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 Options for filters with several inputs (framesync)
317 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
319 Some filters with several inputs support a common set of options.
320 These options can only be set by name, not with the short notation.
324 The action to take when EOF is encountered on the secondary input; it accepts
325 one of the following values:
329 Repeat the last frame (the default).
333 Pass the main input through.
337 If set to 1, force the output to terminate when the shortest input
338 terminates. Default value is 0.
341 If set to 1, force the filter to extend the last frame of secondary streams
342 until the end of the primary stream. A value of 0 disables this behavior.
346 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
348 @chapter Audio Filters
349 @c man begin AUDIO FILTERS
351 When you configure your FFmpeg build, you can disable any of the
352 existing filters using @code{--disable-filters}.
353 The configure output will show the audio filters included in your
356 Below is a description of the currently available audio filters.
360 A compressor is mainly used to reduce the dynamic range of a signal.
361 Especially modern music is mostly compressed at a high ratio to
362 improve the overall loudness. It's done to get the highest attention
363 of a listener, "fatten" the sound and bring more "power" to the track.
364 If a signal is compressed too much it may sound dull or "dead"
365 afterwards or it may start to "pump" (which could be a powerful effect
366 but can also destroy a track completely).
367 The right compression is the key to reach a professional sound and is
368 the high art of mixing and mastering. Because of its complex settings
369 it may take a long time to get the right feeling for this kind of effect.
371 Compression is done by detecting the volume above a chosen level
372 @code{threshold} and dividing it by the factor set with @code{ratio}.
373 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
374 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
375 the signal would cause distortion of the waveform the reduction can be
376 levelled over the time. This is done by setting "Attack" and "Release".
377 @code{attack} determines how long the signal has to rise above the threshold
378 before any reduction will occur and @code{release} sets the time the signal
379 has to fall below the threshold to reduce the reduction again. Shorter signals
380 than the chosen attack time will be left untouched.
381 The overall reduction of the signal can be made up afterwards with the
382 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
383 raising the makeup to this level results in a signal twice as loud than the
384 source. To gain a softer entry in the compression the @code{knee} flattens the
385 hard edge at the threshold in the range of the chosen decibels.
387 The filter accepts the following options:
391 Set input gain. Default is 1. Range is between 0.015625 and 64.
394 If a signal of stream rises above this level it will affect the gain
396 By default it is 0.125. Range is between 0.00097563 and 1.
399 Set a ratio by which the signal is reduced. 1:2 means that if the level
400 rose 4dB above the threshold, it will be only 2dB above after the reduction.
401 Default is 2. Range is between 1 and 20.
404 Amount of milliseconds the signal has to rise above the threshold before gain
405 reduction starts. Default is 20. Range is between 0.01 and 2000.
408 Amount of milliseconds the signal has to fall below the threshold before
409 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
412 Set the amount by how much signal will be amplified after processing.
413 Default is 1. Range is from 1 to 64.
416 Curve the sharp knee around the threshold to enter gain reduction more softly.
417 Default is 2.82843. Range is between 1 and 8.
420 Choose if the @code{average} level between all channels of input stream
421 or the louder(@code{maximum}) channel of input stream affects the
422 reduction. Default is @code{average}.
425 Should the exact signal be taken in case of @code{peak} or an RMS one in case
426 of @code{rms}. Default is @code{rms} which is mostly smoother.
429 How much to use compressed signal in output. Default is 1.
430 Range is between 0 and 1.
434 Simple audio dynamic range commpression/expansion filter.
436 The filter accepts the following options:
440 Set contrast. Default is 33. Allowed range is between 0 and 100.
445 Copy the input audio source unchanged to the output. This is mainly useful for
450 Apply cross fade from one input audio stream to another input audio stream.
451 The cross fade is applied for specified duration near the end of first stream.
453 The filter accepts the following options:
457 Specify the number of samples for which the cross fade effect has to last.
458 At the end of the cross fade effect the first input audio will be completely
459 silent. Default is 44100.
462 Specify the duration of the cross fade effect. See
463 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
464 for the accepted syntax.
465 By default the duration is determined by @var{nb_samples}.
466 If set this option is used instead of @var{nb_samples}.
469 Should first stream end overlap with second stream start. Default is enabled.
472 Set curve for cross fade transition for first stream.
475 Set curve for cross fade transition for second stream.
477 For description of available curve types see @ref{afade} filter description.
484 Cross fade from one input to another:
486 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
490 Cross fade from one input to another but without overlapping:
492 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
498 Reduce audio bit resolution.
500 This filter is bit crusher with enhanced functionality. A bit crusher
501 is used to audibly reduce number of bits an audio signal is sampled
502 with. This doesn't change the bit depth at all, it just produces the
503 effect. Material reduced in bit depth sounds more harsh and "digital".
504 This filter is able to even round to continuous values instead of discrete
506 Additionally it has a D/C offset which results in different crushing of
507 the lower and the upper half of the signal.
508 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
510 Another feature of this filter is the logarithmic mode.
511 This setting switches from linear distances between bits to logarithmic ones.
512 The result is a much more "natural" sounding crusher which doesn't gate low
513 signals for example. The human ear has a logarithmic perception,
514 so this kind of crushing is much more pleasant.
515 Logarithmic crushing is also able to get anti-aliased.
517 The filter accepts the following options:
533 Can be linear: @code{lin} or logarithmic: @code{log}.
542 Set sample reduction.
545 Enable LFO. By default disabled.
556 Delay one or more audio channels.
558 Samples in delayed channel are filled with silence.
560 The filter accepts the following option:
564 Set list of delays in milliseconds for each channel separated by '|'.
565 Unused delays will be silently ignored. If number of given delays is
566 smaller than number of channels all remaining channels will not be delayed.
567 If you want to delay exact number of samples, append 'S' to number.
574 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
575 the second channel (and any other channels that may be present) unchanged.
581 Delay second channel by 500 samples, the third channel by 700 samples and leave
582 the first channel (and any other channels that may be present) unchanged.
590 Apply echoing to the input audio.
592 Echoes are reflected sound and can occur naturally amongst mountains
593 (and sometimes large buildings) when talking or shouting; digital echo
594 effects emulate this behaviour and are often used to help fill out the
595 sound of a single instrument or vocal. The time difference between the
596 original signal and the reflection is the @code{delay}, and the
597 loudness of the reflected signal is the @code{decay}.
598 Multiple echoes can have different delays and decays.
600 A description of the accepted parameters follows.
604 Set input gain of reflected signal. Default is @code{0.6}.
607 Set output gain of reflected signal. Default is @code{0.3}.
610 Set list of time intervals in milliseconds between original signal and reflections
611 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
612 Default is @code{1000}.
615 Set list of loudness of reflected signals separated by '|'.
616 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
617 Default is @code{0.5}.
624 Make it sound as if there are twice as many instruments as are actually playing:
626 aecho=0.8:0.88:60:0.4
630 If delay is very short, then it sound like a (metallic) robot playing music:
636 A longer delay will sound like an open air concert in the mountains:
638 aecho=0.8:0.9:1000:0.3
642 Same as above but with one more mountain:
644 aecho=0.8:0.9:1000|1800:0.3|0.25
649 Audio emphasis filter creates or restores material directly taken from LPs or
650 emphased CDs with different filter curves. E.g. to store music on vinyl the
651 signal has to be altered by a filter first to even out the disadvantages of
652 this recording medium.
653 Once the material is played back the inverse filter has to be applied to
654 restore the distortion of the frequency response.
656 The filter accepts the following options:
666 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
667 use @code{production} mode. Default is @code{reproduction} mode.
670 Set filter type. Selects medium. Can be one of the following:
682 select Compact Disc (CD).
688 select 50µs (FM-KF).
690 select 75µs (FM-KF).
696 Modify an audio signal according to the specified expressions.
698 This filter accepts one or more expressions (one for each channel),
699 which are evaluated and used to modify a corresponding audio signal.
701 It accepts the following parameters:
705 Set the '|'-separated expressions list for each separate channel. If
706 the number of input channels is greater than the number of
707 expressions, the last specified expression is used for the remaining
710 @item channel_layout, c
711 Set output channel layout. If not specified, the channel layout is
712 specified by the number of expressions. If set to @samp{same}, it will
713 use by default the same input channel layout.
716 Each expression in @var{exprs} can contain the following constants and functions:
720 channel number of the current expression
723 number of the evaluated sample, starting from 0
729 time of the evaluated sample expressed in seconds
732 @item nb_out_channels
733 input and output number of channels
736 the value of input channel with number @var{CH}
739 Note: this filter is slow. For faster processing you should use a
748 aeval=val(ch)/2:c=same
752 Invert phase of the second channel:
761 Apply fade-in/out effect to input audio.
763 A description of the accepted parameters follows.
767 Specify the effect type, can be either @code{in} for fade-in, or
768 @code{out} for a fade-out effect. Default is @code{in}.
770 @item start_sample, ss
771 Specify the number of the start sample for starting to apply the fade
772 effect. Default is 0.
775 Specify the number of samples for which the fade effect has to last. At
776 the end of the fade-in effect the output audio will have the same
777 volume as the input audio, at the end of the fade-out transition
778 the output audio will be silence. Default is 44100.
781 Specify the start time of the fade effect. Default is 0.
782 The value must be specified as a time duration; see
783 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
784 for the accepted syntax.
785 If set this option is used instead of @var{start_sample}.
788 Specify the duration of the fade effect. See
789 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
790 for the accepted syntax.
791 At the end of the fade-in effect the output audio will have the same
792 volume as the input audio, at the end of the fade-out transition
793 the output audio will be silence.
794 By default the duration is determined by @var{nb_samples}.
795 If set this option is used instead of @var{nb_samples}.
798 Set curve for fade transition.
800 It accepts the following values:
803 select triangular, linear slope (default)
805 select quarter of sine wave
807 select half of sine wave
809 select exponential sine wave
813 select inverted parabola
827 select inverted quarter of sine wave
829 select inverted half of sine wave
831 select double-exponential seat
833 select double-exponential sigmoid
841 Fade in first 15 seconds of audio:
847 Fade out last 25 seconds of a 900 seconds audio:
849 afade=t=out:st=875:d=25
854 Apply arbitrary expressions to samples in frequency domain.
858 Set frequency domain real expression for each separate channel separated
859 by '|'. Default is "1".
860 If the number of input channels is greater than the number of
861 expressions, the last specified expression is used for the remaining
865 Set frequency domain imaginary expression for each separate channel
866 separated by '|'. If not set, @var{real} option is used.
868 Each expression in @var{real} and @var{imag} can contain the following
876 current frequency bin number
879 number of available bins
882 channel number of the current expression
894 It accepts the following values:
910 Default is @code{w4096}
913 Set window function. Default is @code{hann}.
916 Set window overlap. If set to 1, the recommended overlap for selected
917 window function will be picked. Default is @code{0.75}.
924 Leave almost only low frequencies in audio:
926 afftfilt="1-clip((b/nb)*b,0,1)"
933 Apply an arbitrary Frequency Impulse Response filter.
935 This filter is designed for applying long FIR filters,
936 up to 30 seconds long.
938 It can be used as component for digital crossover filters,
939 room equalization, cross talk cancellation, wavefield synthesis,
940 auralization, ambiophonics and ambisonics.
942 This filter uses second stream as FIR coefficients.
943 If second stream holds single channel, it will be used
944 for all input channels in first stream, otherwise
945 number of channels in second stream must be same as
946 number of channels in first stream.
948 It accepts the following parameters:
952 Set dry gain. This sets input gain.
955 Set wet gain. This sets final output gain.
958 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
961 Enable applying gain measured from power of IR.
968 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
970 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
977 Set output format constraints for the input audio. The framework will
978 negotiate the most appropriate format to minimize conversions.
980 It accepts the following parameters:
984 A '|'-separated list of requested sample formats.
987 A '|'-separated list of requested sample rates.
989 @item channel_layouts
990 A '|'-separated list of requested channel layouts.
992 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
993 for the required syntax.
996 If a parameter is omitted, all values are allowed.
998 Force the output to either unsigned 8-bit or signed 16-bit stereo
1000 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1005 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1006 processing reduces disturbing noise between useful signals.
1008 Gating is done by detecting the volume below a chosen level @var{threshold}
1009 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1010 floor is set via @var{range}. Because an exact manipulation of the signal
1011 would cause distortion of the waveform the reduction can be levelled over
1012 time. This is done by setting @var{attack} and @var{release}.
1014 @var{attack} determines how long the signal has to fall below the threshold
1015 before any reduction will occur and @var{release} sets the time the signal
1016 has to rise above the threshold to reduce the reduction again.
1017 Shorter signals than the chosen attack time will be left untouched.
1021 Set input level before filtering.
1022 Default is 1. Allowed range is from 0.015625 to 64.
1025 Set the level of gain reduction when the signal is below the threshold.
1026 Default is 0.06125. Allowed range is from 0 to 1.
1029 If a signal rises above this level the gain reduction is released.
1030 Default is 0.125. Allowed range is from 0 to 1.
1033 Set a ratio by which the signal is reduced.
1034 Default is 2. Allowed range is from 1 to 9000.
1037 Amount of milliseconds the signal has to rise above the threshold before gain
1039 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1042 Amount of milliseconds the signal has to fall below the threshold before the
1043 reduction is increased again. Default is 250 milliseconds.
1044 Allowed range is from 0.01 to 9000.
1047 Set amount of amplification of signal after processing.
1048 Default is 1. Allowed range is from 1 to 64.
1051 Curve the sharp knee around the threshold to enter gain reduction more softly.
1052 Default is 2.828427125. Allowed range is from 1 to 8.
1055 Choose if exact signal should be taken for detection or an RMS like one.
1056 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1059 Choose if the average level between all channels or the louder channel affects
1061 Default is @code{average}. Can be @code{average} or @code{maximum}.
1066 Apply an arbitrary Infinite Impulse Response filter.
1068 It accepts the following parameters:
1072 Set numerator/zeros coefficients.
1075 Set denominator/poles coefficients.
1087 Set coefficients format.
1093 Z-plane zeros/poles, cartesian (default)
1095 Z-plane zeros/poles, polar radians
1097 Z-plane zeros/poles, polar degrees
1101 Set kind of processing.
1102 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1105 Set filtering precision.
1109 double-precision floating-point (default)
1111 single-precision floating-point
1120 Coefficients in @code{tf} format are separated by spaces and are in ascending
1123 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1124 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1127 Different coefficients and gains can be provided for every channel, in such case
1128 use '|' to separate coefficients or gains. Last provided coefficients will be
1129 used for all remaining channels.
1131 @subsection Examples
1135 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1137 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
1141 Same as above but in @code{zp} format:
1143 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
1149 The limiter prevents an input signal from rising over a desired threshold.
1150 This limiter uses lookahead technology to prevent your signal from distorting.
1151 It means that there is a small delay after the signal is processed. Keep in mind
1152 that the delay it produces is the attack time you set.
1154 The filter accepts the following options:
1158 Set input gain. Default is 1.
1161 Set output gain. Default is 1.
1164 Don't let signals above this level pass the limiter. Default is 1.
1167 The limiter will reach its attenuation level in this amount of time in
1168 milliseconds. Default is 5 milliseconds.
1171 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1172 Default is 50 milliseconds.
1175 When gain reduction is always needed ASC takes care of releasing to an
1176 average reduction level rather than reaching a reduction of 0 in the release
1180 Select how much the release time is affected by ASC, 0 means nearly no changes
1181 in release time while 1 produces higher release times.
1184 Auto level output signal. Default is enabled.
1185 This normalizes audio back to 0dB if enabled.
1188 Depending on picked setting it is recommended to upsample input 2x or 4x times
1189 with @ref{aresample} before applying this filter.
1193 Apply a two-pole all-pass filter with central frequency (in Hz)
1194 @var{frequency}, and filter-width @var{width}.
1195 An all-pass filter changes the audio's frequency to phase relationship
1196 without changing its frequency to amplitude relationship.
1198 The filter accepts the following options:
1202 Set frequency in Hz.
1205 Set method to specify band-width of filter.
1220 Specify the band-width of a filter in width_type units.
1223 Specify which channels to filter, by default all available are filtered.
1226 @subsection Commands
1228 This filter supports the following commands:
1231 Change allpass frequency.
1232 Syntax for the command is : "@var{frequency}"
1235 Change allpass width_type.
1236 Syntax for the command is : "@var{width_type}"
1239 Change allpass width.
1240 Syntax for the command is : "@var{width}"
1247 The filter accepts the following options:
1251 Set the number of loops. Setting this value to -1 will result in infinite loops.
1255 Set maximal number of samples. Default is 0.
1258 Set first sample of loop. Default is 0.
1264 Merge two or more audio streams into a single multi-channel stream.
1266 The filter accepts the following options:
1271 Set the number of inputs. Default is 2.
1275 If the channel layouts of the inputs are disjoint, and therefore compatible,
1276 the channel layout of the output will be set accordingly and the channels
1277 will be reordered as necessary. If the channel layouts of the inputs are not
1278 disjoint, the output will have all the channels of the first input then all
1279 the channels of the second input, in that order, and the channel layout of
1280 the output will be the default value corresponding to the total number of
1283 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1284 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1285 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1286 first input, b1 is the first channel of the second input).
1288 On the other hand, if both input are in stereo, the output channels will be
1289 in the default order: a1, a2, b1, b2, and the channel layout will be
1290 arbitrarily set to 4.0, which may or may not be the expected value.
1292 All inputs must have the same sample rate, and format.
1294 If inputs do not have the same duration, the output will stop with the
1297 @subsection Examples
1301 Merge two mono files into a stereo stream:
1303 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1307 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1309 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
1315 Mixes multiple audio inputs into a single output.
1317 Note that this filter only supports float samples (the @var{amerge}
1318 and @var{pan} audio filters support many formats). If the @var{amix}
1319 input has integer samples then @ref{aresample} will be automatically
1320 inserted to perform the conversion to float samples.
1324 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1326 will mix 3 input audio streams to a single output with the same duration as the
1327 first input and a dropout transition time of 3 seconds.
1329 It accepts the following parameters:
1333 The number of inputs. If unspecified, it defaults to 2.
1336 How to determine the end-of-stream.
1340 The duration of the longest input. (default)
1343 The duration of the shortest input.
1346 The duration of the first input.
1350 @item dropout_transition
1351 The transition time, in seconds, for volume renormalization when an input
1352 stream ends. The default value is 2 seconds.
1355 Specify weight of each input audio stream as sequence.
1356 Each weight is separated by space. By default all inputs have same weight.
1359 @section anequalizer
1361 High-order parametric multiband equalizer for each channel.
1363 It accepts the following parameters:
1367 This option string is in format:
1368 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1369 Each equalizer band is separated by '|'.
1373 Set channel number to which equalization will be applied.
1374 If input doesn't have that channel the entry is ignored.
1377 Set central frequency for band.
1378 If input doesn't have that frequency the entry is ignored.
1381 Set band width in hertz.
1384 Set band gain in dB.
1387 Set filter type for band, optional, can be:
1391 Butterworth, this is default.
1402 With this option activated frequency response of anequalizer is displayed
1406 Set video stream size. Only useful if curves option is activated.
1409 Set max gain that will be displayed. Only useful if curves option is activated.
1410 Setting this to a reasonable value makes it possible to display gain which is derived from
1411 neighbour bands which are too close to each other and thus produce higher gain
1412 when both are activated.
1415 Set frequency scale used to draw frequency response in video output.
1416 Can be linear or logarithmic. Default is logarithmic.
1419 Set color for each channel curve which is going to be displayed in video stream.
1420 This is list of color names separated by space or by '|'.
1421 Unrecognised or missing colors will be replaced by white color.
1424 @subsection Examples
1428 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1429 for first 2 channels using Chebyshev type 1 filter:
1431 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1435 @subsection Commands
1437 This filter supports the following commands:
1440 Alter existing filter parameters.
1441 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1443 @var{fN} is existing filter number, starting from 0, if no such filter is available
1445 @var{freq} set new frequency parameter.
1446 @var{width} set new width parameter in herz.
1447 @var{gain} set new gain parameter in dB.
1449 Full filter invocation with asendcmd may look like this:
1450 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1455 Pass the audio source unchanged to the output.
1459 Pad the end of an audio stream with silence.
1461 This can be used together with @command{ffmpeg} @option{-shortest} to
1462 extend audio streams to the same length as the video stream.
1464 A description of the accepted options follows.
1468 Set silence packet size. Default value is 4096.
1471 Set the number of samples of silence to add to the end. After the
1472 value is reached, the stream is terminated. This option is mutually
1473 exclusive with @option{whole_len}.
1476 Set the minimum total number of samples in the output audio stream. If
1477 the value is longer than the input audio length, silence is added to
1478 the end, until the value is reached. This option is mutually exclusive
1479 with @option{pad_len}.
1482 If neither the @option{pad_len} nor the @option{whole_len} option is
1483 set, the filter will add silence to the end of the input stream
1486 @subsection Examples
1490 Add 1024 samples of silence to the end of the input:
1496 Make sure the audio output will contain at least 10000 samples, pad
1497 the input with silence if required:
1499 apad=whole_len=10000
1503 Use @command{ffmpeg} to pad the audio input with silence, so that the
1504 video stream will always result the shortest and will be converted
1505 until the end in the output file when using the @option{shortest}
1508 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1513 Add a phasing effect to the input audio.
1515 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1516 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1518 A description of the accepted parameters follows.
1522 Set input gain. Default is 0.4.
1525 Set output gain. Default is 0.74
1528 Set delay in milliseconds. Default is 3.0.
1531 Set decay. Default is 0.4.
1534 Set modulation speed in Hz. Default is 0.5.
1537 Set modulation type. Default is triangular.
1539 It accepts the following values:
1548 Audio pulsator is something between an autopanner and a tremolo.
1549 But it can produce funny stereo effects as well. Pulsator changes the volume
1550 of the left and right channel based on a LFO (low frequency oscillator) with
1551 different waveforms and shifted phases.
1552 This filter have the ability to define an offset between left and right
1553 channel. An offset of 0 means that both LFO shapes match each other.
1554 The left and right channel are altered equally - a conventional tremolo.
1555 An offset of 50% means that the shape of the right channel is exactly shifted
1556 in phase (or moved backwards about half of the frequency) - pulsator acts as
1557 an autopanner. At 1 both curves match again. Every setting in between moves the
1558 phase shift gapless between all stages and produces some "bypassing" sounds with
1559 sine and triangle waveforms. The more you set the offset near 1 (starting from
1560 the 0.5) the faster the signal passes from the left to the right speaker.
1562 The filter accepts the following options:
1566 Set input gain. By default it is 1. Range is [0.015625 - 64].
1569 Set output gain. By default it is 1. Range is [0.015625 - 64].
1572 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1573 sawup or sawdown. Default is sine.
1576 Set modulation. Define how much of original signal is affected by the LFO.
1579 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1582 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1585 Set pulse width. Default is 1. Allowed range is [0 - 2].
1588 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1591 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1595 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1599 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1600 if timing is set to hz.
1606 Resample the input audio to the specified parameters, using the
1607 libswresample library. If none are specified then the filter will
1608 automatically convert between its input and output.
1610 This filter is also able to stretch/squeeze the audio data to make it match
1611 the timestamps or to inject silence / cut out audio to make it match the
1612 timestamps, do a combination of both or do neither.
1614 The filter accepts the syntax
1615 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1616 expresses a sample rate and @var{resampler_options} is a list of
1617 @var{key}=@var{value} pairs, separated by ":". See the
1618 @ref{Resampler Options,,"Resampler Options" section in the
1619 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1620 for the complete list of supported options.
1622 @subsection Examples
1626 Resample the input audio to 44100Hz:
1632 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1633 samples per second compensation:
1635 aresample=async=1000
1641 Reverse an audio clip.
1643 Warning: This filter requires memory to buffer the entire clip, so trimming
1646 @subsection Examples
1650 Take the first 5 seconds of a clip, and reverse it.
1652 atrim=end=5,areverse
1656 @section asetnsamples
1658 Set the number of samples per each output audio frame.
1660 The last output packet may contain a different number of samples, as
1661 the filter will flush all the remaining samples when the input audio
1664 The filter accepts the following options:
1668 @item nb_out_samples, n
1669 Set the number of frames per each output audio frame. The number is
1670 intended as the number of samples @emph{per each channel}.
1671 Default value is 1024.
1674 If set to 1, the filter will pad the last audio frame with zeroes, so
1675 that the last frame will contain the same number of samples as the
1676 previous ones. Default value is 1.
1679 For example, to set the number of per-frame samples to 1234 and
1680 disable padding for the last frame, use:
1682 asetnsamples=n=1234:p=0
1687 Set the sample rate without altering the PCM data.
1688 This will result in a change of speed and pitch.
1690 The filter accepts the following options:
1693 @item sample_rate, r
1694 Set the output sample rate. Default is 44100 Hz.
1699 Show a line containing various information for each input audio frame.
1700 The input audio is not modified.
1702 The shown line contains a sequence of key/value pairs of the form
1703 @var{key}:@var{value}.
1705 The following values are shown in the output:
1709 The (sequential) number of the input frame, starting from 0.
1712 The presentation timestamp of the input frame, in time base units; the time base
1713 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1716 The presentation timestamp of the input frame in seconds.
1719 position of the frame in the input stream, -1 if this information in
1720 unavailable and/or meaningless (for example in case of synthetic audio)
1729 The sample rate for the audio frame.
1732 The number of samples (per channel) in the frame.
1735 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1736 audio, the data is treated as if all the planes were concatenated.
1738 @item plane_checksums
1739 A list of Adler-32 checksums for each data plane.
1745 Display time domain statistical information about the audio channels.
1746 Statistics are calculated and displayed for each audio channel and,
1747 where applicable, an overall figure is also given.
1749 It accepts the following option:
1752 Short window length in seconds, used for peak and trough RMS measurement.
1753 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
1757 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1758 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1761 Available keys for each channel are:
1795 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1796 this @code{lavfi.astats.Overall.Peak_count}.
1798 For description what each key means read below.
1801 Set number of frame after which stats are going to be recalculated.
1802 Default is disabled.
1805 A description of each shown parameter follows:
1809 Mean amplitude displacement from zero.
1812 Minimal sample level.
1815 Maximal sample level.
1817 @item Min difference
1818 Minimal difference between two consecutive samples.
1820 @item Max difference
1821 Maximal difference between two consecutive samples.
1823 @item Mean difference
1824 Mean difference between two consecutive samples.
1825 The average of each difference between two consecutive samples.
1827 @item RMS difference
1828 Root Mean Square difference between two consecutive samples.
1832 Standard peak and RMS level measured in dBFS.
1836 Peak and trough values for RMS level measured over a short window.
1839 Standard ratio of peak to RMS level (note: not in dB).
1842 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1843 (i.e. either @var{Min level} or @var{Max level}).
1846 Number of occasions (not the number of samples) that the signal attained either
1847 @var{Min level} or @var{Max level}.
1850 Overall bit depth of audio. Number of bits used for each sample.
1853 Measured dynamic range of audio in dB.
1860 The filter accepts exactly one parameter, the audio tempo. If not
1861 specified then the filter will assume nominal 1.0 tempo. Tempo must
1862 be in the [0.5, 2.0] range.
1864 @subsection Examples
1868 Slow down audio to 80% tempo:
1874 To speed up audio to 125% tempo:
1882 Trim the input so that the output contains one continuous subpart of the input.
1884 It accepts the following parameters:
1887 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1888 sample with the timestamp @var{start} will be the first sample in the output.
1891 Specify time of the first audio sample that will be dropped, i.e. the
1892 audio sample immediately preceding the one with the timestamp @var{end} will be
1893 the last sample in the output.
1896 Same as @var{start}, except this option sets the start timestamp in samples
1900 Same as @var{end}, except this option sets the end timestamp in samples instead
1904 The maximum duration of the output in seconds.
1907 The number of the first sample that should be output.
1910 The number of the first sample that should be dropped.
1913 @option{start}, @option{end}, and @option{duration} are expressed as time
1914 duration specifications; see
1915 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1917 Note that the first two sets of the start/end options and the @option{duration}
1918 option look at the frame timestamp, while the _sample options simply count the
1919 samples that pass through the filter. So start/end_pts and start/end_sample will
1920 give different results when the timestamps are wrong, inexact or do not start at
1921 zero. Also note that this filter does not modify the timestamps. If you wish
1922 to have the output timestamps start at zero, insert the asetpts filter after the
1925 If multiple start or end options are set, this filter tries to be greedy and
1926 keep all samples that match at least one of the specified constraints. To keep
1927 only the part that matches all the constraints at once, chain multiple atrim
1930 The defaults are such that all the input is kept. So it is possible to set e.g.
1931 just the end values to keep everything before the specified time.
1936 Drop everything except the second minute of input:
1938 ffmpeg -i INPUT -af atrim=60:120
1942 Keep only the first 1000 samples:
1944 ffmpeg -i INPUT -af atrim=end_sample=1000
1951 Apply a two-pole Butterworth band-pass filter with central
1952 frequency @var{frequency}, and (3dB-point) band-width width.
1953 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1954 instead of the default: constant 0dB peak gain.
1955 The filter roll off at 6dB per octave (20dB per decade).
1957 The filter accepts the following options:
1961 Set the filter's central frequency. Default is @code{3000}.
1964 Constant skirt gain if set to 1. Defaults to 0.
1967 Set method to specify band-width of filter.
1982 Specify the band-width of a filter in width_type units.
1985 Specify which channels to filter, by default all available are filtered.
1988 @subsection Commands
1990 This filter supports the following commands:
1993 Change bandpass frequency.
1994 Syntax for the command is : "@var{frequency}"
1997 Change bandpass width_type.
1998 Syntax for the command is : "@var{width_type}"
2001 Change bandpass width.
2002 Syntax for the command is : "@var{width}"
2007 Apply a two-pole Butterworth band-reject filter with central
2008 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2009 The filter roll off at 6dB per octave (20dB per decade).
2011 The filter accepts the following options:
2015 Set the filter's central frequency. Default is @code{3000}.
2018 Set method to specify band-width of filter.
2033 Specify the band-width of a filter in width_type units.
2036 Specify which channels to filter, by default all available are filtered.
2039 @subsection Commands
2041 This filter supports the following commands:
2044 Change bandreject frequency.
2045 Syntax for the command is : "@var{frequency}"
2048 Change bandreject width_type.
2049 Syntax for the command is : "@var{width_type}"
2052 Change bandreject width.
2053 Syntax for the command is : "@var{width}"
2058 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2059 shelving filter with a response similar to that of a standard
2060 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2062 The filter accepts the following options:
2066 Give the gain at 0 Hz. Its useful range is about -20
2067 (for a large cut) to +20 (for a large boost).
2068 Beware of clipping when using a positive gain.
2071 Set the filter's central frequency and so can be used
2072 to extend or reduce the frequency range to be boosted or cut.
2073 The default value is @code{100} Hz.
2076 Set method to specify band-width of filter.
2091 Determine how steep is the filter's shelf transition.
2094 Specify which channels to filter, by default all available are filtered.
2097 @subsection Commands
2099 This filter supports the following commands:
2102 Change bass frequency.
2103 Syntax for the command is : "@var{frequency}"
2106 Change bass width_type.
2107 Syntax for the command is : "@var{width_type}"
2111 Syntax for the command is : "@var{width}"
2115 Syntax for the command is : "@var{gain}"
2120 Apply a biquad IIR filter with the given coefficients.
2121 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2122 are the numerator and denominator coefficients respectively.
2123 and @var{channels}, @var{c} specify which channels to filter, by default all
2124 available are filtered.
2126 @subsection Commands
2128 This filter supports the following commands:
2136 Change biquad parameter.
2137 Syntax for the command is : "@var{value}"
2141 Bauer stereo to binaural transformation, which improves headphone listening of
2142 stereo audio records.
2144 To enable compilation of this filter you need to configure FFmpeg with
2145 @code{--enable-libbs2b}.
2147 It accepts the following parameters:
2151 Pre-defined crossfeed level.
2155 Default level (fcut=700, feed=50).
2158 Chu Moy circuit (fcut=700, feed=60).
2161 Jan Meier circuit (fcut=650, feed=95).
2166 Cut frequency (in Hz).
2175 Remap input channels to new locations.
2177 It accepts the following parameters:
2180 Map channels from input to output. The argument is a '|'-separated list of
2181 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2182 @var{in_channel} form. @var{in_channel} can be either the name of the input
2183 channel (e.g. FL for front left) or its index in the input channel layout.
2184 @var{out_channel} is the name of the output channel or its index in the output
2185 channel layout. If @var{out_channel} is not given then it is implicitly an
2186 index, starting with zero and increasing by one for each mapping.
2188 @item channel_layout
2189 The channel layout of the output stream.
2192 If no mapping is present, the filter will implicitly map input channels to
2193 output channels, preserving indices.
2195 @subsection Examples
2199 For example, assuming a 5.1+downmix input MOV file,
2201 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2203 will create an output WAV file tagged as stereo from the downmix channels of
2207 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2209 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2213 @section channelsplit
2215 Split each channel from an input audio stream into a separate output stream.
2217 It accepts the following parameters:
2219 @item channel_layout
2220 The channel layout of the input stream. The default is "stereo".
2222 A channel layout describing the channels to be extracted as separate output streams
2223 or "all" to extract each input channel as a separate stream. The default is "all".
2225 Choosing channels not present in channel layout in the input will result in an error.
2228 @subsection Examples
2232 For example, assuming a stereo input MP3 file,
2234 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2236 will create an output Matroska file with two audio streams, one containing only
2237 the left channel and the other the right channel.
2240 Split a 5.1 WAV file into per-channel files:
2242 ffmpeg -i in.wav -filter_complex
2243 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2244 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2245 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2250 Extract only LFE from a 5.1 WAV file:
2252 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2253 -map '[LFE]' lfe.wav
2258 Add a chorus effect to the audio.
2260 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2262 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2263 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2264 The modulation depth defines the range the modulated delay is played before or after
2265 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2266 sound tuned around the original one, like in a chorus where some vocals are slightly
2269 It accepts the following parameters:
2272 Set input gain. Default is 0.4.
2275 Set output gain. Default is 0.4.
2278 Set delays. A typical delay is around 40ms to 60ms.
2290 @subsection Examples
2296 chorus=0.7:0.9:55:0.4:0.25:2
2302 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2306 Fuller sounding chorus with three delays:
2308 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
2313 Compress or expand the audio's dynamic range.
2315 It accepts the following parameters:
2321 A list of times in seconds for each channel over which the instantaneous level
2322 of the input signal is averaged to determine its volume. @var{attacks} refers to
2323 increase of volume and @var{decays} refers to decrease of volume. For most
2324 situations, the attack time (response to the audio getting louder) should be
2325 shorter than the decay time, because the human ear is more sensitive to sudden
2326 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2327 a typical value for decay is 0.8 seconds.
2328 If specified number of attacks & decays is lower than number of channels, the last
2329 set attack/decay will be used for all remaining channels.
2332 A list of points for the transfer function, specified in dB relative to the
2333 maximum possible signal amplitude. Each key points list must be defined using
2334 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2335 @code{x0/y0 x1/y1 x2/y2 ....}
2337 The input values must be in strictly increasing order but the transfer function
2338 does not have to be monotonically rising. The point @code{0/0} is assumed but
2339 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2340 function are @code{-70/-70|-60/-20|1/0}.
2343 Set the curve radius in dB for all joints. It defaults to 0.01.
2346 Set the additional gain in dB to be applied at all points on the transfer
2347 function. This allows for easy adjustment of the overall gain.
2351 Set an initial volume, in dB, to be assumed for each channel when filtering
2352 starts. This permits the user to supply a nominal level initially, so that, for
2353 example, a very large gain is not applied to initial signal levels before the
2354 companding has begun to operate. A typical value for audio which is initially
2355 quiet is -90 dB. It defaults to 0.
2358 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2359 delayed before being fed to the volume adjuster. Specifying a delay
2360 approximately equal to the attack/decay times allows the filter to effectively
2361 operate in predictive rather than reactive mode. It defaults to 0.
2365 @subsection Examples
2369 Make music with both quiet and loud passages suitable for listening to in a
2372 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2375 Another example for audio with whisper and explosion parts:
2377 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2381 A noise gate for when the noise is at a lower level than the signal:
2383 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2387 Here is another noise gate, this time for when the noise is at a higher level
2388 than the signal (making it, in some ways, similar to squelch):
2390 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2394 2:1 compression starting at -6dB:
2396 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2400 2:1 compression starting at -9dB:
2402 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2406 2:1 compression starting at -12dB:
2408 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2412 2:1 compression starting at -18dB:
2414 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2418 3:1 compression starting at -15dB:
2420 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2426 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2432 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
2436 Hard limiter at -6dB:
2438 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2442 Hard limiter at -12dB:
2444 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2448 Hard noise gate at -35 dB:
2450 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2456 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2460 @section compensationdelay
2462 Compensation Delay Line is a metric based delay to compensate differing
2463 positions of microphones or speakers.
2465 For example, you have recorded guitar with two microphones placed in
2466 different location. Because the front of sound wave has fixed speed in
2467 normal conditions, the phasing of microphones can vary and depends on
2468 their location and interposition. The best sound mix can be achieved when
2469 these microphones are in phase (synchronized). Note that distance of
2470 ~30 cm between microphones makes one microphone to capture signal in
2471 antiphase to another microphone. That makes the final mix sounding moody.
2472 This filter helps to solve phasing problems by adding different delays
2473 to each microphone track and make them synchronized.
2475 The best result can be reached when you take one track as base and
2476 synchronize other tracks one by one with it.
2477 Remember that synchronization/delay tolerance depends on sample rate, too.
2478 Higher sample rates will give more tolerance.
2480 It accepts the following parameters:
2484 Set millimeters distance. This is compensation distance for fine tuning.
2488 Set cm distance. This is compensation distance for tightening distance setup.
2492 Set meters distance. This is compensation distance for hard distance setup.
2496 Set dry amount. Amount of unprocessed (dry) signal.
2500 Set wet amount. Amount of processed (wet) signal.
2504 Set temperature degree in Celsius. This is the temperature of the environment.
2509 Apply headphone crossfeed filter.
2511 Crossfeed is the process of blending the left and right channels of stereo
2513 It is mainly used to reduce extreme stereo separation of low frequencies.
2515 The intent is to produce more speaker like sound to the listener.
2517 The filter accepts the following options:
2521 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2522 This sets gain of low shelf filter for side part of stereo image.
2523 Default is -6dB. Max allowed is -30db when strength is set to 1.
2526 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2527 This sets cut off frequency of low shelf filter. Default is cut off near
2528 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2531 Set input gain. Default is 0.9.
2534 Set output gain. Default is 1.
2537 @section crystalizer
2538 Simple algorithm to expand audio dynamic range.
2540 The filter accepts the following options:
2544 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2545 (unchanged sound) to 10.0 (maximum effect).
2548 Enable clipping. By default is enabled.
2552 Apply a DC shift to the audio.
2554 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2555 in the recording chain) from the audio. The effect of a DC offset is reduced
2556 headroom and hence volume. The @ref{astats} filter can be used to determine if
2557 a signal has a DC offset.
2561 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2565 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2566 used to prevent clipping.
2570 Measure audio dynamic range.
2572 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2573 is found in transition material. And anything less that 8 have very poor dynamics
2574 and is very compressed.
2576 The filter accepts the following options:
2580 Set window length in seconds used to split audio into segments of equal length.
2581 Default is 3 seconds.
2585 Dynamic Audio Normalizer.
2587 This filter applies a certain amount of gain to the input audio in order
2588 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2589 contrast to more "simple" normalization algorithms, the Dynamic Audio
2590 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2591 This allows for applying extra gain to the "quiet" sections of the audio
2592 while avoiding distortions or clipping the "loud" sections. In other words:
2593 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2594 sections, in the sense that the volume of each section is brought to the
2595 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2596 this goal *without* applying "dynamic range compressing". It will retain 100%
2597 of the dynamic range *within* each section of the audio file.
2601 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2602 Default is 500 milliseconds.
2603 The Dynamic Audio Normalizer processes the input audio in small chunks,
2604 referred to as frames. This is required, because a peak magnitude has no
2605 meaning for just a single sample value. Instead, we need to determine the
2606 peak magnitude for a contiguous sequence of sample values. While a "standard"
2607 normalizer would simply use the peak magnitude of the complete file, the
2608 Dynamic Audio Normalizer determines the peak magnitude individually for each
2609 frame. The length of a frame is specified in milliseconds. By default, the
2610 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2611 been found to give good results with most files.
2612 Note that the exact frame length, in number of samples, will be determined
2613 automatically, based on the sampling rate of the individual input audio file.
2616 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2617 number. Default is 31.
2618 Probably the most important parameter of the Dynamic Audio Normalizer is the
2619 @code{window size} of the Gaussian smoothing filter. The filter's window size
2620 is specified in frames, centered around the current frame. For the sake of
2621 simplicity, this must be an odd number. Consequently, the default value of 31
2622 takes into account the current frame, as well as the 15 preceding frames and
2623 the 15 subsequent frames. Using a larger window results in a stronger
2624 smoothing effect and thus in less gain variation, i.e. slower gain
2625 adaptation. Conversely, using a smaller window results in a weaker smoothing
2626 effect and thus in more gain variation, i.e. faster gain adaptation.
2627 In other words, the more you increase this value, the more the Dynamic Audio
2628 Normalizer will behave like a "traditional" normalization filter. On the
2629 contrary, the more you decrease this value, the more the Dynamic Audio
2630 Normalizer will behave like a dynamic range compressor.
2633 Set the target peak value. This specifies the highest permissible magnitude
2634 level for the normalized audio input. This filter will try to approach the
2635 target peak magnitude as closely as possible, but at the same time it also
2636 makes sure that the normalized signal will never exceed the peak magnitude.
2637 A frame's maximum local gain factor is imposed directly by the target peak
2638 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2639 It is not recommended to go above this value.
2642 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2643 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2644 factor for each input frame, i.e. the maximum gain factor that does not
2645 result in clipping or distortion. The maximum gain factor is determined by
2646 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2647 additionally bounds the frame's maximum gain factor by a predetermined
2648 (global) maximum gain factor. This is done in order to avoid excessive gain
2649 factors in "silent" or almost silent frames. By default, the maximum gain
2650 factor is 10.0, For most inputs the default value should be sufficient and
2651 it usually is not recommended to increase this value. Though, for input
2652 with an extremely low overall volume level, it may be necessary to allow even
2653 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2654 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2655 Instead, a "sigmoid" threshold function will be applied. This way, the
2656 gain factors will smoothly approach the threshold value, but never exceed that
2660 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2661 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2662 This means that the maximum local gain factor for each frame is defined
2663 (only) by the frame's highest magnitude sample. This way, the samples can
2664 be amplified as much as possible without exceeding the maximum signal
2665 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2666 Normalizer can also take into account the frame's root mean square,
2667 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2668 determine the power of a time-varying signal. It is therefore considered
2669 that the RMS is a better approximation of the "perceived loudness" than
2670 just looking at the signal's peak magnitude. Consequently, by adjusting all
2671 frames to a constant RMS value, a uniform "perceived loudness" can be
2672 established. If a target RMS value has been specified, a frame's local gain
2673 factor is defined as the factor that would result in exactly that RMS value.
2674 Note, however, that the maximum local gain factor is still restricted by the
2675 frame's highest magnitude sample, in order to prevent clipping.
2678 Enable channels coupling. By default is enabled.
2679 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2680 amount. This means the same gain factor will be applied to all channels, i.e.
2681 the maximum possible gain factor is determined by the "loudest" channel.
2682 However, in some recordings, it may happen that the volume of the different
2683 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2684 In this case, this option can be used to disable the channel coupling. This way,
2685 the gain factor will be determined independently for each channel, depending
2686 only on the individual channel's highest magnitude sample. This allows for
2687 harmonizing the volume of the different channels.
2690 Enable DC bias correction. By default is disabled.
2691 An audio signal (in the time domain) is a sequence of sample values.
2692 In the Dynamic Audio Normalizer these sample values are represented in the
2693 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2694 audio signal, or "waveform", should be centered around the zero point.
2695 That means if we calculate the mean value of all samples in a file, or in a
2696 single frame, then the result should be 0.0 or at least very close to that
2697 value. If, however, there is a significant deviation of the mean value from
2698 0.0, in either positive or negative direction, this is referred to as a
2699 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2700 Audio Normalizer provides optional DC bias correction.
2701 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2702 the mean value, or "DC correction" offset, of each input frame and subtract
2703 that value from all of the frame's sample values which ensures those samples
2704 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2705 boundaries, the DC correction offset values will be interpolated smoothly
2706 between neighbouring frames.
2709 Enable alternative boundary mode. By default is disabled.
2710 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2711 around each frame. This includes the preceding frames as well as the
2712 subsequent frames. However, for the "boundary" frames, located at the very
2713 beginning and at the very end of the audio file, not all neighbouring
2714 frames are available. In particular, for the first few frames in the audio
2715 file, the preceding frames are not known. And, similarly, for the last few
2716 frames in the audio file, the subsequent frames are not known. Thus, the
2717 question arises which gain factors should be assumed for the missing frames
2718 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2719 to deal with this situation. The default boundary mode assumes a gain factor
2720 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2721 "fade out" at the beginning and at the end of the input, respectively.
2724 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2725 By default, the Dynamic Audio Normalizer does not apply "traditional"
2726 compression. This means that signal peaks will not be pruned and thus the
2727 full dynamic range will be retained within each local neighbourhood. However,
2728 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2729 normalization algorithm with a more "traditional" compression.
2730 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2731 (thresholding) function. If (and only if) the compression feature is enabled,
2732 all input frames will be processed by a soft knee thresholding function prior
2733 to the actual normalization process. Put simply, the thresholding function is
2734 going to prune all samples whose magnitude exceeds a certain threshold value.
2735 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2736 value. Instead, the threshold value will be adjusted for each individual
2738 In general, smaller parameters result in stronger compression, and vice versa.
2739 Values below 3.0 are not recommended, because audible distortion may appear.
2744 Make audio easier to listen to on headphones.
2746 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2747 so that when listened to on headphones the stereo image is moved from
2748 inside your head (standard for headphones) to outside and in front of
2749 the listener (standard for speakers).
2755 Apply a two-pole peaking equalisation (EQ) filter. With this
2756 filter, the signal-level at and around a selected frequency can
2757 be increased or decreased, whilst (unlike bandpass and bandreject
2758 filters) that at all other frequencies is unchanged.
2760 In order to produce complex equalisation curves, this filter can
2761 be given several times, each with a different central frequency.
2763 The filter accepts the following options:
2767 Set the filter's central frequency in Hz.
2770 Set method to specify band-width of filter.
2785 Specify the band-width of a filter in width_type units.
2788 Set the required gain or attenuation in dB.
2789 Beware of clipping when using a positive gain.
2792 Specify which channels to filter, by default all available are filtered.
2795 @subsection Examples
2798 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2800 equalizer=f=1000:t=h:width=200:g=-10
2804 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2806 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2810 @subsection Commands
2812 This filter supports the following commands:
2815 Change equalizer frequency.
2816 Syntax for the command is : "@var{frequency}"
2819 Change equalizer width_type.
2820 Syntax for the command is : "@var{width_type}"
2823 Change equalizer width.
2824 Syntax for the command is : "@var{width}"
2827 Change equalizer gain.
2828 Syntax for the command is : "@var{gain}"
2831 @section extrastereo
2833 Linearly increases the difference between left and right channels which
2834 adds some sort of "live" effect to playback.
2836 The filter accepts the following options:
2840 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2841 (average of both channels), with 1.0 sound will be unchanged, with
2842 -1.0 left and right channels will be swapped.
2845 Enable clipping. By default is enabled.
2848 @section firequalizer
2849 Apply FIR Equalization using arbitrary frequency response.
2851 The filter accepts the following option:
2855 Set gain curve equation (in dB). The expression can contain variables:
2858 the evaluated frequency
2862 channel number, set to 0 when multichannels evaluation is disabled
2864 channel id, see libavutil/channel_layout.h, set to the first channel id when
2865 multichannels evaluation is disabled
2869 channel_layout, see libavutil/channel_layout.h
2874 @item gain_interpolate(f)
2875 interpolate gain on frequency f based on gain_entry
2876 @item cubic_interpolate(f)
2877 same as gain_interpolate, but smoother
2879 This option is also available as command. Default is @code{gain_interpolate(f)}.
2882 Set gain entry for gain_interpolate function. The expression can
2886 store gain entry at frequency f with value g
2888 This option is also available as command.
2891 Set filter delay in seconds. Higher value means more accurate.
2892 Default is @code{0.01}.
2895 Set filter accuracy in Hz. Lower value means more accurate.
2896 Default is @code{5}.
2899 Set window function. Acceptable values are:
2902 rectangular window, useful when gain curve is already smooth
2904 hann window (default)
2910 3-terms continuous 1st derivative nuttall window
2912 minimum 3-terms discontinuous nuttall window
2914 4-terms continuous 1st derivative nuttall window
2916 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2918 blackman-harris window
2924 If enabled, use fixed number of audio samples. This improves speed when
2925 filtering with large delay. Default is disabled.
2928 Enable multichannels evaluation on gain. Default is disabled.
2931 Enable zero phase mode by subtracting timestamp to compensate delay.
2932 Default is disabled.
2935 Set scale used by gain. Acceptable values are:
2938 linear frequency, linear gain
2940 linear frequency, logarithmic (in dB) gain (default)
2942 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2944 logarithmic frequency, logarithmic gain
2948 Set file for dumping, suitable for gnuplot.
2951 Set scale for dumpfile. Acceptable values are same with scale option.
2955 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2956 Default is disabled.
2959 Enable minimum phase impulse response. Default is disabled.
2962 @subsection Examples
2967 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2970 lowpass at 1000 Hz with gain_entry:
2972 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2975 custom equalization:
2977 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2980 higher delay with zero phase to compensate delay:
2982 firequalizer=delay=0.1:fixed=on:zero_phase=on
2985 lowpass on left channel, highpass on right channel:
2987 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2988 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2993 Apply a flanging effect to the audio.
2995 The filter accepts the following options:
2999 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3002 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3005 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3009 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3010 Default value is 71.
3013 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3016 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3017 Default value is @var{sinusoidal}.
3020 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3021 Default value is 25.
3024 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3025 Default is @var{linear}.
3029 Apply Haas effect to audio.
3031 Note that this makes most sense to apply on mono signals.
3032 With this filter applied to mono signals it give some directionality and
3033 stretches its stereo image.
3035 The filter accepts the following options:
3039 Set input level. By default is @var{1}, or 0dB
3042 Set output level. By default is @var{1}, or 0dB.
3045 Set gain applied to side part of signal. By default is @var{1}.
3048 Set kind of middle source. Can be one of the following:
3058 Pick middle part signal of stereo image.
3061 Pick side part signal of stereo image.
3065 Change middle phase. By default is disabled.
3068 Set left channel delay. By default is @var{2.05} milliseconds.
3071 Set left channel balance. By default is @var{-1}.
3074 Set left channel gain. By default is @var{1}.
3077 Change left phase. By default is disabled.
3080 Set right channel delay. By defaults is @var{2.12} milliseconds.
3083 Set right channel balance. By default is @var{1}.
3086 Set right channel gain. By default is @var{1}.
3089 Change right phase. By default is enabled.
3094 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3095 embedded HDCD codes is expanded into a 20-bit PCM stream.
3097 The filter supports the Peak Extend and Low-level Gain Adjustment features
3098 of HDCD, and detects the Transient Filter flag.
3101 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3104 When using the filter with wav, note the default encoding for wav is 16-bit,
3105 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3106 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3108 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3109 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3112 The filter accepts the following options:
3115 @item disable_autoconvert
3116 Disable any automatic format conversion or resampling in the filter graph.
3118 @item process_stereo
3119 Process the stereo channels together. If target_gain does not match between
3120 channels, consider it invalid and use the last valid target_gain.
3123 Set the code detect timer period in ms.
3126 Always extend peaks above -3dBFS even if PE isn't signaled.
3129 Replace audio with a solid tone and adjust the amplitude to signal some
3130 specific aspect of the decoding process. The output file can be loaded in
3131 an audio editor alongside the original to aid analysis.
3133 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3140 Gain adjustment level at each sample
3142 Samples where peak extend occurs
3144 Samples where the code detect timer is active
3146 Samples where the target gain does not match between channels
3152 Apply head-related transfer functions (HRTFs) to create virtual
3153 loudspeakers around the user for binaural listening via headphones.
3154 The HRIRs are provided via additional streams, for each channel
3155 one stereo input stream is needed.
3157 The filter accepts the following options:
3161 Set mapping of input streams for convolution.
3162 The argument is a '|'-separated list of channel names in order as they
3163 are given as additional stream inputs for filter.
3164 This also specify number of input streams. Number of input streams
3165 must be not less than number of channels in first stream plus one.
3168 Set gain applied to audio. Value is in dB. Default is 0.
3171 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3172 processing audio in time domain which is slow.
3173 @var{freq} is processing audio in frequency domain which is fast.
3174 Default is @var{freq}.
3177 Set custom gain for LFE channels. Value is in dB. Default is 0.
3180 @subsection Examples
3184 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3185 each amovie filter use stereo file with IR coefficients as input.
3186 The files give coefficients for each position of virtual loudspeaker:
3188 ffmpeg -i input.wav -lavfi-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],[a:0][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3195 Apply a high-pass filter with 3dB point frequency.
3196 The filter can be either single-pole, or double-pole (the default).
3197 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3199 The filter accepts the following options:
3203 Set frequency in Hz. Default is 3000.
3206 Set number of poles. Default is 2.
3209 Set method to specify band-width of filter.
3224 Specify the band-width of a filter in width_type units.
3225 Applies only to double-pole filter.
3226 The default is 0.707q and gives a Butterworth response.
3229 Specify which channels to filter, by default all available are filtered.
3232 @subsection Commands
3234 This filter supports the following commands:
3237 Change highpass frequency.
3238 Syntax for the command is : "@var{frequency}"
3241 Change highpass width_type.
3242 Syntax for the command is : "@var{width_type}"
3245 Change highpass width.
3246 Syntax for the command is : "@var{width}"
3251 Join multiple input streams into one multi-channel stream.
3253 It accepts the following parameters:
3257 The number of input streams. It defaults to 2.
3259 @item channel_layout
3260 The desired output channel layout. It defaults to stereo.
3263 Map channels from inputs to output. The argument is a '|'-separated list of
3264 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3265 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3266 can be either the name of the input channel (e.g. FL for front left) or its
3267 index in the specified input stream. @var{out_channel} is the name of the output
3271 The filter will attempt to guess the mappings when they are not specified
3272 explicitly. It does so by first trying to find an unused matching input channel
3273 and if that fails it picks the first unused input channel.
3275 Join 3 inputs (with properly set channel layouts):
3277 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3280 Build a 5.1 output from 6 single-channel streams:
3282 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3283 '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'
3289 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3291 To enable compilation of this filter you need to configure FFmpeg with
3292 @code{--enable-ladspa}.
3296 Specifies the name of LADSPA plugin library to load. If the environment
3297 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3298 each one of the directories specified by the colon separated list in
3299 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3300 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3301 @file{/usr/lib/ladspa/}.
3304 Specifies the plugin within the library. Some libraries contain only
3305 one plugin, but others contain many of them. If this is not set filter
3306 will list all available plugins within the specified library.
3309 Set the '|' separated list of controls which are zero or more floating point
3310 values that determine the behavior of the loaded plugin (for example delay,
3312 Controls need to be defined using the following syntax:
3313 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3314 @var{valuei} is the value set on the @var{i}-th control.
3315 Alternatively they can be also defined using the following syntax:
3316 @var{value0}|@var{value1}|@var{value2}|..., where
3317 @var{valuei} is the value set on the @var{i}-th control.
3318 If @option{controls} is set to @code{help}, all available controls and
3319 their valid ranges are printed.
3321 @item sample_rate, s
3322 Specify the sample rate, default to 44100. Only used if plugin have
3326 Set the number of samples per channel per each output frame, default
3327 is 1024. Only used if plugin have zero inputs.
3330 Set the minimum duration of the sourced audio. See
3331 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3332 for the accepted syntax.
3333 Note that the resulting duration may be greater than the specified duration,
3334 as the generated audio is always cut at the end of a complete frame.
3335 If not specified, or the expressed duration is negative, the audio is
3336 supposed to be generated forever.
3337 Only used if plugin have zero inputs.
3341 @subsection Examples
3345 List all available plugins within amp (LADSPA example plugin) library:
3351 List all available controls and their valid ranges for @code{vcf_notch}
3352 plugin from @code{VCF} library:
3354 ladspa=f=vcf:p=vcf_notch:c=help
3358 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3361 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3365 Add reverberation to the audio using TAP-plugins
3366 (Tom's Audio Processing plugins):
3368 ladspa=file=tap_reverb:tap_reverb
3372 Generate white noise, with 0.2 amplitude:
3374 ladspa=file=cmt:noise_source_white:c=c0=.2
3378 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3379 @code{C* Audio Plugin Suite} (CAPS) library:
3381 ladspa=file=caps:Click:c=c1=20'
3385 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3387 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3391 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3392 @code{SWH Plugins} collection:
3394 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3398 Attenuate low frequencies using Multiband EQ from Steve Harris
3399 @code{SWH Plugins} collection:
3401 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3405 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3408 ladspa=caps:Narrower
3412 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3414 ladspa=caps:White:.2
3418 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3420 ladspa=caps:Fractal:c=c1=1
3424 Dynamic volume normalization using @code{VLevel} plugin:
3426 ladspa=vlevel-ladspa:vlevel_mono
3430 @subsection Commands
3432 This filter supports the following commands:
3435 Modify the @var{N}-th control value.
3437 If the specified value is not valid, it is ignored and prior one is kept.
3442 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3443 Support for both single pass (livestreams, files) and double pass (files) modes.
3444 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3445 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3446 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3448 The filter accepts the following options:
3452 Set integrated loudness target.
3453 Range is -70.0 - -5.0. Default value is -24.0.
3456 Set loudness range target.
3457 Range is 1.0 - 20.0. Default value is 7.0.
3460 Set maximum true peak.
3461 Range is -9.0 - +0.0. Default value is -2.0.
3463 @item measured_I, measured_i
3464 Measured IL of input file.
3465 Range is -99.0 - +0.0.
3467 @item measured_LRA, measured_lra
3468 Measured LRA of input file.
3469 Range is 0.0 - 99.0.
3471 @item measured_TP, measured_tp
3472 Measured true peak of input file.
3473 Range is -99.0 - +99.0.
3475 @item measured_thresh
3476 Measured threshold of input file.
3477 Range is -99.0 - +0.0.
3480 Set offset gain. Gain is applied before the true-peak limiter.
3481 Range is -99.0 - +99.0. Default is +0.0.
3484 Normalize linearly if possible.
3485 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3486 to be specified in order to use this mode.
3487 Options are true or false. Default is true.
3490 Treat mono input files as "dual-mono". If a mono file is intended for playback
3491 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3492 If set to @code{true}, this option will compensate for this effect.
3493 Multi-channel input files are not affected by this option.
3494 Options are true or false. Default is false.
3497 Set print format for stats. Options are summary, json, or none.
3498 Default value is none.
3503 Apply a low-pass filter with 3dB point frequency.
3504 The filter can be either single-pole or double-pole (the default).
3505 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3507 The filter accepts the following options:
3511 Set frequency in Hz. Default is 500.
3514 Set number of poles. Default is 2.
3517 Set method to specify band-width of filter.
3532 Specify the band-width of a filter in width_type units.
3533 Applies only to double-pole filter.
3534 The default is 0.707q and gives a Butterworth response.
3537 Specify which channels to filter, by default all available are filtered.
3540 @subsection Examples
3543 Lowpass only LFE channel, it LFE is not present it does nothing:
3549 @subsection Commands
3551 This filter supports the following commands:
3554 Change lowpass frequency.
3555 Syntax for the command is : "@var{frequency}"
3558 Change lowpass width_type.
3559 Syntax for the command is : "@var{width_type}"
3562 Change lowpass width.
3563 Syntax for the command is : "@var{width}"
3568 Load a LV2 (LADSPA Version 2) plugin.
3570 To enable compilation of this filter you need to configure FFmpeg with
3571 @code{--enable-lv2}.
3575 Specifies the plugin URI. You may need to escape ':'.
3578 Set the '|' separated list of controls which are zero or more floating point
3579 values that determine the behavior of the loaded plugin (for example delay,
3581 If @option{controls} is set to @code{help}, all available controls and
3582 their valid ranges are printed.
3584 @item sample_rate, s
3585 Specify the sample rate, default to 44100. Only used if plugin have
3589 Set the number of samples per channel per each output frame, default
3590 is 1024. Only used if plugin have zero inputs.
3593 Set the minimum duration of the sourced audio. See
3594 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3595 for the accepted syntax.
3596 Note that the resulting duration may be greater than the specified duration,
3597 as the generated audio is always cut at the end of a complete frame.
3598 If not specified, or the expressed duration is negative, the audio is
3599 supposed to be generated forever.
3600 Only used if plugin have zero inputs.
3603 @subsection Examples
3607 Apply bass enhancer plugin from Calf:
3609 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3613 Apply vinyl plugin from Calf:
3615 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3619 Apply bit crusher plugin from ArtyFX:
3621 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3626 Multiband Compress or expand the audio's dynamic range.
3628 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3629 This is akin to the crossover of a loudspeaker, and results in flat frequency
3630 response when absent compander action.
3632 It accepts the following parameters:
3636 This option syntax is:
3637 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3638 For explanation of each item refer to compand filter documentation.
3644 Mix channels with specific gain levels. The filter accepts the output
3645 channel layout followed by a set of channels definitions.
3647 This filter is also designed to efficiently remap the channels of an audio
3650 The filter accepts parameters of the form:
3651 "@var{l}|@var{outdef}|@var{outdef}|..."
3655 output channel layout or number of channels
3658 output channel specification, of the form:
3659 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3662 output channel to define, either a channel name (FL, FR, etc.) or a channel
3663 number (c0, c1, etc.)
3666 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3669 input channel to use, see out_name for details; it is not possible to mix
3670 named and numbered input channels
3673 If the `=' in a channel specification is replaced by `<', then the gains for
3674 that specification will be renormalized so that the total is 1, thus
3675 avoiding clipping noise.
3677 @subsection Mixing examples
3679 For example, if you want to down-mix from stereo to mono, but with a bigger
3680 factor for the left channel:
3682 pan=1c|c0=0.9*c0+0.1*c1
3685 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3686 7-channels surround:
3688 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3691 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3692 that should be preferred (see "-ac" option) unless you have very specific
3695 @subsection Remapping examples
3697 The channel remapping will be effective if, and only if:
3700 @item gain coefficients are zeroes or ones,
3701 @item only one input per channel output,
3704 If all these conditions are satisfied, the filter will notify the user ("Pure
3705 channel mapping detected"), and use an optimized and lossless method to do the
3708 For example, if you have a 5.1 source and want a stereo audio stream by
3709 dropping the extra channels:
3711 pan="stereo| c0=FL | c1=FR"
3714 Given the same source, you can also switch front left and front right channels
3715 and keep the input channel layout:
3717 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3720 If the input is a stereo audio stream, you can mute the front left channel (and
3721 still keep the stereo channel layout) with:
3726 Still with a stereo audio stream input, you can copy the right channel in both
3727 front left and right:
3729 pan="stereo| c0=FR | c1=FR"
3734 ReplayGain scanner filter. This filter takes an audio stream as an input and
3735 outputs it unchanged.
3736 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3740 Convert the audio sample format, sample rate and channel layout. It is
3741 not meant to be used directly.
3744 Apply time-stretching and pitch-shifting with librubberband.
3746 The filter accepts the following options:
3750 Set tempo scale factor.
3753 Set pitch scale factor.
3756 Set transients detector.
3757 Possible values are:
3766 Possible values are:
3775 Possible values are:
3782 Set processing window size.
3783 Possible values are:
3792 Possible values are:
3799 Enable formant preservation when shift pitching.
3800 Possible values are:
3808 Possible values are:
3817 Possible values are:
3824 @section sidechaincompress
3826 This filter acts like normal compressor but has the ability to compress
3827 detected signal using second input signal.
3828 It needs two input streams and returns one output stream.
3829 First input stream will be processed depending on second stream signal.
3830 The filtered signal then can be filtered with other filters in later stages of
3831 processing. See @ref{pan} and @ref{amerge} filter.
3833 The filter accepts the following options:
3837 Set input gain. Default is 1. Range is between 0.015625 and 64.
3840 If a signal of second stream raises above this level it will affect the gain
3841 reduction of first stream.
3842 By default is 0.125. Range is between 0.00097563 and 1.
3845 Set a ratio about which the signal is reduced. 1:2 means that if the level
3846 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3847 Default is 2. Range is between 1 and 20.
3850 Amount of milliseconds the signal has to rise above the threshold before gain
3851 reduction starts. Default is 20. Range is between 0.01 and 2000.
3854 Amount of milliseconds the signal has to fall below the threshold before
3855 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3858 Set the amount by how much signal will be amplified after processing.
3859 Default is 1. Range is from 1 to 64.
3862 Curve the sharp knee around the threshold to enter gain reduction more softly.
3863 Default is 2.82843. Range is between 1 and 8.
3866 Choose if the @code{average} level between all channels of side-chain stream
3867 or the louder(@code{maximum}) channel of side-chain stream affects the
3868 reduction. Default is @code{average}.
3871 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3872 of @code{rms}. Default is @code{rms} which is mainly smoother.
3875 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3878 How much to use compressed signal in output. Default is 1.
3879 Range is between 0 and 1.
3882 @subsection Examples
3886 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3887 depending on the signal of 2nd input and later compressed signal to be
3888 merged with 2nd input:
3890 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3894 @section sidechaingate
3896 A sidechain gate acts like a normal (wideband) gate but has the ability to
3897 filter the detected signal before sending it to the gain reduction stage.
3898 Normally a gate uses the full range signal to detect a level above the
3900 For example: If you cut all lower frequencies from your sidechain signal
3901 the gate will decrease the volume of your track only if not enough highs
3902 appear. With this technique you are able to reduce the resonation of a
3903 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3905 It needs two input streams and returns one output stream.
3906 First input stream will be processed depending on second stream signal.
3908 The filter accepts the following options:
3912 Set input level before filtering.
3913 Default is 1. Allowed range is from 0.015625 to 64.
3916 Set the level of gain reduction when the signal is below the threshold.
3917 Default is 0.06125. Allowed range is from 0 to 1.
3920 If a signal rises above this level the gain reduction is released.
3921 Default is 0.125. Allowed range is from 0 to 1.
3924 Set a ratio about which the signal is reduced.
3925 Default is 2. Allowed range is from 1 to 9000.
3928 Amount of milliseconds the signal has to rise above the threshold before gain
3930 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3933 Amount of milliseconds the signal has to fall below the threshold before the
3934 reduction is increased again. Default is 250 milliseconds.
3935 Allowed range is from 0.01 to 9000.
3938 Set amount of amplification of signal after processing.
3939 Default is 1. Allowed range is from 1 to 64.
3942 Curve the sharp knee around the threshold to enter gain reduction more softly.
3943 Default is 2.828427125. Allowed range is from 1 to 8.
3946 Choose if exact signal should be taken for detection or an RMS like one.
3947 Default is rms. Can be peak or rms.
3950 Choose if the average level between all channels or the louder channel affects
3952 Default is average. Can be average or maximum.
3955 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3958 @section silencedetect
3960 Detect silence in an audio stream.
3962 This filter logs a message when it detects that the input audio volume is less
3963 or equal to a noise tolerance value for a duration greater or equal to the
3964 minimum detected noise duration.
3966 The printed times and duration are expressed in seconds.
3968 The filter accepts the following options:
3972 Set silence duration until notification (default is 2 seconds).
3975 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3976 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3979 @subsection Examples
3983 Detect 5 seconds of silence with -50dB noise tolerance:
3985 silencedetect=n=-50dB:d=5
3989 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3990 tolerance in @file{silence.mp3}:
3992 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3996 @section silenceremove
3998 Remove silence from the beginning, middle or end of the audio.
4000 The filter accepts the following options:
4004 This value is used to indicate if audio should be trimmed at beginning of
4005 the audio. A value of zero indicates no silence should be trimmed from the
4006 beginning. When specifying a non-zero value, it trims audio up until it
4007 finds non-silence. Normally, when trimming silence from beginning of audio
4008 the @var{start_periods} will be @code{1} but it can be increased to higher
4009 values to trim all audio up to specific count of non-silence periods.
4010 Default value is @code{0}.
4012 @item start_duration
4013 Specify the amount of time that non-silence must be detected before it stops
4014 trimming audio. By increasing the duration, bursts of noises can be treated
4015 as silence and trimmed off. Default value is @code{0}.
4017 @item start_threshold
4018 This indicates what sample value should be treated as silence. For digital
4019 audio, a value of @code{0} may be fine but for audio recorded from analog,
4020 you may wish to increase the value to account for background noise.
4021 Can be specified in dB (in case "dB" is appended to the specified value)
4022 or amplitude ratio. Default value is @code{0}.
4025 Set the count for trimming silence from the end of audio.
4026 To remove silence from the middle of a file, specify a @var{stop_periods}
4027 that is negative. This value is then treated as a positive value and is
4028 used to indicate the effect should restart processing as specified by
4029 @var{start_periods}, making it suitable for removing periods of silence
4030 in the middle of the audio.
4031 Default value is @code{0}.
4034 Specify a duration of silence that must exist before audio is not copied any
4035 more. By specifying a higher duration, silence that is wanted can be left in
4037 Default value is @code{0}.
4039 @item stop_threshold
4040 This is the same as @option{start_threshold} but for trimming silence from
4042 Can be specified in dB (in case "dB" is appended to the specified value)
4043 or amplitude ratio. Default value is @code{0}.
4046 This indicates that @var{stop_duration} length of audio should be left intact
4047 at the beginning of each period of silence.
4048 For example, if you want to remove long pauses between words but do not want
4049 to remove the pauses completely. Default value is @code{0}.
4052 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4053 and works better with digital silence which is exactly 0.
4054 Default value is @code{rms}.
4057 Set ratio used to calculate size of window for detecting silence.
4058 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4061 @subsection Examples
4065 The following example shows how this filter can be used to start a recording
4066 that does not contain the delay at the start which usually occurs between
4067 pressing the record button and the start of the performance:
4069 silenceremove=1:5:0.02
4073 Trim all silence encountered from beginning to end where there is more than 1
4074 second of silence in audio:
4076 silenceremove=0:0:0:-1:1:-90dB
4082 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4083 loudspeakers around the user for binaural listening via headphones (audio
4084 formats up to 9 channels supported).
4085 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4086 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4087 Austrian Academy of Sciences.
4089 To enable compilation of this filter you need to configure FFmpeg with
4090 @code{--enable-libmysofa}.
4092 The filter accepts the following options:
4096 Set the SOFA file used for rendering.
4099 Set gain applied to audio. Value is in dB. Default is 0.
4102 Set rotation of virtual loudspeakers in deg. Default is 0.
4105 Set elevation of virtual speakers in deg. Default is 0.
4108 Set distance in meters between loudspeakers and the listener with near-field
4109 HRTFs. Default is 1.
4112 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4113 processing audio in time domain which is slow.
4114 @var{freq} is processing audio in frequency domain which is fast.
4115 Default is @var{freq}.
4118 Set custom positions of virtual loudspeakers. Syntax for this option is:
4119 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4120 Each virtual loudspeaker is described with short channel name following with
4121 azimuth and elevation in degrees.
4122 Each virtual loudspeaker description is separated by '|'.
4123 For example to override front left and front right channel positions use:
4124 'speakers=FL 45 15|FR 345 15'.
4125 Descriptions with unrecognised channel names are ignored.
4128 Set custom gain for LFE channels. Value is in dB. Default is 0.
4131 @subsection Examples
4135 Using ClubFritz6 sofa file:
4137 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4141 Using ClubFritz12 sofa file and bigger radius with small rotation:
4143 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4147 Similar as above but with custom speaker positions for front left, front right, back left and back right
4148 and also with custom gain:
4150 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4154 @section stereotools
4156 This filter has some handy utilities to manage stereo signals, for converting
4157 M/S stereo recordings to L/R signal while having control over the parameters
4158 or spreading the stereo image of master track.
4160 The filter accepts the following options:
4164 Set input level before filtering for both channels. Defaults is 1.
4165 Allowed range is from 0.015625 to 64.
4168 Set output level after filtering for both channels. Defaults is 1.
4169 Allowed range is from 0.015625 to 64.
4172 Set input balance between both channels. Default is 0.
4173 Allowed range is from -1 to 1.
4176 Set output balance between both channels. Default is 0.
4177 Allowed range is from -1 to 1.
4180 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4181 clipping. Disabled by default.
4184 Mute the left channel. Disabled by default.
4187 Mute the right channel. Disabled by default.
4190 Change the phase of the left channel. Disabled by default.
4193 Change the phase of the right channel. Disabled by default.
4196 Set stereo mode. Available values are:
4200 Left/Right to Left/Right, this is default.
4203 Left/Right to Mid/Side.
4206 Mid/Side to Left/Right.
4209 Left/Right to Left/Left.
4212 Left/Right to Right/Right.
4215 Left/Right to Left + Right.
4218 Left/Right to Right/Left.
4221 Mid/Side to Left/Left.
4224 Mid/Side to Right/Right.
4228 Set level of side signal. Default is 1.
4229 Allowed range is from 0.015625 to 64.
4232 Set balance of side signal. Default is 0.
4233 Allowed range is from -1 to 1.
4236 Set level of the middle signal. Default is 1.
4237 Allowed range is from 0.015625 to 64.
4240 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4243 Set stereo base between mono and inversed channels. Default is 0.
4244 Allowed range is from -1 to 1.
4247 Set delay in milliseconds how much to delay left from right channel and
4248 vice versa. Default is 0. Allowed range is from -20 to 20.
4251 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4254 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4256 @item bmode_in, bmode_out
4257 Set balance mode for balance_in/balance_out option.
4259 Can be one of the following:
4263 Classic balance mode. Attenuate one channel at time.
4264 Gain is raised up to 1.
4267 Similar as classic mode above but gain is raised up to 2.
4270 Equal power distribution, from -6dB to +6dB range.
4274 @subsection Examples
4278 Apply karaoke like effect:
4280 stereotools=mlev=0.015625
4284 Convert M/S signal to L/R:
4286 "stereotools=mode=ms>lr"
4290 @section stereowiden
4292 This filter enhance the stereo effect by suppressing signal common to both
4293 channels and by delaying the signal of left into right and vice versa,
4294 thereby widening the stereo effect.
4296 The filter accepts the following options:
4300 Time in milliseconds of the delay of left signal into right and vice versa.
4301 Default is 20 milliseconds.
4304 Amount of gain in delayed signal into right and vice versa. Gives a delay
4305 effect of left signal in right output and vice versa which gives widening
4306 effect. Default is 0.3.
4309 Cross feed of left into right with inverted phase. This helps in suppressing
4310 the mono. If the value is 1 it will cancel all the signal common to both
4311 channels. Default is 0.3.
4314 Set level of input signal of original channel. Default is 0.8.
4317 @section superequalizer
4318 Apply 18 band equalizer.
4320 The filter accepts the following options:
4327 Set 131Hz band gain.
4329 Set 185Hz band gain.
4331 Set 262Hz band gain.
4333 Set 370Hz band gain.
4335 Set 523Hz band gain.
4337 Set 740Hz band gain.
4339 Set 1047Hz band gain.
4341 Set 1480Hz band gain.
4343 Set 2093Hz band gain.
4345 Set 2960Hz band gain.
4347 Set 4186Hz band gain.
4349 Set 5920Hz band gain.
4351 Set 8372Hz band gain.
4353 Set 11840Hz band gain.
4355 Set 16744Hz band gain.
4357 Set 20000Hz band gain.
4361 Apply audio surround upmix filter.
4363 This filter allows to produce multichannel output from audio stream.
4365 The filter accepts the following options:
4369 Set output channel layout. By default, this is @var{5.1}.
4371 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4372 for the required syntax.
4375 Set input channel layout. By default, this is @var{stereo}.
4377 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4378 for the required syntax.
4381 Set input volume level. By default, this is @var{1}.
4384 Set output volume level. By default, this is @var{1}.
4387 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4390 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4393 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4396 Set front center input volume. By default, this is @var{1}.
4399 Set front center output volume. By default, this is @var{1}.
4402 Set LFE input volume. By default, this is @var{1}.
4405 Set LFE output volume. By default, this is @var{1}.
4410 Boost or cut treble (upper) frequencies of the audio using a two-pole
4411 shelving filter with a response similar to that of a standard
4412 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4414 The filter accepts the following options:
4418 Give the gain at whichever is the lower of ~22 kHz and the
4419 Nyquist frequency. Its useful range is about -20 (for a large cut)
4420 to +20 (for a large boost). Beware of clipping when using a positive gain.
4423 Set the filter's central frequency and so can be used
4424 to extend or reduce the frequency range to be boosted or cut.
4425 The default value is @code{3000} Hz.
4428 Set method to specify band-width of filter.
4443 Determine how steep is the filter's shelf transition.
4446 Specify which channels to filter, by default all available are filtered.
4449 @subsection Commands
4451 This filter supports the following commands:
4454 Change treble frequency.
4455 Syntax for the command is : "@var{frequency}"
4458 Change treble width_type.
4459 Syntax for the command is : "@var{width_type}"
4462 Change treble width.
4463 Syntax for the command is : "@var{width}"
4467 Syntax for the command is : "@var{gain}"
4472 Sinusoidal amplitude modulation.
4474 The filter accepts the following options:
4478 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4479 (20 Hz or lower) will result in a tremolo effect.
4480 This filter may also be used as a ring modulator by specifying
4481 a modulation frequency higher than 20 Hz.
4482 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4485 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4486 Default value is 0.5.
4491 Sinusoidal phase modulation.
4493 The filter accepts the following options:
4497 Modulation frequency in Hertz.
4498 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4501 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4502 Default value is 0.5.
4507 Adjust the input audio volume.
4509 It accepts the following parameters:
4513 Set audio volume expression.
4515 Output values are clipped to the maximum value.
4517 The output audio volume is given by the relation:
4519 @var{output_volume} = @var{volume} * @var{input_volume}
4522 The default value for @var{volume} is "1.0".
4525 This parameter represents the mathematical precision.
4527 It determines which input sample formats will be allowed, which affects the
4528 precision of the volume scaling.
4532 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4534 32-bit floating-point; this limits input sample format to FLT. (default)
4536 64-bit floating-point; this limits input sample format to DBL.
4540 Choose the behaviour on encountering ReplayGain side data in input frames.
4544 Remove ReplayGain side data, ignoring its contents (the default).
4547 Ignore ReplayGain side data, but leave it in the frame.
4550 Prefer the track gain, if present.
4553 Prefer the album gain, if present.
4556 @item replaygain_preamp
4557 Pre-amplification gain in dB to apply to the selected replaygain gain.
4559 Default value for @var{replaygain_preamp} is 0.0.
4562 Set when the volume expression is evaluated.
4564 It accepts the following values:
4567 only evaluate expression once during the filter initialization, or
4568 when the @samp{volume} command is sent
4571 evaluate expression for each incoming frame
4574 Default value is @samp{once}.
4577 The volume expression can contain the following parameters.
4581 frame number (starting at zero)
4584 @item nb_consumed_samples
4585 number of samples consumed by the filter
4587 number of samples in the current frame
4589 original frame position in the file
4595 PTS at start of stream
4597 time at start of stream
4603 last set volume value
4606 Note that when @option{eval} is set to @samp{once} only the
4607 @var{sample_rate} and @var{tb} variables are available, all other
4608 variables will evaluate to NAN.
4610 @subsection Commands
4612 This filter supports the following commands:
4615 Modify the volume expression.
4616 The command accepts the same syntax of the corresponding option.
4618 If the specified expression is not valid, it is kept at its current
4620 @item replaygain_noclip
4621 Prevent clipping by limiting the gain applied.
4623 Default value for @var{replaygain_noclip} is 1.
4627 @subsection Examples
4631 Halve the input audio volume:
4635 volume=volume=-6.0206dB
4638 In all the above example the named key for @option{volume} can be
4639 omitted, for example like in:
4645 Increase input audio power by 6 decibels using fixed-point precision:
4647 volume=volume=6dB:precision=fixed
4651 Fade volume after time 10 with an annihilation period of 5 seconds:
4653 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4657 @section volumedetect
4659 Detect the volume of the input video.
4661 The filter has no parameters. The input is not modified. Statistics about
4662 the volume will be printed in the log when the input stream end is reached.
4664 In particular it will show the mean volume (root mean square), maximum
4665 volume (on a per-sample basis), and the beginning of a histogram of the
4666 registered volume values (from the maximum value to a cumulated 1/1000 of
4669 All volumes are in decibels relative to the maximum PCM value.
4671 @subsection Examples
4673 Here is an excerpt of the output:
4675 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4676 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4677 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4678 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4679 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4680 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4681 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4682 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4683 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4689 The mean square energy is approximately -27 dB, or 10^-2.7.
4691 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4693 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4696 In other words, raising the volume by +4 dB does not cause any clipping,
4697 raising it by +5 dB causes clipping for 6 samples, etc.
4699 @c man end AUDIO FILTERS
4701 @chapter Audio Sources
4702 @c man begin AUDIO SOURCES
4704 Below is a description of the currently available audio sources.
4708 Buffer audio frames, and make them available to the filter chain.
4710 This source is mainly intended for a programmatic use, in particular
4711 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4713 It accepts the following parameters:
4717 The timebase which will be used for timestamps of submitted frames. It must be
4718 either a floating-point number or in @var{numerator}/@var{denominator} form.
4721 The sample rate of the incoming audio buffers.
4724 The sample format of the incoming audio buffers.
4725 Either a sample format name or its corresponding integer representation from
4726 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4728 @item channel_layout
4729 The channel layout of the incoming audio buffers.
4730 Either a channel layout name from channel_layout_map in
4731 @file{libavutil/channel_layout.c} or its corresponding integer representation
4732 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4735 The number of channels of the incoming audio buffers.
4736 If both @var{channels} and @var{channel_layout} are specified, then they
4741 @subsection Examples
4744 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4747 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4748 Since the sample format with name "s16p" corresponds to the number
4749 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4752 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4757 Generate an audio signal specified by an expression.
4759 This source accepts in input one or more expressions (one for each
4760 channel), which are evaluated and used to generate a corresponding
4763 This source accepts the following options:
4767 Set the '|'-separated expressions list for each separate channel. In case the
4768 @option{channel_layout} option is not specified, the selected channel layout
4769 depends on the number of provided expressions. Otherwise the last
4770 specified expression is applied to the remaining output channels.
4772 @item channel_layout, c
4773 Set the channel layout. The number of channels in the specified layout
4774 must be equal to the number of specified expressions.
4777 Set the minimum duration of the sourced audio. See
4778 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4779 for the accepted syntax.
4780 Note that the resulting duration may be greater than the specified
4781 duration, as the generated audio is always cut at the end of a
4784 If not specified, or the expressed duration is negative, the audio is
4785 supposed to be generated forever.
4788 Set the number of samples per channel per each output frame,
4791 @item sample_rate, s
4792 Specify the sample rate, default to 44100.
4795 Each expression in @var{exprs} can contain the following constants:
4799 number of the evaluated sample, starting from 0
4802 time of the evaluated sample expressed in seconds, starting from 0
4809 @subsection Examples
4819 Generate a sin signal with frequency of 440 Hz, set sample rate to
4822 aevalsrc="sin(440*2*PI*t):s=8000"
4826 Generate a two channels signal, specify the channel layout (Front
4827 Center + Back Center) explicitly:
4829 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4833 Generate white noise:
4835 aevalsrc="-2+random(0)"
4839 Generate an amplitude modulated signal:
4841 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4845 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4847 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4854 The null audio source, return unprocessed audio frames. It is mainly useful
4855 as a template and to be employed in analysis / debugging tools, or as
4856 the source for filters which ignore the input data (for example the sox
4859 This source accepts the following options:
4863 @item channel_layout, cl
4865 Specifies the channel layout, and can be either an integer or a string
4866 representing a channel layout. The default value of @var{channel_layout}
4869 Check the channel_layout_map definition in
4870 @file{libavutil/channel_layout.c} for the mapping between strings and
4871 channel layout values.
4873 @item sample_rate, r
4874 Specifies the sample rate, and defaults to 44100.
4877 Set the number of samples per requested frames.
4881 @subsection Examples
4885 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4887 anullsrc=r=48000:cl=4
4891 Do the same operation with a more obvious syntax:
4893 anullsrc=r=48000:cl=mono
4897 All the parameters need to be explicitly defined.
4901 Synthesize a voice utterance using the libflite library.
4903 To enable compilation of this filter you need to configure FFmpeg with
4904 @code{--enable-libflite}.
4906 Note that versions of the flite library prior to 2.0 are not thread-safe.
4908 The filter accepts the following options:
4913 If set to 1, list the names of the available voices and exit
4914 immediately. Default value is 0.
4917 Set the maximum number of samples per frame. Default value is 512.
4920 Set the filename containing the text to speak.
4923 Set the text to speak.
4926 Set the voice to use for the speech synthesis. Default value is
4927 @code{kal}. See also the @var{list_voices} option.
4930 @subsection Examples
4934 Read from file @file{speech.txt}, and synthesize the text using the
4935 standard flite voice:
4937 flite=textfile=speech.txt
4941 Read the specified text selecting the @code{slt} voice:
4943 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4947 Input text to ffmpeg:
4949 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4953 Make @file{ffplay} speak the specified text, using @code{flite} and
4954 the @code{lavfi} device:
4956 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4960 For more information about libflite, check:
4961 @url{http://www.festvox.org/flite/}
4965 Generate a noise audio signal.
4967 The filter accepts the following options:
4970 @item sample_rate, r
4971 Specify the sample rate. Default value is 48000 Hz.
4974 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4978 Specify the duration of the generated audio stream. Not specifying this option
4979 results in noise with an infinite length.
4981 @item color, colour, c
4982 Specify the color of noise. Available noise colors are white, pink, brown,
4983 blue and violet. Default color is white.
4986 Specify a value used to seed the PRNG.
4989 Set the number of samples per each output frame, default is 1024.
4992 @subsection Examples
4997 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4999 anoisesrc=d=60:c=pink:r=44100:a=0.5
5005 Generate odd-tap Hilbert transform FIR coefficients.
5007 The resulting stream can be used with @ref{afir} filter for phase-shifting
5008 the signal by 90 degrees.
5010 This is used in many matrix coding schemes and for analytic signal generation.
5011 The process is often written as a multiplication by i (or j), the imaginary unit.
5013 The filter accepts the following options:
5017 @item sample_rate, s
5018 Set sample rate, default is 44100.
5021 Set length of FIR filter, default is 22051.
5024 Set number of samples per each frame.
5027 Set window function to be used when generating FIR coefficients.
5032 Generate an audio signal made of a sine wave with amplitude 1/8.
5034 The audio signal is bit-exact.
5036 The filter accepts the following options:
5041 Set the carrier frequency. Default is 440 Hz.
5043 @item beep_factor, b
5044 Enable a periodic beep every second with frequency @var{beep_factor} times
5045 the carrier frequency. Default is 0, meaning the beep is disabled.
5047 @item sample_rate, r
5048 Specify the sample rate, default is 44100.
5051 Specify the duration of the generated audio stream.
5053 @item samples_per_frame
5054 Set the number of samples per output frame.
5056 The expression can contain the following constants:
5060 The (sequential) number of the output audio frame, starting from 0.
5063 The PTS (Presentation TimeStamp) of the output audio frame,
5064 expressed in @var{TB} units.
5067 The PTS of the output audio frame, expressed in seconds.
5070 The timebase of the output audio frames.
5073 Default is @code{1024}.
5076 @subsection Examples
5081 Generate a simple 440 Hz sine wave:
5087 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5091 sine=frequency=220:beep_factor=4:duration=5
5095 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5098 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5102 @c man end AUDIO SOURCES
5104 @chapter Audio Sinks
5105 @c man begin AUDIO SINKS
5107 Below is a description of the currently available audio sinks.
5109 @section abuffersink
5111 Buffer audio frames, and make them available to the end of filter chain.
5113 This sink is mainly intended for programmatic use, in particular
5114 through the interface defined in @file{libavfilter/buffersink.h}
5115 or the options system.
5117 It accepts a pointer to an AVABufferSinkContext structure, which
5118 defines the incoming buffers' formats, to be passed as the opaque
5119 parameter to @code{avfilter_init_filter} for initialization.
5122 Null audio sink; do absolutely nothing with the input audio. It is
5123 mainly useful as a template and for use in analysis / debugging
5126 @c man end AUDIO SINKS
5128 @chapter Video Filters
5129 @c man begin VIDEO FILTERS
5131 When you configure your FFmpeg build, you can disable any of the
5132 existing filters using @code{--disable-filters}.
5133 The configure output will show the video filters included in your
5136 Below is a description of the currently available video filters.
5138 @section alphaextract
5140 Extract the alpha component from the input as a grayscale video. This
5141 is especially useful with the @var{alphamerge} filter.
5145 Add or replace the alpha component of the primary input with the
5146 grayscale value of a second input. This is intended for use with
5147 @var{alphaextract} to allow the transmission or storage of frame
5148 sequences that have alpha in a format that doesn't support an alpha
5151 For example, to reconstruct full frames from a normal YUV-encoded video
5152 and a separate video created with @var{alphaextract}, you might use:
5154 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5157 Since this filter is designed for reconstruction, it operates on frame
5158 sequences without considering timestamps, and terminates when either
5159 input reaches end of stream. This will cause problems if your encoding
5160 pipeline drops frames. If you're trying to apply an image as an
5161 overlay to a video stream, consider the @var{overlay} filter instead.
5165 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5166 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5167 Substation Alpha) subtitles files.
5169 This filter accepts the following option in addition to the common options from
5170 the @ref{subtitles} filter:
5174 Set the shaping engine
5176 Available values are:
5179 The default libass shaping engine, which is the best available.
5181 Fast, font-agnostic shaper that can do only substitutions
5183 Slower shaper using OpenType for substitutions and positioning
5186 The default is @code{auto}.
5190 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5192 The filter accepts the following options:
5196 Set threshold A for 1st plane. Default is 0.02.
5197 Valid range is 0 to 0.3.
5200 Set threshold B for 1st plane. Default is 0.04.
5201 Valid range is 0 to 5.
5204 Set threshold A for 2nd plane. Default is 0.02.
5205 Valid range is 0 to 0.3.
5208 Set threshold B for 2nd plane. Default is 0.04.
5209 Valid range is 0 to 5.
5212 Set threshold A for 3rd plane. Default is 0.02.
5213 Valid range is 0 to 0.3.
5216 Set threshold B for 3rd plane. Default is 0.04.
5217 Valid range is 0 to 5.
5219 Threshold A is designed to react on abrupt changes in the input signal and
5220 threshold B is designed to react on continuous changes in the input signal.
5223 Set number of frames filter will use for averaging. Default is 33. Must be odd
5224 number in range [5, 129].
5227 Set what planes of frame filter will use for averaging. Default is all.
5232 Apply average blur filter.
5234 The filter accepts the following options:
5238 Set horizontal kernel size.
5241 Set which planes to filter. By default all planes are filtered.
5244 Set vertical kernel size, if zero it will be same as @code{sizeX}.
5245 Default is @code{0}.
5250 Compute the bounding box for the non-black pixels in the input frame
5253 This filter computes the bounding box containing all the pixels with a
5254 luminance value greater than the minimum allowed value.
5255 The parameters describing the bounding box are printed on the filter
5258 The filter accepts the following option:
5262 Set the minimal luminance value. Default is @code{16}.
5265 @section bitplanenoise
5267 Show and measure bit plane noise.
5269 The filter accepts the following options:
5273 Set which plane to analyze. Default is @code{1}.
5276 Filter out noisy pixels from @code{bitplane} set above.
5277 Default is disabled.
5280 @section blackdetect
5282 Detect video intervals that are (almost) completely black. Can be
5283 useful to detect chapter transitions, commercials, or invalid
5284 recordings. Output lines contains the time for the start, end and
5285 duration of the detected black interval expressed in seconds.
5287 In order to display the output lines, you need to set the loglevel at
5288 least to the AV_LOG_INFO value.
5290 The filter accepts the following options:
5293 @item black_min_duration, d
5294 Set the minimum detected black duration expressed in seconds. It must
5295 be a non-negative floating point number.
5297 Default value is 2.0.
5299 @item picture_black_ratio_th, pic_th
5300 Set the threshold for considering a picture "black".
5301 Express the minimum value for the ratio:
5303 @var{nb_black_pixels} / @var{nb_pixels}
5306 for which a picture is considered black.
5307 Default value is 0.98.
5309 @item pixel_black_th, pix_th
5310 Set the threshold for considering a pixel "black".
5312 The threshold expresses the maximum pixel luminance value for which a
5313 pixel is considered "black". The provided value is scaled according to
5314 the following equation:
5316 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5319 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5320 the input video format, the range is [0-255] for YUV full-range
5321 formats and [16-235] for YUV non full-range formats.
5323 Default value is 0.10.
5326 The following example sets the maximum pixel threshold to the minimum
5327 value, and detects only black intervals of 2 or more seconds:
5329 blackdetect=d=2:pix_th=0.00
5334 Detect frames that are (almost) completely black. Can be useful to
5335 detect chapter transitions or commercials. Output lines consist of
5336 the frame number of the detected frame, the percentage of blackness,
5337 the position in the file if known or -1 and the timestamp in seconds.
5339 In order to display the output lines, you need to set the loglevel at
5340 least to the AV_LOG_INFO value.
5342 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5343 The value represents the percentage of pixels in the picture that
5344 are below the threshold value.
5346 It accepts the following parameters:
5351 The percentage of the pixels that have to be below the threshold; it defaults to
5354 @item threshold, thresh
5355 The threshold below which a pixel value is considered black; it defaults to
5360 @section blend, tblend
5362 Blend two video frames into each other.
5364 The @code{blend} filter takes two input streams and outputs one
5365 stream, the first input is the "top" layer and second input is
5366 "bottom" layer. By default, the output terminates when the longest input terminates.
5368 The @code{tblend} (time blend) filter takes two consecutive frames
5369 from one single stream, and outputs the result obtained by blending
5370 the new frame on top of the old frame.
5372 A description of the accepted options follows.
5380 Set blend mode for specific pixel component or all pixel components in case
5381 of @var{all_mode}. Default value is @code{normal}.
5383 Available values for component modes are:
5425 Set blend opacity for specific pixel component or all pixel components in case
5426 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5433 Set blend expression for specific pixel component or all pixel components in case
5434 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5436 The expressions can use the following variables:
5440 The sequential number of the filtered frame, starting from @code{0}.
5444 the coordinates of the current sample
5448 the width and height of currently filtered plane
5452 Width and height scale depending on the currently filtered plane. It is the
5453 ratio between the corresponding luma plane number of pixels and the current
5454 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5455 @code{0.5,0.5} for chroma planes.
5458 Time of the current frame, expressed in seconds.
5461 Value of pixel component at current location for first video frame (top layer).
5464 Value of pixel component at current location for second video frame (bottom layer).
5468 The @code{blend} filter also supports the @ref{framesync} options.
5470 @subsection Examples
5474 Apply transition from bottom layer to top layer in first 10 seconds:
5476 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5480 Apply linear horizontal transition from top layer to bottom layer:
5482 blend=all_expr='A*(X/W)+B*(1-X/W)'
5486 Apply 1x1 checkerboard effect:
5488 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5492 Apply uncover left effect:
5494 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5498 Apply uncover down effect:
5500 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5504 Apply uncover up-left effect:
5506 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5510 Split diagonally video and shows top and bottom layer on each side:
5512 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5516 Display differences between the current and the previous frame:
5518 tblend=all_mode=grainextract
5524 Apply a boxblur algorithm to the input video.
5526 It accepts the following parameters:
5530 @item luma_radius, lr
5531 @item luma_power, lp
5532 @item chroma_radius, cr
5533 @item chroma_power, cp
5534 @item alpha_radius, ar
5535 @item alpha_power, ap
5539 A description of the accepted options follows.
5542 @item luma_radius, lr
5543 @item chroma_radius, cr
5544 @item alpha_radius, ar
5545 Set an expression for the box radius in pixels used for blurring the
5546 corresponding input plane.
5548 The radius value must be a non-negative number, and must not be
5549 greater than the value of the expression @code{min(w,h)/2} for the
5550 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5553 Default value for @option{luma_radius} is "2". If not specified,
5554 @option{chroma_radius} and @option{alpha_radius} default to the
5555 corresponding value set for @option{luma_radius}.
5557 The expressions can contain the following constants:
5561 The input width and height in pixels.
5565 The input chroma image width and height in pixels.
5569 The horizontal and vertical chroma subsample values. For example, for the
5570 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5573 @item luma_power, lp
5574 @item chroma_power, cp
5575 @item alpha_power, ap
5576 Specify how many times the boxblur filter is applied to the
5577 corresponding plane.
5579 Default value for @option{luma_power} is 2. If not specified,
5580 @option{chroma_power} and @option{alpha_power} default to the
5581 corresponding value set for @option{luma_power}.
5583 A value of 0 will disable the effect.
5586 @subsection Examples
5590 Apply a boxblur filter with the luma, chroma, and alpha radii
5593 boxblur=luma_radius=2:luma_power=1
5598 Set the luma radius to 2, and alpha and chroma radius to 0:
5600 boxblur=2:1:cr=0:ar=0
5604 Set the luma and chroma radii to a fraction of the video dimension:
5606 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5612 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5613 Deinterlacing Filter").
5615 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5616 interpolation algorithms.
5617 It accepts the following parameters:
5621 The interlacing mode to adopt. It accepts one of the following values:
5625 Output one frame for each frame.
5627 Output one frame for each field.
5630 The default value is @code{send_field}.
5633 The picture field parity assumed for the input interlaced video. It accepts one
5634 of the following values:
5638 Assume the top field is first.
5640 Assume the bottom field is first.
5642 Enable automatic detection of field parity.
5645 The default value is @code{auto}.
5646 If the interlacing is unknown or the decoder does not export this information,
5647 top field first will be assumed.
5650 Specify which frames to deinterlace. Accept one of the following
5655 Deinterlace all frames.
5657 Only deinterlace frames marked as interlaced.
5660 The default value is @code{all}.
5664 YUV colorspace color/chroma keying.
5666 The filter accepts the following options:
5670 The color which will be replaced with transparency.
5673 Similarity percentage with the key color.
5675 0.01 matches only the exact key color, while 1.0 matches everything.
5680 0.0 makes pixels either fully transparent, or not transparent at all.
5682 Higher values result in semi-transparent pixels, with a higher transparency
5683 the more similar the pixels color is to the key color.
5686 Signals that the color passed is already in YUV instead of RGB.
5688 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5689 This can be used to pass exact YUV values as hexadecimal numbers.
5692 @subsection Examples
5696 Make every green pixel in the input image transparent:
5698 ffmpeg -i input.png -vf chromakey=green out.png
5702 Overlay a greenscreen-video on top of a static black background.
5704 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
5710 Display CIE color diagram with pixels overlaid onto it.
5712 The filter accepts the following options:
5727 @item uhdtv, rec2020
5740 Set what gamuts to draw.
5742 See @code{system} option for available values.
5745 Set ciescope size, by default set to 512.
5748 Set intensity used to map input pixel values to CIE diagram.
5751 Set contrast used to draw tongue colors that are out of active color system gamut.
5754 Correct gamma displayed on scope, by default enabled.
5757 Show white point on CIE diagram, by default disabled.
5760 Set input gamma. Used only with XYZ input color space.
5765 Visualize information exported by some codecs.
5767 Some codecs can export information through frames using side-data or other
5768 means. For example, some MPEG based codecs export motion vectors through the
5769 @var{export_mvs} flag in the codec @option{flags2} option.
5771 The filter accepts the following option:
5775 Set motion vectors to visualize.
5777 Available flags for @var{mv} are:
5781 forward predicted MVs of P-frames
5783 forward predicted MVs of B-frames
5785 backward predicted MVs of B-frames
5789 Display quantization parameters using the chroma planes.
5792 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5794 Available flags for @var{mv_type} are:
5798 forward predicted MVs
5800 backward predicted MVs
5803 @item frame_type, ft
5804 Set frame type to visualize motion vectors of.
5806 Available flags for @var{frame_type} are:
5810 intra-coded frames (I-frames)
5812 predicted frames (P-frames)
5814 bi-directionally predicted frames (B-frames)
5818 @subsection Examples
5822 Visualize forward predicted MVs of all frames using @command{ffplay}:
5824 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5828 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5830 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5834 @section colorbalance
5835 Modify intensity of primary colors (red, green and blue) of input frames.
5837 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5838 regions for the red-cyan, green-magenta or blue-yellow balance.
5840 A positive adjustment value shifts the balance towards the primary color, a negative
5841 value towards the complementary color.
5843 The filter accepts the following options:
5849 Adjust red, green and blue shadows (darkest pixels).
5854 Adjust red, green and blue midtones (medium pixels).
5859 Adjust red, green and blue highlights (brightest pixels).
5861 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5864 @subsection Examples
5868 Add red color cast to shadows:
5875 RGB colorspace color keying.
5877 The filter accepts the following options:
5881 The color which will be replaced with transparency.
5884 Similarity percentage with the key color.
5886 0.01 matches only the exact key color, while 1.0 matches everything.
5891 0.0 makes pixels either fully transparent, or not transparent at all.
5893 Higher values result in semi-transparent pixels, with a higher transparency
5894 the more similar the pixels color is to the key color.
5897 @subsection Examples
5901 Make every green pixel in the input image transparent:
5903 ffmpeg -i input.png -vf colorkey=green out.png
5907 Overlay a greenscreen-video on top of a static background image.
5909 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
5913 @section colorlevels
5915 Adjust video input frames using levels.
5917 The filter accepts the following options:
5924 Adjust red, green, blue and alpha input black point.
5925 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5931 Adjust red, green, blue and alpha input white point.
5932 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5934 Input levels are used to lighten highlights (bright tones), darken shadows
5935 (dark tones), change the balance of bright and dark tones.
5941 Adjust red, green, blue and alpha output black point.
5942 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5948 Adjust red, green, blue and alpha output white point.
5949 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5951 Output levels allows manual selection of a constrained output level range.
5954 @subsection Examples
5958 Make video output darker:
5960 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5966 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5970 Make video output lighter:
5972 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5976 Increase brightness:
5978 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5982 @section colorchannelmixer
5984 Adjust video input frames by re-mixing color channels.
5986 This filter modifies a color channel by adding the values associated to
5987 the other channels of the same pixels. For example if the value to
5988 modify is red, the output value will be:
5990 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5993 The filter accepts the following options:
6000 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6001 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6007 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6008 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6014 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6015 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6021 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6022 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6024 Allowed ranges for options are @code{[-2.0, 2.0]}.
6027 @subsection Examples
6031 Convert source to grayscale:
6033 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6036 Simulate sepia tones:
6038 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6042 @section colormatrix
6044 Convert color matrix.
6046 The filter accepts the following options:
6051 Specify the source and destination color matrix. Both values must be
6054 The accepted values are:
6082 For example to convert from BT.601 to SMPTE-240M, use the command:
6084 colormatrix=bt601:smpte240m
6089 Convert colorspace, transfer characteristics or color primaries.
6090 Input video needs to have an even size.
6092 The filter accepts the following options:
6097 Specify all color properties at once.
6099 The accepted values are:
6129 Specify output colorspace.
6131 The accepted values are:
6140 BT.470BG or BT.601-6 625
6143 SMPTE-170M or BT.601-6 525
6152 BT.2020 with non-constant luminance
6158 Specify output transfer characteristics.
6160 The accepted values are:
6172 Constant gamma of 2.2
6175 Constant gamma of 2.8
6178 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6196 BT.2020 for 10-bits content
6199 BT.2020 for 12-bits content
6205 Specify output color primaries.
6207 The accepted values are:
6216 BT.470BG or BT.601-6 625
6219 SMPTE-170M or BT.601-6 525
6243 Specify output color range.
6245 The accepted values are:
6248 TV (restricted) range
6251 MPEG (restricted) range
6262 Specify output color format.
6264 The accepted values are:
6267 YUV 4:2:0 planar 8-bits
6270 YUV 4:2:0 planar 10-bits
6273 YUV 4:2:0 planar 12-bits
6276 YUV 4:2:2 planar 8-bits
6279 YUV 4:2:2 planar 10-bits
6282 YUV 4:2:2 planar 12-bits
6285 YUV 4:4:4 planar 8-bits
6288 YUV 4:4:4 planar 10-bits
6291 YUV 4:4:4 planar 12-bits
6296 Do a fast conversion, which skips gamma/primary correction. This will take
6297 significantly less CPU, but will be mathematically incorrect. To get output
6298 compatible with that produced by the colormatrix filter, use fast=1.
6301 Specify dithering mode.
6303 The accepted values are:
6309 Floyd-Steinberg dithering
6313 Whitepoint adaptation mode.
6315 The accepted values are:
6318 Bradford whitepoint adaptation
6321 von Kries whitepoint adaptation
6324 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6328 Override all input properties at once. Same accepted values as @ref{all}.
6331 Override input colorspace. Same accepted values as @ref{space}.
6334 Override input color primaries. Same accepted values as @ref{primaries}.
6337 Override input transfer characteristics. Same accepted values as @ref{trc}.
6340 Override input color range. Same accepted values as @ref{range}.
6344 The filter converts the transfer characteristics, color space and color
6345 primaries to the specified user values. The output value, if not specified,
6346 is set to a default value based on the "all" property. If that property is
6347 also not specified, the filter will log an error. The output color range and
6348 format default to the same value as the input color range and format. The
6349 input transfer characteristics, color space, color primaries and color range
6350 should be set on the input data. If any of these are missing, the filter will
6351 log an error and no conversion will take place.
6353 For example to convert the input to SMPTE-240M, use the command:
6355 colorspace=smpte240m
6358 @section convolution
6360 Apply convolution 3x3, 5x5 or 7x7 filter.
6362 The filter accepts the following options:
6369 Set matrix for each plane.
6370 Matrix is sequence of 9, 25 or 49 signed integers.
6376 Set multiplier for calculated value for each plane.
6382 Set bias for each plane. This value is added to the result of the multiplication.
6383 Useful for making the overall image brighter or darker. Default is 0.0.
6386 @subsection Examples
6392 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"
6398 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"
6404 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"
6410 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"
6414 Apply laplacian edge detector which includes diagonals:
6416 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"
6422 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"
6428 Apply 2D convolution of video stream in frequency domain using second stream
6431 The filter accepts the following options:
6435 Set which planes to process.
6438 Set which impulse video frames will be processed, can be @var{first}
6439 or @var{all}. Default is @var{all}.
6442 The @code{convolve} filter also supports the @ref{framesync} options.
6446 Copy the input video source unchanged to the output. This is mainly useful for
6451 Video filtering on GPU using Apple's CoreImage API on OSX.
6453 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6454 processed by video hardware. However, software-based OpenGL implementations
6455 exist which means there is no guarantee for hardware processing. It depends on
6458 There are many filters and image generators provided by Apple that come with a
6459 large variety of options. The filter has to be referenced by its name along
6462 The coreimage filter accepts the following options:
6465 List all available filters and generators along with all their respective
6466 options as well as possible minimum and maximum values along with the default
6473 Specify all filters by their respective name and options.
6474 Use @var{list_filters} to determine all valid filter names and options.
6475 Numerical options are specified by a float value and are automatically clamped
6476 to their respective value range. Vector and color options have to be specified
6477 by a list of space separated float values. Character escaping has to be done.
6478 A special option name @code{default} is available to use default options for a
6481 It is required to specify either @code{default} or at least one of the filter options.
6482 All omitted options are used with their default values.
6483 The syntax of the filter string is as follows:
6485 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6489 Specify a rectangle where the output of the filter chain is copied into the
6490 input image. It is given by a list of space separated float values:
6492 output_rect=x\ y\ width\ height
6494 If not given, the output rectangle equals the dimensions of the input image.
6495 The output rectangle is automatically cropped at the borders of the input
6496 image. Negative values are valid for each component.
6498 output_rect=25\ 25\ 100\ 100
6502 Several filters can be chained for successive processing without GPU-HOST
6503 transfers allowing for fast processing of complex filter chains.
6504 Currently, only filters with zero (generators) or exactly one (filters) input
6505 image and one output image are supported. Also, transition filters are not yet
6508 Some filters generate output images with additional padding depending on the
6509 respective filter kernel. The padding is automatically removed to ensure the
6510 filter output has the same size as the input image.
6512 For image generators, the size of the output image is determined by the
6513 previous output image of the filter chain or the input image of the whole
6514 filterchain, respectively. The generators do not use the pixel information of
6515 this image to generate their output. However, the generated output is
6516 blended onto this image, resulting in partial or complete coverage of the
6519 The @ref{coreimagesrc} video source can be used for generating input images
6520 which are directly fed into the filter chain. By using it, providing input
6521 images by another video source or an input video is not required.
6523 @subsection Examples
6528 List all filters available:
6530 coreimage=list_filters=true
6534 Use the CIBoxBlur filter with default options to blur an image:
6536 coreimage=filter=CIBoxBlur@@default
6540 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6541 its center at 100x100 and a radius of 50 pixels:
6543 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6547 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6548 given as complete and escaped command-line for Apple's standard bash shell:
6550 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6556 Crop the input video to given dimensions.
6558 It accepts the following parameters:
6562 The width of the output video. It defaults to @code{iw}.
6563 This expression is evaluated only once during the filter
6564 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6567 The height of the output video. It defaults to @code{ih}.
6568 This expression is evaluated only once during the filter
6569 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6572 The horizontal position, in the input video, of the left edge of the output
6573 video. It defaults to @code{(in_w-out_w)/2}.
6574 This expression is evaluated per-frame.
6577 The vertical position, in the input video, of the top edge of the output video.
6578 It defaults to @code{(in_h-out_h)/2}.
6579 This expression is evaluated per-frame.
6582 If set to 1 will force the output display aspect ratio
6583 to be the same of the input, by changing the output sample aspect
6584 ratio. It defaults to 0.
6587 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6588 width/height/x/y as specified and will not be rounded to nearest smaller value.
6592 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6593 expressions containing the following constants:
6598 The computed values for @var{x} and @var{y}. They are evaluated for
6603 The input width and height.
6607 These are the same as @var{in_w} and @var{in_h}.
6611 The output (cropped) width and height.
6615 These are the same as @var{out_w} and @var{out_h}.
6618 same as @var{iw} / @var{ih}
6621 input sample aspect ratio
6624 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6628 horizontal and vertical chroma subsample values. For example for the
6629 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6632 The number of the input frame, starting from 0.
6635 the position in the file of the input frame, NAN if unknown
6638 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6642 The expression for @var{out_w} may depend on the value of @var{out_h},
6643 and the expression for @var{out_h} may depend on @var{out_w}, but they
6644 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6645 evaluated after @var{out_w} and @var{out_h}.
6647 The @var{x} and @var{y} parameters specify the expressions for the
6648 position of the top-left corner of the output (non-cropped) area. They
6649 are evaluated for each frame. If the evaluated value is not valid, it
6650 is approximated to the nearest valid value.
6652 The expression for @var{x} may depend on @var{y}, and the expression
6653 for @var{y} may depend on @var{x}.
6655 @subsection Examples
6659 Crop area with size 100x100 at position (12,34).
6664 Using named options, the example above becomes:
6666 crop=w=100:h=100:x=12:y=34
6670 Crop the central input area with size 100x100:
6676 Crop the central input area with size 2/3 of the input video:
6678 crop=2/3*in_w:2/3*in_h
6682 Crop the input video central square:
6689 Delimit the rectangle with the top-left corner placed at position
6690 100:100 and the right-bottom corner corresponding to the right-bottom
6691 corner of the input image.
6693 crop=in_w-100:in_h-100:100:100
6697 Crop 10 pixels from the left and right borders, and 20 pixels from
6698 the top and bottom borders
6700 crop=in_w-2*10:in_h-2*20
6704 Keep only the bottom right quarter of the input image:
6706 crop=in_w/2:in_h/2:in_w/2:in_h/2
6710 Crop height for getting Greek harmony:
6712 crop=in_w:1/PHI*in_w
6716 Apply trembling effect:
6718 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)
6722 Apply erratic camera effect depending on timestamp:
6724 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)"
6728 Set x depending on the value of y:
6730 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6734 @subsection Commands
6736 This filter supports the following commands:
6742 Set width/height of the output video and the horizontal/vertical position
6744 The command accepts the same syntax of the corresponding option.
6746 If the specified expression is not valid, it is kept at its current
6752 Auto-detect the crop size.
6754 It calculates the necessary cropping parameters and prints the
6755 recommended parameters via the logging system. The detected dimensions
6756 correspond to the non-black area of the input video.
6758 It accepts the following parameters:
6763 Set higher black value threshold, which can be optionally specified
6764 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6765 value greater to the set value is considered non-black. It defaults to 24.
6766 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6767 on the bitdepth of the pixel format.
6770 The value which the width/height should be divisible by. It defaults to
6771 16. The offset is automatically adjusted to center the video. Use 2 to
6772 get only even dimensions (needed for 4:2:2 video). 16 is best when
6773 encoding to most video codecs.
6775 @item reset_count, reset
6776 Set the counter that determines after how many frames cropdetect will
6777 reset the previously detected largest video area and start over to
6778 detect the current optimal crop area. Default value is 0.
6780 This can be useful when channel logos distort the video area. 0
6781 indicates 'never reset', and returns the largest area encountered during
6788 Apply color adjustments using curves.
6790 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6791 component (red, green and blue) has its values defined by @var{N} key points
6792 tied from each other using a smooth curve. The x-axis represents the pixel
6793 values from the input frame, and the y-axis the new pixel values to be set for
6796 By default, a component curve is defined by the two points @var{(0;0)} and
6797 @var{(1;1)}. This creates a straight line where each original pixel value is
6798 "adjusted" to its own value, which means no change to the image.
6800 The filter allows you to redefine these two points and add some more. A new
6801 curve (using a natural cubic spline interpolation) will be define to pass
6802 smoothly through all these new coordinates. The new defined points needs to be
6803 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6804 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6805 the vector spaces, the values will be clipped accordingly.
6807 The filter accepts the following options:
6811 Select one of the available color presets. This option can be used in addition
6812 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6813 options takes priority on the preset values.
6814 Available presets are:
6817 @item color_negative
6820 @item increase_contrast
6822 @item linear_contrast
6823 @item medium_contrast
6825 @item strong_contrast
6828 Default is @code{none}.
6830 Set the master key points. These points will define a second pass mapping. It
6831 is sometimes called a "luminance" or "value" mapping. It can be used with
6832 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6833 post-processing LUT.
6835 Set the key points for the red component.
6837 Set the key points for the green component.
6839 Set the key points for the blue component.
6841 Set the key points for all components (not including master).
6842 Can be used in addition to the other key points component
6843 options. In this case, the unset component(s) will fallback on this
6844 @option{all} setting.
6846 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6848 Save Gnuplot script of the curves in specified file.
6851 To avoid some filtergraph syntax conflicts, each key points list need to be
6852 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6854 @subsection Examples
6858 Increase slightly the middle level of blue:
6860 curves=blue='0/0 0.5/0.58 1/1'
6866 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'
6868 Here we obtain the following coordinates for each components:
6871 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6873 @code{(0;0) (0.50;0.48) (1;1)}
6875 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6879 The previous example can also be achieved with the associated built-in preset:
6881 curves=preset=vintage
6891 Use a Photoshop preset and redefine the points of the green component:
6893 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6897 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6898 and @command{gnuplot}:
6900 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6901 gnuplot -p /tmp/curves.plt
6907 Video data analysis filter.
6909 This filter shows hexadecimal pixel values of part of video.
6911 The filter accepts the following options:
6915 Set output video size.
6918 Set x offset from where to pick pixels.
6921 Set y offset from where to pick pixels.
6924 Set scope mode, can be one of the following:
6927 Draw hexadecimal pixel values with white color on black background.
6930 Draw hexadecimal pixel values with input video pixel color on black
6934 Draw hexadecimal pixel values on color background picked from input video,
6935 the text color is picked in such way so its always visible.
6939 Draw rows and columns numbers on left and top of video.
6942 Set background opacity.
6947 Denoise frames using 2D DCT (frequency domain filtering).
6949 This filter is not designed for real time.
6951 The filter accepts the following options:
6955 Set the noise sigma constant.
6957 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6958 coefficient (absolute value) below this threshold with be dropped.
6960 If you need a more advanced filtering, see @option{expr}.
6962 Default is @code{0}.
6965 Set number overlapping pixels for each block. Since the filter can be slow, you
6966 may want to reduce this value, at the cost of a less effective filter and the
6967 risk of various artefacts.
6969 If the overlapping value doesn't permit processing the whole input width or
6970 height, a warning will be displayed and according borders won't be denoised.
6972 Default value is @var{blocksize}-1, which is the best possible setting.
6975 Set the coefficient factor expression.
6977 For each coefficient of a DCT block, this expression will be evaluated as a
6978 multiplier value for the coefficient.
6980 If this is option is set, the @option{sigma} option will be ignored.
6982 The absolute value of the coefficient can be accessed through the @var{c}
6986 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6987 @var{blocksize}, which is the width and height of the processed blocks.
6989 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6990 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6991 on the speed processing. Also, a larger block size does not necessarily means a
6995 @subsection Examples
6997 Apply a denoise with a @option{sigma} of @code{4.5}:
7002 The same operation can be achieved using the expression system:
7004 dctdnoiz=e='gte(c, 4.5*3)'
7007 Violent denoise using a block size of @code{16x16}:
7014 Remove banding artifacts from input video.
7015 It works by replacing banded pixels with average value of referenced pixels.
7017 The filter accepts the following options:
7024 Set banding detection threshold for each plane. Default is 0.02.
7025 Valid range is 0.00003 to 0.5.
7026 If difference between current pixel and reference pixel is less than threshold,
7027 it will be considered as banded.
7030 Banding detection range in pixels. Default is 16. If positive, random number
7031 in range 0 to set value will be used. If negative, exact absolute value
7033 The range defines square of four pixels around current pixel.
7036 Set direction in radians from which four pixel will be compared. If positive,
7037 random direction from 0 to set direction will be picked. If negative, exact of
7038 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7039 will pick only pixels on same row and -PI/2 will pick only pixels on same
7043 If enabled, current pixel is compared with average value of all four
7044 surrounding pixels. The default is enabled. If disabled current pixel is
7045 compared with all four surrounding pixels. The pixel is considered banded
7046 if only all four differences with surrounding pixels are less than threshold.
7049 If enabled, current pixel is changed if and only if all pixel components are banded,
7050 e.g. banding detection threshold is triggered for all color components.
7051 The default is disabled.
7057 Drop duplicated frames at regular intervals.
7059 The filter accepts the following options:
7063 Set the number of frames from which one will be dropped. Setting this to
7064 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7065 Default is @code{5}.
7068 Set the threshold for duplicate detection. If the difference metric for a frame
7069 is less than or equal to this value, then it is declared as duplicate. Default
7073 Set scene change threshold. Default is @code{15}.
7077 Set the size of the x and y-axis blocks used during metric calculations.
7078 Larger blocks give better noise suppression, but also give worse detection of
7079 small movements. Must be a power of two. Default is @code{32}.
7082 Mark main input as a pre-processed input and activate clean source input
7083 stream. This allows the input to be pre-processed with various filters to help
7084 the metrics calculation while keeping the frame selection lossless. When set to
7085 @code{1}, the first stream is for the pre-processed input, and the second
7086 stream is the clean source from where the kept frames are chosen. Default is
7090 Set whether or not chroma is considered in the metric calculations. Default is
7096 Apply 2D deconvolution of video stream in frequency domain using second stream
7099 The filter accepts the following options:
7103 Set which planes to process.
7106 Set which impulse video frames will be processed, can be @var{first}
7107 or @var{all}. Default is @var{all}.
7110 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7111 and height are not same and not power of 2 or if stream prior to convolving
7115 The @code{deconvolve} filter also supports the @ref{framesync} options.
7119 Apply deflate effect to the video.
7121 This filter replaces the pixel by the local(3x3) average by taking into account
7122 only values lower than the pixel.
7124 It accepts the following options:
7131 Limit the maximum change for each plane, default is 65535.
7132 If 0, plane will remain unchanged.
7137 Remove temporal frame luminance variations.
7139 It accepts the following options:
7143 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7146 Set averaging mode to smooth temporal luminance variations.
7148 Available values are:
7173 Do not actually modify frame. Useful when one only wants metadata.
7178 Remove judder produced by partially interlaced telecined content.
7180 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7181 source was partially telecined content then the output of @code{pullup,dejudder}
7182 will have a variable frame rate. May change the recorded frame rate of the
7183 container. Aside from that change, this filter will not affect constant frame
7186 The option available in this filter is:
7190 Specify the length of the window over which the judder repeats.
7192 Accepts any integer greater than 1. Useful values are:
7196 If the original was telecined from 24 to 30 fps (Film to NTSC).
7199 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7202 If a mixture of the two.
7205 The default is @samp{4}.
7210 Suppress a TV station logo by a simple interpolation of the surrounding
7211 pixels. Just set a rectangle covering the logo and watch it disappear
7212 (and sometimes something even uglier appear - your mileage may vary).
7214 It accepts the following parameters:
7219 Specify the top left corner coordinates of the logo. They must be
7224 Specify the width and height of the logo to clear. They must be
7228 Specify the thickness of the fuzzy edge of the rectangle (added to
7229 @var{w} and @var{h}). The default value is 1. This option is
7230 deprecated, setting higher values should no longer be necessary and
7234 When set to 1, a green rectangle is drawn on the screen to simplify
7235 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7236 The default value is 0.
7238 The rectangle is drawn on the outermost pixels which will be (partly)
7239 replaced with interpolated values. The values of the next pixels
7240 immediately outside this rectangle in each direction will be used to
7241 compute the interpolated pixel values inside the rectangle.
7245 @subsection Examples
7249 Set a rectangle covering the area with top left corner coordinates 0,0
7250 and size 100x77, and a band of size 10:
7252 delogo=x=0:y=0:w=100:h=77:band=10
7259 Attempt to fix small changes in horizontal and/or vertical shift. This
7260 filter helps remove camera shake from hand-holding a camera, bumping a
7261 tripod, moving on a vehicle, etc.
7263 The filter accepts the following options:
7271 Specify a rectangular area where to limit the search for motion
7273 If desired the search for motion vectors can be limited to a
7274 rectangular area of the frame defined by its top left corner, width
7275 and height. These parameters have the same meaning as the drawbox
7276 filter which can be used to visualise the position of the bounding
7279 This is useful when simultaneous movement of subjects within the frame
7280 might be confused for camera motion by the motion vector search.
7282 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7283 then the full frame is used. This allows later options to be set
7284 without specifying the bounding box for the motion vector search.
7286 Default - search the whole frame.
7290 Specify the maximum extent of movement in x and y directions in the
7291 range 0-64 pixels. Default 16.
7294 Specify how to generate pixels to fill blanks at the edge of the
7295 frame. Available values are:
7298 Fill zeroes at blank locations
7300 Original image at blank locations
7302 Extruded edge value at blank locations
7304 Mirrored edge at blank locations
7306 Default value is @samp{mirror}.
7309 Specify the blocksize to use for motion search. Range 4-128 pixels,
7313 Specify the contrast threshold for blocks. Only blocks with more than
7314 the specified contrast (difference between darkest and lightest
7315 pixels) will be considered. Range 1-255, default 125.
7318 Specify the search strategy. Available values are:
7321 Set exhaustive search
7323 Set less exhaustive search.
7325 Default value is @samp{exhaustive}.
7328 If set then a detailed log of the motion search is written to the
7335 Remove unwanted contamination of foreground colors, caused by reflected color of
7336 greenscreen or bluescreen.
7338 This filter accepts the following options:
7342 Set what type of despill to use.
7345 Set how spillmap will be generated.
7348 Set how much to get rid of still remaining spill.
7351 Controls amount of red in spill area.
7354 Controls amount of green in spill area.
7355 Should be -1 for greenscreen.
7358 Controls amount of blue in spill area.
7359 Should be -1 for bluescreen.
7362 Controls brightness of spill area, preserving colors.
7365 Modify alpha from generated spillmap.
7370 Apply an exact inverse of the telecine operation. It requires a predefined
7371 pattern specified using the pattern option which must be the same as that passed
7372 to the telecine filter.
7374 This filter accepts the following options:
7383 The default value is @code{top}.
7387 A string of numbers representing the pulldown pattern you wish to apply.
7388 The default value is @code{23}.
7391 A number representing position of the first frame with respect to the telecine
7392 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7397 Apply dilation effect to the video.
7399 This filter replaces the pixel by the local(3x3) maximum.
7401 It accepts the following options:
7408 Limit the maximum change for each plane, default is 65535.
7409 If 0, plane will remain unchanged.
7412 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7415 Flags to local 3x3 coordinates maps like this:
7424 Displace pixels as indicated by second and third input stream.
7426 It takes three input streams and outputs one stream, the first input is the
7427 source, and second and third input are displacement maps.
7429 The second input specifies how much to displace pixels along the
7430 x-axis, while the third input specifies how much to displace pixels
7432 If one of displacement map streams terminates, last frame from that
7433 displacement map will be used.
7435 Note that once generated, displacements maps can be reused over and over again.
7437 A description of the accepted options follows.
7441 Set displace behavior for pixels that are out of range.
7443 Available values are:
7446 Missing pixels are replaced by black pixels.
7449 Adjacent pixels will spread out to replace missing pixels.
7452 Out of range pixels are wrapped so they point to pixels of other side.
7455 Out of range pixels will be replaced with mirrored pixels.
7457 Default is @samp{smear}.
7461 @subsection Examples
7465 Add ripple effect to rgb input of video size hd720:
7467 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
7471 Add wave effect to rgb input of video size hd720:
7473 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
7479 Draw a colored box on the input image.
7481 It accepts the following parameters:
7486 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7490 The expressions which specify the width and height of the box; if 0 they are interpreted as
7491 the input width and height. It defaults to 0.
7494 Specify the color of the box to write. For the general syntax of this option,
7495 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7496 value @code{invert} is used, the box edge color is the same as the
7497 video with inverted luma.
7500 The expression which sets the thickness of the box edge.
7501 A value of @code{fill} will create a filled box. Default value is @code{3}.
7503 See below for the list of accepted constants.
7506 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
7507 will overwrite the video's color and alpha pixels.
7508 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
7511 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7512 following constants:
7516 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7520 horizontal and vertical chroma subsample values. For example for the
7521 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7525 The input width and height.
7528 The input sample aspect ratio.
7532 The x and y offset coordinates where the box is drawn.
7536 The width and height of the drawn box.
7539 The thickness of the drawn box.
7541 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7542 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7546 @subsection Examples
7550 Draw a black box around the edge of the input image:
7556 Draw a box with color red and an opacity of 50%:
7558 drawbox=10:20:200:60:red@@0.5
7561 The previous example can be specified as:
7563 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7567 Fill the box with pink color:
7569 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
7573 Draw a 2-pixel red 2.40:1 mask:
7575 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
7581 Draw a grid on the input image.
7583 It accepts the following parameters:
7588 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7592 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7593 input width and height, respectively, minus @code{thickness}, so image gets
7594 framed. Default to 0.
7597 Specify the color of the grid. For the general syntax of this option,
7598 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7599 value @code{invert} is used, the grid color is the same as the
7600 video with inverted luma.
7603 The expression which sets the thickness of the grid line. Default value is @code{1}.
7605 See below for the list of accepted constants.
7608 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
7609 will overwrite the video's color and alpha pixels.
7610 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
7613 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7614 following constants:
7618 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7622 horizontal and vertical chroma subsample values. For example for the
7623 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7627 The input grid cell width and height.
7630 The input sample aspect ratio.
7634 The x and y coordinates of some point of grid intersection (meant to configure offset).
7638 The width and height of the drawn cell.
7641 The thickness of the drawn cell.
7643 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7644 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7648 @subsection Examples
7652 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7654 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7658 Draw a white 3x3 grid with an opacity of 50%:
7660 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7667 Draw a text string or text from a specified file on top of a video, using the
7668 libfreetype library.
7670 To enable compilation of this filter, you need to configure FFmpeg with
7671 @code{--enable-libfreetype}.
7672 To enable default font fallback and the @var{font} option you need to
7673 configure FFmpeg with @code{--enable-libfontconfig}.
7674 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7675 @code{--enable-libfribidi}.
7679 It accepts the following parameters:
7684 Used to draw a box around text using the background color.
7685 The value must be either 1 (enable) or 0 (disable).
7686 The default value of @var{box} is 0.
7689 Set the width of the border to be drawn around the box using @var{boxcolor}.
7690 The default value of @var{boxborderw} is 0.
7693 The color to be used for drawing box around text. For the syntax of this
7694 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7696 The default value of @var{boxcolor} is "white".
7699 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7700 The default value of @var{line_spacing} is 0.
7703 Set the width of the border to be drawn around the text using @var{bordercolor}.
7704 The default value of @var{borderw} is 0.
7707 Set the color to be used for drawing border around text. For the syntax of this
7708 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7710 The default value of @var{bordercolor} is "black".
7713 Select how the @var{text} is expanded. Can be either @code{none},
7714 @code{strftime} (deprecated) or
7715 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7719 Set a start time for the count. Value is in microseconds. Only applied
7720 in the deprecated strftime expansion mode. To emulate in normal expansion
7721 mode use the @code{pts} function, supplying the start time (in seconds)
7722 as the second argument.
7725 If true, check and fix text coords to avoid clipping.
7728 The color to be used for drawing fonts. For the syntax of this option, check
7729 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7731 The default value of @var{fontcolor} is "black".
7733 @item fontcolor_expr
7734 String which is expanded the same way as @var{text} to obtain dynamic
7735 @var{fontcolor} value. By default this option has empty value and is not
7736 processed. When this option is set, it overrides @var{fontcolor} option.
7739 The font family to be used for drawing text. By default Sans.
7742 The font file to be used for drawing text. The path must be included.
7743 This parameter is mandatory if the fontconfig support is disabled.
7746 Draw the text applying alpha blending. The value can
7747 be a number between 0.0 and 1.0.
7748 The expression accepts the same variables @var{x, y} as well.
7749 The default value is 1.
7750 Please see @var{fontcolor_expr}.
7753 The font size to be used for drawing text.
7754 The default value of @var{fontsize} is 16.
7757 If set to 1, attempt to shape the text (for example, reverse the order of
7758 right-to-left text and join Arabic characters) before drawing it.
7759 Otherwise, just draw the text exactly as given.
7760 By default 1 (if supported).
7763 The flags to be used for loading the fonts.
7765 The flags map the corresponding flags supported by libfreetype, and are
7766 a combination of the following values:
7773 @item vertical_layout
7774 @item force_autohint
7777 @item ignore_global_advance_width
7779 @item ignore_transform
7785 Default value is "default".
7787 For more information consult the documentation for the FT_LOAD_*
7791 The color to be used for drawing a shadow behind the drawn text. For the
7792 syntax of this option, check the @ref{color syntax,,"Color" section in the
7793 ffmpeg-utils manual,ffmpeg-utils}.
7795 The default value of @var{shadowcolor} is "black".
7799 The x and y offsets for the text shadow position with respect to the
7800 position of the text. They can be either positive or negative
7801 values. The default value for both is "0".
7804 The starting frame number for the n/frame_num variable. The default value
7808 The size in number of spaces to use for rendering the tab.
7812 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7813 format. It can be used with or without text parameter. @var{timecode_rate}
7814 option must be specified.
7816 @item timecode_rate, rate, r
7817 Set the timecode frame rate (timecode only). Value will be rounded to nearest
7818 integer. Minimum value is "1".
7819 Drop-frame timecode is supported for frame rates 30 & 60.
7822 If set to 1, the output of the timecode option will wrap around at 24 hours.
7823 Default is 0 (disabled).
7826 The text string to be drawn. The text must be a sequence of UTF-8
7828 This parameter is mandatory if no file is specified with the parameter
7832 A text file containing text to be drawn. The text must be a sequence
7833 of UTF-8 encoded characters.
7835 This parameter is mandatory if no text string is specified with the
7836 parameter @var{text}.
7838 If both @var{text} and @var{textfile} are specified, an error is thrown.
7841 If set to 1, the @var{textfile} will be reloaded before each frame.
7842 Be sure to update it atomically, or it may be read partially, or even fail.
7846 The expressions which specify the offsets where text will be drawn
7847 within the video frame. They are relative to the top/left border of the
7850 The default value of @var{x} and @var{y} is "0".
7852 See below for the list of accepted constants and functions.
7855 The parameters for @var{x} and @var{y} are expressions containing the
7856 following constants and functions:
7860 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7864 horizontal and vertical chroma subsample values. For example for the
7865 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7868 the height of each text line
7876 @item max_glyph_a, ascent
7877 the maximum distance from the baseline to the highest/upper grid
7878 coordinate used to place a glyph outline point, for all the rendered
7880 It is a positive value, due to the grid's orientation with the Y axis
7883 @item max_glyph_d, descent
7884 the maximum distance from the baseline to the lowest grid coordinate
7885 used to place a glyph outline point, for all the rendered glyphs.
7886 This is a negative value, due to the grid's orientation, with the Y axis
7890 maximum glyph height, that is the maximum height for all the glyphs
7891 contained in the rendered text, it is equivalent to @var{ascent} -
7895 maximum glyph width, that is the maximum width for all the glyphs
7896 contained in the rendered text
7899 the number of input frame, starting from 0
7901 @item rand(min, max)
7902 return a random number included between @var{min} and @var{max}
7905 The input sample aspect ratio.
7908 timestamp expressed in seconds, NAN if the input timestamp is unknown
7911 the height of the rendered text
7914 the width of the rendered text
7918 the x and y offset coordinates where the text is drawn.
7920 These parameters allow the @var{x} and @var{y} expressions to refer
7921 each other, so you can for example specify @code{y=x/dar}.
7924 @anchor{drawtext_expansion}
7925 @subsection Text expansion
7927 If @option{expansion} is set to @code{strftime},
7928 the filter recognizes strftime() sequences in the provided text and
7929 expands them accordingly. Check the documentation of strftime(). This
7930 feature is deprecated.
7932 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7934 If @option{expansion} is set to @code{normal} (which is the default),
7935 the following expansion mechanism is used.
7937 The backslash character @samp{\}, followed by any character, always expands to
7938 the second character.
7940 Sequences of the form @code{%@{...@}} are expanded. The text between the
7941 braces is a function name, possibly followed by arguments separated by ':'.
7942 If the arguments contain special characters or delimiters (':' or '@}'),
7943 they should be escaped.
7945 Note that they probably must also be escaped as the value for the
7946 @option{text} option in the filter argument string and as the filter
7947 argument in the filtergraph description, and possibly also for the shell,
7948 that makes up to four levels of escaping; using a text file avoids these
7951 The following functions are available:
7956 The expression evaluation result.
7958 It must take one argument specifying the expression to be evaluated,
7959 which accepts the same constants and functions as the @var{x} and
7960 @var{y} values. Note that not all constants should be used, for
7961 example the text size is not known when evaluating the expression, so
7962 the constants @var{text_w} and @var{text_h} will have an undefined
7965 @item expr_int_format, eif
7966 Evaluate the expression's value and output as formatted integer.
7968 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7969 The second argument specifies the output format. Allowed values are @samp{x},
7970 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7971 @code{printf} function.
7972 The third parameter is optional and sets the number of positions taken by the output.
7973 It can be used to add padding with zeros from the left.
7976 The time at which the filter is running, expressed in UTC.
7977 It can accept an argument: a strftime() format string.
7980 The time at which the filter is running, expressed in the local time zone.
7981 It can accept an argument: a strftime() format string.
7984 Frame metadata. Takes one or two arguments.
7986 The first argument is mandatory and specifies the metadata key.
7988 The second argument is optional and specifies a default value, used when the
7989 metadata key is not found or empty.
7992 The frame number, starting from 0.
7995 A 1 character description of the current picture type.
7998 The timestamp of the current frame.
7999 It can take up to three arguments.
8001 The first argument is the format of the timestamp; it defaults to @code{flt}
8002 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8003 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8004 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8005 @code{localtime} stands for the timestamp of the frame formatted as
8006 local time zone time.
8008 The second argument is an offset added to the timestamp.
8010 If the format is set to @code{localtime} or @code{gmtime},
8011 a third argument may be supplied: a strftime() format string.
8012 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8015 @subsection Examples
8019 Draw "Test Text" with font FreeSerif, using the default values for the
8020 optional parameters.
8023 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8027 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8028 and y=50 (counting from the top-left corner of the screen), text is
8029 yellow with a red box around it. Both the text and the box have an
8033 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8034 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8037 Note that the double quotes are not necessary if spaces are not used
8038 within the parameter list.
8041 Show the text at the center of the video frame:
8043 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8047 Show the text at a random position, switching to a new position every 30 seconds:
8049 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)"
8053 Show a text line sliding from right to left in the last row of the video
8054 frame. The file @file{LONG_LINE} is assumed to contain a single line
8057 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8061 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8063 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8067 Draw a single green letter "g", at the center of the input video.
8068 The glyph baseline is placed at half screen height.
8070 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8074 Show text for 1 second every 3 seconds:
8076 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8080 Use fontconfig to set the font. Note that the colons need to be escaped.
8082 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8086 Print the date of a real-time encoding (see strftime(3)):
8088 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8092 Show text fading in and out (appearing/disappearing):
8095 DS=1.0 # display start
8096 DE=10.0 # display end
8097 FID=1.5 # fade in duration
8098 FOD=5 # fade out duration
8099 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 @}"
8103 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8104 and the @option{fontsize} value are included in the @option{y} offset.
8106 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8107 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8112 For more information about libfreetype, check:
8113 @url{http://www.freetype.org/}.
8115 For more information about fontconfig, check:
8116 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8118 For more information about libfribidi, check:
8119 @url{http://fribidi.org/}.
8123 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8125 The filter accepts the following options:
8130 Set low and high threshold values used by the Canny thresholding
8133 The high threshold selects the "strong" edge pixels, which are then
8134 connected through 8-connectivity with the "weak" edge pixels selected
8135 by the low threshold.
8137 @var{low} and @var{high} threshold values must be chosen in the range
8138 [0,1], and @var{low} should be lesser or equal to @var{high}.
8140 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8144 Define the drawing mode.
8148 Draw white/gray wires on black background.
8151 Mix the colors to create a paint/cartoon effect.
8154 Default value is @var{wires}.
8157 @subsection Examples
8161 Standard edge detection with custom values for the hysteresis thresholding:
8163 edgedetect=low=0.1:high=0.4
8167 Painting effect without thresholding:
8169 edgedetect=mode=colormix:high=0
8174 Set brightness, contrast, saturation and approximate gamma adjustment.
8176 The filter accepts the following options:
8180 Set the contrast expression. The value must be a float value in range
8181 @code{-2.0} to @code{2.0}. The default value is "1".
8184 Set the brightness expression. The value must be a float value in
8185 range @code{-1.0} to @code{1.0}. The default value is "0".
8188 Set the saturation expression. The value must be a float in
8189 range @code{0.0} to @code{3.0}. The default value is "1".
8192 Set the gamma expression. The value must be a float in range
8193 @code{0.1} to @code{10.0}. The default value is "1".
8196 Set the gamma expression for red. The value must be a float in
8197 range @code{0.1} to @code{10.0}. The default value is "1".
8200 Set the gamma expression for green. The value must be a float in range
8201 @code{0.1} to @code{10.0}. The default value is "1".
8204 Set the gamma expression for blue. The value must be a float in range
8205 @code{0.1} to @code{10.0}. The default value is "1".
8208 Set the gamma weight expression. It can be used to reduce the effect
8209 of a high gamma value on bright image areas, e.g. keep them from
8210 getting overamplified and just plain white. The value must be a float
8211 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8212 gamma correction all the way down while @code{1.0} leaves it at its
8213 full strength. Default is "1".
8216 Set when the expressions for brightness, contrast, saturation and
8217 gamma expressions are evaluated.
8219 It accepts the following values:
8222 only evaluate expressions once during the filter initialization or
8223 when a command is processed
8226 evaluate expressions for each incoming frame
8229 Default value is @samp{init}.
8232 The expressions accept the following parameters:
8235 frame count of the input frame starting from 0
8238 byte position of the corresponding packet in the input file, NAN if
8242 frame rate of the input video, NAN if the input frame rate is unknown
8245 timestamp expressed in seconds, NAN if the input timestamp is unknown
8248 @subsection Commands
8249 The filter supports the following commands:
8253 Set the contrast expression.
8256 Set the brightness expression.
8259 Set the saturation expression.
8262 Set the gamma expression.
8265 Set the gamma_r expression.
8268 Set gamma_g expression.
8271 Set gamma_b expression.
8274 Set gamma_weight expression.
8276 The command accepts the same syntax of the corresponding option.
8278 If the specified expression is not valid, it is kept at its current
8285 Apply erosion effect to the video.
8287 This filter replaces the pixel by the local(3x3) minimum.
8289 It accepts the following options:
8296 Limit the maximum change for each plane, default is 65535.
8297 If 0, plane will remain unchanged.
8300 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8303 Flags to local 3x3 coordinates maps like this:
8310 @section extractplanes
8312 Extract color channel components from input video stream into
8313 separate grayscale video streams.
8315 The filter accepts the following option:
8319 Set plane(s) to extract.
8321 Available values for planes are:
8332 Choosing planes not available in the input will result in an error.
8333 That means you cannot select @code{r}, @code{g}, @code{b} planes
8334 with @code{y}, @code{u}, @code{v} planes at same time.
8337 @subsection Examples
8341 Extract luma, u and v color channel component from input video frame
8342 into 3 grayscale outputs:
8344 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
8350 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8352 For each input image, the filter will compute the optimal mapping from
8353 the input to the output given the codebook length, that is the number
8354 of distinct output colors.
8356 This filter accepts the following options.
8359 @item codebook_length, l
8360 Set codebook length. The value must be a positive integer, and
8361 represents the number of distinct output colors. Default value is 256.
8364 Set the maximum number of iterations to apply for computing the optimal
8365 mapping. The higher the value the better the result and the higher the
8366 computation time. Default value is 1.
8369 Set a random seed, must be an integer included between 0 and
8370 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8371 will try to use a good random seed on a best effort basis.
8374 Set pal8 output pixel format. This option does not work with codebook
8375 length greater than 256.
8380 Measure graylevel entropy in histogram of color channels of video frames.
8382 It accepts the following parameters:
8386 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8388 @var{diff} mode measures entropy of histogram delta values, absolute differences
8389 between neighbour histogram values.
8394 Apply a fade-in/out effect to the input video.
8396 It accepts the following parameters:
8400 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8402 Default is @code{in}.
8404 @item start_frame, s
8405 Specify the number of the frame to start applying the fade
8406 effect at. Default is 0.
8409 The number of frames that the fade effect lasts. At the end of the
8410 fade-in effect, the output video will have the same intensity as the input video.
8411 At the end of the fade-out transition, the output video will be filled with the
8412 selected @option{color}.
8416 If set to 1, fade only alpha channel, if one exists on the input.
8419 @item start_time, st
8420 Specify the timestamp (in seconds) of the frame to start to apply the fade
8421 effect. If both start_frame and start_time are specified, the fade will start at
8422 whichever comes last. Default is 0.
8425 The number of seconds for which the fade effect has to last. At the end of the
8426 fade-in effect the output video will have the same intensity as the input video,
8427 at the end of the fade-out transition the output video will be filled with the
8428 selected @option{color}.
8429 If both duration and nb_frames are specified, duration is used. Default is 0
8430 (nb_frames is used by default).
8433 Specify the color of the fade. Default is "black".
8436 @subsection Examples
8440 Fade in the first 30 frames of video:
8445 The command above is equivalent to:
8451 Fade out the last 45 frames of a 200-frame video:
8454 fade=type=out:start_frame=155:nb_frames=45
8458 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8460 fade=in:0:25, fade=out:975:25
8464 Make the first 5 frames yellow, then fade in from frame 5-24:
8466 fade=in:5:20:color=yellow
8470 Fade in alpha over first 25 frames of video:
8472 fade=in:0:25:alpha=1
8476 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8478 fade=t=in:st=5.5:d=0.5
8484 Apply arbitrary expressions to samples in frequency domain
8488 Adjust the dc value (gain) of the luma plane of the image. The filter
8489 accepts an integer value in range @code{0} to @code{1000}. The default
8490 value is set to @code{0}.
8493 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8494 filter accepts an integer value in range @code{0} to @code{1000}. The
8495 default value is set to @code{0}.
8498 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8499 filter accepts an integer value in range @code{0} to @code{1000}. The
8500 default value is set to @code{0}.
8503 Set the frequency domain weight expression for the luma plane.
8506 Set the frequency domain weight expression for the 1st chroma plane.
8509 Set the frequency domain weight expression for the 2nd chroma plane.
8512 Set when the expressions are evaluated.
8514 It accepts the following values:
8517 Only evaluate expressions once during the filter initialization.
8520 Evaluate expressions for each incoming frame.
8523 Default value is @samp{init}.
8525 The filter accepts the following variables:
8528 The coordinates of the current sample.
8532 The width and height of the image.
8535 The number of input frame, starting from 0.
8538 @subsection Examples
8544 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8550 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8556 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8562 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8569 Extract a single field from an interlaced image using stride
8570 arithmetic to avoid wasting CPU time. The output frames are marked as
8573 The filter accepts the following options:
8577 Specify whether to extract the top (if the value is @code{0} or
8578 @code{top}) or the bottom field (if the value is @code{1} or
8584 Create new frames by copying the top and bottom fields from surrounding frames
8585 supplied as numbers by the hint file.
8589 Set file containing hints: absolute/relative frame numbers.
8591 There must be one line for each frame in a clip. Each line must contain two
8592 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8593 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8594 is current frame number for @code{absolute} mode or out of [-1, 1] range
8595 for @code{relative} mode. First number tells from which frame to pick up top
8596 field and second number tells from which frame to pick up bottom field.
8598 If optionally followed by @code{+} output frame will be marked as interlaced,
8599 else if followed by @code{-} output frame will be marked as progressive, else
8600 it will be marked same as input frame.
8601 If line starts with @code{#} or @code{;} that line is skipped.
8604 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8607 Example of first several lines of @code{hint} file for @code{relative} mode:
8610 1,0 - # second frame, use third's frame top field and second's frame bottom field
8611 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8628 Field matching filter for inverse telecine. It is meant to reconstruct the
8629 progressive frames from a telecined stream. The filter does not drop duplicated
8630 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8631 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8633 The separation of the field matching and the decimation is notably motivated by
8634 the possibility of inserting a de-interlacing filter fallback between the two.
8635 If the source has mixed telecined and real interlaced content,
8636 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8637 But these remaining combed frames will be marked as interlaced, and thus can be
8638 de-interlaced by a later filter such as @ref{yadif} before decimation.
8640 In addition to the various configuration options, @code{fieldmatch} can take an
8641 optional second stream, activated through the @option{ppsrc} option. If
8642 enabled, the frames reconstruction will be based on the fields and frames from
8643 this second stream. This allows the first input to be pre-processed in order to
8644 help the various algorithms of the filter, while keeping the output lossless
8645 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8646 or brightness/contrast adjustments can help.
8648 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8649 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8650 which @code{fieldmatch} is based on. While the semantic and usage are very
8651 close, some behaviour and options names can differ.
8653 The @ref{decimate} filter currently only works for constant frame rate input.
8654 If your input has mixed telecined (30fps) and progressive content with a lower
8655 framerate like 24fps use the following filterchain to produce the necessary cfr
8656 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8658 The filter accepts the following options:
8662 Specify the assumed field order of the input stream. Available values are:
8666 Auto detect parity (use FFmpeg's internal parity value).
8668 Assume bottom field first.
8670 Assume top field first.
8673 Note that it is sometimes recommended not to trust the parity announced by the
8676 Default value is @var{auto}.
8679 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8680 sense that it won't risk creating jerkiness due to duplicate frames when
8681 possible, but if there are bad edits or blended fields it will end up
8682 outputting combed frames when a good match might actually exist. On the other
8683 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8684 but will almost always find a good frame if there is one. The other values are
8685 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8686 jerkiness and creating duplicate frames versus finding good matches in sections
8687 with bad edits, orphaned fields, blended fields, etc.
8689 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8691 Available values are:
8695 2-way matching (p/c)
8697 2-way matching, and trying 3rd match if still combed (p/c + n)
8699 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8701 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8702 still combed (p/c + n + u/b)
8704 3-way matching (p/c/n)
8706 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8707 detected as combed (p/c/n + u/b)
8710 The parenthesis at the end indicate the matches that would be used for that
8711 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8714 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8717 Default value is @var{pc_n}.
8720 Mark the main input stream as a pre-processed input, and enable the secondary
8721 input stream as the clean source to pick the fields from. See the filter
8722 introduction for more details. It is similar to the @option{clip2} feature from
8725 Default value is @code{0} (disabled).
8728 Set the field to match from. It is recommended to set this to the same value as
8729 @option{order} unless you experience matching failures with that setting. In
8730 certain circumstances changing the field that is used to match from can have a
8731 large impact on matching performance. Available values are:
8735 Automatic (same value as @option{order}).
8737 Match from the bottom field.
8739 Match from the top field.
8742 Default value is @var{auto}.
8745 Set whether or not chroma is included during the match comparisons. In most
8746 cases it is recommended to leave this enabled. You should set this to @code{0}
8747 only if your clip has bad chroma problems such as heavy rainbowing or other
8748 artifacts. Setting this to @code{0} could also be used to speed things up at
8749 the cost of some accuracy.
8751 Default value is @code{1}.
8755 These define an exclusion band which excludes the lines between @option{y0} and
8756 @option{y1} from being included in the field matching decision. An exclusion
8757 band can be used to ignore subtitles, a logo, or other things that may
8758 interfere with the matching. @option{y0} sets the starting scan line and
8759 @option{y1} sets the ending line; all lines in between @option{y0} and
8760 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8761 @option{y0} and @option{y1} to the same value will disable the feature.
8762 @option{y0} and @option{y1} defaults to @code{0}.
8765 Set the scene change detection threshold as a percentage of maximum change on
8766 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8767 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8768 @option{scthresh} is @code{[0.0, 100.0]}.
8770 Default value is @code{12.0}.
8773 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8774 account the combed scores of matches when deciding what match to use as the
8775 final match. Available values are:
8779 No final matching based on combed scores.
8781 Combed scores are only used when a scene change is detected.
8783 Use combed scores all the time.
8786 Default is @var{sc}.
8789 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8790 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8791 Available values are:
8795 No forced calculation.
8797 Force p/c/n calculations.
8799 Force p/c/n/u/b calculations.
8802 Default value is @var{none}.
8805 This is the area combing threshold used for combed frame detection. This
8806 essentially controls how "strong" or "visible" combing must be to be detected.
8807 Larger values mean combing must be more visible and smaller values mean combing
8808 can be less visible or strong and still be detected. Valid settings are from
8809 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8810 be detected as combed). This is basically a pixel difference value. A good
8811 range is @code{[8, 12]}.
8813 Default value is @code{9}.
8816 Sets whether or not chroma is considered in the combed frame decision. Only
8817 disable this if your source has chroma problems (rainbowing, etc.) that are
8818 causing problems for the combed frame detection with chroma enabled. Actually,
8819 using @option{chroma}=@var{0} is usually more reliable, except for the case
8820 where there is chroma only combing in the source.
8822 Default value is @code{0}.
8826 Respectively set the x-axis and y-axis size of the window used during combed
8827 frame detection. This has to do with the size of the area in which
8828 @option{combpel} pixels are required to be detected as combed for a frame to be
8829 declared combed. See the @option{combpel} parameter description for more info.
8830 Possible values are any number that is a power of 2 starting at 4 and going up
8833 Default value is @code{16}.
8836 The number of combed pixels inside any of the @option{blocky} by
8837 @option{blockx} size blocks on the frame for the frame to be detected as
8838 combed. While @option{cthresh} controls how "visible" the combing must be, this
8839 setting controls "how much" combing there must be in any localized area (a
8840 window defined by the @option{blockx} and @option{blocky} settings) on the
8841 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8842 which point no frames will ever be detected as combed). This setting is known
8843 as @option{MI} in TFM/VFM vocabulary.
8845 Default value is @code{80}.
8848 @anchor{p/c/n/u/b meaning}
8849 @subsection p/c/n/u/b meaning
8851 @subsubsection p/c/n
8853 We assume the following telecined stream:
8856 Top fields: 1 2 2 3 4
8857 Bottom fields: 1 2 3 4 4
8860 The numbers correspond to the progressive frame the fields relate to. Here, the
8861 first two frames are progressive, the 3rd and 4th are combed, and so on.
8863 When @code{fieldmatch} is configured to run a matching from bottom
8864 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8869 B 1 2 3 4 4 <-- matching reference
8878 As a result of the field matching, we can see that some frames get duplicated.
8879 To perform a complete inverse telecine, you need to rely on a decimation filter
8880 after this operation. See for instance the @ref{decimate} filter.
8882 The same operation now matching from top fields (@option{field}=@var{top})
8887 T 1 2 2 3 4 <-- matching reference
8897 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8898 basically, they refer to the frame and field of the opposite parity:
8901 @item @var{p} matches the field of the opposite parity in the previous frame
8902 @item @var{c} matches the field of the opposite parity in the current frame
8903 @item @var{n} matches the field of the opposite parity in the next frame
8908 The @var{u} and @var{b} matching are a bit special in the sense that they match
8909 from the opposite parity flag. In the following examples, we assume that we are
8910 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8911 'x' is placed above and below each matched fields.
8913 With bottom matching (@option{field}=@var{bottom}):
8918 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8919 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8927 With top matching (@option{field}=@var{top}):
8932 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8933 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8941 @subsection Examples
8943 Simple IVTC of a top field first telecined stream:
8945 fieldmatch=order=tff:combmatch=none, decimate
8948 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8950 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8955 Transform the field order of the input video.
8957 It accepts the following parameters:
8962 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8963 for bottom field first.
8966 The default value is @samp{tff}.
8968 The transformation is done by shifting the picture content up or down
8969 by one line, and filling the remaining line with appropriate picture content.
8970 This method is consistent with most broadcast field order converters.
8972 If the input video is not flagged as being interlaced, or it is already
8973 flagged as being of the required output field order, then this filter does
8974 not alter the incoming video.
8976 It is very useful when converting to or from PAL DV material,
8977 which is bottom field first.
8981 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8984 @section fifo, afifo
8986 Buffer input images and send them when they are requested.
8988 It is mainly useful when auto-inserted by the libavfilter
8991 It does not take parameters.
8993 @section fillborders
8995 Fill borders of the input video, without changing video stream dimensions.
8996 Sometimes video can have garbage at the four edges and you may not want to
8997 crop video input to keep size multiple of some number.
8999 This filter accepts the following options:
9003 Number of pixels to fill from left border.
9006 Number of pixels to fill from right border.
9009 Number of pixels to fill from top border.
9012 Number of pixels to fill from bottom border.
9017 It accepts the following values:
9020 fill pixels using outermost pixels
9023 fill pixels using mirroring
9026 fill pixels with constant value
9029 Default is @var{smear}.
9032 Set color for pixels in fixed mode. Default is @var{black}.
9037 Find a rectangular object
9039 It accepts the following options:
9043 Filepath of the object image, needs to be in gray8.
9046 Detection threshold, default is 0.5.
9049 Number of mipmaps, default is 3.
9051 @item xmin, ymin, xmax, ymax
9052 Specifies the rectangle in which to search.
9055 @subsection Examples
9059 Generate a representative palette of a given video using @command{ffmpeg}:
9061 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9067 Cover a rectangular object
9069 It accepts the following options:
9073 Filepath of the optional cover image, needs to be in yuv420.
9078 It accepts the following values:
9081 cover it by the supplied image
9083 cover it by interpolating the surrounding pixels
9086 Default value is @var{blur}.
9089 @subsection Examples
9093 Generate a representative palette of a given video using @command{ffmpeg}:
9095 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9101 Flood area with values of same pixel components with another values.
9103 It accepts the following options:
9106 Set pixel x coordinate.
9109 Set pixel y coordinate.
9112 Set source #0 component value.
9115 Set source #1 component value.
9118 Set source #2 component value.
9121 Set source #3 component value.
9124 Set destination #0 component value.
9127 Set destination #1 component value.
9130 Set destination #2 component value.
9133 Set destination #3 component value.
9139 Convert the input video to one of the specified pixel formats.
9140 Libavfilter will try to pick one that is suitable as input to
9143 It accepts the following parameters:
9147 A '|'-separated list of pixel format names, such as
9148 "pix_fmts=yuv420p|monow|rgb24".
9152 @subsection Examples
9156 Convert the input video to the @var{yuv420p} format
9158 format=pix_fmts=yuv420p
9161 Convert the input video to any of the formats in the list
9163 format=pix_fmts=yuv420p|yuv444p|yuv410p
9170 Convert the video to specified constant frame rate by duplicating or dropping
9171 frames as necessary.
9173 It accepts the following parameters:
9177 The desired output frame rate. The default is @code{25}.
9180 Assume the first PTS should be the given value, in seconds. This allows for
9181 padding/trimming at the start of stream. By default, no assumption is made
9182 about the first frame's expected PTS, so no padding or trimming is done.
9183 For example, this could be set to 0 to pad the beginning with duplicates of
9184 the first frame if a video stream starts after the audio stream or to trim any
9185 frames with a negative PTS.
9188 Timestamp (PTS) rounding method.
9190 Possible values are:
9197 round towards -infinity
9199 round towards +infinity
9203 The default is @code{near}.
9206 Action performed when reading the last frame.
9208 Possible values are:
9211 Use same timestamp rounding method as used for other frames.
9213 Pass through last frame if input duration has not been reached yet.
9215 The default is @code{round}.
9219 Alternatively, the options can be specified as a flat string:
9220 @var{fps}[:@var{start_time}[:@var{round}]].
9222 See also the @ref{setpts} filter.
9224 @subsection Examples
9228 A typical usage in order to set the fps to 25:
9234 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9236 fps=fps=film:round=near
9242 Pack two different video streams into a stereoscopic video, setting proper
9243 metadata on supported codecs. The two views should have the same size and
9244 framerate and processing will stop when the shorter video ends. Please note
9245 that you may conveniently adjust view properties with the @ref{scale} and
9248 It accepts the following parameters:
9252 The desired packing format. Supported values are:
9257 The views are next to each other (default).
9260 The views are on top of each other.
9263 The views are packed by line.
9266 The views are packed by column.
9269 The views are temporally interleaved.
9278 # Convert left and right views into a frame-sequential video
9279 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9281 # Convert views into a side-by-side video with the same output resolution as the input
9282 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
9287 Change the frame rate by interpolating new video output frames from the source
9290 This filter is not designed to function correctly with interlaced media. If
9291 you wish to change the frame rate of interlaced media then you are required
9292 to deinterlace before this filter and re-interlace after this filter.
9294 A description of the accepted options follows.
9298 Specify the output frames per second. This option can also be specified
9299 as a value alone. The default is @code{50}.
9302 Specify the start of a range where the output frame will be created as a
9303 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9304 the default is @code{15}.
9307 Specify the end of a range where the output frame will be created as a
9308 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9309 the default is @code{240}.
9312 Specify the level at which a scene change is detected as a value between
9313 0 and 100 to indicate a new scene; a low value reflects a low
9314 probability for the current frame to introduce a new scene, while a higher
9315 value means the current frame is more likely to be one.
9316 The default is @code{8.2}.
9319 Specify flags influencing the filter process.
9321 Available value for @var{flags} is:
9324 @item scene_change_detect, scd
9325 Enable scene change detection using the value of the option @var{scene}.
9326 This flag is enabled by default.
9332 Select one frame every N-th frame.
9334 This filter accepts the following option:
9337 Select frame after every @code{step} frames.
9338 Allowed values are positive integers higher than 0. Default value is @code{1}.
9344 Apply a frei0r effect to the input video.
9346 To enable the compilation of this filter, you need to install the frei0r
9347 header and configure FFmpeg with @code{--enable-frei0r}.
9349 It accepts the following parameters:
9354 The name of the frei0r effect to load. If the environment variable
9355 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9356 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9357 Otherwise, the standard frei0r paths are searched, in this order:
9358 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9359 @file{/usr/lib/frei0r-1/}.
9362 A '|'-separated list of parameters to pass to the frei0r effect.
9366 A frei0r effect parameter can be a boolean (its value is either
9367 "y" or "n"), a double, a color (specified as
9368 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9369 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9370 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9371 a position (specified as @var{X}/@var{Y}, where
9372 @var{X} and @var{Y} are floating point numbers) and/or a string.
9374 The number and types of parameters depend on the loaded effect. If an
9375 effect parameter is not specified, the default value is set.
9377 @subsection Examples
9381 Apply the distort0r effect, setting the first two double parameters:
9383 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9387 Apply the colordistance effect, taking a color as the first parameter:
9389 frei0r=colordistance:0.2/0.3/0.4
9390 frei0r=colordistance:violet
9391 frei0r=colordistance:0x112233
9395 Apply the perspective effect, specifying the top left and top right image
9398 frei0r=perspective:0.2/0.2|0.8/0.2
9402 For more information, see
9403 @url{http://frei0r.dyne.org}
9407 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9409 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9410 processing filter, one of them is performed once per block, not per pixel.
9411 This allows for much higher speed.
9413 The filter accepts the following options:
9417 Set quality. This option defines the number of levels for averaging. It accepts
9418 an integer in the range 4-5. Default value is @code{4}.
9421 Force a constant quantization parameter. It accepts an integer in range 0-63.
9422 If not set, the filter will use the QP from the video stream (if available).
9425 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
9426 more details but also more artifacts, while higher values make the image smoother
9427 but also blurrier. Default value is @code{0} − PSNR optimal.
9430 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
9431 option may cause flicker since the B-Frames have often larger QP. Default is
9432 @code{0} (not enabled).
9438 Apply Gaussian blur filter.
9440 The filter accepts the following options:
9444 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
9447 Set number of steps for Gaussian approximation. Defauls is @code{1}.
9450 Set which planes to filter. By default all planes are filtered.
9453 Set vertical sigma, if negative it will be same as @code{sigma}.
9454 Default is @code{-1}.
9459 The filter accepts the following options:
9463 Set the luminance expression.
9465 Set the chrominance blue expression.
9467 Set the chrominance red expression.
9469 Set the alpha expression.
9471 Set the red expression.
9473 Set the green expression.
9475 Set the blue expression.
9478 The colorspace is selected according to the specified options. If one
9479 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
9480 options is specified, the filter will automatically select a YCbCr
9481 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
9482 @option{blue_expr} options is specified, it will select an RGB
9485 If one of the chrominance expression is not defined, it falls back on the other
9486 one. If no alpha expression is specified it will evaluate to opaque value.
9487 If none of chrominance expressions are specified, they will evaluate
9488 to the luminance expression.
9490 The expressions can use the following variables and functions:
9494 The sequential number of the filtered frame, starting from @code{0}.
9498 The coordinates of the current sample.
9502 The width and height of the image.
9506 Width and height scale depending on the currently filtered plane. It is the
9507 ratio between the corresponding luma plane number of pixels and the current
9508 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9509 @code{0.5,0.5} for chroma planes.
9512 Time of the current frame, expressed in seconds.
9515 Return the value of the pixel at location (@var{x},@var{y}) of the current
9519 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9523 Return the value of the pixel at location (@var{x},@var{y}) of the
9524 blue-difference chroma plane. Return 0 if there is no such plane.
9527 Return the value of the pixel at location (@var{x},@var{y}) of the
9528 red-difference chroma plane. Return 0 if there is no such plane.
9533 Return the value of the pixel at location (@var{x},@var{y}) of the
9534 red/green/blue component. Return 0 if there is no such component.
9537 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9538 plane. Return 0 if there is no such plane.
9541 For functions, if @var{x} and @var{y} are outside the area, the value will be
9542 automatically clipped to the closer edge.
9544 @subsection Examples
9548 Flip the image horizontally:
9554 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9555 wavelength of 100 pixels:
9557 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9561 Generate a fancy enigmatic moving light:
9563 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
9567 Generate a quick emboss effect:
9569 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9573 Modify RGB components depending on pixel position:
9575 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9579 Create a radial gradient that is the same size as the input (also see
9580 the @ref{vignette} filter):
9582 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9588 Fix the banding artifacts that are sometimes introduced into nearly flat
9589 regions by truncation to 8-bit color depth.
9590 Interpolate the gradients that should go where the bands are, and
9593 It is designed for playback only. Do not use it prior to
9594 lossy compression, because compression tends to lose the dither and
9595 bring back the bands.
9597 It accepts the following parameters:
9602 The maximum amount by which the filter will change any one pixel. This is also
9603 the threshold for detecting nearly flat regions. Acceptable values range from
9604 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9608 The neighborhood to fit the gradient to. A larger radius makes for smoother
9609 gradients, but also prevents the filter from modifying the pixels near detailed
9610 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9611 values will be clipped to the valid range.
9615 Alternatively, the options can be specified as a flat string:
9616 @var{strength}[:@var{radius}]
9618 @subsection Examples
9622 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9628 Specify radius, omitting the strength (which will fall-back to the default
9639 Apply a Hald CLUT to a video stream.
9641 First input is the video stream to process, and second one is the Hald CLUT.
9642 The Hald CLUT input can be a simple picture or a complete video stream.
9644 The filter accepts the following options:
9648 Force termination when the shortest input terminates. Default is @code{0}.
9650 Continue applying the last CLUT after the end of the stream. A value of
9651 @code{0} disable the filter after the last frame of the CLUT is reached.
9652 Default is @code{1}.
9655 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9656 filters share the same internals).
9658 More information about the Hald CLUT can be found on Eskil Steenberg's website
9659 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9661 @subsection Workflow examples
9663 @subsubsection Hald CLUT video stream
9665 Generate an identity Hald CLUT stream altered with various effects:
9667 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
9670 Note: make sure you use a lossless codec.
9672 Then use it with @code{haldclut} to apply it on some random stream:
9674 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9677 The Hald CLUT will be applied to the 10 first seconds (duration of
9678 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9679 to the remaining frames of the @code{mandelbrot} stream.
9681 @subsubsection Hald CLUT with preview
9683 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9684 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9685 biggest possible square starting at the top left of the picture. The remaining
9686 padding pixels (bottom or right) will be ignored. This area can be used to add
9687 a preview of the Hald CLUT.
9689 Typically, the following generated Hald CLUT will be supported by the
9690 @code{haldclut} filter:
9693 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9694 pad=iw+320 [padded_clut];
9695 smptebars=s=320x256, split [a][b];
9696 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9697 [main][b] overlay=W-320" -frames:v 1 clut.png
9700 It contains the original and a preview of the effect of the CLUT: SMPTE color
9701 bars are displayed on the right-top, and below the same color bars processed by
9704 Then, the effect of this Hald CLUT can be visualized with:
9706 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9711 Flip the input video horizontally.
9713 For example, to horizontally flip the input video with @command{ffmpeg}:
9715 ffmpeg -i in.avi -vf "hflip" out.avi
9719 This filter applies a global color histogram equalization on a
9722 It can be used to correct video that has a compressed range of pixel
9723 intensities. The filter redistributes the pixel intensities to
9724 equalize their distribution across the intensity range. It may be
9725 viewed as an "automatically adjusting contrast filter". This filter is
9726 useful only for correcting degraded or poorly captured source
9729 The filter accepts the following options:
9733 Determine the amount of equalization to be applied. As the strength
9734 is reduced, the distribution of pixel intensities more-and-more
9735 approaches that of the input frame. The value must be a float number
9736 in the range [0,1] and defaults to 0.200.
9739 Set the maximum intensity that can generated and scale the output
9740 values appropriately. The strength should be set as desired and then
9741 the intensity can be limited if needed to avoid washing-out. The value
9742 must be a float number in the range [0,1] and defaults to 0.210.
9745 Set the antibanding level. If enabled the filter will randomly vary
9746 the luminance of output pixels by a small amount to avoid banding of
9747 the histogram. Possible values are @code{none}, @code{weak} or
9748 @code{strong}. It defaults to @code{none}.
9753 Compute and draw a color distribution histogram for the input video.
9755 The computed histogram is a representation of the color component
9756 distribution in an image.
9758 Standard histogram displays the color components distribution in an image.
9759 Displays color graph for each color component. Shows distribution of
9760 the Y, U, V, A or R, G, B components, depending on input format, in the
9761 current frame. Below each graph a color component scale meter is shown.
9763 The filter accepts the following options:
9767 Set height of level. Default value is @code{200}.
9768 Allowed range is [50, 2048].
9771 Set height of color scale. Default value is @code{12}.
9772 Allowed range is [0, 40].
9776 It accepts the following values:
9779 Per color component graphs are placed below each other.
9782 Per color component graphs are placed side by side.
9785 Presents information identical to that in the @code{parade}, except
9786 that the graphs representing color components are superimposed directly
9789 Default is @code{stack}.
9792 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9793 Default is @code{linear}.
9796 Set what color components to display.
9797 Default is @code{7}.
9800 Set foreground opacity. Default is @code{0.7}.
9803 Set background opacity. Default is @code{0.5}.
9806 @subsection Examples
9811 Calculate and draw histogram:
9813 ffplay -i input -vf histogram
9821 This is a high precision/quality 3d denoise filter. It aims to reduce
9822 image noise, producing smooth images and making still images really
9823 still. It should enhance compressibility.
9825 It accepts the following optional parameters:
9829 A non-negative floating point number which specifies spatial luma strength.
9832 @item chroma_spatial
9833 A non-negative floating point number which specifies spatial chroma strength.
9834 It defaults to 3.0*@var{luma_spatial}/4.0.
9837 A floating point number which specifies luma temporal strength. It defaults to
9838 6.0*@var{luma_spatial}/4.0.
9841 A floating point number which specifies chroma temporal strength. It defaults to
9842 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9847 Download hardware frames to system memory.
9849 The input must be in hardware frames, and the output a non-hardware format.
9850 Not all formats will be supported on the output - it may be necessary to insert
9851 an additional @option{format} filter immediately following in the graph to get
9852 the output in a supported format.
9856 Map hardware frames to system memory or to another device.
9858 This filter has several different modes of operation; which one is used depends
9859 on the input and output formats:
9862 Hardware frame input, normal frame output
9864 Map the input frames to system memory and pass them to the output. If the
9865 original hardware frame is later required (for example, after overlaying
9866 something else on part of it), the @option{hwmap} filter can be used again
9867 in the next mode to retrieve it.
9869 Normal frame input, hardware frame output
9871 If the input is actually a software-mapped hardware frame, then unmap it -
9872 that is, return the original hardware frame.
9874 Otherwise, a device must be provided. Create new hardware surfaces on that
9875 device for the output, then map them back to the software format at the input
9876 and give those frames to the preceding filter. This will then act like the
9877 @option{hwupload} filter, but may be able to avoid an additional copy when
9878 the input is already in a compatible format.
9880 Hardware frame input and output
9882 A device must be supplied for the output, either directly or with the
9883 @option{derive_device} option. The input and output devices must be of
9884 different types and compatible - the exact meaning of this is
9885 system-dependent, but typically it means that they must refer to the same
9886 underlying hardware context (for example, refer to the same graphics card).
9888 If the input frames were originally created on the output device, then unmap
9889 to retrieve the original frames.
9891 Otherwise, map the frames to the output device - create new hardware frames
9892 on the output corresponding to the frames on the input.
9895 The following additional parameters are accepted:
9899 Set the frame mapping mode. Some combination of:
9902 The mapped frame should be readable.
9904 The mapped frame should be writeable.
9906 The mapping will always overwrite the entire frame.
9908 This may improve performance in some cases, as the original contents of the
9909 frame need not be loaded.
9911 The mapping must not involve any copying.
9913 Indirect mappings to copies of frames are created in some cases where either
9914 direct mapping is not possible or it would have unexpected properties.
9915 Setting this flag ensures that the mapping is direct and will fail if that is
9918 Defaults to @var{read+write} if not specified.
9920 @item derive_device @var{type}
9921 Rather than using the device supplied at initialisation, instead derive a new
9922 device of type @var{type} from the device the input frames exist on.
9925 In a hardware to hardware mapping, map in reverse - create frames in the sink
9926 and map them back to the source. This may be necessary in some cases where
9927 a mapping in one direction is required but only the opposite direction is
9928 supported by the devices being used.
9930 This option is dangerous - it may break the preceding filter in undefined
9931 ways if there are any additional constraints on that filter's output.
9932 Do not use it without fully understanding the implications of its use.
9937 Upload system memory frames to hardware surfaces.
9939 The device to upload to must be supplied when the filter is initialised. If
9940 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
9943 @anchor{hwupload_cuda}
9944 @section hwupload_cuda
9946 Upload system memory frames to a CUDA device.
9948 It accepts the following optional parameters:
9952 The number of the CUDA device to use
9957 Apply a high-quality magnification filter designed for pixel art. This filter
9958 was originally created by Maxim Stepin.
9960 It accepts the following option:
9964 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
9965 @code{hq3x} and @code{4} for @code{hq4x}.
9966 Default is @code{3}.
9970 Stack input videos horizontally.
9972 All streams must be of same pixel format and of same height.
9974 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
9975 to create same output.
9977 The filter accept the following option:
9981 Set number of input streams. Default is 2.
9984 If set to 1, force the output to terminate when the shortest input
9985 terminates. Default value is 0.
9990 Modify the hue and/or the saturation of the input.
9992 It accepts the following parameters:
9996 Specify the hue angle as a number of degrees. It accepts an expression,
9997 and defaults to "0".
10000 Specify the saturation in the [-10,10] range. It accepts an expression and
10004 Specify the hue angle as a number of radians. It accepts an
10005 expression, and defaults to "0".
10008 Specify the brightness in the [-10,10] range. It accepts an expression and
10012 @option{h} and @option{H} are mutually exclusive, and can't be
10013 specified at the same time.
10015 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10016 expressions containing the following constants:
10020 frame count of the input frame starting from 0
10023 presentation timestamp of the input frame expressed in time base units
10026 frame rate of the input video, NAN if the input frame rate is unknown
10029 timestamp expressed in seconds, NAN if the input timestamp is unknown
10032 time base of the input video
10035 @subsection Examples
10039 Set the hue to 90 degrees and the saturation to 1.0:
10045 Same command but expressing the hue in radians:
10051 Rotate hue and make the saturation swing between 0
10052 and 2 over a period of 1 second:
10054 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10058 Apply a 3 seconds saturation fade-in effect starting at 0:
10060 hue="s=min(t/3\,1)"
10063 The general fade-in expression can be written as:
10065 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10069 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10071 hue="s=max(0\, min(1\, (8-t)/3))"
10074 The general fade-out expression can be written as:
10076 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10081 @subsection Commands
10083 This filter supports the following commands:
10089 Modify the hue and/or the saturation and/or brightness of the input video.
10090 The command accepts the same syntax of the corresponding option.
10092 If the specified expression is not valid, it is kept at its current
10096 @section hysteresis
10098 Grow first stream into second stream by connecting components.
10099 This makes it possible to build more robust edge masks.
10101 This filter accepts the following options:
10105 Set which planes will be processed as bitmap, unprocessed planes will be
10106 copied from first stream.
10107 By default value 0xf, all planes will be processed.
10110 Set threshold which is used in filtering. If pixel component value is higher than
10111 this value filter algorithm for connecting components is activated.
10112 By default value is 0.
10117 Detect video interlacing type.
10119 This filter tries to detect if the input frames are interlaced, progressive,
10120 top or bottom field first. It will also try to detect fields that are
10121 repeated between adjacent frames (a sign of telecine).
10123 Single frame detection considers only immediately adjacent frames when classifying each frame.
10124 Multiple frame detection incorporates the classification history of previous frames.
10126 The filter will log these metadata values:
10129 @item single.current_frame
10130 Detected type of current frame using single-frame detection. One of:
10131 ``tff'' (top field first), ``bff'' (bottom field first),
10132 ``progressive'', or ``undetermined''
10135 Cumulative number of frames detected as top field first using single-frame detection.
10138 Cumulative number of frames detected as top field first using multiple-frame detection.
10141 Cumulative number of frames detected as bottom field first using single-frame detection.
10143 @item multiple.current_frame
10144 Detected type of current frame using multiple-frame detection. One of:
10145 ``tff'' (top field first), ``bff'' (bottom field first),
10146 ``progressive'', or ``undetermined''
10149 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10151 @item single.progressive
10152 Cumulative number of frames detected as progressive using single-frame detection.
10154 @item multiple.progressive
10155 Cumulative number of frames detected as progressive using multiple-frame detection.
10157 @item single.undetermined
10158 Cumulative number of frames that could not be classified using single-frame detection.
10160 @item multiple.undetermined
10161 Cumulative number of frames that could not be classified using multiple-frame detection.
10163 @item repeated.current_frame
10164 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10166 @item repeated.neither
10167 Cumulative number of frames with no repeated field.
10170 Cumulative number of frames with the top field repeated from the previous frame's top field.
10172 @item repeated.bottom
10173 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10176 The filter accepts the following options:
10180 Set interlacing threshold.
10182 Set progressive threshold.
10184 Threshold for repeated field detection.
10186 Number of frames after which a given frame's contribution to the
10187 statistics is halved (i.e., it contributes only 0.5 to its
10188 classification). The default of 0 means that all frames seen are given
10189 full weight of 1.0 forever.
10190 @item analyze_interlaced_flag
10191 When this is not 0 then idet will use the specified number of frames to determine
10192 if the interlaced flag is accurate, it will not count undetermined frames.
10193 If the flag is found to be accurate it will be used without any further
10194 computations, if it is found to be inaccurate it will be cleared without any
10195 further computations. This allows inserting the idet filter as a low computational
10196 method to clean up the interlaced flag
10201 Deinterleave or interleave fields.
10203 This filter allows one to process interlaced images fields without
10204 deinterlacing them. Deinterleaving splits the input frame into 2
10205 fields (so called half pictures). Odd lines are moved to the top
10206 half of the output image, even lines to the bottom half.
10207 You can process (filter) them independently and then re-interleave them.
10209 The filter accepts the following options:
10213 @item chroma_mode, c
10214 @item alpha_mode, a
10215 Available values for @var{luma_mode}, @var{chroma_mode} and
10216 @var{alpha_mode} are:
10222 @item deinterleave, d
10223 Deinterleave fields, placing one above the other.
10225 @item interleave, i
10226 Interleave fields. Reverse the effect of deinterleaving.
10228 Default value is @code{none}.
10230 @item luma_swap, ls
10231 @item chroma_swap, cs
10232 @item alpha_swap, as
10233 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10238 Apply inflate effect to the video.
10240 This filter replaces the pixel by the local(3x3) average by taking into account
10241 only values higher than the pixel.
10243 It accepts the following options:
10250 Limit the maximum change for each plane, default is 65535.
10251 If 0, plane will remain unchanged.
10256 Simple interlacing filter from progressive contents. This interleaves upper (or
10257 lower) lines from odd frames with lower (or upper) lines from even frames,
10258 halving the frame rate and preserving image height.
10261 Original Original New Frame
10262 Frame 'j' Frame 'j+1' (tff)
10263 ========== =========== ==================
10264 Line 0 --------------------> Frame 'j' Line 0
10265 Line 1 Line 1 ----> Frame 'j+1' Line 1
10266 Line 2 ---------------------> Frame 'j' Line 2
10267 Line 3 Line 3 ----> Frame 'j+1' Line 3
10269 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10272 It accepts the following optional parameters:
10276 This determines whether the interlaced frame is taken from the even
10277 (tff - default) or odd (bff) lines of the progressive frame.
10280 Vertical lowpass filter to avoid twitter interlacing and
10281 reduce moire patterns.
10285 Disable vertical lowpass filter
10288 Enable linear filter (default)
10291 Enable complex filter. This will slightly less reduce twitter and moire
10292 but better retain detail and subjective sharpness impression.
10299 Deinterlace input video by applying Donald Graft's adaptive kernel
10300 deinterling. Work on interlaced parts of a video to produce
10301 progressive frames.
10303 The description of the accepted parameters follows.
10307 Set the threshold which affects the filter's tolerance when
10308 determining if a pixel line must be processed. It must be an integer
10309 in the range [0,255] and defaults to 10. A value of 0 will result in
10310 applying the process on every pixels.
10313 Paint pixels exceeding the threshold value to white if set to 1.
10317 Set the fields order. Swap fields if set to 1, leave fields alone if
10321 Enable additional sharpening if set to 1. Default is 0.
10324 Enable twoway sharpening if set to 1. Default is 0.
10327 @subsection Examples
10331 Apply default values:
10333 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10337 Enable additional sharpening:
10343 Paint processed pixels in white:
10349 @section lenscorrection
10351 Correct radial lens distortion
10353 This filter can be used to correct for radial distortion as can result from the use
10354 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10355 one can use tools available for example as part of opencv or simply trial-and-error.
10356 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10357 and extract the k1 and k2 coefficients from the resulting matrix.
10359 Note that effectively the same filter is available in the open-source tools Krita and
10360 Digikam from the KDE project.
10362 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10363 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10364 brightness distribution, so you may want to use both filters together in certain
10365 cases, though you will have to take care of ordering, i.e. whether vignetting should
10366 be applied before or after lens correction.
10368 @subsection Options
10370 The filter accepts the following options:
10374 Relative x-coordinate of the focal point of the image, and thereby the center of the
10375 distortion. This value has a range [0,1] and is expressed as fractions of the image
10378 Relative y-coordinate of the focal point of the image, and thereby the center of the
10379 distortion. This value has a range [0,1] and is expressed as fractions of the image
10382 Coefficient of the quadratic correction term. 0.5 means no correction.
10384 Coefficient of the double quadratic correction term. 0.5 means no correction.
10387 The formula that generates the correction is:
10389 @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)
10391 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
10392 distances from the focal point in the source and target images, respectively.
10396 Obtain the VMAF (Video Multi-Method Assessment Fusion)
10397 score between two input videos.
10399 The obtained VMAF score is printed through the logging system.
10401 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
10402 After installing the library it can be enabled using:
10403 @code{./configure --enable-libvmaf}.
10404 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
10406 The filter has following options:
10410 Set the model path which is to be used for SVM.
10411 Default value: @code{"vmaf_v0.6.1.pkl"}
10414 Set the file path to be used to store logs.
10417 Set the format of the log file (xml or json).
10419 @item enable_transform
10420 Enables transform for computing vmaf.
10423 Invokes the phone model which will generate VMAF scores higher than in the
10424 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
10427 Enables computing psnr along with vmaf.
10430 Enables computing ssim along with vmaf.
10433 Enables computing ms_ssim along with vmaf.
10436 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
10439 This filter also supports the @ref{framesync} options.
10441 On the below examples the input file @file{main.mpg} being processed is
10442 compared with the reference file @file{ref.mpg}.
10445 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
10448 Example with options:
10450 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
10455 Limits the pixel components values to the specified range [min, max].
10457 The filter accepts the following options:
10461 Lower bound. Defaults to the lowest allowed value for the input.
10464 Upper bound. Defaults to the highest allowed value for the input.
10467 Specify which planes will be processed. Defaults to all available.
10474 The filter accepts the following options:
10478 Set the number of loops. Setting this value to -1 will result in infinite loops.
10482 Set maximal size in number of frames. Default is 0.
10485 Set first frame of loop. Default is 0.
10491 Apply a 3D LUT to an input video.
10493 The filter accepts the following options:
10497 Set the 3D LUT file name.
10499 Currently supported formats:
10511 Select interpolation mode.
10513 Available values are:
10517 Use values from the nearest defined point.
10519 Interpolate values using the 8 points defining a cube.
10521 Interpolate values using a tetrahedron.
10525 This filter also supports the @ref{framesync} options.
10529 Turn certain luma values into transparency.
10531 The filter accepts the following options:
10535 Set the luma which will be used as base for transparency.
10536 Default value is @code{0}.
10539 Set the range of luma values to be keyed out.
10540 Default value is @code{0}.
10543 Set the range of softness. Default value is @code{0}.
10544 Use this to control gradual transition from zero to full transparency.
10547 @section lut, lutrgb, lutyuv
10549 Compute a look-up table for binding each pixel component input value
10550 to an output value, and apply it to the input video.
10552 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10553 to an RGB input video.
10555 These filters accept the following parameters:
10558 set first pixel component expression
10560 set second pixel component expression
10562 set third pixel component expression
10564 set fourth pixel component expression, corresponds to the alpha component
10567 set red component expression
10569 set green component expression
10571 set blue component expression
10573 alpha component expression
10576 set Y/luminance component expression
10578 set U/Cb component expression
10580 set V/Cr component expression
10583 Each of them specifies the expression to use for computing the lookup table for
10584 the corresponding pixel component values.
10586 The exact component associated to each of the @var{c*} options depends on the
10589 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10590 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10592 The expressions can contain the following constants and functions:
10597 The input width and height.
10600 The input value for the pixel component.
10603 The input value, clipped to the @var{minval}-@var{maxval} range.
10606 The maximum value for the pixel component.
10609 The minimum value for the pixel component.
10612 The negated value for the pixel component value, clipped to the
10613 @var{minval}-@var{maxval} range; it corresponds to the expression
10614 "maxval-clipval+minval".
10617 The computed value in @var{val}, clipped to the
10618 @var{minval}-@var{maxval} range.
10620 @item gammaval(gamma)
10621 The computed gamma correction value of the pixel component value,
10622 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10624 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10628 All expressions default to "val".
10630 @subsection Examples
10634 Negate input video:
10636 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10637 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10640 The above is the same as:
10642 lutrgb="r=negval:g=negval:b=negval"
10643 lutyuv="y=negval:u=negval:v=negval"
10653 Remove chroma components, turning the video into a graytone image:
10655 lutyuv="u=128:v=128"
10659 Apply a luma burning effect:
10665 Remove green and blue components:
10671 Set a constant alpha channel value on input:
10673 format=rgba,lutrgb=a="maxval-minval/2"
10677 Correct luminance gamma by a factor of 0.5:
10679 lutyuv=y=gammaval(0.5)
10683 Discard least significant bits of luma:
10685 lutyuv=y='bitand(val, 128+64+32)'
10689 Technicolor like effect:
10691 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10695 @section lut2, tlut2
10697 The @code{lut2} filter takes two input streams and outputs one
10700 The @code{tlut2} (time lut2) filter takes two consecutive frames
10701 from one single stream.
10703 This filter accepts the following parameters:
10706 set first pixel component expression
10708 set second pixel component expression
10710 set third pixel component expression
10712 set fourth pixel component expression, corresponds to the alpha component
10715 Each of them specifies the expression to use for computing the lookup table for
10716 the corresponding pixel component values.
10718 The exact component associated to each of the @var{c*} options depends on the
10721 The expressions can contain the following constants:
10726 The input width and height.
10729 The first input value for the pixel component.
10732 The second input value for the pixel component.
10735 The first input video bit depth.
10738 The second input video bit depth.
10741 All expressions default to "x".
10743 @subsection Examples
10747 Highlight differences between two RGB video streams:
10749 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)'
10753 Highlight differences between two YUV video streams:
10755 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)'
10759 Show max difference between two video streams:
10761 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)))'
10765 @section maskedclamp
10767 Clamp the first input stream with the second input and third input stream.
10769 Returns the value of first stream to be between second input
10770 stream - @code{undershoot} and third input stream + @code{overshoot}.
10772 This filter accepts the following options:
10775 Default value is @code{0}.
10778 Default value is @code{0}.
10781 Set which planes will be processed as bitmap, unprocessed planes will be
10782 copied from first stream.
10783 By default value 0xf, all planes will be processed.
10786 @section maskedmerge
10788 Merge the first input stream with the second input stream using per pixel
10789 weights in the third input stream.
10791 A value of 0 in the third stream pixel component means that pixel component
10792 from first stream is returned unchanged, while maximum value (eg. 255 for
10793 8-bit videos) means that pixel component from second stream is returned
10794 unchanged. Intermediate values define the amount of merging between both
10795 input stream's pixel components.
10797 This filter accepts the following options:
10800 Set which planes will be processed as bitmap, unprocessed planes will be
10801 copied from first stream.
10802 By default value 0xf, all planes will be processed.
10807 Apply motion-compensation deinterlacing.
10809 It needs one field per frame as input and must thus be used together
10810 with yadif=1/3 or equivalent.
10812 This filter accepts the following options:
10815 Set the deinterlacing mode.
10817 It accepts one of the following values:
10822 use iterative motion estimation
10824 like @samp{slow}, but use multiple reference frames.
10826 Default value is @samp{fast}.
10829 Set the picture field parity assumed for the input video. It must be
10830 one of the following values:
10834 assume top field first
10836 assume bottom field first
10839 Default value is @samp{bff}.
10842 Set per-block quantization parameter (QP) used by the internal
10845 Higher values should result in a smoother motion vector field but less
10846 optimal individual vectors. Default value is 1.
10849 @section mergeplanes
10851 Merge color channel components from several video streams.
10853 The filter accepts up to 4 input streams, and merge selected input
10854 planes to the output video.
10856 This filter accepts the following options:
10859 Set input to output plane mapping. Default is @code{0}.
10861 The mappings is specified as a bitmap. It should be specified as a
10862 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10863 mapping for the first plane of the output stream. 'A' sets the number of
10864 the input stream to use (from 0 to 3), and 'a' the plane number of the
10865 corresponding input to use (from 0 to 3). The rest of the mappings is
10866 similar, 'Bb' describes the mapping for the output stream second
10867 plane, 'Cc' describes the mapping for the output stream third plane and
10868 'Dd' describes the mapping for the output stream fourth plane.
10871 Set output pixel format. Default is @code{yuva444p}.
10874 @subsection Examples
10878 Merge three gray video streams of same width and height into single video stream:
10880 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10884 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10886 [a0][a1]mergeplanes=0x00010210:yuva444p
10890 Swap Y and A plane in yuva444p stream:
10892 format=yuva444p,mergeplanes=0x03010200:yuva444p
10896 Swap U and V plane in yuv420p stream:
10898 format=yuv420p,mergeplanes=0x000201:yuv420p
10902 Cast a rgb24 clip to yuv444p:
10904 format=rgb24,mergeplanes=0x000102:yuv444p
10910 Estimate and export motion vectors using block matching algorithms.
10911 Motion vectors are stored in frame side data to be used by other filters.
10913 This filter accepts the following options:
10916 Specify the motion estimation method. Accepts one of the following values:
10920 Exhaustive search algorithm.
10922 Three step search algorithm.
10924 Two dimensional logarithmic search algorithm.
10926 New three step search algorithm.
10928 Four step search algorithm.
10930 Diamond search algorithm.
10932 Hexagon-based search algorithm.
10934 Enhanced predictive zonal search algorithm.
10936 Uneven multi-hexagon search algorithm.
10938 Default value is @samp{esa}.
10941 Macroblock size. Default @code{16}.
10944 Search parameter. Default @code{7}.
10947 @section midequalizer
10949 Apply Midway Image Equalization effect using two video streams.
10951 Midway Image Equalization adjusts a pair of images to have the same
10952 histogram, while maintaining their dynamics as much as possible. It's
10953 useful for e.g. matching exposures from a pair of stereo cameras.
10955 This filter has two inputs and one output, which must be of same pixel format, but
10956 may be of different sizes. The output of filter is first input adjusted with
10957 midway histogram of both inputs.
10959 This filter accepts the following option:
10963 Set which planes to process. Default is @code{15}, which is all available planes.
10966 @section minterpolate
10968 Convert the video to specified frame rate using motion interpolation.
10970 This filter accepts the following options:
10973 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}.
10976 Motion interpolation mode. Following values are accepted:
10979 Duplicate previous or next frame for interpolating new ones.
10981 Blend source frames. Interpolated frame is mean of previous and next frames.
10983 Motion compensated interpolation. Following options are effective when this mode is selected:
10987 Motion compensation mode. Following values are accepted:
10990 Overlapped block motion compensation.
10992 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
10994 Default mode is @samp{obmc}.
10997 Motion estimation mode. Following values are accepted:
11000 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
11002 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
11004 Default mode is @samp{bilat}.
11007 The algorithm to be used for motion estimation. Following values are accepted:
11010 Exhaustive search algorithm.
11012 Three step search algorithm.
11014 Two dimensional logarithmic search algorithm.
11016 New three step search algorithm.
11018 Four step search algorithm.
11020 Diamond search algorithm.
11022 Hexagon-based search algorithm.
11024 Enhanced predictive zonal search algorithm.
11026 Uneven multi-hexagon search algorithm.
11028 Default algorithm is @samp{epzs}.
11031 Macroblock size. Default @code{16}.
11034 Motion estimation search parameter. Default @code{32}.
11037 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).
11042 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:
11045 Disable scene change detection.
11047 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11049 Default method is @samp{fdiff}.
11051 @item scd_threshold
11052 Scene change detection threshold. Default is @code{5.0}.
11057 Mix several video input streams into one video stream.
11059 A description of the accepted options follows.
11063 The number of inputs. If unspecified, it defaults to 2.
11066 Specify weight of each input video stream as sequence.
11067 Each weight is separated by space.
11070 Specify how end of stream is determined.
11073 The duration of the longest input. (default)
11076 The duration of the shortest input.
11079 The duration of the first input.
11083 @section mpdecimate
11085 Drop frames that do not differ greatly from the previous frame in
11086 order to reduce frame rate.
11088 The main use of this filter is for very-low-bitrate encoding
11089 (e.g. streaming over dialup modem), but it could in theory be used for
11090 fixing movies that were inverse-telecined incorrectly.
11092 A description of the accepted options follows.
11096 Set the maximum number of consecutive frames which can be dropped (if
11097 positive), or the minimum interval between dropped frames (if
11098 negative). If the value is 0, the frame is dropped disregarding the
11099 number of previous sequentially dropped frames.
11101 Default value is 0.
11106 Set the dropping threshold values.
11108 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11109 represent actual pixel value differences, so a threshold of 64
11110 corresponds to 1 unit of difference for each pixel, or the same spread
11111 out differently over the block.
11113 A frame is a candidate for dropping if no 8x8 blocks differ by more
11114 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11115 meaning the whole image) differ by more than a threshold of @option{lo}.
11117 Default value for @option{hi} is 64*12, default value for @option{lo} is
11118 64*5, and default value for @option{frac} is 0.33.
11124 Negate input video.
11126 It accepts an integer in input; if non-zero it negates the
11127 alpha component (if available). The default value in input is 0.
11131 Denoise frames using Non-Local Means algorithm.
11133 Each pixel is adjusted by looking for other pixels with similar contexts. This
11134 context similarity is defined by comparing their surrounding patches of size
11135 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11138 Note that the research area defines centers for patches, which means some
11139 patches will be made of pixels outside that research area.
11141 The filter accepts the following options.
11145 Set denoising strength.
11151 Same as @option{p} but for chroma planes.
11153 The default value is @var{0} and means automatic.
11159 Same as @option{r} but for chroma planes.
11161 The default value is @var{0} and means automatic.
11166 Deinterlace video using neural network edge directed interpolation.
11168 This filter accepts the following options:
11172 Mandatory option, without binary file filter can not work.
11173 Currently file can be found here:
11174 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11177 Set which frames to deinterlace, by default it is @code{all}.
11178 Can be @code{all} or @code{interlaced}.
11181 Set mode of operation.
11183 Can be one of the following:
11187 Use frame flags, both fields.
11189 Use frame flags, single field.
11191 Use top field only.
11193 Use bottom field only.
11195 Use both fields, top first.
11197 Use both fields, bottom first.
11201 Set which planes to process, by default filter process all frames.
11204 Set size of local neighborhood around each pixel, used by the predictor neural
11207 Can be one of the following:
11220 Set the number of neurons in predictor neural network.
11221 Can be one of the following:
11232 Controls the number of different neural network predictions that are blended
11233 together to compute the final output value. Can be @code{fast}, default or
11237 Set which set of weights to use in the predictor.
11238 Can be one of the following:
11242 weights trained to minimize absolute error
11244 weights trained to minimize squared error
11248 Controls whether or not the prescreener neural network is used to decide
11249 which pixels should be processed by the predictor neural network and which
11250 can be handled by simple cubic interpolation.
11251 The prescreener is trained to know whether cubic interpolation will be
11252 sufficient for a pixel or whether it should be predicted by the predictor nn.
11253 The computational complexity of the prescreener nn is much less than that of
11254 the predictor nn. Since most pixels can be handled by cubic interpolation,
11255 using the prescreener generally results in much faster processing.
11256 The prescreener is pretty accurate, so the difference between using it and not
11257 using it is almost always unnoticeable.
11259 Can be one of the following:
11267 Default is @code{new}.
11270 Set various debugging flags.
11275 Force libavfilter not to use any of the specified pixel formats for the
11276 input to the next filter.
11278 It accepts the following parameters:
11282 A '|'-separated list of pixel format names, such as
11283 pix_fmts=yuv420p|monow|rgb24".
11287 @subsection Examples
11291 Force libavfilter to use a format different from @var{yuv420p} for the
11292 input to the vflip filter:
11294 noformat=pix_fmts=yuv420p,vflip
11298 Convert the input video to any of the formats not contained in the list:
11300 noformat=yuv420p|yuv444p|yuv410p
11306 Add noise on video input frame.
11308 The filter accepts the following options:
11316 Set noise seed for specific pixel component or all pixel components in case
11317 of @var{all_seed}. Default value is @code{123457}.
11319 @item all_strength, alls
11320 @item c0_strength, c0s
11321 @item c1_strength, c1s
11322 @item c2_strength, c2s
11323 @item c3_strength, c3s
11324 Set noise strength for specific pixel component or all pixel components in case
11325 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
11327 @item all_flags, allf
11328 @item c0_flags, c0f
11329 @item c1_flags, c1f
11330 @item c2_flags, c2f
11331 @item c3_flags, c3f
11332 Set pixel component flags or set flags for all components if @var{all_flags}.
11333 Available values for component flags are:
11336 averaged temporal noise (smoother)
11338 mix random noise with a (semi)regular pattern
11340 temporal noise (noise pattern changes between frames)
11342 uniform noise (gaussian otherwise)
11346 @subsection Examples
11348 Add temporal and uniform noise to input video:
11350 noise=alls=20:allf=t+u
11355 Normalize RGB video (aka histogram stretching, contrast stretching).
11356 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
11358 For each channel of each frame, the filter computes the input range and maps
11359 it linearly to the user-specified output range. The output range defaults
11360 to the full dynamic range from pure black to pure white.
11362 Temporal smoothing can be used on the input range to reduce flickering (rapid
11363 changes in brightness) caused when small dark or bright objects enter or leave
11364 the scene. This is similar to the auto-exposure (automatic gain control) on a
11365 video camera, and, like a video camera, it may cause a period of over- or
11366 under-exposure of the video.
11368 The R,G,B channels can be normalized independently, which may cause some
11369 color shifting, or linked together as a single channel, which prevents
11370 color shifting. Linked normalization preserves hue. Independent normalization
11371 does not, so it can be used to remove some color casts. Independent and linked
11372 normalization can be combined in any ratio.
11374 The normalize filter accepts the following options:
11379 Colors which define the output range. The minimum input value is mapped to
11380 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
11381 The defaults are black and white respectively. Specifying white for
11382 @var{blackpt} and black for @var{whitept} will give color-inverted,
11383 normalized video. Shades of grey can be used to reduce the dynamic range
11384 (contrast). Specifying saturated colors here can create some interesting
11388 The number of previous frames to use for temporal smoothing. The input range
11389 of each channel is smoothed using a rolling average over the current frame
11390 and the @var{smoothing} previous frames. The default is 0 (no temporal
11394 Controls the ratio of independent (color shifting) channel normalization to
11395 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
11396 independent. Defaults to 1.0 (fully independent).
11399 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
11400 expensive no-op. Defaults to 1.0 (full strength).
11404 @subsection Examples
11406 Stretch video contrast to use the full dynamic range, with no temporal
11407 smoothing; may flicker depending on the source content:
11409 normalize=blackpt=black:whitept=white:smoothing=0
11412 As above, but with 50 frames of temporal smoothing; flicker should be
11413 reduced, depending on the source content:
11415 normalize=blackpt=black:whitept=white:smoothing=50
11418 As above, but with hue-preserving linked channel normalization:
11420 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
11423 As above, but with half strength:
11425 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
11428 Map the darkest input color to red, the brightest input color to cyan:
11430 normalize=blackpt=red:whitept=cyan
11435 Pass the video source unchanged to the output.
11438 Optical Character Recognition
11440 This filter uses Tesseract for optical character recognition.
11442 It accepts the following options:
11446 Set datapath to tesseract data. Default is to use whatever was
11447 set at installation.
11450 Set language, default is "eng".
11453 Set character whitelist.
11456 Set character blacklist.
11459 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
11463 Apply a video transform using libopencv.
11465 To enable this filter, install the libopencv library and headers and
11466 configure FFmpeg with @code{--enable-libopencv}.
11468 It accepts the following parameters:
11473 The name of the libopencv filter to apply.
11475 @item filter_params
11476 The parameters to pass to the libopencv filter. If not specified, the default
11477 values are assumed.
11481 Refer to the official libopencv documentation for more precise
11483 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11485 Several libopencv filters are supported; see the following subsections.
11490 Dilate an image by using a specific structuring element.
11491 It corresponds to the libopencv function @code{cvDilate}.
11493 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11495 @var{struct_el} represents a structuring element, and has the syntax:
11496 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11498 @var{cols} and @var{rows} represent the number of columns and rows of
11499 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11500 point, and @var{shape} the shape for the structuring element. @var{shape}
11501 must be "rect", "cross", "ellipse", or "custom".
11503 If the value for @var{shape} is "custom", it must be followed by a
11504 string of the form "=@var{filename}". The file with name
11505 @var{filename} is assumed to represent a binary image, with each
11506 printable character corresponding to a bright pixel. When a custom
11507 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11508 or columns and rows of the read file are assumed instead.
11510 The default value for @var{struct_el} is "3x3+0x0/rect".
11512 @var{nb_iterations} specifies the number of times the transform is
11513 applied to the image, and defaults to 1.
11517 # Use the default values
11520 # Dilate using a structuring element with a 5x5 cross, iterating two times
11521 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
11523 # Read the shape from the file diamond.shape, iterating two times.
11524 # The file diamond.shape may contain a pattern of characters like this
11530 # The specified columns and rows are ignored
11531 # but the anchor point coordinates are not
11532 ocv=dilate:0x0+2x2/custom=diamond.shape|2
11537 Erode an image by using a specific structuring element.
11538 It corresponds to the libopencv function @code{cvErode}.
11540 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
11541 with the same syntax and semantics as the @ref{dilate} filter.
11545 Smooth the input video.
11547 The filter takes the following parameters:
11548 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
11550 @var{type} is the type of smooth filter to apply, and must be one of
11551 the following values: "blur", "blur_no_scale", "median", "gaussian",
11552 or "bilateral". The default value is "gaussian".
11554 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
11555 depend on the smooth type. @var{param1} and
11556 @var{param2} accept integer positive values or 0. @var{param3} and
11557 @var{param4} accept floating point values.
11559 The default value for @var{param1} is 3. The default value for the
11560 other parameters is 0.
11562 These parameters correspond to the parameters assigned to the
11563 libopencv function @code{cvSmooth}.
11565 @section oscilloscope
11567 2D Video Oscilloscope.
11569 Useful to measure spatial impulse, step responses, chroma delays, etc.
11571 It accepts the following parameters:
11575 Set scope center x position.
11578 Set scope center y position.
11581 Set scope size, relative to frame diagonal.
11584 Set scope tilt/rotation.
11590 Set trace center x position.
11593 Set trace center y position.
11596 Set trace width, relative to width of frame.
11599 Set trace height, relative to height of frame.
11602 Set which components to trace. By default it traces first three components.
11605 Draw trace grid. By default is enabled.
11608 Draw some statistics. By default is enabled.
11611 Draw scope. By default is enabled.
11614 @subsection Examples
11618 Inspect full first row of video frame.
11620 oscilloscope=x=0.5:y=0:s=1
11624 Inspect full last row of video frame.
11626 oscilloscope=x=0.5:y=1:s=1
11630 Inspect full 5th line of video frame of height 1080.
11632 oscilloscope=x=0.5:y=5/1080:s=1
11636 Inspect full last column of video frame.
11638 oscilloscope=x=1:y=0.5:s=1:t=1
11646 Overlay one video on top of another.
11648 It takes two inputs and has one output. The first input is the "main"
11649 video on which the second input is overlaid.
11651 It accepts the following parameters:
11653 A description of the accepted options follows.
11658 Set the expression for the x and y coordinates of the overlaid video
11659 on the main video. Default value is "0" for both expressions. In case
11660 the expression is invalid, it is set to a huge value (meaning that the
11661 overlay will not be displayed within the output visible area).
11664 See @ref{framesync}.
11667 Set when the expressions for @option{x}, and @option{y} are evaluated.
11669 It accepts the following values:
11672 only evaluate expressions once during the filter initialization or
11673 when a command is processed
11676 evaluate expressions for each incoming frame
11679 Default value is @samp{frame}.
11682 See @ref{framesync}.
11685 Set the format for the output video.
11687 It accepts the following values:
11690 force YUV420 output
11693 force YUV422 output
11696 force YUV444 output
11699 force packed RGB output
11702 force planar RGB output
11705 automatically pick format
11708 Default value is @samp{yuv420}.
11711 See @ref{framesync}.
11714 Set format of alpha of the overlaid video, it can be @var{straight} or
11715 @var{premultiplied}. Default is @var{straight}.
11718 The @option{x}, and @option{y} expressions can contain the following
11724 The main input width and height.
11728 The overlay input width and height.
11732 The computed values for @var{x} and @var{y}. They are evaluated for
11737 horizontal and vertical chroma subsample values of the output
11738 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11742 the number of input frame, starting from 0
11745 the position in the file of the input frame, NAN if unknown
11748 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11752 This filter also supports the @ref{framesync} options.
11754 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11755 when evaluation is done @emph{per frame}, and will evaluate to NAN
11756 when @option{eval} is set to @samp{init}.
11758 Be aware that frames are taken from each input video in timestamp
11759 order, hence, if their initial timestamps differ, it is a good idea
11760 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11761 have them begin in the same zero timestamp, as the example for
11762 the @var{movie} filter does.
11764 You can chain together more overlays but you should test the
11765 efficiency of such approach.
11767 @subsection Commands
11769 This filter supports the following commands:
11773 Modify the x and y of the overlay input.
11774 The command accepts the same syntax of the corresponding option.
11776 If the specified expression is not valid, it is kept at its current
11780 @subsection Examples
11784 Draw the overlay at 10 pixels from the bottom right corner of the main
11787 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11790 Using named options the example above becomes:
11792 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11796 Insert a transparent PNG logo in the bottom left corner of the input,
11797 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11799 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11803 Insert 2 different transparent PNG logos (second logo on bottom
11804 right corner) using the @command{ffmpeg} tool:
11806 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
11810 Add a transparent color layer on top of the main video; @code{WxH}
11811 must specify the size of the main input to the overlay filter:
11813 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11817 Play an original video and a filtered version (here with the deshake
11818 filter) side by side using the @command{ffplay} tool:
11820 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11823 The above command is the same as:
11825 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11829 Make a sliding overlay appearing from the left to the right top part of the
11830 screen starting since time 2:
11832 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11836 Compose output by putting two input videos side to side:
11838 ffmpeg -i left.avi -i right.avi -filter_complex "
11839 nullsrc=size=200x100 [background];
11840 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11841 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11842 [background][left] overlay=shortest=1 [background+left];
11843 [background+left][right] overlay=shortest=1:x=100 [left+right]
11848 Mask 10-20 seconds of a video by applying the delogo filter to a section
11850 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11851 -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]'
11856 Chain several overlays in cascade:
11858 nullsrc=s=200x200 [bg];
11859 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11860 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11861 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11862 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11863 [in3] null, [mid2] overlay=100:100 [out0]
11870 Apply Overcomplete Wavelet denoiser.
11872 The filter accepts the following options:
11878 Larger depth values will denoise lower frequency components more, but
11879 slow down filtering.
11881 Must be an int in the range 8-16, default is @code{8}.
11883 @item luma_strength, ls
11886 Must be a double value in the range 0-1000, default is @code{1.0}.
11888 @item chroma_strength, cs
11889 Set chroma strength.
11891 Must be a double value in the range 0-1000, default is @code{1.0}.
11897 Add paddings to the input image, and place the original input at the
11898 provided @var{x}, @var{y} coordinates.
11900 It accepts the following parameters:
11905 Specify an expression for the size of the output image with the
11906 paddings added. If the value for @var{width} or @var{height} is 0, the
11907 corresponding input size is used for the output.
11909 The @var{width} expression can reference the value set by the
11910 @var{height} expression, and vice versa.
11912 The default value of @var{width} and @var{height} is 0.
11916 Specify the offsets to place the input image at within the padded area,
11917 with respect to the top/left border of the output image.
11919 The @var{x} expression can reference the value set by the @var{y}
11920 expression, and vice versa.
11922 The default value of @var{x} and @var{y} is 0.
11924 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
11925 so the input image is centered on the padded area.
11928 Specify the color of the padded area. For the syntax of this option,
11929 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
11930 manual,ffmpeg-utils}.
11932 The default value of @var{color} is "black".
11935 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
11937 It accepts the following values:
11941 Only evaluate expressions once during the filter initialization or when
11942 a command is processed.
11945 Evaluate expressions for each incoming frame.
11949 Default value is @samp{init}.
11952 Pad to aspect instead to a resolution.
11956 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
11957 options are expressions containing the following constants:
11962 The input video width and height.
11966 These are the same as @var{in_w} and @var{in_h}.
11970 The output width and height (the size of the padded area), as
11971 specified by the @var{width} and @var{height} expressions.
11975 These are the same as @var{out_w} and @var{out_h}.
11979 The x and y offsets as specified by the @var{x} and @var{y}
11980 expressions, or NAN if not yet specified.
11983 same as @var{iw} / @var{ih}
11986 input sample aspect ratio
11989 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
11993 The horizontal and vertical chroma subsample values. For example for the
11994 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11997 @subsection Examples
12001 Add paddings with the color "violet" to the input video. The output video
12002 size is 640x480, and the top-left corner of the input video is placed at
12005 pad=640:480:0:40:violet
12008 The example above is equivalent to the following command:
12010 pad=width=640:height=480:x=0:y=40:color=violet
12014 Pad the input to get an output with dimensions increased by 3/2,
12015 and put the input video at the center of the padded area:
12017 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
12021 Pad the input to get a squared output with size equal to the maximum
12022 value between the input width and height, and put the input video at
12023 the center of the padded area:
12025 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12029 Pad the input to get a final w/h ratio of 16:9:
12031 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12035 In case of anamorphic video, in order to set the output display aspect
12036 correctly, it is necessary to use @var{sar} in the expression,
12037 according to the relation:
12039 (ih * X / ih) * sar = output_dar
12040 X = output_dar / sar
12043 Thus the previous example needs to be modified to:
12045 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12049 Double the output size and put the input video in the bottom-right
12050 corner of the output padded area:
12052 pad="2*iw:2*ih:ow-iw:oh-ih"
12056 @anchor{palettegen}
12057 @section palettegen
12059 Generate one palette for a whole video stream.
12061 It accepts the following options:
12065 Set the maximum number of colors to quantize in the palette.
12066 Note: the palette will still contain 256 colors; the unused palette entries
12069 @item reserve_transparent
12070 Create a palette of 255 colors maximum and reserve the last one for
12071 transparency. Reserving the transparency color is useful for GIF optimization.
12072 If not set, the maximum of colors in the palette will be 256. You probably want
12073 to disable this option for a standalone image.
12076 @item transparency_color
12077 Set the color that will be used as background for transparency.
12080 Set statistics mode.
12082 It accepts the following values:
12085 Compute full frame histograms.
12087 Compute histograms only for the part that differs from previous frame. This
12088 might be relevant to give more importance to the moving part of your input if
12089 the background is static.
12091 Compute new histogram for each frame.
12094 Default value is @var{full}.
12097 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12098 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12099 color quantization of the palette. This information is also visible at
12100 @var{info} logging level.
12102 @subsection Examples
12106 Generate a representative palette of a given video using @command{ffmpeg}:
12108 ffmpeg -i input.mkv -vf palettegen palette.png
12112 @section paletteuse
12114 Use a palette to downsample an input video stream.
12116 The filter takes two inputs: one video stream and a palette. The palette must
12117 be a 256 pixels image.
12119 It accepts the following options:
12123 Select dithering mode. Available algorithms are:
12126 Ordered 8x8 bayer dithering (deterministic)
12128 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12129 Note: this dithering is sometimes considered "wrong" and is included as a
12131 @item floyd_steinberg
12132 Floyd and Steingberg dithering (error diffusion)
12134 Frankie Sierra dithering v2 (error diffusion)
12136 Frankie Sierra dithering v2 "Lite" (error diffusion)
12139 Default is @var{sierra2_4a}.
12142 When @var{bayer} dithering is selected, this option defines the scale of the
12143 pattern (how much the crosshatch pattern is visible). A low value means more
12144 visible pattern for less banding, and higher value means less visible pattern
12145 at the cost of more banding.
12147 The option must be an integer value in the range [0,5]. Default is @var{2}.
12150 If set, define the zone to process
12154 Only the changing rectangle will be reprocessed. This is similar to GIF
12155 cropping/offsetting compression mechanism. This option can be useful for speed
12156 if only a part of the image is changing, and has use cases such as limiting the
12157 scope of the error diffusal @option{dither} to the rectangle that bounds the
12158 moving scene (it leads to more deterministic output if the scene doesn't change
12159 much, and as a result less moving noise and better GIF compression).
12162 Default is @var{none}.
12165 Take new palette for each output frame.
12167 @item alpha_threshold
12168 Sets the alpha threshold for transparency. Alpha values above this threshold
12169 will be treated as completely opaque, and values below this threshold will be
12170 treated as completely transparent.
12172 The option must be an integer value in the range [0,255]. Default is @var{128}.
12175 @subsection Examples
12179 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12180 using @command{ffmpeg}:
12182 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12186 @section perspective
12188 Correct perspective of video not recorded perpendicular to the screen.
12190 A description of the accepted parameters follows.
12201 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12202 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
12203 If the @code{sense} option is set to @code{source}, then the specified points will be sent
12204 to the corners of the destination. If the @code{sense} option is set to @code{destination},
12205 then the corners of the source will be sent to the specified coordinates.
12207 The expressions can use the following variables:
12212 the width and height of video frame.
12216 Output frame count.
12219 @item interpolation
12220 Set interpolation for perspective correction.
12222 It accepts the following values:
12228 Default value is @samp{linear}.
12231 Set interpretation of coordinate options.
12233 It accepts the following values:
12237 Send point in the source specified by the given coordinates to
12238 the corners of the destination.
12240 @item 1, destination
12242 Send the corners of the source to the point in the destination specified
12243 by the given coordinates.
12245 Default value is @samp{source}.
12249 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
12251 It accepts the following values:
12254 only evaluate expressions once during the filter initialization or
12255 when a command is processed
12258 evaluate expressions for each incoming frame
12261 Default value is @samp{init}.
12266 Delay interlaced video by one field time so that the field order changes.
12268 The intended use is to fix PAL movies that have been captured with the
12269 opposite field order to the film-to-video transfer.
12271 A description of the accepted parameters follows.
12277 It accepts the following values:
12280 Capture field order top-first, transfer bottom-first.
12281 Filter will delay the bottom field.
12284 Capture field order bottom-first, transfer top-first.
12285 Filter will delay the top field.
12288 Capture and transfer with the same field order. This mode only exists
12289 for the documentation of the other options to refer to, but if you
12290 actually select it, the filter will faithfully do nothing.
12293 Capture field order determined automatically by field flags, transfer
12295 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
12296 basis using field flags. If no field information is available,
12297 then this works just like @samp{u}.
12300 Capture unknown or varying, transfer opposite.
12301 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
12302 analyzing the images and selecting the alternative that produces best
12303 match between the fields.
12306 Capture top-first, transfer unknown or varying.
12307 Filter selects among @samp{t} and @samp{p} using image analysis.
12310 Capture bottom-first, transfer unknown or varying.
12311 Filter selects among @samp{b} and @samp{p} using image analysis.
12314 Capture determined by field flags, transfer unknown or varying.
12315 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
12316 image analysis. If no field information is available, then this works just
12317 like @samp{U}. This is the default mode.
12320 Both capture and transfer unknown or varying.
12321 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
12325 @section pixdesctest
12327 Pixel format descriptor test filter, mainly useful for internal
12328 testing. The output video should be equal to the input video.
12332 format=monow, pixdesctest
12335 can be used to test the monowhite pixel format descriptor definition.
12339 Display sample values of color channels. Mainly useful for checking color
12340 and levels. Minimum supported resolution is 640x480.
12342 The filters accept the following options:
12346 Set scope X position, relative offset on X axis.
12349 Set scope Y position, relative offset on Y axis.
12358 Set window opacity. This window also holds statistics about pixel area.
12361 Set window X position, relative offset on X axis.
12364 Set window Y position, relative offset on Y axis.
12369 Enable the specified chain of postprocessing subfilters using libpostproc. This
12370 library should be automatically selected with a GPL build (@code{--enable-gpl}).
12371 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
12372 Each subfilter and some options have a short and a long name that can be used
12373 interchangeably, i.e. dr/dering are the same.
12375 The filters accept the following options:
12379 Set postprocessing subfilters string.
12382 All subfilters share common options to determine their scope:
12386 Honor the quality commands for this subfilter.
12389 Do chrominance filtering, too (default).
12392 Do luminance filtering only (no chrominance).
12395 Do chrominance filtering only (no luminance).
12398 These options can be appended after the subfilter name, separated by a '|'.
12400 Available subfilters are:
12403 @item hb/hdeblock[|difference[|flatness]]
12404 Horizontal deblocking filter
12407 Difference factor where higher values mean more deblocking (default: @code{32}).
12409 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12412 @item vb/vdeblock[|difference[|flatness]]
12413 Vertical deblocking filter
12416 Difference factor where higher values mean more deblocking (default: @code{32}).
12418 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12421 @item ha/hadeblock[|difference[|flatness]]
12422 Accurate horizontal deblocking filter
12425 Difference factor where higher values mean more deblocking (default: @code{32}).
12427 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12430 @item va/vadeblock[|difference[|flatness]]
12431 Accurate vertical deblocking filter
12434 Difference factor where higher values mean more deblocking (default: @code{32}).
12436 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12440 The horizontal and vertical deblocking filters share the difference and
12441 flatness values so you cannot set different horizontal and vertical
12445 @item h1/x1hdeblock
12446 Experimental horizontal deblocking filter
12448 @item v1/x1vdeblock
12449 Experimental vertical deblocking filter
12454 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
12457 larger -> stronger filtering
12459 larger -> stronger filtering
12461 larger -> stronger filtering
12464 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
12467 Stretch luminance to @code{0-255}.
12470 @item lb/linblenddeint
12471 Linear blend deinterlacing filter that deinterlaces the given block by
12472 filtering all lines with a @code{(1 2 1)} filter.
12474 @item li/linipoldeint
12475 Linear interpolating deinterlacing filter that deinterlaces the given block by
12476 linearly interpolating every second line.
12478 @item ci/cubicipoldeint
12479 Cubic interpolating deinterlacing filter deinterlaces the given block by
12480 cubically interpolating every second line.
12482 @item md/mediandeint
12483 Median deinterlacing filter that deinterlaces the given block by applying a
12484 median filter to every second line.
12486 @item fd/ffmpegdeint
12487 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12488 second line with a @code{(-1 4 2 4 -1)} filter.
12491 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12492 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12494 @item fq/forceQuant[|quantizer]
12495 Overrides the quantizer table from the input with the constant quantizer you
12503 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12506 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
12509 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
12512 @subsection Examples
12516 Apply horizontal and vertical deblocking, deringing and automatic
12517 brightness/contrast:
12523 Apply default filters without brightness/contrast correction:
12529 Apply default filters and temporal denoiser:
12531 pp=default/tmpnoise|1|2|3
12535 Apply deblocking on luminance only, and switch vertical deblocking on or off
12536 automatically depending on available CPU time:
12543 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
12544 similar to spp = 6 with 7 point DCT, where only the center sample is
12547 The filter accepts the following options:
12551 Force a constant quantization parameter. It accepts an integer in range
12552 0 to 63. If not set, the filter will use the QP from the video stream
12556 Set thresholding mode. Available modes are:
12560 Set hard thresholding.
12562 Set soft thresholding (better de-ringing effect, but likely blurrier).
12564 Set medium thresholding (good results, default).
12568 @section premultiply
12569 Apply alpha premultiply effect to input video stream using first plane
12570 of second stream as alpha.
12572 Both streams must have same dimensions and same pixel format.
12574 The filter accepts the following option:
12578 Set which planes will be processed, unprocessed planes will be copied.
12579 By default value 0xf, all planes will be processed.
12582 Do not require 2nd input for processing, instead use alpha plane from input stream.
12586 Apply prewitt operator to input video stream.
12588 The filter accepts the following option:
12592 Set which planes will be processed, unprocessed planes will be copied.
12593 By default value 0xf, all planes will be processed.
12596 Set value which will be multiplied with filtered result.
12599 Set value which will be added to filtered result.
12602 @anchor{program_opencl}
12603 @section program_opencl
12605 Filter video using an OpenCL program.
12610 OpenCL program source file.
12613 Kernel name in program.
12616 Number of inputs to the filter. Defaults to 1.
12619 Size of output frames. Defaults to the same as the first input.
12623 The program source file must contain a kernel function with the given name,
12624 which will be run once for each plane of the output. Each run on a plane
12625 gets enqueued as a separate 2D global NDRange with one work-item for each
12626 pixel to be generated. The global ID offset for each work-item is therefore
12627 the coordinates of a pixel in the destination image.
12629 The kernel function needs to take the following arguments:
12632 Destination image, @var{__write_only image2d_t}.
12634 This image will become the output; the kernel should write all of it.
12636 Frame index, @var{unsigned int}.
12638 This is a counter starting from zero and increasing by one for each frame.
12640 Source images, @var{__read_only image2d_t}.
12642 These are the most recent images on each input. The kernel may read from
12643 them to generate the output, but they can't be written to.
12650 Copy the input to the output (output must be the same size as the input).
12652 __kernel void copy(__write_only image2d_t destination,
12653 unsigned int index,
12654 __read_only image2d_t source)
12656 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
12658 int2 location = (int2)(get_global_id(0), get_global_id(1));
12660 float4 value = read_imagef(source, sampler, location);
12662 write_imagef(destination, location, value);
12667 Apply a simple transformation, rotating the input by an amount increasing
12668 with the index counter. Pixel values are linearly interpolated by the
12669 sampler, and the output need not have the same dimensions as the input.
12671 __kernel void rotate_image(__write_only image2d_t dst,
12672 unsigned int index,
12673 __read_only image2d_t src)
12675 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12676 CLK_FILTER_LINEAR);
12678 float angle = (float)index / 100.0f;
12680 float2 dst_dim = convert_float2(get_image_dim(dst));
12681 float2 src_dim = convert_float2(get_image_dim(src));
12683 float2 dst_cen = dst_dim / 2.0f;
12684 float2 src_cen = src_dim / 2.0f;
12686 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12688 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
12690 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
12691 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
12693 src_pos = src_pos * src_dim / dst_dim;
12695 float2 src_loc = src_pos + src_cen;
12697 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
12698 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
12699 write_imagef(dst, dst_loc, 0.5f);
12701 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
12706 Blend two inputs together, with the amount of each input used varying
12707 with the index counter.
12709 __kernel void blend_images(__write_only image2d_t dst,
12710 unsigned int index,
12711 __read_only image2d_t src1,
12712 __read_only image2d_t src2)
12714 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12715 CLK_FILTER_LINEAR);
12717 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
12719 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12720 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
12721 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
12723 float4 val1 = read_imagef(src1, sampler, src1_loc);
12724 float4 val2 = read_imagef(src2, sampler, src2_loc);
12726 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
12732 @section pseudocolor
12734 Alter frame colors in video with pseudocolors.
12736 This filter accept the following options:
12740 set pixel first component expression
12743 set pixel second component expression
12746 set pixel third component expression
12749 set pixel fourth component expression, corresponds to the alpha component
12752 set component to use as base for altering colors
12755 Each of them specifies the expression to use for computing the lookup table for
12756 the corresponding pixel component values.
12758 The expressions can contain the following constants and functions:
12763 The input width and height.
12766 The input value for the pixel component.
12768 @item ymin, umin, vmin, amin
12769 The minimum allowed component value.
12771 @item ymax, umax, vmax, amax
12772 The maximum allowed component value.
12775 All expressions default to "val".
12777 @subsection Examples
12781 Change too high luma values to gradient:
12783 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'"
12789 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12790 Ratio) between two input videos.
12792 This filter takes in input two input videos, the first input is
12793 considered the "main" source and is passed unchanged to the
12794 output. The second input is used as a "reference" video for computing
12797 Both video inputs must have the same resolution and pixel format for
12798 this filter to work correctly. Also it assumes that both inputs
12799 have the same number of frames, which are compared one by one.
12801 The obtained average PSNR is printed through the logging system.
12803 The filter stores the accumulated MSE (mean squared error) of each
12804 frame, and at the end of the processing it is averaged across all frames
12805 equally, and the following formula is applied to obtain the PSNR:
12808 PSNR = 10*log10(MAX^2/MSE)
12811 Where MAX is the average of the maximum values of each component of the
12814 The description of the accepted parameters follows.
12817 @item stats_file, f
12818 If specified the filter will use the named file to save the PSNR of
12819 each individual frame. When filename equals "-" the data is sent to
12822 @item stats_version
12823 Specifies which version of the stats file format to use. Details of
12824 each format are written below.
12825 Default value is 1.
12827 @item stats_add_max
12828 Determines whether the max value is output to the stats log.
12829 Default value is 0.
12830 Requires stats_version >= 2. If this is set and stats_version < 2,
12831 the filter will return an error.
12834 This filter also supports the @ref{framesync} options.
12836 The file printed if @var{stats_file} is selected, contains a sequence of
12837 key/value pairs of the form @var{key}:@var{value} for each compared
12840 If a @var{stats_version} greater than 1 is specified, a header line precedes
12841 the list of per-frame-pair stats, with key value pairs following the frame
12842 format with the following parameters:
12845 @item psnr_log_version
12846 The version of the log file format. Will match @var{stats_version}.
12849 A comma separated list of the per-frame-pair parameters included in
12853 A description of each shown per-frame-pair parameter follows:
12857 sequential number of the input frame, starting from 1
12860 Mean Square Error pixel-by-pixel average difference of the compared
12861 frames, averaged over all the image components.
12863 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
12864 Mean Square Error pixel-by-pixel average difference of the compared
12865 frames for the component specified by the suffix.
12867 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12868 Peak Signal to Noise ratio of the compared frames for the component
12869 specified by the suffix.
12871 @item max_avg, max_y, max_u, max_v
12872 Maximum allowed value for each channel, and average over all
12878 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12879 [main][ref] psnr="stats_file=stats.log" [out]
12882 On this example the input file being processed is compared with the
12883 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12884 is stored in @file{stats.log}.
12889 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12890 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12893 The pullup filter is designed to take advantage of future context in making
12894 its decisions. This filter is stateless in the sense that it does not lock
12895 onto a pattern to follow, but it instead looks forward to the following
12896 fields in order to identify matches and rebuild progressive frames.
12898 To produce content with an even framerate, insert the fps filter after
12899 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12900 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12902 The filter accepts the following options:
12909 These options set the amount of "junk" to ignore at the left, right, top, and
12910 bottom of the image, respectively. Left and right are in units of 8 pixels,
12911 while top and bottom are in units of 2 lines.
12912 The default is 8 pixels on each side.
12915 Set the strict breaks. Setting this option to 1 will reduce the chances of
12916 filter generating an occasional mismatched frame, but it may also cause an
12917 excessive number of frames to be dropped during high motion sequences.
12918 Conversely, setting it to -1 will make filter match fields more easily.
12919 This may help processing of video where there is slight blurring between
12920 the fields, but may also cause there to be interlaced frames in the output.
12921 Default value is @code{0}.
12924 Set the metric plane to use. It accepts the following values:
12930 Use chroma blue plane.
12933 Use chroma red plane.
12936 This option may be set to use chroma plane instead of the default luma plane
12937 for doing filter's computations. This may improve accuracy on very clean
12938 source material, but more likely will decrease accuracy, especially if there
12939 is chroma noise (rainbow effect) or any grayscale video.
12940 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
12941 load and make pullup usable in realtime on slow machines.
12944 For best results (without duplicated frames in the output file) it is
12945 necessary to change the output frame rate. For example, to inverse
12946 telecine NTSC input:
12948 ffmpeg -i input -vf pullup -r 24000/1001 ...
12953 Change video quantization parameters (QP).
12955 The filter accepts the following option:
12959 Set expression for quantization parameter.
12962 The expression is evaluated through the eval API and can contain, among others,
12963 the following constants:
12967 1 if index is not 129, 0 otherwise.
12970 Sequential index starting from -129 to 128.
12973 @subsection Examples
12977 Some equation like:
12985 Flush video frames from internal cache of frames into a random order.
12986 No frame is discarded.
12987 Inspired by @ref{frei0r} nervous filter.
12991 Set size in number of frames of internal cache, in range from @code{2} to
12992 @code{512}. Default is @code{30}.
12995 Set seed for random number generator, must be an integer included between
12996 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
12997 less than @code{0}, the filter will try to use a good random seed on a
13001 @section readeia608
13003 Read closed captioning (EIA-608) information from the top lines of a video frame.
13005 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
13006 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
13007 with EIA-608 data (starting from 0). A description of each metadata value follows:
13010 @item lavfi.readeia608.X.cc
13011 The two bytes stored as EIA-608 data (printed in hexadecimal).
13013 @item lavfi.readeia608.X.line
13014 The number of the line on which the EIA-608 data was identified and read.
13017 This filter accepts the following options:
13021 Set the line to start scanning for EIA-608 data. Default is @code{0}.
13024 Set the line to end scanning for EIA-608 data. Default is @code{29}.
13027 Set minimal acceptable amplitude change for sync codes detection.
13028 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13031 Set the ratio of width reserved for sync code detection.
13032 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13035 Set the max peaks height difference for sync code detection.
13036 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13039 Set max peaks period difference for sync code detection.
13040 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13043 Set the first two max start code bits differences.
13044 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13047 Set the minimum ratio of bits height compared to 3rd start code bit.
13048 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13051 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13054 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13057 Enable checking the parity bit. In the event of a parity error, the filter will output
13058 @code{0x00} for that character. Default is false.
13061 @subsection Examples
13065 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13067 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
13073 Read vertical interval timecode (VITC) information from the top lines of a
13076 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13077 timecode value, if a valid timecode has been detected. Further metadata key
13078 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13079 timecode data has been found or not.
13081 This filter accepts the following options:
13085 Set the maximum number of lines to scan for VITC data. If the value is set to
13086 @code{-1} the full video frame is scanned. Default is @code{45}.
13089 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13090 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13093 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13094 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13097 @subsection Examples
13101 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13102 draw @code{--:--:--:--} as a placeholder:
13104 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13110 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13112 Destination pixel at position (X, Y) will be picked from source (x, y) position
13113 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13114 value for pixel will be used for destination pixel.
13116 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13117 will have Xmap/Ymap video stream dimensions.
13118 Xmap and Ymap input video streams are 16bit depth, single channel.
13120 @section removegrain
13122 The removegrain filter is a spatial denoiser for progressive video.
13126 Set mode for the first plane.
13129 Set mode for the second plane.
13132 Set mode for the third plane.
13135 Set mode for the fourth plane.
13138 Range of mode is from 0 to 24. Description of each mode follows:
13142 Leave input plane unchanged. Default.
13145 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13148 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13151 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13154 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13155 This is equivalent to a median filter.
13158 Line-sensitive clipping giving the minimal change.
13161 Line-sensitive clipping, intermediate.
13164 Line-sensitive clipping, intermediate.
13167 Line-sensitive clipping, intermediate.
13170 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13173 Replaces the target pixel with the closest neighbour.
13176 [1 2 1] horizontal and vertical kernel blur.
13182 Bob mode, interpolates top field from the line where the neighbours
13183 pixels are the closest.
13186 Bob mode, interpolates bottom field from the line where the neighbours
13187 pixels are the closest.
13190 Bob mode, interpolates top field. Same as 13 but with a more complicated
13191 interpolation formula.
13194 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13195 interpolation formula.
13198 Clips the pixel with the minimum and maximum of respectively the maximum and
13199 minimum of each pair of opposite neighbour pixels.
13202 Line-sensitive clipping using opposite neighbours whose greatest distance from
13203 the current pixel is minimal.
13206 Replaces the pixel with the average of its 8 neighbours.
13209 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
13212 Clips pixels using the averages of opposite neighbour.
13215 Same as mode 21 but simpler and faster.
13218 Small edge and halo removal, but reputed useless.
13224 @section removelogo
13226 Suppress a TV station logo, using an image file to determine which
13227 pixels comprise the logo. It works by filling in the pixels that
13228 comprise the logo with neighboring pixels.
13230 The filter accepts the following options:
13234 Set the filter bitmap file, which can be any image format supported by
13235 libavformat. The width and height of the image file must match those of the
13236 video stream being processed.
13239 Pixels in the provided bitmap image with a value of zero are not
13240 considered part of the logo, non-zero pixels are considered part of
13241 the logo. If you use white (255) for the logo and black (0) for the
13242 rest, you will be safe. For making the filter bitmap, it is
13243 recommended to take a screen capture of a black frame with the logo
13244 visible, and then using a threshold filter followed by the erode
13245 filter once or twice.
13247 If needed, little splotches can be fixed manually. Remember that if
13248 logo pixels are not covered, the filter quality will be much
13249 reduced. Marking too many pixels as part of the logo does not hurt as
13250 much, but it will increase the amount of blurring needed to cover over
13251 the image and will destroy more information than necessary, and extra
13252 pixels will slow things down on a large logo.
13254 @section repeatfields
13256 This filter uses the repeat_field flag from the Video ES headers and hard repeats
13257 fields based on its value.
13261 Reverse a video clip.
13263 Warning: This filter requires memory to buffer the entire clip, so trimming
13266 @subsection Examples
13270 Take the first 5 seconds of a clip, and reverse it.
13277 Apply roberts cross operator to input video stream.
13279 The filter accepts the following option:
13283 Set which planes will be processed, unprocessed planes will be copied.
13284 By default value 0xf, all planes will be processed.
13287 Set value which will be multiplied with filtered result.
13290 Set value which will be added to filtered result.
13295 Rotate video by an arbitrary angle expressed in radians.
13297 The filter accepts the following options:
13299 A description of the optional parameters follows.
13302 Set an expression for the angle by which to rotate the input video
13303 clockwise, expressed as a number of radians. A negative value will
13304 result in a counter-clockwise rotation. By default it is set to "0".
13306 This expression is evaluated for each frame.
13309 Set the output width expression, default value is "iw".
13310 This expression is evaluated just once during configuration.
13313 Set the output height expression, default value is "ih".
13314 This expression is evaluated just once during configuration.
13317 Enable bilinear interpolation if set to 1, a value of 0 disables
13318 it. Default value is 1.
13321 Set the color used to fill the output area not covered by the rotated
13322 image. For the general syntax of this option, check the
13323 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
13324 If the special value "none" is selected then no
13325 background is printed (useful for example if the background is never shown).
13327 Default value is "black".
13330 The expressions for the angle and the output size can contain the
13331 following constants and functions:
13335 sequential number of the input frame, starting from 0. It is always NAN
13336 before the first frame is filtered.
13339 time in seconds of the input frame, it is set to 0 when the filter is
13340 configured. It is always NAN before the first frame is filtered.
13344 horizontal and vertical chroma subsample values. For example for the
13345 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13349 the input video width and height
13353 the output width and height, that is the size of the padded area as
13354 specified by the @var{width} and @var{height} expressions
13358 the minimal width/height required for completely containing the input
13359 video rotated by @var{a} radians.
13361 These are only available when computing the @option{out_w} and
13362 @option{out_h} expressions.
13365 @subsection Examples
13369 Rotate the input by PI/6 radians clockwise:
13375 Rotate the input by PI/6 radians counter-clockwise:
13381 Rotate the input by 45 degrees clockwise:
13387 Apply a constant rotation with period T, starting from an angle of PI/3:
13389 rotate=PI/3+2*PI*t/T
13393 Make the input video rotation oscillating with a period of T
13394 seconds and an amplitude of A radians:
13396 rotate=A*sin(2*PI/T*t)
13400 Rotate the video, output size is chosen so that the whole rotating
13401 input video is always completely contained in the output:
13403 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
13407 Rotate the video, reduce the output size so that no background is ever
13410 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
13414 @subsection Commands
13416 The filter supports the following commands:
13420 Set the angle expression.
13421 The command accepts the same syntax of the corresponding option.
13423 If the specified expression is not valid, it is kept at its current
13429 Apply Shape Adaptive Blur.
13431 The filter accepts the following options:
13434 @item luma_radius, lr
13435 Set luma blur filter strength, must be a value in range 0.1-4.0, default
13436 value is 1.0. A greater value will result in a more blurred image, and
13437 in slower processing.
13439 @item luma_pre_filter_radius, lpfr
13440 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
13443 @item luma_strength, ls
13444 Set luma maximum difference between pixels to still be considered, must
13445 be a value in the 0.1-100.0 range, default value is 1.0.
13447 @item chroma_radius, cr
13448 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
13449 greater value will result in a more blurred image, and in slower
13452 @item chroma_pre_filter_radius, cpfr
13453 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
13455 @item chroma_strength, cs
13456 Set chroma maximum difference between pixels to still be considered,
13457 must be a value in the -0.9-100.0 range.
13460 Each chroma option value, if not explicitly specified, is set to the
13461 corresponding luma option value.
13466 Scale (resize) the input video, using the libswscale library.
13468 The scale filter forces the output display aspect ratio to be the same
13469 of the input, by changing the output sample aspect ratio.
13471 If the input image format is different from the format requested by
13472 the next filter, the scale filter will convert the input to the
13475 @subsection Options
13476 The filter accepts the following options, or any of the options
13477 supported by the libswscale scaler.
13479 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
13480 the complete list of scaler options.
13485 Set the output video dimension expression. Default value is the input
13488 If the @var{width} or @var{w} value is 0, the input width is used for
13489 the output. If the @var{height} or @var{h} value is 0, the input height
13490 is used for the output.
13492 If one and only one of the values is -n with n >= 1, the scale filter
13493 will use a value that maintains the aspect ratio of the input image,
13494 calculated from the other specified dimension. After that it will,
13495 however, make sure that the calculated dimension is divisible by n and
13496 adjust the value if necessary.
13498 If both values are -n with n >= 1, the behavior will be identical to
13499 both values being set to 0 as previously detailed.
13501 See below for the list of accepted constants for use in the dimension
13505 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
13509 Only evaluate expressions once during the filter initialization or when a command is processed.
13512 Evaluate expressions for each incoming frame.
13516 Default value is @samp{init}.
13520 Set the interlacing mode. It accepts the following values:
13524 Force interlaced aware scaling.
13527 Do not apply interlaced scaling.
13530 Select interlaced aware scaling depending on whether the source frames
13531 are flagged as interlaced or not.
13534 Default value is @samp{0}.
13537 Set libswscale scaling flags. See
13538 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13539 complete list of values. If not explicitly specified the filter applies
13543 @item param0, param1
13544 Set libswscale input parameters for scaling algorithms that need them. See
13545 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13546 complete documentation. If not explicitly specified the filter applies
13552 Set the video size. For the syntax of this option, check the
13553 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13555 @item in_color_matrix
13556 @item out_color_matrix
13557 Set in/output YCbCr color space type.
13559 This allows the autodetected value to be overridden as well as allows forcing
13560 a specific value used for the output and encoder.
13562 If not specified, the color space type depends on the pixel format.
13568 Choose automatically.
13571 Format conforming to International Telecommunication Union (ITU)
13572 Recommendation BT.709.
13575 Set color space conforming to the United States Federal Communications
13576 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
13579 Set color space conforming to:
13583 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
13586 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
13589 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
13594 Set color space conforming to SMPTE ST 240:1999.
13599 Set in/output YCbCr sample range.
13601 This allows the autodetected value to be overridden as well as allows forcing
13602 a specific value used for the output and encoder. If not specified, the
13603 range depends on the pixel format. Possible values:
13607 Choose automatically.
13610 Set full range (0-255 in case of 8-bit luma).
13612 @item mpeg/limited/tv
13613 Set "MPEG" range (16-235 in case of 8-bit luma).
13616 @item force_original_aspect_ratio
13617 Enable decreasing or increasing output video width or height if necessary to
13618 keep the original aspect ratio. Possible values:
13622 Scale the video as specified and disable this feature.
13625 The output video dimensions will automatically be decreased if needed.
13628 The output video dimensions will automatically be increased if needed.
13632 One useful instance of this option is that when you know a specific device's
13633 maximum allowed resolution, you can use this to limit the output video to
13634 that, while retaining the aspect ratio. For example, device A allows
13635 1280x720 playback, and your video is 1920x800. Using this option (set it to
13636 decrease) and specifying 1280x720 to the command line makes the output
13639 Please note that this is a different thing than specifying -1 for @option{w}
13640 or @option{h}, you still need to specify the output resolution for this option
13645 The values of the @option{w} and @option{h} options are expressions
13646 containing the following constants:
13651 The input width and height
13655 These are the same as @var{in_w} and @var{in_h}.
13659 The output (scaled) width and height
13663 These are the same as @var{out_w} and @var{out_h}
13666 The same as @var{iw} / @var{ih}
13669 input sample aspect ratio
13672 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
13676 horizontal and vertical input chroma subsample values. For example for the
13677 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13681 horizontal and vertical output chroma subsample values. For example for the
13682 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13685 @subsection Examples
13689 Scale the input video to a size of 200x100
13694 This is equivalent to:
13705 Specify a size abbreviation for the output size:
13710 which can also be written as:
13716 Scale the input to 2x:
13718 scale=w=2*iw:h=2*ih
13722 The above is the same as:
13724 scale=2*in_w:2*in_h
13728 Scale the input to 2x with forced interlaced scaling:
13730 scale=2*iw:2*ih:interl=1
13734 Scale the input to half size:
13736 scale=w=iw/2:h=ih/2
13740 Increase the width, and set the height to the same size:
13746 Seek Greek harmony:
13753 Increase the height, and set the width to 3/2 of the height:
13755 scale=w=3/2*oh:h=3/5*ih
13759 Increase the size, making the size a multiple of the chroma
13762 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13766 Increase the width to a maximum of 500 pixels,
13767 keeping the same aspect ratio as the input:
13769 scale=w='min(500\, iw*3/2):h=-1'
13773 Make pixels square by combining scale and setsar:
13775 scale='trunc(ih*dar):ih',setsar=1/1
13779 Make pixels square by combining scale and setsar,
13780 making sure the resulting resolution is even (required by some codecs):
13782 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
13786 @subsection Commands
13788 This filter supports the following commands:
13792 Set the output video dimension expression.
13793 The command accepts the same syntax of the corresponding option.
13795 If the specified expression is not valid, it is kept at its current
13801 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13802 format conversion on CUDA video frames. Setting the output width and height
13803 works in the same way as for the @var{scale} filter.
13805 The following additional options are accepted:
13808 The pixel format of the output CUDA frames. If set to the string "same" (the
13809 default), the input format will be kept. Note that automatic format negotiation
13810 and conversion is not yet supported for hardware frames
13813 The interpolation algorithm used for resizing. One of the following:
13820 @item cubic2p_bspline
13821 2-parameter cubic (B=1, C=0)
13823 @item cubic2p_catmullrom
13824 2-parameter cubic (B=0, C=1/2)
13826 @item cubic2p_b05c03
13827 2-parameter cubic (B=1/2, C=3/10)
13839 Scale (resize) the input video, based on a reference video.
13841 See the scale filter for available options, scale2ref supports the same but
13842 uses the reference video instead of the main input as basis. scale2ref also
13843 supports the following additional constants for the @option{w} and
13844 @option{h} options:
13849 The main input video's width and height
13852 The same as @var{main_w} / @var{main_h}
13855 The main input video's sample aspect ratio
13857 @item main_dar, mdar
13858 The main input video's display aspect ratio. Calculated from
13859 @code{(main_w / main_h) * main_sar}.
13863 The main input video's horizontal and vertical chroma subsample values.
13864 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13868 @subsection Examples
13872 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13874 'scale2ref[b][a];[a][b]overlay'
13878 @anchor{selectivecolor}
13879 @section selectivecolor
13881 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13882 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13883 by the "purity" of the color (that is, how saturated it already is).
13885 This filter is similar to the Adobe Photoshop Selective Color tool.
13887 The filter accepts the following options:
13890 @item correction_method
13891 Select color correction method.
13893 Available values are:
13896 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13899 Specified adjustments are relative to the original component value.
13901 Default is @code{absolute}.
13903 Adjustments for red pixels (pixels where the red component is the maximum)
13905 Adjustments for yellow pixels (pixels where the blue component is the minimum)
13907 Adjustments for green pixels (pixels where the green component is the maximum)
13909 Adjustments for cyan pixels (pixels where the red component is the minimum)
13911 Adjustments for blue pixels (pixels where the blue component is the maximum)
13913 Adjustments for magenta pixels (pixels where the green component is the minimum)
13915 Adjustments for white pixels (pixels where all components are greater than 128)
13917 Adjustments for all pixels except pure black and pure white
13919 Adjustments for black pixels (pixels where all components are lesser than 128)
13921 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
13924 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
13925 4 space separated floating point adjustment values in the [-1,1] range,
13926 respectively to adjust the amount of cyan, magenta, yellow and black for the
13927 pixels of its range.
13929 @subsection Examples
13933 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
13934 increase magenta by 27% in blue areas:
13936 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
13940 Use a Photoshop selective color preset:
13942 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
13946 @anchor{separatefields}
13947 @section separatefields
13949 The @code{separatefields} takes a frame-based video input and splits
13950 each frame into its components fields, producing a new half height clip
13951 with twice the frame rate and twice the frame count.
13953 This filter use field-dominance information in frame to decide which
13954 of each pair of fields to place first in the output.
13955 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
13957 @section setdar, setsar
13959 The @code{setdar} filter sets the Display Aspect Ratio for the filter
13962 This is done by changing the specified Sample (aka Pixel) Aspect
13963 Ratio, according to the following equation:
13965 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
13968 Keep in mind that the @code{setdar} filter does not modify the pixel
13969 dimensions of the video frame. Also, the display aspect ratio set by
13970 this filter may be changed by later filters in the filterchain,
13971 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
13974 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
13975 the filter output video.
13977 Note that as a consequence of the application of this filter, the
13978 output display aspect ratio will change according to the equation
13981 Keep in mind that the sample aspect ratio set by the @code{setsar}
13982 filter may be changed by later filters in the filterchain, e.g. if
13983 another "setsar" or a "setdar" filter is applied.
13985 It accepts the following parameters:
13988 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
13989 Set the aspect ratio used by the filter.
13991 The parameter can be a floating point number string, an expression, or
13992 a string of the form @var{num}:@var{den}, where @var{num} and
13993 @var{den} are the numerator and denominator of the aspect ratio. If
13994 the parameter is not specified, it is assumed the value "0".
13995 In case the form "@var{num}:@var{den}" is used, the @code{:} character
13999 Set the maximum integer value to use for expressing numerator and
14000 denominator when reducing the expressed aspect ratio to a rational.
14001 Default value is @code{100}.
14005 The parameter @var{sar} is an expression containing
14006 the following constants:
14010 These are approximated values for the mathematical constants e
14011 (Euler's number), pi (Greek pi), and phi (the golden ratio).
14014 The input width and height.
14017 These are the same as @var{w} / @var{h}.
14020 The input sample aspect ratio.
14023 The input display aspect ratio. It is the same as
14024 (@var{w} / @var{h}) * @var{sar}.
14027 Horizontal and vertical chroma subsample values. For example, for the
14028 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14031 @subsection Examples
14036 To change the display aspect ratio to 16:9, specify one of the following:
14043 To change the sample aspect ratio to 10:11, specify:
14049 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14050 1000 in the aspect ratio reduction, use the command:
14052 setdar=ratio=16/9:max=1000
14060 Force field for the output video frame.
14062 The @code{setfield} filter marks the interlace type field for the
14063 output frames. It does not change the input frame, but only sets the
14064 corresponding property, which affects how the frame is treated by
14065 following filters (e.g. @code{fieldorder} or @code{yadif}).
14067 The filter accepts the following options:
14072 Available values are:
14076 Keep the same field property.
14079 Mark the frame as bottom-field-first.
14082 Mark the frame as top-field-first.
14085 Mark the frame as progressive.
14091 Show a line containing various information for each input video frame.
14092 The input video is not modified.
14094 The shown line contains a sequence of key/value pairs of the form
14095 @var{key}:@var{value}.
14097 The following values are shown in the output:
14101 The (sequential) number of the input frame, starting from 0.
14104 The Presentation TimeStamp of the input frame, expressed as a number of
14105 time base units. The time base unit depends on the filter input pad.
14108 The Presentation TimeStamp of the input frame, expressed as a number of
14112 The position of the frame in the input stream, or -1 if this information is
14113 unavailable and/or meaningless (for example in case of synthetic video).
14116 The pixel format name.
14119 The sample aspect ratio of the input frame, expressed in the form
14120 @var{num}/@var{den}.
14123 The size of the input frame. For the syntax of this option, check the
14124 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14127 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14128 for bottom field first).
14131 This is 1 if the frame is a key frame, 0 otherwise.
14134 The picture type of the input frame ("I" for an I-frame, "P" for a
14135 P-frame, "B" for a B-frame, or "?" for an unknown type).
14136 Also refer to the documentation of the @code{AVPictureType} enum and of
14137 the @code{av_get_picture_type_char} function defined in
14138 @file{libavutil/avutil.h}.
14141 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14143 @item plane_checksum
14144 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14145 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14148 @section showpalette
14150 Displays the 256 colors palette of each frame. This filter is only relevant for
14151 @var{pal8} pixel format frames.
14153 It accepts the following option:
14157 Set the size of the box used to represent one palette color entry. Default is
14158 @code{30} (for a @code{30x30} pixel box).
14161 @section shuffleframes
14163 Reorder and/or duplicate and/or drop video frames.
14165 It accepts the following parameters:
14169 Set the destination indexes of input frames.
14170 This is space or '|' separated list of indexes that maps input frames to output
14171 frames. Number of indexes also sets maximal value that each index may have.
14172 '-1' index have special meaning and that is to drop frame.
14175 The first frame has the index 0. The default is to keep the input unchanged.
14177 @subsection Examples
14181 Swap second and third frame of every three frames of the input:
14183 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14187 Swap 10th and 1st frame of every ten frames of the input:
14189 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14193 @section shuffleplanes
14195 Reorder and/or duplicate video planes.
14197 It accepts the following parameters:
14202 The index of the input plane to be used as the first output plane.
14205 The index of the input plane to be used as the second output plane.
14208 The index of the input plane to be used as the third output plane.
14211 The index of the input plane to be used as the fourth output plane.
14215 The first plane has the index 0. The default is to keep the input unchanged.
14217 @subsection Examples
14221 Swap the second and third planes of the input:
14223 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
14227 @anchor{signalstats}
14228 @section signalstats
14229 Evaluate various visual metrics that assist in determining issues associated
14230 with the digitization of analog video media.
14232 By default the filter will log these metadata values:
14236 Display the minimal Y value contained within the input frame. Expressed in
14240 Display the Y value at the 10% percentile within the input frame. Expressed in
14244 Display the average Y value within the input frame. Expressed in range of
14248 Display the Y value at the 90% percentile within the input frame. Expressed in
14252 Display the maximum Y value contained within the input frame. Expressed in
14256 Display the minimal U value contained within the input frame. Expressed in
14260 Display the U value at the 10% percentile within the input frame. Expressed in
14264 Display the average U value within the input frame. Expressed in range of
14268 Display the U value at the 90% percentile within the input frame. Expressed in
14272 Display the maximum U value contained within the input frame. Expressed in
14276 Display the minimal V value contained within the input frame. Expressed in
14280 Display the V value at the 10% percentile within the input frame. Expressed in
14284 Display the average V value within the input frame. Expressed in range of
14288 Display the V value at the 90% percentile within the input frame. Expressed in
14292 Display the maximum V value contained within the input frame. Expressed in
14296 Display the minimal saturation value contained within the input frame.
14297 Expressed in range of [0-~181.02].
14300 Display the saturation value at the 10% percentile within the input frame.
14301 Expressed in range of [0-~181.02].
14304 Display the average saturation value within the input frame. Expressed in range
14308 Display the saturation value at the 90% percentile within the input frame.
14309 Expressed in range of [0-~181.02].
14312 Display the maximum saturation value contained within the input frame.
14313 Expressed in range of [0-~181.02].
14316 Display the median value for hue within the input frame. Expressed in range of
14320 Display the average value for hue within the input frame. Expressed in range of
14324 Display the average of sample value difference between all values of the Y
14325 plane in the current frame and corresponding values of the previous input frame.
14326 Expressed in range of [0-255].
14329 Display the average of sample value difference between all values of the U
14330 plane in the current frame and corresponding values of the previous input frame.
14331 Expressed in range of [0-255].
14334 Display the average of sample value difference between all values of the V
14335 plane in the current frame and corresponding values of the previous input frame.
14336 Expressed in range of [0-255].
14339 Display bit depth of Y plane in current frame.
14340 Expressed in range of [0-16].
14343 Display bit depth of U plane in current frame.
14344 Expressed in range of [0-16].
14347 Display bit depth of V plane in current frame.
14348 Expressed in range of [0-16].
14351 The filter accepts the following options:
14357 @option{stat} specify an additional form of image analysis.
14358 @option{out} output video with the specified type of pixel highlighted.
14360 Both options accept the following values:
14364 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
14365 unlike the neighboring pixels of the same field. Examples of temporal outliers
14366 include the results of video dropouts, head clogs, or tape tracking issues.
14369 Identify @var{vertical line repetition}. Vertical line repetition includes
14370 similar rows of pixels within a frame. In born-digital video vertical line
14371 repetition is common, but this pattern is uncommon in video digitized from an
14372 analog source. When it occurs in video that results from the digitization of an
14373 analog source it can indicate concealment from a dropout compensator.
14376 Identify pixels that fall outside of legal broadcast range.
14380 Set the highlight color for the @option{out} option. The default color is
14384 @subsection Examples
14388 Output data of various video metrics:
14390 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
14394 Output specific data about the minimum and maximum values of the Y plane per frame:
14396 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
14400 Playback video while highlighting pixels that are outside of broadcast range in red.
14402 ffplay example.mov -vf signalstats="out=brng:color=red"
14406 Playback video with signalstats metadata drawn over the frame.
14408 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
14411 The contents of signalstat_drawtext.txt used in the command are:
14414 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
14415 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
14416 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
14417 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
14425 Calculates the MPEG-7 Video Signature. The filter can handle more than one
14426 input. In this case the matching between the inputs can be calculated additionally.
14427 The filter always passes through the first input. The signature of each stream can
14428 be written into a file.
14430 It accepts the following options:
14434 Enable or disable the matching process.
14436 Available values are:
14440 Disable the calculation of a matching (default).
14442 Calculate the matching for the whole video and output whether the whole video
14443 matches or only parts.
14445 Calculate only until a matching is found or the video ends. Should be faster in
14450 Set the number of inputs. The option value must be a non negative integer.
14451 Default value is 1.
14454 Set the path to which the output is written. If there is more than one input,
14455 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
14456 integer), that will be replaced with the input number. If no filename is
14457 specified, no output will be written. This is the default.
14460 Choose the output format.
14462 Available values are:
14466 Use the specified binary representation (default).
14468 Use the specified xml representation.
14472 Set threshold to detect one word as similar. The option value must be an integer
14473 greater than zero. The default value is 9000.
14476 Set threshold to detect all words as similar. The option value must be an integer
14477 greater than zero. The default value is 60000.
14480 Set threshold to detect frames as similar. The option value must be an integer
14481 greater than zero. The default value is 116.
14484 Set the minimum length of a sequence in frames to recognize it as matching
14485 sequence. The option value must be a non negative integer value.
14486 The default value is 0.
14489 Set the minimum relation, that matching frames to all frames must have.
14490 The option value must be a double value between 0 and 1. The default value is 0.5.
14493 @subsection Examples
14497 To calculate the signature of an input video and store it in signature.bin:
14499 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
14503 To detect whether two videos match and store the signatures in XML format in
14504 signature0.xml and signature1.xml:
14506 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 -
14514 Blur the input video without impacting the outlines.
14516 It accepts the following options:
14519 @item luma_radius, lr
14520 Set the luma radius. The option value must be a float number in
14521 the range [0.1,5.0] that specifies the variance of the gaussian filter
14522 used to blur the image (slower if larger). Default value is 1.0.
14524 @item luma_strength, ls
14525 Set the luma strength. The option value must be a float number
14526 in the range [-1.0,1.0] that configures the blurring. A value included
14527 in [0.0,1.0] will blur the image whereas a value included in
14528 [-1.0,0.0] will sharpen the image. Default value is 1.0.
14530 @item luma_threshold, lt
14531 Set the luma threshold used as a coefficient to determine
14532 whether a pixel should be blurred or not. The option value must be an
14533 integer in the range [-30,30]. A value of 0 will filter all the image,
14534 a value included in [0,30] will filter flat areas and a value included
14535 in [-30,0] will filter edges. Default value is 0.
14537 @item chroma_radius, cr
14538 Set the chroma radius. The option value must be a float number in
14539 the range [0.1,5.0] that specifies the variance of the gaussian filter
14540 used to blur the image (slower if larger). Default value is @option{luma_radius}.
14542 @item chroma_strength, cs
14543 Set the chroma strength. The option value must be a float number
14544 in the range [-1.0,1.0] that configures the blurring. A value included
14545 in [0.0,1.0] will blur the image whereas a value included in
14546 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
14548 @item chroma_threshold, ct
14549 Set the chroma threshold used as a coefficient to determine
14550 whether a pixel should be blurred or not. The option value must be an
14551 integer in the range [-30,30]. A value of 0 will filter all the image,
14552 a value included in [0,30] will filter flat areas and a value included
14553 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
14556 If a chroma option is not explicitly set, the corresponding luma value
14561 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
14563 This filter takes in input two input videos, the first input is
14564 considered the "main" source and is passed unchanged to the
14565 output. The second input is used as a "reference" video for computing
14568 Both video inputs must have the same resolution and pixel format for
14569 this filter to work correctly. Also it assumes that both inputs
14570 have the same number of frames, which are compared one by one.
14572 The filter stores the calculated SSIM of each frame.
14574 The description of the accepted parameters follows.
14577 @item stats_file, f
14578 If specified the filter will use the named file to save the SSIM of
14579 each individual frame. When filename equals "-" the data is sent to
14583 The file printed if @var{stats_file} is selected, contains a sequence of
14584 key/value pairs of the form @var{key}:@var{value} for each compared
14587 A description of each shown parameter follows:
14591 sequential number of the input frame, starting from 1
14593 @item Y, U, V, R, G, B
14594 SSIM of the compared frames for the component specified by the suffix.
14597 SSIM of the compared frames for the whole frame.
14600 Same as above but in dB representation.
14603 This filter also supports the @ref{framesync} options.
14607 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14608 [main][ref] ssim="stats_file=stats.log" [out]
14611 On this example the input file being processed is compared with the
14612 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
14613 is stored in @file{stats.log}.
14615 Another example with both psnr and ssim at same time:
14617 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
14622 Convert between different stereoscopic image formats.
14624 The filters accept the following options:
14628 Set stereoscopic image format of input.
14630 Available values for input image formats are:
14633 side by side parallel (left eye left, right eye right)
14636 side by side crosseye (right eye left, left eye right)
14639 side by side parallel with half width resolution
14640 (left eye left, right eye right)
14643 side by side crosseye with half width resolution
14644 (right eye left, left eye right)
14647 above-below (left eye above, right eye below)
14650 above-below (right eye above, left eye below)
14653 above-below with half height resolution
14654 (left eye above, right eye below)
14657 above-below with half height resolution
14658 (right eye above, left eye below)
14661 alternating frames (left eye first, right eye second)
14664 alternating frames (right eye first, left eye second)
14667 interleaved rows (left eye has top row, right eye starts on next row)
14670 interleaved rows (right eye has top row, left eye starts on next row)
14673 interleaved columns, left eye first
14676 interleaved columns, right eye first
14678 Default value is @samp{sbsl}.
14682 Set stereoscopic image format of output.
14686 side by side parallel (left eye left, right eye right)
14689 side by side crosseye (right eye left, left eye right)
14692 side by side parallel with half width resolution
14693 (left eye left, right eye right)
14696 side by side crosseye with half width resolution
14697 (right eye left, left eye right)
14700 above-below (left eye above, right eye below)
14703 above-below (right eye above, left eye below)
14706 above-below with half height resolution
14707 (left eye above, right eye below)
14710 above-below with half height resolution
14711 (right eye above, left eye below)
14714 alternating frames (left eye first, right eye second)
14717 alternating frames (right eye first, left eye second)
14720 interleaved rows (left eye has top row, right eye starts on next row)
14723 interleaved rows (right eye has top row, left eye starts on next row)
14726 anaglyph red/blue gray
14727 (red filter on left eye, blue filter on right eye)
14730 anaglyph red/green gray
14731 (red filter on left eye, green filter on right eye)
14734 anaglyph red/cyan gray
14735 (red filter on left eye, cyan filter on right eye)
14738 anaglyph red/cyan half colored
14739 (red filter on left eye, cyan filter on right eye)
14742 anaglyph red/cyan color
14743 (red filter on left eye, cyan filter on right eye)
14746 anaglyph red/cyan color optimized with the least squares projection of dubois
14747 (red filter on left eye, cyan filter on right eye)
14750 anaglyph green/magenta gray
14751 (green filter on left eye, magenta filter on right eye)
14754 anaglyph green/magenta half colored
14755 (green filter on left eye, magenta filter on right eye)
14758 anaglyph green/magenta colored
14759 (green filter on left eye, magenta filter on right eye)
14762 anaglyph green/magenta color optimized with the least squares projection of dubois
14763 (green filter on left eye, magenta filter on right eye)
14766 anaglyph yellow/blue gray
14767 (yellow filter on left eye, blue filter on right eye)
14770 anaglyph yellow/blue half colored
14771 (yellow filter on left eye, blue filter on right eye)
14774 anaglyph yellow/blue colored
14775 (yellow filter on left eye, blue filter on right eye)
14778 anaglyph yellow/blue color optimized with the least squares projection of dubois
14779 (yellow filter on left eye, blue filter on right eye)
14782 mono output (left eye only)
14785 mono output (right eye only)
14788 checkerboard, left eye first
14791 checkerboard, right eye first
14794 interleaved columns, left eye first
14797 interleaved columns, right eye first
14803 Default value is @samp{arcd}.
14806 @subsection Examples
14810 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14816 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14822 @section streamselect, astreamselect
14823 Select video or audio streams.
14825 The filter accepts the following options:
14829 Set number of inputs. Default is 2.
14832 Set input indexes to remap to outputs.
14835 @subsection Commands
14837 The @code{streamselect} and @code{astreamselect} filter supports the following
14842 Set input indexes to remap to outputs.
14845 @subsection Examples
14849 Select first 5 seconds 1st stream and rest of time 2nd stream:
14851 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14855 Same as above, but for audio:
14857 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14862 Apply sobel operator to input video stream.
14864 The filter accepts the following option:
14868 Set which planes will be processed, unprocessed planes will be copied.
14869 By default value 0xf, all planes will be processed.
14872 Set value which will be multiplied with filtered result.
14875 Set value which will be added to filtered result.
14881 Apply a simple postprocessing filter that compresses and decompresses the image
14882 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14883 and average the results.
14885 The filter accepts the following options:
14889 Set quality. This option defines the number of levels for averaging. It accepts
14890 an integer in the range 0-6. If set to @code{0}, the filter will have no
14891 effect. A value of @code{6} means the higher quality. For each increment of
14892 that value the speed drops by a factor of approximately 2. Default value is
14896 Force a constant quantization parameter. If not set, the filter will use the QP
14897 from the video stream (if available).
14900 Set thresholding mode. Available modes are:
14904 Set hard thresholding (default).
14906 Set soft thresholding (better de-ringing effect, but likely blurrier).
14909 @item use_bframe_qp
14910 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
14911 option may cause flicker since the B-Frames have often larger QP. Default is
14912 @code{0} (not enabled).
14918 Draw subtitles on top of input video using the libass library.
14920 To enable compilation of this filter you need to configure FFmpeg with
14921 @code{--enable-libass}. This filter also requires a build with libavcodec and
14922 libavformat to convert the passed subtitles file to ASS (Advanced Substation
14923 Alpha) subtitles format.
14925 The filter accepts the following options:
14929 Set the filename of the subtitle file to read. It must be specified.
14931 @item original_size
14932 Specify the size of the original video, the video for which the ASS file
14933 was composed. For the syntax of this option, check the
14934 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14935 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
14936 correctly scale the fonts if the aspect ratio has been changed.
14939 Set a directory path containing fonts that can be used by the filter.
14940 These fonts will be used in addition to whatever the font provider uses.
14943 Process alpha channel, by default alpha channel is untouched.
14946 Set subtitles input character encoding. @code{subtitles} filter only. Only
14947 useful if not UTF-8.
14949 @item stream_index, si
14950 Set subtitles stream index. @code{subtitles} filter only.
14953 Override default style or script info parameters of the subtitles. It accepts a
14954 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
14957 If the first key is not specified, it is assumed that the first value
14958 specifies the @option{filename}.
14960 For example, to render the file @file{sub.srt} on top of the input
14961 video, use the command:
14966 which is equivalent to:
14968 subtitles=filename=sub.srt
14971 To render the default subtitles stream from file @file{video.mkv}, use:
14973 subtitles=video.mkv
14976 To render the second subtitles stream from that file, use:
14978 subtitles=video.mkv:si=1
14981 To make the subtitles stream from @file{sub.srt} appear in transparent green
14982 @code{DejaVu Serif}, use:
14984 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
14987 @section super2xsai
14989 Scale the input by 2x and smooth using the Super2xSaI (Scale and
14990 Interpolate) pixel art scaling algorithm.
14992 Useful for enlarging pixel art images without reducing sharpness.
14996 Swap two rectangular objects in video.
14998 This filter accepts the following options:
15008 Set 1st rect x coordinate.
15011 Set 1st rect y coordinate.
15014 Set 2nd rect x coordinate.
15017 Set 2nd rect y coordinate.
15019 All expressions are evaluated once for each frame.
15022 The all options are expressions containing the following constants:
15027 The input width and height.
15030 same as @var{w} / @var{h}
15033 input sample aspect ratio
15036 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15039 The number of the input frame, starting from 0.
15042 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15045 the position in the file of the input frame, NAN if unknown
15053 Apply telecine process to the video.
15055 This filter accepts the following options:
15064 The default value is @code{top}.
15068 A string of numbers representing the pulldown pattern you wish to apply.
15069 The default value is @code{23}.
15073 Some typical patterns:
15078 24p: 2332 (preferred)
15085 24p: 222222222223 ("Euro pulldown")
15092 Apply threshold effect to video stream.
15094 This filter needs four video streams to perform thresholding.
15095 First stream is stream we are filtering.
15096 Second stream is holding threshold values, third stream is holding min values,
15097 and last, fourth stream is holding max values.
15099 The filter accepts the following option:
15103 Set which planes will be processed, unprocessed planes will be copied.
15104 By default value 0xf, all planes will be processed.
15107 For example if first stream pixel's component value is less then threshold value
15108 of pixel component from 2nd threshold stream, third stream value will picked,
15109 otherwise fourth stream pixel component value will be picked.
15111 Using color source filter one can perform various types of thresholding:
15113 @subsection Examples
15117 Binary threshold, using gray color as threshold:
15119 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15123 Inverted binary threshold, using gray color as threshold:
15125 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15129 Truncate binary threshold, using gray color as threshold:
15131 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15135 Threshold to zero, using gray color as threshold:
15137 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15141 Inverted threshold to zero, using gray color as threshold:
15143 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15148 Select the most representative frame in a given sequence of consecutive frames.
15150 The filter accepts the following options:
15154 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
15155 will pick one of them, and then handle the next batch of @var{n} frames until
15156 the end. Default is @code{100}.
15159 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
15160 value will result in a higher memory usage, so a high value is not recommended.
15162 @subsection Examples
15166 Extract one picture each 50 frames:
15172 Complete example of a thumbnail creation with @command{ffmpeg}:
15174 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
15180 Tile several successive frames together.
15182 The filter accepts the following options:
15187 Set the grid size (i.e. the number of lines and columns). For the syntax of
15188 this option, check the
15189 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15192 Set the maximum number of frames to render in the given area. It must be less
15193 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
15194 the area will be used.
15197 Set the outer border margin in pixels.
15200 Set the inner border thickness (i.e. the number of pixels between frames). For
15201 more advanced padding options (such as having different values for the edges),
15202 refer to the pad video filter.
15205 Specify the color of the unused area. For the syntax of this option, check the
15206 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15207 The default value of @var{color} is "black".
15210 Set the number of frames to overlap when tiling several successive frames together.
15211 The value must be between @code{0} and @var{nb_frames - 1}.
15214 Set the number of frames to initially be empty before displaying first output frame.
15215 This controls how soon will one get first output frame.
15216 The value must be between @code{0} and @var{nb_frames - 1}.
15219 @subsection Examples
15223 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
15225 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
15227 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
15228 duplicating each output frame to accommodate the originally detected frame
15232 Display @code{5} pictures in an area of @code{3x2} frames,
15233 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
15234 mixed flat and named options:
15236 tile=3x2:nb_frames=5:padding=7:margin=2
15240 @section tinterlace
15242 Perform various types of temporal field interlacing.
15244 Frames are counted starting from 1, so the first input frame is
15247 The filter accepts the following options:
15252 Specify the mode of the interlacing. This option can also be specified
15253 as a value alone. See below for a list of values for this option.
15255 Available values are:
15259 Move odd frames into the upper field, even into the lower field,
15260 generating a double height frame at half frame rate.
15264 Frame 1 Frame 2 Frame 3 Frame 4
15266 11111 22222 33333 44444
15267 11111 22222 33333 44444
15268 11111 22222 33333 44444
15269 11111 22222 33333 44444
15283 Only output odd frames, even frames are dropped, generating a frame with
15284 unchanged height at half frame rate.
15289 Frame 1 Frame 2 Frame 3 Frame 4
15291 11111 22222 33333 44444
15292 11111 22222 33333 44444
15293 11111 22222 33333 44444
15294 11111 22222 33333 44444
15304 Only output even frames, odd frames are dropped, generating a frame with
15305 unchanged height at half frame rate.
15310 Frame 1 Frame 2 Frame 3 Frame 4
15312 11111 22222 33333 44444
15313 11111 22222 33333 44444
15314 11111 22222 33333 44444
15315 11111 22222 33333 44444
15325 Expand each frame to full height, but pad alternate lines with black,
15326 generating a frame with double height at the same input frame rate.
15331 Frame 1 Frame 2 Frame 3 Frame 4
15333 11111 22222 33333 44444
15334 11111 22222 33333 44444
15335 11111 22222 33333 44444
15336 11111 22222 33333 44444
15339 11111 ..... 33333 .....
15340 ..... 22222 ..... 44444
15341 11111 ..... 33333 .....
15342 ..... 22222 ..... 44444
15343 11111 ..... 33333 .....
15344 ..... 22222 ..... 44444
15345 11111 ..... 33333 .....
15346 ..... 22222 ..... 44444
15350 @item interleave_top, 4
15351 Interleave the upper field from odd frames with the lower field from
15352 even frames, generating a frame with unchanged height at half frame rate.
15357 Frame 1 Frame 2 Frame 3 Frame 4
15359 11111<- 22222 33333<- 44444
15360 11111 22222<- 33333 44444<-
15361 11111<- 22222 33333<- 44444
15362 11111 22222<- 33333 44444<-
15372 @item interleave_bottom, 5
15373 Interleave the lower field from odd frames with the upper field from
15374 even frames, generating a frame with unchanged height at half frame rate.
15379 Frame 1 Frame 2 Frame 3 Frame 4
15381 11111 22222<- 33333 44444<-
15382 11111<- 22222 33333<- 44444
15383 11111 22222<- 33333 44444<-
15384 11111<- 22222 33333<- 44444
15394 @item interlacex2, 6
15395 Double frame rate with unchanged height. Frames are inserted each
15396 containing the second temporal field from the previous input frame and
15397 the first temporal field from the next input frame. This mode relies on
15398 the top_field_first flag. Useful for interlaced video displays with no
15399 field synchronisation.
15404 Frame 1 Frame 2 Frame 3 Frame 4
15406 11111 22222 33333 44444
15407 11111 22222 33333 44444
15408 11111 22222 33333 44444
15409 11111 22222 33333 44444
15412 11111 22222 22222 33333 33333 44444 44444
15413 11111 11111 22222 22222 33333 33333 44444
15414 11111 22222 22222 33333 33333 44444 44444
15415 11111 11111 22222 22222 33333 33333 44444
15420 Move odd frames into the upper field, even into the lower field,
15421 generating a double height frame at same frame rate.
15426 Frame 1 Frame 2 Frame 3 Frame 4
15428 11111 22222 33333 44444
15429 11111 22222 33333 44444
15430 11111 22222 33333 44444
15431 11111 22222 33333 44444
15434 11111 33333 33333 55555
15435 22222 22222 44444 44444
15436 11111 33333 33333 55555
15437 22222 22222 44444 44444
15438 11111 33333 33333 55555
15439 22222 22222 44444 44444
15440 11111 33333 33333 55555
15441 22222 22222 44444 44444
15446 Numeric values are deprecated but are accepted for backward
15447 compatibility reasons.
15449 Default mode is @code{merge}.
15452 Specify flags influencing the filter process.
15454 Available value for @var{flags} is:
15457 @item low_pass_filter, vlfp
15458 Enable linear vertical low-pass filtering in the filter.
15459 Vertical low-pass filtering is required when creating an interlaced
15460 destination from a progressive source which contains high-frequency
15461 vertical detail. Filtering will reduce interlace 'twitter' and Moire
15464 @item complex_filter, cvlfp
15465 Enable complex vertical low-pass filtering.
15466 This will slightly less reduce interlace 'twitter' and Moire
15467 patterning but better retain detail and subjective sharpness impression.
15471 Vertical low-pass filtering can only be enabled for @option{mode}
15472 @var{interleave_top} and @var{interleave_bottom}.
15477 Tone map colors from different dynamic ranges.
15479 This filter expects data in single precision floating point, as it needs to
15480 operate on (and can output) out-of-range values. Another filter, such as
15481 @ref{zscale}, is needed to convert the resulting frame to a usable format.
15483 The tonemapping algorithms implemented only work on linear light, so input
15484 data should be linearized beforehand (and possibly correctly tagged).
15487 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
15490 @subsection Options
15491 The filter accepts the following options.
15495 Set the tone map algorithm to use.
15497 Possible values are:
15500 Do not apply any tone map, only desaturate overbright pixels.
15503 Hard-clip any out-of-range values. Use it for perfect color accuracy for
15504 in-range values, while distorting out-of-range values.
15507 Stretch the entire reference gamut to a linear multiple of the display.
15510 Fit a logarithmic transfer between the tone curves.
15513 Preserve overall image brightness with a simple curve, using nonlinear
15514 contrast, which results in flattening details and degrading color accuracy.
15517 Preserve both dark and bright details better than @var{reinhard}, at the cost
15518 of slightly darkening everything. Use it when detail preservation is more
15519 important than color and brightness accuracy.
15522 Smoothly map out-of-range values, while retaining contrast and colors for
15523 in-range material as much as possible. Use it when color accuracy is more
15524 important than detail preservation.
15530 Tune the tone mapping algorithm.
15532 This affects the following algorithms:
15538 Specifies the scale factor to use while stretching.
15542 Specifies the exponent of the function.
15546 Specify an extra linear coefficient to multiply into the signal before clipping.
15550 Specify the local contrast coefficient at the display peak.
15551 Default to 0.5, which means that in-gamut values will be about half as bright
15558 Specify the transition point from linear to mobius transform. Every value
15559 below this point is guaranteed to be mapped 1:1. The higher the value, the
15560 more accurate the result will be, at the cost of losing bright details.
15561 Default to 0.3, which due to the steep initial slope still preserves in-range
15562 colors fairly accurately.
15566 Apply desaturation for highlights that exceed this level of brightness. The
15567 higher the parameter, the more color information will be preserved. This
15568 setting helps prevent unnaturally blown-out colors for super-highlights, by
15569 (smoothly) turning into white instead. This makes images feel more natural,
15570 at the cost of reducing information about out-of-range colors.
15572 The default of 2.0 is somewhat conservative and will mostly just apply to
15573 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
15575 This option works only if the input frame has a supported color tag.
15578 Override signal/nominal/reference peak with this value. Useful when the
15579 embedded peak information in display metadata is not reliable or when tone
15580 mapping from a lower range to a higher range.
15585 Transpose rows with columns in the input video and optionally flip it.
15587 It accepts the following parameters:
15592 Specify the transposition direction.
15594 Can assume the following values:
15596 @item 0, 4, cclock_flip
15597 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
15605 Rotate by 90 degrees clockwise, that is:
15613 Rotate by 90 degrees counterclockwise, that is:
15620 @item 3, 7, clock_flip
15621 Rotate by 90 degrees clockwise and vertically flip, that is:
15629 For values between 4-7, the transposition is only done if the input
15630 video geometry is portrait and not landscape. These values are
15631 deprecated, the @code{passthrough} option should be used instead.
15633 Numerical values are deprecated, and should be dropped in favor of
15634 symbolic constants.
15637 Do not apply the transposition if the input geometry matches the one
15638 specified by the specified value. It accepts the following values:
15641 Always apply transposition.
15643 Preserve portrait geometry (when @var{height} >= @var{width}).
15645 Preserve landscape geometry (when @var{width} >= @var{height}).
15648 Default value is @code{none}.
15651 For example to rotate by 90 degrees clockwise and preserve portrait
15654 transpose=dir=1:passthrough=portrait
15657 The command above can also be specified as:
15659 transpose=1:portrait
15663 Trim the input so that the output contains one continuous subpart of the input.
15665 It accepts the following parameters:
15668 Specify the time of the start of the kept section, i.e. the frame with the
15669 timestamp @var{start} will be the first frame in the output.
15672 Specify the time of the first frame that will be dropped, i.e. the frame
15673 immediately preceding the one with the timestamp @var{end} will be the last
15674 frame in the output.
15677 This is the same as @var{start}, except this option sets the start timestamp
15678 in timebase units instead of seconds.
15681 This is the same as @var{end}, except this option sets the end timestamp
15682 in timebase units instead of seconds.
15685 The maximum duration of the output in seconds.
15688 The number of the first frame that should be passed to the output.
15691 The number of the first frame that should be dropped.
15694 @option{start}, @option{end}, and @option{duration} are expressed as time
15695 duration specifications; see
15696 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15697 for the accepted syntax.
15699 Note that the first two sets of the start/end options and the @option{duration}
15700 option look at the frame timestamp, while the _frame variants simply count the
15701 frames that pass through the filter. Also note that this filter does not modify
15702 the timestamps. If you wish for the output timestamps to start at zero, insert a
15703 setpts filter after the trim filter.
15705 If multiple start or end options are set, this filter tries to be greedy and
15706 keep all the frames that match at least one of the specified constraints. To keep
15707 only the part that matches all the constraints at once, chain multiple trim
15710 The defaults are such that all the input is kept. So it is possible to set e.g.
15711 just the end values to keep everything before the specified time.
15716 Drop everything except the second minute of input:
15718 ffmpeg -i INPUT -vf trim=60:120
15722 Keep only the first second:
15724 ffmpeg -i INPUT -vf trim=duration=1
15729 @section unpremultiply
15730 Apply alpha unpremultiply effect to input video stream using first plane
15731 of second stream as alpha.
15733 Both streams must have same dimensions and same pixel format.
15735 The filter accepts the following option:
15739 Set which planes will be processed, unprocessed planes will be copied.
15740 By default value 0xf, all planes will be processed.
15742 If the format has 1 or 2 components, then luma is bit 0.
15743 If the format has 3 or 4 components:
15744 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15745 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15746 If present, the alpha channel is always the last bit.
15749 Do not require 2nd input for processing, instead use alpha plane from input stream.
15755 Sharpen or blur the input video.
15757 It accepts the following parameters:
15760 @item luma_msize_x, lx
15761 Set the luma matrix horizontal size. It must be an odd integer between
15762 3 and 23. The default value is 5.
15764 @item luma_msize_y, ly
15765 Set the luma matrix vertical size. It must be an odd integer between 3
15766 and 23. The default value is 5.
15768 @item luma_amount, la
15769 Set the luma effect strength. It must be a floating point number, reasonable
15770 values lay between -1.5 and 1.5.
15772 Negative values will blur the input video, while positive values will
15773 sharpen it, a value of zero will disable the effect.
15775 Default value is 1.0.
15777 @item chroma_msize_x, cx
15778 Set the chroma matrix horizontal size. It must be an odd integer
15779 between 3 and 23. The default value is 5.
15781 @item chroma_msize_y, cy
15782 Set the chroma matrix vertical size. It must be an odd integer
15783 between 3 and 23. The default value is 5.
15785 @item chroma_amount, ca
15786 Set the chroma effect strength. It must be a floating point number, reasonable
15787 values lay between -1.5 and 1.5.
15789 Negative values will blur the input video, while positive values will
15790 sharpen it, a value of zero will disable the effect.
15792 Default value is 0.0.
15796 All parameters are optional and default to the equivalent of the
15797 string '5:5:1.0:5:5:0.0'.
15799 @subsection Examples
15803 Apply strong luma sharpen effect:
15805 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15809 Apply a strong blur of both luma and chroma parameters:
15811 unsharp=7:7:-2:7:7:-2
15817 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15818 the image at several (or - in the case of @option{quality} level @code{8} - all)
15819 shifts and average the results.
15821 The way this differs from the behavior of spp is that uspp actually encodes &
15822 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15823 DCT similar to MJPEG.
15825 The filter accepts the following options:
15829 Set quality. This option defines the number of levels for averaging. It accepts
15830 an integer in the range 0-8. If set to @code{0}, the filter will have no
15831 effect. A value of @code{8} means the higher quality. For each increment of
15832 that value the speed drops by a factor of approximately 2. Default value is
15836 Force a constant quantization parameter. If not set, the filter will use the QP
15837 from the video stream (if available).
15840 @section vaguedenoiser
15842 Apply a wavelet based denoiser.
15844 It transforms each frame from the video input into the wavelet domain,
15845 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15846 the obtained coefficients. It does an inverse wavelet transform after.
15847 Due to wavelet properties, it should give a nice smoothed result, and
15848 reduced noise, without blurring picture features.
15850 This filter accepts the following options:
15854 The filtering strength. The higher, the more filtered the video will be.
15855 Hard thresholding can use a higher threshold than soft thresholding
15856 before the video looks overfiltered. Default value is 2.
15859 The filtering method the filter will use.
15861 It accepts the following values:
15864 All values under the threshold will be zeroed.
15867 All values under the threshold will be zeroed. All values above will be
15868 reduced by the threshold.
15871 Scales or nullifies coefficients - intermediary between (more) soft and
15872 (less) hard thresholding.
15875 Default is garrote.
15878 Number of times, the wavelet will decompose the picture. Picture can't
15879 be decomposed beyond a particular point (typically, 8 for a 640x480
15880 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15883 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
15886 A list of the planes to process. By default all planes are processed.
15889 @section vectorscope
15891 Display 2 color component values in the two dimensional graph (which is called
15894 This filter accepts the following options:
15898 Set vectorscope mode.
15900 It accepts the following values:
15903 Gray values are displayed on graph, higher brightness means more pixels have
15904 same component color value on location in graph. This is the default mode.
15907 Gray values are displayed on graph. Surrounding pixels values which are not
15908 present in video frame are drawn in gradient of 2 color components which are
15909 set by option @code{x} and @code{y}. The 3rd color component is static.
15912 Actual color components values present in video frame are displayed on graph.
15915 Similar as color2 but higher frequency of same values @code{x} and @code{y}
15916 on graph increases value of another color component, which is luminance by
15917 default values of @code{x} and @code{y}.
15920 Actual colors present in video frame are displayed on graph. If two different
15921 colors map to same position on graph then color with higher value of component
15922 not present in graph is picked.
15925 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
15926 component picked from radial gradient.
15930 Set which color component will be represented on X-axis. Default is @code{1}.
15933 Set which color component will be represented on Y-axis. Default is @code{2}.
15936 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
15937 of color component which represents frequency of (X, Y) location in graph.
15942 No envelope, this is default.
15945 Instant envelope, even darkest single pixel will be clearly highlighted.
15948 Hold maximum and minimum values presented in graph over time. This way you
15949 can still spot out of range values without constantly looking at vectorscope.
15952 Peak and instant envelope combined together.
15956 Set what kind of graticule to draw.
15964 Set graticule opacity.
15967 Set graticule flags.
15971 Draw graticule for white point.
15974 Draw graticule for black point.
15977 Draw color points short names.
15981 Set background opacity.
15983 @item lthreshold, l
15984 Set low threshold for color component not represented on X or Y axis.
15985 Values lower than this value will be ignored. Default is 0.
15986 Note this value is multiplied with actual max possible value one pixel component
15987 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
15990 @item hthreshold, h
15991 Set high threshold for color component not represented on X or Y axis.
15992 Values higher than this value will be ignored. Default is 1.
15993 Note this value is multiplied with actual max possible value one pixel component
15994 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
15995 is 0.9 * 255 = 230.
15997 @item colorspace, c
15998 Set what kind of colorspace to use when drawing graticule.
16007 @anchor{vidstabdetect}
16008 @section vidstabdetect
16010 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
16011 @ref{vidstabtransform} for pass 2.
16013 This filter generates a file with relative translation and rotation
16014 transform information about subsequent frames, which is then used by
16015 the @ref{vidstabtransform} filter.
16017 To enable compilation of this filter you need to configure FFmpeg with
16018 @code{--enable-libvidstab}.
16020 This filter accepts the following options:
16024 Set the path to the file used to write the transforms information.
16025 Default value is @file{transforms.trf}.
16028 Set how shaky the video is and how quick the camera is. It accepts an
16029 integer in the range 1-10, a value of 1 means little shakiness, a
16030 value of 10 means strong shakiness. Default value is 5.
16033 Set the accuracy of the detection process. It must be a value in the
16034 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16035 accuracy. Default value is 15.
16038 Set stepsize of the search process. The region around minimum is
16039 scanned with 1 pixel resolution. Default value is 6.
16042 Set minimum contrast. Below this value a local measurement field is
16043 discarded. Must be a floating point value in the range 0-1. Default
16047 Set reference frame number for tripod mode.
16049 If enabled, the motion of the frames is compared to a reference frame
16050 in the filtered stream, identified by the specified number. The idea
16051 is to compensate all movements in a more-or-less static scene and keep
16052 the camera view absolutely still.
16054 If set to 0, it is disabled. The frames are counted starting from 1.
16057 Show fields and transforms in the resulting frames. It accepts an
16058 integer in the range 0-2. Default value is 0, which disables any
16062 @subsection Examples
16066 Use default values:
16072 Analyze strongly shaky movie and put the results in file
16073 @file{mytransforms.trf}:
16075 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
16079 Visualize the result of internal transformations in the resulting
16082 vidstabdetect=show=1
16086 Analyze a video with medium shakiness using @command{ffmpeg}:
16088 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
16092 @anchor{vidstabtransform}
16093 @section vidstabtransform
16095 Video stabilization/deshaking: pass 2 of 2,
16096 see @ref{vidstabdetect} for pass 1.
16098 Read a file with transform information for each frame and
16099 apply/compensate them. Together with the @ref{vidstabdetect}
16100 filter this can be used to deshake videos. See also
16101 @url{http://public.hronopik.de/vid.stab}. It is important to also use
16102 the @ref{unsharp} filter, see below.
16104 To enable compilation of this filter you need to configure FFmpeg with
16105 @code{--enable-libvidstab}.
16107 @subsection Options
16111 Set path to the file used to read the transforms. Default value is
16112 @file{transforms.trf}.
16115 Set the number of frames (value*2 + 1) used for lowpass filtering the
16116 camera movements. Default value is 10.
16118 For example a number of 10 means that 21 frames are used (10 in the
16119 past and 10 in the future) to smoothen the motion in the video. A
16120 larger value leads to a smoother video, but limits the acceleration of
16121 the camera (pan/tilt movements). 0 is a special case where a static
16122 camera is simulated.
16125 Set the camera path optimization algorithm.
16127 Accepted values are:
16130 gaussian kernel low-pass filter on camera motion (default)
16132 averaging on transformations
16136 Set maximal number of pixels to translate frames. Default value is -1,
16140 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
16141 value is -1, meaning no limit.
16144 Specify how to deal with borders that may be visible due to movement
16147 Available values are:
16150 keep image information from previous frame (default)
16152 fill the border black
16156 Invert transforms if set to 1. Default value is 0.
16159 Consider transforms as relative to previous frame if set to 1,
16160 absolute if set to 0. Default value is 0.
16163 Set percentage to zoom. A positive value will result in a zoom-in
16164 effect, a negative value in a zoom-out effect. Default value is 0 (no
16168 Set optimal zooming to avoid borders.
16170 Accepted values are:
16175 optimal static zoom value is determined (only very strong movements
16176 will lead to visible borders) (default)
16178 optimal adaptive zoom value is determined (no borders will be
16179 visible), see @option{zoomspeed}
16182 Note that the value given at zoom is added to the one calculated here.
16185 Set percent to zoom maximally each frame (enabled when
16186 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
16190 Specify type of interpolation.
16192 Available values are:
16197 linear only horizontal
16199 linear in both directions (default)
16201 cubic in both directions (slow)
16205 Enable virtual tripod mode if set to 1, which is equivalent to
16206 @code{relative=0:smoothing=0}. Default value is 0.
16208 Use also @code{tripod} option of @ref{vidstabdetect}.
16211 Increase log verbosity if set to 1. Also the detected global motions
16212 are written to the temporary file @file{global_motions.trf}. Default
16216 @subsection Examples
16220 Use @command{ffmpeg} for a typical stabilization with default values:
16222 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
16225 Note the use of the @ref{unsharp} filter which is always recommended.
16228 Zoom in a bit more and load transform data from a given file:
16230 vidstabtransform=zoom=5:input="mytransforms.trf"
16234 Smoothen the video even more:
16236 vidstabtransform=smoothing=30
16242 Flip the input video vertically.
16244 For example, to vertically flip a video with @command{ffmpeg}:
16246 ffmpeg -i in.avi -vf "vflip" out.avi
16252 Make or reverse a natural vignetting effect.
16254 The filter accepts the following options:
16258 Set lens angle expression as a number of radians.
16260 The value is clipped in the @code{[0,PI/2]} range.
16262 Default value: @code{"PI/5"}
16266 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
16270 Set forward/backward mode.
16272 Available modes are:
16275 The larger the distance from the central point, the darker the image becomes.
16278 The larger the distance from the central point, the brighter the image becomes.
16279 This can be used to reverse a vignette effect, though there is no automatic
16280 detection to extract the lens @option{angle} and other settings (yet). It can
16281 also be used to create a burning effect.
16284 Default value is @samp{forward}.
16287 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
16289 It accepts the following values:
16292 Evaluate expressions only once during the filter initialization.
16295 Evaluate expressions for each incoming frame. This is way slower than the
16296 @samp{init} mode since it requires all the scalers to be re-computed, but it
16297 allows advanced dynamic expressions.
16300 Default value is @samp{init}.
16303 Set dithering to reduce the circular banding effects. Default is @code{1}
16307 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
16308 Setting this value to the SAR of the input will make a rectangular vignetting
16309 following the dimensions of the video.
16311 Default is @code{1/1}.
16314 @subsection Expressions
16316 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
16317 following parameters.
16322 input width and height
16325 the number of input frame, starting from 0
16328 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
16329 @var{TB} units, NAN if undefined
16332 frame rate of the input video, NAN if the input frame rate is unknown
16335 the PTS (Presentation TimeStamp) of the filtered video frame,
16336 expressed in seconds, NAN if undefined
16339 time base of the input video
16343 @subsection Examples
16347 Apply simple strong vignetting effect:
16353 Make a flickering vignetting:
16355 vignette='PI/4+random(1)*PI/50':eval=frame
16360 @section vmafmotion
16362 Obtain the average vmaf motion score of a video.
16363 It is one of the component filters of VMAF.
16365 The obtained average motion score is printed through the logging system.
16367 In the below example the input file @file{ref.mpg} is being processed and score
16371 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
16375 Stack input videos vertically.
16377 All streams must be of same pixel format and of same width.
16379 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
16380 to create same output.
16382 The filter accept the following option:
16386 Set number of input streams. Default is 2.
16389 If set to 1, force the output to terminate when the shortest input
16390 terminates. Default value is 0.
16395 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
16396 Deinterlacing Filter").
16398 Based on the process described by Martin Weston for BBC R&D, and
16399 implemented based on the de-interlace algorithm written by Jim
16400 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
16401 uses filter coefficients calculated by BBC R&D.
16403 There are two sets of filter coefficients, so called "simple":
16404 and "complex". Which set of filter coefficients is used can
16405 be set by passing an optional parameter:
16409 Set the interlacing filter coefficients. Accepts one of the following values:
16413 Simple filter coefficient set.
16415 More-complex filter coefficient set.
16417 Default value is @samp{complex}.
16420 Specify which frames to deinterlace. Accept one of the following values:
16424 Deinterlace all frames,
16426 Only deinterlace frames marked as interlaced.
16429 Default value is @samp{all}.
16433 Video waveform monitor.
16435 The waveform monitor plots color component intensity. By default luminance
16436 only. Each column of the waveform corresponds to a column of pixels in the
16439 It accepts the following options:
16443 Can be either @code{row}, or @code{column}. Default is @code{column}.
16444 In row mode, the graph on the left side represents color component value 0 and
16445 the right side represents value = 255. In column mode, the top side represents
16446 color component value = 0 and bottom side represents value = 255.
16449 Set intensity. Smaller values are useful to find out how many values of the same
16450 luminance are distributed across input rows/columns.
16451 Default value is @code{0.04}. Allowed range is [0, 1].
16454 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
16455 In mirrored mode, higher values will be represented on the left
16456 side for @code{row} mode and at the top for @code{column} mode. Default is
16457 @code{1} (mirrored).
16461 It accepts the following values:
16464 Presents information identical to that in the @code{parade}, except
16465 that the graphs representing color components are superimposed directly
16468 This display mode makes it easier to spot relative differences or similarities
16469 in overlapping areas of the color components that are supposed to be identical,
16470 such as neutral whites, grays, or blacks.
16473 Display separate graph for the color components side by side in
16474 @code{row} mode or one below the other in @code{column} mode.
16477 Display separate graph for the color components side by side in
16478 @code{column} mode or one below the other in @code{row} mode.
16480 Using this display mode makes it easy to spot color casts in the highlights
16481 and shadows of an image, by comparing the contours of the top and the bottom
16482 graphs of each waveform. Since whites, grays, and blacks are characterized
16483 by exactly equal amounts of red, green, and blue, neutral areas of the picture
16484 should display three waveforms of roughly equal width/height. If not, the
16485 correction is easy to perform by making level adjustments the three waveforms.
16487 Default is @code{stack}.
16489 @item components, c
16490 Set which color components to display. Default is 1, which means only luminance
16491 or red color component if input is in RGB colorspace. If is set for example to
16492 7 it will display all 3 (if) available color components.
16497 No envelope, this is default.
16500 Instant envelope, minimum and maximum values presented in graph will be easily
16501 visible even with small @code{step} value.
16504 Hold minimum and maximum values presented in graph across time. This way you
16505 can still spot out of range values without constantly looking at waveforms.
16508 Peak and instant envelope combined together.
16514 No filtering, this is default.
16517 Luma and chroma combined together.
16520 Similar as above, but shows difference between blue and red chroma.
16523 Similar as above, but use different colors.
16526 Displays only chroma.
16529 Displays actual color value on waveform.
16532 Similar as above, but with luma showing frequency of chroma values.
16536 Set which graticule to display.
16540 Do not display graticule.
16543 Display green graticule showing legal broadcast ranges.
16546 Display orange graticule showing legal broadcast ranges.
16550 Set graticule opacity.
16553 Set graticule flags.
16557 Draw numbers above lines. By default enabled.
16560 Draw dots instead of lines.
16564 Set scale used for displaying graticule.
16571 Default is digital.
16574 Set background opacity.
16577 @section weave, doubleweave
16579 The @code{weave} takes a field-based video input and join
16580 each two sequential fields into single frame, producing a new double
16581 height clip with half the frame rate and half the frame count.
16583 The @code{doubleweave} works same as @code{weave} but without
16584 halving frame rate and frame count.
16586 It accepts the following option:
16590 Set first field. Available values are:
16594 Set the frame as top-field-first.
16597 Set the frame as bottom-field-first.
16601 @subsection Examples
16605 Interlace video using @ref{select} and @ref{separatefields} filter:
16607 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
16612 Apply the xBR high-quality magnification filter which is designed for pixel
16613 art. It follows a set of edge-detection rules, see
16614 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
16616 It accepts the following option:
16620 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
16621 @code{3xBR} and @code{4} for @code{4xBR}.
16622 Default is @code{3}.
16628 Deinterlace the input video ("yadif" means "yet another deinterlacing
16631 It accepts the following parameters:
16637 The interlacing mode to adopt. It accepts one of the following values:
16640 @item 0, send_frame
16641 Output one frame for each frame.
16642 @item 1, send_field
16643 Output one frame for each field.
16644 @item 2, send_frame_nospatial
16645 Like @code{send_frame}, but it skips the spatial interlacing check.
16646 @item 3, send_field_nospatial
16647 Like @code{send_field}, but it skips the spatial interlacing check.
16650 The default value is @code{send_frame}.
16653 The picture field parity assumed for the input interlaced video. It accepts one
16654 of the following values:
16658 Assume the top field is first.
16660 Assume the bottom field is first.
16662 Enable automatic detection of field parity.
16665 The default value is @code{auto}.
16666 If the interlacing is unknown or the decoder does not export this information,
16667 top field first will be assumed.
16670 Specify which frames to deinterlace. Accept one of the following
16675 Deinterlace all frames.
16676 @item 1, interlaced
16677 Only deinterlace frames marked as interlaced.
16680 The default value is @code{all}.
16685 Apply Zoom & Pan effect.
16687 This filter accepts the following options:
16691 Set the zoom expression. Default is 1.
16695 Set the x and y expression. Default is 0.
16698 Set the duration expression in number of frames.
16699 This sets for how many number of frames effect will last for
16700 single input image.
16703 Set the output image size, default is 'hd720'.
16706 Set the output frame rate, default is '25'.
16709 Each expression can contain the following constants:
16728 Output frame count.
16732 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16733 for current input frame.
16737 'x' and 'y' of last output frame of previous input frame or 0 when there was
16738 not yet such frame (first input frame).
16741 Last calculated zoom from 'z' expression for current input frame.
16744 Last calculated zoom of last output frame of previous input frame.
16747 Number of output frames for current input frame. Calculated from 'd' expression
16748 for each input frame.
16751 number of output frames created for previous input frame
16754 Rational number: input width / input height
16757 sample aspect ratio
16760 display aspect ratio
16764 @subsection Examples
16768 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
16770 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
16774 Zoom-in up to 1.5 and pan always at center of picture:
16776 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16780 Same as above but without pausing:
16782 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16788 Scale (resize) the input video, using the z.lib library:
16789 https://github.com/sekrit-twc/zimg.
16791 The zscale filter forces the output display aspect ratio to be the same
16792 as the input, by changing the output sample aspect ratio.
16794 If the input image format is different from the format requested by
16795 the next filter, the zscale filter will convert the input to the
16798 @subsection Options
16799 The filter accepts the following options.
16804 Set the output video dimension expression. Default value is the input
16807 If the @var{width} or @var{w} value is 0, the input width is used for
16808 the output. If the @var{height} or @var{h} value is 0, the input height
16809 is used for the output.
16811 If one and only one of the values is -n with n >= 1, the zscale filter
16812 will use a value that maintains the aspect ratio of the input image,
16813 calculated from the other specified dimension. After that it will,
16814 however, make sure that the calculated dimension is divisible by n and
16815 adjust the value if necessary.
16817 If both values are -n with n >= 1, the behavior will be identical to
16818 both values being set to 0 as previously detailed.
16820 See below for the list of accepted constants for use in the dimension
16824 Set the video size. For the syntax of this option, check the
16825 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16828 Set the dither type.
16830 Possible values are:
16835 @item error_diffusion
16841 Set the resize filter type.
16843 Possible values are:
16853 Default is bilinear.
16856 Set the color range.
16858 Possible values are:
16865 Default is same as input.
16868 Set the color primaries.
16870 Possible values are:
16880 Default is same as input.
16883 Set the transfer characteristics.
16885 Possible values are:
16899 Default is same as input.
16902 Set the colorspace matrix.
16904 Possible value are:
16915 Default is same as input.
16918 Set the input color range.
16920 Possible values are:
16927 Default is same as input.
16929 @item primariesin, pin
16930 Set the input color primaries.
16932 Possible values are:
16942 Default is same as input.
16944 @item transferin, tin
16945 Set the input transfer characteristics.
16947 Possible values are:
16958 Default is same as input.
16960 @item matrixin, min
16961 Set the input colorspace matrix.
16963 Possible value are:
16975 Set the output chroma location.
16977 Possible values are:
16988 @item chromalin, cin
16989 Set the input chroma location.
16991 Possible values are:
17003 Set the nominal peak luminance.
17006 The values of the @option{w} and @option{h} options are expressions
17007 containing the following constants:
17012 The input width and height
17016 These are the same as @var{in_w} and @var{in_h}.
17020 The output (scaled) width and height
17024 These are the same as @var{out_w} and @var{out_h}
17027 The same as @var{iw} / @var{ih}
17030 input sample aspect ratio
17033 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17037 horizontal and vertical input chroma subsample values. For example for the
17038 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17042 horizontal and vertical output chroma subsample values. For example for the
17043 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17049 @c man end VIDEO FILTERS
17051 @chapter Video Sources
17052 @c man begin VIDEO SOURCES
17054 Below is a description of the currently available video sources.
17058 Buffer video frames, and make them available to the filter chain.
17060 This source is mainly intended for a programmatic use, in particular
17061 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
17063 It accepts the following parameters:
17068 Specify the size (width and height) of the buffered video frames. For the
17069 syntax of this option, check the
17070 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17073 The input video width.
17076 The input video height.
17079 A string representing the pixel format of the buffered video frames.
17080 It may be a number corresponding to a pixel format, or a pixel format
17084 Specify the timebase assumed by the timestamps of the buffered frames.
17087 Specify the frame rate expected for the video stream.
17089 @item pixel_aspect, sar
17090 The sample (pixel) aspect ratio of the input video.
17093 Specify the optional parameters to be used for the scale filter which
17094 is automatically inserted when an input change is detected in the
17095 input size or format.
17097 @item hw_frames_ctx
17098 When using a hardware pixel format, this should be a reference to an
17099 AVHWFramesContext describing input frames.
17104 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
17107 will instruct the source to accept video frames with size 320x240 and
17108 with format "yuv410p", assuming 1/24 as the timestamps timebase and
17109 square pixels (1:1 sample aspect ratio).
17110 Since the pixel format with name "yuv410p" corresponds to the number 6
17111 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
17112 this example corresponds to:
17114 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
17117 Alternatively, the options can be specified as a flat string, but this
17118 syntax is deprecated:
17120 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}[:@var{sws_param}]
17124 Create a pattern generated by an elementary cellular automaton.
17126 The initial state of the cellular automaton can be defined through the
17127 @option{filename} and @option{pattern} options. If such options are
17128 not specified an initial state is created randomly.
17130 At each new frame a new row in the video is filled with the result of
17131 the cellular automaton next generation. The behavior when the whole
17132 frame is filled is defined by the @option{scroll} option.
17134 This source accepts the following options:
17138 Read the initial cellular automaton state, i.e. the starting row, from
17139 the specified file.
17140 In the file, each non-whitespace character is considered an alive
17141 cell, a newline will terminate the row, and further characters in the
17142 file will be ignored.
17145 Read the initial cellular automaton state, i.e. the starting row, from
17146 the specified string.
17148 Each non-whitespace character in the string is considered an alive
17149 cell, a newline will terminate the row, and further characters in the
17150 string will be ignored.
17153 Set the video rate, that is the number of frames generated per second.
17156 @item random_fill_ratio, ratio
17157 Set the random fill ratio for the initial cellular automaton row. It
17158 is a floating point number value ranging from 0 to 1, defaults to
17161 This option is ignored when a file or a pattern is specified.
17163 @item random_seed, seed
17164 Set the seed for filling randomly the initial row, must be an integer
17165 included between 0 and UINT32_MAX. If not specified, or if explicitly
17166 set to -1, the filter will try to use a good random seed on a best
17170 Set the cellular automaton rule, it is a number ranging from 0 to 255.
17171 Default value is 110.
17174 Set the size of the output video. For the syntax of this option, check the
17175 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17177 If @option{filename} or @option{pattern} is specified, the size is set
17178 by default to the width of the specified initial state row, and the
17179 height is set to @var{width} * PHI.
17181 If @option{size} is set, it must contain the width of the specified
17182 pattern string, and the specified pattern will be centered in the
17185 If a filename or a pattern string is not specified, the size value
17186 defaults to "320x518" (used for a randomly generated initial state).
17189 If set to 1, scroll the output upward when all the rows in the output
17190 have been already filled. If set to 0, the new generated row will be
17191 written over the top row just after the bottom row is filled.
17194 @item start_full, full
17195 If set to 1, completely fill the output with generated rows before
17196 outputting the first frame.
17197 This is the default behavior, for disabling set the value to 0.
17200 If set to 1, stitch the left and right row edges together.
17201 This is the default behavior, for disabling set the value to 0.
17204 @subsection Examples
17208 Read the initial state from @file{pattern}, and specify an output of
17211 cellauto=f=pattern:s=200x400
17215 Generate a random initial row with a width of 200 cells, with a fill
17218 cellauto=ratio=2/3:s=200x200
17222 Create a pattern generated by rule 18 starting by a single alive cell
17223 centered on an initial row with width 100:
17225 cellauto=p=@@:s=100x400:full=0:rule=18
17229 Specify a more elaborated initial pattern:
17231 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
17236 @anchor{coreimagesrc}
17237 @section coreimagesrc
17238 Video source generated on GPU using Apple's CoreImage API on OSX.
17240 This video source is a specialized version of the @ref{coreimage} video filter.
17241 Use a core image generator at the beginning of the applied filterchain to
17242 generate the content.
17244 The coreimagesrc video source accepts the following options:
17246 @item list_generators
17247 List all available generators along with all their respective options as well as
17248 possible minimum and maximum values along with the default values.
17250 list_generators=true
17254 Specify the size of the sourced video. For the syntax of this option, check the
17255 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17256 The default value is @code{320x240}.
17259 Specify the frame rate of the sourced video, as the number of frames
17260 generated per second. It has to be a string in the format
17261 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17262 number or a valid video frame rate abbreviation. The default value is
17266 Set the sample aspect ratio of the sourced video.
17269 Set the duration of the sourced video. See
17270 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17271 for the accepted syntax.
17273 If not specified, or the expressed duration is negative, the video is
17274 supposed to be generated forever.
17277 Additionally, all options of the @ref{coreimage} video filter are accepted.
17278 A complete filterchain can be used for further processing of the
17279 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
17280 and examples for details.
17282 @subsection Examples
17287 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
17288 given as complete and escaped command-line for Apple's standard bash shell:
17290 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
17292 This example is equivalent to the QRCode example of @ref{coreimage} without the
17293 need for a nullsrc video source.
17297 @section mandelbrot
17299 Generate a Mandelbrot set fractal, and progressively zoom towards the
17300 point specified with @var{start_x} and @var{start_y}.
17302 This source accepts the following options:
17307 Set the terminal pts value. Default value is 400.
17310 Set the terminal scale value.
17311 Must be a floating point value. Default value is 0.3.
17314 Set the inner coloring mode, that is the algorithm used to draw the
17315 Mandelbrot fractal internal region.
17317 It shall assume one of the following values:
17322 Show time until convergence.
17324 Set color based on point closest to the origin of the iterations.
17329 Default value is @var{mincol}.
17332 Set the bailout value. Default value is 10.0.
17335 Set the maximum of iterations performed by the rendering
17336 algorithm. Default value is 7189.
17339 Set outer coloring mode.
17340 It shall assume one of following values:
17342 @item iteration_count
17343 Set iteration cound mode.
17344 @item normalized_iteration_count
17345 set normalized iteration count mode.
17347 Default value is @var{normalized_iteration_count}.
17350 Set frame rate, expressed as number of frames per second. Default
17354 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
17355 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
17358 Set the initial scale value. Default value is 3.0.
17361 Set the initial x position. Must be a floating point value between
17362 -100 and 100. Default value is -0.743643887037158704752191506114774.
17365 Set the initial y position. Must be a floating point value between
17366 -100 and 100. Default value is -0.131825904205311970493132056385139.
17371 Generate various test patterns, as generated by the MPlayer test filter.
17373 The size of the generated video is fixed, and is 256x256.
17374 This source is useful in particular for testing encoding features.
17376 This source accepts the following options:
17381 Specify the frame rate of the sourced video, as the number of frames
17382 generated per second. It has to be a string in the format
17383 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17384 number or a valid video frame rate abbreviation. The default value is
17388 Set the duration of the sourced video. See
17389 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17390 for the accepted syntax.
17392 If not specified, or the expressed duration is negative, the video is
17393 supposed to be generated forever.
17397 Set the number or the name of the test to perform. Supported tests are:
17413 Default value is "all", which will cycle through the list of all tests.
17418 mptestsrc=t=dc_luma
17421 will generate a "dc_luma" test pattern.
17423 @section frei0r_src
17425 Provide a frei0r source.
17427 To enable compilation of this filter you need to install the frei0r
17428 header and configure FFmpeg with @code{--enable-frei0r}.
17430 This source accepts the following parameters:
17435 The size of the video to generate. For the syntax of this option, check the
17436 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17439 The framerate of the generated video. It may be a string of the form
17440 @var{num}/@var{den} or a frame rate abbreviation.
17443 The name to the frei0r source to load. For more information regarding frei0r and
17444 how to set the parameters, read the @ref{frei0r} section in the video filters
17447 @item filter_params
17448 A '|'-separated list of parameters to pass to the frei0r source.
17452 For example, to generate a frei0r partik0l source with size 200x200
17453 and frame rate 10 which is overlaid on the overlay filter main input:
17455 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
17460 Generate a life pattern.
17462 This source is based on a generalization of John Conway's life game.
17464 The sourced input represents a life grid, each pixel represents a cell
17465 which can be in one of two possible states, alive or dead. Every cell
17466 interacts with its eight neighbours, which are the cells that are
17467 horizontally, vertically, or diagonally adjacent.
17469 At each interaction the grid evolves according to the adopted rule,
17470 which specifies the number of neighbor alive cells which will make a
17471 cell stay alive or born. The @option{rule} option allows one to specify
17474 This source accepts the following options:
17478 Set the file from which to read the initial grid state. In the file,
17479 each non-whitespace character is considered an alive cell, and newline
17480 is used to delimit the end of each row.
17482 If this option is not specified, the initial grid is generated
17486 Set the video rate, that is the number of frames generated per second.
17489 @item random_fill_ratio, ratio
17490 Set the random fill ratio for the initial random grid. It is a
17491 floating point number value ranging from 0 to 1, defaults to 1/PHI.
17492 It is ignored when a file is specified.
17494 @item random_seed, seed
17495 Set the seed for filling the initial random grid, must be an integer
17496 included between 0 and UINT32_MAX. If not specified, or if explicitly
17497 set to -1, the filter will try to use a good random seed on a best
17503 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
17504 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
17505 @var{NS} specifies the number of alive neighbor cells which make a
17506 live cell stay alive, and @var{NB} the number of alive neighbor cells
17507 which make a dead cell to become alive (i.e. to "born").
17508 "s" and "b" can be used in place of "S" and "B", respectively.
17510 Alternatively a rule can be specified by an 18-bits integer. The 9
17511 high order bits are used to encode the next cell state if it is alive
17512 for each number of neighbor alive cells, the low order bits specify
17513 the rule for "borning" new cells. Higher order bits encode for an
17514 higher number of neighbor cells.
17515 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
17516 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
17518 Default value is "S23/B3", which is the original Conway's game of life
17519 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
17520 cells, and will born a new cell if there are three alive cells around
17524 Set the size of the output video. For the syntax of this option, check the
17525 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17527 If @option{filename} is specified, the size is set by default to the
17528 same size of the input file. If @option{size} is set, it must contain
17529 the size specified in the input file, and the initial grid defined in
17530 that file is centered in the larger resulting area.
17532 If a filename is not specified, the size value defaults to "320x240"
17533 (used for a randomly generated initial grid).
17536 If set to 1, stitch the left and right grid edges together, and the
17537 top and bottom edges also. Defaults to 1.
17540 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
17541 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
17542 value from 0 to 255.
17545 Set the color of living (or new born) cells.
17548 Set the color of dead cells. If @option{mold} is set, this is the first color
17549 used to represent a dead cell.
17552 Set mold color, for definitely dead and moldy cells.
17554 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
17555 ffmpeg-utils manual,ffmpeg-utils}.
17558 @subsection Examples
17562 Read a grid from @file{pattern}, and center it on a grid of size
17565 life=f=pattern:s=300x300
17569 Generate a random grid of size 200x200, with a fill ratio of 2/3:
17571 life=ratio=2/3:s=200x200
17575 Specify a custom rule for evolving a randomly generated grid:
17581 Full example with slow death effect (mold) using @command{ffplay}:
17583 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
17590 @anchor{haldclutsrc}
17592 @anchor{rgbtestsrc}
17594 @anchor{smptehdbars}
17597 @anchor{yuvtestsrc}
17598 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
17600 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
17602 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
17604 The @code{color} source provides an uniformly colored input.
17606 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
17607 @ref{haldclut} filter.
17609 The @code{nullsrc} source returns unprocessed video frames. It is
17610 mainly useful to be employed in analysis / debugging tools, or as the
17611 source for filters which ignore the input data.
17613 The @code{rgbtestsrc} source generates an RGB test pattern useful for
17614 detecting RGB vs BGR issues. You should see a red, green and blue
17615 stripe from top to bottom.
17617 The @code{smptebars} source generates a color bars pattern, based on
17618 the SMPTE Engineering Guideline EG 1-1990.
17620 The @code{smptehdbars} source generates a color bars pattern, based on
17621 the SMPTE RP 219-2002.
17623 The @code{testsrc} source generates a test video pattern, showing a
17624 color pattern, a scrolling gradient and a timestamp. This is mainly
17625 intended for testing purposes.
17627 The @code{testsrc2} source is similar to testsrc, but supports more
17628 pixel formats instead of just @code{rgb24}. This allows using it as an
17629 input for other tests without requiring a format conversion.
17631 The @code{yuvtestsrc} source generates an YUV test pattern. You should
17632 see a y, cb and cr stripe from top to bottom.
17634 The sources accept the following parameters:
17639 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
17640 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
17641 pixels to be used as identity matrix for 3D lookup tables. Each component is
17642 coded on a @code{1/(N*N)} scale.
17645 Specify the color of the source, only available in the @code{color}
17646 source. For the syntax of this option, check the
17647 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17650 Specify the size of the sourced video. For the syntax of this option, check the
17651 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17652 The default value is @code{320x240}.
17654 This option is not available with the @code{allrgb}, @code{allyuv}, and
17655 @code{haldclutsrc} filters.
17658 Specify the frame rate of the sourced video, as the number of frames
17659 generated per second. It has to be a string in the format
17660 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17661 number or a valid video frame rate abbreviation. The default value is
17665 Set the duration of the sourced video. See
17666 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17667 for the accepted syntax.
17669 If not specified, or the expressed duration is negative, the video is
17670 supposed to be generated forever.
17673 Set the sample aspect ratio of the sourced video.
17676 Specify the alpha (opacity) of the background, only available in the
17677 @code{testsrc2} source. The value must be between 0 (fully transparent) and
17678 255 (fully opaque, the default).
17681 Set the number of decimals to show in the timestamp, only available in the
17682 @code{testsrc} source.
17684 The displayed timestamp value will correspond to the original
17685 timestamp value multiplied by the power of 10 of the specified
17686 value. Default value is 0.
17689 @subsection Examples
17693 Generate a video with a duration of 5.3 seconds, with size
17694 176x144 and a frame rate of 10 frames per second:
17696 testsrc=duration=5.3:size=qcif:rate=10
17700 The following graph description will generate a red source
17701 with an opacity of 0.2, with size "qcif" and a frame rate of 10
17704 color=c=red@@0.2:s=qcif:r=10
17708 If the input content is to be ignored, @code{nullsrc} can be used. The
17709 following command generates noise in the luminance plane by employing
17710 the @code{geq} filter:
17712 nullsrc=s=256x256, geq=random(1)*255:128:128
17716 @subsection Commands
17718 The @code{color} source supports the following commands:
17722 Set the color of the created image. Accepts the same syntax of the
17723 corresponding @option{color} option.
17728 Generate video using an OpenCL program.
17733 OpenCL program source file.
17736 Kernel name in program.
17739 Size of frames to generate. This must be set.
17742 Pixel format to use for the generated frames. This must be set.
17745 Number of frames generated every second. Default value is '25'.
17749 For details of how the program loading works, see the @ref{program_opencl}
17756 Generate a colour ramp by setting pixel values from the position of the pixel
17757 in the output image. (Note that this will work with all pixel formats, but
17758 the generated output will not be the same.)
17760 __kernel void ramp(__write_only image2d_t dst,
17761 unsigned int index)
17763 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17766 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
17768 write_imagef(dst, loc, val);
17773 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
17775 __kernel void sierpinski_carpet(__write_only image2d_t dst,
17776 unsigned int index)
17778 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17780 float4 value = 0.0f;
17781 int x = loc.x + index;
17782 int y = loc.y + index;
17783 while (x > 0 || y > 0) {
17784 if (x % 3 == 1 && y % 3 == 1) {
17792 write_imagef(dst, loc, value);
17798 @c man end VIDEO SOURCES
17800 @chapter Video Sinks
17801 @c man begin VIDEO SINKS
17803 Below is a description of the currently available video sinks.
17805 @section buffersink
17807 Buffer video frames, and make them available to the end of the filter
17810 This sink is mainly intended for programmatic use, in particular
17811 through the interface defined in @file{libavfilter/buffersink.h}
17812 or the options system.
17814 It accepts a pointer to an AVBufferSinkContext structure, which
17815 defines the incoming buffers' formats, to be passed as the opaque
17816 parameter to @code{avfilter_init_filter} for initialization.
17820 Null video sink: do absolutely nothing with the input video. It is
17821 mainly useful as a template and for use in analysis / debugging
17824 @c man end VIDEO SINKS
17826 @chapter Multimedia Filters
17827 @c man begin MULTIMEDIA FILTERS
17829 Below is a description of the currently available multimedia filters.
17833 Convert input audio to a video output, displaying the audio bit scope.
17835 The filter accepts the following options:
17839 Set frame rate, expressed as number of frames per second. Default
17843 Specify the video size for the output. For the syntax of this option, check the
17844 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17845 Default value is @code{1024x256}.
17848 Specify list of colors separated by space or by '|' which will be used to
17849 draw channels. Unrecognized or missing colors will be replaced
17853 @section ahistogram
17855 Convert input audio to a video output, displaying the volume histogram.
17857 The filter accepts the following options:
17861 Specify how histogram is calculated.
17863 It accepts the following values:
17866 Use single histogram for all channels.
17868 Use separate histogram for each channel.
17870 Default is @code{single}.
17873 Set frame rate, expressed as number of frames per second. Default
17877 Specify the video size for the output. For the syntax of this option, check the
17878 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17879 Default value is @code{hd720}.
17884 It accepts the following values:
17895 reverse logarithmic
17897 Default is @code{log}.
17900 Set amplitude scale.
17902 It accepts the following values:
17909 Default is @code{log}.
17912 Set how much frames to accumulate in histogram.
17913 Defauls is 1. Setting this to -1 accumulates all frames.
17916 Set histogram ratio of window height.
17919 Set sonogram sliding.
17921 It accepts the following values:
17924 replace old rows with new ones.
17926 scroll from top to bottom.
17928 Default is @code{replace}.
17931 @section aphasemeter
17933 Convert input audio to a video output, displaying the audio phase.
17935 The filter accepts the following options:
17939 Set the output frame rate. Default value is @code{25}.
17942 Set the video size for the output. For the syntax of this option, check the
17943 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17944 Default value is @code{800x400}.
17949 Specify the red, green, blue contrast. Default values are @code{2},
17950 @code{7} and @code{1}.
17951 Allowed range is @code{[0, 255]}.
17954 Set color which will be used for drawing median phase. If color is
17955 @code{none} which is default, no median phase value will be drawn.
17958 Enable video output. Default is enabled.
17961 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
17962 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
17963 The @code{-1} means left and right channels are completely out of phase and
17964 @code{1} means channels are in phase.
17966 @section avectorscope
17968 Convert input audio to a video output, representing the audio vector
17971 The filter is used to measure the difference between channels of stereo
17972 audio stream. A monoaural signal, consisting of identical left and right
17973 signal, results in straight vertical line. Any stereo separation is visible
17974 as a deviation from this line, creating a Lissajous figure.
17975 If the straight (or deviation from it) but horizontal line appears this
17976 indicates that the left and right channels are out of phase.
17978 The filter accepts the following options:
17982 Set the vectorscope mode.
17984 Available values are:
17987 Lissajous rotated by 45 degrees.
17990 Same as above but not rotated.
17993 Shape resembling half of circle.
17996 Default value is @samp{lissajous}.
17999 Set the video size for the output. For the syntax of this option, check the
18000 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18001 Default value is @code{400x400}.
18004 Set the output frame rate. Default value is @code{25}.
18010 Specify the red, green, blue and alpha contrast. Default values are @code{40},
18011 @code{160}, @code{80} and @code{255}.
18012 Allowed range is @code{[0, 255]}.
18018 Specify the red, green, blue and alpha fade. Default values are @code{15},
18019 @code{10}, @code{5} and @code{5}.
18020 Allowed range is @code{[0, 255]}.
18023 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
18024 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
18027 Set the vectorscope drawing mode.
18029 Available values are:
18032 Draw dot for each sample.
18035 Draw line between previous and current sample.
18038 Default value is @samp{dot}.
18041 Specify amplitude scale of audio samples.
18043 Available values are:
18059 Swap left channel axis with right channel axis.
18069 Mirror only x axis.
18072 Mirror only y axis.
18080 @subsection Examples
18084 Complete example using @command{ffplay}:
18086 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18087 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
18091 @section bench, abench
18093 Benchmark part of a filtergraph.
18095 The filter accepts the following options:
18099 Start or stop a timer.
18101 Available values are:
18104 Get the current time, set it as frame metadata (using the key
18105 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
18108 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
18109 the input frame metadata to get the time difference. Time difference, average,
18110 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
18111 @code{min}) are then printed. The timestamps are expressed in seconds.
18115 @subsection Examples
18119 Benchmark @ref{selectivecolor} filter:
18121 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
18127 Concatenate audio and video streams, joining them together one after the
18130 The filter works on segments of synchronized video and audio streams. All
18131 segments must have the same number of streams of each type, and that will
18132 also be the number of streams at output.
18134 The filter accepts the following options:
18139 Set the number of segments. Default is 2.
18142 Set the number of output video streams, that is also the number of video
18143 streams in each segment. Default is 1.
18146 Set the number of output audio streams, that is also the number of audio
18147 streams in each segment. Default is 0.
18150 Activate unsafe mode: do not fail if segments have a different format.
18154 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
18155 @var{a} audio outputs.
18157 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
18158 segment, in the same order as the outputs, then the inputs for the second
18161 Related streams do not always have exactly the same duration, for various
18162 reasons including codec frame size or sloppy authoring. For that reason,
18163 related synchronized streams (e.g. a video and its audio track) should be
18164 concatenated at once. The concat filter will use the duration of the longest
18165 stream in each segment (except the last one), and if necessary pad shorter
18166 audio streams with silence.
18168 For this filter to work correctly, all segments must start at timestamp 0.
18170 All corresponding streams must have the same parameters in all segments; the
18171 filtering system will automatically select a common pixel format for video
18172 streams, and a common sample format, sample rate and channel layout for
18173 audio streams, but other settings, such as resolution, must be converted
18174 explicitly by the user.
18176 Different frame rates are acceptable but will result in variable frame rate
18177 at output; be sure to configure the output file to handle it.
18179 @subsection Examples
18183 Concatenate an opening, an episode and an ending, all in bilingual version
18184 (video in stream 0, audio in streams 1 and 2):
18186 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
18187 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
18188 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
18189 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
18193 Concatenate two parts, handling audio and video separately, using the
18194 (a)movie sources, and adjusting the resolution:
18196 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
18197 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
18198 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
18200 Note that a desync will happen at the stitch if the audio and video streams
18201 do not have exactly the same duration in the first file.
18205 @subsection Commands
18207 This filter supports the following commands:
18210 Close the current segment and step to the next one
18213 @section drawgraph, adrawgraph
18215 Draw a graph using input video or audio metadata.
18217 It accepts the following parameters:
18221 Set 1st frame metadata key from which metadata values will be used to draw a graph.
18224 Set 1st foreground color expression.
18227 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
18230 Set 2nd foreground color expression.
18233 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
18236 Set 3rd foreground color expression.
18239 Set 4th frame metadata key from which metadata values will be used to draw a graph.
18242 Set 4th foreground color expression.
18245 Set minimal value of metadata value.
18248 Set maximal value of metadata value.
18251 Set graph background color. Default is white.
18256 Available values for mode is:
18263 Default is @code{line}.
18268 Available values for slide is:
18271 Draw new frame when right border is reached.
18274 Replace old columns with new ones.
18277 Scroll from right to left.
18280 Scroll from left to right.
18283 Draw single picture.
18286 Default is @code{frame}.
18289 Set size of graph video. For the syntax of this option, check the
18290 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18291 The default value is @code{900x256}.
18293 The foreground color expressions can use the following variables:
18296 Minimal value of metadata value.
18299 Maximal value of metadata value.
18302 Current metadata key value.
18305 The color is defined as 0xAABBGGRR.
18308 Example using metadata from @ref{signalstats} filter:
18310 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
18313 Example using metadata from @ref{ebur128} filter:
18315 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
18321 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
18322 it unchanged. By default, it logs a message at a frequency of 10Hz with the
18323 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
18324 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
18326 The filter also has a video output (see the @var{video} option) with a real
18327 time graph to observe the loudness evolution. The graphic contains the logged
18328 message mentioned above, so it is not printed anymore when this option is set,
18329 unless the verbose logging is set. The main graphing area contains the
18330 short-term loudness (3 seconds of analysis), and the gauge on the right is for
18331 the momentary loudness (400 milliseconds).
18333 More information about the Loudness Recommendation EBU R128 on
18334 @url{http://tech.ebu.ch/loudness}.
18336 The filter accepts the following options:
18341 Activate the video output. The audio stream is passed unchanged whether this
18342 option is set or no. The video stream will be the first output stream if
18343 activated. Default is @code{0}.
18346 Set the video size. This option is for video only. For the syntax of this
18348 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18349 Default and minimum resolution is @code{640x480}.
18352 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
18353 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
18354 other integer value between this range is allowed.
18357 Set metadata injection. If set to @code{1}, the audio input will be segmented
18358 into 100ms output frames, each of them containing various loudness information
18359 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
18361 Default is @code{0}.
18364 Force the frame logging level.
18366 Available values are:
18369 information logging level
18371 verbose logging level
18374 By default, the logging level is set to @var{info}. If the @option{video} or
18375 the @option{metadata} options are set, it switches to @var{verbose}.
18380 Available modes can be cumulated (the option is a @code{flag} type). Possible
18384 Disable any peak mode (default).
18386 Enable sample-peak mode.
18388 Simple peak mode looking for the higher sample value. It logs a message
18389 for sample-peak (identified by @code{SPK}).
18391 Enable true-peak mode.
18393 If enabled, the peak lookup is done on an over-sampled version of the input
18394 stream for better peak accuracy. It logs a message for true-peak.
18395 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
18396 This mode requires a build with @code{libswresample}.
18400 Treat mono input files as "dual mono". If a mono file is intended for playback
18401 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
18402 If set to @code{true}, this option will compensate for this effect.
18403 Multi-channel input files are not affected by this option.
18406 Set a specific pan law to be used for the measurement of dual mono files.
18407 This parameter is optional, and has a default value of -3.01dB.
18410 @subsection Examples
18414 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
18416 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
18420 Run an analysis with @command{ffmpeg}:
18422 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
18426 @section interleave, ainterleave
18428 Temporally interleave frames from several inputs.
18430 @code{interleave} works with video inputs, @code{ainterleave} with audio.
18432 These filters read frames from several inputs and send the oldest
18433 queued frame to the output.
18435 Input streams must have well defined, monotonically increasing frame
18438 In order to submit one frame to output, these filters need to enqueue
18439 at least one frame for each input, so they cannot work in case one
18440 input is not yet terminated and will not receive incoming frames.
18442 For example consider the case when one input is a @code{select} filter
18443 which always drops input frames. The @code{interleave} filter will keep
18444 reading from that input, but it will never be able to send new frames
18445 to output until the input sends an end-of-stream signal.
18447 Also, depending on inputs synchronization, the filters will drop
18448 frames in case one input receives more frames than the other ones, and
18449 the queue is already filled.
18451 These filters accept the following options:
18455 Set the number of different inputs, it is 2 by default.
18458 @subsection Examples
18462 Interleave frames belonging to different streams using @command{ffmpeg}:
18464 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
18468 Add flickering blur effect:
18470 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
18474 @section metadata, ametadata
18476 Manipulate frame metadata.
18478 This filter accepts the following options:
18482 Set mode of operation of the filter.
18484 Can be one of the following:
18488 If both @code{value} and @code{key} is set, select frames
18489 which have such metadata. If only @code{key} is set, select
18490 every frame that has such key in metadata.
18493 Add new metadata @code{key} and @code{value}. If key is already available
18497 Modify value of already present key.
18500 If @code{value} is set, delete only keys that have such value.
18501 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
18505 Print key and its value if metadata was found. If @code{key} is not set print all
18506 metadata values available in frame.
18510 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
18513 Set metadata value which will be used. This option is mandatory for
18514 @code{modify} and @code{add} mode.
18517 Which function to use when comparing metadata value and @code{value}.
18519 Can be one of following:
18523 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
18526 Values are interpreted as strings, returns true if metadata value starts with
18527 the @code{value} option string.
18530 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
18533 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
18536 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
18539 Values are interpreted as floats, returns true if expression from option @code{expr}
18544 Set expression which is used when @code{function} is set to @code{expr}.
18545 The expression is evaluated through the eval API and can contain the following
18550 Float representation of @code{value} from metadata key.
18553 Float representation of @code{value} as supplied by user in @code{value} option.
18557 If specified in @code{print} mode, output is written to the named file. Instead of
18558 plain filename any writable url can be specified. Filename ``-'' is a shorthand
18559 for standard output. If @code{file} option is not set, output is written to the log
18560 with AV_LOG_INFO loglevel.
18564 @subsection Examples
18568 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
18571 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
18574 Print silencedetect output to file @file{metadata.txt}.
18576 silencedetect,ametadata=mode=print:file=metadata.txt
18579 Direct all metadata to a pipe with file descriptor 4.
18581 metadata=mode=print:file='pipe\:4'
18585 @section perms, aperms
18587 Set read/write permissions for the output frames.
18589 These filters are mainly aimed at developers to test direct path in the
18590 following filter in the filtergraph.
18592 The filters accept the following options:
18596 Select the permissions mode.
18598 It accepts the following values:
18601 Do nothing. This is the default.
18603 Set all the output frames read-only.
18605 Set all the output frames directly writable.
18607 Make the frame read-only if writable, and writable if read-only.
18609 Set each output frame read-only or writable randomly.
18613 Set the seed for the @var{random} mode, must be an integer included between
18614 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
18615 @code{-1}, the filter will try to use a good random seed on a best effort
18619 Note: in case of auto-inserted filter between the permission filter and the
18620 following one, the permission might not be received as expected in that
18621 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
18622 perms/aperms filter can avoid this problem.
18624 @section realtime, arealtime
18626 Slow down filtering to match real time approximately.
18628 These filters will pause the filtering for a variable amount of time to
18629 match the output rate with the input timestamps.
18630 They are similar to the @option{re} option to @code{ffmpeg}.
18632 They accept the following options:
18636 Time limit for the pauses. Any pause longer than that will be considered
18637 a timestamp discontinuity and reset the timer. Default is 2 seconds.
18641 @section select, aselect
18643 Select frames to pass in output.
18645 This filter accepts the following options:
18650 Set expression, which is evaluated for each input frame.
18652 If the expression is evaluated to zero, the frame is discarded.
18654 If the evaluation result is negative or NaN, the frame is sent to the
18655 first output; otherwise it is sent to the output with index
18656 @code{ceil(val)-1}, assuming that the input index starts from 0.
18658 For example a value of @code{1.2} corresponds to the output with index
18659 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
18662 Set the number of outputs. The output to which to send the selected
18663 frame is based on the result of the evaluation. Default value is 1.
18666 The expression can contain the following constants:
18670 The (sequential) number of the filtered frame, starting from 0.
18673 The (sequential) number of the selected frame, starting from 0.
18675 @item prev_selected_n
18676 The sequential number of the last selected frame. It's NAN if undefined.
18679 The timebase of the input timestamps.
18682 The PTS (Presentation TimeStamp) of the filtered video frame,
18683 expressed in @var{TB} units. It's NAN if undefined.
18686 The PTS of the filtered video frame,
18687 expressed in seconds. It's NAN if undefined.
18690 The PTS of the previously filtered video frame. It's NAN if undefined.
18692 @item prev_selected_pts
18693 The PTS of the last previously filtered video frame. It's NAN if undefined.
18695 @item prev_selected_t
18696 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
18699 The PTS of the first video frame in the video. It's NAN if undefined.
18702 The time of the first video frame in the video. It's NAN if undefined.
18704 @item pict_type @emph{(video only)}
18705 The type of the filtered frame. It can assume one of the following
18717 @item interlace_type @emph{(video only)}
18718 The frame interlace type. It can assume one of the following values:
18721 The frame is progressive (not interlaced).
18723 The frame is top-field-first.
18725 The frame is bottom-field-first.
18728 @item consumed_sample_n @emph{(audio only)}
18729 the number of selected samples before the current frame
18731 @item samples_n @emph{(audio only)}
18732 the number of samples in the current frame
18734 @item sample_rate @emph{(audio only)}
18735 the input sample rate
18738 This is 1 if the filtered frame is a key-frame, 0 otherwise.
18741 the position in the file of the filtered frame, -1 if the information
18742 is not available (e.g. for synthetic video)
18744 @item scene @emph{(video only)}
18745 value between 0 and 1 to indicate a new scene; a low value reflects a low
18746 probability for the current frame to introduce a new scene, while a higher
18747 value means the current frame is more likely to be one (see the example below)
18749 @item concatdec_select
18750 The concat demuxer can select only part of a concat input file by setting an
18751 inpoint and an outpoint, but the output packets may not be entirely contained
18752 in the selected interval. By using this variable, it is possible to skip frames
18753 generated by the concat demuxer which are not exactly contained in the selected
18756 This works by comparing the frame pts against the @var{lavf.concat.start_time}
18757 and the @var{lavf.concat.duration} packet metadata values which are also
18758 present in the decoded frames.
18760 The @var{concatdec_select} variable is -1 if the frame pts is at least
18761 start_time and either the duration metadata is missing or the frame pts is less
18762 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
18765 That basically means that an input frame is selected if its pts is within the
18766 interval set by the concat demuxer.
18770 The default value of the select expression is "1".
18772 @subsection Examples
18776 Select all frames in input:
18781 The example above is the same as:
18793 Select only I-frames:
18795 select='eq(pict_type\,I)'
18799 Select one frame every 100:
18801 select='not(mod(n\,100))'
18805 Select only frames contained in the 10-20 time interval:
18807 select=between(t\,10\,20)
18811 Select only I-frames contained in the 10-20 time interval:
18813 select=between(t\,10\,20)*eq(pict_type\,I)
18817 Select frames with a minimum distance of 10 seconds:
18819 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
18823 Use aselect to select only audio frames with samples number > 100:
18825 aselect='gt(samples_n\,100)'
18829 Create a mosaic of the first scenes:
18831 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
18834 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
18838 Send even and odd frames to separate outputs, and compose them:
18840 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
18844 Select useful frames from an ffconcat file which is using inpoints and
18845 outpoints but where the source files are not intra frame only.
18847 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
18851 @section sendcmd, asendcmd
18853 Send commands to filters in the filtergraph.
18855 These filters read commands to be sent to other filters in the
18858 @code{sendcmd} must be inserted between two video filters,
18859 @code{asendcmd} must be inserted between two audio filters, but apart
18860 from that they act the same way.
18862 The specification of commands can be provided in the filter arguments
18863 with the @var{commands} option, or in a file specified by the
18864 @var{filename} option.
18866 These filters accept the following options:
18869 Set the commands to be read and sent to the other filters.
18871 Set the filename of the commands to be read and sent to the other
18875 @subsection Commands syntax
18877 A commands description consists of a sequence of interval
18878 specifications, comprising a list of commands to be executed when a
18879 particular event related to that interval occurs. The occurring event
18880 is typically the current frame time entering or leaving a given time
18883 An interval is specified by the following syntax:
18885 @var{START}[-@var{END}] @var{COMMANDS};
18888 The time interval is specified by the @var{START} and @var{END} times.
18889 @var{END} is optional and defaults to the maximum time.
18891 The current frame time is considered within the specified interval if
18892 it is included in the interval [@var{START}, @var{END}), that is when
18893 the time is greater or equal to @var{START} and is lesser than
18896 @var{COMMANDS} consists of a sequence of one or more command
18897 specifications, separated by ",", relating to that interval. The
18898 syntax of a command specification is given by:
18900 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
18903 @var{FLAGS} is optional and specifies the type of events relating to
18904 the time interval which enable sending the specified command, and must
18905 be a non-null sequence of identifier flags separated by "+" or "|" and
18906 enclosed between "[" and "]".
18908 The following flags are recognized:
18911 The command is sent when the current frame timestamp enters the
18912 specified interval. In other words, the command is sent when the
18913 previous frame timestamp was not in the given interval, and the
18917 The command is sent when the current frame timestamp leaves the
18918 specified interval. In other words, the command is sent when the
18919 previous frame timestamp was in the given interval, and the
18923 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
18926 @var{TARGET} specifies the target of the command, usually the name of
18927 the filter class or a specific filter instance name.
18929 @var{COMMAND} specifies the name of the command for the target filter.
18931 @var{ARG} is optional and specifies the optional list of argument for
18932 the given @var{COMMAND}.
18934 Between one interval specification and another, whitespaces, or
18935 sequences of characters starting with @code{#} until the end of line,
18936 are ignored and can be used to annotate comments.
18938 A simplified BNF description of the commands specification syntax
18941 @var{COMMAND_FLAG} ::= "enter" | "leave"
18942 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
18943 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
18944 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
18945 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
18946 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
18949 @subsection Examples
18953 Specify audio tempo change at second 4:
18955 asendcmd=c='4.0 atempo tempo 1.5',atempo
18959 Target a specific filter instance:
18961 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
18965 Specify a list of drawtext and hue commands in a file.
18967 # show text in the interval 5-10
18968 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
18969 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
18971 # desaturate the image in the interval 15-20
18972 15.0-20.0 [enter] hue s 0,
18973 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
18975 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
18977 # apply an exponential saturation fade-out effect, starting from time 25
18978 25 [enter] hue s exp(25-t)
18981 A filtergraph allowing to read and process the above command list
18982 stored in a file @file{test.cmd}, can be specified with:
18984 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
18989 @section setpts, asetpts
18991 Change the PTS (presentation timestamp) of the input frames.
18993 @code{setpts} works on video frames, @code{asetpts} on audio frames.
18995 This filter accepts the following options:
19000 The expression which is evaluated for each frame to construct its timestamp.
19004 The expression is evaluated through the eval API and can contain the following
19009 frame rate, only defined for constant frame-rate video
19012 The presentation timestamp in input
19015 The count of the input frame for video or the number of consumed samples,
19016 not including the current frame for audio, starting from 0.
19018 @item NB_CONSUMED_SAMPLES
19019 The number of consumed samples, not including the current frame (only
19022 @item NB_SAMPLES, S
19023 The number of samples in the current frame (only audio)
19025 @item SAMPLE_RATE, SR
19026 The audio sample rate.
19029 The PTS of the first frame.
19032 the time in seconds of the first frame
19035 State whether the current frame is interlaced.
19038 the time in seconds of the current frame
19041 original position in the file of the frame, or undefined if undefined
19042 for the current frame
19045 The previous input PTS.
19048 previous input time in seconds
19051 The previous output PTS.
19054 previous output time in seconds
19057 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
19061 The wallclock (RTC) time at the start of the movie in microseconds.
19064 The timebase of the input timestamps.
19068 @subsection Examples
19072 Start counting PTS from zero
19074 setpts=PTS-STARTPTS
19078 Apply fast motion effect:
19084 Apply slow motion effect:
19090 Set fixed rate of 25 frames per second:
19096 Set fixed rate 25 fps with some jitter:
19098 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
19102 Apply an offset of 10 seconds to the input PTS:
19108 Generate timestamps from a "live source" and rebase onto the current timebase:
19110 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
19114 Generate timestamps by counting samples:
19123 Force color range for the output video frame.
19125 The @code{setrange} filter marks the color range property for the
19126 output frames. It does not change the input frame, but only sets the
19127 corresponding property, which affects how the frame is treated by
19130 The filter accepts the following options:
19135 Available values are:
19139 Keep the same color range property.
19141 @item unspecified, unknown
19142 Set the color range as unspecified.
19144 @item limited, tv, mpeg
19145 Set the color range as limited.
19147 @item full, pc, jpeg
19148 Set the color range as full.
19152 @section settb, asettb
19154 Set the timebase to use for the output frames timestamps.
19155 It is mainly useful for testing timebase configuration.
19157 It accepts the following parameters:
19162 The expression which is evaluated into the output timebase.
19166 The value for @option{tb} is an arithmetic expression representing a
19167 rational. The expression can contain the constants "AVTB" (the default
19168 timebase), "intb" (the input timebase) and "sr" (the sample rate,
19169 audio only). Default value is "intb".
19171 @subsection Examples
19175 Set the timebase to 1/25:
19181 Set the timebase to 1/10:
19187 Set the timebase to 1001/1000:
19193 Set the timebase to 2*intb:
19199 Set the default timebase value:
19206 Convert input audio to a video output representing frequency spectrum
19207 logarithmically using Brown-Puckette constant Q transform algorithm with
19208 direct frequency domain coefficient calculation (but the transform itself
19209 is not really constant Q, instead the Q factor is actually variable/clamped),
19210 with musical tone scale, from E0 to D#10.
19212 The filter accepts the following options:
19216 Specify the video size for the output. It must be even. For the syntax of this option,
19217 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19218 Default value is @code{1920x1080}.
19221 Set the output frame rate. Default value is @code{25}.
19224 Set the bargraph height. It must be even. Default value is @code{-1} which
19225 computes the bargraph height automatically.
19228 Set the axis height. It must be even. Default value is @code{-1} which computes
19229 the axis height automatically.
19232 Set the sonogram height. It must be even. Default value is @code{-1} which
19233 computes the sonogram height automatically.
19236 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
19237 instead. Default value is @code{1}.
19239 @item sono_v, volume
19240 Specify the sonogram volume expression. It can contain variables:
19243 the @var{bar_v} evaluated expression
19244 @item frequency, freq, f
19245 the frequency where it is evaluated
19246 @item timeclamp, tc
19247 the value of @var{timeclamp} option
19251 @item a_weighting(f)
19252 A-weighting of equal loudness
19253 @item b_weighting(f)
19254 B-weighting of equal loudness
19255 @item c_weighting(f)
19256 C-weighting of equal loudness.
19258 Default value is @code{16}.
19260 @item bar_v, volume2
19261 Specify the bargraph volume expression. It can contain variables:
19264 the @var{sono_v} evaluated expression
19265 @item frequency, freq, f
19266 the frequency where it is evaluated
19267 @item timeclamp, tc
19268 the value of @var{timeclamp} option
19272 @item a_weighting(f)
19273 A-weighting of equal loudness
19274 @item b_weighting(f)
19275 B-weighting of equal loudness
19276 @item c_weighting(f)
19277 C-weighting of equal loudness.
19279 Default value is @code{sono_v}.
19281 @item sono_g, gamma
19282 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
19283 higher gamma makes the spectrum having more range. Default value is @code{3}.
19284 Acceptable range is @code{[1, 7]}.
19286 @item bar_g, gamma2
19287 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
19291 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
19292 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
19294 @item timeclamp, tc
19295 Specify the transform timeclamp. At low frequency, there is trade-off between
19296 accuracy in time domain and frequency domain. If timeclamp is lower,
19297 event in time domain is represented more accurately (such as fast bass drum),
19298 otherwise event in frequency domain is represented more accurately
19299 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
19302 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
19303 limits future samples by applying asymmetric windowing in time domain, useful
19304 when low latency is required. Accepted range is @code{[0, 1]}.
19307 Specify the transform base frequency. Default value is @code{20.01523126408007475},
19308 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
19311 Specify the transform end frequency. Default value is @code{20495.59681441799654},
19312 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
19315 This option is deprecated and ignored.
19318 Specify the transform length in time domain. Use this option to control accuracy
19319 trade-off between time domain and frequency domain at every frequency sample.
19320 It can contain variables:
19322 @item frequency, freq, f
19323 the frequency where it is evaluated
19324 @item timeclamp, tc
19325 the value of @var{timeclamp} option.
19327 Default value is @code{384*tc/(384+tc*f)}.
19330 Specify the transform count for every video frame. Default value is @code{6}.
19331 Acceptable range is @code{[1, 30]}.
19334 Specify the transform count for every single pixel. Default value is @code{0},
19335 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
19338 Specify font file for use with freetype to draw the axis. If not specified,
19339 use embedded font. Note that drawing with font file or embedded font is not
19340 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
19344 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
19345 The : in the pattern may be replaced by | to avoid unnecessary escaping.
19348 Specify font color expression. This is arithmetic expression that should return
19349 integer value 0xRRGGBB. It can contain variables:
19351 @item frequency, freq, f
19352 the frequency where it is evaluated
19353 @item timeclamp, tc
19354 the value of @var{timeclamp} option
19359 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
19360 @item r(x), g(x), b(x)
19361 red, green, and blue value of intensity x.
19363 Default value is @code{st(0, (midi(f)-59.5)/12);
19364 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
19365 r(1-ld(1)) + b(ld(1))}.
19368 Specify image file to draw the axis. This option override @var{fontfile} and
19369 @var{fontcolor} option.
19372 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
19373 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
19374 Default value is @code{1}.
19377 Set colorspace. The accepted values are:
19380 Unspecified (default)
19389 BT.470BG or BT.601-6 625
19392 SMPTE-170M or BT.601-6 525
19398 BT.2020 with non-constant luminance
19403 Set spectrogram color scheme. This is list of floating point values with format
19404 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
19405 The default is @code{1|0.5|0|0|0.5|1}.
19409 @subsection Examples
19413 Playing audio while showing the spectrum:
19415 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
19419 Same as above, but with frame rate 30 fps:
19421 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
19425 Playing at 1280x720:
19427 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
19431 Disable sonogram display:
19437 A1 and its harmonics: A1, A2, (near)E3, A3:
19439 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),
19440 asplit[a][out1]; [a] showcqt [out0]'
19444 Same as above, but with more accuracy in frequency domain:
19446 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),
19447 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
19453 bar_v=10:sono_v=bar_v*a_weighting(f)
19457 Custom gamma, now spectrum is linear to the amplitude.
19463 Custom tlength equation:
19465 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)))'
19469 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
19471 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
19475 Custom font using fontconfig:
19477 font='Courier New,Monospace,mono|bold'
19481 Custom frequency range with custom axis using image file:
19483 axisfile=myaxis.png:basefreq=40:endfreq=10000
19489 Convert input audio to video output representing the audio power spectrum.
19490 Audio amplitude is on Y-axis while frequency is on X-axis.
19492 The filter accepts the following options:
19496 Specify size of video. For the syntax of this option, check the
19497 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19498 Default is @code{1024x512}.
19502 This set how each frequency bin will be represented.
19504 It accepts the following values:
19510 Default is @code{bar}.
19513 Set amplitude scale.
19515 It accepts the following values:
19529 Default is @code{log}.
19532 Set frequency scale.
19534 It accepts the following values:
19543 Reverse logarithmic scale.
19545 Default is @code{lin}.
19550 It accepts the following values:
19566 Default is @code{w2048}
19569 Set windowing function.
19571 It accepts the following values:
19593 Default is @code{hanning}.
19596 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19597 which means optimal overlap for selected window function will be picked.
19600 Set time averaging. Setting this to 0 will display current maximal peaks.
19601 Default is @code{1}, which means time averaging is disabled.
19604 Specify list of colors separated by space or by '|' which will be used to
19605 draw channel frequencies. Unrecognized or missing colors will be replaced
19609 Set channel display mode.
19611 It accepts the following values:
19616 Default is @code{combined}.
19619 Set minimum amplitude used in @code{log} amplitude scaler.
19623 @anchor{showspectrum}
19624 @section showspectrum
19626 Convert input audio to a video output, representing the audio frequency
19629 The filter accepts the following options:
19633 Specify the video size for the output. For the syntax of this option, check the
19634 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19635 Default value is @code{640x512}.
19638 Specify how the spectrum should slide along the window.
19640 It accepts the following values:
19643 the samples start again on the left when they reach the right
19645 the samples scroll from right to left
19647 frames are only produced when the samples reach the right
19649 the samples scroll from left to right
19652 Default value is @code{replace}.
19655 Specify display mode.
19657 It accepts the following values:
19660 all channels are displayed in the same row
19662 all channels are displayed in separate rows
19665 Default value is @samp{combined}.
19668 Specify display color mode.
19670 It accepts the following values:
19673 each channel is displayed in a separate color
19675 each channel is displayed using the same color scheme
19677 each channel is displayed using the rainbow color scheme
19679 each channel is displayed using the moreland color scheme
19681 each channel is displayed using the nebulae color scheme
19683 each channel is displayed using the fire color scheme
19685 each channel is displayed using the fiery color scheme
19687 each channel is displayed using the fruit color scheme
19689 each channel is displayed using the cool color scheme
19692 Default value is @samp{channel}.
19695 Specify scale used for calculating intensity color values.
19697 It accepts the following values:
19702 square root, default
19713 Default value is @samp{sqrt}.
19716 Set saturation modifier for displayed colors. Negative values provide
19717 alternative color scheme. @code{0} is no saturation at all.
19718 Saturation must be in [-10.0, 10.0] range.
19719 Default value is @code{1}.
19722 Set window function.
19724 It accepts the following values:
19748 Default value is @code{hann}.
19751 Set orientation of time vs frequency axis. Can be @code{vertical} or
19752 @code{horizontal}. Default is @code{vertical}.
19755 Set ratio of overlap window. Default value is @code{0}.
19756 When value is @code{1} overlap is set to recommended size for specific
19757 window function currently used.
19760 Set scale gain for calculating intensity color values.
19761 Default value is @code{1}.
19764 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
19767 Set color rotation, must be in [-1.0, 1.0] range.
19768 Default value is @code{0}.
19771 The usage is very similar to the showwaves filter; see the examples in that
19774 @subsection Examples
19778 Large window with logarithmic color scaling:
19780 showspectrum=s=1280x480:scale=log
19784 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
19786 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
19787 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
19791 @section showspectrumpic
19793 Convert input audio to a single video frame, representing the audio frequency
19796 The filter accepts the following options:
19800 Specify the video size for the output. For the syntax of this option, check the
19801 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19802 Default value is @code{4096x2048}.
19805 Specify display mode.
19807 It accepts the following values:
19810 all channels are displayed in the same row
19812 all channels are displayed in separate rows
19814 Default value is @samp{combined}.
19817 Specify display color mode.
19819 It accepts the following values:
19822 each channel is displayed in a separate color
19824 each channel is displayed using the same color scheme
19826 each channel is displayed using the rainbow color scheme
19828 each channel is displayed using the moreland color scheme
19830 each channel is displayed using the nebulae color scheme
19832 each channel is displayed using the fire color scheme
19834 each channel is displayed using the fiery color scheme
19836 each channel is displayed using the fruit color scheme
19838 each channel is displayed using the cool color scheme
19840 Default value is @samp{intensity}.
19843 Specify scale used for calculating intensity color values.
19845 It accepts the following values:
19850 square root, default
19860 Default value is @samp{log}.
19863 Set saturation modifier for displayed colors. Negative values provide
19864 alternative color scheme. @code{0} is no saturation at all.
19865 Saturation must be in [-10.0, 10.0] range.
19866 Default value is @code{1}.
19869 Set window function.
19871 It accepts the following values:
19894 Default value is @code{hann}.
19897 Set orientation of time vs frequency axis. Can be @code{vertical} or
19898 @code{horizontal}. Default is @code{vertical}.
19901 Set scale gain for calculating intensity color values.
19902 Default value is @code{1}.
19905 Draw time and frequency axes and legends. Default is enabled.
19908 Set color rotation, must be in [-1.0, 1.0] range.
19909 Default value is @code{0}.
19912 @subsection Examples
19916 Extract an audio spectrogram of a whole audio track
19917 in a 1024x1024 picture using @command{ffmpeg}:
19919 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
19923 @section showvolume
19925 Convert input audio volume to a video output.
19927 The filter accepts the following options:
19934 Set border width, allowed range is [0, 5]. Default is 1.
19937 Set channel width, allowed range is [80, 8192]. Default is 400.
19940 Set channel height, allowed range is [1, 900]. Default is 20.
19943 Set fade, allowed range is [0, 1]. Default is 0.95.
19946 Set volume color expression.
19948 The expression can use the following variables:
19952 Current max volume of channel in dB.
19958 Current channel number, starting from 0.
19962 If set, displays channel names. Default is enabled.
19965 If set, displays volume values. Default is enabled.
19968 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
19969 default is @code{h}.
19972 Set step size, allowed range is [0, 5]. Default is 0, which means
19976 Set background opacity, allowed range is [0, 1]. Default is 0.
19979 Set metering mode, can be peak: @code{p} or rms: @code{r},
19980 default is @code{p}.
19985 Convert input audio to a video output, representing the samples waves.
19987 The filter accepts the following options:
19991 Specify the video size for the output. For the syntax of this option, check the
19992 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19993 Default value is @code{600x240}.
19998 Available values are:
20001 Draw a point for each sample.
20004 Draw a vertical line for each sample.
20007 Draw a point for each sample and a line between them.
20010 Draw a centered vertical line for each sample.
20013 Default value is @code{point}.
20016 Set the number of samples which are printed on the same column. A
20017 larger value will decrease the frame rate. Must be a positive
20018 integer. This option can be set only if the value for @var{rate}
20019 is not explicitly specified.
20022 Set the (approximate) output frame rate. This is done by setting the
20023 option @var{n}. Default value is "25".
20025 @item split_channels
20026 Set if channels should be drawn separately or overlap. Default value is 0.
20029 Set colors separated by '|' which are going to be used for drawing of each channel.
20032 Set amplitude scale.
20034 Available values are:
20052 Set the draw mode. This is mostly useful to set for high @var{n}.
20054 Available values are:
20057 Scale pixel values for each drawn sample.
20060 Draw every sample directly.
20063 Default value is @code{scale}.
20066 @subsection Examples
20070 Output the input file audio and the corresponding video representation
20073 amovie=a.mp3,asplit[out0],showwaves[out1]
20077 Create a synthetic signal and show it with showwaves, forcing a
20078 frame rate of 30 frames per second:
20080 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
20084 @section showwavespic
20086 Convert input audio to a single video frame, representing the samples waves.
20088 The filter accepts the following options:
20092 Specify the video size for the output. For the syntax of this option, check the
20093 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20094 Default value is @code{600x240}.
20096 @item split_channels
20097 Set if channels should be drawn separately or overlap. Default value is 0.
20100 Set colors separated by '|' which are going to be used for drawing of each channel.
20103 Set amplitude scale.
20105 Available values are:
20123 @subsection Examples
20127 Extract a channel split representation of the wave form of a whole audio track
20128 in a 1024x800 picture using @command{ffmpeg}:
20130 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
20134 @section sidedata, asidedata
20136 Delete frame side data, or select frames based on it.
20138 This filter accepts the following options:
20142 Set mode of operation of the filter.
20144 Can be one of the following:
20148 Select every frame with side data of @code{type}.
20151 Delete side data of @code{type}. If @code{type} is not set, delete all side
20157 Set side data type used with all modes. Must be set for @code{select} mode. For
20158 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
20159 in @file{libavutil/frame.h}. For example, to choose
20160 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
20164 @section spectrumsynth
20166 Sythesize audio from 2 input video spectrums, first input stream represents
20167 magnitude across time and second represents phase across time.
20168 The filter will transform from frequency domain as displayed in videos back
20169 to time domain as presented in audio output.
20171 This filter is primarily created for reversing processed @ref{showspectrum}
20172 filter outputs, but can synthesize sound from other spectrograms too.
20173 But in such case results are going to be poor if the phase data is not
20174 available, because in such cases phase data need to be recreated, usually
20175 its just recreated from random noise.
20176 For best results use gray only output (@code{channel} color mode in
20177 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
20178 @code{lin} scale for phase video. To produce phase, for 2nd video, use
20179 @code{data} option. Inputs videos should generally use @code{fullframe}
20180 slide mode as that saves resources needed for decoding video.
20182 The filter accepts the following options:
20186 Specify sample rate of output audio, the sample rate of audio from which
20187 spectrum was generated may differ.
20190 Set number of channels represented in input video spectrums.
20193 Set scale which was used when generating magnitude input spectrum.
20194 Can be @code{lin} or @code{log}. Default is @code{log}.
20197 Set slide which was used when generating inputs spectrums.
20198 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
20199 Default is @code{fullframe}.
20202 Set window function used for resynthesis.
20205 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20206 which means optimal overlap for selected window function will be picked.
20209 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
20210 Default is @code{vertical}.
20213 @subsection Examples
20217 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
20218 then resynthesize videos back to audio with spectrumsynth:
20220 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
20221 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
20222 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
20226 @section split, asplit
20228 Split input into several identical outputs.
20230 @code{asplit} works with audio input, @code{split} with video.
20232 The filter accepts a single parameter which specifies the number of outputs. If
20233 unspecified, it defaults to 2.
20235 @subsection Examples
20239 Create two separate outputs from the same input:
20241 [in] split [out0][out1]
20245 To create 3 or more outputs, you need to specify the number of
20248 [in] asplit=3 [out0][out1][out2]
20252 Create two separate outputs from the same input, one cropped and
20255 [in] split [splitout1][splitout2];
20256 [splitout1] crop=100:100:0:0 [cropout];
20257 [splitout2] pad=200:200:100:100 [padout];
20261 Create 5 copies of the input audio with @command{ffmpeg}:
20263 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
20269 Receive commands sent through a libzmq client, and forward them to
20270 filters in the filtergraph.
20272 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
20273 must be inserted between two video filters, @code{azmq} between two
20274 audio filters. Both are capable to send messages to any filter type.
20276 To enable these filters you need to install the libzmq library and
20277 headers and configure FFmpeg with @code{--enable-libzmq}.
20279 For more information about libzmq see:
20280 @url{http://www.zeromq.org/}
20282 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
20283 receives messages sent through a network interface defined by the
20284 @option{bind_address} (or the abbreviation "@option{b}") option.
20285 Default value of this option is @file{tcp://localhost:5555}. You may
20286 want to alter this value to your needs, but do not forget to escape any
20287 ':' signs (see @ref{filtergraph escaping}).
20289 The received message must be in the form:
20291 @var{TARGET} @var{COMMAND} [@var{ARG}]
20294 @var{TARGET} specifies the target of the command, usually the name of
20295 the filter class or a specific filter instance name. The default
20296 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
20297 but you can override this by using the @samp{filter_name@@id} syntax
20298 (see @ref{Filtergraph syntax}).
20300 @var{COMMAND} specifies the name of the command for the target filter.
20302 @var{ARG} is optional and specifies the optional argument list for the
20303 given @var{COMMAND}.
20305 Upon reception, the message is processed and the corresponding command
20306 is injected into the filtergraph. Depending on the result, the filter
20307 will send a reply to the client, adopting the format:
20309 @var{ERROR_CODE} @var{ERROR_REASON}
20313 @var{MESSAGE} is optional.
20315 @subsection Examples
20317 Look at @file{tools/zmqsend} for an example of a zmq client which can
20318 be used to send commands processed by these filters.
20320 Consider the following filtergraph generated by @command{ffplay}.
20321 In this example the last overlay filter has an instance name. All other
20322 filters will have default instance names.
20325 ffplay -dumpgraph 1 -f lavfi "
20326 color=s=100x100:c=red [l];
20327 color=s=100x100:c=blue [r];
20328 nullsrc=s=200x100, zmq [bg];
20329 [bg][l] overlay [bg+l];
20330 [bg+l][r] overlay@@my=x=100 "
20333 To change the color of the left side of the video, the following
20334 command can be used:
20336 echo Parsed_color_0 c yellow | tools/zmqsend
20339 To change the right side:
20341 echo Parsed_color_1 c pink | tools/zmqsend
20344 To change the position of the right side:
20346 echo overlay@@my x 150 | tools/zmqsend
20350 @c man end MULTIMEDIA FILTERS
20352 @chapter Multimedia Sources
20353 @c man begin MULTIMEDIA SOURCES
20355 Below is a description of the currently available multimedia sources.
20359 This is the same as @ref{movie} source, except it selects an audio
20365 Read audio and/or video stream(s) from a movie container.
20367 It accepts the following parameters:
20371 The name of the resource to read (not necessarily a file; it can also be a
20372 device or a stream accessed through some protocol).
20374 @item format_name, f
20375 Specifies the format assumed for the movie to read, and can be either
20376 the name of a container or an input device. If not specified, the
20377 format is guessed from @var{movie_name} or by probing.
20379 @item seek_point, sp
20380 Specifies the seek point in seconds. The frames will be output
20381 starting from this seek point. The parameter is evaluated with
20382 @code{av_strtod}, so the numerical value may be suffixed by an IS
20383 postfix. The default value is "0".
20386 Specifies the streams to read. Several streams can be specified,
20387 separated by "+". The source will then have as many outputs, in the
20388 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
20389 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
20390 respectively the default (best suited) video and audio stream. Default
20391 is "dv", or "da" if the filter is called as "amovie".
20393 @item stream_index, si
20394 Specifies the index of the video stream to read. If the value is -1,
20395 the most suitable video stream will be automatically selected. The default
20396 value is "-1". Deprecated. If the filter is called "amovie", it will select
20397 audio instead of video.
20400 Specifies how many times to read the stream in sequence.
20401 If the value is 0, the stream will be looped infinitely.
20402 Default value is "1".
20404 Note that when the movie is looped the source timestamps are not
20405 changed, so it will generate non monotonically increasing timestamps.
20407 @item discontinuity
20408 Specifies the time difference between frames above which the point is
20409 considered a timestamp discontinuity which is removed by adjusting the later
20413 It allows overlaying a second video on top of the main input of
20414 a filtergraph, as shown in this graph:
20416 input -----------> deltapts0 --> overlay --> output
20419 movie --> scale--> deltapts1 -------+
20421 @subsection Examples
20425 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
20426 on top of the input labelled "in":
20428 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
20429 [in] setpts=PTS-STARTPTS [main];
20430 [main][over] overlay=16:16 [out]
20434 Read from a video4linux2 device, and overlay it on top of the input
20437 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
20438 [in] setpts=PTS-STARTPTS [main];
20439 [main][over] overlay=16:16 [out]
20443 Read the first video stream and the audio stream with id 0x81 from
20444 dvd.vob; the video is connected to the pad named "video" and the audio is
20445 connected to the pad named "audio":
20447 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
20451 @subsection Commands
20453 Both movie and amovie support the following commands:
20456 Perform seek using "av_seek_frame".
20457 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
20460 @var{stream_index}: If stream_index is -1, a default
20461 stream is selected, and @var{timestamp} is automatically converted
20462 from AV_TIME_BASE units to the stream specific time_base.
20464 @var{timestamp}: Timestamp in AVStream.time_base units
20465 or, if no stream is specified, in AV_TIME_BASE units.
20467 @var{flags}: Flags which select direction and seeking mode.
20471 Get movie duration in AV_TIME_BASE units.
20475 @c man end MULTIMEDIA SOURCES