,
25765 event in time domain is represented more accurately (such as fast bass drum),
25766 otherwise event in frequency domain is represented more accurately
25767 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
25770 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
25771 limits future samples by applying asymmetric windowing in time domain, useful
25772 when low latency is required. Accepted range is @code{[0, 1]}.
25775 Specify the transform base frequency. Default value is @code{20.01523126408007475},
25776 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
25779 Specify the transform end frequency. Default value is @code{20495.59681441799654},
25780 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
25783 This option is deprecated and ignored.
25786 Specify the transform length in time domain. Use this option to control accuracy
25787 trade-off between time domain and frequency domain at every frequency sample.
25788 It can contain variables:
25790 @item frequency, freq, f
25791 the frequency where it is evaluated
25792 @item timeclamp, tc
25793 the value of @var{timeclamp} option.
25795 Default value is @code{384*tc/(384+tc*f)}.
25798 Specify the transform count for every video frame. Default value is @code{6}.
25799 Acceptable range is @code{[1, 30]}.
25802 Specify the transform count for every single pixel. Default value is @code{0},
25803 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
25806 Specify font file for use with freetype to draw the axis. If not specified,
25807 use embedded font. Note that drawing with font file or embedded font is not
25808 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
25812 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
25813 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
25817 Specify font color expression. This is arithmetic expression that should return
25818 integer value 0xRRGGBB. It can contain variables:
25820 @item frequency, freq, f
25821 the frequency where it is evaluated
25822 @item timeclamp, tc
25823 the value of @var{timeclamp} option
25828 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
25829 @item r(x), g(x), b(x)
25830 red, green, and blue value of intensity x.
25832 Default value is @code{st(0, (midi(f)-59.5)/12);
25833 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
25834 r(1-ld(1)) + b(ld(1))}.
25837 Specify image file to draw the axis. This option override @var{fontfile} and
25838 @var{fontcolor} option.
25841 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
25842 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
25843 Default value is @code{1}.
25846 Set colorspace. The accepted values are:
25849 Unspecified (default)
25858 BT.470BG or BT.601-6 625
25861 SMPTE-170M or BT.601-6 525
25867 BT.2020 with non-constant luminance
25872 Set spectrogram color scheme. This is list of floating point values with format
25873 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
25874 The default is @code{1|0.5|0|0|0.5|1}.
25878 @subsection Examples
25882 Playing audio while showing the spectrum:
25884 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
25888 Same as above, but with frame rate 30 fps:
25890 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
25894 Playing at 1280x720:
25896 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
25900 Disable sonogram display:
25906 A1 and its harmonics: A1, A2, (near)E3, A3:
25908 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),
25909 asplit[a][out1]; [a] showcqt [out0]'
25913 Same as above, but with more accuracy in frequency domain:
25915 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),
25916 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
25922 bar_v=10:sono_v=bar_v*a_weighting(f)
25926 Custom gamma, now spectrum is linear to the amplitude.
25932 Custom tlength equation:
25934 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)))'
25938 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
25940 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
25944 Custom font using fontconfig:
25946 font='Courier New,Monospace,mono|bold'
25950 Custom frequency range with custom axis using image file:
25952 axisfile=myaxis.png:basefreq=40:endfreq=10000
25958 Convert input audio to video output representing the audio power spectrum.
25959 Audio amplitude is on Y-axis while frequency is on X-axis.
25961 The filter accepts the following options:
25965 Specify size of video. For the syntax of this option, check the
25966 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25967 Default is @code{1024x512}.
25971 This set how each frequency bin will be represented.
25973 It accepts the following values:
25979 Default is @code{bar}.
25982 Set amplitude scale.
25984 It accepts the following values:
25998 Default is @code{log}.
26001 Set frequency scale.
26003 It accepts the following values:
26012 Reverse logarithmic scale.
26014 Default is @code{lin}.
26017 Set window size. Allowed range is from 16 to 65536.
26019 Default is @code{2048}
26022 Set windowing function.
26024 It accepts the following values:
26047 Default is @code{hanning}.
26050 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26051 which means optimal overlap for selected window function will be picked.
26054 Set time averaging. Setting this to 0 will display current maximal peaks.
26055 Default is @code{1}, which means time averaging is disabled.
26058 Specify list of colors separated by space or by '|' which will be used to
26059 draw channel frequencies. Unrecognized or missing colors will be replaced
26063 Set channel display mode.
26065 It accepts the following values:
26070 Default is @code{combined}.
26073 Set minimum amplitude used in @code{log} amplitude scaler.
26076 Set data display mode.
26078 It accepts the following values:
26084 Default is @code{magnitude}.
26087 @section showspatial
26089 Convert stereo input audio to a video output, representing the spatial relationship
26090 between two channels.
26092 The filter accepts the following options:
26096 Specify the video size for the output. For the syntax of this option, check the
26097 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26098 Default value is @code{512x512}.
26101 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
26104 Set window function.
26106 It accepts the following values:
26131 Default value is @code{hann}.
26134 Set ratio of overlap window. Default value is @code{0.5}.
26135 When value is @code{1} overlap is set to recommended size for specific
26136 window function currently used.
26139 @anchor{showspectrum}
26140 @section showspectrum
26142 Convert input audio to a video output, representing the audio frequency
26145 The filter accepts the following options:
26149 Specify the video size for the output. For the syntax of this option, check the
26150 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26151 Default value is @code{640x512}.
26154 Specify how the spectrum should slide along the window.
26156 It accepts the following values:
26159 the samples start again on the left when they reach the right
26161 the samples scroll from right to left
26163 frames are only produced when the samples reach the right
26165 the samples scroll from left to right
26168 Default value is @code{replace}.
26171 Specify display mode.
26173 It accepts the following values:
26176 all channels are displayed in the same row
26178 all channels are displayed in separate rows
26181 Default value is @samp{combined}.
26184 Specify display color mode.
26186 It accepts the following values:
26189 each channel is displayed in a separate color
26191 each channel is displayed using the same color scheme
26193 each channel is displayed using the rainbow color scheme
26195 each channel is displayed using the moreland color scheme
26197 each channel is displayed using the nebulae color scheme
26199 each channel is displayed using the fire color scheme
26201 each channel is displayed using the fiery color scheme
26203 each channel is displayed using the fruit color scheme
26205 each channel is displayed using the cool color scheme
26207 each channel is displayed using the magma color scheme
26209 each channel is displayed using the green color scheme
26211 each channel is displayed using the viridis color scheme
26213 each channel is displayed using the plasma color scheme
26215 each channel is displayed using the cividis color scheme
26217 each channel is displayed using the terrain color scheme
26220 Default value is @samp{channel}.
26223 Specify scale used for calculating intensity color values.
26225 It accepts the following values:
26230 square root, default
26241 Default value is @samp{sqrt}.
26244 Specify frequency scale.
26246 It accepts the following values:
26254 Default value is @samp{lin}.
26257 Set saturation modifier for displayed colors. Negative values provide
26258 alternative color scheme. @code{0} is no saturation at all.
26259 Saturation must be in [-10.0, 10.0] range.
26260 Default value is @code{1}.
26263 Set window function.
26265 It accepts the following values:
26290 Default value is @code{hann}.
26293 Set orientation of time vs frequency axis. Can be @code{vertical} or
26294 @code{horizontal}. Default is @code{vertical}.
26297 Set ratio of overlap window. Default value is @code{0}.
26298 When value is @code{1} overlap is set to recommended size for specific
26299 window function currently used.
26302 Set scale gain for calculating intensity color values.
26303 Default value is @code{1}.
26306 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
26309 Set color rotation, must be in [-1.0, 1.0] range.
26310 Default value is @code{0}.
26313 Set start frequency from which to display spectrogram. Default is @code{0}.
26316 Set stop frequency to which to display spectrogram. Default is @code{0}.
26319 Set upper frame rate limit. Default is @code{auto}, unlimited.
26322 Draw time and frequency axes and legends. Default is disabled.
26325 The usage is very similar to the showwaves filter; see the examples in that
26328 @subsection Examples
26332 Large window with logarithmic color scaling:
26334 showspectrum=s=1280x480:scale=log
26338 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
26340 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
26341 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
26345 @section showspectrumpic
26347 Convert input audio to a single video frame, representing the audio frequency
26350 The filter accepts the following options:
26354 Specify the video size for the output. For the syntax of this option, check the
26355 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26356 Default value is @code{4096x2048}.
26359 Specify display mode.
26361 It accepts the following values:
26364 all channels are displayed in the same row
26366 all channels are displayed in separate rows
26368 Default value is @samp{combined}.
26371 Specify display color mode.
26373 It accepts the following values:
26376 each channel is displayed in a separate color
26378 each channel is displayed using the same color scheme
26380 each channel is displayed using the rainbow color scheme
26382 each channel is displayed using the moreland color scheme
26384 each channel is displayed using the nebulae color scheme
26386 each channel is displayed using the fire color scheme
26388 each channel is displayed using the fiery color scheme
26390 each channel is displayed using the fruit color scheme
26392 each channel is displayed using the cool color scheme
26394 each channel is displayed using the magma color scheme
26396 each channel is displayed using the green color scheme
26398 each channel is displayed using the viridis color scheme
26400 each channel is displayed using the plasma color scheme
26402 each channel is displayed using the cividis color scheme
26404 each channel is displayed using the terrain color scheme
26406 Default value is @samp{intensity}.
26409 Specify scale used for calculating intensity color values.
26411 It accepts the following values:
26416 square root, default
26426 Default value is @samp{log}.
26429 Specify frequency scale.
26431 It accepts the following values:
26439 Default value is @samp{lin}.
26442 Set saturation modifier for displayed colors. Negative values provide
26443 alternative color scheme. @code{0} is no saturation at all.
26444 Saturation must be in [-10.0, 10.0] range.
26445 Default value is @code{1}.
26448 Set window function.
26450 It accepts the following values:
26474 Default value is @code{hann}.
26477 Set orientation of time vs frequency axis. Can be @code{vertical} or
26478 @code{horizontal}. Default is @code{vertical}.
26481 Set scale gain for calculating intensity color values.
26482 Default value is @code{1}.
26485 Draw time and frequency axes and legends. Default is enabled.
26488 Set color rotation, must be in [-1.0, 1.0] range.
26489 Default value is @code{0}.
26492 Set start frequency from which to display spectrogram. Default is @code{0}.
26495 Set stop frequency to which to display spectrogram. Default is @code{0}.
26498 @subsection Examples
26502 Extract an audio spectrogram of a whole audio track
26503 in a 1024x1024 picture using @command{ffmpeg}:
26505 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
26509 @section showvolume
26511 Convert input audio volume to a video output.
26513 The filter accepts the following options:
26520 Set border width, allowed range is [0, 5]. Default is 1.
26523 Set channel width, allowed range is [80, 8192]. Default is 400.
26526 Set channel height, allowed range is [1, 900]. Default is 20.
26529 Set fade, allowed range is [0, 1]. Default is 0.95.
26532 Set volume color expression.
26534 The expression can use the following variables:
26538 Current max volume of channel in dB.
26544 Current channel number, starting from 0.
26548 If set, displays channel names. Default is enabled.
26551 If set, displays volume values. Default is enabled.
26554 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
26555 default is @code{h}.
26558 Set step size, allowed range is [0, 5]. Default is 0, which means
26562 Set background opacity, allowed range is [0, 1]. Default is 0.
26565 Set metering mode, can be peak: @code{p} or rms: @code{r},
26566 default is @code{p}.
26569 Set display scale, can be linear: @code{lin} or log: @code{log},
26570 default is @code{lin}.
26574 If set to > 0., display a line for the max level
26575 in the previous seconds.
26576 default is disabled: @code{0.}
26579 The color of the max line. Use when @code{dm} option is set to > 0.
26580 default is: @code{orange}
26585 Convert input audio to a video output, representing the samples waves.
26587 The filter accepts the following options:
26591 Specify the video size for the output. For the syntax of this option, check the
26592 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26593 Default value is @code{600x240}.
26598 Available values are:
26601 Draw a point for each sample.
26604 Draw a vertical line for each sample.
26607 Draw a point for each sample and a line between them.
26610 Draw a centered vertical line for each sample.
26613 Default value is @code{point}.
26616 Set the number of samples which are printed on the same column. A
26617 larger value will decrease the frame rate. Must be a positive
26618 integer. This option can be set only if the value for @var{rate}
26619 is not explicitly specified.
26622 Set the (approximate) output frame rate. This is done by setting the
26623 option @var{n}. Default value is "25".
26625 @item split_channels
26626 Set if channels should be drawn separately or overlap. Default value is 0.
26629 Set colors separated by '|' which are going to be used for drawing of each channel.
26632 Set amplitude scale.
26634 Available values are:
26652 Set the draw mode. This is mostly useful to set for high @var{n}.
26654 Available values are:
26657 Scale pixel values for each drawn sample.
26660 Draw every sample directly.
26663 Default value is @code{scale}.
26666 @subsection Examples
26670 Output the input file audio and the corresponding video representation
26673 amovie=a.mp3,asplit[out0],showwaves[out1]
26677 Create a synthetic signal and show it with showwaves, forcing a
26678 frame rate of 30 frames per second:
26680 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
26684 @section showwavespic
26686 Convert input audio to a single video frame, representing the samples waves.
26688 The filter accepts the following options:
26692 Specify the video size for the output. For the syntax of this option, check the
26693 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26694 Default value is @code{600x240}.
26696 @item split_channels
26697 Set if channels should be drawn separately or overlap. Default value is 0.
26700 Set colors separated by '|' which are going to be used for drawing of each channel.
26703 Set amplitude scale.
26705 Available values are:
26725 Available values are:
26728 Scale pixel values for each drawn sample.
26731 Draw every sample directly.
26734 Default value is @code{scale}.
26737 Set the filter mode.
26739 Available values are:
26742 Use average samples values for each drawn sample.
26745 Use peak samples values for each drawn sample.
26748 Default value is @code{average}.
26751 @subsection Examples
26755 Extract a channel split representation of the wave form of a whole audio track
26756 in a 1024x800 picture using @command{ffmpeg}:
26758 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
26762 @section sidedata, asidedata
26764 Delete frame side data, or select frames based on it.
26766 This filter accepts the following options:
26770 Set mode of operation of the filter.
26772 Can be one of the following:
26776 Select every frame with side data of @code{type}.
26779 Delete side data of @code{type}. If @code{type} is not set, delete all side
26785 Set side data type used with all modes. Must be set for @code{select} mode. For
26786 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
26787 in @file{libavutil/frame.h}. For example, to choose
26788 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
26792 @section spectrumsynth
26794 Synthesize audio from 2 input video spectrums, first input stream represents
26795 magnitude across time and second represents phase across time.
26796 The filter will transform from frequency domain as displayed in videos back
26797 to time domain as presented in audio output.
26799 This filter is primarily created for reversing processed @ref{showspectrum}
26800 filter outputs, but can synthesize sound from other spectrograms too.
26801 But in such case results are going to be poor if the phase data is not
26802 available, because in such cases phase data need to be recreated, usually
26803 it's just recreated from random noise.
26804 For best results use gray only output (@code{channel} color mode in
26805 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
26806 @code{lin} scale for phase video. To produce phase, for 2nd video, use
26807 @code{data} option. Inputs videos should generally use @code{fullframe}
26808 slide mode as that saves resources needed for decoding video.
26810 The filter accepts the following options:
26814 Specify sample rate of output audio, the sample rate of audio from which
26815 spectrum was generated may differ.
26818 Set number of channels represented in input video spectrums.
26821 Set scale which was used when generating magnitude input spectrum.
26822 Can be @code{lin} or @code{log}. Default is @code{log}.
26825 Set slide which was used when generating inputs spectrums.
26826 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
26827 Default is @code{fullframe}.
26830 Set window function used for resynthesis.
26833 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26834 which means optimal overlap for selected window function will be picked.
26837 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
26838 Default is @code{vertical}.
26841 @subsection Examples
26845 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
26846 then resynthesize videos back to audio with spectrumsynth:
26848 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
26849 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
26850 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
26854 @section split, asplit
26856 Split input into several identical outputs.
26858 @code{asplit} works with audio input, @code{split} with video.
26860 The filter accepts a single parameter which specifies the number of outputs. If
26861 unspecified, it defaults to 2.
26863 @subsection Examples
26867 Create two separate outputs from the same input:
26869 [in] split [out0][out1]
26873 To create 3 or more outputs, you need to specify the number of
26876 [in] asplit=3 [out0][out1][out2]
26880 Create two separate outputs from the same input, one cropped and
26883 [in] split [splitout1][splitout2];
26884 [splitout1] crop=100:100:0:0 [cropout];
26885 [splitout2] pad=200:200:100:100 [padout];
26889 Create 5 copies of the input audio with @command{ffmpeg}:
26891 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
26897 Receive commands sent through a libzmq client, and forward them to
26898 filters in the filtergraph.
26900 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
26901 must be inserted between two video filters, @code{azmq} between two
26902 audio filters. Both are capable to send messages to any filter type.
26904 To enable these filters you need to install the libzmq library and
26905 headers and configure FFmpeg with @code{--enable-libzmq}.
26907 For more information about libzmq see:
26908 @url{http://www.zeromq.org/}
26910 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
26911 receives messages sent through a network interface defined by the
26912 @option{bind_address} (or the abbreviation "@option{b}") option.
26913 Default value of this option is @file{tcp://localhost:5555}. You may
26914 want to alter this value to your needs, but do not forget to escape any
26915 ':' signs (see @ref{filtergraph escaping}).
26917 The received message must be in the form:
26919 @var{TARGET} @var{COMMAND} [@var{ARG}]
26922 @var{TARGET} specifies the target of the command, usually the name of
26923 the filter class or a specific filter instance name. The default
26924 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
26925 but you can override this by using the @samp{filter_name@@id} syntax
26926 (see @ref{Filtergraph syntax}).
26928 @var{COMMAND} specifies the name of the command for the target filter.
26930 @var{ARG} is optional and specifies the optional argument list for the
26931 given @var{COMMAND}.
26933 Upon reception, the message is processed and the corresponding command
26934 is injected into the filtergraph. Depending on the result, the filter
26935 will send a reply to the client, adopting the format:
26937 @var{ERROR_CODE} @var{ERROR_REASON}
26941 @var{MESSAGE} is optional.
26943 @subsection Examples
26945 Look at @file{tools/zmqsend} for an example of a zmq client which can
26946 be used to send commands processed by these filters.
26948 Consider the following filtergraph generated by @command{ffplay}.
26949 In this example the last overlay filter has an instance name. All other
26950 filters will have default instance names.
26953 ffplay -dumpgraph 1 -f lavfi "
26954 color=s=100x100:c=red [l];
26955 color=s=100x100:c=blue [r];
26956 nullsrc=s=200x100, zmq [bg];
26957 [bg][l] overlay [bg+l];
26958 [bg+l][r] overlay@@my=x=100 "
26961 To change the color of the left side of the video, the following
26962 command can be used:
26964 echo Parsed_color_0 c yellow | tools/zmqsend
26967 To change the right side:
26969 echo Parsed_color_1 c pink | tools/zmqsend
26972 To change the position of the right side:
26974 echo overlay@@my x 150 | tools/zmqsend
26978 @c man end MULTIMEDIA FILTERS
26980 @chapter Multimedia Sources
26981 @c man begin MULTIMEDIA SOURCES
26983 Below is a description of the currently available multimedia sources.
26987 This is the same as @ref{movie} source, except it selects an audio
26993 Read audio and/or video stream(s) from a movie container.
26995 It accepts the following parameters:
26999 The name of the resource to read (not necessarily a file; it can also be a
27000 device or a stream accessed through some protocol).
27002 @item format_name, f
27003 Specifies the format assumed for the movie to read, and can be either
27004 the name of a container or an input device. If not specified, the
27005 format is guessed from @var{movie_name} or by probing.
27007 @item seek_point, sp
27008 Specifies the seek point in seconds. The frames will be output
27009 starting from this seek point. The parameter is evaluated with
27010 @code{av_strtod}, so the numerical value may be suffixed by an IS
27011 postfix. The default value is "0".
27014 Specifies the streams to read. Several streams can be specified,
27015 separated by "+". The source will then have as many outputs, in the
27016 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
27017 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
27018 respectively the default (best suited) video and audio stream. Default
27019 is "dv", or "da" if the filter is called as "amovie".
27021 @item stream_index, si
27022 Specifies the index of the video stream to read. If the value is -1,
27023 the most suitable video stream will be automatically selected. The default
27024 value is "-1". Deprecated. If the filter is called "amovie", it will select
27025 audio instead of video.
27028 Specifies how many times to read the stream in sequence.
27029 If the value is 0, the stream will be looped infinitely.
27030 Default value is "1".
27032 Note that when the movie is looped the source timestamps are not
27033 changed, so it will generate non monotonically increasing timestamps.
27035 @item discontinuity
27036 Specifies the time difference between frames above which the point is
27037 considered a timestamp discontinuity which is removed by adjusting the later
27041 It allows overlaying a second video on top of the main input of
27042 a filtergraph, as shown in this graph:
27044 input -----------> deltapts0 --> overlay --> output
27047 movie --> scale--> deltapts1 -------+
27049 @subsection Examples
27053 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
27054 on top of the input labelled "in":
27056 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
27057 [in] setpts=PTS-STARTPTS [main];
27058 [main][over] overlay=16:16 [out]
27062 Read from a video4linux2 device, and overlay it on top of the input
27065 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
27066 [in] setpts=PTS-STARTPTS [main];
27067 [main][over] overlay=16:16 [out]
27071 Read the first video stream and the audio stream with id 0x81 from
27072 dvd.vob; the video is connected to the pad named "video" and the audio is
27073 connected to the pad named "audio":
27075 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
27079 @subsection Commands
27081 Both movie and amovie support the following commands:
27084 Perform seek using "av_seek_frame".
27085 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
27088 @var{stream_index}: If stream_index is -1, a default
27089 stream is selected, and @var{timestamp} is automatically converted
27090 from AV_TIME_BASE units to the stream specific time_base.
27092 @var{timestamp}: Timestamp in AVStream.time_base units
27093 or, if no stream is specified, in AV_TIME_BASE units.
27095 @var{flags}: Flags which select direction and seeking mode.
27099 Get movie duration in AV_TIME_BASE units.
27103 @c man end MULTIMEDIA SOURCES