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 RGB colorspace color keying.
8176 The filter accepts the following options:
8180 The color which will be replaced with transparency.
8183 Similarity percentage with the key color.
8185 0.01 matches only the exact key color, while 1.0 matches everything.
8190 0.0 makes pixels either fully transparent, or not transparent at all.
8192 Higher values result in semi-transparent pixels, with a higher transparency
8193 the more similar the pixels color is to the key color.
8196 @subsection Examples
8200 Make every green pixel in the input image transparent:
8202 ffmpeg -i input.png -vf colorkey=green out.png
8206 Overlay a greenscreen-video on top of a static background image.
8208 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
8212 @subsection Commands
8213 This filter supports same @ref{commands} as options.
8214 The command accepts the same syntax of the corresponding option.
8216 If the specified expression is not valid, it is kept at its current
8220 Remove all color information for all RGB colors except for certain one.
8222 The filter accepts the following options:
8226 The color which will not be replaced with neutral gray.
8229 Similarity percentage with the above color.
8230 0.01 matches only the exact key color, while 1.0 matches everything.
8233 Blend percentage. 0.0 makes pixels fully gray.
8234 Higher values result in more preserved color.
8237 @subsection Commands
8238 This filter supports same @ref{commands} as options.
8239 The command accepts the same syntax of the corresponding option.
8241 If the specified expression is not valid, it is kept at its current
8244 @section colorlevels
8246 Adjust video input frames using levels.
8248 The filter accepts the following options:
8255 Adjust red, green, blue and alpha input black point.
8256 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
8262 Adjust red, green, blue and alpha input white point.
8263 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
8265 Input levels are used to lighten highlights (bright tones), darken shadows
8266 (dark tones), change the balance of bright and dark tones.
8272 Adjust red, green, blue and alpha output black point.
8273 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
8279 Adjust red, green, blue and alpha output white point.
8280 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
8282 Output levels allows manual selection of a constrained output level range.
8285 @subsection Examples
8289 Make video output darker:
8291 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
8297 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
8301 Make video output lighter:
8303 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
8307 Increase brightness:
8309 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
8313 @subsection Commands
8315 This filter supports the all above options as @ref{commands}.
8317 @section colormatrix
8319 Convert color matrix.
8321 The filter accepts the following options:
8326 Specify the source and destination color matrix. Both values must be
8329 The accepted values are:
8357 For example to convert from BT.601 to SMPTE-240M, use the command:
8359 colormatrix=bt601:smpte240m
8364 Convert colorspace, transfer characteristics or color primaries.
8365 Input video needs to have an even size.
8367 The filter accepts the following options:
8372 Specify all color properties at once.
8374 The accepted values are:
8404 Specify output colorspace.
8406 The accepted values are:
8415 BT.470BG or BT.601-6 625
8418 SMPTE-170M or BT.601-6 525
8427 BT.2020 with non-constant luminance
8433 Specify output transfer characteristics.
8435 The accepted values are:
8447 Constant gamma of 2.2
8450 Constant gamma of 2.8
8453 SMPTE-170M, BT.601-6 625 or BT.601-6 525
8471 BT.2020 for 10-bits content
8474 BT.2020 for 12-bits content
8480 Specify output color primaries.
8482 The accepted values are:
8491 BT.470BG or BT.601-6 625
8494 SMPTE-170M or BT.601-6 525
8518 Specify output color range.
8520 The accepted values are:
8523 TV (restricted) range
8526 MPEG (restricted) range
8537 Specify output color format.
8539 The accepted values are:
8542 YUV 4:2:0 planar 8-bits
8545 YUV 4:2:0 planar 10-bits
8548 YUV 4:2:0 planar 12-bits
8551 YUV 4:2:2 planar 8-bits
8554 YUV 4:2:2 planar 10-bits
8557 YUV 4:2:2 planar 12-bits
8560 YUV 4:4:4 planar 8-bits
8563 YUV 4:4:4 planar 10-bits
8566 YUV 4:4:4 planar 12-bits
8571 Do a fast conversion, which skips gamma/primary correction. This will take
8572 significantly less CPU, but will be mathematically incorrect. To get output
8573 compatible with that produced by the colormatrix filter, use fast=1.
8576 Specify dithering mode.
8578 The accepted values are:
8584 Floyd-Steinberg dithering
8588 Whitepoint adaptation mode.
8590 The accepted values are:
8593 Bradford whitepoint adaptation
8596 von Kries whitepoint adaptation
8599 identity whitepoint adaptation (i.e. no whitepoint adaptation)
8603 Override all input properties at once. Same accepted values as @ref{all}.
8606 Override input colorspace. Same accepted values as @ref{space}.
8609 Override input color primaries. Same accepted values as @ref{primaries}.
8612 Override input transfer characteristics. Same accepted values as @ref{trc}.
8615 Override input color range. Same accepted values as @ref{range}.
8619 The filter converts the transfer characteristics, color space and color
8620 primaries to the specified user values. The output value, if not specified,
8621 is set to a default value based on the "all" property. If that property is
8622 also not specified, the filter will log an error. The output color range and
8623 format default to the same value as the input color range and format. The
8624 input transfer characteristics, color space, color primaries and color range
8625 should be set on the input data. If any of these are missing, the filter will
8626 log an error and no conversion will take place.
8628 For example to convert the input to SMPTE-240M, use the command:
8630 colorspace=smpte240m
8633 @section colortemperature
8634 Adjust color temperature in video to simulate variations in ambient color temperature.
8636 The filter accepts the following options:
8640 Set the temperature in Kelvin. Allowed range is from 1000 to 40000.
8641 Default value is 6500 K.
8644 Set mixing with filtered output. Allowed range is from 0 to 1.
8648 Set the amount of preserving lightness. Allowed range is from 0 to 1.
8652 @subsection Commands
8653 This filter supports same @ref{commands} as options.
8655 @section convolution
8657 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
8659 The filter accepts the following options:
8666 Set matrix for each plane.
8667 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
8668 and from 1 to 49 odd number of signed integers in @var{row} mode.
8674 Set multiplier for calculated value for each plane.
8675 If unset or 0, it will be sum of all matrix elements.
8681 Set bias for each plane. This value is added to the result of the multiplication.
8682 Useful for making the overall image brighter or darker. Default is 0.0.
8688 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
8689 Default is @var{square}.
8692 @subsection Commands
8694 This filter supports the all above options as @ref{commands}.
8696 @subsection Examples
8702 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"
8708 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"
8714 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"
8720 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"
8724 Apply laplacian edge detector which includes diagonals:
8726 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"
8732 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"
8738 Apply 2D convolution of video stream in frequency domain using second stream
8741 The filter accepts the following options:
8745 Set which planes to process.
8748 Set which impulse video frames will be processed, can be @var{first}
8749 or @var{all}. Default is @var{all}.
8752 The @code{convolve} filter also supports the @ref{framesync} options.
8756 Copy the input video source unchanged to the output. This is mainly useful for
8761 Video filtering on GPU using Apple's CoreImage API on OSX.
8763 Hardware acceleration is based on an OpenGL context. Usually, this means it is
8764 processed by video hardware. However, software-based OpenGL implementations
8765 exist which means there is no guarantee for hardware processing. It depends on
8768 There are many filters and image generators provided by Apple that come with a
8769 large variety of options. The filter has to be referenced by its name along
8772 The coreimage filter accepts the following options:
8775 List all available filters and generators along with all their respective
8776 options as well as possible minimum and maximum values along with the default
8783 Specify all filters by their respective name and options.
8784 Use @var{list_filters} to determine all valid filter names and options.
8785 Numerical options are specified by a float value and are automatically clamped
8786 to their respective value range. Vector and color options have to be specified
8787 by a list of space separated float values. Character escaping has to be done.
8788 A special option name @code{default} is available to use default options for a
8791 It is required to specify either @code{default} or at least one of the filter options.
8792 All omitted options are used with their default values.
8793 The syntax of the filter string is as follows:
8795 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
8799 Specify a rectangle where the output of the filter chain is copied into the
8800 input image. It is given by a list of space separated float values:
8802 output_rect=x\ y\ width\ height
8804 If not given, the output rectangle equals the dimensions of the input image.
8805 The output rectangle is automatically cropped at the borders of the input
8806 image. Negative values are valid for each component.
8808 output_rect=25\ 25\ 100\ 100
8812 Several filters can be chained for successive processing without GPU-HOST
8813 transfers allowing for fast processing of complex filter chains.
8814 Currently, only filters with zero (generators) or exactly one (filters) input
8815 image and one output image are supported. Also, transition filters are not yet
8818 Some filters generate output images with additional padding depending on the
8819 respective filter kernel. The padding is automatically removed to ensure the
8820 filter output has the same size as the input image.
8822 For image generators, the size of the output image is determined by the
8823 previous output image of the filter chain or the input image of the whole
8824 filterchain, respectively. The generators do not use the pixel information of
8825 this image to generate their output. However, the generated output is
8826 blended onto this image, resulting in partial or complete coverage of the
8829 The @ref{coreimagesrc} video source can be used for generating input images
8830 which are directly fed into the filter chain. By using it, providing input
8831 images by another video source or an input video is not required.
8833 @subsection Examples
8838 List all filters available:
8840 coreimage=list_filters=true
8844 Use the CIBoxBlur filter with default options to blur an image:
8846 coreimage=filter=CIBoxBlur@@default
8850 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
8851 its center at 100x100 and a radius of 50 pixels:
8853 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
8857 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
8858 given as complete and escaped command-line for Apple's standard bash shell:
8860 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
8866 Cover a rectangular object
8868 It accepts the following options:
8872 Filepath of the optional cover image, needs to be in yuv420.
8877 It accepts the following values:
8880 cover it by the supplied image
8882 cover it by interpolating the surrounding pixels
8885 Default value is @var{blur}.
8888 @subsection Examples
8892 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
8894 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8900 Crop the input video to given dimensions.
8902 It accepts the following parameters:
8906 The width of the output video. It defaults to @code{iw}.
8907 This expression is evaluated only once during the filter
8908 configuration, or when the @samp{w} or @samp{out_w} command is sent.
8911 The height of the output video. It defaults to @code{ih}.
8912 This expression is evaluated only once during the filter
8913 configuration, or when the @samp{h} or @samp{out_h} command is sent.
8916 The horizontal position, in the input video, of the left edge of the output
8917 video. It defaults to @code{(in_w-out_w)/2}.
8918 This expression is evaluated per-frame.
8921 The vertical position, in the input video, of the top edge of the output video.
8922 It defaults to @code{(in_h-out_h)/2}.
8923 This expression is evaluated per-frame.
8926 If set to 1 will force the output display aspect ratio
8927 to be the same of the input, by changing the output sample aspect
8928 ratio. It defaults to 0.
8931 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
8932 width/height/x/y as specified and will not be rounded to nearest smaller value.
8936 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
8937 expressions containing the following constants:
8942 The computed values for @var{x} and @var{y}. They are evaluated for
8947 The input width and height.
8951 These are the same as @var{in_w} and @var{in_h}.
8955 The output (cropped) width and height.
8959 These are the same as @var{out_w} and @var{out_h}.
8962 same as @var{iw} / @var{ih}
8965 input sample aspect ratio
8968 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
8972 horizontal and vertical chroma subsample values. For example for the
8973 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8976 The number of the input frame, starting from 0.
8979 the position in the file of the input frame, NAN if unknown
8982 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
8986 The expression for @var{out_w} may depend on the value of @var{out_h},
8987 and the expression for @var{out_h} may depend on @var{out_w}, but they
8988 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
8989 evaluated after @var{out_w} and @var{out_h}.
8991 The @var{x} and @var{y} parameters specify the expressions for the
8992 position of the top-left corner of the output (non-cropped) area. They
8993 are evaluated for each frame. If the evaluated value is not valid, it
8994 is approximated to the nearest valid value.
8996 The expression for @var{x} may depend on @var{y}, and the expression
8997 for @var{y} may depend on @var{x}.
8999 @subsection Examples
9003 Crop area with size 100x100 at position (12,34).
9008 Using named options, the example above becomes:
9010 crop=w=100:h=100:x=12:y=34
9014 Crop the central input area with size 100x100:
9020 Crop the central input area with size 2/3 of the input video:
9022 crop=2/3*in_w:2/3*in_h
9026 Crop the input video central square:
9033 Delimit the rectangle with the top-left corner placed at position
9034 100:100 and the right-bottom corner corresponding to the right-bottom
9035 corner of the input image.
9037 crop=in_w-100:in_h-100:100:100
9041 Crop 10 pixels from the left and right borders, and 20 pixels from
9042 the top and bottom borders
9044 crop=in_w-2*10:in_h-2*20
9048 Keep only the bottom right quarter of the input image:
9050 crop=in_w/2:in_h/2:in_w/2:in_h/2
9054 Crop height for getting Greek harmony:
9056 crop=in_w:1/PHI*in_w
9060 Apply trembling effect:
9062 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)
9066 Apply erratic camera effect depending on timestamp:
9068 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)"
9072 Set x depending on the value of y:
9074 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
9078 @subsection Commands
9080 This filter supports the following commands:
9086 Set width/height of the output video and the horizontal/vertical position
9088 The command accepts the same syntax of the corresponding option.
9090 If the specified expression is not valid, it is kept at its current
9096 Auto-detect the crop size.
9098 It calculates the necessary cropping parameters and prints the
9099 recommended parameters via the logging system. The detected dimensions
9100 correspond to the non-black area of the input video.
9102 It accepts the following parameters:
9107 Set higher black value threshold, which can be optionally specified
9108 from nothing (0) to everything (255 for 8-bit based formats). An intensity
9109 value greater to the set value is considered non-black. It defaults to 24.
9110 You can also specify a value between 0.0 and 1.0 which will be scaled depending
9111 on the bitdepth of the pixel format.
9114 The value which the width/height should be divisible by. It defaults to
9115 16. The offset is automatically adjusted to center the video. Use 2 to
9116 get only even dimensions (needed for 4:2:2 video). 16 is best when
9117 encoding to most video codecs.
9120 Set the number of initial frames for which evaluation is skipped.
9121 Default is 2. Range is 0 to INT_MAX.
9123 @item reset_count, reset
9124 Set the counter that determines after how many frames cropdetect will
9125 reset the previously detected largest video area and start over to
9126 detect the current optimal crop area. Default value is 0.
9128 This can be useful when channel logos distort the video area. 0
9129 indicates 'never reset', and returns the largest area encountered during
9136 Delay video filtering until a given wallclock timestamp. The filter first
9137 passes on @option{preroll} amount of frames, then it buffers at most
9138 @option{buffer} amount of frames and waits for the cue. After reaching the cue
9139 it forwards the buffered frames and also any subsequent frames coming in its
9142 The filter can be used synchronize the output of multiple ffmpeg processes for
9143 realtime output devices like decklink. By putting the delay in the filtering
9144 chain and pre-buffering frames the process can pass on data to output almost
9145 immediately after the target wallclock timestamp is reached.
9147 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
9153 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
9156 The duration of content to pass on as preroll expressed in seconds. Default is 0.
9159 The maximum duration of content to buffer before waiting for the cue expressed
9160 in seconds. Default is 0.
9167 Apply color adjustments using curves.
9169 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
9170 component (red, green and blue) has its values defined by @var{N} key points
9171 tied from each other using a smooth curve. The x-axis represents the pixel
9172 values from the input frame, and the y-axis the new pixel values to be set for
9175 By default, a component curve is defined by the two points @var{(0;0)} and
9176 @var{(1;1)}. This creates a straight line where each original pixel value is
9177 "adjusted" to its own value, which means no change to the image.
9179 The filter allows you to redefine these two points and add some more. A new
9180 curve (using a natural cubic spline interpolation) will be define to pass
9181 smoothly through all these new coordinates. The new defined points needs to be
9182 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
9183 be in the @var{[0;1]} interval. If the computed curves happened to go outside
9184 the vector spaces, the values will be clipped accordingly.
9186 The filter accepts the following options:
9190 Select one of the available color presets. This option can be used in addition
9191 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
9192 options takes priority on the preset values.
9193 Available presets are:
9196 @item color_negative
9199 @item increase_contrast
9201 @item linear_contrast
9202 @item medium_contrast
9204 @item strong_contrast
9207 Default is @code{none}.
9209 Set the master key points. These points will define a second pass mapping. It
9210 is sometimes called a "luminance" or "value" mapping. It can be used with
9211 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
9212 post-processing LUT.
9214 Set the key points for the red component.
9216 Set the key points for the green component.
9218 Set the key points for the blue component.
9220 Set the key points for all components (not including master).
9221 Can be used in addition to the other key points component
9222 options. In this case, the unset component(s) will fallback on this
9223 @option{all} setting.
9225 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
9227 Save Gnuplot script of the curves in specified file.
9230 To avoid some filtergraph syntax conflicts, each key points list need to be
9231 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
9233 @subsection Examples
9237 Increase slightly the middle level of blue:
9239 curves=blue='0/0 0.5/0.58 1/1'
9245 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'
9247 Here we obtain the following coordinates for each components:
9250 @code{(0;0.11) (0.42;0.51) (1;0.95)}
9252 @code{(0;0) (0.50;0.48) (1;1)}
9254 @code{(0;0.22) (0.49;0.44) (1;0.80)}
9258 The previous example can also be achieved with the associated built-in preset:
9260 curves=preset=vintage
9270 Use a Photoshop preset and redefine the points of the green component:
9272 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
9276 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
9277 and @command{gnuplot}:
9279 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
9280 gnuplot -p /tmp/curves.plt
9286 Video data analysis filter.
9288 This filter shows hexadecimal pixel values of part of video.
9290 The filter accepts the following options:
9294 Set output video size.
9297 Set x offset from where to pick pixels.
9300 Set y offset from where to pick pixels.
9303 Set scope mode, can be one of the following:
9306 Draw hexadecimal pixel values with white color on black background.
9309 Draw hexadecimal pixel values with input video pixel color on black
9313 Draw hexadecimal pixel values on color background picked from input video,
9314 the text color is picked in such way so its always visible.
9318 Draw rows and columns numbers on left and top of video.
9321 Set background opacity.
9324 Set display number format. Can be @code{hex}, or @code{dec}. Default is @code{hex}.
9327 Set pixel components to display. By default all pixel components are displayed.
9331 Apply Directional blur filter.
9333 The filter accepts the following options:
9337 Set angle of directional blur. Default is @code{45}.
9340 Set radius of directional blur. Default is @code{5}.
9343 Set which planes to filter. By default all planes are filtered.
9346 @subsection Commands
9347 This filter supports same @ref{commands} as options.
9348 The command accepts the same syntax of the corresponding option.
9350 If the specified expression is not valid, it is kept at its current
9355 Denoise frames using 2D DCT (frequency domain filtering).
9357 This filter is not designed for real time.
9359 The filter accepts the following options:
9363 Set the noise sigma constant.
9365 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
9366 coefficient (absolute value) below this threshold with be dropped.
9368 If you need a more advanced filtering, see @option{expr}.
9370 Default is @code{0}.
9373 Set number overlapping pixels for each block. Since the filter can be slow, you
9374 may want to reduce this value, at the cost of a less effective filter and the
9375 risk of various artefacts.
9377 If the overlapping value doesn't permit processing the whole input width or
9378 height, a warning will be displayed and according borders won't be denoised.
9380 Default value is @var{blocksize}-1, which is the best possible setting.
9383 Set the coefficient factor expression.
9385 For each coefficient of a DCT block, this expression will be evaluated as a
9386 multiplier value for the coefficient.
9388 If this is option is set, the @option{sigma} option will be ignored.
9390 The absolute value of the coefficient can be accessed through the @var{c}
9394 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
9395 @var{blocksize}, which is the width and height of the processed blocks.
9397 The default value is @var{3} (8x8) and can be raised to @var{4} for a
9398 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
9399 on the speed processing. Also, a larger block size does not necessarily means a
9403 @subsection Examples
9405 Apply a denoise with a @option{sigma} of @code{4.5}:
9410 The same operation can be achieved using the expression system:
9412 dctdnoiz=e='gte(c, 4.5*3)'
9415 Violent denoise using a block size of @code{16x16}:
9422 Remove banding artifacts from input video.
9423 It works by replacing banded pixels with average value of referenced pixels.
9425 The filter accepts the following options:
9432 Set banding detection threshold for each plane. Default is 0.02.
9433 Valid range is 0.00003 to 0.5.
9434 If difference between current pixel and reference pixel is less than threshold,
9435 it will be considered as banded.
9438 Banding detection range in pixels. Default is 16. If positive, random number
9439 in range 0 to set value will be used. If negative, exact absolute value
9441 The range defines square of four pixels around current pixel.
9444 Set direction in radians from which four pixel will be compared. If positive,
9445 random direction from 0 to set direction will be picked. If negative, exact of
9446 absolute value will be picked. For example direction 0, -PI or -2*PI radians
9447 will pick only pixels on same row and -PI/2 will pick only pixels on same
9451 If enabled, current pixel is compared with average value of all four
9452 surrounding pixels. The default is enabled. If disabled current pixel is
9453 compared with all four surrounding pixels. The pixel is considered banded
9454 if only all four differences with surrounding pixels are less than threshold.
9457 If enabled, current pixel is changed if and only if all pixel components are banded,
9458 e.g. banding detection threshold is triggered for all color components.
9459 The default is disabled.
9464 Remove blocking artifacts from input video.
9466 The filter accepts the following options:
9470 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
9471 This controls what kind of deblocking is applied.
9474 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
9480 Set blocking detection thresholds. Allowed range is 0 to 1.
9481 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
9482 Using higher threshold gives more deblocking strength.
9483 Setting @var{alpha} controls threshold detection at exact edge of block.
9484 Remaining options controls threshold detection near the edge. Each one for
9485 below/above or left/right. Setting any of those to @var{0} disables
9489 Set planes to filter. Default is to filter all available planes.
9492 @subsection Examples
9496 Deblock using weak filter and block size of 4 pixels.
9498 deblock=filter=weak:block=4
9502 Deblock using strong filter, block size of 4 pixels and custom thresholds for
9503 deblocking more edges.
9505 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
9509 Similar as above, but filter only first plane.
9511 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
9515 Similar as above, but filter only second and third plane.
9517 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
9524 Drop duplicated frames at regular intervals.
9526 The filter accepts the following options:
9530 Set the number of frames from which one will be dropped. Setting this to
9531 @var{N} means one frame in every batch of @var{N} frames will be dropped.
9532 Default is @code{5}.
9535 Set the threshold for duplicate detection. If the difference metric for a frame
9536 is less than or equal to this value, then it is declared as duplicate. Default
9540 Set scene change threshold. Default is @code{15}.
9544 Set the size of the x and y-axis blocks used during metric calculations.
9545 Larger blocks give better noise suppression, but also give worse detection of
9546 small movements. Must be a power of two. Default is @code{32}.
9549 Mark main input as a pre-processed input and activate clean source input
9550 stream. This allows the input to be pre-processed with various filters to help
9551 the metrics calculation while keeping the frame selection lossless. When set to
9552 @code{1}, the first stream is for the pre-processed input, and the second
9553 stream is the clean source from where the kept frames are chosen. Default is
9557 Set whether or not chroma is considered in the metric calculations. Default is
9563 Apply 2D deconvolution of video stream in frequency domain using second stream
9566 The filter accepts the following options:
9570 Set which planes to process.
9573 Set which impulse video frames will be processed, can be @var{first}
9574 or @var{all}. Default is @var{all}.
9577 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
9578 and height are not same and not power of 2 or if stream prior to convolving
9582 The @code{deconvolve} filter also supports the @ref{framesync} options.
9586 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
9588 It accepts the following options:
9592 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
9593 @var{rainbows} for cross-color reduction.
9596 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
9599 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
9602 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
9605 Set temporal chroma threshold. Lower values increases reduction of cross-color.
9610 Apply deflate effect to the video.
9612 This filter replaces the pixel by the local(3x3) average by taking into account
9613 only values lower than the pixel.
9615 It accepts the following options:
9622 Limit the maximum change for each plane, default is 65535.
9623 If 0, plane will remain unchanged.
9626 @subsection Commands
9628 This filter supports the all above options as @ref{commands}.
9632 Remove temporal frame luminance variations.
9634 It accepts the following options:
9638 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
9641 Set averaging mode to smooth temporal luminance variations.
9643 Available values are:
9668 Do not actually modify frame. Useful when one only wants metadata.
9673 Remove judder produced by partially interlaced telecined content.
9675 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
9676 source was partially telecined content then the output of @code{pullup,dejudder}
9677 will have a variable frame rate. May change the recorded frame rate of the
9678 container. Aside from that change, this filter will not affect constant frame
9681 The option available in this filter is:
9685 Specify the length of the window over which the judder repeats.
9687 Accepts any integer greater than 1. Useful values are:
9691 If the original was telecined from 24 to 30 fps (Film to NTSC).
9694 If the original was telecined from 25 to 30 fps (PAL to NTSC).
9697 If a mixture of the two.
9700 The default is @samp{4}.
9705 Suppress a TV station logo by a simple interpolation of the surrounding
9706 pixels. Just set a rectangle covering the logo and watch it disappear
9707 (and sometimes something even uglier appear - your mileage may vary).
9709 It accepts the following parameters:
9714 Specify the top left corner coordinates of the logo. They must be
9719 Specify the width and height of the logo to clear. They must be
9723 Specify the thickness of the fuzzy edge of the rectangle (added to
9724 @var{w} and @var{h}). The default value is 1. This option is
9725 deprecated, setting higher values should no longer be necessary and
9729 When set to 1, a green rectangle is drawn on the screen to simplify
9730 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
9731 The default value is 0.
9733 The rectangle is drawn on the outermost pixels which will be (partly)
9734 replaced with interpolated values. The values of the next pixels
9735 immediately outside this rectangle in each direction will be used to
9736 compute the interpolated pixel values inside the rectangle.
9740 @subsection Examples
9744 Set a rectangle covering the area with top left corner coordinates 0,0
9745 and size 100x77, and a band of size 10:
9747 delogo=x=0:y=0:w=100:h=77:band=10
9755 Remove the rain in the input image/video by applying the derain methods based on
9756 convolutional neural networks. Supported models:
9760 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
9761 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
9764 Training as well as model generation scripts are provided in
9765 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
9767 Native model files (.model) can be generated from TensorFlow model
9768 files (.pb) by using tools/python/convert.py
9770 The filter accepts the following options:
9774 Specify which filter to use. This option accepts the following values:
9778 Derain filter. To conduct derain filter, you need to use a derain model.
9781 Dehaze filter. To conduct dehaze filter, you need to use a dehaze model.
9783 Default value is @samp{derain}.
9786 Specify which DNN backend to use for model loading and execution. This option accepts
9787 the following values:
9791 Native implementation of DNN loading and execution.
9794 TensorFlow backend. To enable this backend you
9795 need to install the TensorFlow for C library (see
9796 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
9797 @code{--enable-libtensorflow}
9799 Default value is @samp{native}.
9802 Set path to model file specifying network architecture and its parameters.
9803 Note that different backends use different file formats. TensorFlow and native
9804 backend can load files for only its format.
9807 It can also be finished with @ref{dnn_processing} filter.
9811 Attempt to fix small changes in horizontal and/or vertical shift. This
9812 filter helps remove camera shake from hand-holding a camera, bumping a
9813 tripod, moving on a vehicle, etc.
9815 The filter accepts the following options:
9823 Specify a rectangular area where to limit the search for motion
9825 If desired the search for motion vectors can be limited to a
9826 rectangular area of the frame defined by its top left corner, width
9827 and height. These parameters have the same meaning as the drawbox
9828 filter which can be used to visualise the position of the bounding
9831 This is useful when simultaneous movement of subjects within the frame
9832 might be confused for camera motion by the motion vector search.
9834 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
9835 then the full frame is used. This allows later options to be set
9836 without specifying the bounding box for the motion vector search.
9838 Default - search the whole frame.
9842 Specify the maximum extent of movement in x and y directions in the
9843 range 0-64 pixels. Default 16.
9846 Specify how to generate pixels to fill blanks at the edge of the
9847 frame. Available values are:
9850 Fill zeroes at blank locations
9852 Original image at blank locations
9854 Extruded edge value at blank locations
9856 Mirrored edge at blank locations
9858 Default value is @samp{mirror}.
9861 Specify the blocksize to use for motion search. Range 4-128 pixels,
9865 Specify the contrast threshold for blocks. Only blocks with more than
9866 the specified contrast (difference between darkest and lightest
9867 pixels) will be considered. Range 1-255, default 125.
9870 Specify the search strategy. Available values are:
9873 Set exhaustive search
9875 Set less exhaustive search.
9877 Default value is @samp{exhaustive}.
9880 If set then a detailed log of the motion search is written to the
9887 Remove unwanted contamination of foreground colors, caused by reflected color of
9888 greenscreen or bluescreen.
9890 This filter accepts the following options:
9894 Set what type of despill to use.
9897 Set how spillmap will be generated.
9900 Set how much to get rid of still remaining spill.
9903 Controls amount of red in spill area.
9906 Controls amount of green in spill area.
9907 Should be -1 for greenscreen.
9910 Controls amount of blue in spill area.
9911 Should be -1 for bluescreen.
9914 Controls brightness of spill area, preserving colors.
9917 Modify alpha from generated spillmap.
9920 @subsection Commands
9922 This filter supports the all above options as @ref{commands}.
9926 Apply an exact inverse of the telecine operation. It requires a predefined
9927 pattern specified using the pattern option which must be the same as that passed
9928 to the telecine filter.
9930 This filter accepts the following options:
9939 The default value is @code{top}.
9943 A string of numbers representing the pulldown pattern you wish to apply.
9944 The default value is @code{23}.
9947 A number representing position of the first frame with respect to the telecine
9948 pattern. This is to be used if the stream is cut. The default value is @code{0}.
9953 Apply dilation effect to the video.
9955 This filter replaces the pixel by the local(3x3) maximum.
9957 It accepts the following options:
9964 Limit the maximum change for each plane, default is 65535.
9965 If 0, plane will remain unchanged.
9968 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9971 Flags to local 3x3 coordinates maps like this:
9978 @subsection Commands
9980 This filter supports the all above options as @ref{commands}.
9984 Displace pixels as indicated by second and third input stream.
9986 It takes three input streams and outputs one stream, the first input is the
9987 source, and second and third input are displacement maps.
9989 The second input specifies how much to displace pixels along the
9990 x-axis, while the third input specifies how much to displace pixels
9992 If one of displacement map streams terminates, last frame from that
9993 displacement map will be used.
9995 Note that once generated, displacements maps can be reused over and over again.
9997 A description of the accepted options follows.
10001 Set displace behavior for pixels that are out of range.
10003 Available values are:
10006 Missing pixels are replaced by black pixels.
10009 Adjacent pixels will spread out to replace missing pixels.
10012 Out of range pixels are wrapped so they point to pixels of other side.
10015 Out of range pixels will be replaced with mirrored pixels.
10017 Default is @samp{smear}.
10021 @subsection Examples
10025 Add ripple effect to rgb input of video size hd720:
10027 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
10031 Add wave effect to rgb input of video size hd720:
10033 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
10037 @anchor{dnn_processing}
10038 @section dnn_processing
10040 Do image processing with deep neural networks. It works together with another filter
10041 which converts the pixel format of the Frame to what the dnn network requires.
10043 The filter accepts the following options:
10047 Specify which DNN backend to use for model loading and execution. This option accepts
10048 the following values:
10052 Native implementation of DNN loading and execution.
10055 TensorFlow backend. To enable this backend you
10056 need to install the TensorFlow for C library (see
10057 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
10058 @code{--enable-libtensorflow}
10061 OpenVINO backend. To enable this backend you
10062 need to build and install the OpenVINO for C library (see
10063 @url{https://github.com/openvinotoolkit/openvino/blob/master/build-instruction.md}) and configure FFmpeg with
10064 @code{--enable-libopenvino} (--extra-cflags=-I... --extra-ldflags=-L... might
10065 be needed if the header files and libraries are not installed into system path)
10069 Default value is @samp{native}.
10072 Set path to model file specifying network architecture and its parameters.
10073 Note that different backends use different file formats. TensorFlow, OpenVINO and native
10074 backend can load files for only its format.
10076 Native model file (.model) can be generated from TensorFlow model file (.pb) by using tools/python/convert.py
10079 Set the input name of the dnn network.
10082 Set the output name of the dnn network.
10085 use DNN async execution if set (default: set),
10086 roll back to sync execution if the backend does not support async.
10090 @subsection Examples
10094 Remove rain in rgb24 frame with can.pb (see @ref{derain} filter):
10096 ./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
10100 Halve the pixel value of the frame with format gray32f:
10102 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
10106 Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
10108 ./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
10112 Handle the Y channel with espcn.pb (see @ref{sr} filter), which changes frame size, for format yuv420p (planar YUV formats supported):
10114 ./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y -y tmp.espcn.jpg
10121 Draw a colored box on the input image.
10123 It accepts the following parameters:
10128 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
10132 The expressions which specify the width and height of the box; if 0 they are interpreted as
10133 the input width and height. It defaults to 0.
10136 Specify the color of the box to write. For the general syntax of this option,
10137 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10138 value @code{invert} is used, the box edge color is the same as the
10139 video with inverted luma.
10142 The expression which sets the thickness of the box edge.
10143 A value of @code{fill} will create a filled box. Default value is @code{3}.
10145 See below for the list of accepted constants.
10148 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
10149 will overwrite the video's color and alpha pixels.
10150 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
10153 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10154 following constants:
10158 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10162 horizontal and vertical chroma subsample values. For example for the
10163 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10167 The input width and height.
10170 The input sample aspect ratio.
10174 The x and y offset coordinates where the box is drawn.
10178 The width and height of the drawn box.
10181 The thickness of the drawn box.
10183 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10184 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10188 @subsection Examples
10192 Draw a black box around the edge of the input image:
10198 Draw a box with color red and an opacity of 50%:
10200 drawbox=10:20:200:60:red@@0.5
10203 The previous example can be specified as:
10205 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
10209 Fill the box with pink color:
10211 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
10215 Draw a 2-pixel red 2.40:1 mask:
10217 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
10221 @subsection Commands
10222 This filter supports same commands as options.
10223 The command accepts the same syntax of the corresponding option.
10225 If the specified expression is not valid, it is kept at its current
10230 Draw a graph using input video metadata.
10232 It accepts the following parameters:
10236 Set 1st frame metadata key from which metadata values will be used to draw a graph.
10239 Set 1st foreground color expression.
10242 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
10245 Set 2nd foreground color expression.
10248 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
10251 Set 3rd foreground color expression.
10254 Set 4th frame metadata key from which metadata values will be used to draw a graph.
10257 Set 4th foreground color expression.
10260 Set minimal value of metadata value.
10263 Set maximal value of metadata value.
10266 Set graph background color. Default is white.
10271 Available values for mode is:
10278 Default is @code{line}.
10283 Available values for slide is:
10286 Draw new frame when right border is reached.
10289 Replace old columns with new ones.
10292 Scroll from right to left.
10295 Scroll from left to right.
10298 Draw single picture.
10301 Default is @code{frame}.
10304 Set size of graph video. For the syntax of this option, check the
10305 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
10306 The default value is @code{900x256}.
10309 Set the output frame rate. Default value is @code{25}.
10311 The foreground color expressions can use the following variables:
10314 Minimal value of metadata value.
10317 Maximal value of metadata value.
10320 Current metadata key value.
10323 The color is defined as 0xAABBGGRR.
10326 Example using metadata from @ref{signalstats} filter:
10328 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
10331 Example using metadata from @ref{ebur128} filter:
10333 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
10338 Draw a grid on the input image.
10340 It accepts the following parameters:
10345 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
10349 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
10350 input width and height, respectively, minus @code{thickness}, so image gets
10351 framed. Default to 0.
10354 Specify the color of the grid. For the general syntax of this option,
10355 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
10356 value @code{invert} is used, the grid color is the same as the
10357 video with inverted luma.
10360 The expression which sets the thickness of the grid line. Default value is @code{1}.
10362 See below for the list of accepted constants.
10365 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
10366 will overwrite the video's color and alpha pixels.
10367 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
10370 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
10371 following constants:
10375 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
10379 horizontal and vertical chroma subsample values. For example for the
10380 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10384 The input grid cell width and height.
10387 The input sample aspect ratio.
10391 The x and y coordinates of some point of grid intersection (meant to configure offset).
10395 The width and height of the drawn cell.
10398 The thickness of the drawn cell.
10400 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
10401 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
10405 @subsection Examples
10409 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
10411 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
10415 Draw a white 3x3 grid with an opacity of 50%:
10417 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
10421 @subsection Commands
10422 This filter supports same commands as options.
10423 The command accepts the same syntax of the corresponding option.
10425 If the specified expression is not valid, it is kept at its current
10431 Draw a text string or text from a specified file on top of a video, using the
10432 libfreetype library.
10434 To enable compilation of this filter, you need to configure FFmpeg with
10435 @code{--enable-libfreetype}.
10436 To enable default font fallback and the @var{font} option you need to
10437 configure FFmpeg with @code{--enable-libfontconfig}.
10438 To enable the @var{text_shaping} option, you need to configure FFmpeg with
10439 @code{--enable-libfribidi}.
10443 It accepts the following parameters:
10448 Used to draw a box around text using the background color.
10449 The value must be either 1 (enable) or 0 (disable).
10450 The default value of @var{box} is 0.
10453 Set the width of the border to be drawn around the box using @var{boxcolor}.
10454 The default value of @var{boxborderw} is 0.
10457 The color to be used for drawing box around text. For the syntax of this
10458 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10460 The default value of @var{boxcolor} is "white".
10463 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
10464 The default value of @var{line_spacing} is 0.
10467 Set the width of the border to be drawn around the text using @var{bordercolor}.
10468 The default value of @var{borderw} is 0.
10471 Set the color to be used for drawing border around text. For the syntax of this
10472 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10474 The default value of @var{bordercolor} is "black".
10477 Select how the @var{text} is expanded. Can be either @code{none},
10478 @code{strftime} (deprecated) or
10479 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
10483 Set a start time for the count. Value is in microseconds. Only applied
10484 in the deprecated strftime expansion mode. To emulate in normal expansion
10485 mode use the @code{pts} function, supplying the start time (in seconds)
10486 as the second argument.
10489 If true, check and fix text coords to avoid clipping.
10492 The color to be used for drawing fonts. For the syntax of this option, check
10493 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
10495 The default value of @var{fontcolor} is "black".
10497 @item fontcolor_expr
10498 String which is expanded the same way as @var{text} to obtain dynamic
10499 @var{fontcolor} value. By default this option has empty value and is not
10500 processed. When this option is set, it overrides @var{fontcolor} option.
10503 The font family to be used for drawing text. By default Sans.
10506 The font file to be used for drawing text. The path must be included.
10507 This parameter is mandatory if the fontconfig support is disabled.
10510 Draw the text applying alpha blending. The value can
10511 be a number between 0.0 and 1.0.
10512 The expression accepts the same variables @var{x, y} as well.
10513 The default value is 1.
10514 Please see @var{fontcolor_expr}.
10517 The font size to be used for drawing text.
10518 The default value of @var{fontsize} is 16.
10521 If set to 1, attempt to shape the text (for example, reverse the order of
10522 right-to-left text and join Arabic characters) before drawing it.
10523 Otherwise, just draw the text exactly as given.
10524 By default 1 (if supported).
10526 @item ft_load_flags
10527 The flags to be used for loading the fonts.
10529 The flags map the corresponding flags supported by libfreetype, and are
10530 a combination of the following values:
10537 @item vertical_layout
10538 @item force_autohint
10541 @item ignore_global_advance_width
10543 @item ignore_transform
10545 @item linear_design
10549 Default value is "default".
10551 For more information consult the documentation for the FT_LOAD_*
10555 The color to be used for drawing a shadow behind the drawn text. For the
10556 syntax of this option, check the @ref{color syntax,,"Color" section in the
10557 ffmpeg-utils manual,ffmpeg-utils}.
10559 The default value of @var{shadowcolor} is "black".
10563 The x and y offsets for the text shadow position with respect to the
10564 position of the text. They can be either positive or negative
10565 values. The default value for both is "0".
10568 The starting frame number for the n/frame_num variable. The default value
10572 The size in number of spaces to use for rendering the tab.
10573 Default value is 4.
10576 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
10577 format. It can be used with or without text parameter. @var{timecode_rate}
10578 option must be specified.
10580 @item timecode_rate, rate, r
10581 Set the timecode frame rate (timecode only). Value will be rounded to nearest
10582 integer. Minimum value is "1".
10583 Drop-frame timecode is supported for frame rates 30 & 60.
10586 If set to 1, the output of the timecode option will wrap around at 24 hours.
10587 Default is 0 (disabled).
10590 The text string to be drawn. The text must be a sequence of UTF-8
10591 encoded characters.
10592 This parameter is mandatory if no file is specified with the parameter
10596 A text file containing text to be drawn. The text must be a sequence
10597 of UTF-8 encoded characters.
10599 This parameter is mandatory if no text string is specified with the
10600 parameter @var{text}.
10602 If both @var{text} and @var{textfile} are specified, an error is thrown.
10605 If set to 1, the @var{textfile} will be reloaded before each frame.
10606 Be sure to update it atomically, or it may be read partially, or even fail.
10610 The expressions which specify the offsets where text will be drawn
10611 within the video frame. They are relative to the top/left border of the
10614 The default value of @var{x} and @var{y} is "0".
10616 See below for the list of accepted constants and functions.
10619 The parameters for @var{x} and @var{y} are expressions containing the
10620 following constants and functions:
10624 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
10628 horizontal and vertical chroma subsample values. For example for the
10629 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10632 the height of each text line
10640 @item max_glyph_a, ascent
10641 the maximum distance from the baseline to the highest/upper grid
10642 coordinate used to place a glyph outline point, for all the rendered
10644 It is a positive value, due to the grid's orientation with the Y axis
10647 @item max_glyph_d, descent
10648 the maximum distance from the baseline to the lowest grid coordinate
10649 used to place a glyph outline point, for all the rendered glyphs.
10650 This is a negative value, due to the grid's orientation, with the Y axis
10654 maximum glyph height, that is the maximum height for all the glyphs
10655 contained in the rendered text, it is equivalent to @var{ascent} -
10659 maximum glyph width, that is the maximum width for all the glyphs
10660 contained in the rendered text
10663 the number of input frame, starting from 0
10665 @item rand(min, max)
10666 return a random number included between @var{min} and @var{max}
10669 The input sample aspect ratio.
10672 timestamp expressed in seconds, NAN if the input timestamp is unknown
10675 the height of the rendered text
10678 the width of the rendered text
10682 the x and y offset coordinates where the text is drawn.
10684 These parameters allow the @var{x} and @var{y} expressions to refer
10685 to each other, so you can for example specify @code{y=x/dar}.
10688 A one character description of the current frame's picture type.
10691 The current packet's position in the input file or stream
10692 (in bytes, from the start of the input). A value of -1 indicates
10693 this info is not available.
10696 The current packet's duration, in seconds.
10699 The current packet's size (in bytes).
10702 @anchor{drawtext_expansion}
10703 @subsection Text expansion
10705 If @option{expansion} is set to @code{strftime},
10706 the filter recognizes strftime() sequences in the provided text and
10707 expands them accordingly. Check the documentation of strftime(). This
10708 feature is deprecated.
10710 If @option{expansion} is set to @code{none}, the text is printed verbatim.
10712 If @option{expansion} is set to @code{normal} (which is the default),
10713 the following expansion mechanism is used.
10715 The backslash character @samp{\}, followed by any character, always expands to
10716 the second character.
10718 Sequences of the form @code{%@{...@}} are expanded. The text between the
10719 braces is a function name, possibly followed by arguments separated by ':'.
10720 If the arguments contain special characters or delimiters (':' or '@}'),
10721 they should be escaped.
10723 Note that they probably must also be escaped as the value for the
10724 @option{text} option in the filter argument string and as the filter
10725 argument in the filtergraph description, and possibly also for the shell,
10726 that makes up to four levels of escaping; using a text file avoids these
10729 The following functions are available:
10734 The expression evaluation result.
10736 It must take one argument specifying the expression to be evaluated,
10737 which accepts the same constants and functions as the @var{x} and
10738 @var{y} values. Note that not all constants should be used, for
10739 example the text size is not known when evaluating the expression, so
10740 the constants @var{text_w} and @var{text_h} will have an undefined
10743 @item expr_int_format, eif
10744 Evaluate the expression's value and output as formatted integer.
10746 The first argument is the expression to be evaluated, just as for the @var{expr} function.
10747 The second argument specifies the output format. Allowed values are @samp{x},
10748 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
10749 @code{printf} function.
10750 The third parameter is optional and sets the number of positions taken by the output.
10751 It can be used to add padding with zeros from the left.
10754 The time at which the filter is running, expressed in UTC.
10755 It can accept an argument: a strftime() format string.
10758 The time at which the filter is running, expressed in the local time zone.
10759 It can accept an argument: a strftime() format string.
10762 Frame metadata. Takes one or two arguments.
10764 The first argument is mandatory and specifies the metadata key.
10766 The second argument is optional and specifies a default value, used when the
10767 metadata key is not found or empty.
10769 Available metadata can be identified by inspecting entries
10770 starting with TAG included within each frame section
10771 printed by running @code{ffprobe -show_frames}.
10773 String metadata generated in filters leading to
10774 the drawtext filter are also available.
10777 The frame number, starting from 0.
10780 A one character description of the current picture type.
10783 The timestamp of the current frame.
10784 It can take up to three arguments.
10786 The first argument is the format of the timestamp; it defaults to @code{flt}
10787 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
10788 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
10789 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
10790 @code{localtime} stands for the timestamp of the frame formatted as
10791 local time zone time.
10793 The second argument is an offset added to the timestamp.
10795 If the format is set to @code{hms}, a third argument @code{24HH} may be
10796 supplied to present the hour part of the formatted timestamp in 24h format
10799 If the format is set to @code{localtime} or @code{gmtime},
10800 a third argument may be supplied: a strftime() format string.
10801 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
10804 @subsection Commands
10806 This filter supports altering parameters via commands:
10809 Alter existing filter parameters.
10811 Syntax for the argument is the same as for filter invocation, e.g.
10814 fontsize=56:fontcolor=green:text='Hello World'
10817 Full filter invocation with sendcmd would look like this:
10820 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
10824 If the entire argument can't be parsed or applied as valid values then the filter will
10825 continue with its existing parameters.
10827 @subsection Examples
10831 Draw "Test Text" with font FreeSerif, using the default values for the
10832 optional parameters.
10835 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
10839 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
10840 and y=50 (counting from the top-left corner of the screen), text is
10841 yellow with a red box around it. Both the text and the box have an
10845 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
10846 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
10849 Note that the double quotes are not necessary if spaces are not used
10850 within the parameter list.
10853 Show the text at the center of the video frame:
10855 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
10859 Show the text at a random position, switching to a new position every 30 seconds:
10861 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)"
10865 Show a text line sliding from right to left in the last row of the video
10866 frame. The file @file{LONG_LINE} is assumed to contain a single line
10869 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
10873 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
10875 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
10879 Draw a single green letter "g", at the center of the input video.
10880 The glyph baseline is placed at half screen height.
10882 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
10886 Show text for 1 second every 3 seconds:
10888 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
10892 Use fontconfig to set the font. Note that the colons need to be escaped.
10894 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
10898 Draw "Test Text" with font size dependent on height of the video.
10900 drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
10904 Print the date of a real-time encoding (see strftime(3)):
10906 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
10910 Show text fading in and out (appearing/disappearing):
10913 DS=1.0 # display start
10914 DE=10.0 # display end
10915 FID=1.5 # fade in duration
10916 FOD=5 # fade out duration
10917 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 @}"
10921 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
10922 and the @option{fontsize} value are included in the @option{y} offset.
10924 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
10925 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
10929 Plot special @var{lavf.image2dec.source_basename} metadata onto each frame if
10930 such metadata exists. Otherwise, plot the string "NA". Note that image2 demuxer
10931 must have option @option{-export_path_metadata 1} for the special metadata fields
10932 to be available for filters.
10934 drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%@{metadata\:lavf.image2dec.source_basename\:NA@}':x=10:y=10"
10939 For more information about libfreetype, check:
10940 @url{http://www.freetype.org/}.
10942 For more information about fontconfig, check:
10943 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
10945 For more information about libfribidi, check:
10946 @url{http://fribidi.org/}.
10948 @section edgedetect
10950 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
10952 The filter accepts the following options:
10957 Set low and high threshold values used by the Canny thresholding
10960 The high threshold selects the "strong" edge pixels, which are then
10961 connected through 8-connectivity with the "weak" edge pixels selected
10962 by the low threshold.
10964 @var{low} and @var{high} threshold values must be chosen in the range
10965 [0,1], and @var{low} should be lesser or equal to @var{high}.
10967 Default value for @var{low} is @code{20/255}, and default value for @var{high}
10971 Define the drawing mode.
10975 Draw white/gray wires on black background.
10978 Mix the colors to create a paint/cartoon effect.
10981 Apply Canny edge detector on all selected planes.
10983 Default value is @var{wires}.
10986 Select planes for filtering. By default all available planes are filtered.
10989 @subsection Examples
10993 Standard edge detection with custom values for the hysteresis thresholding:
10995 edgedetect=low=0.1:high=0.4
10999 Painting effect without thresholding:
11001 edgedetect=mode=colormix:high=0
11007 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
11009 For each input image, the filter will compute the optimal mapping from
11010 the input to the output given the codebook length, that is the number
11011 of distinct output colors.
11013 This filter accepts the following options.
11016 @item codebook_length, l
11017 Set codebook length. The value must be a positive integer, and
11018 represents the number of distinct output colors. Default value is 256.
11021 Set the maximum number of iterations to apply for computing the optimal
11022 mapping. The higher the value the better the result and the higher the
11023 computation time. Default value is 1.
11026 Set a random seed, must be an integer included between 0 and
11027 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
11028 will try to use a good random seed on a best effort basis.
11031 Set pal8 output pixel format. This option does not work with codebook
11032 length greater than 256.
11037 Measure graylevel entropy in histogram of color channels of video frames.
11039 It accepts the following parameters:
11043 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
11045 @var{diff} mode measures entropy of histogram delta values, absolute differences
11046 between neighbour histogram values.
11050 Apply the EPX magnification filter which is designed for pixel art.
11052 It accepts the following option:
11056 Set the scaling dimension: @code{2} for @code{2xEPX}, @code{3} for
11058 Default is @code{3}.
11062 Set brightness, contrast, saturation and approximate gamma adjustment.
11064 The filter accepts the following options:
11068 Set the contrast expression. The value must be a float value in range
11069 @code{-1000.0} to @code{1000.0}. The default value is "1".
11072 Set the brightness expression. The value must be a float value in
11073 range @code{-1.0} to @code{1.0}. The default value is "0".
11076 Set the saturation expression. The value must be a float in
11077 range @code{0.0} to @code{3.0}. The default value is "1".
11080 Set the gamma expression. The value must be a float in range
11081 @code{0.1} to @code{10.0}. The default value is "1".
11084 Set the gamma expression for red. The value must be a float in
11085 range @code{0.1} to @code{10.0}. The default value is "1".
11088 Set the gamma expression for green. The value must be a float in range
11089 @code{0.1} to @code{10.0}. The default value is "1".
11092 Set the gamma expression for blue. The value must be a float in range
11093 @code{0.1} to @code{10.0}. The default value is "1".
11096 Set the gamma weight expression. It can be used to reduce the effect
11097 of a high gamma value on bright image areas, e.g. keep them from
11098 getting overamplified and just plain white. The value must be a float
11099 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
11100 gamma correction all the way down while @code{1.0} leaves it at its
11101 full strength. Default is "1".
11104 Set when the expressions for brightness, contrast, saturation and
11105 gamma expressions are evaluated.
11107 It accepts the following values:
11110 only evaluate expressions once during the filter initialization or
11111 when a command is processed
11114 evaluate expressions for each incoming frame
11117 Default value is @samp{init}.
11120 The expressions accept the following parameters:
11123 frame count of the input frame starting from 0
11126 byte position of the corresponding packet in the input file, NAN if
11130 frame rate of the input video, NAN if the input frame rate is unknown
11133 timestamp expressed in seconds, NAN if the input timestamp is unknown
11136 @subsection Commands
11137 The filter supports the following commands:
11141 Set the contrast expression.
11144 Set the brightness expression.
11147 Set the saturation expression.
11150 Set the gamma expression.
11153 Set the gamma_r expression.
11156 Set gamma_g expression.
11159 Set gamma_b expression.
11162 Set gamma_weight expression.
11164 The command accepts the same syntax of the corresponding option.
11166 If the specified expression is not valid, it is kept at its current
11173 Apply erosion effect to the video.
11175 This filter replaces the pixel by the local(3x3) minimum.
11177 It accepts the following options:
11184 Limit the maximum change for each plane, default is 65535.
11185 If 0, plane will remain unchanged.
11188 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
11191 Flags to local 3x3 coordinates maps like this:
11198 @subsection Commands
11200 This filter supports the all above options as @ref{commands}.
11204 Deinterlace the input video ("estdif" stands for "Edge Slope
11205 Tracing Deinterlacing Filter").
11207 Spatial only filter that uses edge slope tracing algorithm
11208 to interpolate missing lines.
11209 It accepts the following parameters:
11213 The interlacing mode to adopt. It accepts one of the following values:
11217 Output one frame for each frame.
11219 Output one frame for each field.
11222 The default value is @code{field}.
11225 The picture field parity assumed for the input interlaced video. It accepts one
11226 of the following values:
11230 Assume the top field is first.
11232 Assume the bottom field is first.
11234 Enable automatic detection of field parity.
11237 The default value is @code{auto}.
11238 If the interlacing is unknown or the decoder does not export this information,
11239 top field first will be assumed.
11242 Specify which frames to deinterlace. Accepts one of the following
11247 Deinterlace all frames.
11249 Only deinterlace frames marked as interlaced.
11252 The default value is @code{all}.
11255 Specify the search radius for edge slope tracing. Default value is 1.
11256 Allowed range is from 1 to 15.
11259 Specify the search radius for best edge matching. Default value is 2.
11260 Allowed range is from 0 to 15.
11263 Specify the interpolation used. Default is 4-point interpolation. It accepts one
11264 of the following values:
11268 Two-point interpolation.
11270 Four-point interpolation.
11272 Six-point interpolation.
11276 @subsection Commands
11277 This filter supports same @ref{commands} as options.
11279 @section extractplanes
11281 Extract color channel components from input video stream into
11282 separate grayscale video streams.
11284 The filter accepts the following option:
11288 Set plane(s) to extract.
11290 Available values for planes are:
11301 Choosing planes not available in the input will result in an error.
11302 That means you cannot select @code{r}, @code{g}, @code{b} planes
11303 with @code{y}, @code{u}, @code{v} planes at same time.
11306 @subsection Examples
11310 Extract luma, u and v color channel component from input video frame
11311 into 3 grayscale outputs:
11313 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
11319 Apply a fade-in/out effect to the input video.
11321 It accepts the following parameters:
11325 The effect type can be either "in" for a fade-in, or "out" for a fade-out
11327 Default is @code{in}.
11329 @item start_frame, s
11330 Specify the number of the frame to start applying the fade
11331 effect at. Default is 0.
11334 The number of frames that the fade effect lasts. At the end of the
11335 fade-in effect, the output video will have the same intensity as the input video.
11336 At the end of the fade-out transition, the output video will be filled with the
11337 selected @option{color}.
11341 If set to 1, fade only alpha channel, if one exists on the input.
11342 Default value is 0.
11344 @item start_time, st
11345 Specify the timestamp (in seconds) of the frame to start to apply the fade
11346 effect. If both start_frame and start_time are specified, the fade will start at
11347 whichever comes last. Default is 0.
11350 The number of seconds for which the fade effect has to last. At the end of the
11351 fade-in effect the output video will have the same intensity as the input video,
11352 at the end of the fade-out transition the output video will be filled with the
11353 selected @option{color}.
11354 If both duration and nb_frames are specified, duration is used. Default is 0
11355 (nb_frames is used by default).
11358 Specify the color of the fade. Default is "black".
11361 @subsection Examples
11365 Fade in the first 30 frames of video:
11370 The command above is equivalent to:
11376 Fade out the last 45 frames of a 200-frame video:
11379 fade=type=out:start_frame=155:nb_frames=45
11383 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
11385 fade=in:0:25, fade=out:975:25
11389 Make the first 5 frames yellow, then fade in from frame 5-24:
11391 fade=in:5:20:color=yellow
11395 Fade in alpha over first 25 frames of video:
11397 fade=in:0:25:alpha=1
11401 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
11403 fade=t=in:st=5.5:d=0.5
11409 Denoise frames using 3D FFT (frequency domain filtering).
11411 The filter accepts the following options:
11415 Set the noise sigma constant. This sets denoising strength.
11416 Default value is 1. Allowed range is from 0 to 30.
11417 Using very high sigma with low overlap may give blocking artifacts.
11420 Set amount of denoising. By default all detected noise is reduced.
11421 Default value is 1. Allowed range is from 0 to 1.
11424 Set size of block, Default is 4, can be 3, 4, 5 or 6.
11425 Actual size of block in pixels is 2 to power of @var{block}, so by default
11426 block size in pixels is 2^4 which is 16.
11429 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
11432 Set number of previous frames to use for denoising. By default is set to 0.
11435 Set number of next frames to to use for denoising. By default is set to 0.
11438 Set planes which will be filtered, by default are all available filtered
11443 Apply arbitrary expressions to samples in frequency domain
11447 Adjust the dc value (gain) of the luma plane of the image. The filter
11448 accepts an integer value in range @code{0} to @code{1000}. The default
11449 value is set to @code{0}.
11452 Adjust the dc value (gain) of the 1st chroma plane of the image. The
11453 filter accepts an integer value in range @code{0} to @code{1000}. The
11454 default value is set to @code{0}.
11457 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
11458 filter accepts an integer value in range @code{0} to @code{1000}. The
11459 default value is set to @code{0}.
11462 Set the frequency domain weight expression for the luma plane.
11465 Set the frequency domain weight expression for the 1st chroma plane.
11468 Set the frequency domain weight expression for the 2nd chroma plane.
11471 Set when the expressions are evaluated.
11473 It accepts the following values:
11476 Only evaluate expressions once during the filter initialization.
11479 Evaluate expressions for each incoming frame.
11482 Default value is @samp{init}.
11484 The filter accepts the following variables:
11487 The coordinates of the current sample.
11491 The width and height of the image.
11494 The number of input frame, starting from 0.
11497 @subsection Examples
11503 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
11509 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
11515 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
11521 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
11528 Extract a single field from an interlaced image using stride
11529 arithmetic to avoid wasting CPU time. The output frames are marked as
11532 The filter accepts the following options:
11536 Specify whether to extract the top (if the value is @code{0} or
11537 @code{top}) or the bottom field (if the value is @code{1} or
11543 Create new frames by copying the top and bottom fields from surrounding frames
11544 supplied as numbers by the hint file.
11548 Set file containing hints: absolute/relative frame numbers.
11550 There must be one line for each frame in a clip. Each line must contain two
11551 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
11552 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
11553 is current frame number for @code{absolute} mode or out of [-1, 1] range
11554 for @code{relative} mode. First number tells from which frame to pick up top
11555 field and second number tells from which frame to pick up bottom field.
11557 If optionally followed by @code{+} output frame will be marked as interlaced,
11558 else if followed by @code{-} output frame will be marked as progressive, else
11559 it will be marked same as input frame.
11560 If optionally followed by @code{t} output frame will use only top field, or in
11561 case of @code{b} it will use only bottom field.
11562 If line starts with @code{#} or @code{;} that line is skipped.
11565 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
11568 Example of first several lines of @code{hint} file for @code{relative} mode:
11570 0,0 - # first frame
11571 1,0 - # second frame, use third's frame top field and second's frame bottom field
11572 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
11587 @section fieldmatch
11589 Field matching filter for inverse telecine. It is meant to reconstruct the
11590 progressive frames from a telecined stream. The filter does not drop duplicated
11591 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
11592 followed by a decimation filter such as @ref{decimate} in the filtergraph.
11594 The separation of the field matching and the decimation is notably motivated by
11595 the possibility of inserting a de-interlacing filter fallback between the two.
11596 If the source has mixed telecined and real interlaced content,
11597 @code{fieldmatch} will not be able to match fields for the interlaced parts.
11598 But these remaining combed frames will be marked as interlaced, and thus can be
11599 de-interlaced by a later filter such as @ref{yadif} before decimation.
11601 In addition to the various configuration options, @code{fieldmatch} can take an
11602 optional second stream, activated through the @option{ppsrc} option. If
11603 enabled, the frames reconstruction will be based on the fields and frames from
11604 this second stream. This allows the first input to be pre-processed in order to
11605 help the various algorithms of the filter, while keeping the output lossless
11606 (assuming the fields are matched properly). Typically, a field-aware denoiser,
11607 or brightness/contrast adjustments can help.
11609 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
11610 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
11611 which @code{fieldmatch} is based on. While the semantic and usage are very
11612 close, some behaviour and options names can differ.
11614 The @ref{decimate} filter currently only works for constant frame rate input.
11615 If your input has mixed telecined (30fps) and progressive content with a lower
11616 framerate like 24fps use the following filterchain to produce the necessary cfr
11617 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
11619 The filter accepts the following options:
11623 Specify the assumed field order of the input stream. Available values are:
11627 Auto detect parity (use FFmpeg's internal parity value).
11629 Assume bottom field first.
11631 Assume top field first.
11634 Note that it is sometimes recommended not to trust the parity announced by the
11637 Default value is @var{auto}.
11640 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
11641 sense that it won't risk creating jerkiness due to duplicate frames when
11642 possible, but if there are bad edits or blended fields it will end up
11643 outputting combed frames when a good match might actually exist. On the other
11644 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
11645 but will almost always find a good frame if there is one. The other values are
11646 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
11647 jerkiness and creating duplicate frames versus finding good matches in sections
11648 with bad edits, orphaned fields, blended fields, etc.
11650 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
11652 Available values are:
11656 2-way matching (p/c)
11658 2-way matching, and trying 3rd match if still combed (p/c + n)
11660 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
11662 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
11663 still combed (p/c + n + u/b)
11665 3-way matching (p/c/n)
11667 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
11668 detected as combed (p/c/n + u/b)
11671 The parenthesis at the end indicate the matches that would be used for that
11672 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
11675 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
11678 Default value is @var{pc_n}.
11681 Mark the main input stream as a pre-processed input, and enable the secondary
11682 input stream as the clean source to pick the fields from. See the filter
11683 introduction for more details. It is similar to the @option{clip2} feature from
11686 Default value is @code{0} (disabled).
11689 Set the field to match from. It is recommended to set this to the same value as
11690 @option{order} unless you experience matching failures with that setting. In
11691 certain circumstances changing the field that is used to match from can have a
11692 large impact on matching performance. Available values are:
11696 Automatic (same value as @option{order}).
11698 Match from the bottom field.
11700 Match from the top field.
11703 Default value is @var{auto}.
11706 Set whether or not chroma is included during the match comparisons. In most
11707 cases it is recommended to leave this enabled. You should set this to @code{0}
11708 only if your clip has bad chroma problems such as heavy rainbowing or other
11709 artifacts. Setting this to @code{0} could also be used to speed things up at
11710 the cost of some accuracy.
11712 Default value is @code{1}.
11716 These define an exclusion band which excludes the lines between @option{y0} and
11717 @option{y1} from being included in the field matching decision. An exclusion
11718 band can be used to ignore subtitles, a logo, or other things that may
11719 interfere with the matching. @option{y0} sets the starting scan line and
11720 @option{y1} sets the ending line; all lines in between @option{y0} and
11721 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
11722 @option{y0} and @option{y1} to the same value will disable the feature.
11723 @option{y0} and @option{y1} defaults to @code{0}.
11726 Set the scene change detection threshold as a percentage of maximum change on
11727 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
11728 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
11729 @option{scthresh} is @code{[0.0, 100.0]}.
11731 Default value is @code{12.0}.
11734 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
11735 account the combed scores of matches when deciding what match to use as the
11736 final match. Available values are:
11740 No final matching based on combed scores.
11742 Combed scores are only used when a scene change is detected.
11744 Use combed scores all the time.
11747 Default is @var{sc}.
11750 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
11751 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
11752 Available values are:
11756 No forced calculation.
11758 Force p/c/n calculations.
11760 Force p/c/n/u/b calculations.
11763 Default value is @var{none}.
11766 This is the area combing threshold used for combed frame detection. This
11767 essentially controls how "strong" or "visible" combing must be to be detected.
11768 Larger values mean combing must be more visible and smaller values mean combing
11769 can be less visible or strong and still be detected. Valid settings are from
11770 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
11771 be detected as combed). This is basically a pixel difference value. A good
11772 range is @code{[8, 12]}.
11774 Default value is @code{9}.
11777 Sets whether or not chroma is considered in the combed frame decision. Only
11778 disable this if your source has chroma problems (rainbowing, etc.) that are
11779 causing problems for the combed frame detection with chroma enabled. Actually,
11780 using @option{chroma}=@var{0} is usually more reliable, except for the case
11781 where there is chroma only combing in the source.
11783 Default value is @code{0}.
11787 Respectively set the x-axis and y-axis size of the window used during combed
11788 frame detection. This has to do with the size of the area in which
11789 @option{combpel} pixels are required to be detected as combed for a frame to be
11790 declared combed. See the @option{combpel} parameter description for more info.
11791 Possible values are any number that is a power of 2 starting at 4 and going up
11794 Default value is @code{16}.
11797 The number of combed pixels inside any of the @option{blocky} by
11798 @option{blockx} size blocks on the frame for the frame to be detected as
11799 combed. While @option{cthresh} controls how "visible" the combing must be, this
11800 setting controls "how much" combing there must be in any localized area (a
11801 window defined by the @option{blockx} and @option{blocky} settings) on the
11802 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
11803 which point no frames will ever be detected as combed). This setting is known
11804 as @option{MI} in TFM/VFM vocabulary.
11806 Default value is @code{80}.
11809 @anchor{p/c/n/u/b meaning}
11810 @subsection p/c/n/u/b meaning
11812 @subsubsection p/c/n
11814 We assume the following telecined stream:
11817 Top fields: 1 2 2 3 4
11818 Bottom fields: 1 2 3 4 4
11821 The numbers correspond to the progressive frame the fields relate to. Here, the
11822 first two frames are progressive, the 3rd and 4th are combed, and so on.
11824 When @code{fieldmatch} is configured to run a matching from bottom
11825 (@option{field}=@var{bottom}) this is how this input stream get transformed:
11830 B 1 2 3 4 4 <-- matching reference
11839 As a result of the field matching, we can see that some frames get duplicated.
11840 To perform a complete inverse telecine, you need to rely on a decimation filter
11841 after this operation. See for instance the @ref{decimate} filter.
11843 The same operation now matching from top fields (@option{field}=@var{top})
11848 T 1 2 2 3 4 <-- matching reference
11858 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
11859 basically, they refer to the frame and field of the opposite parity:
11862 @item @var{p} matches the field of the opposite parity in the previous frame
11863 @item @var{c} matches the field of the opposite parity in the current frame
11864 @item @var{n} matches the field of the opposite parity in the next frame
11869 The @var{u} and @var{b} matching are a bit special in the sense that they match
11870 from the opposite parity flag. In the following examples, we assume that we are
11871 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
11872 'x' is placed above and below each matched fields.
11874 With bottom matching (@option{field}=@var{bottom}):
11879 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11880 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11888 With top matching (@option{field}=@var{top}):
11893 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
11894 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
11902 @subsection Examples
11904 Simple IVTC of a top field first telecined stream:
11906 fieldmatch=order=tff:combmatch=none, decimate
11909 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
11911 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
11914 @section fieldorder
11916 Transform the field order of the input video.
11918 It accepts the following parameters:
11923 The output field order. Valid values are @var{tff} for top field first or @var{bff}
11924 for bottom field first.
11927 The default value is @samp{tff}.
11929 The transformation is done by shifting the picture content up or down
11930 by one line, and filling the remaining line with appropriate picture content.
11931 This method is consistent with most broadcast field order converters.
11933 If the input video is not flagged as being interlaced, or it is already
11934 flagged as being of the required output field order, then this filter does
11935 not alter the incoming video.
11937 It is very useful when converting to or from PAL DV material,
11938 which is bottom field first.
11942 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
11945 @section fifo, afifo
11947 Buffer input images and send them when they are requested.
11949 It is mainly useful when auto-inserted by the libavfilter
11952 It does not take parameters.
11954 @section fillborders
11956 Fill borders of the input video, without changing video stream dimensions.
11957 Sometimes video can have garbage at the four edges and you may not want to
11958 crop video input to keep size multiple of some number.
11960 This filter accepts the following options:
11964 Number of pixels to fill from left border.
11967 Number of pixels to fill from right border.
11970 Number of pixels to fill from top border.
11973 Number of pixels to fill from bottom border.
11978 It accepts the following values:
11981 fill pixels using outermost pixels
11984 fill pixels using mirroring (half sample symmetric)
11987 fill pixels with constant value
11990 fill pixels using reflecting (whole sample symmetric)
11993 fill pixels using wrapping
11996 fade pixels to constant value
11999 Default is @var{smear}.
12002 Set color for pixels in fixed or fade mode. Default is @var{black}.
12005 @subsection Commands
12006 This filter supports same @ref{commands} as options.
12007 The command accepts the same syntax of the corresponding option.
12009 If the specified expression is not valid, it is kept at its current
12014 Find a rectangular object
12016 It accepts the following options:
12020 Filepath of the object image, needs to be in gray8.
12023 Detection threshold, default is 0.5.
12026 Number of mipmaps, default is 3.
12028 @item xmin, ymin, xmax, ymax
12029 Specifies the rectangle in which to search.
12032 @subsection Examples
12036 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
12038 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
12044 Flood area with values of same pixel components with another values.
12046 It accepts the following options:
12049 Set pixel x coordinate.
12052 Set pixel y coordinate.
12055 Set source #0 component value.
12058 Set source #1 component value.
12061 Set source #2 component value.
12064 Set source #3 component value.
12067 Set destination #0 component value.
12070 Set destination #1 component value.
12073 Set destination #2 component value.
12076 Set destination #3 component value.
12082 Convert the input video to one of the specified pixel formats.
12083 Libavfilter will try to pick one that is suitable as input to
12086 It accepts the following parameters:
12090 A '|'-separated list of pixel format names, such as
12091 "pix_fmts=yuv420p|monow|rgb24".
12095 @subsection Examples
12099 Convert the input video to the @var{yuv420p} format
12101 format=pix_fmts=yuv420p
12104 Convert the input video to any of the formats in the list
12106 format=pix_fmts=yuv420p|yuv444p|yuv410p
12113 Convert the video to specified constant frame rate by duplicating or dropping
12114 frames as necessary.
12116 It accepts the following parameters:
12120 The desired output frame rate. The default is @code{25}.
12123 Assume the first PTS should be the given value, in seconds. This allows for
12124 padding/trimming at the start of stream. By default, no assumption is made
12125 about the first frame's expected PTS, so no padding or trimming is done.
12126 For example, this could be set to 0 to pad the beginning with duplicates of
12127 the first frame if a video stream starts after the audio stream or to trim any
12128 frames with a negative PTS.
12131 Timestamp (PTS) rounding method.
12133 Possible values are:
12140 round towards -infinity
12142 round towards +infinity
12146 The default is @code{near}.
12149 Action performed when reading the last frame.
12151 Possible values are:
12154 Use same timestamp rounding method as used for other frames.
12156 Pass through last frame if input duration has not been reached yet.
12158 The default is @code{round}.
12162 Alternatively, the options can be specified as a flat string:
12163 @var{fps}[:@var{start_time}[:@var{round}]].
12165 See also the @ref{setpts} filter.
12167 @subsection Examples
12171 A typical usage in order to set the fps to 25:
12177 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
12179 fps=fps=film:round=near
12185 Pack two different video streams into a stereoscopic video, setting proper
12186 metadata on supported codecs. The two views should have the same size and
12187 framerate and processing will stop when the shorter video ends. Please note
12188 that you may conveniently adjust view properties with the @ref{scale} and
12191 It accepts the following parameters:
12195 The desired packing format. Supported values are:
12200 The views are next to each other (default).
12203 The views are on top of each other.
12206 The views are packed by line.
12209 The views are packed by column.
12212 The views are temporally interleaved.
12221 # Convert left and right views into a frame-sequential video
12222 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
12224 # Convert views into a side-by-side video with the same output resolution as the input
12225 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
12230 Change the frame rate by interpolating new video output frames from the source
12233 This filter is not designed to function correctly with interlaced media. If
12234 you wish to change the frame rate of interlaced media then you are required
12235 to deinterlace before this filter and re-interlace after this filter.
12237 A description of the accepted options follows.
12241 Specify the output frames per second. This option can also be specified
12242 as a value alone. The default is @code{50}.
12245 Specify the start of a range where the output frame will be created as a
12246 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12247 the default is @code{15}.
12250 Specify the end of a range where the output frame will be created as a
12251 linear interpolation of two frames. The range is [@code{0}-@code{255}],
12252 the default is @code{240}.
12255 Specify the level at which a scene change is detected as a value between
12256 0 and 100 to indicate a new scene; a low value reflects a low
12257 probability for the current frame to introduce a new scene, while a higher
12258 value means the current frame is more likely to be one.
12259 The default is @code{8.2}.
12262 Specify flags influencing the filter process.
12264 Available value for @var{flags} is:
12267 @item scene_change_detect, scd
12268 Enable scene change detection using the value of the option @var{scene}.
12269 This flag is enabled by default.
12275 Select one frame every N-th frame.
12277 This filter accepts the following option:
12280 Select frame after every @code{step} frames.
12281 Allowed values are positive integers higher than 0. Default value is @code{1}.
12284 @section freezedetect
12286 Detect frozen video.
12288 This filter logs a message and sets frame metadata when it detects that the
12289 input video has no significant change in content during a specified duration.
12290 Video freeze detection calculates the mean average absolute difference of all
12291 the components of video frames and compares it to a noise floor.
12293 The printed times and duration are expressed in seconds. The
12294 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
12295 whose timestamp equals or exceeds the detection duration and it contains the
12296 timestamp of the first frame of the freeze. The
12297 @code{lavfi.freezedetect.freeze_duration} and
12298 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
12301 The filter accepts the following options:
12305 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
12306 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
12310 Set freeze duration until notification (default is 2 seconds).
12313 @section freezeframes
12315 Freeze video frames.
12317 This filter freezes video frames using frame from 2nd input.
12319 The filter accepts the following options:
12323 Set number of first frame from which to start freeze.
12326 Set number of last frame from which to end freeze.
12329 Set number of frame from 2nd input which will be used instead of replaced frames.
12335 Apply a frei0r effect to the input video.
12337 To enable the compilation of this filter, you need to install the frei0r
12338 header and configure FFmpeg with @code{--enable-frei0r}.
12340 It accepts the following parameters:
12345 The name of the frei0r effect to load. If the environment variable
12346 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
12347 directories specified by the colon-separated list in @env{FREI0R_PATH}.
12348 Otherwise, the standard frei0r paths are searched, in this order:
12349 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
12350 @file{/usr/lib/frei0r-1/}.
12352 @item filter_params
12353 A '|'-separated list of parameters to pass to the frei0r effect.
12357 A frei0r effect parameter can be a boolean (its value is either
12358 "y" or "n"), a double, a color (specified as
12359 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
12360 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
12361 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
12362 a position (specified as @var{X}/@var{Y}, where
12363 @var{X} and @var{Y} are floating point numbers) and/or a string.
12365 The number and types of parameters depend on the loaded effect. If an
12366 effect parameter is not specified, the default value is set.
12368 @subsection Examples
12372 Apply the distort0r effect, setting the first two double parameters:
12374 frei0r=filter_name=distort0r:filter_params=0.5|0.01
12378 Apply the colordistance effect, taking a color as the first parameter:
12380 frei0r=colordistance:0.2/0.3/0.4
12381 frei0r=colordistance:violet
12382 frei0r=colordistance:0x112233
12386 Apply the perspective effect, specifying the top left and top right image
12389 frei0r=perspective:0.2/0.2|0.8/0.2
12393 For more information, see
12394 @url{http://frei0r.dyne.org}
12396 @subsection Commands
12398 This filter supports the @option{filter_params} option as @ref{commands}.
12402 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
12404 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
12405 processing filter, one of them is performed once per block, not per pixel.
12406 This allows for much higher speed.
12408 The filter accepts the following options:
12412 Set quality. This option defines the number of levels for averaging. It accepts
12413 an integer in the range 4-5. Default value is @code{4}.
12416 Force a constant quantization parameter. It accepts an integer in range 0-63.
12417 If not set, the filter will use the QP from the video stream (if available).
12420 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
12421 more details but also more artifacts, while higher values make the image smoother
12422 but also blurrier. Default value is @code{0} − PSNR optimal.
12424 @item use_bframe_qp
12425 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12426 option may cause flicker since the B-Frames have often larger QP. Default is
12427 @code{0} (not enabled).
12433 Apply Gaussian blur filter.
12435 The filter accepts the following options:
12439 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
12442 Set number of steps for Gaussian approximation. Default is @code{1}.
12445 Set which planes to filter. By default all planes are filtered.
12448 Set vertical sigma, if negative it will be same as @code{sigma}.
12449 Default is @code{-1}.
12452 @subsection Commands
12453 This filter supports same commands as options.
12454 The command accepts the same syntax of the corresponding option.
12456 If the specified expression is not valid, it is kept at its current
12461 Apply generic equation to each pixel.
12463 The filter accepts the following options:
12466 @item lum_expr, lum
12467 Set the luminance expression.
12469 Set the chrominance blue expression.
12471 Set the chrominance red expression.
12472 @item alpha_expr, a
12473 Set the alpha expression.
12475 Set the red expression.
12476 @item green_expr, g
12477 Set the green expression.
12479 Set the blue expression.
12482 The colorspace is selected according to the specified options. If one
12483 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
12484 options is specified, the filter will automatically select a YCbCr
12485 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
12486 @option{blue_expr} options is specified, it will select an RGB
12489 If one of the chrominance expression is not defined, it falls back on the other
12490 one. If no alpha expression is specified it will evaluate to opaque value.
12491 If none of chrominance expressions are specified, they will evaluate
12492 to the luminance expression.
12494 The expressions can use the following variables and functions:
12498 The sequential number of the filtered frame, starting from @code{0}.
12502 The coordinates of the current sample.
12506 The width and height of the image.
12510 Width and height scale depending on the currently filtered plane. It is the
12511 ratio between the corresponding luma plane number of pixels and the current
12512 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
12513 @code{0.5,0.5} for chroma planes.
12516 Time of the current frame, expressed in seconds.
12519 Return the value of the pixel at location (@var{x},@var{y}) of the current
12523 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
12527 Return the value of the pixel at location (@var{x},@var{y}) of the
12528 blue-difference chroma plane. Return 0 if there is no such plane.
12531 Return the value of the pixel at location (@var{x},@var{y}) of the
12532 red-difference chroma plane. Return 0 if there is no such plane.
12537 Return the value of the pixel at location (@var{x},@var{y}) of the
12538 red/green/blue component. Return 0 if there is no such component.
12541 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
12542 plane. Return 0 if there is no such plane.
12544 @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)
12545 Sum of sample values in the rectangle from (0,0) to (x,y), this allows obtaining
12546 sums of samples within a rectangle. See the functions without the sum postfix.
12548 @item interpolation
12549 Set one of interpolation methods:
12554 Default is bilinear.
12557 For functions, if @var{x} and @var{y} are outside the area, the value will be
12558 automatically clipped to the closer edge.
12560 Please note that this filter can use multiple threads in which case each slice
12561 will have its own expression state. If you want to use only a single expression
12562 state because your expressions depend on previous state then you should limit
12563 the number of filter threads to 1.
12565 @subsection Examples
12569 Flip the image horizontally:
12575 Generate a bidimensional sine wave, with angle @code{PI/3} and a
12576 wavelength of 100 pixels:
12578 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
12582 Generate a fancy enigmatic moving light:
12584 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
12588 Generate a quick emboss effect:
12590 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
12594 Modify RGB components depending on pixel position:
12596 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
12600 Create a radial gradient that is the same size as the input (also see
12601 the @ref{vignette} filter):
12603 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
12609 Fix the banding artifacts that are sometimes introduced into nearly flat
12610 regions by truncation to 8-bit color depth.
12611 Interpolate the gradients that should go where the bands are, and
12614 It is designed for playback only. Do not use it prior to
12615 lossy compression, because compression tends to lose the dither and
12616 bring back the bands.
12618 It accepts the following parameters:
12623 The maximum amount by which the filter will change any one pixel. This is also
12624 the threshold for detecting nearly flat regions. Acceptable values range from
12625 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
12629 The neighborhood to fit the gradient to. A larger radius makes for smoother
12630 gradients, but also prevents the filter from modifying the pixels near detailed
12631 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
12632 values will be clipped to the valid range.
12636 Alternatively, the options can be specified as a flat string:
12637 @var{strength}[:@var{radius}]
12639 @subsection Examples
12643 Apply the filter with a @code{3.5} strength and radius of @code{8}:
12649 Specify radius, omitting the strength (which will fall-back to the default
12657 @anchor{graphmonitor}
12658 @section graphmonitor
12659 Show various filtergraph stats.
12661 With this filter one can debug complete filtergraph.
12662 Especially issues with links filling with queued frames.
12664 The filter accepts the following options:
12668 Set video output size. Default is @var{hd720}.
12671 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
12674 Set output mode, can be @var{fulll} or @var{compact}.
12675 In @var{compact} mode only filters with some queued frames have displayed stats.
12678 Set flags which enable which stats are shown in video.
12680 Available values for flags are:
12683 Display number of queued frames in each link.
12685 @item frame_count_in
12686 Display number of frames taken from filter.
12688 @item frame_count_out
12689 Display number of frames given out from filter.
12692 Display current filtered frame pts.
12695 Display current filtered frame time.
12698 Display time base for filter link.
12701 Display used format for filter link.
12704 Display video size or number of audio channels in case of audio used by filter link.
12707 Display video frame rate or sample rate in case of audio used by filter link.
12710 Display link output status.
12714 Set upper limit for video rate of output stream, Default value is @var{25}.
12715 This guarantee that output video frame rate will not be higher than this value.
12719 A color constancy variation filter which estimates scene illumination via grey edge algorithm
12720 and corrects the scene colors accordingly.
12722 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
12724 The filter accepts the following options:
12728 The order of differentiation to be applied on the scene. Must be chosen in the range
12729 [0,2] and default value is 1.
12732 The Minkowski parameter to be used for calculating the Minkowski distance. Must
12733 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
12734 max value instead of calculating Minkowski distance.
12737 The standard deviation of Gaussian blur to be applied on the scene. Must be
12738 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
12739 can't be equal to 0 if @var{difford} is greater than 0.
12742 @subsection Examples
12748 greyedge=difford=1:minknorm=5:sigma=2
12754 greyedge=difford=1:minknorm=0:sigma=2
12762 Apply a Hald CLUT to a video stream.
12764 First input is the video stream to process, and second one is the Hald CLUT.
12765 The Hald CLUT input can be a simple picture or a complete video stream.
12767 The filter accepts the following options:
12771 Force termination when the shortest input terminates. Default is @code{0}.
12773 Continue applying the last CLUT after the end of the stream. A value of
12774 @code{0} disable the filter after the last frame of the CLUT is reached.
12775 Default is @code{1}.
12778 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
12779 filters share the same internals).
12781 This filter also supports the @ref{framesync} options.
12783 More information about the Hald CLUT can be found on Eskil Steenberg's website
12784 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
12786 @subsection Workflow examples
12788 @subsubsection Hald CLUT video stream
12790 Generate an identity Hald CLUT stream altered with various effects:
12792 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
12795 Note: make sure you use a lossless codec.
12797 Then use it with @code{haldclut} to apply it on some random stream:
12799 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
12802 The Hald CLUT will be applied to the 10 first seconds (duration of
12803 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
12804 to the remaining frames of the @code{mandelbrot} stream.
12806 @subsubsection Hald CLUT with preview
12808 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
12809 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
12810 biggest possible square starting at the top left of the picture. The remaining
12811 padding pixels (bottom or right) will be ignored. This area can be used to add
12812 a preview of the Hald CLUT.
12814 Typically, the following generated Hald CLUT will be supported by the
12815 @code{haldclut} filter:
12818 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
12819 pad=iw+320 [padded_clut];
12820 smptebars=s=320x256, split [a][b];
12821 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
12822 [main][b] overlay=W-320" -frames:v 1 clut.png
12825 It contains the original and a preview of the effect of the CLUT: SMPTE color
12826 bars are displayed on the right-top, and below the same color bars processed by
12829 Then, the effect of this Hald CLUT can be visualized with:
12831 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
12836 Flip the input video horizontally.
12838 For example, to horizontally flip the input video with @command{ffmpeg}:
12840 ffmpeg -i in.avi -vf "hflip" out.avi
12844 This filter applies a global color histogram equalization on a
12847 It can be used to correct video that has a compressed range of pixel
12848 intensities. The filter redistributes the pixel intensities to
12849 equalize their distribution across the intensity range. It may be
12850 viewed as an "automatically adjusting contrast filter". This filter is
12851 useful only for correcting degraded or poorly captured source
12854 The filter accepts the following options:
12858 Determine the amount of equalization to be applied. As the strength
12859 is reduced, the distribution of pixel intensities more-and-more
12860 approaches that of the input frame. The value must be a float number
12861 in the range [0,1] and defaults to 0.200.
12864 Set the maximum intensity that can generated and scale the output
12865 values appropriately. The strength should be set as desired and then
12866 the intensity can be limited if needed to avoid washing-out. The value
12867 must be a float number in the range [0,1] and defaults to 0.210.
12870 Set the antibanding level. If enabled the filter will randomly vary
12871 the luminance of output pixels by a small amount to avoid banding of
12872 the histogram. Possible values are @code{none}, @code{weak} or
12873 @code{strong}. It defaults to @code{none}.
12879 Compute and draw a color distribution histogram for the input video.
12881 The computed histogram is a representation of the color component
12882 distribution in an image.
12884 Standard histogram displays the color components distribution in an image.
12885 Displays color graph for each color component. Shows distribution of
12886 the Y, U, V, A or R, G, B components, depending on input format, in the
12887 current frame. Below each graph a color component scale meter is shown.
12889 The filter accepts the following options:
12893 Set height of level. Default value is @code{200}.
12894 Allowed range is [50, 2048].
12897 Set height of color scale. Default value is @code{12}.
12898 Allowed range is [0, 40].
12902 It accepts the following values:
12905 Per color component graphs are placed below each other.
12908 Per color component graphs are placed side by side.
12911 Presents information identical to that in the @code{parade}, except
12912 that the graphs representing color components are superimposed directly
12915 Default is @code{stack}.
12918 Set mode. Can be either @code{linear}, or @code{logarithmic}.
12919 Default is @code{linear}.
12922 Set what color components to display.
12923 Default is @code{7}.
12926 Set foreground opacity. Default is @code{0.7}.
12929 Set background opacity. Default is @code{0.5}.
12932 @subsection Examples
12937 Calculate and draw histogram:
12939 ffplay -i input -vf histogram
12947 This is a high precision/quality 3d denoise filter. It aims to reduce
12948 image noise, producing smooth images and making still images really
12949 still. It should enhance compressibility.
12951 It accepts the following optional parameters:
12955 A non-negative floating point number which specifies spatial luma strength.
12956 It defaults to 4.0.
12958 @item chroma_spatial
12959 A non-negative floating point number which specifies spatial chroma strength.
12960 It defaults to 3.0*@var{luma_spatial}/4.0.
12963 A floating point number which specifies luma temporal strength. It defaults to
12964 6.0*@var{luma_spatial}/4.0.
12967 A floating point number which specifies chroma temporal strength. It defaults to
12968 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
12971 @subsection Commands
12972 This filter supports same @ref{commands} as options.
12973 The command accepts the same syntax of the corresponding option.
12975 If the specified expression is not valid, it is kept at its current
12978 @anchor{hwdownload}
12979 @section hwdownload
12981 Download hardware frames to system memory.
12983 The input must be in hardware frames, and the output a non-hardware format.
12984 Not all formats will be supported on the output - it may be necessary to insert
12985 an additional @option{format} filter immediately following in the graph to get
12986 the output in a supported format.
12990 Map hardware frames to system memory or to another device.
12992 This filter has several different modes of operation; which one is used depends
12993 on the input and output formats:
12996 Hardware frame input, normal frame output
12998 Map the input frames to system memory and pass them to the output. If the
12999 original hardware frame is later required (for example, after overlaying
13000 something else on part of it), the @option{hwmap} filter can be used again
13001 in the next mode to retrieve it.
13003 Normal frame input, hardware frame output
13005 If the input is actually a software-mapped hardware frame, then unmap it -
13006 that is, return the original hardware frame.
13008 Otherwise, a device must be provided. Create new hardware surfaces on that
13009 device for the output, then map them back to the software format at the input
13010 and give those frames to the preceding filter. This will then act like the
13011 @option{hwupload} filter, but may be able to avoid an additional copy when
13012 the input is already in a compatible format.
13014 Hardware frame input and output
13016 A device must be supplied for the output, either directly or with the
13017 @option{derive_device} option. The input and output devices must be of
13018 different types and compatible - the exact meaning of this is
13019 system-dependent, but typically it means that they must refer to the same
13020 underlying hardware context (for example, refer to the same graphics card).
13022 If the input frames were originally created on the output device, then unmap
13023 to retrieve the original frames.
13025 Otherwise, map the frames to the output device - create new hardware frames
13026 on the output corresponding to the frames on the input.
13029 The following additional parameters are accepted:
13033 Set the frame mapping mode. Some combination of:
13036 The mapped frame should be readable.
13038 The mapped frame should be writeable.
13040 The mapping will always overwrite the entire frame.
13042 This may improve performance in some cases, as the original contents of the
13043 frame need not be loaded.
13045 The mapping must not involve any copying.
13047 Indirect mappings to copies of frames are created in some cases where either
13048 direct mapping is not possible or it would have unexpected properties.
13049 Setting this flag ensures that the mapping is direct and will fail if that is
13052 Defaults to @var{read+write} if not specified.
13054 @item derive_device @var{type}
13055 Rather than using the device supplied at initialisation, instead derive a new
13056 device of type @var{type} from the device the input frames exist on.
13059 In a hardware to hardware mapping, map in reverse - create frames in the sink
13060 and map them back to the source. This may be necessary in some cases where
13061 a mapping in one direction is required but only the opposite direction is
13062 supported by the devices being used.
13064 This option is dangerous - it may break the preceding filter in undefined
13065 ways if there are any additional constraints on that filter's output.
13066 Do not use it without fully understanding the implications of its use.
13072 Upload system memory frames to hardware surfaces.
13074 The device to upload to must be supplied when the filter is initialised. If
13075 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
13076 option or with the @option{derive_device} option. The input and output devices
13077 must be of different types and compatible - the exact meaning of this is
13078 system-dependent, but typically it means that they must refer to the same
13079 underlying hardware context (for example, refer to the same graphics card).
13081 The following additional parameters are accepted:
13084 @item derive_device @var{type}
13085 Rather than using the device supplied at initialisation, instead derive a new
13086 device of type @var{type} from the device the input frames exist on.
13089 @anchor{hwupload_cuda}
13090 @section hwupload_cuda
13092 Upload system memory frames to a CUDA device.
13094 It accepts the following optional parameters:
13098 The number of the CUDA device to use
13103 Apply a high-quality magnification filter designed for pixel art. This filter
13104 was originally created by Maxim Stepin.
13106 It accepts the following option:
13110 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
13111 @code{hq3x} and @code{4} for @code{hq4x}.
13112 Default is @code{3}.
13116 Stack input videos horizontally.
13118 All streams must be of same pixel format and of same height.
13120 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13121 to create same output.
13123 The filter accepts the following option:
13127 Set number of input streams. Default is 2.
13130 If set to 1, force the output to terminate when the shortest input
13131 terminates. Default value is 0.
13136 Modify the hue and/or the saturation of the input.
13138 It accepts the following parameters:
13142 Specify the hue angle as a number of degrees. It accepts an expression,
13143 and defaults to "0".
13146 Specify the saturation in the [-10,10] range. It accepts an expression and
13150 Specify the hue angle as a number of radians. It accepts an
13151 expression, and defaults to "0".
13154 Specify the brightness in the [-10,10] range. It accepts an expression and
13158 @option{h} and @option{H} are mutually exclusive, and can't be
13159 specified at the same time.
13161 The @option{b}, @option{h}, @option{H} and @option{s} option values are
13162 expressions containing the following constants:
13166 frame count of the input frame starting from 0
13169 presentation timestamp of the input frame expressed in time base units
13172 frame rate of the input video, NAN if the input frame rate is unknown
13175 timestamp expressed in seconds, NAN if the input timestamp is unknown
13178 time base of the input video
13181 @subsection Examples
13185 Set the hue to 90 degrees and the saturation to 1.0:
13191 Same command but expressing the hue in radians:
13197 Rotate hue and make the saturation swing between 0
13198 and 2 over a period of 1 second:
13200 hue="H=2*PI*t: s=sin(2*PI*t)+1"
13204 Apply a 3 seconds saturation fade-in effect starting at 0:
13206 hue="s=min(t/3\,1)"
13209 The general fade-in expression can be written as:
13211 hue="s=min(0\, max((t-START)/DURATION\, 1))"
13215 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
13217 hue="s=max(0\, min(1\, (8-t)/3))"
13220 The general fade-out expression can be written as:
13222 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
13227 @subsection Commands
13229 This filter supports the following commands:
13235 Modify the hue and/or the saturation and/or brightness of the input video.
13236 The command accepts the same syntax of the corresponding option.
13238 If the specified expression is not valid, it is kept at its current
13242 @section hysteresis
13244 Grow first stream into second stream by connecting components.
13245 This makes it possible to build more robust edge masks.
13247 This filter accepts the following options:
13251 Set which planes will be processed as bitmap, unprocessed planes will be
13252 copied from first stream.
13253 By default value 0xf, all planes will be processed.
13256 Set threshold which is used in filtering. If pixel component value is higher than
13257 this value filter algorithm for connecting components is activated.
13258 By default value is 0.
13261 The @code{hysteresis} filter also supports the @ref{framesync} options.
13265 Detect video interlacing type.
13267 This filter tries to detect if the input frames are interlaced, progressive,
13268 top or bottom field first. It will also try to detect fields that are
13269 repeated between adjacent frames (a sign of telecine).
13271 Single frame detection considers only immediately adjacent frames when classifying each frame.
13272 Multiple frame detection incorporates the classification history of previous frames.
13274 The filter will log these metadata values:
13277 @item single.current_frame
13278 Detected type of current frame using single-frame detection. One of:
13279 ``tff'' (top field first), ``bff'' (bottom field first),
13280 ``progressive'', or ``undetermined''
13283 Cumulative number of frames detected as top field first using single-frame detection.
13286 Cumulative number of frames detected as top field first using multiple-frame detection.
13289 Cumulative number of frames detected as bottom field first using single-frame detection.
13291 @item multiple.current_frame
13292 Detected type of current frame using multiple-frame detection. One of:
13293 ``tff'' (top field first), ``bff'' (bottom field first),
13294 ``progressive'', or ``undetermined''
13297 Cumulative number of frames detected as bottom field first using multiple-frame detection.
13299 @item single.progressive
13300 Cumulative number of frames detected as progressive using single-frame detection.
13302 @item multiple.progressive
13303 Cumulative number of frames detected as progressive using multiple-frame detection.
13305 @item single.undetermined
13306 Cumulative number of frames that could not be classified using single-frame detection.
13308 @item multiple.undetermined
13309 Cumulative number of frames that could not be classified using multiple-frame detection.
13311 @item repeated.current_frame
13312 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
13314 @item repeated.neither
13315 Cumulative number of frames with no repeated field.
13318 Cumulative number of frames with the top field repeated from the previous frame's top field.
13320 @item repeated.bottom
13321 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
13324 The filter accepts the following options:
13328 Set interlacing threshold.
13330 Set progressive threshold.
13332 Threshold for repeated field detection.
13334 Number of frames after which a given frame's contribution to the
13335 statistics is halved (i.e., it contributes only 0.5 to its
13336 classification). The default of 0 means that all frames seen are given
13337 full weight of 1.0 forever.
13338 @item analyze_interlaced_flag
13339 When this is not 0 then idet will use the specified number of frames to determine
13340 if the interlaced flag is accurate, it will not count undetermined frames.
13341 If the flag is found to be accurate it will be used without any further
13342 computations, if it is found to be inaccurate it will be cleared without any
13343 further computations. This allows inserting the idet filter as a low computational
13344 method to clean up the interlaced flag
13349 Deinterleave or interleave fields.
13351 This filter allows one to process interlaced images fields without
13352 deinterlacing them. Deinterleaving splits the input frame into 2
13353 fields (so called half pictures). Odd lines are moved to the top
13354 half of the output image, even lines to the bottom half.
13355 You can process (filter) them independently and then re-interleave them.
13357 The filter accepts the following options:
13361 @item chroma_mode, c
13362 @item alpha_mode, a
13363 Available values for @var{luma_mode}, @var{chroma_mode} and
13364 @var{alpha_mode} are:
13370 @item deinterleave, d
13371 Deinterleave fields, placing one above the other.
13373 @item interleave, i
13374 Interleave fields. Reverse the effect of deinterleaving.
13376 Default value is @code{none}.
13378 @item luma_swap, ls
13379 @item chroma_swap, cs
13380 @item alpha_swap, as
13381 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
13384 @subsection Commands
13386 This filter supports the all above options as @ref{commands}.
13390 Apply inflate effect to the video.
13392 This filter replaces the pixel by the local(3x3) average by taking into account
13393 only values higher than the pixel.
13395 It accepts the following options:
13402 Limit the maximum change for each plane, default is 65535.
13403 If 0, plane will remain unchanged.
13406 @subsection Commands
13408 This filter supports the all above options as @ref{commands}.
13412 Simple interlacing filter from progressive contents. This interleaves upper (or
13413 lower) lines from odd frames with lower (or upper) lines from even frames,
13414 halving the frame rate and preserving image height.
13417 Original Original New Frame
13418 Frame 'j' Frame 'j+1' (tff)
13419 ========== =========== ==================
13420 Line 0 --------------------> Frame 'j' Line 0
13421 Line 1 Line 1 ----> Frame 'j+1' Line 1
13422 Line 2 ---------------------> Frame 'j' Line 2
13423 Line 3 Line 3 ----> Frame 'j+1' Line 3
13425 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
13428 It accepts the following optional parameters:
13432 This determines whether the interlaced frame is taken from the even
13433 (tff - default) or odd (bff) lines of the progressive frame.
13436 Vertical lowpass filter to avoid twitter interlacing and
13437 reduce moire patterns.
13441 Disable vertical lowpass filter
13444 Enable linear filter (default)
13447 Enable complex filter. This will slightly less reduce twitter and moire
13448 but better retain detail and subjective sharpness impression.
13455 Deinterlace input video by applying Donald Graft's adaptive kernel
13456 deinterling. Work on interlaced parts of a video to produce
13457 progressive frames.
13459 The description of the accepted parameters follows.
13463 Set the threshold which affects the filter's tolerance when
13464 determining if a pixel line must be processed. It must be an integer
13465 in the range [0,255] and defaults to 10. A value of 0 will result in
13466 applying the process on every pixels.
13469 Paint pixels exceeding the threshold value to white if set to 1.
13473 Set the fields order. Swap fields if set to 1, leave fields alone if
13477 Enable additional sharpening if set to 1. Default is 0.
13480 Enable twoway sharpening if set to 1. Default is 0.
13483 @subsection Examples
13487 Apply default values:
13489 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
13493 Enable additional sharpening:
13499 Paint processed pixels in white:
13506 Apply kirsch operator to input video stream.
13508 The filter accepts the following option:
13512 Set which planes will be processed, unprocessed planes will be copied.
13513 By default value 0xf, all planes will be processed.
13516 Set value which will be multiplied with filtered result.
13519 Set value which will be added to filtered result.
13522 @subsection Commands
13524 This filter supports the all above options as @ref{commands}.
13528 Slowly update darker pixels.
13530 This filter makes short flashes of light appear longer.
13531 This filter accepts the following options:
13535 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
13538 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
13541 @subsection Commands
13543 This filter supports the all above options as @ref{commands}.
13545 @section lenscorrection
13547 Correct radial lens distortion
13549 This filter can be used to correct for radial distortion as can result from the use
13550 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
13551 one can use tools available for example as part of opencv or simply trial-and-error.
13552 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
13553 and extract the k1 and k2 coefficients from the resulting matrix.
13555 Note that effectively the same filter is available in the open-source tools Krita and
13556 Digikam from the KDE project.
13558 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
13559 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
13560 brightness distribution, so you may want to use both filters together in certain
13561 cases, though you will have to take care of ordering, i.e. whether vignetting should
13562 be applied before or after lens correction.
13564 @subsection Options
13566 The filter accepts the following options:
13570 Relative x-coordinate of the focal point of the image, and thereby the center of the
13571 distortion. This value has a range [0,1] and is expressed as fractions of the image
13572 width. Default is 0.5.
13574 Relative y-coordinate of the focal point of the image, and thereby the center of the
13575 distortion. This value has a range [0,1] and is expressed as fractions of the image
13576 height. Default is 0.5.
13578 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
13579 no correction. Default is 0.
13581 Coefficient of the double quadratic correction term. This value has a range [-1,1].
13582 0 means no correction. Default is 0.
13584 Set interpolation type. Can be @code{nearest} or @code{bilinear}.
13585 Default is @code{nearest}.
13587 Specify the color of the unmapped pixels. For the syntax of this option,
13588 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13589 manual,ffmpeg-utils}. Default color is @code{black@@0}.
13592 The formula that generates the correction is:
13594 @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)
13596 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
13597 distances from the focal point in the source and target images, respectively.
13599 @subsection Commands
13601 This filter supports the all above options as @ref{commands}.
13605 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
13607 The @code{lensfun} filter requires the camera make, camera model, and lens model
13608 to apply the lens correction. The filter will load the lensfun database and
13609 query it to find the corresponding camera and lens entries in the database. As
13610 long as these entries can be found with the given options, the filter can
13611 perform corrections on frames. Note that incomplete strings will result in the
13612 filter choosing the best match with the given options, and the filter will
13613 output the chosen camera and lens models (logged with level "info"). You must
13614 provide the make, camera model, and lens model as they are required.
13616 The filter accepts the following options:
13620 The make of the camera (for example, "Canon"). This option is required.
13623 The model of the camera (for example, "Canon EOS 100D"). This option is
13627 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
13628 option is required.
13631 The type of correction to apply. The following values are valid options:
13635 Enables fixing lens vignetting.
13638 Enables fixing lens geometry. This is the default.
13641 Enables fixing chromatic aberrations.
13644 Enables fixing lens vignetting and lens geometry.
13647 Enables fixing lens vignetting and chromatic aberrations.
13650 Enables fixing both lens geometry and chromatic aberrations.
13653 Enables all possible corrections.
13657 The focal length of the image/video (zoom; expected constant for video). For
13658 example, a 18--55mm lens has focal length range of [18--55], so a value in that
13659 range should be chosen when using that lens. Default 18.
13662 The aperture of the image/video (expected constant for video). Note that
13663 aperture is only used for vignetting correction. Default 3.5.
13665 @item focus_distance
13666 The focus distance of the image/video (expected constant for video). Note that
13667 focus distance is only used for vignetting and only slightly affects the
13668 vignetting correction process. If unknown, leave it at the default value (which
13672 The scale factor which is applied after transformation. After correction the
13673 video is no longer necessarily rectangular. This parameter controls how much of
13674 the resulting image is visible. The value 0 means that a value will be chosen
13675 automatically such that there is little or no unmapped area in the output
13676 image. 1.0 means that no additional scaling is done. Lower values may result
13677 in more of the corrected image being visible, while higher values may avoid
13678 unmapped areas in the output.
13680 @item target_geometry
13681 The target geometry of the output image/video. The following values are valid
13685 @item rectilinear (default)
13688 @item equirectangular
13689 @item fisheye_orthographic
13690 @item fisheye_stereographic
13691 @item fisheye_equisolid
13692 @item fisheye_thoby
13695 Apply the reverse of image correction (instead of correcting distortion, apply
13698 @item interpolation
13699 The type of interpolation used when correcting distortion. The following values
13704 @item linear (default)
13709 @subsection Examples
13713 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
13714 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
13718 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
13722 Apply the same as before, but only for the first 5 seconds of video.
13725 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
13732 Obtain the VMAF (Video Multi-Method Assessment Fusion)
13733 score between two input videos.
13735 The obtained VMAF score is printed through the logging system.
13737 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
13738 After installing the library it can be enabled using:
13739 @code{./configure --enable-libvmaf}.
13740 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
13742 The filter has following options:
13746 Set the model path which is to be used for SVM.
13747 Default value: @code{"/usr/local/share/model/vmaf_v0.6.1.pkl"}
13750 Set the file path to be used to store logs.
13753 Set the format of the log file (csv, json or xml).
13755 @item enable_transform
13756 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
13757 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
13758 Default value: @code{false}
13761 Invokes the phone model which will generate VMAF scores higher than in the
13762 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
13763 Default value: @code{false}
13766 Enables computing psnr along with vmaf.
13767 Default value: @code{false}
13770 Enables computing ssim along with vmaf.
13771 Default value: @code{false}
13774 Enables computing ms_ssim along with vmaf.
13775 Default value: @code{false}
13778 Set the pool method to be used for computing vmaf.
13779 Options are @code{min}, @code{harmonic_mean} or @code{mean} (default).
13782 Set number of threads to be used when computing vmaf.
13783 Default value: @code{0}, which makes use of all available logical processors.
13786 Set interval for frame subsampling used when computing vmaf.
13787 Default value: @code{1}
13789 @item enable_conf_interval
13790 Enables confidence interval.
13791 Default value: @code{false}
13794 This filter also supports the @ref{framesync} options.
13796 @subsection Examples
13799 On the below examples the input file @file{main.mpg} being processed is
13800 compared with the reference file @file{ref.mpg}.
13803 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
13807 Example with options:
13809 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
13813 Example with options and different containers:
13815 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 -
13821 Limits the pixel components values to the specified range [min, max].
13823 The filter accepts the following options:
13827 Lower bound. Defaults to the lowest allowed value for the input.
13830 Upper bound. Defaults to the highest allowed value for the input.
13833 Specify which planes will be processed. Defaults to all available.
13836 @subsection Commands
13838 This filter supports the all above options as @ref{commands}.
13844 The filter accepts the following options:
13848 Set the number of loops. Setting this value to -1 will result in infinite loops.
13852 Set maximal size in number of frames. Default is 0.
13855 Set first frame of loop. Default is 0.
13858 @subsection Examples
13862 Loop single first frame infinitely:
13864 loop=loop=-1:size=1:start=0
13868 Loop single first frame 10 times:
13870 loop=loop=10:size=1:start=0
13874 Loop 10 first frames 5 times:
13876 loop=loop=5:size=10:start=0
13882 Apply a 1D LUT to an input video.
13884 The filter accepts the following options:
13888 Set the 1D LUT file name.
13890 Currently supported formats:
13899 Select interpolation mode.
13901 Available values are:
13905 Use values from the nearest defined point.
13907 Interpolate values using the linear interpolation.
13909 Interpolate values using the cosine interpolation.
13911 Interpolate values using the cubic interpolation.
13913 Interpolate values using the spline interpolation.
13920 Apply a 3D LUT to an input video.
13922 The filter accepts the following options:
13926 Set the 3D LUT file name.
13928 Currently supported formats:
13942 Select interpolation mode.
13944 Available values are:
13948 Use values from the nearest defined point.
13950 Interpolate values using the 8 points defining a cube.
13952 Interpolate values using a tetrahedron.
13954 Interpolate values using a pyramid.
13956 Interpolate values using a prism.
13962 Turn certain luma values into transparency.
13964 The filter accepts the following options:
13968 Set the luma which will be used as base for transparency.
13969 Default value is @code{0}.
13972 Set the range of luma values to be keyed out.
13973 Default value is @code{0.01}.
13976 Set the range of softness. Default value is @code{0}.
13977 Use this to control gradual transition from zero to full transparency.
13980 @subsection Commands
13981 This filter supports same @ref{commands} as options.
13982 The command accepts the same syntax of the corresponding option.
13984 If the specified expression is not valid, it is kept at its current
13987 @section lut, lutrgb, lutyuv
13989 Compute a look-up table for binding each pixel component input value
13990 to an output value, and apply it to the input video.
13992 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
13993 to an RGB input video.
13995 These filters accept the following parameters:
13998 set first pixel component expression
14000 set second pixel component expression
14002 set third pixel component expression
14004 set fourth pixel component expression, corresponds to the alpha component
14007 set red component expression
14009 set green component expression
14011 set blue component expression
14013 alpha component expression
14016 set Y/luminance component expression
14018 set U/Cb component expression
14020 set V/Cr component expression
14023 Each of them specifies the expression to use for computing the lookup table for
14024 the corresponding pixel component values.
14026 The exact component associated to each of the @var{c*} options depends on the
14029 The @var{lut} filter requires either YUV or RGB pixel formats in input,
14030 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
14032 The expressions can contain the following constants and functions:
14037 The input width and height.
14040 The input value for the pixel component.
14043 The input value, clipped to the @var{minval}-@var{maxval} range.
14046 The maximum value for the pixel component.
14049 The minimum value for the pixel component.
14052 The negated value for the pixel component value, clipped to the
14053 @var{minval}-@var{maxval} range; it corresponds to the expression
14054 "maxval-clipval+minval".
14057 The computed value in @var{val}, clipped to the
14058 @var{minval}-@var{maxval} range.
14060 @item gammaval(gamma)
14061 The computed gamma correction value of the pixel component value,
14062 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
14064 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
14068 All expressions default to "val".
14070 @subsection Examples
14074 Negate input video:
14076 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
14077 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
14080 The above is the same as:
14082 lutrgb="r=negval:g=negval:b=negval"
14083 lutyuv="y=negval:u=negval:v=negval"
14093 Remove chroma components, turning the video into a graytone image:
14095 lutyuv="u=128:v=128"
14099 Apply a luma burning effect:
14105 Remove green and blue components:
14111 Set a constant alpha channel value on input:
14113 format=rgba,lutrgb=a="maxval-minval/2"
14117 Correct luminance gamma by a factor of 0.5:
14119 lutyuv=y=gammaval(0.5)
14123 Discard least significant bits of luma:
14125 lutyuv=y='bitand(val, 128+64+32)'
14129 Technicolor like effect:
14131 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
14135 @section lut2, tlut2
14137 The @code{lut2} filter takes two input streams and outputs one
14140 The @code{tlut2} (time lut2) filter takes two consecutive frames
14141 from one single stream.
14143 This filter accepts the following parameters:
14146 set first pixel component expression
14148 set second pixel component expression
14150 set third pixel component expression
14152 set fourth pixel component expression, corresponds to the alpha component
14155 set output bit depth, only available for @code{lut2} filter. By default is 0,
14156 which means bit depth is automatically picked from first input format.
14159 The @code{lut2} filter also supports the @ref{framesync} options.
14161 Each of them specifies the expression to use for computing the lookup table for
14162 the corresponding pixel component values.
14164 The exact component associated to each of the @var{c*} options depends on the
14167 The expressions can contain the following constants:
14172 The input width and height.
14175 The first input value for the pixel component.
14178 The second input value for the pixel component.
14181 The first input video bit depth.
14184 The second input video bit depth.
14187 All expressions default to "x".
14189 @subsection Examples
14193 Highlight differences between two RGB video streams:
14195 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)'
14199 Highlight differences between two YUV video streams:
14201 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)'
14205 Show max difference between two video streams:
14207 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)))'
14211 @section maskedclamp
14213 Clamp the first input stream with the second input and third input stream.
14215 Returns the value of first stream to be between second input
14216 stream - @code{undershoot} and third input stream + @code{overshoot}.
14218 This filter accepts the following options:
14221 Default value is @code{0}.
14224 Default value is @code{0}.
14227 Set which planes will be processed as bitmap, unprocessed planes will be
14228 copied from first stream.
14229 By default value 0xf, all planes will be processed.
14232 @subsection Commands
14234 This filter supports the all above options as @ref{commands}.
14238 Merge the second and third input stream into output stream using absolute differences
14239 between second input stream and first input stream and absolute difference between
14240 third input stream and first input stream. The picked value will be from second input
14241 stream if second absolute difference is greater than first one or from third input stream
14244 This filter accepts the following options:
14247 Set which planes will be processed as bitmap, unprocessed planes will be
14248 copied from first stream.
14249 By default value 0xf, all planes will be processed.
14252 @subsection Commands
14254 This filter supports the all above options as @ref{commands}.
14256 @section maskedmerge
14258 Merge the first input stream with the second input stream using per pixel
14259 weights in the third input stream.
14261 A value of 0 in the third stream pixel component means that pixel component
14262 from first stream is returned unchanged, while maximum value (eg. 255 for
14263 8-bit videos) means that pixel component from second stream is returned
14264 unchanged. Intermediate values define the amount of merging between both
14265 input stream's pixel components.
14267 This filter accepts the following options:
14270 Set which planes will be processed as bitmap, unprocessed planes will be
14271 copied from first stream.
14272 By default value 0xf, all planes will be processed.
14275 @subsection Commands
14277 This filter supports the all above options as @ref{commands}.
14281 Merge the second and third input stream into output stream using absolute differences
14282 between second input stream and first input stream and absolute difference between
14283 third input stream and first input stream. The picked value will be from second input
14284 stream if second absolute difference is less than first one or from third input stream
14287 This filter accepts the following options:
14290 Set which planes will be processed as bitmap, unprocessed planes will be
14291 copied from first stream.
14292 By default value 0xf, all planes will be processed.
14295 @subsection Commands
14297 This filter supports the all above options as @ref{commands}.
14299 @section maskedthreshold
14300 Pick pixels comparing absolute difference of two video streams with fixed
14303 If absolute difference between pixel component of first and second video
14304 stream is equal or lower than user supplied threshold than pixel component
14305 from first video stream is picked, otherwise pixel component from second
14306 video stream is picked.
14308 This filter accepts the following options:
14311 Set threshold used when picking pixels from absolute difference from two input
14315 Set which planes will be processed as bitmap, unprocessed planes will be
14316 copied from second stream.
14317 By default value 0xf, all planes will be processed.
14320 @subsection Commands
14322 This filter supports the all above options as @ref{commands}.
14325 Create mask from input video.
14327 For example it is useful to create motion masks after @code{tblend} filter.
14329 This filter accepts the following options:
14333 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
14336 Set high threshold. Any pixel component higher than this value will be set to max value
14337 allowed for current pixel format.
14340 Set planes to filter, by default all available planes are filtered.
14343 Fill all frame pixels with this value.
14346 Set max average pixel value for frame. If sum of all pixel components is higher that this
14347 average, output frame will be completely filled with value set by @var{fill} option.
14348 Typically useful for scene changes when used in combination with @code{tblend} filter.
14353 Apply motion-compensation deinterlacing.
14355 It needs one field per frame as input and must thus be used together
14356 with yadif=1/3 or equivalent.
14358 This filter accepts the following options:
14361 Set the deinterlacing mode.
14363 It accepts one of the following values:
14368 use iterative motion estimation
14370 like @samp{slow}, but use multiple reference frames.
14372 Default value is @samp{fast}.
14375 Set the picture field parity assumed for the input video. It must be
14376 one of the following values:
14380 assume top field first
14382 assume bottom field first
14385 Default value is @samp{bff}.
14388 Set per-block quantization parameter (QP) used by the internal
14391 Higher values should result in a smoother motion vector field but less
14392 optimal individual vectors. Default value is 1.
14397 Pick median pixel from certain rectangle defined by radius.
14399 This filter accepts the following options:
14403 Set horizontal radius size. Default value is @code{1}.
14404 Allowed range is integer from 1 to 127.
14407 Set which planes to process. Default is @code{15}, which is all available planes.
14410 Set vertical radius size. Default value is @code{0}.
14411 Allowed range is integer from 0 to 127.
14412 If it is 0, value will be picked from horizontal @code{radius} option.
14415 Set median percentile. Default value is @code{0.5}.
14416 Default value of @code{0.5} will pick always median values, while @code{0} will pick
14417 minimum values, and @code{1} maximum values.
14420 @subsection Commands
14421 This filter supports same @ref{commands} as options.
14422 The command accepts the same syntax of the corresponding option.
14424 If the specified expression is not valid, it is kept at its current
14427 @section mergeplanes
14429 Merge color channel components from several video streams.
14431 The filter accepts up to 4 input streams, and merge selected input
14432 planes to the output video.
14434 This filter accepts the following options:
14437 Set input to output plane mapping. Default is @code{0}.
14439 The mappings is specified as a bitmap. It should be specified as a
14440 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
14441 mapping for the first plane of the output stream. 'A' sets the number of
14442 the input stream to use (from 0 to 3), and 'a' the plane number of the
14443 corresponding input to use (from 0 to 3). The rest of the mappings is
14444 similar, 'Bb' describes the mapping for the output stream second
14445 plane, 'Cc' describes the mapping for the output stream third plane and
14446 'Dd' describes the mapping for the output stream fourth plane.
14449 Set output pixel format. Default is @code{yuva444p}.
14452 @subsection Examples
14456 Merge three gray video streams of same width and height into single video stream:
14458 [a0][a1][a2]mergeplanes=0x001020:yuv444p
14462 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
14464 [a0][a1]mergeplanes=0x00010210:yuva444p
14468 Swap Y and A plane in yuva444p stream:
14470 format=yuva444p,mergeplanes=0x03010200:yuva444p
14474 Swap U and V plane in yuv420p stream:
14476 format=yuv420p,mergeplanes=0x000201:yuv420p
14480 Cast a rgb24 clip to yuv444p:
14482 format=rgb24,mergeplanes=0x000102:yuv444p
14488 Estimate and export motion vectors using block matching algorithms.
14489 Motion vectors are stored in frame side data to be used by other filters.
14491 This filter accepts the following options:
14494 Specify the motion estimation method. Accepts one of the following values:
14498 Exhaustive search algorithm.
14500 Three step search algorithm.
14502 Two dimensional logarithmic search algorithm.
14504 New three step search algorithm.
14506 Four step search algorithm.
14508 Diamond search algorithm.
14510 Hexagon-based search algorithm.
14512 Enhanced predictive zonal search algorithm.
14514 Uneven multi-hexagon search algorithm.
14516 Default value is @samp{esa}.
14519 Macroblock size. Default @code{16}.
14522 Search parameter. Default @code{7}.
14525 @section midequalizer
14527 Apply Midway Image Equalization effect using two video streams.
14529 Midway Image Equalization adjusts a pair of images to have the same
14530 histogram, while maintaining their dynamics as much as possible. It's
14531 useful for e.g. matching exposures from a pair of stereo cameras.
14533 This filter has two inputs and one output, which must be of same pixel format, but
14534 may be of different sizes. The output of filter is first input adjusted with
14535 midway histogram of both inputs.
14537 This filter accepts the following option:
14541 Set which planes to process. Default is @code{15}, which is all available planes.
14544 @section minterpolate
14546 Convert the video to specified frame rate using motion interpolation.
14548 This filter accepts the following options:
14551 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}.
14554 Motion interpolation mode. Following values are accepted:
14557 Duplicate previous or next frame for interpolating new ones.
14559 Blend source frames. Interpolated frame is mean of previous and next frames.
14561 Motion compensated interpolation. Following options are effective when this mode is selected:
14565 Motion compensation mode. Following values are accepted:
14568 Overlapped block motion compensation.
14570 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
14572 Default mode is @samp{obmc}.
14575 Motion estimation mode. Following values are accepted:
14578 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
14580 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
14582 Default mode is @samp{bilat}.
14585 The algorithm to be used for motion estimation. Following values are accepted:
14588 Exhaustive search algorithm.
14590 Three step search algorithm.
14592 Two dimensional logarithmic search algorithm.
14594 New three step search algorithm.
14596 Four step search algorithm.
14598 Diamond search algorithm.
14600 Hexagon-based search algorithm.
14602 Enhanced predictive zonal search algorithm.
14604 Uneven multi-hexagon search algorithm.
14606 Default algorithm is @samp{epzs}.
14609 Macroblock size. Default @code{16}.
14612 Motion estimation search parameter. Default @code{32}.
14615 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).
14620 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:
14623 Disable scene change detection.
14625 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
14627 Default method is @samp{fdiff}.
14629 @item scd_threshold
14630 Scene change detection threshold. Default is @code{10.}.
14635 Mix several video input streams into one video stream.
14637 A description of the accepted options follows.
14641 The number of inputs. If unspecified, it defaults to 2.
14644 Specify weight of each input video stream as sequence.
14645 Each weight is separated by space. If number of weights
14646 is smaller than number of @var{frames} last specified
14647 weight will be used for all remaining unset weights.
14650 Specify scale, if it is set it will be multiplied with sum
14651 of each weight multiplied with pixel values to give final destination
14652 pixel value. By default @var{scale} is auto scaled to sum of weights.
14655 Specify how end of stream is determined.
14658 The duration of the longest input. (default)
14661 The duration of the shortest input.
14664 The duration of the first input.
14668 @subsection Commands
14670 This filter supports the following commands:
14674 Syntax is same as option with same name.
14677 @section mpdecimate
14679 Drop frames that do not differ greatly from the previous frame in
14680 order to reduce frame rate.
14682 The main use of this filter is for very-low-bitrate encoding
14683 (e.g. streaming over dialup modem), but it could in theory be used for
14684 fixing movies that were inverse-telecined incorrectly.
14686 A description of the accepted options follows.
14690 Set the maximum number of consecutive frames which can be dropped (if
14691 positive), or the minimum interval between dropped frames (if
14692 negative). If the value is 0, the frame is dropped disregarding the
14693 number of previous sequentially dropped frames.
14695 Default value is 0.
14700 Set the dropping threshold values.
14702 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
14703 represent actual pixel value differences, so a threshold of 64
14704 corresponds to 1 unit of difference for each pixel, or the same spread
14705 out differently over the block.
14707 A frame is a candidate for dropping if no 8x8 blocks differ by more
14708 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
14709 meaning the whole image) differ by more than a threshold of @option{lo}.
14711 Default value for @option{hi} is 64*12, default value for @option{lo} is
14712 64*5, and default value for @option{frac} is 0.33.
14718 Negate (invert) the input video.
14720 It accepts the following option:
14725 With value 1, it negates the alpha component, if present. Default value is 0.
14731 Denoise frames using Non-Local Means algorithm.
14733 Each pixel is adjusted by looking for other pixels with similar contexts. This
14734 context similarity is defined by comparing their surrounding patches of size
14735 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
14738 Note that the research area defines centers for patches, which means some
14739 patches will be made of pixels outside that research area.
14741 The filter accepts the following options.
14745 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
14748 Set patch size. Default is 7. Must be odd number in range [0, 99].
14751 Same as @option{p} but for chroma planes.
14753 The default value is @var{0} and means automatic.
14756 Set research size. Default is 15. Must be odd number in range [0, 99].
14759 Same as @option{r} but for chroma planes.
14761 The default value is @var{0} and means automatic.
14766 Deinterlace video using neural network edge directed interpolation.
14768 This filter accepts the following options:
14772 Mandatory option, without binary file filter can not work.
14773 Currently file can be found here:
14774 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
14777 Set which frames to deinterlace, by default it is @code{all}.
14778 Can be @code{all} or @code{interlaced}.
14781 Set mode of operation.
14783 Can be one of the following:
14787 Use frame flags, both fields.
14789 Use frame flags, single field.
14791 Use top field only.
14793 Use bottom field only.
14795 Use both fields, top first.
14797 Use both fields, bottom first.
14801 Set which planes to process, by default filter process all frames.
14804 Set size of local neighborhood around each pixel, used by the predictor neural
14807 Can be one of the following:
14820 Set the number of neurons in predictor neural network.
14821 Can be one of the following:
14832 Controls the number of different neural network predictions that are blended
14833 together to compute the final output value. Can be @code{fast}, default or
14837 Set which set of weights to use in the predictor.
14838 Can be one of the following:
14842 weights trained to minimize absolute error
14844 weights trained to minimize squared error
14848 Controls whether or not the prescreener neural network is used to decide
14849 which pixels should be processed by the predictor neural network and which
14850 can be handled by simple cubic interpolation.
14851 The prescreener is trained to know whether cubic interpolation will be
14852 sufficient for a pixel or whether it should be predicted by the predictor nn.
14853 The computational complexity of the prescreener nn is much less than that of
14854 the predictor nn. Since most pixels can be handled by cubic interpolation,
14855 using the prescreener generally results in much faster processing.
14856 The prescreener is pretty accurate, so the difference between using it and not
14857 using it is almost always unnoticeable.
14859 Can be one of the following:
14869 Default is @code{new}.
14872 @subsection Commands
14873 This filter supports same @ref{commands} as options, excluding @var{weights} option.
14877 Force libavfilter not to use any of the specified pixel formats for the
14878 input to the next filter.
14880 It accepts the following parameters:
14884 A '|'-separated list of pixel format names, such as
14885 pix_fmts=yuv420p|monow|rgb24".
14889 @subsection Examples
14893 Force libavfilter to use a format different from @var{yuv420p} for the
14894 input to the vflip filter:
14896 noformat=pix_fmts=yuv420p,vflip
14900 Convert the input video to any of the formats not contained in the list:
14902 noformat=yuv420p|yuv444p|yuv410p
14908 Add noise on video input frame.
14910 The filter accepts the following options:
14918 Set noise seed for specific pixel component or all pixel components in case
14919 of @var{all_seed}. Default value is @code{123457}.
14921 @item all_strength, alls
14922 @item c0_strength, c0s
14923 @item c1_strength, c1s
14924 @item c2_strength, c2s
14925 @item c3_strength, c3s
14926 Set noise strength for specific pixel component or all pixel components in case
14927 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
14929 @item all_flags, allf
14930 @item c0_flags, c0f
14931 @item c1_flags, c1f
14932 @item c2_flags, c2f
14933 @item c3_flags, c3f
14934 Set pixel component flags or set flags for all components if @var{all_flags}.
14935 Available values for component flags are:
14938 averaged temporal noise (smoother)
14940 mix random noise with a (semi)regular pattern
14942 temporal noise (noise pattern changes between frames)
14944 uniform noise (gaussian otherwise)
14948 @subsection Examples
14950 Add temporal and uniform noise to input video:
14952 noise=alls=20:allf=t+u
14957 Normalize RGB video (aka histogram stretching, contrast stretching).
14958 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
14960 For each channel of each frame, the filter computes the input range and maps
14961 it linearly to the user-specified output range. The output range defaults
14962 to the full dynamic range from pure black to pure white.
14964 Temporal smoothing can be used on the input range to reduce flickering (rapid
14965 changes in brightness) caused when small dark or bright objects enter or leave
14966 the scene. This is similar to the auto-exposure (automatic gain control) on a
14967 video camera, and, like a video camera, it may cause a period of over- or
14968 under-exposure of the video.
14970 The R,G,B channels can be normalized independently, which may cause some
14971 color shifting, or linked together as a single channel, which prevents
14972 color shifting. Linked normalization preserves hue. Independent normalization
14973 does not, so it can be used to remove some color casts. Independent and linked
14974 normalization can be combined in any ratio.
14976 The normalize filter accepts the following options:
14981 Colors which define the output range. The minimum input value is mapped to
14982 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
14983 The defaults are black and white respectively. Specifying white for
14984 @var{blackpt} and black for @var{whitept} will give color-inverted,
14985 normalized video. Shades of grey can be used to reduce the dynamic range
14986 (contrast). Specifying saturated colors here can create some interesting
14990 The number of previous frames to use for temporal smoothing. The input range
14991 of each channel is smoothed using a rolling average over the current frame
14992 and the @var{smoothing} previous frames. The default is 0 (no temporal
14996 Controls the ratio of independent (color shifting) channel normalization to
14997 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
14998 independent. Defaults to 1.0 (fully independent).
15001 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
15002 expensive no-op. Defaults to 1.0 (full strength).
15006 @subsection Commands
15007 This filter supports same @ref{commands} as options, excluding @var{smoothing} option.
15008 The command accepts the same syntax of the corresponding option.
15010 If the specified expression is not valid, it is kept at its current
15013 @subsection Examples
15015 Stretch video contrast to use the full dynamic range, with no temporal
15016 smoothing; may flicker depending on the source content:
15018 normalize=blackpt=black:whitept=white:smoothing=0
15021 As above, but with 50 frames of temporal smoothing; flicker should be
15022 reduced, depending on the source content:
15024 normalize=blackpt=black:whitept=white:smoothing=50
15027 As above, but with hue-preserving linked channel normalization:
15029 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
15032 As above, but with half strength:
15034 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
15037 Map the darkest input color to red, the brightest input color to cyan:
15039 normalize=blackpt=red:whitept=cyan
15044 Pass the video source unchanged to the output.
15047 Optical Character Recognition
15049 This filter uses Tesseract for optical character recognition. To enable
15050 compilation of this filter, you need to configure FFmpeg with
15051 @code{--enable-libtesseract}.
15053 It accepts the following options:
15057 Set datapath to tesseract data. Default is to use whatever was
15058 set at installation.
15061 Set language, default is "eng".
15064 Set character whitelist.
15067 Set character blacklist.
15070 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
15071 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
15075 Apply a video transform using libopencv.
15077 To enable this filter, install the libopencv library and headers and
15078 configure FFmpeg with @code{--enable-libopencv}.
15080 It accepts the following parameters:
15085 The name of the libopencv filter to apply.
15087 @item filter_params
15088 The parameters to pass to the libopencv filter. If not specified, the default
15089 values are assumed.
15093 Refer to the official libopencv documentation for more precise
15095 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
15097 Several libopencv filters are supported; see the following subsections.
15102 Dilate an image by using a specific structuring element.
15103 It corresponds to the libopencv function @code{cvDilate}.
15105 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
15107 @var{struct_el} represents a structuring element, and has the syntax:
15108 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
15110 @var{cols} and @var{rows} represent the number of columns and rows of
15111 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
15112 point, and @var{shape} the shape for the structuring element. @var{shape}
15113 must be "rect", "cross", "ellipse", or "custom".
15115 If the value for @var{shape} is "custom", it must be followed by a
15116 string of the form "=@var{filename}". The file with name
15117 @var{filename} is assumed to represent a binary image, with each
15118 printable character corresponding to a bright pixel. When a custom
15119 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
15120 or columns and rows of the read file are assumed instead.
15122 The default value for @var{struct_el} is "3x3+0x0/rect".
15124 @var{nb_iterations} specifies the number of times the transform is
15125 applied to the image, and defaults to 1.
15129 # Use the default values
15132 # Dilate using a structuring element with a 5x5 cross, iterating two times
15133 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
15135 # Read the shape from the file diamond.shape, iterating two times.
15136 # The file diamond.shape may contain a pattern of characters like this
15142 # The specified columns and rows are ignored
15143 # but the anchor point coordinates are not
15144 ocv=dilate:0x0+2x2/custom=diamond.shape|2
15149 Erode an image by using a specific structuring element.
15150 It corresponds to the libopencv function @code{cvErode}.
15152 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
15153 with the same syntax and semantics as the @ref{dilate} filter.
15157 Smooth the input video.
15159 The filter takes the following parameters:
15160 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
15162 @var{type} is the type of smooth filter to apply, and must be one of
15163 the following values: "blur", "blur_no_scale", "median", "gaussian",
15164 or "bilateral". The default value is "gaussian".
15166 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
15167 depends on the smooth type. @var{param1} and
15168 @var{param2} accept integer positive values or 0. @var{param3} and
15169 @var{param4} accept floating point values.
15171 The default value for @var{param1} is 3. The default value for the
15172 other parameters is 0.
15174 These parameters correspond to the parameters assigned to the
15175 libopencv function @code{cvSmooth}.
15177 @section oscilloscope
15179 2D Video Oscilloscope.
15181 Useful to measure spatial impulse, step responses, chroma delays, etc.
15183 It accepts the following parameters:
15187 Set scope center x position.
15190 Set scope center y position.
15193 Set scope size, relative to frame diagonal.
15196 Set scope tilt/rotation.
15202 Set trace center x position.
15205 Set trace center y position.
15208 Set trace width, relative to width of frame.
15211 Set trace height, relative to height of frame.
15214 Set which components to trace. By default it traces first three components.
15217 Draw trace grid. By default is enabled.
15220 Draw some statistics. By default is enabled.
15223 Draw scope. By default is enabled.
15226 @subsection Commands
15227 This filter supports same @ref{commands} as options.
15228 The command accepts the same syntax of the corresponding option.
15230 If the specified expression is not valid, it is kept at its current
15233 @subsection Examples
15237 Inspect full first row of video frame.
15239 oscilloscope=x=0.5:y=0:s=1
15243 Inspect full last row of video frame.
15245 oscilloscope=x=0.5:y=1:s=1
15249 Inspect full 5th line of video frame of height 1080.
15251 oscilloscope=x=0.5:y=5/1080:s=1
15255 Inspect full last column of video frame.
15257 oscilloscope=x=1:y=0.5:s=1:t=1
15265 Overlay one video on top of another.
15267 It takes two inputs and has one output. The first input is the "main"
15268 video on which the second input is overlaid.
15270 It accepts the following parameters:
15272 A description of the accepted options follows.
15277 Set the expression for the x and y coordinates of the overlaid video
15278 on the main video. Default value is "0" for both expressions. In case
15279 the expression is invalid, it is set to a huge value (meaning that the
15280 overlay will not be displayed within the output visible area).
15283 See @ref{framesync}.
15286 Set when the expressions for @option{x}, and @option{y} are evaluated.
15288 It accepts the following values:
15291 only evaluate expressions once during the filter initialization or
15292 when a command is processed
15295 evaluate expressions for each incoming frame
15298 Default value is @samp{frame}.
15301 See @ref{framesync}.
15304 Set the format for the output video.
15306 It accepts the following values:
15309 force YUV420 output
15312 force YUV420p10 output
15315 force YUV422 output
15318 force YUV422p10 output
15321 force YUV444 output
15324 force packed RGB output
15327 force planar RGB output
15330 automatically pick format
15333 Default value is @samp{yuv420}.
15336 See @ref{framesync}.
15339 Set format of alpha of the overlaid video, it can be @var{straight} or
15340 @var{premultiplied}. Default is @var{straight}.
15343 The @option{x}, and @option{y} expressions can contain the following
15349 The main input width and height.
15353 The overlay input width and height.
15357 The computed values for @var{x} and @var{y}. They are evaluated for
15362 horizontal and vertical chroma subsample values of the output
15363 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
15367 the number of input frame, starting from 0
15370 the position in the file of the input frame, NAN if unknown
15373 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
15377 This filter also supports the @ref{framesync} options.
15379 Note that the @var{n}, @var{pos}, @var{t} variables are available only
15380 when evaluation is done @emph{per frame}, and will evaluate to NAN
15381 when @option{eval} is set to @samp{init}.
15383 Be aware that frames are taken from each input video in timestamp
15384 order, hence, if their initial timestamps differ, it is a good idea
15385 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
15386 have them begin in the same zero timestamp, as the example for
15387 the @var{movie} filter does.
15389 You can chain together more overlays but you should test the
15390 efficiency of such approach.
15392 @subsection Commands
15394 This filter supports the following commands:
15398 Modify the x and y of the overlay input.
15399 The command accepts the same syntax of the corresponding option.
15401 If the specified expression is not valid, it is kept at its current
15405 @subsection Examples
15409 Draw the overlay at 10 pixels from the bottom right corner of the main
15412 overlay=main_w-overlay_w-10:main_h-overlay_h-10
15415 Using named options the example above becomes:
15417 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
15421 Insert a transparent PNG logo in the bottom left corner of the input,
15422 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
15424 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
15428 Insert 2 different transparent PNG logos (second logo on bottom
15429 right corner) using the @command{ffmpeg} tool:
15431 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
15435 Add a transparent color layer on top of the main video; @code{WxH}
15436 must specify the size of the main input to the overlay filter:
15438 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
15442 Play an original video and a filtered version (here with the deshake
15443 filter) side by side using the @command{ffplay} tool:
15445 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
15448 The above command is the same as:
15450 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
15454 Make a sliding overlay appearing from the left to the right top part of the
15455 screen starting since time 2:
15457 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
15461 Compose output by putting two input videos side to side:
15463 ffmpeg -i left.avi -i right.avi -filter_complex "
15464 nullsrc=size=200x100 [background];
15465 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
15466 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
15467 [background][left] overlay=shortest=1 [background+left];
15468 [background+left][right] overlay=shortest=1:x=100 [left+right]
15473 Mask 10-20 seconds of a video by applying the delogo filter to a section
15475 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
15476 -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]'
15481 Chain several overlays in cascade:
15483 nullsrc=s=200x200 [bg];
15484 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
15485 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
15486 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
15487 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
15488 [in3] null, [mid2] overlay=100:100 [out0]
15493 @anchor{overlay_cuda}
15494 @section overlay_cuda
15496 Overlay one video on top of another.
15498 This is the CUDA variant of the @ref{overlay} filter.
15499 It only accepts CUDA frames. The underlying input pixel formats have to match.
15501 It takes two inputs and has one output. The first input is the "main"
15502 video on which the second input is overlaid.
15504 It accepts the following parameters:
15509 Set the x and y coordinates of the overlaid video on the main video.
15510 Default value is "0" for both expressions.
15513 See @ref{framesync}.
15516 See @ref{framesync}.
15519 See @ref{framesync}.
15523 This filter also supports the @ref{framesync} options.
15527 Apply Overcomplete Wavelet denoiser.
15529 The filter accepts the following options:
15535 Larger depth values will denoise lower frequency components more, but
15536 slow down filtering.
15538 Must be an int in the range 8-16, default is @code{8}.
15540 @item luma_strength, ls
15543 Must be a double value in the range 0-1000, default is @code{1.0}.
15545 @item chroma_strength, cs
15546 Set chroma strength.
15548 Must be a double value in the range 0-1000, default is @code{1.0}.
15554 Add paddings to the input image, and place the original input at the
15555 provided @var{x}, @var{y} coordinates.
15557 It accepts the following parameters:
15562 Specify an expression for the size of the output image with the
15563 paddings added. If the value for @var{width} or @var{height} is 0, the
15564 corresponding input size is used for the output.
15566 The @var{width} expression can reference the value set by the
15567 @var{height} expression, and vice versa.
15569 The default value of @var{width} and @var{height} is 0.
15573 Specify the offsets to place the input image at within the padded area,
15574 with respect to the top/left border of the output image.
15576 The @var{x} expression can reference the value set by the @var{y}
15577 expression, and vice versa.
15579 The default value of @var{x} and @var{y} is 0.
15581 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
15582 so the input image is centered on the padded area.
15585 Specify the color of the padded area. For the syntax of this option,
15586 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
15587 manual,ffmpeg-utils}.
15589 The default value of @var{color} is "black".
15592 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
15594 It accepts the following values:
15598 Only evaluate expressions once during the filter initialization or when
15599 a command is processed.
15602 Evaluate expressions for each incoming frame.
15606 Default value is @samp{init}.
15609 Pad to aspect instead to a resolution.
15613 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
15614 options are expressions containing the following constants:
15619 The input video width and height.
15623 These are the same as @var{in_w} and @var{in_h}.
15627 The output width and height (the size of the padded area), as
15628 specified by the @var{width} and @var{height} expressions.
15632 These are the same as @var{out_w} and @var{out_h}.
15636 The x and y offsets as specified by the @var{x} and @var{y}
15637 expressions, or NAN if not yet specified.
15640 same as @var{iw} / @var{ih}
15643 input sample aspect ratio
15646 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
15650 The horizontal and vertical chroma subsample values. For example for the
15651 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15654 @subsection Examples
15658 Add paddings with the color "violet" to the input video. The output video
15659 size is 640x480, and the top-left corner of the input video is placed at
15662 pad=640:480:0:40:violet
15665 The example above is equivalent to the following command:
15667 pad=width=640:height=480:x=0:y=40:color=violet
15671 Pad the input to get an output with dimensions increased by 3/2,
15672 and put the input video at the center of the padded area:
15674 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
15678 Pad the input to get a squared output with size equal to the maximum
15679 value between the input width and height, and put the input video at
15680 the center of the padded area:
15682 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
15686 Pad the input to get a final w/h ratio of 16:9:
15688 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
15692 In case of anamorphic video, in order to set the output display aspect
15693 correctly, it is necessary to use @var{sar} in the expression,
15694 according to the relation:
15696 (ih * X / ih) * sar = output_dar
15697 X = output_dar / sar
15700 Thus the previous example needs to be modified to:
15702 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
15706 Double the output size and put the input video in the bottom-right
15707 corner of the output padded area:
15709 pad="2*iw:2*ih:ow-iw:oh-ih"
15713 @anchor{palettegen}
15714 @section palettegen
15716 Generate one palette for a whole video stream.
15718 It accepts the following options:
15722 Set the maximum number of colors to quantize in the palette.
15723 Note: the palette will still contain 256 colors; the unused palette entries
15726 @item reserve_transparent
15727 Create a palette of 255 colors maximum and reserve the last one for
15728 transparency. Reserving the transparency color is useful for GIF optimization.
15729 If not set, the maximum of colors in the palette will be 256. You probably want
15730 to disable this option for a standalone image.
15733 @item transparency_color
15734 Set the color that will be used as background for transparency.
15737 Set statistics mode.
15739 It accepts the following values:
15742 Compute full frame histograms.
15744 Compute histograms only for the part that differs from previous frame. This
15745 might be relevant to give more importance to the moving part of your input if
15746 the background is static.
15748 Compute new histogram for each frame.
15751 Default value is @var{full}.
15754 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
15755 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
15756 color quantization of the palette. This information is also visible at
15757 @var{info} logging level.
15759 @subsection Examples
15763 Generate a representative palette of a given video using @command{ffmpeg}:
15765 ffmpeg -i input.mkv -vf palettegen palette.png
15769 @section paletteuse
15771 Use a palette to downsample an input video stream.
15773 The filter takes two inputs: one video stream and a palette. The palette must
15774 be a 256 pixels image.
15776 It accepts the following options:
15780 Select dithering mode. Available algorithms are:
15783 Ordered 8x8 bayer dithering (deterministic)
15785 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
15786 Note: this dithering is sometimes considered "wrong" and is included as a
15788 @item floyd_steinberg
15789 Floyd and Steingberg dithering (error diffusion)
15791 Frankie Sierra dithering v2 (error diffusion)
15793 Frankie Sierra dithering v2 "Lite" (error diffusion)
15796 Default is @var{sierra2_4a}.
15799 When @var{bayer} dithering is selected, this option defines the scale of the
15800 pattern (how much the crosshatch pattern is visible). A low value means more
15801 visible pattern for less banding, and higher value means less visible pattern
15802 at the cost of more banding.
15804 The option must be an integer value in the range [0,5]. Default is @var{2}.
15807 If set, define the zone to process
15811 Only the changing rectangle will be reprocessed. This is similar to GIF
15812 cropping/offsetting compression mechanism. This option can be useful for speed
15813 if only a part of the image is changing, and has use cases such as limiting the
15814 scope of the error diffusal @option{dither} to the rectangle that bounds the
15815 moving scene (it leads to more deterministic output if the scene doesn't change
15816 much, and as a result less moving noise and better GIF compression).
15819 Default is @var{none}.
15822 Take new palette for each output frame.
15824 @item alpha_threshold
15825 Sets the alpha threshold for transparency. Alpha values above this threshold
15826 will be treated as completely opaque, and values below this threshold will be
15827 treated as completely transparent.
15829 The option must be an integer value in the range [0,255]. Default is @var{128}.
15832 @subsection Examples
15836 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
15837 using @command{ffmpeg}:
15839 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
15843 @section perspective
15845 Correct perspective of video not recorded perpendicular to the screen.
15847 A description of the accepted parameters follows.
15858 Set coordinates expression for top left, top right, bottom left and bottom right corners.
15859 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
15860 If the @code{sense} option is set to @code{source}, then the specified points will be sent
15861 to the corners of the destination. If the @code{sense} option is set to @code{destination},
15862 then the corners of the source will be sent to the specified coordinates.
15864 The expressions can use the following variables:
15869 the width and height of video frame.
15873 Output frame count.
15876 @item interpolation
15877 Set interpolation for perspective correction.
15879 It accepts the following values:
15885 Default value is @samp{linear}.
15888 Set interpretation of coordinate options.
15890 It accepts the following values:
15894 Send point in the source specified by the given coordinates to
15895 the corners of the destination.
15897 @item 1, destination
15899 Send the corners of the source to the point in the destination specified
15900 by the given coordinates.
15902 Default value is @samp{source}.
15906 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
15908 It accepts the following values:
15911 only evaluate expressions once during the filter initialization or
15912 when a command is processed
15915 evaluate expressions for each incoming frame
15918 Default value is @samp{init}.
15923 Delay interlaced video by one field time so that the field order changes.
15925 The intended use is to fix PAL movies that have been captured with the
15926 opposite field order to the film-to-video transfer.
15928 A description of the accepted parameters follows.
15934 It accepts the following values:
15937 Capture field order top-first, transfer bottom-first.
15938 Filter will delay the bottom field.
15941 Capture field order bottom-first, transfer top-first.
15942 Filter will delay the top field.
15945 Capture and transfer with the same field order. This mode only exists
15946 for the documentation of the other options to refer to, but if you
15947 actually select it, the filter will faithfully do nothing.
15950 Capture field order determined automatically by field flags, transfer
15952 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
15953 basis using field flags. If no field information is available,
15954 then this works just like @samp{u}.
15957 Capture unknown or varying, transfer opposite.
15958 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
15959 analyzing the images and selecting the alternative that produces best
15960 match between the fields.
15963 Capture top-first, transfer unknown or varying.
15964 Filter selects among @samp{t} and @samp{p} using image analysis.
15967 Capture bottom-first, transfer unknown or varying.
15968 Filter selects among @samp{b} and @samp{p} using image analysis.
15971 Capture determined by field flags, transfer unknown or varying.
15972 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
15973 image analysis. If no field information is available, then this works just
15974 like @samp{U}. This is the default mode.
15977 Both capture and transfer unknown or varying.
15978 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
15982 @subsection Commands
15984 This filter supports the all above options as @ref{commands}.
15986 @section photosensitivity
15987 Reduce various flashes in video, so to help users with epilepsy.
15989 It accepts the following options:
15992 Set how many frames to use when filtering. Default is 30.
15995 Set detection threshold factor. Default is 1.
15999 Set how many pixels to skip when sampling frames. Default is 1.
16000 Allowed range is from 1 to 1024.
16003 Leave frames unchanged. Default is disabled.
16006 @section pixdesctest
16008 Pixel format descriptor test filter, mainly useful for internal
16009 testing. The output video should be equal to the input video.
16013 format=monow, pixdesctest
16016 can be used to test the monowhite pixel format descriptor definition.
16020 Display sample values of color channels. Mainly useful for checking color
16021 and levels. Minimum supported resolution is 640x480.
16023 The filters accept the following options:
16027 Set scope X position, relative offset on X axis.
16030 Set scope Y position, relative offset on Y axis.
16039 Set window opacity. This window also holds statistics about pixel area.
16042 Set window X position, relative offset on X axis.
16045 Set window Y position, relative offset on Y axis.
16050 Enable the specified chain of postprocessing subfilters using libpostproc. This
16051 library should be automatically selected with a GPL build (@code{--enable-gpl}).
16052 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
16053 Each subfilter and some options have a short and a long name that can be used
16054 interchangeably, i.e. dr/dering are the same.
16056 The filters accept the following options:
16060 Set postprocessing subfilters string.
16063 All subfilters share common options to determine their scope:
16067 Honor the quality commands for this subfilter.
16070 Do chrominance filtering, too (default).
16073 Do luminance filtering only (no chrominance).
16076 Do chrominance filtering only (no luminance).
16079 These options can be appended after the subfilter name, separated by a '|'.
16081 Available subfilters are:
16084 @item hb/hdeblock[|difference[|flatness]]
16085 Horizontal deblocking filter
16088 Difference factor where higher values mean more deblocking (default: @code{32}).
16090 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16093 @item vb/vdeblock[|difference[|flatness]]
16094 Vertical deblocking filter
16097 Difference factor where higher values mean more deblocking (default: @code{32}).
16099 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16102 @item ha/hadeblock[|difference[|flatness]]
16103 Accurate horizontal deblocking filter
16106 Difference factor where higher values mean more deblocking (default: @code{32}).
16108 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16111 @item va/vadeblock[|difference[|flatness]]
16112 Accurate vertical deblocking filter
16115 Difference factor where higher values mean more deblocking (default: @code{32}).
16117 Flatness threshold where lower values mean more deblocking (default: @code{39}).
16121 The horizontal and vertical deblocking filters share the difference and
16122 flatness values so you cannot set different horizontal and vertical
16126 @item h1/x1hdeblock
16127 Experimental horizontal deblocking filter
16129 @item v1/x1vdeblock
16130 Experimental vertical deblocking filter
16135 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
16138 larger -> stronger filtering
16140 larger -> stronger filtering
16142 larger -> stronger filtering
16145 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
16148 Stretch luminance to @code{0-255}.
16151 @item lb/linblenddeint
16152 Linear blend deinterlacing filter that deinterlaces the given block by
16153 filtering all lines with a @code{(1 2 1)} filter.
16155 @item li/linipoldeint
16156 Linear interpolating deinterlacing filter that deinterlaces the given block by
16157 linearly interpolating every second line.
16159 @item ci/cubicipoldeint
16160 Cubic interpolating deinterlacing filter deinterlaces the given block by
16161 cubically interpolating every second line.
16163 @item md/mediandeint
16164 Median deinterlacing filter that deinterlaces the given block by applying a
16165 median filter to every second line.
16167 @item fd/ffmpegdeint
16168 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
16169 second line with a @code{(-1 4 2 4 -1)} filter.
16172 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
16173 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
16175 @item fq/forceQuant[|quantizer]
16176 Overrides the quantizer table from the input with the constant quantizer you
16184 Default pp filter combination (@code{hb|a,vb|a,dr|a})
16187 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
16190 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
16193 @subsection Examples
16197 Apply horizontal and vertical deblocking, deringing and automatic
16198 brightness/contrast:
16204 Apply default filters without brightness/contrast correction:
16210 Apply default filters and temporal denoiser:
16212 pp=default/tmpnoise|1|2|3
16216 Apply deblocking on luminance only, and switch vertical deblocking on or off
16217 automatically depending on available CPU time:
16224 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
16225 similar to spp = 6 with 7 point DCT, where only the center sample is
16228 The filter accepts the following options:
16232 Force a constant quantization parameter. It accepts an integer in range
16233 0 to 63. If not set, the filter will use the QP from the video stream
16237 Set thresholding mode. Available modes are:
16241 Set hard thresholding.
16243 Set soft thresholding (better de-ringing effect, but likely blurrier).
16245 Set medium thresholding (good results, default).
16249 @section premultiply
16250 Apply alpha premultiply effect to input video stream using first plane
16251 of second stream as alpha.
16253 Both streams must have same dimensions and same pixel format.
16255 The filter accepts the following option:
16259 Set which planes will be processed, unprocessed planes will be copied.
16260 By default value 0xf, all planes will be processed.
16263 Do not require 2nd input for processing, instead use alpha plane from input stream.
16267 Apply prewitt operator to input video stream.
16269 The filter accepts the following option:
16273 Set which planes will be processed, unprocessed planes will be copied.
16274 By default value 0xf, all planes will be processed.
16277 Set value which will be multiplied with filtered result.
16280 Set value which will be added to filtered result.
16283 @subsection Commands
16285 This filter supports the all above options as @ref{commands}.
16287 @section pseudocolor
16289 Alter frame colors in video with pseudocolors.
16291 This filter accepts the following options:
16295 set pixel first component expression
16298 set pixel second component expression
16301 set pixel third component expression
16304 set pixel fourth component expression, corresponds to the alpha component
16307 set component to use as base for altering colors
16310 Pick one of built-in LUTs. By default is set to none.
16326 Each of them specifies the expression to use for computing the lookup table for
16327 the corresponding pixel component values.
16329 The expressions can contain the following constants and functions:
16334 The input width and height.
16337 The input value for the pixel component.
16339 @item ymin, umin, vmin, amin
16340 The minimum allowed component value.
16342 @item ymax, umax, vmax, amax
16343 The maximum allowed component value.
16346 All expressions default to "val".
16348 @subsection Commands
16350 This filter supports the all above options as @ref{commands}.
16352 @subsection Examples
16356 Change too high luma values to gradient:
16358 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'"
16364 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
16365 Ratio) between two input videos.
16367 This filter takes in input two input videos, the first input is
16368 considered the "main" source and is passed unchanged to the
16369 output. The second input is used as a "reference" video for computing
16372 Both video inputs must have the same resolution and pixel format for
16373 this filter to work correctly. Also it assumes that both inputs
16374 have the same number of frames, which are compared one by one.
16376 The obtained average PSNR is printed through the logging system.
16378 The filter stores the accumulated MSE (mean squared error) of each
16379 frame, and at the end of the processing it is averaged across all frames
16380 equally, and the following formula is applied to obtain the PSNR:
16383 PSNR = 10*log10(MAX^2/MSE)
16386 Where MAX is the average of the maximum values of each component of the
16389 The description of the accepted parameters follows.
16392 @item stats_file, f
16393 If specified the filter will use the named file to save the PSNR of
16394 each individual frame. When filename equals "-" the data is sent to
16397 @item stats_version
16398 Specifies which version of the stats file format to use. Details of
16399 each format are written below.
16400 Default value is 1.
16402 @item stats_add_max
16403 Determines whether the max value is output to the stats log.
16404 Default value is 0.
16405 Requires stats_version >= 2. If this is set and stats_version < 2,
16406 the filter will return an error.
16409 This filter also supports the @ref{framesync} options.
16411 The file printed if @var{stats_file} is selected, contains a sequence of
16412 key/value pairs of the form @var{key}:@var{value} for each compared
16415 If a @var{stats_version} greater than 1 is specified, a header line precedes
16416 the list of per-frame-pair stats, with key value pairs following the frame
16417 format with the following parameters:
16420 @item psnr_log_version
16421 The version of the log file format. Will match @var{stats_version}.
16424 A comma separated list of the per-frame-pair parameters included in
16428 A description of each shown per-frame-pair parameter follows:
16432 sequential number of the input frame, starting from 1
16435 Mean Square Error pixel-by-pixel average difference of the compared
16436 frames, averaged over all the image components.
16438 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
16439 Mean Square Error pixel-by-pixel average difference of the compared
16440 frames for the component specified by the suffix.
16442 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
16443 Peak Signal to Noise ratio of the compared frames for the component
16444 specified by the suffix.
16446 @item max_avg, max_y, max_u, max_v
16447 Maximum allowed value for each channel, and average over all
16451 @subsection Examples
16456 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16457 [main][ref] psnr="stats_file=stats.log" [out]
16460 On this example the input file being processed is compared with the
16461 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
16462 is stored in @file{stats.log}.
16465 Another example with different containers:
16467 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 -
16474 Pulldown reversal (inverse telecine) filter, capable of handling mixed
16475 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
16478 The pullup filter is designed to take advantage of future context in making
16479 its decisions. This filter is stateless in the sense that it does not lock
16480 onto a pattern to follow, but it instead looks forward to the following
16481 fields in order to identify matches and rebuild progressive frames.
16483 To produce content with an even framerate, insert the fps filter after
16484 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
16485 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
16487 The filter accepts the following options:
16494 These options set the amount of "junk" to ignore at the left, right, top, and
16495 bottom of the image, respectively. Left and right are in units of 8 pixels,
16496 while top and bottom are in units of 2 lines.
16497 The default is 8 pixels on each side.
16500 Set the strict breaks. Setting this option to 1 will reduce the chances of
16501 filter generating an occasional mismatched frame, but it may also cause an
16502 excessive number of frames to be dropped during high motion sequences.
16503 Conversely, setting it to -1 will make filter match fields more easily.
16504 This may help processing of video where there is slight blurring between
16505 the fields, but may also cause there to be interlaced frames in the output.
16506 Default value is @code{0}.
16509 Set the metric plane to use. It accepts the following values:
16515 Use chroma blue plane.
16518 Use chroma red plane.
16521 This option may be set to use chroma plane instead of the default luma plane
16522 for doing filter's computations. This may improve accuracy on very clean
16523 source material, but more likely will decrease accuracy, especially if there
16524 is chroma noise (rainbow effect) or any grayscale video.
16525 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
16526 load and make pullup usable in realtime on slow machines.
16529 For best results (without duplicated frames in the output file) it is
16530 necessary to change the output frame rate. For example, to inverse
16531 telecine NTSC input:
16533 ffmpeg -i input -vf pullup -r 24000/1001 ...
16538 Change video quantization parameters (QP).
16540 The filter accepts the following option:
16544 Set expression for quantization parameter.
16547 The expression is evaluated through the eval API and can contain, among others,
16548 the following constants:
16552 1 if index is not 129, 0 otherwise.
16555 Sequential index starting from -129 to 128.
16558 @subsection Examples
16562 Some equation like:
16570 Flush video frames from internal cache of frames into a random order.
16571 No frame is discarded.
16572 Inspired by @ref{frei0r} nervous filter.
16576 Set size in number of frames of internal cache, in range from @code{2} to
16577 @code{512}. Default is @code{30}.
16580 Set seed for random number generator, must be an integer included between
16581 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16582 less than @code{0}, the filter will try to use a good random seed on a
16586 @section readeia608
16588 Read closed captioning (EIA-608) information from the top lines of a video frame.
16590 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
16591 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
16592 with EIA-608 data (starting from 0). A description of each metadata value follows:
16595 @item lavfi.readeia608.X.cc
16596 The two bytes stored as EIA-608 data (printed in hexadecimal).
16598 @item lavfi.readeia608.X.line
16599 The number of the line on which the EIA-608 data was identified and read.
16602 This filter accepts the following options:
16606 Set the line to start scanning for EIA-608 data. Default is @code{0}.
16609 Set the line to end scanning for EIA-608 data. Default is @code{29}.
16612 Set the ratio of width reserved for sync code detection.
16613 Default is @code{0.27}. Allowed range is @code{[0.1 - 0.7]}.
16616 Enable checking the parity bit. In the event of a parity error, the filter will output
16617 @code{0x00} for that character. Default is false.
16620 Lowpass lines prior to further processing. Default is enabled.
16623 @subsection Commands
16625 This filter supports the all above options as @ref{commands}.
16627 @subsection Examples
16631 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
16633 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
16639 Read vertical interval timecode (VITC) information from the top lines of a
16642 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
16643 timecode value, if a valid timecode has been detected. Further metadata key
16644 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
16645 timecode data has been found or not.
16647 This filter accepts the following options:
16651 Set the maximum number of lines to scan for VITC data. If the value is set to
16652 @code{-1} the full video frame is scanned. Default is @code{45}.
16655 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
16656 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
16659 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
16660 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
16663 @subsection Examples
16667 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
16668 draw @code{--:--:--:--} as a placeholder:
16670 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
16676 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
16678 Destination pixel at position (X, Y) will be picked from source (x, y) position
16679 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
16680 value for pixel will be used for destination pixel.
16682 Xmap and Ymap input video streams must be of same dimensions. Output video stream
16683 will have Xmap/Ymap video stream dimensions.
16684 Xmap and Ymap input video streams are 16bit depth, single channel.
16688 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
16689 Default is @code{color}.
16692 Specify the color of the unmapped pixels. For the syntax of this option,
16693 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16694 manual,ffmpeg-utils}. Default color is @code{black}.
16697 @section removegrain
16699 The removegrain filter is a spatial denoiser for progressive video.
16703 Set mode for the first plane.
16706 Set mode for the second plane.
16709 Set mode for the third plane.
16712 Set mode for the fourth plane.
16715 Range of mode is from 0 to 24. Description of each mode follows:
16719 Leave input plane unchanged. Default.
16722 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
16725 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
16728 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
16731 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
16732 This is equivalent to a median filter.
16735 Line-sensitive clipping giving the minimal change.
16738 Line-sensitive clipping, intermediate.
16741 Line-sensitive clipping, intermediate.
16744 Line-sensitive clipping, intermediate.
16747 Line-sensitive clipping on a line where the neighbours pixels are the closest.
16750 Replaces the target pixel with the closest neighbour.
16753 [1 2 1] horizontal and vertical kernel blur.
16759 Bob mode, interpolates top field from the line where the neighbours
16760 pixels are the closest.
16763 Bob mode, interpolates bottom field from the line where the neighbours
16764 pixels are the closest.
16767 Bob mode, interpolates top field. Same as 13 but with a more complicated
16768 interpolation formula.
16771 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
16772 interpolation formula.
16775 Clips the pixel with the minimum and maximum of respectively the maximum and
16776 minimum of each pair of opposite neighbour pixels.
16779 Line-sensitive clipping using opposite neighbours whose greatest distance from
16780 the current pixel is minimal.
16783 Replaces the pixel with the average of its 8 neighbours.
16786 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
16789 Clips pixels using the averages of opposite neighbour.
16792 Same as mode 21 but simpler and faster.
16795 Small edge and halo removal, but reputed useless.
16801 @section removelogo
16803 Suppress a TV station logo, using an image file to determine which
16804 pixels comprise the logo. It works by filling in the pixels that
16805 comprise the logo with neighboring pixels.
16807 The filter accepts the following options:
16811 Set the filter bitmap file, which can be any image format supported by
16812 libavformat. The width and height of the image file must match those of the
16813 video stream being processed.
16816 Pixels in the provided bitmap image with a value of zero are not
16817 considered part of the logo, non-zero pixels are considered part of
16818 the logo. If you use white (255) for the logo and black (0) for the
16819 rest, you will be safe. For making the filter bitmap, it is
16820 recommended to take a screen capture of a black frame with the logo
16821 visible, and then using a threshold filter followed by the erode
16822 filter once or twice.
16824 If needed, little splotches can be fixed manually. Remember that if
16825 logo pixels are not covered, the filter quality will be much
16826 reduced. Marking too many pixels as part of the logo does not hurt as
16827 much, but it will increase the amount of blurring needed to cover over
16828 the image and will destroy more information than necessary, and extra
16829 pixels will slow things down on a large logo.
16831 @section repeatfields
16833 This filter uses the repeat_field flag from the Video ES headers and hard repeats
16834 fields based on its value.
16838 Reverse a video clip.
16840 Warning: This filter requires memory to buffer the entire clip, so trimming
16843 @subsection Examples
16847 Take the first 5 seconds of a clip, and reverse it.
16854 Shift R/G/B/A pixels horizontally and/or vertically.
16856 The filter accepts the following options:
16859 Set amount to shift red horizontally.
16861 Set amount to shift red vertically.
16863 Set amount to shift green horizontally.
16865 Set amount to shift green vertically.
16867 Set amount to shift blue horizontally.
16869 Set amount to shift blue vertically.
16871 Set amount to shift alpha horizontally.
16873 Set amount to shift alpha vertically.
16875 Set edge mode, can be @var{smear}, default, or @var{warp}.
16878 @subsection Commands
16880 This filter supports the all above options as @ref{commands}.
16883 Apply roberts cross operator to input video stream.
16885 The filter accepts the following option:
16889 Set which planes will be processed, unprocessed planes will be copied.
16890 By default value 0xf, all planes will be processed.
16893 Set value which will be multiplied with filtered result.
16896 Set value which will be added to filtered result.
16899 @subsection Commands
16901 This filter supports the all above options as @ref{commands}.
16905 Rotate video by an arbitrary angle expressed in radians.
16907 The filter accepts the following options:
16909 A description of the optional parameters follows.
16912 Set an expression for the angle by which to rotate the input video
16913 clockwise, expressed as a number of radians. A negative value will
16914 result in a counter-clockwise rotation. By default it is set to "0".
16916 This expression is evaluated for each frame.
16919 Set the output width expression, default value is "iw".
16920 This expression is evaluated just once during configuration.
16923 Set the output height expression, default value is "ih".
16924 This expression is evaluated just once during configuration.
16927 Enable bilinear interpolation if set to 1, a value of 0 disables
16928 it. Default value is 1.
16931 Set the color used to fill the output area not covered by the rotated
16932 image. For the general syntax of this option, check the
16933 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16934 If the special value "none" is selected then no
16935 background is printed (useful for example if the background is never shown).
16937 Default value is "black".
16940 The expressions for the angle and the output size can contain the
16941 following constants and functions:
16945 sequential number of the input frame, starting from 0. It is always NAN
16946 before the first frame is filtered.
16949 time in seconds of the input frame, it is set to 0 when the filter is
16950 configured. It is always NAN before the first frame is filtered.
16954 horizontal and vertical chroma subsample values. For example for the
16955 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
16959 the input video width and height
16963 the output width and height, that is the size of the padded area as
16964 specified by the @var{width} and @var{height} expressions
16968 the minimal width/height required for completely containing the input
16969 video rotated by @var{a} radians.
16971 These are only available when computing the @option{out_w} and
16972 @option{out_h} expressions.
16975 @subsection Examples
16979 Rotate the input by PI/6 radians clockwise:
16985 Rotate the input by PI/6 radians counter-clockwise:
16991 Rotate the input by 45 degrees clockwise:
16997 Apply a constant rotation with period T, starting from an angle of PI/3:
16999 rotate=PI/3+2*PI*t/T
17003 Make the input video rotation oscillating with a period of T
17004 seconds and an amplitude of A radians:
17006 rotate=A*sin(2*PI/T*t)
17010 Rotate the video, output size is chosen so that the whole rotating
17011 input video is always completely contained in the output:
17013 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
17017 Rotate the video, reduce the output size so that no background is ever
17020 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
17024 @subsection Commands
17026 The filter supports the following commands:
17030 Set the angle expression.
17031 The command accepts the same syntax of the corresponding option.
17033 If the specified expression is not valid, it is kept at its current
17039 Apply Shape Adaptive Blur.
17041 The filter accepts the following options:
17044 @item luma_radius, lr
17045 Set luma blur filter strength, must be a value in range 0.1-4.0, default
17046 value is 1.0. A greater value will result in a more blurred image, and
17047 in slower processing.
17049 @item luma_pre_filter_radius, lpfr
17050 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
17053 @item luma_strength, ls
17054 Set luma maximum difference between pixels to still be considered, must
17055 be a value in the 0.1-100.0 range, default value is 1.0.
17057 @item chroma_radius, cr
17058 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
17059 greater value will result in a more blurred image, and in slower
17062 @item chroma_pre_filter_radius, cpfr
17063 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
17065 @item chroma_strength, cs
17066 Set chroma maximum difference between pixels to still be considered,
17067 must be a value in the -0.9-100.0 range.
17070 Each chroma option value, if not explicitly specified, is set to the
17071 corresponding luma option value.
17076 Scale (resize) the input video, using the libswscale library.
17078 The scale filter forces the output display aspect ratio to be the same
17079 of the input, by changing the output sample aspect ratio.
17081 If the input image format is different from the format requested by
17082 the next filter, the scale filter will convert the input to the
17085 @subsection Options
17086 The filter accepts the following options, or any of the options
17087 supported by the libswscale scaler.
17089 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
17090 the complete list of scaler options.
17095 Set the output video dimension expression. Default value is the input
17098 If the @var{width} or @var{w} value is 0, the input width is used for
17099 the output. If the @var{height} or @var{h} value is 0, the input height
17100 is used for the output.
17102 If one and only one of the values is -n with n >= 1, the scale filter
17103 will use a value that maintains the aspect ratio of the input image,
17104 calculated from the other specified dimension. After that it will,
17105 however, make sure that the calculated dimension is divisible by n and
17106 adjust the value if necessary.
17108 If both values are -n with n >= 1, the behavior will be identical to
17109 both values being set to 0 as previously detailed.
17111 See below for the list of accepted constants for use in the dimension
17115 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
17119 Only evaluate expressions once during the filter initialization or when a command is processed.
17122 Evaluate expressions for each incoming frame.
17126 Default value is @samp{init}.
17130 Set the interlacing mode. It accepts the following values:
17134 Force interlaced aware scaling.
17137 Do not apply interlaced scaling.
17140 Select interlaced aware scaling depending on whether the source frames
17141 are flagged as interlaced or not.
17144 Default value is @samp{0}.
17147 Set libswscale scaling flags. See
17148 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17149 complete list of values. If not explicitly specified the filter applies
17153 @item param0, param1
17154 Set libswscale input parameters for scaling algorithms that need them. See
17155 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
17156 complete documentation. If not explicitly specified the filter applies
17162 Set the video size. For the syntax of this option, check the
17163 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17165 @item in_color_matrix
17166 @item out_color_matrix
17167 Set in/output YCbCr color space type.
17169 This allows the autodetected value to be overridden as well as allows forcing
17170 a specific value used for the output and encoder.
17172 If not specified, the color space type depends on the pixel format.
17178 Choose automatically.
17181 Format conforming to International Telecommunication Union (ITU)
17182 Recommendation BT.709.
17185 Set color space conforming to the United States Federal Communications
17186 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
17191 Set color space conforming to:
17195 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
17198 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
17201 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
17206 Set color space conforming to SMPTE ST 240:1999.
17209 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
17214 Set in/output YCbCr sample range.
17216 This allows the autodetected value to be overridden as well as allows forcing
17217 a specific value used for the output and encoder. If not specified, the
17218 range depends on the pixel format. Possible values:
17222 Choose automatically.
17225 Set full range (0-255 in case of 8-bit luma).
17227 @item mpeg/limited/tv
17228 Set "MPEG" range (16-235 in case of 8-bit luma).
17231 @item force_original_aspect_ratio
17232 Enable decreasing or increasing output video width or height if necessary to
17233 keep the original aspect ratio. Possible values:
17237 Scale the video as specified and disable this feature.
17240 The output video dimensions will automatically be decreased if needed.
17243 The output video dimensions will automatically be increased if needed.
17247 One useful instance of this option is that when you know a specific device's
17248 maximum allowed resolution, you can use this to limit the output video to
17249 that, while retaining the aspect ratio. For example, device A allows
17250 1280x720 playback, and your video is 1920x800. Using this option (set it to
17251 decrease) and specifying 1280x720 to the command line makes the output
17254 Please note that this is a different thing than specifying -1 for @option{w}
17255 or @option{h}, you still need to specify the output resolution for this option
17258 @item force_divisible_by
17259 Ensures that both the output dimensions, width and height, are divisible by the
17260 given integer when used together with @option{force_original_aspect_ratio}. This
17261 works similar to using @code{-n} in the @option{w} and @option{h} options.
17263 This option respects the value set for @option{force_original_aspect_ratio},
17264 increasing or decreasing the resolution accordingly. The video's aspect ratio
17265 may be slightly modified.
17267 This option can be handy if you need to have a video fit within or exceed
17268 a defined resolution using @option{force_original_aspect_ratio} but also have
17269 encoder restrictions on width or height divisibility.
17273 The values of the @option{w} and @option{h} options are expressions
17274 containing the following constants:
17279 The input width and height
17283 These are the same as @var{in_w} and @var{in_h}.
17287 The output (scaled) width and height
17291 These are the same as @var{out_w} and @var{out_h}
17294 The same as @var{iw} / @var{ih}
17297 input sample aspect ratio
17300 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17304 horizontal and vertical input chroma subsample values. For example for the
17305 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17309 horizontal and vertical output chroma subsample values. For example for the
17310 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17313 The (sequential) number of the input frame, starting from 0.
17314 Only available with @code{eval=frame}.
17317 The presentation timestamp of the input frame, expressed as a number of
17318 seconds. Only available with @code{eval=frame}.
17321 The position (byte offset) of the frame in the input stream, or NaN if
17322 this information is unavailable and/or meaningless (for example in case of synthetic video).
17323 Only available with @code{eval=frame}.
17326 @subsection Examples
17330 Scale the input video to a size of 200x100
17335 This is equivalent to:
17346 Specify a size abbreviation for the output size:
17351 which can also be written as:
17357 Scale the input to 2x:
17359 scale=w=2*iw:h=2*ih
17363 The above is the same as:
17365 scale=2*in_w:2*in_h
17369 Scale the input to 2x with forced interlaced scaling:
17371 scale=2*iw:2*ih:interl=1
17375 Scale the input to half size:
17377 scale=w=iw/2:h=ih/2
17381 Increase the width, and set the height to the same size:
17387 Seek Greek harmony:
17394 Increase the height, and set the width to 3/2 of the height:
17396 scale=w=3/2*oh:h=3/5*ih
17400 Increase the size, making the size a multiple of the chroma
17403 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
17407 Increase the width to a maximum of 500 pixels,
17408 keeping the same aspect ratio as the input:
17410 scale=w='min(500\, iw*3/2):h=-1'
17414 Make pixels square by combining scale and setsar:
17416 scale='trunc(ih*dar):ih',setsar=1/1
17420 Make pixels square by combining scale and setsar,
17421 making sure the resulting resolution is even (required by some codecs):
17423 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
17427 @subsection Commands
17429 This filter supports the following commands:
17433 Set the output video dimension expression.
17434 The command accepts the same syntax of the corresponding option.
17436 If the specified expression is not valid, it is kept at its current
17442 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
17443 format conversion on CUDA video frames. Setting the output width and height
17444 works in the same way as for the @var{scale} filter.
17446 The following additional options are accepted:
17449 The pixel format of the output CUDA frames. If set to the string "same" (the
17450 default), the input format will be kept. Note that automatic format negotiation
17451 and conversion is not yet supported for hardware frames
17454 The interpolation algorithm used for resizing. One of the following:
17461 @item cubic2p_bspline
17462 2-parameter cubic (B=1, C=0)
17464 @item cubic2p_catmullrom
17465 2-parameter cubic (B=0, C=1/2)
17467 @item cubic2p_b05c03
17468 2-parameter cubic (B=1/2, C=3/10)
17476 @item force_original_aspect_ratio
17477 Enable decreasing or increasing output video width or height if necessary to
17478 keep the original aspect ratio. Possible values:
17482 Scale the video as specified and disable this feature.
17485 The output video dimensions will automatically be decreased if needed.
17488 The output video dimensions will automatically be increased if needed.
17492 One useful instance of this option is that when you know a specific device's
17493 maximum allowed resolution, you can use this to limit the output video to
17494 that, while retaining the aspect ratio. For example, device A allows
17495 1280x720 playback, and your video is 1920x800. Using this option (set it to
17496 decrease) and specifying 1280x720 to the command line makes the output
17499 Please note that this is a different thing than specifying -1 for @option{w}
17500 or @option{h}, you still need to specify the output resolution for this option
17503 @item force_divisible_by
17504 Ensures that both the output dimensions, width and height, are divisible by the
17505 given integer when used together with @option{force_original_aspect_ratio}. This
17506 works similar to using @code{-n} in the @option{w} and @option{h} options.
17508 This option respects the value set for @option{force_original_aspect_ratio},
17509 increasing or decreasing the resolution accordingly. The video's aspect ratio
17510 may be slightly modified.
17512 This option can be handy if you need to have a video fit within or exceed
17513 a defined resolution using @option{force_original_aspect_ratio} but also have
17514 encoder restrictions on width or height divisibility.
17520 Scale (resize) the input video, based on a reference video.
17522 See the scale filter for available options, scale2ref supports the same but
17523 uses the reference video instead of the main input as basis. scale2ref also
17524 supports the following additional constants for the @option{w} and
17525 @option{h} options:
17530 The main input video's width and height
17533 The same as @var{main_w} / @var{main_h}
17536 The main input video's sample aspect ratio
17538 @item main_dar, mdar
17539 The main input video's display aspect ratio. Calculated from
17540 @code{(main_w / main_h) * main_sar}.
17544 The main input video's horizontal and vertical chroma subsample values.
17545 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
17549 The (sequential) number of the main input frame, starting from 0.
17550 Only available with @code{eval=frame}.
17553 The presentation timestamp of the main input frame, expressed as a number of
17554 seconds. Only available with @code{eval=frame}.
17557 The position (byte offset) of the frame in the main input stream, or NaN if
17558 this information is unavailable and/or meaningless (for example in case of synthetic video).
17559 Only available with @code{eval=frame}.
17562 @subsection Examples
17566 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
17568 'scale2ref[b][a];[a][b]overlay'
17572 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
17574 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
17578 @subsection Commands
17580 This filter supports the following commands:
17584 Set the output video dimension expression.
17585 The command accepts the same syntax of the corresponding option.
17587 If the specified expression is not valid, it is kept at its current
17592 Scroll input video horizontally and/or vertically by constant speed.
17594 The filter accepts the following options:
17596 @item horizontal, h
17597 Set the horizontal scrolling speed. Default is 0. Allowed range is from -1 to 1.
17598 Negative values changes scrolling direction.
17601 Set the vertical scrolling speed. Default is 0. Allowed range is from -1 to 1.
17602 Negative values changes scrolling direction.
17605 Set the initial horizontal scrolling position. Default is 0. Allowed range is from 0 to 1.
17608 Set the initial vertical scrolling position. Default is 0. Allowed range is from 0 to 1.
17611 @subsection Commands
17613 This filter supports the following @ref{commands}:
17615 @item horizontal, h
17616 Set the horizontal scrolling speed.
17618 Set the vertical scrolling speed.
17624 Detect video scene change.
17626 This filter sets frame metadata with mafd between frame, the scene score, and
17627 forward the frame to the next filter, so they can use these metadata to detect
17628 scene change or others.
17630 In addition, this filter logs a message and sets frame metadata when it detects
17631 a scene change by @option{threshold}.
17633 @code{lavfi.scd.mafd} metadata keys are set with mafd for every frame.
17635 @code{lavfi.scd.score} metadata keys are set with scene change score for every frame
17636 to detect scene change.
17638 @code{lavfi.scd.time} metadata keys are set with current filtered frame time which
17639 detect scene change with @option{threshold}.
17641 The filter accepts the following options:
17645 Set the scene change detection threshold as a percentage of maximum change. Good
17646 values are in the @code{[8.0, 14.0]} range. The range for @option{threshold} is
17649 Default value is @code{10.}.
17652 Set the flag to pass scene change frames to the next filter. Default value is @code{0}
17653 You can enable it if you want to get snapshot of scene change frames only.
17656 @anchor{selectivecolor}
17657 @section selectivecolor
17659 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
17660 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
17661 by the "purity" of the color (that is, how saturated it already is).
17663 This filter is similar to the Adobe Photoshop Selective Color tool.
17665 The filter accepts the following options:
17668 @item correction_method
17669 Select color correction method.
17671 Available values are:
17674 Specified adjustments are applied "as-is" (added/subtracted to original pixel
17677 Specified adjustments are relative to the original component value.
17679 Default is @code{absolute}.
17681 Adjustments for red pixels (pixels where the red component is the maximum)
17683 Adjustments for yellow pixels (pixels where the blue component is the minimum)
17685 Adjustments for green pixels (pixels where the green component is the maximum)
17687 Adjustments for cyan pixels (pixels where the red component is the minimum)
17689 Adjustments for blue pixels (pixels where the blue component is the maximum)
17691 Adjustments for magenta pixels (pixels where the green component is the minimum)
17693 Adjustments for white pixels (pixels where all components are greater than 128)
17695 Adjustments for all pixels except pure black and pure white
17697 Adjustments for black pixels (pixels where all components are lesser than 128)
17699 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
17702 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
17703 4 space separated floating point adjustment values in the [-1,1] range,
17704 respectively to adjust the amount of cyan, magenta, yellow and black for the
17705 pixels of its range.
17707 @subsection Examples
17711 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
17712 increase magenta by 27% in blue areas:
17714 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
17718 Use a Photoshop selective color preset:
17720 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
17724 @anchor{separatefields}
17725 @section separatefields
17727 The @code{separatefields} takes a frame-based video input and splits
17728 each frame into its components fields, producing a new half height clip
17729 with twice the frame rate and twice the frame count.
17731 This filter use field-dominance information in frame to decide which
17732 of each pair of fields to place first in the output.
17733 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
17735 @section setdar, setsar
17737 The @code{setdar} filter sets the Display Aspect Ratio for the filter
17740 This is done by changing the specified Sample (aka Pixel) Aspect
17741 Ratio, according to the following equation:
17743 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
17746 Keep in mind that the @code{setdar} filter does not modify the pixel
17747 dimensions of the video frame. Also, the display aspect ratio set by
17748 this filter may be changed by later filters in the filterchain,
17749 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
17752 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
17753 the filter output video.
17755 Note that as a consequence of the application of this filter, the
17756 output display aspect ratio will change according to the equation
17759 Keep in mind that the sample aspect ratio set by the @code{setsar}
17760 filter may be changed by later filters in the filterchain, e.g. if
17761 another "setsar" or a "setdar" filter is applied.
17763 It accepts the following parameters:
17766 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
17767 Set the aspect ratio used by the filter.
17769 The parameter can be a floating point number string, an expression, or
17770 a string of the form @var{num}:@var{den}, where @var{num} and
17771 @var{den} are the numerator and denominator of the aspect ratio. If
17772 the parameter is not specified, it is assumed the value "0".
17773 In case the form "@var{num}:@var{den}" is used, the @code{:} character
17777 Set the maximum integer value to use for expressing numerator and
17778 denominator when reducing the expressed aspect ratio to a rational.
17779 Default value is @code{100}.
17783 The parameter @var{sar} is an expression containing
17784 the following constants:
17788 These are approximated values for the mathematical constants e
17789 (Euler's number), pi (Greek pi), and phi (the golden ratio).
17792 The input width and height.
17795 These are the same as @var{w} / @var{h}.
17798 The input sample aspect ratio.
17801 The input display aspect ratio. It is the same as
17802 (@var{w} / @var{h}) * @var{sar}.
17805 Horizontal and vertical chroma subsample values. For example, for the
17806 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17809 @subsection Examples
17814 To change the display aspect ratio to 16:9, specify one of the following:
17821 To change the sample aspect ratio to 10:11, specify:
17827 To set a display aspect ratio of 16:9, and specify a maximum integer value of
17828 1000 in the aspect ratio reduction, use the command:
17830 setdar=ratio=16/9:max=1000
17838 Force field for the output video frame.
17840 The @code{setfield} filter marks the interlace type field for the
17841 output frames. It does not change the input frame, but only sets the
17842 corresponding property, which affects how the frame is treated by
17843 following filters (e.g. @code{fieldorder} or @code{yadif}).
17845 The filter accepts the following options:
17850 Available values are:
17854 Keep the same field property.
17857 Mark the frame as bottom-field-first.
17860 Mark the frame as top-field-first.
17863 Mark the frame as progressive.
17870 Force frame parameter for the output video frame.
17872 The @code{setparams} filter marks interlace and color range for the
17873 output frames. It does not change the input frame, but only sets the
17874 corresponding property, which affects how the frame is treated by
17879 Available values are:
17883 Keep the same field property (default).
17886 Mark the frame as bottom-field-first.
17889 Mark the frame as top-field-first.
17892 Mark the frame as progressive.
17896 Available values are:
17900 Keep the same color range property (default).
17902 @item unspecified, unknown
17903 Mark the frame as unspecified color range.
17905 @item limited, tv, mpeg
17906 Mark the frame as limited range.
17908 @item full, pc, jpeg
17909 Mark the frame as full range.
17912 @item color_primaries
17913 Set the color primaries.
17914 Available values are:
17918 Keep the same color primaries property (default).
17935 Set the color transfer.
17936 Available values are:
17940 Keep the same color trc property (default).
17962 Set the colorspace.
17963 Available values are:
17967 Keep the same colorspace property (default).
17980 @item chroma-derived-nc
17981 @item chroma-derived-c
17987 Apply shear transform to input video.
17989 This filter supports the following options:
17993 Shear factor in X-direction. Default value is 0.
17994 Allowed range is from -2 to 2.
17997 Shear factor in Y-direction. Default value is 0.
17998 Allowed range is from -2 to 2.
18001 Set the color used to fill the output area not covered by the transformed
18002 video. For the general syntax of this option, check the
18003 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
18004 If the special value "none" is selected then no
18005 background is printed (useful for example if the background is never shown).
18007 Default value is "black".
18010 Set interpolation type. Can be @code{bilinear} or @code{nearest}. Default is @code{bilinear}.
18013 @subsection Commands
18015 This filter supports the all above options as @ref{commands}.
18019 Show a line containing various information for each input video frame.
18020 The input video is not modified.
18022 This filter supports the following options:
18026 Calculate checksums of each plane. By default enabled.
18029 The shown line contains a sequence of key/value pairs of the form
18030 @var{key}:@var{value}.
18032 The following values are shown in the output:
18036 The (sequential) number of the input frame, starting from 0.
18039 The Presentation TimeStamp of the input frame, expressed as a number of
18040 time base units. The time base unit depends on the filter input pad.
18043 The Presentation TimeStamp of the input frame, expressed as a number of
18047 The position of the frame in the input stream, or -1 if this information is
18048 unavailable and/or meaningless (for example in case of synthetic video).
18051 The pixel format name.
18054 The sample aspect ratio of the input frame, expressed in the form
18055 @var{num}/@var{den}.
18058 The size of the input frame. For the syntax of this option, check the
18059 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18062 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
18063 for bottom field first).
18066 This is 1 if the frame is a key frame, 0 otherwise.
18069 The picture type of the input frame ("I" for an I-frame, "P" for a
18070 P-frame, "B" for a B-frame, or "?" for an unknown type).
18071 Also refer to the documentation of the @code{AVPictureType} enum and of
18072 the @code{av_get_picture_type_char} function defined in
18073 @file{libavutil/avutil.h}.
18076 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
18078 @item plane_checksum
18079 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
18080 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
18083 The mean value of pixels in each plane of the input frame, expressed in the form
18084 "[@var{mean0} @var{mean1} @var{mean2} @var{mean3}]".
18087 The standard deviation of pixel values in each plane of the input frame, expressed
18088 in the form "[@var{stdev0} @var{stdev1} @var{stdev2} @var{stdev3}]".
18092 @section showpalette
18094 Displays the 256 colors palette of each frame. This filter is only relevant for
18095 @var{pal8} pixel format frames.
18097 It accepts the following option:
18101 Set the size of the box used to represent one palette color entry. Default is
18102 @code{30} (for a @code{30x30} pixel box).
18105 @section shuffleframes
18107 Reorder and/or duplicate and/or drop video frames.
18109 It accepts the following parameters:
18113 Set the destination indexes of input frames.
18114 This is space or '|' separated list of indexes that maps input frames to output
18115 frames. Number of indexes also sets maximal value that each index may have.
18116 '-1' index have special meaning and that is to drop frame.
18119 The first frame has the index 0. The default is to keep the input unchanged.
18121 @subsection Examples
18125 Swap second and third frame of every three frames of the input:
18127 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
18131 Swap 10th and 1st frame of every ten frames of the input:
18133 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
18137 @section shufflepixels
18139 Reorder pixels in video frames.
18141 This filter accepts the following options:
18145 Set shuffle direction. Can be forward or inverse direction.
18146 Default direction is forward.
18149 Set shuffle mode. Can be horizontal, vertical or block mode.
18153 Set shuffle block_size. In case of horizontal shuffle mode only width
18154 part of size is used, and in case of vertical shuffle mode only height
18155 part of size is used.
18158 Set random seed used with shuffling pixels. Mainly useful to set to be able
18159 to reverse filtering process to get original input.
18160 For example, to reverse forward shuffle you need to use same parameters
18161 and exact same seed and to set direction to inverse.
18164 @section shuffleplanes
18166 Reorder and/or duplicate video planes.
18168 It accepts the following parameters:
18173 The index of the input plane to be used as the first output plane.
18176 The index of the input plane to be used as the second output plane.
18179 The index of the input plane to be used as the third output plane.
18182 The index of the input plane to be used as the fourth output plane.
18186 The first plane has the index 0. The default is to keep the input unchanged.
18188 @subsection Examples
18192 Swap the second and third planes of the input:
18194 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
18198 @anchor{signalstats}
18199 @section signalstats
18200 Evaluate various visual metrics that assist in determining issues associated
18201 with the digitization of analog video media.
18203 By default the filter will log these metadata values:
18207 Display the minimal Y value contained within the input frame. Expressed in
18211 Display the Y value at the 10% percentile within the input frame. Expressed in
18215 Display the average Y value within the input frame. Expressed in range of
18219 Display the Y value at the 90% percentile within the input frame. Expressed in
18223 Display the maximum Y value contained within the input frame. Expressed in
18227 Display the minimal U value contained within the input frame. Expressed in
18231 Display the U value at the 10% percentile within the input frame. Expressed in
18235 Display the average U value within the input frame. Expressed in range of
18239 Display the U value at the 90% percentile within the input frame. Expressed in
18243 Display the maximum U value contained within the input frame. Expressed in
18247 Display the minimal V value contained within the input frame. Expressed in
18251 Display the V value at the 10% percentile within the input frame. Expressed in
18255 Display the average V value within the input frame. Expressed in range of
18259 Display the V value at the 90% percentile within the input frame. Expressed in
18263 Display the maximum V value contained within the input frame. Expressed in
18267 Display the minimal saturation value contained within the input frame.
18268 Expressed in range of [0-~181.02].
18271 Display the saturation value at the 10% percentile within the input frame.
18272 Expressed in range of [0-~181.02].
18275 Display the average saturation value within the input frame. Expressed in range
18279 Display the saturation value at the 90% percentile within the input frame.
18280 Expressed in range of [0-~181.02].
18283 Display the maximum saturation value contained within the input frame.
18284 Expressed in range of [0-~181.02].
18287 Display the median value for hue within the input frame. Expressed in range of
18291 Display the average value for hue within the input frame. Expressed in range of
18295 Display the average of sample value difference between all values of the Y
18296 plane in the current frame and corresponding values of the previous input frame.
18297 Expressed in range of [0-255].
18300 Display the average of sample value difference between all values of the U
18301 plane in the current frame and corresponding values of the previous input frame.
18302 Expressed in range of [0-255].
18305 Display the average of sample value difference between all values of the V
18306 plane in the current frame and corresponding values of the previous input frame.
18307 Expressed in range of [0-255].
18310 Display bit depth of Y plane in current frame.
18311 Expressed in range of [0-16].
18314 Display bit depth of U plane in current frame.
18315 Expressed in range of [0-16].
18318 Display bit depth of V plane in current frame.
18319 Expressed in range of [0-16].
18322 The filter accepts the following options:
18328 @option{stat} specify an additional form of image analysis.
18329 @option{out} output video with the specified type of pixel highlighted.
18331 Both options accept the following values:
18335 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
18336 unlike the neighboring pixels of the same field. Examples of temporal outliers
18337 include the results of video dropouts, head clogs, or tape tracking issues.
18340 Identify @var{vertical line repetition}. Vertical line repetition includes
18341 similar rows of pixels within a frame. In born-digital video vertical line
18342 repetition is common, but this pattern is uncommon in video digitized from an
18343 analog source. When it occurs in video that results from the digitization of an
18344 analog source it can indicate concealment from a dropout compensator.
18347 Identify pixels that fall outside of legal broadcast range.
18351 Set the highlight color for the @option{out} option. The default color is
18355 @subsection Examples
18359 Output data of various video metrics:
18361 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
18365 Output specific data about the minimum and maximum values of the Y plane per frame:
18367 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
18371 Playback video while highlighting pixels that are outside of broadcast range in red.
18373 ffplay example.mov -vf signalstats="out=brng:color=red"
18377 Playback video with signalstats metadata drawn over the frame.
18379 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
18382 The contents of signalstat_drawtext.txt used in the command are:
18385 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
18386 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
18387 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
18388 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
18396 Calculates the MPEG-7 Video Signature. The filter can handle more than one
18397 input. In this case the matching between the inputs can be calculated additionally.
18398 The filter always passes through the first input. The signature of each stream can
18399 be written into a file.
18401 It accepts the following options:
18405 Enable or disable the matching process.
18407 Available values are:
18411 Disable the calculation of a matching (default).
18413 Calculate the matching for the whole video and output whether the whole video
18414 matches or only parts.
18416 Calculate only until a matching is found or the video ends. Should be faster in
18421 Set the number of inputs. The option value must be a non negative integer.
18422 Default value is 1.
18425 Set the path to which the output is written. If there is more than one input,
18426 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
18427 integer), that will be replaced with the input number. If no filename is
18428 specified, no output will be written. This is the default.
18431 Choose the output format.
18433 Available values are:
18437 Use the specified binary representation (default).
18439 Use the specified xml representation.
18443 Set threshold to detect one word as similar. The option value must be an integer
18444 greater than zero. The default value is 9000.
18447 Set threshold to detect all words as similar. The option value must be an integer
18448 greater than zero. The default value is 60000.
18451 Set threshold to detect frames as similar. The option value must be an integer
18452 greater than zero. The default value is 116.
18455 Set the minimum length of a sequence in frames to recognize it as matching
18456 sequence. The option value must be a non negative integer value.
18457 The default value is 0.
18460 Set the minimum relation, that matching frames to all frames must have.
18461 The option value must be a double value between 0 and 1. The default value is 0.5.
18464 @subsection Examples
18468 To calculate the signature of an input video and store it in signature.bin:
18470 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
18474 To detect whether two videos match and store the signatures in XML format in
18475 signature0.xml and signature1.xml:
18477 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 -
18485 Blur the input video without impacting the outlines.
18487 It accepts the following options:
18490 @item luma_radius, lr
18491 Set the luma radius. The option value must be a float number in
18492 the range [0.1,5.0] that specifies the variance of the gaussian filter
18493 used to blur the image (slower if larger). Default value is 1.0.
18495 @item luma_strength, ls
18496 Set the luma strength. The option value must be a float number
18497 in the range [-1.0,1.0] that configures the blurring. A value included
18498 in [0.0,1.0] will blur the image whereas a value included in
18499 [-1.0,0.0] will sharpen the image. Default value is 1.0.
18501 @item luma_threshold, lt
18502 Set the luma threshold used as a coefficient to determine
18503 whether a pixel should be blurred or not. The option value must be an
18504 integer in the range [-30,30]. A value of 0 will filter all the image,
18505 a value included in [0,30] will filter flat areas and a value included
18506 in [-30,0] will filter edges. Default value is 0.
18508 @item chroma_radius, cr
18509 Set the chroma radius. The option value must be a float number in
18510 the range [0.1,5.0] that specifies the variance of the gaussian filter
18511 used to blur the image (slower if larger). Default value is @option{luma_radius}.
18513 @item chroma_strength, cs
18514 Set the chroma strength. The option value must be a float number
18515 in the range [-1.0,1.0] that configures the blurring. A value included
18516 in [0.0,1.0] will blur the image whereas a value included in
18517 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
18519 @item chroma_threshold, ct
18520 Set the chroma threshold used as a coefficient to determine
18521 whether a pixel should be blurred or not. The option value must be an
18522 integer in the range [-30,30]. A value of 0 will filter all the image,
18523 a value included in [0,30] will filter flat areas and a value included
18524 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
18527 If a chroma option is not explicitly set, the corresponding luma value
18531 Apply sobel operator to input video stream.
18533 The filter accepts the following option:
18537 Set which planes will be processed, unprocessed planes will be copied.
18538 By default value 0xf, all planes will be processed.
18541 Set value which will be multiplied with filtered result.
18544 Set value which will be added to filtered result.
18547 @subsection Commands
18549 This filter supports the all above options as @ref{commands}.
18554 Apply a simple postprocessing filter that compresses and decompresses the image
18555 at several (or - in the case of @option{quality} level @code{6} - all) shifts
18556 and average the results.
18558 The filter accepts the following options:
18562 Set quality. This option defines the number of levels for averaging. It accepts
18563 an integer in the range 0-6. If set to @code{0}, the filter will have no
18564 effect. A value of @code{6} means the higher quality. For each increment of
18565 that value the speed drops by a factor of approximately 2. Default value is
18569 Force a constant quantization parameter. If not set, the filter will use the QP
18570 from the video stream (if available).
18573 Set thresholding mode. Available modes are:
18577 Set hard thresholding (default).
18579 Set soft thresholding (better de-ringing effect, but likely blurrier).
18582 @item use_bframe_qp
18583 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
18584 option may cause flicker since the B-Frames have often larger QP. Default is
18585 @code{0} (not enabled).
18588 @subsection Commands
18590 This filter supports the following commands:
18592 @item quality, level
18593 Set quality level. The value @code{max} can be used to set the maximum level,
18594 currently @code{6}.
18600 Scale the input by applying one of the super-resolution methods based on
18601 convolutional neural networks. Supported models:
18605 Super-Resolution Convolutional Neural Network model (SRCNN).
18606 See @url{https://arxiv.org/abs/1501.00092}.
18609 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
18610 See @url{https://arxiv.org/abs/1609.05158}.
18613 Training scripts as well as scripts for model file (.pb) saving can be found at
18614 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
18615 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
18617 Native model files (.model) can be generated from TensorFlow model
18618 files (.pb) by using tools/python/convert.py
18620 The filter accepts the following options:
18624 Specify which DNN backend to use for model loading and execution. This option accepts
18625 the following values:
18629 Native implementation of DNN loading and execution.
18632 TensorFlow backend. To enable this backend you
18633 need to install the TensorFlow for C library (see
18634 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
18635 @code{--enable-libtensorflow}
18638 Default value is @samp{native}.
18641 Set path to model file specifying network architecture and its parameters.
18642 Note that different backends use different file formats. TensorFlow backend
18643 can load files for both formats, while native backend can load files for only
18647 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
18648 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
18649 input upscaled using bicubic upscaling with proper scale factor.
18652 This feature can also be finished with @ref{dnn_processing} filter.
18656 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
18658 This filter takes in input two input videos, the first input is
18659 considered the "main" source and is passed unchanged to the
18660 output. The second input is used as a "reference" video for computing
18663 Both video inputs must have the same resolution and pixel format for
18664 this filter to work correctly. Also it assumes that both inputs
18665 have the same number of frames, which are compared one by one.
18667 The filter stores the calculated SSIM of each frame.
18669 The description of the accepted parameters follows.
18672 @item stats_file, f
18673 If specified the filter will use the named file to save the SSIM of
18674 each individual frame. When filename equals "-" the data is sent to
18678 The file printed if @var{stats_file} is selected, contains a sequence of
18679 key/value pairs of the form @var{key}:@var{value} for each compared
18682 A description of each shown parameter follows:
18686 sequential number of the input frame, starting from 1
18688 @item Y, U, V, R, G, B
18689 SSIM of the compared frames for the component specified by the suffix.
18692 SSIM of the compared frames for the whole frame.
18695 Same as above but in dB representation.
18698 This filter also supports the @ref{framesync} options.
18700 @subsection Examples
18705 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
18706 [main][ref] ssim="stats_file=stats.log" [out]
18709 On this example the input file being processed is compared with the
18710 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
18711 is stored in @file{stats.log}.
18714 Another example with both psnr and ssim at same time:
18716 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
18720 Another example with different containers:
18722 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 -
18728 Convert between different stereoscopic image formats.
18730 The filters accept the following options:
18734 Set stereoscopic image format of input.
18736 Available values for input image formats are:
18739 side by side parallel (left eye left, right eye right)
18742 side by side crosseye (right eye left, left eye right)
18745 side by side parallel with half width resolution
18746 (left eye left, right eye right)
18749 side by side crosseye with half width resolution
18750 (right eye left, left eye right)
18754 above-below (left eye above, right eye below)
18758 above-below (right eye above, left eye below)
18762 above-below with half height resolution
18763 (left eye above, right eye below)
18767 above-below with half height resolution
18768 (right eye above, left eye below)
18771 alternating frames (left eye first, right eye second)
18774 alternating frames (right eye first, left eye second)
18777 interleaved rows (left eye has top row, right eye starts on next row)
18780 interleaved rows (right eye has top row, left eye starts on next row)
18783 interleaved columns, left eye first
18786 interleaved columns, right eye first
18788 Default value is @samp{sbsl}.
18792 Set stereoscopic image format of output.
18796 side by side parallel (left eye left, right eye right)
18799 side by side crosseye (right eye left, left eye right)
18802 side by side parallel with half width resolution
18803 (left eye left, right eye right)
18806 side by side crosseye with half width resolution
18807 (right eye left, left eye right)
18811 above-below (left eye above, right eye below)
18815 above-below (right eye above, left eye below)
18819 above-below with half height resolution
18820 (left eye above, right eye below)
18824 above-below with half height resolution
18825 (right eye above, left eye below)
18828 alternating frames (left eye first, right eye second)
18831 alternating frames (right eye first, left eye second)
18834 interleaved rows (left eye has top row, right eye starts on next row)
18837 interleaved rows (right eye has top row, left eye starts on next row)
18840 anaglyph red/blue gray
18841 (red filter on left eye, blue filter on right eye)
18844 anaglyph red/green gray
18845 (red filter on left eye, green filter on right eye)
18848 anaglyph red/cyan gray
18849 (red filter on left eye, cyan filter on right eye)
18852 anaglyph red/cyan half colored
18853 (red filter on left eye, cyan filter on right eye)
18856 anaglyph red/cyan color
18857 (red filter on left eye, cyan filter on right eye)
18860 anaglyph red/cyan color optimized with the least squares projection of dubois
18861 (red filter on left eye, cyan filter on right eye)
18864 anaglyph green/magenta gray
18865 (green filter on left eye, magenta filter on right eye)
18868 anaglyph green/magenta half colored
18869 (green filter on left eye, magenta filter on right eye)
18872 anaglyph green/magenta colored
18873 (green filter on left eye, magenta filter on right eye)
18876 anaglyph green/magenta color optimized with the least squares projection of dubois
18877 (green filter on left eye, magenta filter on right eye)
18880 anaglyph yellow/blue gray
18881 (yellow filter on left eye, blue filter on right eye)
18884 anaglyph yellow/blue half colored
18885 (yellow filter on left eye, blue filter on right eye)
18888 anaglyph yellow/blue colored
18889 (yellow filter on left eye, blue filter on right eye)
18892 anaglyph yellow/blue color optimized with the least squares projection of dubois
18893 (yellow filter on left eye, blue filter on right eye)
18896 mono output (left eye only)
18899 mono output (right eye only)
18902 checkerboard, left eye first
18905 checkerboard, right eye first
18908 interleaved columns, left eye first
18911 interleaved columns, right eye first
18917 Default value is @samp{arcd}.
18920 @subsection Examples
18924 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
18930 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
18936 @section streamselect, astreamselect
18937 Select video or audio streams.
18939 The filter accepts the following options:
18943 Set number of inputs. Default is 2.
18946 Set input indexes to remap to outputs.
18949 @subsection Commands
18951 The @code{streamselect} and @code{astreamselect} filter supports the following
18956 Set input indexes to remap to outputs.
18959 @subsection Examples
18963 Select first 5 seconds 1st stream and rest of time 2nd stream:
18965 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
18969 Same as above, but for audio:
18971 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
18978 Draw subtitles on top of input video using the libass library.
18980 To enable compilation of this filter you need to configure FFmpeg with
18981 @code{--enable-libass}. This filter also requires a build with libavcodec and
18982 libavformat to convert the passed subtitles file to ASS (Advanced Substation
18983 Alpha) subtitles format.
18985 The filter accepts the following options:
18989 Set the filename of the subtitle file to read. It must be specified.
18991 @item original_size
18992 Specify the size of the original video, the video for which the ASS file
18993 was composed. For the syntax of this option, check the
18994 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18995 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
18996 correctly scale the fonts if the aspect ratio has been changed.
18999 Set a directory path containing fonts that can be used by the filter.
19000 These fonts will be used in addition to whatever the font provider uses.
19003 Process alpha channel, by default alpha channel is untouched.
19006 Set subtitles input character encoding. @code{subtitles} filter only. Only
19007 useful if not UTF-8.
19009 @item stream_index, si
19010 Set subtitles stream index. @code{subtitles} filter only.
19013 Override default style or script info parameters of the subtitles. It accepts a
19014 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
19017 If the first key is not specified, it is assumed that the first value
19018 specifies the @option{filename}.
19020 For example, to render the file @file{sub.srt} on top of the input
19021 video, use the command:
19026 which is equivalent to:
19028 subtitles=filename=sub.srt
19031 To render the default subtitles stream from file @file{video.mkv}, use:
19033 subtitles=video.mkv
19036 To render the second subtitles stream from that file, use:
19038 subtitles=video.mkv:si=1
19041 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
19042 @code{DejaVu Serif}, use:
19044 subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
19047 @section super2xsai
19049 Scale the input by 2x and smooth using the Super2xSaI (Scale and
19050 Interpolate) pixel art scaling algorithm.
19052 Useful for enlarging pixel art images without reducing sharpness.
19056 Swap two rectangular objects in video.
19058 This filter accepts the following options:
19068 Set 1st rect x coordinate.
19071 Set 1st rect y coordinate.
19074 Set 2nd rect x coordinate.
19077 Set 2nd rect y coordinate.
19079 All expressions are evaluated once for each frame.
19082 The all options are expressions containing the following constants:
19087 The input width and height.
19090 same as @var{w} / @var{h}
19093 input sample aspect ratio
19096 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
19099 The number of the input frame, starting from 0.
19102 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
19105 the position in the file of the input frame, NAN if unknown
19112 Blend successive video frames.
19118 Apply telecine process to the video.
19120 This filter accepts the following options:
19129 The default value is @code{top}.
19133 A string of numbers representing the pulldown pattern you wish to apply.
19134 The default value is @code{23}.
19138 Some typical patterns:
19143 24p: 2332 (preferred)
19150 24p: 222222222223 ("Euro pulldown")
19155 @section thistogram
19157 Compute and draw a color distribution histogram for the input video across time.
19159 Unlike @ref{histogram} video filter which only shows histogram of single input frame
19160 at certain time, this filter shows also past histograms of number of frames defined
19161 by @code{width} option.
19163 The computed histogram is a representation of the color component
19164 distribution in an image.
19166 The filter accepts the following options:
19170 Set width of single color component output. Default value is @code{0}.
19171 Value of @code{0} means width will be picked from input video.
19172 This also set number of passed histograms to keep.
19173 Allowed range is [0, 8192].
19175 @item display_mode, d
19177 It accepts the following values:
19180 Per color component graphs are placed below each other.
19183 Per color component graphs are placed side by side.
19186 Presents information identical to that in the @code{parade}, except
19187 that the graphs representing color components are superimposed directly
19190 Default is @code{stack}.
19192 @item levels_mode, m
19193 Set mode. Can be either @code{linear}, or @code{logarithmic}.
19194 Default is @code{linear}.
19196 @item components, c
19197 Set what color components to display.
19198 Default is @code{7}.
19201 Set background opacity. Default is @code{0.9}.
19204 Show envelope. Default is disabled.
19207 Set envelope color. Default is @code{gold}.
19212 Available values for slide is:
19215 Draw new frame when right border is reached.
19218 Replace old columns with new ones.
19221 Scroll from right to left.
19224 Scroll from left to right.
19227 Draw single picture.
19230 Default is @code{replace}.
19235 Apply threshold effect to video stream.
19237 This filter needs four video streams to perform thresholding.
19238 First stream is stream we are filtering.
19239 Second stream is holding threshold values, third stream is holding min values,
19240 and last, fourth stream is holding max values.
19242 The filter accepts the following option:
19246 Set which planes will be processed, unprocessed planes will be copied.
19247 By default value 0xf, all planes will be processed.
19250 For example if first stream pixel's component value is less then threshold value
19251 of pixel component from 2nd threshold stream, third stream value will picked,
19252 otherwise fourth stream pixel component value will be picked.
19254 Using color source filter one can perform various types of thresholding:
19256 @subsection Examples
19260 Binary threshold, using gray color as threshold:
19262 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
19266 Inverted binary threshold, using gray color as threshold:
19268 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
19272 Truncate binary threshold, using gray color as threshold:
19274 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
19278 Threshold to zero, using gray color as threshold:
19280 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
19284 Inverted threshold to zero, using gray color as threshold:
19286 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
19291 Select the most representative frame in a given sequence of consecutive frames.
19293 The filter accepts the following options:
19297 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
19298 will pick one of them, and then handle the next batch of @var{n} frames until
19299 the end. Default is @code{100}.
19302 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
19303 value will result in a higher memory usage, so a high value is not recommended.
19305 @subsection Examples
19309 Extract one picture each 50 frames:
19315 Complete example of a thumbnail creation with @command{ffmpeg}:
19317 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
19324 Tile several successive frames together.
19326 The @ref{untile} filter can do the reverse.
19328 The filter accepts the following options:
19333 Set the grid size (i.e. the number of lines and columns). For the syntax of
19334 this option, check the
19335 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19338 Set the maximum number of frames to render in the given area. It must be less
19339 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
19340 the area will be used.
19343 Set the outer border margin in pixels.
19346 Set the inner border thickness (i.e. the number of pixels between frames). For
19347 more advanced padding options (such as having different values for the edges),
19348 refer to the pad video filter.
19351 Specify the color of the unused area. For the syntax of this option, check the
19352 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19353 The default value of @var{color} is "black".
19356 Set the number of frames to overlap when tiling several successive frames together.
19357 The value must be between @code{0} and @var{nb_frames - 1}.
19360 Set the number of frames to initially be empty before displaying first output frame.
19361 This controls how soon will one get first output frame.
19362 The value must be between @code{0} and @var{nb_frames - 1}.
19365 @subsection Examples
19369 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
19371 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
19373 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
19374 duplicating each output frame to accommodate the originally detected frame
19378 Display @code{5} pictures in an area of @code{3x2} frames,
19379 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
19380 mixed flat and named options:
19382 tile=3x2:nb_frames=5:padding=7:margin=2
19386 @section tinterlace
19388 Perform various types of temporal field interlacing.
19390 Frames are counted starting from 1, so the first input frame is
19393 The filter accepts the following options:
19398 Specify the mode of the interlacing. This option can also be specified
19399 as a value alone. See below for a list of values for this option.
19401 Available values are:
19405 Move odd frames into the upper field, even into the lower field,
19406 generating a double height frame at half frame rate.
19410 Frame 1 Frame 2 Frame 3 Frame 4
19412 11111 22222 33333 44444
19413 11111 22222 33333 44444
19414 11111 22222 33333 44444
19415 11111 22222 33333 44444
19429 Only output odd frames, even frames are dropped, generating a frame with
19430 unchanged height at half frame rate.
19435 Frame 1 Frame 2 Frame 3 Frame 4
19437 11111 22222 33333 44444
19438 11111 22222 33333 44444
19439 11111 22222 33333 44444
19440 11111 22222 33333 44444
19450 Only output even frames, odd frames are dropped, generating a frame with
19451 unchanged height at half frame rate.
19456 Frame 1 Frame 2 Frame 3 Frame 4
19458 11111 22222 33333 44444
19459 11111 22222 33333 44444
19460 11111 22222 33333 44444
19461 11111 22222 33333 44444
19471 Expand each frame to full height, but pad alternate lines with black,
19472 generating a frame with double height at the same input 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
19485 11111 ..... 33333 .....
19486 ..... 22222 ..... 44444
19487 11111 ..... 33333 .....
19488 ..... 22222 ..... 44444
19489 11111 ..... 33333 .....
19490 ..... 22222 ..... 44444
19491 11111 ..... 33333 .....
19492 ..... 22222 ..... 44444
19496 @item interleave_top, 4
19497 Interleave the upper field from odd frames with the lower field from
19498 even frames, generating a frame with unchanged height at half frame rate.
19503 Frame 1 Frame 2 Frame 3 Frame 4
19505 11111<- 22222 33333<- 44444
19506 11111 22222<- 33333 44444<-
19507 11111<- 22222 33333<- 44444
19508 11111 22222<- 33333 44444<-
19518 @item interleave_bottom, 5
19519 Interleave the lower field from odd frames with the upper field from
19520 even frames, generating a frame with unchanged height at half frame rate.
19525 Frame 1 Frame 2 Frame 3 Frame 4
19527 11111 22222<- 33333 44444<-
19528 11111<- 22222 33333<- 44444
19529 11111 22222<- 33333 44444<-
19530 11111<- 22222 33333<- 44444
19540 @item interlacex2, 6
19541 Double frame rate with unchanged height. Frames are inserted each
19542 containing the second temporal field from the previous input frame and
19543 the first temporal field from the next input frame. This mode relies on
19544 the top_field_first flag. Useful for interlaced video displays with no
19545 field synchronisation.
19550 Frame 1 Frame 2 Frame 3 Frame 4
19552 11111 22222 33333 44444
19553 11111 22222 33333 44444
19554 11111 22222 33333 44444
19555 11111 22222 33333 44444
19558 11111 22222 22222 33333 33333 44444 44444
19559 11111 11111 22222 22222 33333 33333 44444
19560 11111 22222 22222 33333 33333 44444 44444
19561 11111 11111 22222 22222 33333 33333 44444
19566 Move odd frames into the upper field, even into the lower field,
19567 generating a double height frame at same frame rate.
19572 Frame 1 Frame 2 Frame 3 Frame 4
19574 11111 22222 33333 44444
19575 11111 22222 33333 44444
19576 11111 22222 33333 44444
19577 11111 22222 33333 44444
19580 11111 33333 33333 55555
19581 22222 22222 44444 44444
19582 11111 33333 33333 55555
19583 22222 22222 44444 44444
19584 11111 33333 33333 55555
19585 22222 22222 44444 44444
19586 11111 33333 33333 55555
19587 22222 22222 44444 44444
19592 Numeric values are deprecated but are accepted for backward
19593 compatibility reasons.
19595 Default mode is @code{merge}.
19598 Specify flags influencing the filter process.
19600 Available value for @var{flags} is:
19603 @item low_pass_filter, vlpf
19604 Enable linear vertical low-pass filtering in the filter.
19605 Vertical low-pass filtering is required when creating an interlaced
19606 destination from a progressive source which contains high-frequency
19607 vertical detail. Filtering will reduce interlace 'twitter' and Moire
19610 @item complex_filter, cvlpf
19611 Enable complex vertical low-pass filtering.
19612 This will slightly less reduce interlace 'twitter' and Moire
19613 patterning but better retain detail and subjective sharpness impression.
19616 Bypass already interlaced frames, only adjust the frame rate.
19619 Vertical low-pass filtering and bypassing already interlaced frames can only be
19620 enabled for @option{mode} @var{interleave_top} and @var{interleave_bottom}.
19625 Pick median pixels from several successive input video frames.
19627 The filter accepts the following options:
19631 Set radius of median filter.
19632 Default is 1. Allowed range is from 1 to 127.
19635 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
19638 Set median percentile. Default value is @code{0.5}.
19639 Default value of @code{0.5} will pick always median values, while @code{0} will pick
19640 minimum values, and @code{1} maximum values.
19643 @subsection Commands
19645 This filter supports all above options as @ref{commands}, excluding option @code{radius}.
19647 @section tmidequalizer
19649 Apply Temporal Midway Video Equalization effect.
19651 Midway Video Equalization adjusts a sequence of video frames to have the same
19652 histograms, while maintaining their dynamics as much as possible. It's
19653 useful for e.g. matching exposures from a video frames sequence.
19655 This filter accepts the following option:
19659 Set filtering radius. Default is @code{5}. Allowed range is from 1 to 127.
19662 Set filtering sigma. Default is @code{0.5}. This controls strength of filtering.
19663 Setting this option to 0 effectively does nothing.
19666 Set which planes to process. Default is @code{15}, which is all available planes.
19671 Mix successive video frames.
19673 A description of the accepted options follows.
19677 The number of successive frames to mix. If unspecified, it defaults to 3.
19680 Specify weight of each input video frame.
19681 Each weight is separated by space. If number of weights is smaller than
19682 number of @var{frames} last specified weight will be used for all remaining
19686 Specify scale, if it is set it will be multiplied with sum
19687 of each weight multiplied with pixel values to give final destination
19688 pixel value. By default @var{scale} is auto scaled to sum of weights.
19691 @subsection Examples
19695 Average 7 successive frames:
19697 tmix=frames=7:weights="1 1 1 1 1 1 1"
19701 Apply simple temporal convolution:
19703 tmix=frames=3:weights="-1 3 -1"
19707 Similar as above but only showing temporal differences:
19709 tmix=frames=3:weights="-1 2 -1":scale=1
19713 @subsection Commands
19715 This filter supports the following commands:
19719 Syntax is same as option with same name.
19724 Tone map colors from different dynamic ranges.
19726 This filter expects data in single precision floating point, as it needs to
19727 operate on (and can output) out-of-range values. Another filter, such as
19728 @ref{zscale}, is needed to convert the resulting frame to a usable format.
19730 The tonemapping algorithms implemented only work on linear light, so input
19731 data should be linearized beforehand (and possibly correctly tagged).
19734 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
19737 @subsection Options
19738 The filter accepts the following options.
19742 Set the tone map algorithm to use.
19744 Possible values are:
19747 Do not apply any tone map, only desaturate overbright pixels.
19750 Hard-clip any out-of-range values. Use it for perfect color accuracy for
19751 in-range values, while distorting out-of-range values.
19754 Stretch the entire reference gamut to a linear multiple of the display.
19757 Fit a logarithmic transfer between the tone curves.
19760 Preserve overall image brightness with a simple curve, using nonlinear
19761 contrast, which results in flattening details and degrading color accuracy.
19764 Preserve both dark and bright details better than @var{reinhard}, at the cost
19765 of slightly darkening everything. Use it when detail preservation is more
19766 important than color and brightness accuracy.
19769 Smoothly map out-of-range values, while retaining contrast and colors for
19770 in-range material as much as possible. Use it when color accuracy is more
19771 important than detail preservation.
19777 Tune the tone mapping algorithm.
19779 This affects the following algorithms:
19785 Specifies the scale factor to use while stretching.
19789 Specifies the exponent of the function.
19793 Specify an extra linear coefficient to multiply into the signal before clipping.
19797 Specify the local contrast coefficient at the display peak.
19798 Default to 0.5, which means that in-gamut values will be about half as bright
19805 Specify the transition point from linear to mobius transform. Every value
19806 below this point is guaranteed to be mapped 1:1. The higher the value, the
19807 more accurate the result will be, at the cost of losing bright details.
19808 Default to 0.3, which due to the steep initial slope still preserves in-range
19809 colors fairly accurately.
19813 Apply desaturation for highlights that exceed this level of brightness. The
19814 higher the parameter, the more color information will be preserved. This
19815 setting helps prevent unnaturally blown-out colors for super-highlights, by
19816 (smoothly) turning into white instead. This makes images feel more natural,
19817 at the cost of reducing information about out-of-range colors.
19819 The default of 2.0 is somewhat conservative and will mostly just apply to
19820 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
19822 This option works only if the input frame has a supported color tag.
19825 Override signal/nominal/reference peak with this value. Useful when the
19826 embedded peak information in display metadata is not reliable or when tone
19827 mapping from a lower range to a higher range.
19832 Temporarily pad video frames.
19834 The filter accepts the following options:
19838 Specify number of delay frames before input video stream. Default is 0.
19841 Specify number of padding frames after input video stream.
19842 Set to -1 to pad indefinitely. Default is 0.
19845 Set kind of frames added to beginning of stream.
19846 Can be either @var{add} or @var{clone}.
19847 With @var{add} frames of solid-color are added.
19848 With @var{clone} frames are clones of first frame.
19849 Default is @var{add}.
19852 Set kind of frames added to end of stream.
19853 Can be either @var{add} or @var{clone}.
19854 With @var{add} frames of solid-color are added.
19855 With @var{clone} frames are clones of last frame.
19856 Default is @var{add}.
19858 @item start_duration, stop_duration
19859 Specify the duration of the start/stop delay. See
19860 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19861 for the accepted syntax.
19862 These options override @var{start} and @var{stop}. Default is 0.
19865 Specify the color of the padded area. For the syntax of this option,
19866 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
19867 manual,ffmpeg-utils}.
19869 The default value of @var{color} is "black".
19875 Transpose rows with columns in the input video and optionally flip it.
19877 It accepts the following parameters:
19882 Specify the transposition direction.
19884 Can assume the following values:
19886 @item 0, 4, cclock_flip
19887 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
19895 Rotate by 90 degrees clockwise, that is:
19903 Rotate by 90 degrees counterclockwise, that is:
19910 @item 3, 7, clock_flip
19911 Rotate by 90 degrees clockwise and vertically flip, that is:
19919 For values between 4-7, the transposition is only done if the input
19920 video geometry is portrait and not landscape. These values are
19921 deprecated, the @code{passthrough} option should be used instead.
19923 Numerical values are deprecated, and should be dropped in favor of
19924 symbolic constants.
19927 Do not apply the transposition if the input geometry matches the one
19928 specified by the specified value. It accepts the following values:
19931 Always apply transposition.
19933 Preserve portrait geometry (when @var{height} >= @var{width}).
19935 Preserve landscape geometry (when @var{width} >= @var{height}).
19938 Default value is @code{none}.
19941 For example to rotate by 90 degrees clockwise and preserve portrait
19944 transpose=dir=1:passthrough=portrait
19947 The command above can also be specified as:
19949 transpose=1:portrait
19952 @section transpose_npp
19954 Transpose rows with columns in the input video and optionally flip it.
19955 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
19957 It accepts the following parameters:
19962 Specify the transposition direction.
19964 Can assume the following values:
19967 Rotate by 90 degrees counterclockwise and vertically flip. (default)
19970 Rotate by 90 degrees clockwise.
19973 Rotate by 90 degrees counterclockwise.
19976 Rotate by 90 degrees clockwise and vertically flip.
19980 Do not apply the transposition if the input geometry matches the one
19981 specified by the specified value. It accepts the following values:
19984 Always apply transposition. (default)
19986 Preserve portrait geometry (when @var{height} >= @var{width}).
19988 Preserve landscape geometry (when @var{width} >= @var{height}).
19994 Trim the input so that the output contains one continuous subpart of the input.
19996 It accepts the following parameters:
19999 Specify the time of the start of the kept section, i.e. the frame with the
20000 timestamp @var{start} will be the first frame in the output.
20003 Specify the time of the first frame that will be dropped, i.e. the frame
20004 immediately preceding the one with the timestamp @var{end} will be the last
20005 frame in the output.
20008 This is the same as @var{start}, except this option sets the start timestamp
20009 in timebase units instead of seconds.
20012 This is the same as @var{end}, except this option sets the end timestamp
20013 in timebase units instead of seconds.
20016 The maximum duration of the output in seconds.
20019 The number of the first frame that should be passed to the output.
20022 The number of the first frame that should be dropped.
20025 @option{start}, @option{end}, and @option{duration} are expressed as time
20026 duration specifications; see
20027 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20028 for the accepted syntax.
20030 Note that the first two sets of the start/end options and the @option{duration}
20031 option look at the frame timestamp, while the _frame variants simply count the
20032 frames that pass through the filter. Also note that this filter does not modify
20033 the timestamps. If you wish for the output timestamps to start at zero, insert a
20034 setpts filter after the trim filter.
20036 If multiple start or end options are set, this filter tries to be greedy and
20037 keep all the frames that match at least one of the specified constraints. To keep
20038 only the part that matches all the constraints at once, chain multiple trim
20041 The defaults are such that all the input is kept. So it is possible to set e.g.
20042 just the end values to keep everything before the specified time.
20047 Drop everything except the second minute of input:
20049 ffmpeg -i INPUT -vf trim=60:120
20053 Keep only the first second:
20055 ffmpeg -i INPUT -vf trim=duration=1
20060 @section unpremultiply
20061 Apply alpha unpremultiply effect to input video stream using first plane
20062 of second stream as alpha.
20064 Both streams must have same dimensions and same pixel format.
20066 The filter accepts the following option:
20070 Set which planes will be processed, unprocessed planes will be copied.
20071 By default value 0xf, all planes will be processed.
20073 If the format has 1 or 2 components, then luma is bit 0.
20074 If the format has 3 or 4 components:
20075 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
20076 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
20077 If present, the alpha channel is always the last bit.
20080 Do not require 2nd input for processing, instead use alpha plane from input stream.
20086 Sharpen or blur the input video.
20088 It accepts the following parameters:
20091 @item luma_msize_x, lx
20092 Set the luma matrix horizontal size. It must be an odd integer between
20093 3 and 23. The default value is 5.
20095 @item luma_msize_y, ly
20096 Set the luma matrix vertical size. It must be an odd integer between 3
20097 and 23. The default value is 5.
20099 @item luma_amount, la
20100 Set the luma effect strength. It must be a floating point number, reasonable
20101 values lay between -1.5 and 1.5.
20103 Negative values will blur the input video, while positive values will
20104 sharpen it, a value of zero will disable the effect.
20106 Default value is 1.0.
20108 @item chroma_msize_x, cx
20109 Set the chroma matrix horizontal size. It must be an odd integer
20110 between 3 and 23. The default value is 5.
20112 @item chroma_msize_y, cy
20113 Set the chroma matrix vertical size. It must be an odd integer
20114 between 3 and 23. The default value is 5.
20116 @item chroma_amount, ca
20117 Set the chroma effect strength. It must be a floating point number, reasonable
20118 values lay between -1.5 and 1.5.
20120 Negative values will blur the input video, while positive values will
20121 sharpen it, a value of zero will disable the effect.
20123 Default value is 0.0.
20127 All parameters are optional and default to the equivalent of the
20128 string '5:5:1.0:5:5:0.0'.
20130 @subsection Examples
20134 Apply strong luma sharpen effect:
20136 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
20140 Apply a strong blur of both luma and chroma parameters:
20142 unsharp=7:7:-2:7:7:-2
20149 Decompose a video made of tiled images into the individual images.
20151 The frame rate of the output video is the frame rate of the input video
20152 multiplied by the number of tiles.
20154 This filter does the reverse of @ref{tile}.
20156 The filter accepts the following options:
20161 Set the grid size (i.e. the number of lines and columns). For the syntax of
20162 this option, check the
20163 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20166 @subsection Examples
20170 Produce a 1-second video from a still image file made of 25 frames stacked
20171 vertically, like an analogic film reel:
20173 ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
20179 Apply ultra slow/simple postprocessing filter that compresses and decompresses
20180 the image at several (or - in the case of @option{quality} level @code{8} - all)
20181 shifts and average the results.
20183 The way this differs from the behavior of spp is that uspp actually encodes &
20184 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
20185 DCT similar to MJPEG.
20187 The filter accepts the following options:
20191 Set quality. This option defines the number of levels for averaging. It accepts
20192 an integer in the range 0-8. If set to @code{0}, the filter will have no
20193 effect. A value of @code{8} means the higher quality. For each increment of
20194 that value the speed drops by a factor of approximately 2. Default value is
20198 Force a constant quantization parameter. If not set, the filter will use the QP
20199 from the video stream (if available).
20204 Convert 360 videos between various formats.
20206 The filter accepts the following options:
20212 Set format of the input/output video.
20220 Equirectangular projection.
20225 Cubemap with 3x2/6x1/1x6 layout.
20227 Format specific options:
20232 Set padding proportion for the input/output cubemap. Values in decimals.
20239 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)
20242 Default value is @b{@samp{0}}.
20243 Maximum value is @b{@samp{0.1}}.
20247 Set fixed padding for the input/output cubemap. Values in pixels.
20249 Default value is @b{@samp{0}}. If greater than zero it overrides other padding options.
20253 Set order of faces for the input/output cubemap. Choose one direction for each position.
20255 Designation of directions:
20271 Default value is @b{@samp{rludfb}}.
20275 Set rotation of faces for the input/output cubemap. Choose one angle for each position.
20277 Designation of angles:
20280 0 degrees clockwise
20282 90 degrees clockwise
20284 180 degrees clockwise
20286 270 degrees clockwise
20289 Default value is @b{@samp{000000}}.
20293 Equi-Angular Cubemap.
20300 Format specific options:
20305 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20307 If diagonal field of view is set it overrides horizontal and vertical field of view.
20312 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20314 If diagonal field of view is set it overrides horizontal and vertical field of view.
20320 Format specific options:
20325 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20327 If diagonal field of view is set it overrides horizontal and vertical field of view.
20332 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20334 If diagonal field of view is set it overrides horizontal and vertical field of view.
20340 Facebook's 360 formats.
20343 Stereographic format.
20345 Format specific options:
20350 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20352 If diagonal field of view is set it overrides horizontal and vertical field of view.
20357 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20359 If diagonal field of view is set it overrides horizontal and vertical field of view.
20366 Ball format, gives significant distortion toward the back.
20369 Hammer-Aitoff map projection format.
20372 Sinusoidal map projection format.
20375 Fisheye projection.
20377 Format specific options:
20382 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20384 If diagonal field of view is set it overrides horizontal and vertical field of view.
20389 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20391 If diagonal field of view is set it overrides horizontal and vertical field of view.
20395 Pannini projection.
20397 Format specific options:
20400 Set output pannini parameter.
20403 Set input pannini parameter.
20407 Cylindrical projection.
20409 Format specific options:
20414 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20416 If diagonal field of view is set it overrides horizontal and vertical field of view.
20421 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20423 If diagonal field of view is set it overrides horizontal and vertical field of view.
20427 Perspective projection. @i{(output only)}
20429 Format specific options:
20432 Set perspective parameter.
20436 Tetrahedron projection.
20439 Truncated square pyramid projection.
20443 Half equirectangular projection.
20448 Format specific options:
20453 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20455 If diagonal field of view is set it overrides horizontal and vertical field of view.
20460 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20462 If diagonal field of view is set it overrides horizontal and vertical field of view.
20466 Orthographic format.
20468 Format specific options:
20473 Set output horizontal/vertical/diagonal field of view. Values in degrees.
20475 If diagonal field of view is set it overrides horizontal and vertical field of view.
20480 Set input horizontal/vertical/diagonal field of view. Values in degrees.
20482 If diagonal field of view is set it overrides horizontal and vertical field of view.
20486 Octahedron projection.
20490 Set interpolation method.@*
20491 @i{Note: more complex interpolation methods require much more memory to run.}
20501 Bilinear interpolation.
20503 Lagrange9 interpolation.
20506 Bicubic interpolation.
20509 Lanczos interpolation.
20512 Spline16 interpolation.
20515 Gaussian interpolation.
20517 Mitchell interpolation.
20520 Default value is @b{@samp{line}}.
20524 Set the output video resolution.
20526 Default resolution depends on formats.
20530 Set the input/output stereo format.
20541 Default value is @b{@samp{2d}} for input and output format.
20546 Set rotation for the output video. Values in degrees.
20549 Set rotation order for the output video. Choose one item for each position.
20560 Default value is @b{@samp{ypr}}.
20565 Flip the output video horizontally(swaps left-right)/vertically(swaps up-down)/in-depth(swaps back-forward). Boolean values.
20569 Set if input video is flipped horizontally/vertically. Boolean values.
20572 Set if input video is transposed. Boolean value, by default disabled.
20575 Set if output video needs to be transposed. Boolean value, by default disabled.
20578 Build mask in alpha plane for all unmapped pixels by marking them fully transparent. Boolean value, by default disabled.
20581 @subsection Examples
20585 Convert equirectangular video to cubemap with 3x2 layout and 1% padding using bicubic interpolation:
20587 ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
20590 Extract back view of Equi-Angular Cubemap:
20592 ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
20595 Convert transposed and horizontally flipped Equi-Angular Cubemap in side-by-side stereo format to equirectangular top-bottom stereo format:
20597 v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
20601 @subsection Commands
20603 This filter supports subset of above options as @ref{commands}.
20605 @section vaguedenoiser
20607 Apply a wavelet based denoiser.
20609 It transforms each frame from the video input into the wavelet domain,
20610 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
20611 the obtained coefficients. It does an inverse wavelet transform after.
20612 Due to wavelet properties, it should give a nice smoothed result, and
20613 reduced noise, without blurring picture features.
20615 This filter accepts the following options:
20619 The filtering strength. The higher, the more filtered the video will be.
20620 Hard thresholding can use a higher threshold than soft thresholding
20621 before the video looks overfiltered. Default value is 2.
20624 The filtering method the filter will use.
20626 It accepts the following values:
20629 All values under the threshold will be zeroed.
20632 All values under the threshold will be zeroed. All values above will be
20633 reduced by the threshold.
20636 Scales or nullifies coefficients - intermediary between (more) soft and
20637 (less) hard thresholding.
20640 Default is garrote.
20643 Number of times, the wavelet will decompose the picture. Picture can't
20644 be decomposed beyond a particular point (typically, 8 for a 640x480
20645 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
20648 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
20651 A list of the planes to process. By default all planes are processed.
20654 The threshold type the filter will use.
20656 It accepts the following values:
20659 Threshold used is same for all decompositions.
20662 Threshold used depends also on each decomposition coefficients.
20665 Default is universal.
20668 @section vectorscope
20670 Display 2 color component values in the two dimensional graph (which is called
20673 This filter accepts the following options:
20677 Set vectorscope mode.
20679 It accepts the following values:
20683 Gray values are displayed on graph, higher brightness means more pixels have
20684 same component color value on location in graph. This is the default mode.
20687 Gray values are displayed on graph. Surrounding pixels values which are not
20688 present in video frame are drawn in gradient of 2 color components which are
20689 set by option @code{x} and @code{y}. The 3rd color component is static.
20692 Actual color components values present in video frame are displayed on graph.
20695 Similar as color2 but higher frequency of same values @code{x} and @code{y}
20696 on graph increases value of another color component, which is luminance by
20697 default values of @code{x} and @code{y}.
20700 Actual colors present in video frame are displayed on graph. If two different
20701 colors map to same position on graph then color with higher value of component
20702 not present in graph is picked.
20705 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
20706 component picked from radial gradient.
20710 Set which color component will be represented on X-axis. Default is @code{1}.
20713 Set which color component will be represented on Y-axis. Default is @code{2}.
20716 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
20717 of color component which represents frequency of (X, Y) location in graph.
20722 No envelope, this is default.
20725 Instant envelope, even darkest single pixel will be clearly highlighted.
20728 Hold maximum and minimum values presented in graph over time. This way you
20729 can still spot out of range values without constantly looking at vectorscope.
20732 Peak and instant envelope combined together.
20736 Set what kind of graticule to draw.
20745 Set graticule opacity.
20748 Set graticule flags.
20752 Draw graticule for white point.
20755 Draw graticule for black point.
20758 Draw color points short names.
20762 Set background opacity.
20764 @item lthreshold, l
20765 Set low threshold for color component not represented on X or Y axis.
20766 Values lower than this value will be ignored. Default is 0.
20767 Note this value is multiplied with actual max possible value one pixel component
20768 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
20771 @item hthreshold, h
20772 Set high threshold for color component not represented on X or Y axis.
20773 Values higher than this value will be ignored. Default is 1.
20774 Note this value is multiplied with actual max possible value one pixel component
20775 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
20776 is 0.9 * 255 = 230.
20778 @item colorspace, c
20779 Set what kind of colorspace to use when drawing graticule.
20789 Set color tint for gray/tint vectorscope mode. By default both options are zero.
20790 This means no tint, and output will remain gray.
20793 @anchor{vidstabdetect}
20794 @section vidstabdetect
20796 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
20797 @ref{vidstabtransform} for pass 2.
20799 This filter generates a file with relative translation and rotation
20800 transform information about subsequent frames, which is then used by
20801 the @ref{vidstabtransform} filter.
20803 To enable compilation of this filter you need to configure FFmpeg with
20804 @code{--enable-libvidstab}.
20806 This filter accepts the following options:
20810 Set the path to the file used to write the transforms information.
20811 Default value is @file{transforms.trf}.
20814 Set how shaky the video is and how quick the camera is. It accepts an
20815 integer in the range 1-10, a value of 1 means little shakiness, a
20816 value of 10 means strong shakiness. Default value is 5.
20819 Set the accuracy of the detection process. It must be a value in the
20820 range 1-15. A value of 1 means low accuracy, a value of 15 means high
20821 accuracy. Default value is 15.
20824 Set stepsize of the search process. The region around minimum is
20825 scanned with 1 pixel resolution. Default value is 6.
20828 Set minimum contrast. Below this value a local measurement field is
20829 discarded. Must be a floating point value in the range 0-1. Default
20833 Set reference frame number for tripod mode.
20835 If enabled, the motion of the frames is compared to a reference frame
20836 in the filtered stream, identified by the specified number. The idea
20837 is to compensate all movements in a more-or-less static scene and keep
20838 the camera view absolutely still.
20840 If set to 0, it is disabled. The frames are counted starting from 1.
20843 Show fields and transforms in the resulting frames. It accepts an
20844 integer in the range 0-2. Default value is 0, which disables any
20848 @subsection Examples
20852 Use default values:
20858 Analyze strongly shaky movie and put the results in file
20859 @file{mytransforms.trf}:
20861 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
20865 Visualize the result of internal transformations in the resulting
20868 vidstabdetect=show=1
20872 Analyze a video with medium shakiness using @command{ffmpeg}:
20874 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
20878 @anchor{vidstabtransform}
20879 @section vidstabtransform
20881 Video stabilization/deshaking: pass 2 of 2,
20882 see @ref{vidstabdetect} for pass 1.
20884 Read a file with transform information for each frame and
20885 apply/compensate them. Together with the @ref{vidstabdetect}
20886 filter this can be used to deshake videos. See also
20887 @url{http://public.hronopik.de/vid.stab}. It is important to also use
20888 the @ref{unsharp} filter, see below.
20890 To enable compilation of this filter you need to configure FFmpeg with
20891 @code{--enable-libvidstab}.
20893 @subsection Options
20897 Set path to the file used to read the transforms. Default value is
20898 @file{transforms.trf}.
20901 Set the number of frames (value*2 + 1) used for lowpass filtering the
20902 camera movements. Default value is 10.
20904 For example a number of 10 means that 21 frames are used (10 in the
20905 past and 10 in the future) to smoothen the motion in the video. A
20906 larger value leads to a smoother video, but limits the acceleration of
20907 the camera (pan/tilt movements). 0 is a special case where a static
20908 camera is simulated.
20911 Set the camera path optimization algorithm.
20913 Accepted values are:
20916 gaussian kernel low-pass filter on camera motion (default)
20918 averaging on transformations
20922 Set maximal number of pixels to translate frames. Default value is -1,
20926 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
20927 value is -1, meaning no limit.
20930 Specify how to deal with borders that may be visible due to movement
20933 Available values are:
20936 keep image information from previous frame (default)
20938 fill the border black
20942 Invert transforms if set to 1. Default value is 0.
20945 Consider transforms as relative to previous frame if set to 1,
20946 absolute if set to 0. Default value is 0.
20949 Set percentage to zoom. A positive value will result in a zoom-in
20950 effect, a negative value in a zoom-out effect. Default value is 0 (no
20954 Set optimal zooming to avoid borders.
20956 Accepted values are:
20961 optimal static zoom value is determined (only very strong movements
20962 will lead to visible borders) (default)
20964 optimal adaptive zoom value is determined (no borders will be
20965 visible), see @option{zoomspeed}
20968 Note that the value given at zoom is added to the one calculated here.
20971 Set percent to zoom maximally each frame (enabled when
20972 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
20976 Specify type of interpolation.
20978 Available values are:
20983 linear only horizontal
20985 linear in both directions (default)
20987 cubic in both directions (slow)
20991 Enable virtual tripod mode if set to 1, which is equivalent to
20992 @code{relative=0:smoothing=0}. Default value is 0.
20994 Use also @code{tripod} option of @ref{vidstabdetect}.
20997 Increase log verbosity if set to 1. Also the detected global motions
20998 are written to the temporary file @file{global_motions.trf}. Default
21002 @subsection Examples
21006 Use @command{ffmpeg} for a typical stabilization with default values:
21008 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
21011 Note the use of the @ref{unsharp} filter which is always recommended.
21014 Zoom in a bit more and load transform data from a given file:
21016 vidstabtransform=zoom=5:input="mytransforms.trf"
21020 Smoothen the video even more:
21022 vidstabtransform=smoothing=30
21028 Flip the input video vertically.
21030 For example, to vertically flip a video with @command{ffmpeg}:
21032 ffmpeg -i in.avi -vf "vflip" out.avi
21037 Detect variable frame rate video.
21039 This filter tries to detect if the input is variable or constant frame rate.
21041 At end it will output number of frames detected as having variable delta pts,
21042 and ones with constant delta pts.
21043 If there was frames with variable delta, than it will also show min, max and
21044 average delta encountered.
21048 Boost or alter saturation.
21050 The filter accepts the following options:
21053 Set strength of boost if positive value or strength of alter if negative value.
21054 Default is 0. Allowed range is from -2 to 2.
21057 Set the red balance. Default is 1. Allowed range is from -10 to 10.
21060 Set the green balance. Default is 1. Allowed range is from -10 to 10.
21063 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
21066 Set the red luma coefficient.
21069 Set the green luma coefficient.
21072 Set the blue luma coefficient.
21075 If @code{intensity} is negative and this is set to 1, colors will change,
21076 otherwise colors will be less saturated, more towards gray.
21079 @subsection Commands
21081 This filter supports the all above options as @ref{commands}.
21086 Make or reverse a natural vignetting effect.
21088 The filter accepts the following options:
21092 Set lens angle expression as a number of radians.
21094 The value is clipped in the @code{[0,PI/2]} range.
21096 Default value: @code{"PI/5"}
21100 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
21104 Set forward/backward mode.
21106 Available modes are:
21109 The larger the distance from the central point, the darker the image becomes.
21112 The larger the distance from the central point, the brighter the image becomes.
21113 This can be used to reverse a vignette effect, though there is no automatic
21114 detection to extract the lens @option{angle} and other settings (yet). It can
21115 also be used to create a burning effect.
21118 Default value is @samp{forward}.
21121 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
21123 It accepts the following values:
21126 Evaluate expressions only once during the filter initialization.
21129 Evaluate expressions for each incoming frame. This is way slower than the
21130 @samp{init} mode since it requires all the scalers to be re-computed, but it
21131 allows advanced dynamic expressions.
21134 Default value is @samp{init}.
21137 Set dithering to reduce the circular banding effects. Default is @code{1}
21141 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
21142 Setting this value to the SAR of the input will make a rectangular vignetting
21143 following the dimensions of the video.
21145 Default is @code{1/1}.
21148 @subsection Expressions
21150 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
21151 following parameters.
21156 input width and height
21159 the number of input frame, starting from 0
21162 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
21163 @var{TB} units, NAN if undefined
21166 frame rate of the input video, NAN if the input frame rate is unknown
21169 the PTS (Presentation TimeStamp) of the filtered video frame,
21170 expressed in seconds, NAN if undefined
21173 time base of the input video
21177 @subsection Examples
21181 Apply simple strong vignetting effect:
21187 Make a flickering vignetting:
21189 vignette='PI/4+random(1)*PI/50':eval=frame
21194 @section vmafmotion
21196 Obtain the average VMAF motion score of a video.
21197 It is one of the component metrics of VMAF.
21199 The obtained average motion score is printed through the logging system.
21201 The filter accepts the following options:
21205 If specified, the filter will use the named file to save the motion score of
21206 each frame with respect to the previous frame.
21207 When filename equals "-" the data is sent to standard output.
21212 ffmpeg -i ref.mpg -vf vmafmotion -f null -
21216 Stack input videos vertically.
21218 All streams must be of same pixel format and of same width.
21220 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
21221 to create same output.
21223 The filter accepts the following options:
21227 Set number of input streams. Default is 2.
21230 If set to 1, force the output to terminate when the shortest input
21231 terminates. Default value is 0.
21236 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
21237 Deinterlacing Filter").
21239 Based on the process described by Martin Weston for BBC R&D, and
21240 implemented based on the de-interlace algorithm written by Jim
21241 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
21242 uses filter coefficients calculated by BBC R&D.
21244 This filter uses field-dominance information in frame to decide which
21245 of each pair of fields to place first in the output.
21246 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
21248 There are two sets of filter coefficients, so called "simple"
21249 and "complex". Which set of filter coefficients is used can
21250 be set by passing an optional parameter:
21254 Set the interlacing filter coefficients. Accepts one of the following values:
21258 Simple filter coefficient set.
21260 More-complex filter coefficient set.
21262 Default value is @samp{complex}.
21265 The interlacing mode to adopt. It accepts one of the following values:
21269 Output one frame for each frame.
21271 Output one frame for each field.
21274 The default value is @code{field}.
21277 The picture field parity assumed for the input interlaced video. It accepts one
21278 of the following values:
21282 Assume the top field is first.
21284 Assume the bottom field is first.
21286 Enable automatic detection of field parity.
21289 The default value is @code{auto}.
21290 If the interlacing is unknown or the decoder does not export this information,
21291 top field first will be assumed.
21294 Specify which frames to deinterlace. Accepts one of the following values:
21298 Deinterlace all frames,
21300 Only deinterlace frames marked as interlaced.
21303 Default value is @samp{all}.
21306 @subsection Commands
21307 This filter supports same @ref{commands} as options.
21310 Video waveform monitor.
21312 The waveform monitor plots color component intensity. By default luminance
21313 only. Each column of the waveform corresponds to a column of pixels in the
21316 It accepts the following options:
21320 Can be either @code{row}, or @code{column}. Default is @code{column}.
21321 In row mode, the graph on the left side represents color component value 0 and
21322 the right side represents value = 255. In column mode, the top side represents
21323 color component value = 0 and bottom side represents value = 255.
21326 Set intensity. Smaller values are useful to find out how many values of the same
21327 luminance are distributed across input rows/columns.
21328 Default value is @code{0.04}. Allowed range is [0, 1].
21331 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
21332 In mirrored mode, higher values will be represented on the left
21333 side for @code{row} mode and at the top for @code{column} mode. Default is
21334 @code{1} (mirrored).
21338 It accepts the following values:
21341 Presents information identical to that in the @code{parade}, except
21342 that the graphs representing color components are superimposed directly
21345 This display mode makes it easier to spot relative differences or similarities
21346 in overlapping areas of the color components that are supposed to be identical,
21347 such as neutral whites, grays, or blacks.
21350 Display separate graph for the color components side by side in
21351 @code{row} mode or one below the other in @code{column} mode.
21354 Display separate graph for the color components side by side in
21355 @code{column} mode or one below the other in @code{row} mode.
21357 Using this display mode makes it easy to spot color casts in the highlights
21358 and shadows of an image, by comparing the contours of the top and the bottom
21359 graphs of each waveform. Since whites, grays, and blacks are characterized
21360 by exactly equal amounts of red, green, and blue, neutral areas of the picture
21361 should display three waveforms of roughly equal width/height. If not, the
21362 correction is easy to perform by making level adjustments the three waveforms.
21364 Default is @code{stack}.
21366 @item components, c
21367 Set which color components to display. Default is 1, which means only luminance
21368 or red color component if input is in RGB colorspace. If is set for example to
21369 7 it will display all 3 (if) available color components.
21374 No envelope, this is default.
21377 Instant envelope, minimum and maximum values presented in graph will be easily
21378 visible even with small @code{step} value.
21381 Hold minimum and maximum values presented in graph across time. This way you
21382 can still spot out of range values without constantly looking at waveforms.
21385 Peak and instant envelope combined together.
21391 No filtering, this is default.
21394 Luma and chroma combined together.
21397 Similar as above, but shows difference between blue and red chroma.
21400 Similar as above, but use different colors.
21403 Similar as above, but again with different colors.
21406 Displays only chroma.
21409 Displays actual color value on waveform.
21412 Similar as above, but with luma showing frequency of chroma values.
21416 Set which graticule to display.
21420 Do not display graticule.
21423 Display green graticule showing legal broadcast ranges.
21426 Display orange graticule showing legal broadcast ranges.
21429 Display invert graticule showing legal broadcast ranges.
21433 Set graticule opacity.
21436 Set graticule flags.
21440 Draw numbers above lines. By default enabled.
21443 Draw dots instead of lines.
21447 Set scale used for displaying graticule.
21454 Default is digital.
21457 Set background opacity.
21461 Set tint for output.
21462 Only used with lowpass filter and when display is not overlay and input
21463 pixel formats are not RGB.
21466 @section weave, doubleweave
21468 The @code{weave} takes a field-based video input and join
21469 each two sequential fields into single frame, producing a new double
21470 height clip with half the frame rate and half the frame count.
21472 The @code{doubleweave} works same as @code{weave} but without
21473 halving frame rate and frame count.
21475 It accepts the following option:
21479 Set first field. Available values are:
21483 Set the frame as top-field-first.
21486 Set the frame as bottom-field-first.
21490 @subsection Examples
21494 Interlace video using @ref{select} and @ref{separatefields} filter:
21496 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
21501 Apply the xBR high-quality magnification filter which is designed for pixel
21502 art. It follows a set of edge-detection rules, see
21503 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
21505 It accepts the following option:
21509 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
21510 @code{3xBR} and @code{4} for @code{4xBR}.
21511 Default is @code{3}.
21516 Apply cross fade from one input video stream to another input video stream.
21517 The cross fade is applied for specified duration.
21519 The filter accepts the following options:
21523 Set one of available transition effects:
21571 Default transition effect is fade.
21574 Set cross fade duration in seconds.
21575 Default duration is 1 second.
21578 Set cross fade start relative to first input stream in seconds.
21579 Default offset is 0.
21582 Set expression for custom transition effect.
21584 The expressions can use the following variables and functions:
21589 The coordinates of the current sample.
21593 The width and height of the image.
21596 Progress of transition effect.
21599 Currently processed plane.
21602 Return value of first input at current location and plane.
21605 Return value of second input at current location and plane.
21611 Return the value of the pixel at location (@var{x},@var{y}) of the
21612 first/second/third/fourth component of first input.
21618 Return the value of the pixel at location (@var{x},@var{y}) of the
21619 first/second/third/fourth component of second input.
21623 @subsection Examples
21627 Cross fade from one input video to another input video, with fade transition and duration of transition
21628 of 2 seconds starting at offset of 5 seconds:
21630 ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
21635 Pick median pixels from several input videos.
21637 The filter accepts the following options:
21641 Set number of inputs.
21642 Default is 3. Allowed range is from 3 to 255.
21643 If number of inputs is even number, than result will be mean value between two median values.
21646 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
21649 Set median percentile. Default value is @code{0.5}.
21650 Default value of @code{0.5} will pick always median values, while @code{0} will pick
21651 minimum values, and @code{1} maximum values.
21654 @subsection Commands
21656 This filter supports all above options as @ref{commands}, excluding option @code{inputs}.
21659 Stack video inputs into custom layout.
21661 All streams must be of same pixel format.
21663 The filter accepts the following options:
21667 Set number of input streams. Default is 2.
21670 Specify layout of inputs.
21671 This option requires the desired layout configuration to be explicitly set by the user.
21672 This sets position of each video input in output. Each input
21673 is separated by '|'.
21674 The first number represents the column, and the second number represents the row.
21675 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
21676 where X is video input from which to take width or height.
21677 Multiple values can be used when separated by '+'. In such
21678 case values are summed together.
21680 Note that if inputs are of different sizes gaps may appear, as not all of
21681 the output video frame will be filled. Similarly, videos can overlap each
21682 other if their position doesn't leave enough space for the full frame of
21685 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
21686 a layout must be set by the user.
21689 If set to 1, force the output to terminate when the shortest input
21690 terminates. Default value is 0.
21693 If set to valid color, all unused pixels will be filled with that color.
21694 By default fill is set to none, so it is disabled.
21697 @subsection Examples
21701 Display 4 inputs into 2x2 grid.
21705 input1(0, 0) | input3(w0, 0)
21706 input2(0, h0) | input4(w0, h0)
21710 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
21713 Note that if inputs are of different sizes, gaps or overlaps may occur.
21716 Display 4 inputs into 1x4 grid.
21723 input4(0, h0+h1+h2)
21727 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
21730 Note that if inputs are of different widths, unused space will appear.
21733 Display 9 inputs into 3x3 grid.
21737 input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
21738 input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
21739 input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
21743 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
21746 Note that if inputs are of different sizes, gaps or overlaps may occur.
21749 Display 16 inputs into 4x4 grid.
21753 input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
21754 input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
21755 input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
21756 input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
21760 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|
21761 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
21764 Note that if inputs are of different sizes, gaps or overlaps may occur.
21771 Deinterlace the input video ("yadif" means "yet another deinterlacing
21774 It accepts the following parameters:
21780 The interlacing mode to adopt. It accepts one of the following values:
21783 @item 0, send_frame
21784 Output one frame for each frame.
21785 @item 1, send_field
21786 Output one frame for each field.
21787 @item 2, send_frame_nospatial
21788 Like @code{send_frame}, but it skips the spatial interlacing check.
21789 @item 3, send_field_nospatial
21790 Like @code{send_field}, but it skips the spatial interlacing check.
21793 The default value is @code{send_frame}.
21796 The picture field parity assumed for the input interlaced video. It accepts one
21797 of the following values:
21801 Assume the top field is first.
21803 Assume the bottom field is first.
21805 Enable automatic detection of field parity.
21808 The default value is @code{auto}.
21809 If the interlacing is unknown or the decoder does not export this information,
21810 top field first will be assumed.
21813 Specify which frames to deinterlace. Accepts one of the following
21818 Deinterlace all frames.
21819 @item 1, interlaced
21820 Only deinterlace frames marked as interlaced.
21823 The default value is @code{all}.
21826 @section yadif_cuda
21828 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
21829 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
21832 It accepts the following parameters:
21838 The interlacing mode to adopt. It accepts one of the following values:
21841 @item 0, send_frame
21842 Output one frame for each frame.
21843 @item 1, send_field
21844 Output one frame for each field.
21845 @item 2, send_frame_nospatial
21846 Like @code{send_frame}, but it skips the spatial interlacing check.
21847 @item 3, send_field_nospatial
21848 Like @code{send_field}, but it skips the spatial interlacing check.
21851 The default value is @code{send_frame}.
21854 The picture field parity assumed for the input interlaced video. It accepts one
21855 of the following values:
21859 Assume the top field is first.
21861 Assume the bottom field is first.
21863 Enable automatic detection of field parity.
21866 The default value is @code{auto}.
21867 If the interlacing is unknown or the decoder does not export this information,
21868 top field first will be assumed.
21871 Specify which frames to deinterlace. Accepts one of the following
21876 Deinterlace all frames.
21877 @item 1, interlaced
21878 Only deinterlace frames marked as interlaced.
21881 The default value is @code{all}.
21886 Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter").
21887 The algorithm is described in
21888 "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
21890 It accepts the following parameters:
21894 Set the window radius. Default value is 3.
21897 Set which planes to filter. Default is only the first plane.
21900 Set blur strength. Default value is 128.
21903 @subsection Commands
21904 This filter supports same @ref{commands} as options.
21908 Apply Zoom & Pan effect.
21910 This filter accepts the following options:
21914 Set the zoom expression. Range is 1-10. Default is 1.
21918 Set the x and y expression. Default is 0.
21921 Set the duration expression in number of frames.
21922 This sets for how many number of frames effect will last for
21923 single input image.
21926 Set the output image size, default is 'hd720'.
21929 Set the output frame rate, default is '25'.
21932 Each expression can contain the following constants:
21951 Output frame count.
21954 The input timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
21956 @item out_time, time, ot
21957 The output timestamp expressed in seconds.
21961 Last calculated 'x' and 'y' position from 'x' and 'y' expression
21962 for current input frame.
21966 'x' and 'y' of last output frame of previous input frame or 0 when there was
21967 not yet such frame (first input frame).
21970 Last calculated zoom from 'z' expression for current input frame.
21973 Last calculated zoom of last output frame of previous input frame.
21976 Number of output frames for current input frame. Calculated from 'd' expression
21977 for each input frame.
21980 number of output frames created for previous input frame
21983 Rational number: input width / input height
21986 sample aspect ratio
21989 display aspect ratio
21993 @subsection Examples
21997 Zoom in up to 1.5x and pan at same time to some spot near center of picture:
21999 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
22003 Zoom in up to 1.5x and pan always at center of picture:
22005 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22009 Same as above but without pausing:
22011 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22015 Zoom in 2x into center of picture only for the first second of the input video:
22017 zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
22024 Scale (resize) the input video, using the z.lib library:
22025 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
22026 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
22028 The zscale filter forces the output display aspect ratio to be the same
22029 as the input, by changing the output sample aspect ratio.
22031 If the input image format is different from the format requested by
22032 the next filter, the zscale filter will convert the input to the
22035 @subsection Options
22036 The filter accepts the following options.
22041 Set the output video dimension expression. Default value is the input
22044 If the @var{width} or @var{w} value is 0, the input width is used for
22045 the output. If the @var{height} or @var{h} value is 0, the input height
22046 is used for the output.
22048 If one and only one of the values is -n with n >= 1, the zscale filter
22049 will use a value that maintains the aspect ratio of the input image,
22050 calculated from the other specified dimension. After that it will,
22051 however, make sure that the calculated dimension is divisible by n and
22052 adjust the value if necessary.
22054 If both values are -n with n >= 1, the behavior will be identical to
22055 both values being set to 0 as previously detailed.
22057 See below for the list of accepted constants for use in the dimension
22061 Set the video size. For the syntax of this option, check the
22062 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22065 Set the dither type.
22067 Possible values are:
22072 @item error_diffusion
22078 Set the resize filter type.
22080 Possible values are:
22090 Default is bilinear.
22093 Set the color range.
22095 Possible values are:
22102 Default is same as input.
22105 Set the color primaries.
22107 Possible values are:
22117 Default is same as input.
22120 Set the transfer characteristics.
22122 Possible values are:
22136 Default is same as input.
22139 Set the colorspace matrix.
22141 Possible value are:
22152 Default is same as input.
22155 Set the input color range.
22157 Possible values are:
22164 Default is same as input.
22166 @item primariesin, pin
22167 Set the input color primaries.
22169 Possible values are:
22179 Default is same as input.
22181 @item transferin, tin
22182 Set the input transfer characteristics.
22184 Possible values are:
22195 Default is same as input.
22197 @item matrixin, min
22198 Set the input colorspace matrix.
22200 Possible value are:
22212 Set the output chroma location.
22214 Possible values are:
22225 @item chromalin, cin
22226 Set the input chroma location.
22228 Possible values are:
22240 Set the nominal peak luminance.
22243 The values of the @option{w} and @option{h} options are expressions
22244 containing the following constants:
22249 The input width and height
22253 These are the same as @var{in_w} and @var{in_h}.
22257 The output (scaled) width and height
22261 These are the same as @var{out_w} and @var{out_h}
22264 The same as @var{iw} / @var{ih}
22267 input sample aspect ratio
22270 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
22274 horizontal and vertical input chroma subsample values. For example for the
22275 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22279 horizontal and vertical output chroma subsample values. For example for the
22280 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
22283 @subsection Commands
22285 This filter supports the following commands:
22289 Set the output video dimension expression.
22290 The command accepts the same syntax of the corresponding option.
22292 If the specified expression is not valid, it is kept at its current
22296 @c man end VIDEO FILTERS
22298 @chapter OpenCL Video Filters
22299 @c man begin OPENCL VIDEO FILTERS
22301 Below is a description of the currently available OpenCL video filters.
22303 To enable compilation of these filters you need to configure FFmpeg with
22304 @code{--enable-opencl}.
22306 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
22309 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
22310 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
22311 given device parameters.
22313 @item -filter_hw_device @var{name}
22314 Pass the hardware device called @var{name} to all filters in any filter graph.
22318 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
22322 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
22324 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
22328 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.
22330 @section avgblur_opencl
22332 Apply average blur filter.
22334 The filter accepts the following options:
22338 Set horizontal radius size.
22339 Range is @code{[1, 1024]} and default value is @code{1}.
22342 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22345 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
22348 @subsection Example
22352 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.
22354 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
22358 @section boxblur_opencl
22360 Apply a boxblur algorithm to the input video.
22362 It accepts the following parameters:
22366 @item luma_radius, lr
22367 @item luma_power, lp
22368 @item chroma_radius, cr
22369 @item chroma_power, cp
22370 @item alpha_radius, ar
22371 @item alpha_power, ap
22375 A description of the accepted options follows.
22378 @item luma_radius, lr
22379 @item chroma_radius, cr
22380 @item alpha_radius, ar
22381 Set an expression for the box radius in pixels used for blurring the
22382 corresponding input plane.
22384 The radius value must be a non-negative number, and must not be
22385 greater than the value of the expression @code{min(w,h)/2} for the
22386 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
22389 Default value for @option{luma_radius} is "2". If not specified,
22390 @option{chroma_radius} and @option{alpha_radius} default to the
22391 corresponding value set for @option{luma_radius}.
22393 The expressions can contain the following constants:
22397 The input width and height in pixels.
22401 The input chroma image width and height in pixels.
22405 The horizontal and vertical chroma subsample values. For example, for the
22406 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
22409 @item luma_power, lp
22410 @item chroma_power, cp
22411 @item alpha_power, ap
22412 Specify how many times the boxblur filter is applied to the
22413 corresponding plane.
22415 Default value for @option{luma_power} is 2. If not specified,
22416 @option{chroma_power} and @option{alpha_power} default to the
22417 corresponding value set for @option{luma_power}.
22419 A value of 0 will disable the effect.
22422 @subsection Examples
22424 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.
22428 Apply a boxblur filter with the luma, chroma, and alpha radius
22429 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.
22431 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
22432 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
22436 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.
22438 For the luma plane, a 2x2 box radius will be run once.
22440 For the chroma plane, a 4x4 box radius will be run 5 times.
22442 For the alpha plane, a 3x3 box radius will be run 7 times.
22444 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
22448 @section colorkey_opencl
22449 RGB colorspace color keying.
22451 The filter accepts the following options:
22455 The color which will be replaced with transparency.
22458 Similarity percentage with the key color.
22460 0.01 matches only the exact key color, while 1.0 matches everything.
22465 0.0 makes pixels either fully transparent, or not transparent at all.
22467 Higher values result in semi-transparent pixels, with a higher transparency
22468 the more similar the pixels color is to the key color.
22471 @subsection Examples
22475 Make every semi-green pixel in the input transparent with some slight blending:
22477 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
22481 @section convolution_opencl
22483 Apply convolution of 3x3, 5x5, 7x7 matrix.
22485 The filter accepts the following options:
22492 Set matrix for each plane.
22493 Matrix is sequence of 9, 25 or 49 signed numbers.
22494 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
22500 Set multiplier for calculated value for each plane.
22501 If unset or 0, it will be sum of all matrix elements.
22502 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
22508 Set bias for each plane. This value is added to the result of the multiplication.
22509 Useful for making the overall image brighter or darker.
22510 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
22514 @subsection Examples
22520 -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
22526 -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
22530 Apply edge enhance:
22532 -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
22538 -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
22542 Apply laplacian edge detector which includes diagonals:
22544 -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
22550 -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
22554 @section erosion_opencl
22556 Apply erosion effect to the video.
22558 This filter replaces the pixel by the local(3x3) minimum.
22560 It accepts the following options:
22567 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22568 If @code{0}, plane will remain unchanged.
22571 Flag which specifies the pixel to refer to.
22572 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22574 Flags to local 3x3 coordinates region centered on @code{x}:
22583 @subsection Example
22587 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.
22589 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22593 @section deshake_opencl
22594 Feature-point based video stabilization filter.
22596 The filter accepts the following options:
22600 Simulates a tripod by preventing any camera movement whatsoever from the original frame. Defaults to @code{0}.
22603 Whether or not additional debug info should be displayed, both in the processed output and in the console.
22605 Note that in order to see console debug output you will also need to pass @code{-v verbose} to ffmpeg.
22607 Viewing point matches in the output video is only supported for RGB input.
22609 Defaults to @code{0}.
22611 @item adaptive_crop
22612 Whether or not to do a tiny bit of cropping at the borders to cut down on the amount of mirrored pixels.
22614 Defaults to @code{1}.
22616 @item refine_features
22617 Whether or not feature points should be refined at a sub-pixel level.
22619 This can be turned off for a slight performance gain at the cost of precision.
22621 Defaults to @code{1}.
22623 @item smooth_strength
22624 The strength of the smoothing applied to the camera path from @code{0.0} to @code{1.0}.
22626 @code{1.0} is the maximum smoothing strength while values less than that result in less smoothing.
22628 @code{0.0} causes the filter to adaptively choose a smoothing strength on a per-frame basis.
22630 Defaults to @code{0.0}.
22632 @item smooth_window_multiplier
22633 Controls the size of the smoothing window (the number of frames buffered to determine motion information from).
22635 The size of the smoothing window is determined by multiplying the framerate of the video by this number.
22637 Acceptable values range from @code{0.1} to @code{10.0}.
22639 Larger values increase the amount of motion data available for determining how to smooth the camera path,
22640 potentially improving smoothness, but also increase latency and memory usage.
22642 Defaults to @code{2.0}.
22646 @subsection Examples
22650 Stabilize a video with a fixed, medium smoothing strength:
22652 -i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
22656 Stabilize a video with debugging (both in console and in rendered video):
22658 -i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
22662 @section dilation_opencl
22664 Apply dilation effect to the video.
22666 This filter replaces the pixel by the local(3x3) maximum.
22668 It accepts the following options:
22675 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
22676 If @code{0}, plane will remain unchanged.
22679 Flag which specifies the pixel to refer to.
22680 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
22682 Flags to local 3x3 coordinates region centered on @code{x}:
22691 @subsection Example
22695 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.
22697 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
22701 @section nlmeans_opencl
22703 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
22705 @section overlay_opencl
22707 Overlay one video on top of another.
22709 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
22710 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
22712 The filter accepts the following options:
22717 Set the x coordinate of the overlaid video on the main video.
22718 Default value is @code{0}.
22721 Set the y coordinate of the overlaid video on the main video.
22722 Default value is @code{0}.
22726 @subsection Examples
22730 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
22732 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22735 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
22737 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
22742 @section pad_opencl
22744 Add paddings to the input image, and place the original input at the
22745 provided @var{x}, @var{y} coordinates.
22747 It accepts the following options:
22752 Specify an expression for the size of the output image with the
22753 paddings added. If the value for @var{width} or @var{height} is 0, the
22754 corresponding input size is used for the output.
22756 The @var{width} expression can reference the value set by the
22757 @var{height} expression, and vice versa.
22759 The default value of @var{width} and @var{height} is 0.
22763 Specify the offsets to place the input image at within the padded area,
22764 with respect to the top/left border of the output image.
22766 The @var{x} expression can reference the value set by the @var{y}
22767 expression, and vice versa.
22769 The default value of @var{x} and @var{y} is 0.
22771 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
22772 so the input image is centered on the padded area.
22775 Specify the color of the padded area. For the syntax of this option,
22776 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
22777 manual,ffmpeg-utils}.
22780 Pad to an aspect instead to a resolution.
22783 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
22784 options are expressions containing the following constants:
22789 The input video width and height.
22793 These are the same as @var{in_w} and @var{in_h}.
22797 The output width and height (the size of the padded area), as
22798 specified by the @var{width} and @var{height} expressions.
22802 These are the same as @var{out_w} and @var{out_h}.
22806 The x and y offsets as specified by the @var{x} and @var{y}
22807 expressions, or NAN if not yet specified.
22810 same as @var{iw} / @var{ih}
22813 input sample aspect ratio
22816 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
22819 @section prewitt_opencl
22821 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
22823 The filter accepts the following option:
22827 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22830 Set value which will be multiplied with filtered result.
22831 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22834 Set value which will be added to filtered result.
22835 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22838 @subsection Example
22842 Apply the Prewitt operator with scale set to 2 and delta set to 10.
22844 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
22848 @anchor{program_opencl}
22849 @section program_opencl
22851 Filter video using an OpenCL program.
22856 OpenCL program source file.
22859 Kernel name in program.
22862 Number of inputs to the filter. Defaults to 1.
22865 Size of output frames. Defaults to the same as the first input.
22869 The @code{program_opencl} filter also supports the @ref{framesync} options.
22871 The program source file must contain a kernel function with the given name,
22872 which will be run once for each plane of the output. Each run on a plane
22873 gets enqueued as a separate 2D global NDRange with one work-item for each
22874 pixel to be generated. The global ID offset for each work-item is therefore
22875 the coordinates of a pixel in the destination image.
22877 The kernel function needs to take the following arguments:
22880 Destination image, @var{__write_only image2d_t}.
22882 This image will become the output; the kernel should write all of it.
22884 Frame index, @var{unsigned int}.
22886 This is a counter starting from zero and increasing by one for each frame.
22888 Source images, @var{__read_only image2d_t}.
22890 These are the most recent images on each input. The kernel may read from
22891 them to generate the output, but they can't be written to.
22898 Copy the input to the output (output must be the same size as the input).
22900 __kernel void copy(__write_only image2d_t destination,
22901 unsigned int index,
22902 __read_only image2d_t source)
22904 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
22906 int2 location = (int2)(get_global_id(0), get_global_id(1));
22908 float4 value = read_imagef(source, sampler, location);
22910 write_imagef(destination, location, value);
22915 Apply a simple transformation, rotating the input by an amount increasing
22916 with the index counter. Pixel values are linearly interpolated by the
22917 sampler, and the output need not have the same dimensions as the input.
22919 __kernel void rotate_image(__write_only image2d_t dst,
22920 unsigned int index,
22921 __read_only image2d_t src)
22923 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22924 CLK_FILTER_LINEAR);
22926 float angle = (float)index / 100.0f;
22928 float2 dst_dim = convert_float2(get_image_dim(dst));
22929 float2 src_dim = convert_float2(get_image_dim(src));
22931 float2 dst_cen = dst_dim / 2.0f;
22932 float2 src_cen = src_dim / 2.0f;
22934 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
22936 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
22938 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
22939 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
22941 src_pos = src_pos * src_dim / dst_dim;
22943 float2 src_loc = src_pos + src_cen;
22945 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
22946 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
22947 write_imagef(dst, dst_loc, 0.5f);
22949 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
22954 Blend two inputs together, with the amount of each input used varying
22955 with the index counter.
22957 __kernel void blend_images(__write_only image2d_t dst,
22958 unsigned int index,
22959 __read_only image2d_t src1,
22960 __read_only image2d_t src2)
22962 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
22963 CLK_FILTER_LINEAR);
22965 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
22967 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
22968 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
22969 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
22971 float4 val1 = read_imagef(src1, sampler, src1_loc);
22972 float4 val2 = read_imagef(src2, sampler, src2_loc);
22974 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
22980 @section roberts_opencl
22981 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
22983 The filter accepts the following option:
22987 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
22990 Set value which will be multiplied with filtered result.
22991 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
22994 Set value which will be added to filtered result.
22995 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
22998 @subsection Example
23002 Apply the Roberts cross operator with scale set to 2 and delta set to 10
23004 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
23008 @section sobel_opencl
23010 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
23012 The filter accepts the following option:
23016 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
23019 Set value which will be multiplied with filtered result.
23020 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
23023 Set value which will be added to filtered result.
23024 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
23027 @subsection Example
23031 Apply sobel operator with scale set to 2 and delta set to 10
23033 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
23037 @section tonemap_opencl
23039 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
23041 It accepts the following parameters:
23045 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
23048 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
23051 Apply desaturation for highlights that exceed this level of brightness. The
23052 higher the parameter, the more color information will be preserved. This
23053 setting helps prevent unnaturally blown-out colors for super-highlights, by
23054 (smoothly) turning into white instead. This makes images feel more natural,
23055 at the cost of reducing information about out-of-range colors.
23057 The default value is 0.5, and the algorithm here is a little different from
23058 the cpu version tonemap currently. A setting of 0.0 disables this option.
23061 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
23062 is used to detect whether the scene has changed or not. If the distance between
23063 the current frame average brightness and the current running average exceeds
23064 a threshold value, we would re-calculate scene average and peak brightness.
23065 The default value is 0.2.
23068 Specify the output pixel format.
23070 Currently supported formats are:
23077 Set the output color range.
23079 Possible values are:
23085 Default is same as input.
23088 Set the output color primaries.
23090 Possible values are:
23096 Default is same as input.
23099 Set the output transfer characteristics.
23101 Possible values are:
23110 Set the output colorspace matrix.
23112 Possible value are:
23118 Default is same as input.
23122 @subsection Example
23126 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
23128 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
23132 @section unsharp_opencl
23134 Sharpen or blur the input video.
23136 It accepts the following parameters:
23139 @item luma_msize_x, lx
23140 Set the luma matrix horizontal size.
23141 Range is @code{[1, 23]} and default value is @code{5}.
23143 @item luma_msize_y, ly
23144 Set the luma matrix vertical size.
23145 Range is @code{[1, 23]} and default value is @code{5}.
23147 @item luma_amount, la
23148 Set the luma effect strength.
23149 Range is @code{[-10, 10]} and default value is @code{1.0}.
23151 Negative values will blur the input video, while positive values will
23152 sharpen it, a value of zero will disable the effect.
23154 @item chroma_msize_x, cx
23155 Set the chroma matrix horizontal size.
23156 Range is @code{[1, 23]} and default value is @code{5}.
23158 @item chroma_msize_y, cy
23159 Set the chroma matrix vertical size.
23160 Range is @code{[1, 23]} and default value is @code{5}.
23162 @item chroma_amount, ca
23163 Set the chroma effect strength.
23164 Range is @code{[-10, 10]} and default value is @code{0.0}.
23166 Negative values will blur the input video, while positive values will
23167 sharpen it, a value of zero will disable the effect.
23171 All parameters are optional and default to the equivalent of the
23172 string '5:5:1.0:5:5:0.0'.
23174 @subsection Examples
23178 Apply strong luma sharpen effect:
23180 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
23184 Apply a strong blur of both luma and chroma parameters:
23186 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
23190 @section xfade_opencl
23192 Cross fade two videos with custom transition effect by using OpenCL.
23194 It accepts the following options:
23198 Set one of possible transition effects.
23202 Select custom transition effect, the actual transition description
23203 will be picked from source and kernel options.
23215 Default transition is fade.
23219 OpenCL program source file for custom transition.
23222 Set name of kernel to use for custom transition from program source file.
23225 Set duration of video transition.
23228 Set time of start of transition relative to first video.
23231 The program source file must contain a kernel function with the given name,
23232 which will be run once for each plane of the output. Each run on a plane
23233 gets enqueued as a separate 2D global NDRange with one work-item for each
23234 pixel to be generated. The global ID offset for each work-item is therefore
23235 the coordinates of a pixel in the destination image.
23237 The kernel function needs to take the following arguments:
23240 Destination image, @var{__write_only image2d_t}.
23242 This image will become the output; the kernel should write all of it.
23245 First Source image, @var{__read_only image2d_t}.
23246 Second Source image, @var{__read_only image2d_t}.
23248 These are the most recent images on each input. The kernel may read from
23249 them to generate the output, but they can't be written to.
23252 Transition progress, @var{float}. This value is always between 0 and 1 inclusive.
23259 Apply dots curtain transition effect:
23261 __kernel void blend_images(__write_only image2d_t dst,
23262 __read_only image2d_t src1,
23263 __read_only image2d_t src2,
23266 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
23267 CLK_FILTER_LINEAR);
23268 int2 p = (int2)(get_global_id(0), get_global_id(1));
23269 float2 rp = (float2)(get_global_id(0), get_global_id(1));
23270 float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
23273 float2 dots = (float2)(20.0, 20.0);
23274 float2 center = (float2)(0,0);
23277 float4 val1 = read_imagef(src1, sampler, p);
23278 float4 val2 = read_imagef(src2, sampler, p);
23279 bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
23281 write_imagef(dst, p, next ? val1 : val2);
23287 @c man end OPENCL VIDEO FILTERS
23289 @chapter VAAPI Video Filters
23290 @c man begin VAAPI VIDEO FILTERS
23292 VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
23294 To enable compilation of these filters you need to configure FFmpeg with
23295 @code{--enable-vaapi}.
23297 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}
23299 @section tonemap_vaapi
23301 Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping.
23302 It maps the dynamic range of HDR10 content to the SDR content.
23303 It currently only accepts HDR10 as input.
23305 It accepts the following parameters:
23309 Specify the output pixel format.
23311 Currently supported formats are:
23320 Set the output color primaries.
23322 Default is same as input.
23325 Set the output transfer characteristics.
23330 Set the output colorspace matrix.
23332 Default is same as input.
23336 @subsection Example
23340 Convert HDR(HDR10) video to bt2020-transfer-characteristic p010 format
23342 tonemap_vaapi=format=p010:t=bt2020-10
23346 @c man end VAAPI VIDEO FILTERS
23348 @chapter Video Sources
23349 @c man begin VIDEO SOURCES
23351 Below is a description of the currently available video sources.
23355 Buffer video frames, and make them available to the filter chain.
23357 This source is mainly intended for a programmatic use, in particular
23358 through the interface defined in @file{libavfilter/buffersrc.h}.
23360 It accepts the following parameters:
23365 Specify the size (width and height) of the buffered video frames. For the
23366 syntax of this option, check the
23367 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23370 The input video width.
23373 The input video height.
23376 A string representing the pixel format of the buffered video frames.
23377 It may be a number corresponding to a pixel format, or a pixel format
23381 Specify the timebase assumed by the timestamps of the buffered frames.
23384 Specify the frame rate expected for the video stream.
23386 @item pixel_aspect, sar
23387 The sample (pixel) aspect ratio of the input video.
23390 This option is deprecated and ignored. Prepend @code{sws_flags=@var{flags};}
23391 to the filtergraph description to specify swscale flags for automatically
23392 inserted scalers. See @ref{Filtergraph syntax}.
23394 @item hw_frames_ctx
23395 When using a hardware pixel format, this should be a reference to an
23396 AVHWFramesContext describing input frames.
23401 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
23404 will instruct the source to accept video frames with size 320x240 and
23405 with format "yuv410p", assuming 1/24 as the timestamps timebase and
23406 square pixels (1:1 sample aspect ratio).
23407 Since the pixel format with name "yuv410p" corresponds to the number 6
23408 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
23409 this example corresponds to:
23411 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
23414 Alternatively, the options can be specified as a flat string, but this
23415 syntax is deprecated:
23417 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}
23421 Create a pattern generated by an elementary cellular automaton.
23423 The initial state of the cellular automaton can be defined through the
23424 @option{filename} and @option{pattern} options. If such options are
23425 not specified an initial state is created randomly.
23427 At each new frame a new row in the video is filled with the result of
23428 the cellular automaton next generation. The behavior when the whole
23429 frame is filled is defined by the @option{scroll} option.
23431 This source accepts the following options:
23435 Read the initial cellular automaton state, i.e. the starting row, from
23436 the specified file.
23437 In the file, each non-whitespace character is considered an alive
23438 cell, a newline will terminate the row, and further characters in the
23439 file will be ignored.
23442 Read the initial cellular automaton state, i.e. the starting row, from
23443 the specified string.
23445 Each non-whitespace character in the string is considered an alive
23446 cell, a newline will terminate the row, and further characters in the
23447 string will be ignored.
23450 Set the video rate, that is the number of frames generated per second.
23453 @item random_fill_ratio, ratio
23454 Set the random fill ratio for the initial cellular automaton row. It
23455 is a floating point number value ranging from 0 to 1, defaults to
23458 This option is ignored when a file or a pattern is specified.
23460 @item random_seed, seed
23461 Set the seed for filling randomly the initial row, must be an integer
23462 included between 0 and UINT32_MAX. If not specified, or if explicitly
23463 set to -1, the filter will try to use a good random seed on a best
23467 Set the cellular automaton rule, it is a number ranging from 0 to 255.
23468 Default value is 110.
23471 Set the size of the output video. For the syntax of this option, check the
23472 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23474 If @option{filename} or @option{pattern} is specified, the size is set
23475 by default to the width of the specified initial state row, and the
23476 height is set to @var{width} * PHI.
23478 If @option{size} is set, it must contain the width of the specified
23479 pattern string, and the specified pattern will be centered in the
23482 If a filename or a pattern string is not specified, the size value
23483 defaults to "320x518" (used for a randomly generated initial state).
23486 If set to 1, scroll the output upward when all the rows in the output
23487 have been already filled. If set to 0, the new generated row will be
23488 written over the top row just after the bottom row is filled.
23491 @item start_full, full
23492 If set to 1, completely fill the output with generated rows before
23493 outputting the first frame.
23494 This is the default behavior, for disabling set the value to 0.
23497 If set to 1, stitch the left and right row edges together.
23498 This is the default behavior, for disabling set the value to 0.
23501 @subsection Examples
23505 Read the initial state from @file{pattern}, and specify an output of
23508 cellauto=f=pattern:s=200x400
23512 Generate a random initial row with a width of 200 cells, with a fill
23515 cellauto=ratio=2/3:s=200x200
23519 Create a pattern generated by rule 18 starting by a single alive cell
23520 centered on an initial row with width 100:
23522 cellauto=p=@@:s=100x400:full=0:rule=18
23526 Specify a more elaborated initial pattern:
23528 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
23533 @anchor{coreimagesrc}
23534 @section coreimagesrc
23535 Video source generated on GPU using Apple's CoreImage API on OSX.
23537 This video source is a specialized version of the @ref{coreimage} video filter.
23538 Use a core image generator at the beginning of the applied filterchain to
23539 generate the content.
23541 The coreimagesrc video source accepts the following options:
23543 @item list_generators
23544 List all available generators along with all their respective options as well as
23545 possible minimum and maximum values along with the default values.
23547 list_generators=true
23551 Specify the size of the sourced video. For the syntax of this option, check the
23552 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23553 The default value is @code{320x240}.
23556 Specify the frame rate of the sourced video, as the number of frames
23557 generated per second. It has to be a string in the format
23558 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23559 number or a valid video frame rate abbreviation. The default value is
23563 Set the sample aspect ratio of the sourced video.
23566 Set the duration of the sourced video. See
23567 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23568 for the accepted syntax.
23570 If not specified, or the expressed duration is negative, the video is
23571 supposed to be generated forever.
23574 Additionally, all options of the @ref{coreimage} video filter are accepted.
23575 A complete filterchain can be used for further processing of the
23576 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
23577 and examples for details.
23579 @subsection Examples
23584 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
23585 given as complete and escaped command-line for Apple's standard bash shell:
23587 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
23589 This example is equivalent to the QRCode example of @ref{coreimage} without the
23590 need for a nullsrc video source.
23595 Generate several gradients.
23599 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23600 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23603 Set frame rate, expressed as number of frames per second. Default
23606 @item c0, c1, c2, c3, c4, c5, c6, c7
23607 Set 8 colors. Default values for colors is to pick random one.
23609 @item x0, y0, y0, y1
23610 Set gradient line source and destination points. If negative or out of range, random ones
23614 Set number of colors to use at once. Allowed range is from 2 to 8. Default value is 2.
23617 Set seed for picking gradient line points.
23620 Set the duration of the sourced video. See
23621 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23622 for the accepted syntax.
23624 If not specified, or the expressed duration is negative, the video is
23625 supposed to be generated forever.
23628 Set speed of gradients rotation.
23632 @section mandelbrot
23634 Generate a Mandelbrot set fractal, and progressively zoom towards the
23635 point specified with @var{start_x} and @var{start_y}.
23637 This source accepts the following options:
23642 Set the terminal pts value. Default value is 400.
23645 Set the terminal scale value.
23646 Must be a floating point value. Default value is 0.3.
23649 Set the inner coloring mode, that is the algorithm used to draw the
23650 Mandelbrot fractal internal region.
23652 It shall assume one of the following values:
23657 Show time until convergence.
23659 Set color based on point closest to the origin of the iterations.
23664 Default value is @var{mincol}.
23667 Set the bailout value. Default value is 10.0.
23670 Set the maximum of iterations performed by the rendering
23671 algorithm. Default value is 7189.
23674 Set outer coloring mode.
23675 It shall assume one of following values:
23677 @item iteration_count
23678 Set iteration count mode.
23679 @item normalized_iteration_count
23680 set normalized iteration count mode.
23682 Default value is @var{normalized_iteration_count}.
23685 Set frame rate, expressed as number of frames per second. Default
23689 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
23690 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
23693 Set the initial scale value. Default value is 3.0.
23696 Set the initial x position. Must be a floating point value between
23697 -100 and 100. Default value is -0.743643887037158704752191506114774.
23700 Set the initial y position. Must be a floating point value between
23701 -100 and 100. Default value is -0.131825904205311970493132056385139.
23706 Generate various test patterns, as generated by the MPlayer test filter.
23708 The size of the generated video is fixed, and is 256x256.
23709 This source is useful in particular for testing encoding features.
23711 This source accepts the following options:
23716 Specify the frame rate of the sourced video, as the number of frames
23717 generated per second. It has to be a string in the format
23718 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
23719 number or a valid video frame rate abbreviation. The default value is
23723 Set the duration of the sourced video. See
23724 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
23725 for the accepted syntax.
23727 If not specified, or the expressed duration is negative, the video is
23728 supposed to be generated forever.
23732 Set the number or the name of the test to perform. Supported tests are:
23746 @item max_frames, m
23747 Set the maximum number of frames generated for each test, default value is 30.
23751 Default value is "all", which will cycle through the list of all tests.
23756 mptestsrc=t=dc_luma
23759 will generate a "dc_luma" test pattern.
23761 @section frei0r_src
23763 Provide a frei0r source.
23765 To enable compilation of this filter you need to install the frei0r
23766 header and configure FFmpeg with @code{--enable-frei0r}.
23768 This source accepts the following parameters:
23773 The size of the video to generate. For the syntax of this option, check the
23774 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23777 The framerate of the generated video. It may be a string of the form
23778 @var{num}/@var{den} or a frame rate abbreviation.
23781 The name to the frei0r source to load. For more information regarding frei0r and
23782 how to set the parameters, read the @ref{frei0r} section in the video filters
23785 @item filter_params
23786 A '|'-separated list of parameters to pass to the frei0r source.
23790 For example, to generate a frei0r partik0l source with size 200x200
23791 and frame rate 10 which is overlaid on the overlay filter main input:
23793 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
23798 Generate a life pattern.
23800 This source is based on a generalization of John Conway's life game.
23802 The sourced input represents a life grid, each pixel represents a cell
23803 which can be in one of two possible states, alive or dead. Every cell
23804 interacts with its eight neighbours, which are the cells that are
23805 horizontally, vertically, or diagonally adjacent.
23807 At each interaction the grid evolves according to the adopted rule,
23808 which specifies the number of neighbor alive cells which will make a
23809 cell stay alive or born. The @option{rule} option allows one to specify
23812 This source accepts the following options:
23816 Set the file from which to read the initial grid state. In the file,
23817 each non-whitespace character is considered an alive cell, and newline
23818 is used to delimit the end of each row.
23820 If this option is not specified, the initial grid is generated
23824 Set the video rate, that is the number of frames generated per second.
23827 @item random_fill_ratio, ratio
23828 Set the random fill ratio for the initial random grid. It is a
23829 floating point number value ranging from 0 to 1, defaults to 1/PHI.
23830 It is ignored when a file is specified.
23832 @item random_seed, seed
23833 Set the seed for filling the initial random grid, must be an integer
23834 included between 0 and UINT32_MAX. If not specified, or if explicitly
23835 set to -1, the filter will try to use a good random seed on a best
23841 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
23842 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
23843 @var{NS} specifies the number of alive neighbor cells which make a
23844 live cell stay alive, and @var{NB} the number of alive neighbor cells
23845 which make a dead cell to become alive (i.e. to "born").
23846 "s" and "b" can be used in place of "S" and "B", respectively.
23848 Alternatively a rule can be specified by an 18-bits integer. The 9
23849 high order bits are used to encode the next cell state if it is alive
23850 for each number of neighbor alive cells, the low order bits specify
23851 the rule for "borning" new cells. Higher order bits encode for an
23852 higher number of neighbor cells.
23853 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
23854 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
23856 Default value is "S23/B3", which is the original Conway's game of life
23857 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
23858 cells, and will born a new cell if there are three alive cells around
23862 Set the size of the output video. For the syntax of this option, check the
23863 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23865 If @option{filename} is specified, the size is set by default to the
23866 same size of the input file. If @option{size} is set, it must contain
23867 the size specified in the input file, and the initial grid defined in
23868 that file is centered in the larger resulting area.
23870 If a filename is not specified, the size value defaults to "320x240"
23871 (used for a randomly generated initial grid).
23874 If set to 1, stitch the left and right grid edges together, and the
23875 top and bottom edges also. Defaults to 1.
23878 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
23879 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
23880 value from 0 to 255.
23883 Set the color of living (or new born) cells.
23886 Set the color of dead cells. If @option{mold} is set, this is the first color
23887 used to represent a dead cell.
23890 Set mold color, for definitely dead and moldy cells.
23892 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
23893 ffmpeg-utils manual,ffmpeg-utils}.
23896 @subsection Examples
23900 Read a grid from @file{pattern}, and center it on a grid of size
23903 life=f=pattern:s=300x300
23907 Generate a random grid of size 200x200, with a fill ratio of 2/3:
23909 life=ratio=2/3:s=200x200
23913 Specify a custom rule for evolving a randomly generated grid:
23919 Full example with slow death effect (mold) using @command{ffplay}:
23921 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
23928 @anchor{haldclutsrc}
23931 @anchor{pal100bars}
23932 @anchor{rgbtestsrc}
23934 @anchor{smptehdbars}
23937 @anchor{yuvtestsrc}
23938 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
23940 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
23942 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
23944 The @code{color} source provides an uniformly colored input.
23946 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
23947 @ref{haldclut} filter.
23949 The @code{nullsrc} source returns unprocessed video frames. It is
23950 mainly useful to be employed in analysis / debugging tools, or as the
23951 source for filters which ignore the input data.
23953 The @code{pal75bars} source generates a color bars pattern, based on
23954 EBU PAL recommendations with 75% color levels.
23956 The @code{pal100bars} source generates a color bars pattern, based on
23957 EBU PAL recommendations with 100% color levels.
23959 The @code{rgbtestsrc} source generates an RGB test pattern useful for
23960 detecting RGB vs BGR issues. You should see a red, green and blue
23961 stripe from top to bottom.
23963 The @code{smptebars} source generates a color bars pattern, based on
23964 the SMPTE Engineering Guideline EG 1-1990.
23966 The @code{smptehdbars} source generates a color bars pattern, based on
23967 the SMPTE RP 219-2002.
23969 The @code{testsrc} source generates a test video pattern, showing a
23970 color pattern, a scrolling gradient and a timestamp. This is mainly
23971 intended for testing purposes.
23973 The @code{testsrc2} source is similar to testsrc, but supports more
23974 pixel formats instead of just @code{rgb24}. This allows using it as an
23975 input for other tests without requiring a format conversion.
23977 The @code{yuvtestsrc} source generates an YUV test pattern. You should
23978 see a y, cb and cr stripe from top to bottom.
23980 The sources accept the following parameters:
23985 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
23986 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
23987 pixels to be used as identity matrix for 3D lookup tables. Each component is
23988 coded on a @code{1/(N*N)} scale.
23991 Specify the color of the source, only available in the @code{color}
23992 source. For the syntax of this option, check the
23993 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
23996 Specify the size of the sourced video. For the syntax of this option, check the
23997 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23998 The default value is @code{320x240}.
24000 This option is not available with the @code{allrgb}, @code{allyuv}, and
24001 @code{haldclutsrc} filters.
24004 Specify the frame rate of the sourced video, as the number of frames
24005 generated per second. It has to be a string in the format
24006 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
24007 number or a valid video frame rate abbreviation. The default value is
24011 Set the duration of the sourced video. See
24012 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
24013 for the accepted syntax.
24015 If not specified, or the expressed duration is negative, the video is
24016 supposed to be generated forever.
24018 Since the frame rate is used as time base, all frames including the last one
24019 will have their full duration. If the specified duration is not a multiple
24020 of the frame duration, it will be rounded up.
24023 Set the sample aspect ratio of the sourced video.
24026 Specify the alpha (opacity) of the background, only available in the
24027 @code{testsrc2} source. The value must be between 0 (fully transparent) and
24028 255 (fully opaque, the default).
24031 Set the number of decimals to show in the timestamp, only available in the
24032 @code{testsrc} source.
24034 The displayed timestamp value will correspond to the original
24035 timestamp value multiplied by the power of 10 of the specified
24036 value. Default value is 0.
24039 @subsection Examples
24043 Generate a video with a duration of 5.3 seconds, with size
24044 176x144 and a frame rate of 10 frames per second:
24046 testsrc=duration=5.3:size=qcif:rate=10
24050 The following graph description will generate a red source
24051 with an opacity of 0.2, with size "qcif" and a frame rate of 10
24054 color=c=red@@0.2:s=qcif:r=10
24058 If the input content is to be ignored, @code{nullsrc} can be used. The
24059 following command generates noise in the luminance plane by employing
24060 the @code{geq} filter:
24062 nullsrc=s=256x256, geq=random(1)*255:128:128
24066 @subsection Commands
24068 The @code{color} source supports the following commands:
24072 Set the color of the created image. Accepts the same syntax of the
24073 corresponding @option{color} option.
24078 Generate video using an OpenCL program.
24083 OpenCL program source file.
24086 Kernel name in program.
24089 Size of frames to generate. This must be set.
24092 Pixel format to use for the generated frames. This must be set.
24095 Number of frames generated every second. Default value is '25'.
24099 For details of how the program loading works, see the @ref{program_opencl}
24106 Generate a colour ramp by setting pixel values from the position of the pixel
24107 in the output image. (Note that this will work with all pixel formats, but
24108 the generated output will not be the same.)
24110 __kernel void ramp(__write_only image2d_t dst,
24111 unsigned int index)
24113 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24116 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
24118 write_imagef(dst, loc, val);
24123 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
24125 __kernel void sierpinski_carpet(__write_only image2d_t dst,
24126 unsigned int index)
24128 int2 loc = (int2)(get_global_id(0), get_global_id(1));
24130 float4 value = 0.0f;
24131 int x = loc.x + index;
24132 int y = loc.y + index;
24133 while (x > 0 || y > 0) {
24134 if (x % 3 == 1 && y % 3 == 1) {
24142 write_imagef(dst, loc, value);
24148 @section sierpinski
24150 Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
24152 This source accepts the following options:
24156 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
24157 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
24160 Set frame rate, expressed as number of frames per second. Default
24164 Set seed which is used for random panning.
24167 Set max jump for single pan destination. Allowed range is from 1 to 10000.
24170 Set fractal type, can be default @code{carpet} or @code{triangle}.
24173 @c man end VIDEO SOURCES
24175 @chapter Video Sinks
24176 @c man begin VIDEO SINKS
24178 Below is a description of the currently available video sinks.
24180 @section buffersink
24182 Buffer video frames, and make them available to the end of the filter
24185 This sink is mainly intended for programmatic use, in particular
24186 through the interface defined in @file{libavfilter/buffersink.h}
24187 or the options system.
24189 It accepts a pointer to an AVBufferSinkContext structure, which
24190 defines the incoming buffers' formats, to be passed as the opaque
24191 parameter to @code{avfilter_init_filter} for initialization.
24195 Null video sink: do absolutely nothing with the input video. It is
24196 mainly useful as a template and for use in analysis / debugging
24199 @c man end VIDEO SINKS
24201 @chapter Multimedia Filters
24202 @c man begin MULTIMEDIA FILTERS
24204 Below is a description of the currently available multimedia filters.
24208 Convert input audio to a video output, displaying the audio bit scope.
24210 The filter accepts the following options:
24214 Set frame rate, expressed as number of frames per second. Default
24218 Specify the video size for the output. For the syntax of this option, check the
24219 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24220 Default value is @code{1024x256}.
24223 Specify list of colors separated by space or by '|' which will be used to
24224 draw channels. Unrecognized or missing colors will be replaced
24228 @section adrawgraph
24229 Draw a graph using input audio metadata.
24231 See @ref{drawgraph}
24233 @section agraphmonitor
24235 See @ref{graphmonitor}.
24237 @section ahistogram
24239 Convert input audio to a video output, displaying the volume histogram.
24241 The filter accepts the following options:
24245 Specify how histogram is calculated.
24247 It accepts the following values:
24250 Use single histogram for all channels.
24252 Use separate histogram for each channel.
24254 Default is @code{single}.
24257 Set frame rate, expressed as number of frames per second. Default
24261 Specify the video size for the output. For the syntax of this option, check the
24262 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24263 Default value is @code{hd720}.
24268 It accepts the following values:
24279 reverse logarithmic
24281 Default is @code{log}.
24284 Set amplitude scale.
24286 It accepts the following values:
24293 Default is @code{log}.
24296 Set how much frames to accumulate in histogram.
24297 Default is 1. Setting this to -1 accumulates all frames.
24300 Set histogram ratio of window height.
24303 Set sonogram sliding.
24305 It accepts the following values:
24308 replace old rows with new ones.
24310 scroll from top to bottom.
24312 Default is @code{replace}.
24315 @section aphasemeter
24317 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
24318 representing mean phase of current audio frame. A video output can also be produced and is
24319 enabled by default. The audio is passed through as first output.
24321 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
24322 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
24323 and @code{1} means channels are in phase.
24325 The filter accepts the following options, all related to its video output:
24329 Set the output frame rate. Default value is @code{25}.
24332 Set the video size for the output. For the syntax of this option, check the
24333 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24334 Default value is @code{800x400}.
24339 Specify the red, green, blue contrast. Default values are @code{2},
24340 @code{7} and @code{1}.
24341 Allowed range is @code{[0, 255]}.
24344 Set color which will be used for drawing median phase. If color is
24345 @code{none} which is default, no median phase value will be drawn.
24348 Enable video output. Default is enabled.
24351 @subsection phasing detection
24353 The filter also detects out of phase and mono sequences in stereo streams.
24354 It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
24356 The filter accepts the following options for this detection:
24360 Enable mono and out of phase detection. Default is disabled.
24363 Set phase tolerance for mono detection, in amplitude ratio. Default is @code{0}.
24364 Allowed range is @code{[0, 1]}.
24367 Set angle threshold for out of phase detection, in degree. Default is @code{170}.
24368 Allowed range is @code{[90, 180]}.
24371 Set mono or out of phase duration until notification, expressed in seconds. Default is @code{2}.
24374 @subsection Examples
24378 Complete example with @command{ffmpeg} to detect 1 second of mono with 0.001 phase tolerance:
24380 ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
24384 @section avectorscope
24386 Convert input audio to a video output, representing the audio vector
24389 The filter is used to measure the difference between channels of stereo
24390 audio stream. A monaural signal, consisting of identical left and right
24391 signal, results in straight vertical line. Any stereo separation is visible
24392 as a deviation from this line, creating a Lissajous figure.
24393 If the straight (or deviation from it) but horizontal line appears this
24394 indicates that the left and right channels are out of phase.
24396 The filter accepts the following options:
24400 Set the vectorscope mode.
24402 Available values are:
24405 Lissajous rotated by 45 degrees.
24408 Same as above but not rotated.
24411 Shape resembling half of circle.
24414 Default value is @samp{lissajous}.
24417 Set the video size for the output. For the syntax of this option, check the
24418 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24419 Default value is @code{400x400}.
24422 Set the output frame rate. Default value is @code{25}.
24428 Specify the red, green, blue and alpha contrast. Default values are @code{40},
24429 @code{160}, @code{80} and @code{255}.
24430 Allowed range is @code{[0, 255]}.
24436 Specify the red, green, blue and alpha fade. Default values are @code{15},
24437 @code{10}, @code{5} and @code{5}.
24438 Allowed range is @code{[0, 255]}.
24441 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
24442 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
24445 Set the vectorscope drawing mode.
24447 Available values are:
24450 Draw dot for each sample.
24453 Draw line between previous and current sample.
24456 Default value is @samp{dot}.
24459 Specify amplitude scale of audio samples.
24461 Available values are:
24477 Swap left channel axis with right channel axis.
24487 Mirror only x axis.
24490 Mirror only y axis.
24498 @subsection Examples
24502 Complete example using @command{ffplay}:
24504 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
24505 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
24509 @section bench, abench
24511 Benchmark part of a filtergraph.
24513 The filter accepts the following options:
24517 Start or stop a timer.
24519 Available values are:
24522 Get the current time, set it as frame metadata (using the key
24523 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
24526 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
24527 the input frame metadata to get the time difference. Time difference, average,
24528 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
24529 @code{min}) are then printed. The timestamps are expressed in seconds.
24533 @subsection Examples
24537 Benchmark @ref{selectivecolor} filter:
24539 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
24545 Concatenate audio and video streams, joining them together one after the
24548 The filter works on segments of synchronized video and audio streams. All
24549 segments must have the same number of streams of each type, and that will
24550 also be the number of streams at output.
24552 The filter accepts the following options:
24557 Set the number of segments. Default is 2.
24560 Set the number of output video streams, that is also the number of video
24561 streams in each segment. Default is 1.
24564 Set the number of output audio streams, that is also the number of audio
24565 streams in each segment. Default is 0.
24568 Activate unsafe mode: do not fail if segments have a different format.
24572 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
24573 @var{a} audio outputs.
24575 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
24576 segment, in the same order as the outputs, then the inputs for the second
24579 Related streams do not always have exactly the same duration, for various
24580 reasons including codec frame size or sloppy authoring. For that reason,
24581 related synchronized streams (e.g. a video and its audio track) should be
24582 concatenated at once. The concat filter will use the duration of the longest
24583 stream in each segment (except the last one), and if necessary pad shorter
24584 audio streams with silence.
24586 For this filter to work correctly, all segments must start at timestamp 0.
24588 All corresponding streams must have the same parameters in all segments; the
24589 filtering system will automatically select a common pixel format for video
24590 streams, and a common sample format, sample rate and channel layout for
24591 audio streams, but other settings, such as resolution, must be converted
24592 explicitly by the user.
24594 Different frame rates are acceptable but will result in variable frame rate
24595 at output; be sure to configure the output file to handle it.
24597 @subsection Examples
24601 Concatenate an opening, an episode and an ending, all in bilingual version
24602 (video in stream 0, audio in streams 1 and 2):
24604 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
24605 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
24606 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
24607 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
24611 Concatenate two parts, handling audio and video separately, using the
24612 (a)movie sources, and adjusting the resolution:
24614 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
24615 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
24616 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
24618 Note that a desync will happen at the stitch if the audio and video streams
24619 do not have exactly the same duration in the first file.
24623 @subsection Commands
24625 This filter supports the following commands:
24628 Close the current segment and step to the next one
24634 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
24635 level. By default, it logs a message at a frequency of 10Hz with the
24636 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
24637 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
24639 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
24640 sample format is double-precision floating point. The input stream will be converted to
24641 this specification, if needed. Users may need to insert aformat and/or aresample filters
24642 after this filter to obtain the original parameters.
24644 The filter also has a video output (see the @var{video} option) with a real
24645 time graph to observe the loudness evolution. The graphic contains the logged
24646 message mentioned above, so it is not printed anymore when this option is set,
24647 unless the verbose logging is set. The main graphing area contains the
24648 short-term loudness (3 seconds of analysis), and the gauge on the right is for
24649 the momentary loudness (400 milliseconds), but can optionally be configured
24650 to instead display short-term loudness (see @var{gauge}).
24652 The green area marks a +/- 1LU target range around the target loudness
24653 (-23LUFS by default, unless modified through @var{target}).
24655 More information about the Loudness Recommendation EBU R128 on
24656 @url{http://tech.ebu.ch/loudness}.
24658 The filter accepts the following options:
24663 Activate the video output. The audio stream is passed unchanged whether this
24664 option is set or no. The video stream will be the first output stream if
24665 activated. Default is @code{0}.
24668 Set the video size. This option is for video only. For the syntax of this
24670 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
24671 Default and minimum resolution is @code{640x480}.
24674 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
24675 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
24676 other integer value between this range is allowed.
24679 Set metadata injection. If set to @code{1}, the audio input will be segmented
24680 into 100ms output frames, each of them containing various loudness information
24681 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
24683 Default is @code{0}.
24686 Force the frame logging level.
24688 Available values are:
24691 information logging level
24693 verbose logging level
24696 By default, the logging level is set to @var{info}. If the @option{video} or
24697 the @option{metadata} options are set, it switches to @var{verbose}.
24702 Available modes can be cumulated (the option is a @code{flag} type). Possible
24706 Disable any peak mode (default).
24708 Enable sample-peak mode.
24710 Simple peak mode looking for the higher sample value. It logs a message
24711 for sample-peak (identified by @code{SPK}).
24713 Enable true-peak mode.
24715 If enabled, the peak lookup is done on an over-sampled version of the input
24716 stream for better peak accuracy. It logs a message for true-peak.
24717 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
24718 This mode requires a build with @code{libswresample}.
24722 Treat mono input files as "dual mono". If a mono file is intended for playback
24723 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
24724 If set to @code{true}, this option will compensate for this effect.
24725 Multi-channel input files are not affected by this option.
24728 Set a specific pan law to be used for the measurement of dual mono files.
24729 This parameter is optional, and has a default value of -3.01dB.
24732 Set a specific target level (in LUFS) used as relative zero in the visualization.
24733 This parameter is optional and has a default value of -23LUFS as specified
24734 by EBU R128. However, material published online may prefer a level of -16LUFS
24735 (e.g. for use with podcasts or video platforms).
24738 Set the value displayed by the gauge. Valid values are @code{momentary} and s
24739 @code{shortterm}. By default the momentary value will be used, but in certain
24740 scenarios it may be more useful to observe the short term value instead (e.g.
24744 Sets the display scale for the loudness. Valid parameters are @code{absolute}
24745 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
24746 video output, not the summary or continuous log output.
24749 @subsection Examples
24753 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
24755 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
24759 Run an analysis with @command{ffmpeg}:
24761 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
24765 @section interleave, ainterleave
24767 Temporally interleave frames from several inputs.
24769 @code{interleave} works with video inputs, @code{ainterleave} with audio.
24771 These filters read frames from several inputs and send the oldest
24772 queued frame to the output.
24774 Input streams must have well defined, monotonically increasing frame
24777 In order to submit one frame to output, these filters need to enqueue
24778 at least one frame for each input, so they cannot work in case one
24779 input is not yet terminated and will not receive incoming frames.
24781 For example consider the case when one input is a @code{select} filter
24782 which always drops input frames. The @code{interleave} filter will keep
24783 reading from that input, but it will never be able to send new frames
24784 to output until the input sends an end-of-stream signal.
24786 Also, depending on inputs synchronization, the filters will drop
24787 frames in case one input receives more frames than the other ones, and
24788 the queue is already filled.
24790 These filters accept the following options:
24794 Set the number of different inputs, it is 2 by default.
24797 How to determine the end-of-stream.
24801 The duration of the longest input. (default)
24804 The duration of the shortest input.
24807 The duration of the first input.
24812 @subsection Examples
24816 Interleave frames belonging to different streams using @command{ffmpeg}:
24818 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
24822 Add flickering blur effect:
24824 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
24828 @section metadata, ametadata
24830 Manipulate frame metadata.
24832 This filter accepts the following options:
24836 Set mode of operation of the filter.
24838 Can be one of the following:
24842 If both @code{value} and @code{key} is set, select frames
24843 which have such metadata. If only @code{key} is set, select
24844 every frame that has such key in metadata.
24847 Add new metadata @code{key} and @code{value}. If key is already available
24851 Modify value of already present key.
24854 If @code{value} is set, delete only keys that have such value.
24855 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
24859 Print key and its value if metadata was found. If @code{key} is not set print all
24860 metadata values available in frame.
24864 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
24867 Set metadata value which will be used. This option is mandatory for
24868 @code{modify} and @code{add} mode.
24871 Which function to use when comparing metadata value and @code{value}.
24873 Can be one of following:
24877 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
24880 Values are interpreted as strings, returns true if metadata value starts with
24881 the @code{value} option string.
24884 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
24887 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
24890 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
24893 Values are interpreted as floats, returns true if expression from option @code{expr}
24897 Values are interpreted as strings, returns true if metadata value ends with
24898 the @code{value} option string.
24902 Set expression which is used when @code{function} is set to @code{expr}.
24903 The expression is evaluated through the eval API and can contain the following
24908 Float representation of @code{value} from metadata key.
24911 Float representation of @code{value} as supplied by user in @code{value} option.
24915 If specified in @code{print} mode, output is written to the named file. Instead of
24916 plain filename any writable url can be specified. Filename ``-'' is a shorthand
24917 for standard output. If @code{file} option is not set, output is written to the log
24918 with AV_LOG_INFO loglevel.
24921 Reduces buffering in print mode when output is written to a URL set using @var{file}.
24925 @subsection Examples
24929 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
24932 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
24935 Print silencedetect output to file @file{metadata.txt}.
24937 silencedetect,ametadata=mode=print:file=metadata.txt
24940 Direct all metadata to a pipe with file descriptor 4.
24942 metadata=mode=print:file='pipe\:4'
24946 @section perms, aperms
24948 Set read/write permissions for the output frames.
24950 These filters are mainly aimed at developers to test direct path in the
24951 following filter in the filtergraph.
24953 The filters accept the following options:
24957 Select the permissions mode.
24959 It accepts the following values:
24962 Do nothing. This is the default.
24964 Set all the output frames read-only.
24966 Set all the output frames directly writable.
24968 Make the frame read-only if writable, and writable if read-only.
24970 Set each output frame read-only or writable randomly.
24974 Set the seed for the @var{random} mode, must be an integer included between
24975 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
24976 @code{-1}, the filter will try to use a good random seed on a best effort
24980 Note: in case of auto-inserted filter between the permission filter and the
24981 following one, the permission might not be received as expected in that
24982 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
24983 perms/aperms filter can avoid this problem.
24985 @section realtime, arealtime
24987 Slow down filtering to match real time approximately.
24989 These filters will pause the filtering for a variable amount of time to
24990 match the output rate with the input timestamps.
24991 They are similar to the @option{re} option to @code{ffmpeg}.
24993 They accept the following options:
24997 Time limit for the pauses. Any pause longer than that will be considered
24998 a timestamp discontinuity and reset the timer. Default is 2 seconds.
25000 Speed factor for processing. The value must be a float larger than zero.
25001 Values larger than 1.0 will result in faster than realtime processing,
25002 smaller will slow processing down. The @var{limit} is automatically adapted
25003 accordingly. Default is 1.0.
25005 A processing speed faster than what is possible without these filters cannot
25010 @section select, aselect
25012 Select frames to pass in output.
25014 This filter accepts the following options:
25019 Set expression, which is evaluated for each input frame.
25021 If the expression is evaluated to zero, the frame is discarded.
25023 If the evaluation result is negative or NaN, the frame is sent to the
25024 first output; otherwise it is sent to the output with index
25025 @code{ceil(val)-1}, assuming that the input index starts from 0.
25027 For example a value of @code{1.2} corresponds to the output with index
25028 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
25031 Set the number of outputs. The output to which to send the selected
25032 frame is based on the result of the evaluation. Default value is 1.
25035 The expression can contain the following constants:
25039 The (sequential) number of the filtered frame, starting from 0.
25042 The (sequential) number of the selected frame, starting from 0.
25044 @item prev_selected_n
25045 The sequential number of the last selected frame. It's NAN if undefined.
25048 The timebase of the input timestamps.
25051 The PTS (Presentation TimeStamp) of the filtered video frame,
25052 expressed in @var{TB} units. It's NAN if undefined.
25055 The PTS of the filtered video frame,
25056 expressed in seconds. It's NAN if undefined.
25059 The PTS of the previously filtered video frame. It's NAN if undefined.
25061 @item prev_selected_pts
25062 The PTS of the last previously filtered video frame. It's NAN if undefined.
25064 @item prev_selected_t
25065 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
25068 The PTS of the first video frame in the video. It's NAN if undefined.
25071 The time of the first video frame in the video. It's NAN if undefined.
25073 @item pict_type @emph{(video only)}
25074 The type of the filtered frame. It can assume one of the following
25086 @item interlace_type @emph{(video only)}
25087 The frame interlace type. It can assume one of the following values:
25090 The frame is progressive (not interlaced).
25092 The frame is top-field-first.
25094 The frame is bottom-field-first.
25097 @item consumed_sample_n @emph{(audio only)}
25098 the number of selected samples before the current frame
25100 @item samples_n @emph{(audio only)}
25101 the number of samples in the current frame
25103 @item sample_rate @emph{(audio only)}
25104 the input sample rate
25107 This is 1 if the filtered frame is a key-frame, 0 otherwise.
25110 the position in the file of the filtered frame, -1 if the information
25111 is not available (e.g. for synthetic video)
25113 @item scene @emph{(video only)}
25114 value between 0 and 1 to indicate a new scene; a low value reflects a low
25115 probability for the current frame to introduce a new scene, while a higher
25116 value means the current frame is more likely to be one (see the example below)
25118 @item concatdec_select
25119 The concat demuxer can select only part of a concat input file by setting an
25120 inpoint and an outpoint, but the output packets may not be entirely contained
25121 in the selected interval. By using this variable, it is possible to skip frames
25122 generated by the concat demuxer which are not exactly contained in the selected
25125 This works by comparing the frame pts against the @var{lavf.concat.start_time}
25126 and the @var{lavf.concat.duration} packet metadata values which are also
25127 present in the decoded frames.
25129 The @var{concatdec_select} variable is -1 if the frame pts is at least
25130 start_time and either the duration metadata is missing or the frame pts is less
25131 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
25134 That basically means that an input frame is selected if its pts is within the
25135 interval set by the concat demuxer.
25139 The default value of the select expression is "1".
25141 @subsection Examples
25145 Select all frames in input:
25150 The example above is the same as:
25162 Select only I-frames:
25164 select='eq(pict_type\,I)'
25168 Select one frame every 100:
25170 select='not(mod(n\,100))'
25174 Select only frames contained in the 10-20 time interval:
25176 select=between(t\,10\,20)
25180 Select only I-frames contained in the 10-20 time interval:
25182 select=between(t\,10\,20)*eq(pict_type\,I)
25186 Select frames with a minimum distance of 10 seconds:
25188 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
25192 Use aselect to select only audio frames with samples number > 100:
25194 aselect='gt(samples_n\,100)'
25198 Create a mosaic of the first scenes:
25200 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
25203 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
25207 Send even and odd frames to separate outputs, and compose them:
25209 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
25213 Select useful frames from an ffconcat file which is using inpoints and
25214 outpoints but where the source files are not intra frame only.
25216 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
25220 @section sendcmd, asendcmd
25222 Send commands to filters in the filtergraph.
25224 These filters read commands to be sent to other filters in the
25227 @code{sendcmd} must be inserted between two video filters,
25228 @code{asendcmd} must be inserted between two audio filters, but apart
25229 from that they act the same way.
25231 The specification of commands can be provided in the filter arguments
25232 with the @var{commands} option, or in a file specified by the
25233 @var{filename} option.
25235 These filters accept the following options:
25238 Set the commands to be read and sent to the other filters.
25240 Set the filename of the commands to be read and sent to the other
25244 @subsection Commands syntax
25246 A commands description consists of a sequence of interval
25247 specifications, comprising a list of commands to be executed when a
25248 particular event related to that interval occurs. The occurring event
25249 is typically the current frame time entering or leaving a given time
25252 An interval is specified by the following syntax:
25254 @var{START}[-@var{END}] @var{COMMANDS};
25257 The time interval is specified by the @var{START} and @var{END} times.
25258 @var{END} is optional and defaults to the maximum time.
25260 The current frame time is considered within the specified interval if
25261 it is included in the interval [@var{START}, @var{END}), that is when
25262 the time is greater or equal to @var{START} and is lesser than
25265 @var{COMMANDS} consists of a sequence of one or more command
25266 specifications, separated by ",", relating to that interval. The
25267 syntax of a command specification is given by:
25269 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
25272 @var{FLAGS} is optional and specifies the type of events relating to
25273 the time interval which enable sending the specified command, and must
25274 be a non-null sequence of identifier flags separated by "+" or "|" and
25275 enclosed between "[" and "]".
25277 The following flags are recognized:
25280 The command is sent when the current frame timestamp enters the
25281 specified interval. In other words, the command is sent when the
25282 previous frame timestamp was not in the given interval, and the
25286 The command is sent when the current frame timestamp leaves the
25287 specified interval. In other words, the command is sent when the
25288 previous frame timestamp was in the given interval, and the
25292 The command @var{ARG} is interpreted as expression and result of
25293 expression is passed as @var{ARG}.
25295 The expression is evaluated through the eval API and can contain the following
25300 Original position in the file of the frame, or undefined if undefined
25301 for the current frame.
25304 The presentation timestamp in input.
25307 The count of the input frame for video or audio, starting from 0.
25310 The time in seconds of the current frame.
25313 The start time in seconds of the current command interval.
25316 The end time in seconds of the current command interval.
25319 The interpolated time of the current command interval, TI = (T - TS) / (TE - TS).
25324 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
25327 @var{TARGET} specifies the target of the command, usually the name of
25328 the filter class or a specific filter instance name.
25330 @var{COMMAND} specifies the name of the command for the target filter.
25332 @var{ARG} is optional and specifies the optional list of argument for
25333 the given @var{COMMAND}.
25335 Between one interval specification and another, whitespaces, or
25336 sequences of characters starting with @code{#} until the end of line,
25337 are ignored and can be used to annotate comments.
25339 A simplified BNF description of the commands specification syntax
25342 @var{COMMAND_FLAG} ::= "enter" | "leave"
25343 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
25344 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
25345 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
25346 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
25347 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
25350 @subsection Examples
25354 Specify audio tempo change at second 4:
25356 asendcmd=c='4.0 atempo tempo 1.5',atempo
25360 Target a specific filter instance:
25362 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
25366 Specify a list of drawtext and hue commands in a file.
25368 # show text in the interval 5-10
25369 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
25370 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
25372 # desaturate the image in the interval 15-20
25373 15.0-20.0 [enter] hue s 0,
25374 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
25376 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
25378 # apply an exponential saturation fade-out effect, starting from time 25
25379 25 [enter] hue s exp(25-t)
25382 A filtergraph allowing to read and process the above command list
25383 stored in a file @file{test.cmd}, can be specified with:
25385 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
25390 @section setpts, asetpts
25392 Change the PTS (presentation timestamp) of the input frames.
25394 @code{setpts} works on video frames, @code{asetpts} on audio frames.
25396 This filter accepts the following options:
25401 The expression which is evaluated for each frame to construct its timestamp.
25405 The expression is evaluated through the eval API and can contain the following
25409 @item FRAME_RATE, FR
25410 frame rate, only defined for constant frame-rate video
25413 The presentation timestamp in input
25416 The count of the input frame for video or the number of consumed samples,
25417 not including the current frame for audio, starting from 0.
25419 @item NB_CONSUMED_SAMPLES
25420 The number of consumed samples, not including the current frame (only
25423 @item NB_SAMPLES, S
25424 The number of samples in the current frame (only audio)
25426 @item SAMPLE_RATE, SR
25427 The audio sample rate.
25430 The PTS of the first frame.
25433 the time in seconds of the first frame
25436 State whether the current frame is interlaced.
25439 the time in seconds of the current frame
25442 original position in the file of the frame, or undefined if undefined
25443 for the current frame
25446 The previous input PTS.
25449 previous input time in seconds
25452 The previous output PTS.
25455 previous output time in seconds
25458 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
25462 The wallclock (RTC) time at the start of the movie in microseconds.
25465 The timebase of the input timestamps.
25469 @subsection Examples
25473 Start counting PTS from zero
25475 setpts=PTS-STARTPTS
25479 Apply fast motion effect:
25485 Apply slow motion effect:
25491 Set fixed rate of 25 frames per second:
25497 Set fixed rate 25 fps with some jitter:
25499 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
25503 Apply an offset of 10 seconds to the input PTS:
25509 Generate timestamps from a "live source" and rebase onto the current timebase:
25511 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
25515 Generate timestamps by counting samples:
25524 Force color range for the output video frame.
25526 The @code{setrange} filter marks the color range property for the
25527 output frames. It does not change the input frame, but only sets the
25528 corresponding property, which affects how the frame is treated by
25531 The filter accepts the following options:
25536 Available values are:
25540 Keep the same color range property.
25542 @item unspecified, unknown
25543 Set the color range as unspecified.
25545 @item limited, tv, mpeg
25546 Set the color range as limited.
25548 @item full, pc, jpeg
25549 Set the color range as full.
25553 @section settb, asettb
25555 Set the timebase to use for the output frames timestamps.
25556 It is mainly useful for testing timebase configuration.
25558 It accepts the following parameters:
25563 The expression which is evaluated into the output timebase.
25567 The value for @option{tb} is an arithmetic expression representing a
25568 rational. The expression can contain the constants "AVTB" (the default
25569 timebase), "intb" (the input timebase) and "sr" (the sample rate,
25570 audio only). Default value is "intb".
25572 @subsection Examples
25576 Set the timebase to 1/25:
25582 Set the timebase to 1/10:
25588 Set the timebase to 1001/1000:
25594 Set the timebase to 2*intb:
25600 Set the default timebase value:
25607 Convert input audio to a video output representing frequency spectrum
25608 logarithmically using Brown-Puckette constant Q transform algorithm with
25609 direct frequency domain coefficient calculation (but the transform itself
25610 is not really constant Q, instead the Q factor is actually variable/clamped),
25611 with musical tone scale, from E0 to D#10.
25613 The filter accepts the following options:
25617 Specify the video size for the output. It must be even. For the syntax of this option,
25618 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25619 Default value is @code{1920x1080}.
25622 Set the output frame rate. Default value is @code{25}.
25625 Set the bargraph height. It must be even. Default value is @code{-1} which
25626 computes the bargraph height automatically.
25629 Set the axis height. It must be even. Default value is @code{-1} which computes
25630 the axis height automatically.
25633 Set the sonogram height. It must be even. Default value is @code{-1} which
25634 computes the sonogram height automatically.
25637 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
25638 instead. Default value is @code{1}.
25640 @item sono_v, volume
25641 Specify the sonogram volume expression. It can contain variables:
25644 the @var{bar_v} evaluated expression
25645 @item frequency, freq, f
25646 the frequency where it is evaluated
25647 @item timeclamp, tc
25648 the value of @var{timeclamp} option
25652 @item a_weighting(f)
25653 A-weighting of equal loudness
25654 @item b_weighting(f)
25655 B-weighting of equal loudness
25656 @item c_weighting(f)
25657 C-weighting of equal loudness.
25659 Default value is @code{16}.
25661 @item bar_v, volume2
25662 Specify the bargraph volume expression. It can contain variables:
25665 the @var{sono_v} evaluated expression
25666 @item frequency, freq, f
25667 the frequency where it is evaluated
25668 @item timeclamp, tc
25669 the value of @var{timeclamp} option
25673 @item a_weighting(f)
25674 A-weighting of equal loudness
25675 @item b_weighting(f)
25676 B-weighting of equal loudness
25677 @item c_weighting(f)
25678 C-weighting of equal loudness.
25680 Default value is @code{sono_v}.
25682 @item sono_g, gamma
25683 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
25684 higher gamma makes the spectrum having more range. Default value is @code{3}.
25685 Acceptable range is @code{[1, 7]}.
25687 @item bar_g, gamma2
25688 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
25692 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
25693 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
25695 @item timeclamp, tc
25696 Specify the transform timeclamp. At low frequency, there is trade-off between
25697 accuracy in time domain and frequency domain. If timeclamp is lower,
25698 event in time domain is represented more accurately (such as fast bass drum),
25699 otherwise event in frequency domain is represented more accurately
25700 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
25703 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
25704 limits future samples by applying asymmetric windowing in time domain, useful
25705 when low latency is required. Accepted range is @code{[0, 1]}.
25708 Specify the transform base frequency. Default value is @code{20.01523126408007475},
25709 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
25712 Specify the transform end frequency. Default value is @code{20495.59681441799654},
25713 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
25716 This option is deprecated and ignored.
25719 Specify the transform length in time domain. Use this option to control accuracy
25720 trade-off between time domain and frequency domain at every frequency sample.
25721 It can contain variables:
25723 @item frequency, freq, f
25724 the frequency where it is evaluated
25725 @item timeclamp, tc
25726 the value of @var{timeclamp} option.
25728 Default value is @code{384*tc/(384+tc*f)}.
25731 Specify the transform count for every video frame. Default value is @code{6}.
25732 Acceptable range is @code{[1, 30]}.
25735 Specify the transform count for every single pixel. Default value is @code{0},
25736 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
25739 Specify font file for use with freetype to draw the axis. If not specified,
25740 use embedded font. Note that drawing with font file or embedded font is not
25741 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
25745 Specify fontconfig pattern. This has lower priority than @var{fontfile}. The
25746 @code{:} in the pattern may be replaced by @code{|} to avoid unnecessary
25750 Specify font color expression. This is arithmetic expression that should return
25751 integer value 0xRRGGBB. It can contain variables:
25753 @item frequency, freq, f
25754 the frequency where it is evaluated
25755 @item timeclamp, tc
25756 the value of @var{timeclamp} option
25761 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
25762 @item r(x), g(x), b(x)
25763 red, green, and blue value of intensity x.
25765 Default value is @code{st(0, (midi(f)-59.5)/12);
25766 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
25767 r(1-ld(1)) + b(ld(1))}.
25770 Specify image file to draw the axis. This option override @var{fontfile} and
25771 @var{fontcolor} option.
25774 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
25775 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
25776 Default value is @code{1}.
25779 Set colorspace. The accepted values are:
25782 Unspecified (default)
25791 BT.470BG or BT.601-6 625
25794 SMPTE-170M or BT.601-6 525
25800 BT.2020 with non-constant luminance
25805 Set spectrogram color scheme. This is list of floating point values with format
25806 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
25807 The default is @code{1|0.5|0|0|0.5|1}.
25811 @subsection Examples
25815 Playing audio while showing the spectrum:
25817 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
25821 Same as above, but with frame rate 30 fps:
25823 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
25827 Playing at 1280x720:
25829 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
25833 Disable sonogram display:
25839 A1 and its harmonics: A1, A2, (near)E3, A3:
25841 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),
25842 asplit[a][out1]; [a] showcqt [out0]'
25846 Same as above, but with more accuracy in frequency domain:
25848 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),
25849 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
25855 bar_v=10:sono_v=bar_v*a_weighting(f)
25859 Custom gamma, now spectrum is linear to the amplitude.
25865 Custom tlength equation:
25867 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)))'
25871 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
25873 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
25877 Custom font using fontconfig:
25879 font='Courier New,Monospace,mono|bold'
25883 Custom frequency range with custom axis using image file:
25885 axisfile=myaxis.png:basefreq=40:endfreq=10000
25891 Convert input audio to video output representing the audio power spectrum.
25892 Audio amplitude is on Y-axis while frequency is on X-axis.
25894 The filter accepts the following options:
25898 Specify size of video. For the syntax of this option, check the
25899 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
25900 Default is @code{1024x512}.
25904 This set how each frequency bin will be represented.
25906 It accepts the following values:
25912 Default is @code{bar}.
25915 Set amplitude scale.
25917 It accepts the following values:
25931 Default is @code{log}.
25934 Set frequency scale.
25936 It accepts the following values:
25945 Reverse logarithmic scale.
25947 Default is @code{lin}.
25950 Set window size. Allowed range is from 16 to 65536.
25952 Default is @code{2048}
25955 Set windowing function.
25957 It accepts the following values:
25980 Default is @code{hanning}.
25983 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
25984 which means optimal overlap for selected window function will be picked.
25987 Set time averaging. Setting this to 0 will display current maximal peaks.
25988 Default is @code{1}, which means time averaging is disabled.
25991 Specify list of colors separated by space or by '|' which will be used to
25992 draw channel frequencies. Unrecognized or missing colors will be replaced
25996 Set channel display mode.
25998 It accepts the following values:
26003 Default is @code{combined}.
26006 Set minimum amplitude used in @code{log} amplitude scaler.
26009 Set data display mode.
26011 It accepts the following values:
26017 Default is @code{magnitude}.
26020 @section showspatial
26022 Convert stereo input audio to a video output, representing the spatial relationship
26023 between two channels.
26025 The filter accepts the following options:
26029 Specify the video size for the output. For the syntax of this option, check the
26030 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26031 Default value is @code{512x512}.
26034 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
26037 Set window function.
26039 It accepts the following values:
26064 Default value is @code{hann}.
26067 Set ratio of overlap window. Default value is @code{0.5}.
26068 When value is @code{1} overlap is set to recommended size for specific
26069 window function currently used.
26072 @anchor{showspectrum}
26073 @section showspectrum
26075 Convert input audio to a video output, representing the audio frequency
26078 The filter accepts the following options:
26082 Specify the video size for the output. For the syntax of this option, check the
26083 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26084 Default value is @code{640x512}.
26087 Specify how the spectrum should slide along the window.
26089 It accepts the following values:
26092 the samples start again on the left when they reach the right
26094 the samples scroll from right to left
26096 frames are only produced when the samples reach the right
26098 the samples scroll from left to right
26101 Default value is @code{replace}.
26104 Specify display mode.
26106 It accepts the following values:
26109 all channels are displayed in the same row
26111 all channels are displayed in separate rows
26114 Default value is @samp{combined}.
26117 Specify display color mode.
26119 It accepts the following values:
26122 each channel is displayed in a separate color
26124 each channel is displayed using the same color scheme
26126 each channel is displayed using the rainbow color scheme
26128 each channel is displayed using the moreland color scheme
26130 each channel is displayed using the nebulae color scheme
26132 each channel is displayed using the fire color scheme
26134 each channel is displayed using the fiery color scheme
26136 each channel is displayed using the fruit color scheme
26138 each channel is displayed using the cool color scheme
26140 each channel is displayed using the magma color scheme
26142 each channel is displayed using the green color scheme
26144 each channel is displayed using the viridis color scheme
26146 each channel is displayed using the plasma color scheme
26148 each channel is displayed using the cividis color scheme
26150 each channel is displayed using the terrain color scheme
26153 Default value is @samp{channel}.
26156 Specify scale used for calculating intensity color values.
26158 It accepts the following values:
26163 square root, default
26174 Default value is @samp{sqrt}.
26177 Specify frequency scale.
26179 It accepts the following values:
26187 Default value is @samp{lin}.
26190 Set saturation modifier for displayed colors. Negative values provide
26191 alternative color scheme. @code{0} is no saturation at all.
26192 Saturation must be in [-10.0, 10.0] range.
26193 Default value is @code{1}.
26196 Set window function.
26198 It accepts the following values:
26223 Default value is @code{hann}.
26226 Set orientation of time vs frequency axis. Can be @code{vertical} or
26227 @code{horizontal}. Default is @code{vertical}.
26230 Set ratio of overlap window. Default value is @code{0}.
26231 When value is @code{1} overlap is set to recommended size for specific
26232 window function currently used.
26235 Set scale gain for calculating intensity color values.
26236 Default value is @code{1}.
26239 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
26242 Set color rotation, must be in [-1.0, 1.0] range.
26243 Default value is @code{0}.
26246 Set start frequency from which to display spectrogram. Default is @code{0}.
26249 Set stop frequency to which to display spectrogram. Default is @code{0}.
26252 Set upper frame rate limit. Default is @code{auto}, unlimited.
26255 Draw time and frequency axes and legends. Default is disabled.
26258 The usage is very similar to the showwaves filter; see the examples in that
26261 @subsection Examples
26265 Large window with logarithmic color scaling:
26267 showspectrum=s=1280x480:scale=log
26271 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
26273 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
26274 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
26278 @section showspectrumpic
26280 Convert input audio to a single video frame, representing the audio frequency
26283 The filter accepts the following options:
26287 Specify the video size for the output. For the syntax of this option, check the
26288 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26289 Default value is @code{4096x2048}.
26292 Specify display mode.
26294 It accepts the following values:
26297 all channels are displayed in the same row
26299 all channels are displayed in separate rows
26301 Default value is @samp{combined}.
26304 Specify display color mode.
26306 It accepts the following values:
26309 each channel is displayed in a separate color
26311 each channel is displayed using the same color scheme
26313 each channel is displayed using the rainbow color scheme
26315 each channel is displayed using the moreland color scheme
26317 each channel is displayed using the nebulae color scheme
26319 each channel is displayed using the fire color scheme
26321 each channel is displayed using the fiery color scheme
26323 each channel is displayed using the fruit color scheme
26325 each channel is displayed using the cool color scheme
26327 each channel is displayed using the magma color scheme
26329 each channel is displayed using the green color scheme
26331 each channel is displayed using the viridis color scheme
26333 each channel is displayed using the plasma color scheme
26335 each channel is displayed using the cividis color scheme
26337 each channel is displayed using the terrain color scheme
26339 Default value is @samp{intensity}.
26342 Specify scale used for calculating intensity color values.
26344 It accepts the following values:
26349 square root, default
26359 Default value is @samp{log}.
26362 Specify frequency scale.
26364 It accepts the following values:
26372 Default value is @samp{lin}.
26375 Set saturation modifier for displayed colors. Negative values provide
26376 alternative color scheme. @code{0} is no saturation at all.
26377 Saturation must be in [-10.0, 10.0] range.
26378 Default value is @code{1}.
26381 Set window function.
26383 It accepts the following values:
26407 Default value is @code{hann}.
26410 Set orientation of time vs frequency axis. Can be @code{vertical} or
26411 @code{horizontal}. Default is @code{vertical}.
26414 Set scale gain for calculating intensity color values.
26415 Default value is @code{1}.
26418 Draw time and frequency axes and legends. Default is enabled.
26421 Set color rotation, must be in [-1.0, 1.0] range.
26422 Default value is @code{0}.
26425 Set start frequency from which to display spectrogram. Default is @code{0}.
26428 Set stop frequency to which to display spectrogram. Default is @code{0}.
26431 @subsection Examples
26435 Extract an audio spectrogram of a whole audio track
26436 in a 1024x1024 picture using @command{ffmpeg}:
26438 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
26442 @section showvolume
26444 Convert input audio volume to a video output.
26446 The filter accepts the following options:
26453 Set border width, allowed range is [0, 5]. Default is 1.
26456 Set channel width, allowed range is [80, 8192]. Default is 400.
26459 Set channel height, allowed range is [1, 900]. Default is 20.
26462 Set fade, allowed range is [0, 1]. Default is 0.95.
26465 Set volume color expression.
26467 The expression can use the following variables:
26471 Current max volume of channel in dB.
26477 Current channel number, starting from 0.
26481 If set, displays channel names. Default is enabled.
26484 If set, displays volume values. Default is enabled.
26487 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
26488 default is @code{h}.
26491 Set step size, allowed range is [0, 5]. Default is 0, which means
26495 Set background opacity, allowed range is [0, 1]. Default is 0.
26498 Set metering mode, can be peak: @code{p} or rms: @code{r},
26499 default is @code{p}.
26502 Set display scale, can be linear: @code{lin} or log: @code{log},
26503 default is @code{lin}.
26507 If set to > 0., display a line for the max level
26508 in the previous seconds.
26509 default is disabled: @code{0.}
26512 The color of the max line. Use when @code{dm} option is set to > 0.
26513 default is: @code{orange}
26518 Convert input audio to a video output, representing the samples waves.
26520 The filter accepts the following options:
26524 Specify the video size for the output. For the syntax of this option, check the
26525 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26526 Default value is @code{600x240}.
26531 Available values are:
26534 Draw a point for each sample.
26537 Draw a vertical line for each sample.
26540 Draw a point for each sample and a line between them.
26543 Draw a centered vertical line for each sample.
26546 Default value is @code{point}.
26549 Set the number of samples which are printed on the same column. A
26550 larger value will decrease the frame rate. Must be a positive
26551 integer. This option can be set only if the value for @var{rate}
26552 is not explicitly specified.
26555 Set the (approximate) output frame rate. This is done by setting the
26556 option @var{n}. Default value is "25".
26558 @item split_channels
26559 Set if channels should be drawn separately or overlap. Default value is 0.
26562 Set colors separated by '|' which are going to be used for drawing of each channel.
26565 Set amplitude scale.
26567 Available values are:
26585 Set the draw mode. This is mostly useful to set for high @var{n}.
26587 Available values are:
26590 Scale pixel values for each drawn sample.
26593 Draw every sample directly.
26596 Default value is @code{scale}.
26599 @subsection Examples
26603 Output the input file audio and the corresponding video representation
26606 amovie=a.mp3,asplit[out0],showwaves[out1]
26610 Create a synthetic signal and show it with showwaves, forcing a
26611 frame rate of 30 frames per second:
26613 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
26617 @section showwavespic
26619 Convert input audio to a single video frame, representing the samples waves.
26621 The filter accepts the following options:
26625 Specify the video size for the output. For the syntax of this option, check the
26626 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
26627 Default value is @code{600x240}.
26629 @item split_channels
26630 Set if channels should be drawn separately or overlap. Default value is 0.
26633 Set colors separated by '|' which are going to be used for drawing of each channel.
26636 Set amplitude scale.
26638 Available values are:
26658 Available values are:
26661 Scale pixel values for each drawn sample.
26664 Draw every sample directly.
26667 Default value is @code{scale}.
26670 Set the filter mode.
26672 Available values are:
26675 Use average samples values for each drawn sample.
26678 Use peak samples values for each drawn sample.
26681 Default value is @code{average}.
26684 @subsection Examples
26688 Extract a channel split representation of the wave form of a whole audio track
26689 in a 1024x800 picture using @command{ffmpeg}:
26691 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
26695 @section sidedata, asidedata
26697 Delete frame side data, or select frames based on it.
26699 This filter accepts the following options:
26703 Set mode of operation of the filter.
26705 Can be one of the following:
26709 Select every frame with side data of @code{type}.
26712 Delete side data of @code{type}. If @code{type} is not set, delete all side
26718 Set side data type used with all modes. Must be set for @code{select} mode. For
26719 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
26720 in @file{libavutil/frame.h}. For example, to choose
26721 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
26725 @section spectrumsynth
26727 Synthesize audio from 2 input video spectrums, first input stream represents
26728 magnitude across time and second represents phase across time.
26729 The filter will transform from frequency domain as displayed in videos back
26730 to time domain as presented in audio output.
26732 This filter is primarily created for reversing processed @ref{showspectrum}
26733 filter outputs, but can synthesize sound from other spectrograms too.
26734 But in such case results are going to be poor if the phase data is not
26735 available, because in such cases phase data need to be recreated, usually
26736 it's just recreated from random noise.
26737 For best results use gray only output (@code{channel} color mode in
26738 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
26739 @code{lin} scale for phase video. To produce phase, for 2nd video, use
26740 @code{data} option. Inputs videos should generally use @code{fullframe}
26741 slide mode as that saves resources needed for decoding video.
26743 The filter accepts the following options:
26747 Specify sample rate of output audio, the sample rate of audio from which
26748 spectrum was generated may differ.
26751 Set number of channels represented in input video spectrums.
26754 Set scale which was used when generating magnitude input spectrum.
26755 Can be @code{lin} or @code{log}. Default is @code{log}.
26758 Set slide which was used when generating inputs spectrums.
26759 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
26760 Default is @code{fullframe}.
26763 Set window function used for resynthesis.
26766 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
26767 which means optimal overlap for selected window function will be picked.
26770 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
26771 Default is @code{vertical}.
26774 @subsection Examples
26778 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
26779 then resynthesize videos back to audio with spectrumsynth:
26781 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
26782 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
26783 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
26787 @section split, asplit
26789 Split input into several identical outputs.
26791 @code{asplit} works with audio input, @code{split} with video.
26793 The filter accepts a single parameter which specifies the number of outputs. If
26794 unspecified, it defaults to 2.
26796 @subsection Examples
26800 Create two separate outputs from the same input:
26802 [in] split [out0][out1]
26806 To create 3 or more outputs, you need to specify the number of
26809 [in] asplit=3 [out0][out1][out2]
26813 Create two separate outputs from the same input, one cropped and
26816 [in] split [splitout1][splitout2];
26817 [splitout1] crop=100:100:0:0 [cropout];
26818 [splitout2] pad=200:200:100:100 [padout];
26822 Create 5 copies of the input audio with @command{ffmpeg}:
26824 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
26830 Receive commands sent through a libzmq client, and forward them to
26831 filters in the filtergraph.
26833 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
26834 must be inserted between two video filters, @code{azmq} between two
26835 audio filters. Both are capable to send messages to any filter type.
26837 To enable these filters you need to install the libzmq library and
26838 headers and configure FFmpeg with @code{--enable-libzmq}.
26840 For more information about libzmq see:
26841 @url{http://www.zeromq.org/}
26843 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
26844 receives messages sent through a network interface defined by the
26845 @option{bind_address} (or the abbreviation "@option{b}") option.
26846 Default value of this option is @file{tcp://localhost:5555}. You may
26847 want to alter this value to your needs, but do not forget to escape any
26848 ':' signs (see @ref{filtergraph escaping}).
26850 The received message must be in the form:
26852 @var{TARGET} @var{COMMAND} [@var{ARG}]
26855 @var{TARGET} specifies the target of the command, usually the name of
26856 the filter class or a specific filter instance name. The default
26857 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
26858 but you can override this by using the @samp{filter_name@@id} syntax
26859 (see @ref{Filtergraph syntax}).
26861 @var{COMMAND} specifies the name of the command for the target filter.
26863 @var{ARG} is optional and specifies the optional argument list for the
26864 given @var{COMMAND}.
26866 Upon reception, the message is processed and the corresponding command
26867 is injected into the filtergraph. Depending on the result, the filter
26868 will send a reply to the client, adopting the format:
26870 @var{ERROR_CODE} @var{ERROR_REASON}
26874 @var{MESSAGE} is optional.
26876 @subsection Examples
26878 Look at @file{tools/zmqsend} for an example of a zmq client which can
26879 be used to send commands processed by these filters.
26881 Consider the following filtergraph generated by @command{ffplay}.
26882 In this example the last overlay filter has an instance name. All other
26883 filters will have default instance names.
26886 ffplay -dumpgraph 1 -f lavfi "
26887 color=s=100x100:c=red [l];
26888 color=s=100x100:c=blue [r];
26889 nullsrc=s=200x100, zmq [bg];
26890 [bg][l] overlay [bg+l];
26891 [bg+l][r] overlay@@my=x=100 "
26894 To change the color of the left side of the video, the following
26895 command can be used:
26897 echo Parsed_color_0 c yellow | tools/zmqsend
26900 To change the right side:
26902 echo Parsed_color_1 c pink | tools/zmqsend
26905 To change the position of the right side:
26907 echo overlay@@my x 150 | tools/zmqsend
26911 @c man end MULTIMEDIA FILTERS
26913 @chapter Multimedia Sources
26914 @c man begin MULTIMEDIA SOURCES
26916 Below is a description of the currently available multimedia sources.
26920 This is the same as @ref{movie} source, except it selects an audio
26926 Read audio and/or video stream(s) from a movie container.
26928 It accepts the following parameters:
26932 The name of the resource to read (not necessarily a file; it can also be a
26933 device or a stream accessed through some protocol).
26935 @item format_name, f
26936 Specifies the format assumed for the movie to read, and can be either
26937 the name of a container or an input device. If not specified, the
26938 format is guessed from @var{movie_name} or by probing.
26940 @item seek_point, sp
26941 Specifies the seek point in seconds. The frames will be output
26942 starting from this seek point. The parameter is evaluated with
26943 @code{av_strtod}, so the numerical value may be suffixed by an IS
26944 postfix. The default value is "0".
26947 Specifies the streams to read. Several streams can be specified,
26948 separated by "+". The source will then have as many outputs, in the
26949 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
26950 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
26951 respectively the default (best suited) video and audio stream. Default
26952 is "dv", or "da" if the filter is called as "amovie".
26954 @item stream_index, si
26955 Specifies the index of the video stream to read. If the value is -1,
26956 the most suitable video stream will be automatically selected. The default
26957 value is "-1". Deprecated. If the filter is called "amovie", it will select
26958 audio instead of video.
26961 Specifies how many times to read the stream in sequence.
26962 If the value is 0, the stream will be looped infinitely.
26963 Default value is "1".
26965 Note that when the movie is looped the source timestamps are not
26966 changed, so it will generate non monotonically increasing timestamps.
26968 @item discontinuity
26969 Specifies the time difference between frames above which the point is
26970 considered a timestamp discontinuity which is removed by adjusting the later
26974 It allows overlaying a second video on top of the main input of
26975 a filtergraph, as shown in this graph:
26977 input -----------> deltapts0 --> overlay --> output
26980 movie --> scale--> deltapts1 -------+
26982 @subsection Examples
26986 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
26987 on top of the input labelled "in":
26989 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
26990 [in] setpts=PTS-STARTPTS [main];
26991 [main][over] overlay=16:16 [out]
26995 Read from a video4linux2 device, and overlay it on top of the input
26998 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
26999 [in] setpts=PTS-STARTPTS [main];
27000 [main][over] overlay=16:16 [out]
27004 Read the first video stream and the audio stream with id 0x81 from
27005 dvd.vob; the video is connected to the pad named "video" and the audio is
27006 connected to the pad named "audio":
27008 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
27012 @subsection Commands
27014 Both movie and amovie support the following commands:
27017 Perform seek using "av_seek_frame".
27018 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
27021 @var{stream_index}: If stream_index is -1, a default
27022 stream is selected, and @var{timestamp} is automatically converted
27023 from AV_TIME_BASE units to the stream specific time_base.
27025 @var{timestamp}: Timestamp in AVStream.time_base units
27026 or, if no stream is specified, in AV_TIME_BASE units.
27028 @var{flags}: Flags which select direction and seeking mode.
27032 Get movie duration in AV_TIME_BASE units.
27036 @c man end MULTIMEDIA SOURCES