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.
964 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
965 Allowed range is 0.1 to 60 seconds.
972 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
974 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
981 Set output format constraints for the input audio. The framework will
982 negotiate the most appropriate format to minimize conversions.
984 It accepts the following parameters:
988 A '|'-separated list of requested sample formats.
991 A '|'-separated list of requested sample rates.
993 @item channel_layouts
994 A '|'-separated list of requested channel layouts.
996 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
997 for the required syntax.
1000 If a parameter is omitted, all values are allowed.
1002 Force the output to either unsigned 8-bit or signed 16-bit stereo
1004 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1009 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1010 processing reduces disturbing noise between useful signals.
1012 Gating is done by detecting the volume below a chosen level @var{threshold}
1013 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1014 floor is set via @var{range}. Because an exact manipulation of the signal
1015 would cause distortion of the waveform the reduction can be levelled over
1016 time. This is done by setting @var{attack} and @var{release}.
1018 @var{attack} determines how long the signal has to fall below the threshold
1019 before any reduction will occur and @var{release} sets the time the signal
1020 has to rise above the threshold to reduce the reduction again.
1021 Shorter signals than the chosen attack time will be left untouched.
1025 Set input level before filtering.
1026 Default is 1. Allowed range is from 0.015625 to 64.
1029 Set the level of gain reduction when the signal is below the threshold.
1030 Default is 0.06125. Allowed range is from 0 to 1.
1033 If a signal rises above this level the gain reduction is released.
1034 Default is 0.125. Allowed range is from 0 to 1.
1037 Set a ratio by which the signal is reduced.
1038 Default is 2. Allowed range is from 1 to 9000.
1041 Amount of milliseconds the signal has to rise above the threshold before gain
1043 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1046 Amount of milliseconds the signal has to fall below the threshold before the
1047 reduction is increased again. Default is 250 milliseconds.
1048 Allowed range is from 0.01 to 9000.
1051 Set amount of amplification of signal after processing.
1052 Default is 1. Allowed range is from 1 to 64.
1055 Curve the sharp knee around the threshold to enter gain reduction more softly.
1056 Default is 2.828427125. Allowed range is from 1 to 8.
1059 Choose if exact signal should be taken for detection or an RMS like one.
1060 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1063 Choose if the average level between all channels or the louder channel affects
1065 Default is @code{average}. Can be @code{average} or @code{maximum}.
1070 Apply an arbitrary Infinite Impulse Response filter.
1072 It accepts the following parameters:
1076 Set numerator/zeros coefficients.
1079 Set denominator/poles coefficients.
1091 Set coefficients format.
1097 Z-plane zeros/poles, cartesian (default)
1099 Z-plane zeros/poles, polar radians
1101 Z-plane zeros/poles, polar degrees
1105 Set kind of processing.
1106 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1109 Set filtering precision.
1113 double-precision floating-point (default)
1115 single-precision floating-point
1124 Coefficients in @code{tf} format are separated by spaces and are in ascending
1127 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1128 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1131 Different coefficients and gains can be provided for every channel, in such case
1132 use '|' to separate coefficients or gains. Last provided coefficients will be
1133 used for all remaining channels.
1135 @subsection Examples
1139 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1141 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
1145 Same as above but in @code{zp} format:
1147 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
1153 The limiter prevents an input signal from rising over a desired threshold.
1154 This limiter uses lookahead technology to prevent your signal from distorting.
1155 It means that there is a small delay after the signal is processed. Keep in mind
1156 that the delay it produces is the attack time you set.
1158 The filter accepts the following options:
1162 Set input gain. Default is 1.
1165 Set output gain. Default is 1.
1168 Don't let signals above this level pass the limiter. Default is 1.
1171 The limiter will reach its attenuation level in this amount of time in
1172 milliseconds. Default is 5 milliseconds.
1175 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1176 Default is 50 milliseconds.
1179 When gain reduction is always needed ASC takes care of releasing to an
1180 average reduction level rather than reaching a reduction of 0 in the release
1184 Select how much the release time is affected by ASC, 0 means nearly no changes
1185 in release time while 1 produces higher release times.
1188 Auto level output signal. Default is enabled.
1189 This normalizes audio back to 0dB if enabled.
1192 Depending on picked setting it is recommended to upsample input 2x or 4x times
1193 with @ref{aresample} before applying this filter.
1197 Apply a two-pole all-pass filter with central frequency (in Hz)
1198 @var{frequency}, and filter-width @var{width}.
1199 An all-pass filter changes the audio's frequency to phase relationship
1200 without changing its frequency to amplitude relationship.
1202 The filter accepts the following options:
1206 Set frequency in Hz.
1209 Set method to specify band-width of filter.
1224 Specify the band-width of a filter in width_type units.
1227 Specify which channels to filter, by default all available are filtered.
1230 @subsection Commands
1232 This filter supports the following commands:
1235 Change allpass frequency.
1236 Syntax for the command is : "@var{frequency}"
1239 Change allpass width_type.
1240 Syntax for the command is : "@var{width_type}"
1243 Change allpass width.
1244 Syntax for the command is : "@var{width}"
1251 The filter accepts the following options:
1255 Set the number of loops. Setting this value to -1 will result in infinite loops.
1259 Set maximal number of samples. Default is 0.
1262 Set first sample of loop. Default is 0.
1268 Merge two or more audio streams into a single multi-channel stream.
1270 The filter accepts the following options:
1275 Set the number of inputs. Default is 2.
1279 If the channel layouts of the inputs are disjoint, and therefore compatible,
1280 the channel layout of the output will be set accordingly and the channels
1281 will be reordered as necessary. If the channel layouts of the inputs are not
1282 disjoint, the output will have all the channels of the first input then all
1283 the channels of the second input, in that order, and the channel layout of
1284 the output will be the default value corresponding to the total number of
1287 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1288 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1289 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1290 first input, b1 is the first channel of the second input).
1292 On the other hand, if both input are in stereo, the output channels will be
1293 in the default order: a1, a2, b1, b2, and the channel layout will be
1294 arbitrarily set to 4.0, which may or may not be the expected value.
1296 All inputs must have the same sample rate, and format.
1298 If inputs do not have the same duration, the output will stop with the
1301 @subsection Examples
1305 Merge two mono files into a stereo stream:
1307 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1311 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1313 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
1319 Mixes multiple audio inputs into a single output.
1321 Note that this filter only supports float samples (the @var{amerge}
1322 and @var{pan} audio filters support many formats). If the @var{amix}
1323 input has integer samples then @ref{aresample} will be automatically
1324 inserted to perform the conversion to float samples.
1328 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1330 will mix 3 input audio streams to a single output with the same duration as the
1331 first input and a dropout transition time of 3 seconds.
1333 It accepts the following parameters:
1337 The number of inputs. If unspecified, it defaults to 2.
1340 How to determine the end-of-stream.
1344 The duration of the longest input. (default)
1347 The duration of the shortest input.
1350 The duration of the first input.
1354 @item dropout_transition
1355 The transition time, in seconds, for volume renormalization when an input
1356 stream ends. The default value is 2 seconds.
1359 Specify weight of each input audio stream as sequence.
1360 Each weight is separated by space. By default all inputs have same weight.
1363 @section anequalizer
1365 High-order parametric multiband equalizer for each channel.
1367 It accepts the following parameters:
1371 This option string is in format:
1372 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1373 Each equalizer band is separated by '|'.
1377 Set channel number to which equalization will be applied.
1378 If input doesn't have that channel the entry is ignored.
1381 Set central frequency for band.
1382 If input doesn't have that frequency the entry is ignored.
1385 Set band width in hertz.
1388 Set band gain in dB.
1391 Set filter type for band, optional, can be:
1395 Butterworth, this is default.
1406 With this option activated frequency response of anequalizer is displayed
1410 Set video stream size. Only useful if curves option is activated.
1413 Set max gain that will be displayed. Only useful if curves option is activated.
1414 Setting this to a reasonable value makes it possible to display gain which is derived from
1415 neighbour bands which are too close to each other and thus produce higher gain
1416 when both are activated.
1419 Set frequency scale used to draw frequency response in video output.
1420 Can be linear or logarithmic. Default is logarithmic.
1423 Set color for each channel curve which is going to be displayed in video stream.
1424 This is list of color names separated by space or by '|'.
1425 Unrecognised or missing colors will be replaced by white color.
1428 @subsection Examples
1432 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1433 for first 2 channels using Chebyshev type 1 filter:
1435 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1439 @subsection Commands
1441 This filter supports the following commands:
1444 Alter existing filter parameters.
1445 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1447 @var{fN} is existing filter number, starting from 0, if no such filter is available
1449 @var{freq} set new frequency parameter.
1450 @var{width} set new width parameter in herz.
1451 @var{gain} set new gain parameter in dB.
1453 Full filter invocation with asendcmd may look like this:
1454 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1459 Pass the audio source unchanged to the output.
1463 Pad the end of an audio stream with silence.
1465 This can be used together with @command{ffmpeg} @option{-shortest} to
1466 extend audio streams to the same length as the video stream.
1468 A description of the accepted options follows.
1472 Set silence packet size. Default value is 4096.
1475 Set the number of samples of silence to add to the end. After the
1476 value is reached, the stream is terminated. This option is mutually
1477 exclusive with @option{whole_len}.
1480 Set the minimum total number of samples in the output audio stream. If
1481 the value is longer than the input audio length, silence is added to
1482 the end, until the value is reached. This option is mutually exclusive
1483 with @option{pad_len}.
1486 If neither the @option{pad_len} nor the @option{whole_len} option is
1487 set, the filter will add silence to the end of the input stream
1490 @subsection Examples
1494 Add 1024 samples of silence to the end of the input:
1500 Make sure the audio output will contain at least 10000 samples, pad
1501 the input with silence if required:
1503 apad=whole_len=10000
1507 Use @command{ffmpeg} to pad the audio input with silence, so that the
1508 video stream will always result the shortest and will be converted
1509 until the end in the output file when using the @option{shortest}
1512 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1517 Add a phasing effect to the input audio.
1519 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1520 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1522 A description of the accepted parameters follows.
1526 Set input gain. Default is 0.4.
1529 Set output gain. Default is 0.74
1532 Set delay in milliseconds. Default is 3.0.
1535 Set decay. Default is 0.4.
1538 Set modulation speed in Hz. Default is 0.5.
1541 Set modulation type. Default is triangular.
1543 It accepts the following values:
1552 Audio pulsator is something between an autopanner and a tremolo.
1553 But it can produce funny stereo effects as well. Pulsator changes the volume
1554 of the left and right channel based on a LFO (low frequency oscillator) with
1555 different waveforms and shifted phases.
1556 This filter have the ability to define an offset between left and right
1557 channel. An offset of 0 means that both LFO shapes match each other.
1558 The left and right channel are altered equally - a conventional tremolo.
1559 An offset of 50% means that the shape of the right channel is exactly shifted
1560 in phase (or moved backwards about half of the frequency) - pulsator acts as
1561 an autopanner. At 1 both curves match again. Every setting in between moves the
1562 phase shift gapless between all stages and produces some "bypassing" sounds with
1563 sine and triangle waveforms. The more you set the offset near 1 (starting from
1564 the 0.5) the faster the signal passes from the left to the right speaker.
1566 The filter accepts the following options:
1570 Set input gain. By default it is 1. Range is [0.015625 - 64].
1573 Set output gain. By default it is 1. Range is [0.015625 - 64].
1576 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1577 sawup or sawdown. Default is sine.
1580 Set modulation. Define how much of original signal is affected by the LFO.
1583 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1586 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1589 Set pulse width. Default is 1. Allowed range is [0 - 2].
1592 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1595 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1599 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1603 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1604 if timing is set to hz.
1610 Resample the input audio to the specified parameters, using the
1611 libswresample library. If none are specified then the filter will
1612 automatically convert between its input and output.
1614 This filter is also able to stretch/squeeze the audio data to make it match
1615 the timestamps or to inject silence / cut out audio to make it match the
1616 timestamps, do a combination of both or do neither.
1618 The filter accepts the syntax
1619 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1620 expresses a sample rate and @var{resampler_options} is a list of
1621 @var{key}=@var{value} pairs, separated by ":". See the
1622 @ref{Resampler Options,,"Resampler Options" section in the
1623 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1624 for the complete list of supported options.
1626 @subsection Examples
1630 Resample the input audio to 44100Hz:
1636 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1637 samples per second compensation:
1639 aresample=async=1000
1645 Reverse an audio clip.
1647 Warning: This filter requires memory to buffer the entire clip, so trimming
1650 @subsection Examples
1654 Take the first 5 seconds of a clip, and reverse it.
1656 atrim=end=5,areverse
1660 @section asetnsamples
1662 Set the number of samples per each output audio frame.
1664 The last output packet may contain a different number of samples, as
1665 the filter will flush all the remaining samples when the input audio
1668 The filter accepts the following options:
1672 @item nb_out_samples, n
1673 Set the number of frames per each output audio frame. The number is
1674 intended as the number of samples @emph{per each channel}.
1675 Default value is 1024.
1678 If set to 1, the filter will pad the last audio frame with zeroes, so
1679 that the last frame will contain the same number of samples as the
1680 previous ones. Default value is 1.
1683 For example, to set the number of per-frame samples to 1234 and
1684 disable padding for the last frame, use:
1686 asetnsamples=n=1234:p=0
1691 Set the sample rate without altering the PCM data.
1692 This will result in a change of speed and pitch.
1694 The filter accepts the following options:
1697 @item sample_rate, r
1698 Set the output sample rate. Default is 44100 Hz.
1703 Show a line containing various information for each input audio frame.
1704 The input audio is not modified.
1706 The shown line contains a sequence of key/value pairs of the form
1707 @var{key}:@var{value}.
1709 The following values are shown in the output:
1713 The (sequential) number of the input frame, starting from 0.
1716 The presentation timestamp of the input frame, in time base units; the time base
1717 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1720 The presentation timestamp of the input frame in seconds.
1723 position of the frame in the input stream, -1 if this information in
1724 unavailable and/or meaningless (for example in case of synthetic audio)
1733 The sample rate for the audio frame.
1736 The number of samples (per channel) in the frame.
1739 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1740 audio, the data is treated as if all the planes were concatenated.
1742 @item plane_checksums
1743 A list of Adler-32 checksums for each data plane.
1749 Display time domain statistical information about the audio channels.
1750 Statistics are calculated and displayed for each audio channel and,
1751 where applicable, an overall figure is also given.
1753 It accepts the following option:
1756 Short window length in seconds, used for peak and trough RMS measurement.
1757 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
1761 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1762 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1765 Available keys for each channel are:
1799 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1800 this @code{lavfi.astats.Overall.Peak_count}.
1802 For description what each key means read below.
1805 Set number of frame after which stats are going to be recalculated.
1806 Default is disabled.
1809 A description of each shown parameter follows:
1813 Mean amplitude displacement from zero.
1816 Minimal sample level.
1819 Maximal sample level.
1821 @item Min difference
1822 Minimal difference between two consecutive samples.
1824 @item Max difference
1825 Maximal difference between two consecutive samples.
1827 @item Mean difference
1828 Mean difference between two consecutive samples.
1829 The average of each difference between two consecutive samples.
1831 @item RMS difference
1832 Root Mean Square difference between two consecutive samples.
1836 Standard peak and RMS level measured in dBFS.
1840 Peak and trough values for RMS level measured over a short window.
1843 Standard ratio of peak to RMS level (note: not in dB).
1846 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1847 (i.e. either @var{Min level} or @var{Max level}).
1850 Number of occasions (not the number of samples) that the signal attained either
1851 @var{Min level} or @var{Max level}.
1854 Overall bit depth of audio. Number of bits used for each sample.
1857 Measured dynamic range of audio in dB.
1864 The filter accepts exactly one parameter, the audio tempo. If not
1865 specified then the filter will assume nominal 1.0 tempo. Tempo must
1866 be in the [0.5, 2.0] range.
1868 @subsection Examples
1872 Slow down audio to 80% tempo:
1878 To speed up audio to 125% tempo:
1886 Trim the input so that the output contains one continuous subpart of the input.
1888 It accepts the following parameters:
1891 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1892 sample with the timestamp @var{start} will be the first sample in the output.
1895 Specify time of the first audio sample that will be dropped, i.e. the
1896 audio sample immediately preceding the one with the timestamp @var{end} will be
1897 the last sample in the output.
1900 Same as @var{start}, except this option sets the start timestamp in samples
1904 Same as @var{end}, except this option sets the end timestamp in samples instead
1908 The maximum duration of the output in seconds.
1911 The number of the first sample that should be output.
1914 The number of the first sample that should be dropped.
1917 @option{start}, @option{end}, and @option{duration} are expressed as time
1918 duration specifications; see
1919 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1921 Note that the first two sets of the start/end options and the @option{duration}
1922 option look at the frame timestamp, while the _sample options simply count the
1923 samples that pass through the filter. So start/end_pts and start/end_sample will
1924 give different results when the timestamps are wrong, inexact or do not start at
1925 zero. Also note that this filter does not modify the timestamps. If you wish
1926 to have the output timestamps start at zero, insert the asetpts filter after the
1929 If multiple start or end options are set, this filter tries to be greedy and
1930 keep all samples that match at least one of the specified constraints. To keep
1931 only the part that matches all the constraints at once, chain multiple atrim
1934 The defaults are such that all the input is kept. So it is possible to set e.g.
1935 just the end values to keep everything before the specified time.
1940 Drop everything except the second minute of input:
1942 ffmpeg -i INPUT -af atrim=60:120
1946 Keep only the first 1000 samples:
1948 ffmpeg -i INPUT -af atrim=end_sample=1000
1955 Apply a two-pole Butterworth band-pass filter with central
1956 frequency @var{frequency}, and (3dB-point) band-width width.
1957 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1958 instead of the default: constant 0dB peak gain.
1959 The filter roll off at 6dB per octave (20dB per decade).
1961 The filter accepts the following options:
1965 Set the filter's central frequency. Default is @code{3000}.
1968 Constant skirt gain if set to 1. Defaults to 0.
1971 Set method to specify band-width of filter.
1986 Specify the band-width of a filter in width_type units.
1989 Specify which channels to filter, by default all available are filtered.
1992 @subsection Commands
1994 This filter supports the following commands:
1997 Change bandpass frequency.
1998 Syntax for the command is : "@var{frequency}"
2001 Change bandpass width_type.
2002 Syntax for the command is : "@var{width_type}"
2005 Change bandpass width.
2006 Syntax for the command is : "@var{width}"
2011 Apply a two-pole Butterworth band-reject filter with central
2012 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2013 The filter roll off at 6dB per octave (20dB per decade).
2015 The filter accepts the following options:
2019 Set the filter's central frequency. Default is @code{3000}.
2022 Set method to specify band-width of filter.
2037 Specify the band-width of a filter in width_type units.
2040 Specify which channels to filter, by default all available are filtered.
2043 @subsection Commands
2045 This filter supports the following commands:
2048 Change bandreject frequency.
2049 Syntax for the command is : "@var{frequency}"
2052 Change bandreject width_type.
2053 Syntax for the command is : "@var{width_type}"
2056 Change bandreject width.
2057 Syntax for the command is : "@var{width}"
2060 @section bass, lowshelf
2062 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2063 shelving filter with a response similar to that of a standard
2064 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2066 The filter accepts the following options:
2070 Give the gain at 0 Hz. Its useful range is about -20
2071 (for a large cut) to +20 (for a large boost).
2072 Beware of clipping when using a positive gain.
2075 Set the filter's central frequency and so can be used
2076 to extend or reduce the frequency range to be boosted or cut.
2077 The default value is @code{100} Hz.
2080 Set method to specify band-width of filter.
2095 Determine how steep is the filter's shelf transition.
2098 Specify which channels to filter, by default all available are filtered.
2101 @subsection Commands
2103 This filter supports the following commands:
2106 Change bass frequency.
2107 Syntax for the command is : "@var{frequency}"
2110 Change bass width_type.
2111 Syntax for the command is : "@var{width_type}"
2115 Syntax for the command is : "@var{width}"
2119 Syntax for the command is : "@var{gain}"
2124 Apply a biquad IIR filter with the given coefficients.
2125 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2126 are the numerator and denominator coefficients respectively.
2127 and @var{channels}, @var{c} specify which channels to filter, by default all
2128 available are filtered.
2130 @subsection Commands
2132 This filter supports the following commands:
2140 Change biquad parameter.
2141 Syntax for the command is : "@var{value}"
2145 Bauer stereo to binaural transformation, which improves headphone listening of
2146 stereo audio records.
2148 To enable compilation of this filter you need to configure FFmpeg with
2149 @code{--enable-libbs2b}.
2151 It accepts the following parameters:
2155 Pre-defined crossfeed level.
2159 Default level (fcut=700, feed=50).
2162 Chu Moy circuit (fcut=700, feed=60).
2165 Jan Meier circuit (fcut=650, feed=95).
2170 Cut frequency (in Hz).
2179 Remap input channels to new locations.
2181 It accepts the following parameters:
2184 Map channels from input to output. The argument is a '|'-separated list of
2185 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2186 @var{in_channel} form. @var{in_channel} can be either the name of the input
2187 channel (e.g. FL for front left) or its index in the input channel layout.
2188 @var{out_channel} is the name of the output channel or its index in the output
2189 channel layout. If @var{out_channel} is not given then it is implicitly an
2190 index, starting with zero and increasing by one for each mapping.
2192 @item channel_layout
2193 The channel layout of the output stream.
2196 If no mapping is present, the filter will implicitly map input channels to
2197 output channels, preserving indices.
2199 @subsection Examples
2203 For example, assuming a 5.1+downmix input MOV file,
2205 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2207 will create an output WAV file tagged as stereo from the downmix channels of
2211 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2213 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2217 @section channelsplit
2219 Split each channel from an input audio stream into a separate output stream.
2221 It accepts the following parameters:
2223 @item channel_layout
2224 The channel layout of the input stream. The default is "stereo".
2226 A channel layout describing the channels to be extracted as separate output streams
2227 or "all" to extract each input channel as a separate stream. The default is "all".
2229 Choosing channels not present in channel layout in the input will result in an error.
2232 @subsection Examples
2236 For example, assuming a stereo input MP3 file,
2238 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2240 will create an output Matroska file with two audio streams, one containing only
2241 the left channel and the other the right channel.
2244 Split a 5.1 WAV file into per-channel files:
2246 ffmpeg -i in.wav -filter_complex
2247 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2248 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2249 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2254 Extract only LFE from a 5.1 WAV file:
2256 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2257 -map '[LFE]' lfe.wav
2262 Add a chorus effect to the audio.
2264 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2266 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2267 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2268 The modulation depth defines the range the modulated delay is played before or after
2269 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2270 sound tuned around the original one, like in a chorus where some vocals are slightly
2273 It accepts the following parameters:
2276 Set input gain. Default is 0.4.
2279 Set output gain. Default is 0.4.
2282 Set delays. A typical delay is around 40ms to 60ms.
2294 @subsection Examples
2300 chorus=0.7:0.9:55:0.4:0.25:2
2306 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2310 Fuller sounding chorus with three delays:
2312 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
2317 Compress or expand the audio's dynamic range.
2319 It accepts the following parameters:
2325 A list of times in seconds for each channel over which the instantaneous level
2326 of the input signal is averaged to determine its volume. @var{attacks} refers to
2327 increase of volume and @var{decays} refers to decrease of volume. For most
2328 situations, the attack time (response to the audio getting louder) should be
2329 shorter than the decay time, because the human ear is more sensitive to sudden
2330 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2331 a typical value for decay is 0.8 seconds.
2332 If specified number of attacks & decays is lower than number of channels, the last
2333 set attack/decay will be used for all remaining channels.
2336 A list of points for the transfer function, specified in dB relative to the
2337 maximum possible signal amplitude. Each key points list must be defined using
2338 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2339 @code{x0/y0 x1/y1 x2/y2 ....}
2341 The input values must be in strictly increasing order but the transfer function
2342 does not have to be monotonically rising. The point @code{0/0} is assumed but
2343 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2344 function are @code{-70/-70|-60/-20|1/0}.
2347 Set the curve radius in dB for all joints. It defaults to 0.01.
2350 Set the additional gain in dB to be applied at all points on the transfer
2351 function. This allows for easy adjustment of the overall gain.
2355 Set an initial volume, in dB, to be assumed for each channel when filtering
2356 starts. This permits the user to supply a nominal level initially, so that, for
2357 example, a very large gain is not applied to initial signal levels before the
2358 companding has begun to operate. A typical value for audio which is initially
2359 quiet is -90 dB. It defaults to 0.
2362 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2363 delayed before being fed to the volume adjuster. Specifying a delay
2364 approximately equal to the attack/decay times allows the filter to effectively
2365 operate in predictive rather than reactive mode. It defaults to 0.
2369 @subsection Examples
2373 Make music with both quiet and loud passages suitable for listening to in a
2376 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2379 Another example for audio with whisper and explosion parts:
2381 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2385 A noise gate for when the noise is at a lower level than the signal:
2387 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2391 Here is another noise gate, this time for when the noise is at a higher level
2392 than the signal (making it, in some ways, similar to squelch):
2394 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2398 2:1 compression starting at -6dB:
2400 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2404 2:1 compression starting at -9dB:
2406 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2410 2:1 compression starting at -12dB:
2412 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2416 2:1 compression starting at -18dB:
2418 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2422 3:1 compression starting at -15dB:
2424 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2430 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2436 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
2440 Hard limiter at -6dB:
2442 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2446 Hard limiter at -12dB:
2448 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2452 Hard noise gate at -35 dB:
2454 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2460 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2464 @section compensationdelay
2466 Compensation Delay Line is a metric based delay to compensate differing
2467 positions of microphones or speakers.
2469 For example, you have recorded guitar with two microphones placed in
2470 different location. Because the front of sound wave has fixed speed in
2471 normal conditions, the phasing of microphones can vary and depends on
2472 their location and interposition. The best sound mix can be achieved when
2473 these microphones are in phase (synchronized). Note that distance of
2474 ~30 cm between microphones makes one microphone to capture signal in
2475 antiphase to another microphone. That makes the final mix sounding moody.
2476 This filter helps to solve phasing problems by adding different delays
2477 to each microphone track and make them synchronized.
2479 The best result can be reached when you take one track as base and
2480 synchronize other tracks one by one with it.
2481 Remember that synchronization/delay tolerance depends on sample rate, too.
2482 Higher sample rates will give more tolerance.
2484 It accepts the following parameters:
2488 Set millimeters distance. This is compensation distance for fine tuning.
2492 Set cm distance. This is compensation distance for tightening distance setup.
2496 Set meters distance. This is compensation distance for hard distance setup.
2500 Set dry amount. Amount of unprocessed (dry) signal.
2504 Set wet amount. Amount of processed (wet) signal.
2508 Set temperature degree in Celsius. This is the temperature of the environment.
2513 Apply headphone crossfeed filter.
2515 Crossfeed is the process of blending the left and right channels of stereo
2517 It is mainly used to reduce extreme stereo separation of low frequencies.
2519 The intent is to produce more speaker like sound to the listener.
2521 The filter accepts the following options:
2525 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2526 This sets gain of low shelf filter for side part of stereo image.
2527 Default is -6dB. Max allowed is -30db when strength is set to 1.
2530 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2531 This sets cut off frequency of low shelf filter. Default is cut off near
2532 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2535 Set input gain. Default is 0.9.
2538 Set output gain. Default is 1.
2541 @section crystalizer
2542 Simple algorithm to expand audio dynamic range.
2544 The filter accepts the following options:
2548 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2549 (unchanged sound) to 10.0 (maximum effect).
2552 Enable clipping. By default is enabled.
2556 Apply a DC shift to the audio.
2558 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2559 in the recording chain) from the audio. The effect of a DC offset is reduced
2560 headroom and hence volume. The @ref{astats} filter can be used to determine if
2561 a signal has a DC offset.
2565 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2569 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2570 used to prevent clipping.
2574 Measure audio dynamic range.
2576 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2577 is found in transition material. And anything less that 8 have very poor dynamics
2578 and is very compressed.
2580 The filter accepts the following options:
2584 Set window length in seconds used to split audio into segments of equal length.
2585 Default is 3 seconds.
2589 Dynamic Audio Normalizer.
2591 This filter applies a certain amount of gain to the input audio in order
2592 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2593 contrast to more "simple" normalization algorithms, the Dynamic Audio
2594 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2595 This allows for applying extra gain to the "quiet" sections of the audio
2596 while avoiding distortions or clipping the "loud" sections. In other words:
2597 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2598 sections, in the sense that the volume of each section is brought to the
2599 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2600 this goal *without* applying "dynamic range compressing". It will retain 100%
2601 of the dynamic range *within* each section of the audio file.
2605 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2606 Default is 500 milliseconds.
2607 The Dynamic Audio Normalizer processes the input audio in small chunks,
2608 referred to as frames. This is required, because a peak magnitude has no
2609 meaning for just a single sample value. Instead, we need to determine the
2610 peak magnitude for a contiguous sequence of sample values. While a "standard"
2611 normalizer would simply use the peak magnitude of the complete file, the
2612 Dynamic Audio Normalizer determines the peak magnitude individually for each
2613 frame. The length of a frame is specified in milliseconds. By default, the
2614 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2615 been found to give good results with most files.
2616 Note that the exact frame length, in number of samples, will be determined
2617 automatically, based on the sampling rate of the individual input audio file.
2620 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2621 number. Default is 31.
2622 Probably the most important parameter of the Dynamic Audio Normalizer is the
2623 @code{window size} of the Gaussian smoothing filter. The filter's window size
2624 is specified in frames, centered around the current frame. For the sake of
2625 simplicity, this must be an odd number. Consequently, the default value of 31
2626 takes into account the current frame, as well as the 15 preceding frames and
2627 the 15 subsequent frames. Using a larger window results in a stronger
2628 smoothing effect and thus in less gain variation, i.e. slower gain
2629 adaptation. Conversely, using a smaller window results in a weaker smoothing
2630 effect and thus in more gain variation, i.e. faster gain adaptation.
2631 In other words, the more you increase this value, the more the Dynamic Audio
2632 Normalizer will behave like a "traditional" normalization filter. On the
2633 contrary, the more you decrease this value, the more the Dynamic Audio
2634 Normalizer will behave like a dynamic range compressor.
2637 Set the target peak value. This specifies the highest permissible magnitude
2638 level for the normalized audio input. This filter will try to approach the
2639 target peak magnitude as closely as possible, but at the same time it also
2640 makes sure that the normalized signal will never exceed the peak magnitude.
2641 A frame's maximum local gain factor is imposed directly by the target peak
2642 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2643 It is not recommended to go above this value.
2646 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2647 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2648 factor for each input frame, i.e. the maximum gain factor that does not
2649 result in clipping or distortion. The maximum gain factor is determined by
2650 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2651 additionally bounds the frame's maximum gain factor by a predetermined
2652 (global) maximum gain factor. This is done in order to avoid excessive gain
2653 factors in "silent" or almost silent frames. By default, the maximum gain
2654 factor is 10.0, For most inputs the default value should be sufficient and
2655 it usually is not recommended to increase this value. Though, for input
2656 with an extremely low overall volume level, it may be necessary to allow even
2657 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2658 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2659 Instead, a "sigmoid" threshold function will be applied. This way, the
2660 gain factors will smoothly approach the threshold value, but never exceed that
2664 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2665 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2666 This means that the maximum local gain factor for each frame is defined
2667 (only) by the frame's highest magnitude sample. This way, the samples can
2668 be amplified as much as possible without exceeding the maximum signal
2669 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2670 Normalizer can also take into account the frame's root mean square,
2671 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2672 determine the power of a time-varying signal. It is therefore considered
2673 that the RMS is a better approximation of the "perceived loudness" than
2674 just looking at the signal's peak magnitude. Consequently, by adjusting all
2675 frames to a constant RMS value, a uniform "perceived loudness" can be
2676 established. If a target RMS value has been specified, a frame's local gain
2677 factor is defined as the factor that would result in exactly that RMS value.
2678 Note, however, that the maximum local gain factor is still restricted by the
2679 frame's highest magnitude sample, in order to prevent clipping.
2682 Enable channels coupling. By default is enabled.
2683 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2684 amount. This means the same gain factor will be applied to all channels, i.e.
2685 the maximum possible gain factor is determined by the "loudest" channel.
2686 However, in some recordings, it may happen that the volume of the different
2687 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2688 In this case, this option can be used to disable the channel coupling. This way,
2689 the gain factor will be determined independently for each channel, depending
2690 only on the individual channel's highest magnitude sample. This allows for
2691 harmonizing the volume of the different channels.
2694 Enable DC bias correction. By default is disabled.
2695 An audio signal (in the time domain) is a sequence of sample values.
2696 In the Dynamic Audio Normalizer these sample values are represented in the
2697 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2698 audio signal, or "waveform", should be centered around the zero point.
2699 That means if we calculate the mean value of all samples in a file, or in a
2700 single frame, then the result should be 0.0 or at least very close to that
2701 value. If, however, there is a significant deviation of the mean value from
2702 0.0, in either positive or negative direction, this is referred to as a
2703 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2704 Audio Normalizer provides optional DC bias correction.
2705 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2706 the mean value, or "DC correction" offset, of each input frame and subtract
2707 that value from all of the frame's sample values which ensures those samples
2708 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2709 boundaries, the DC correction offset values will be interpolated smoothly
2710 between neighbouring frames.
2713 Enable alternative boundary mode. By default is disabled.
2714 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2715 around each frame. This includes the preceding frames as well as the
2716 subsequent frames. However, for the "boundary" frames, located at the very
2717 beginning and at the very end of the audio file, not all neighbouring
2718 frames are available. In particular, for the first few frames in the audio
2719 file, the preceding frames are not known. And, similarly, for the last few
2720 frames in the audio file, the subsequent frames are not known. Thus, the
2721 question arises which gain factors should be assumed for the missing frames
2722 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2723 to deal with this situation. The default boundary mode assumes a gain factor
2724 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2725 "fade out" at the beginning and at the end of the input, respectively.
2728 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2729 By default, the Dynamic Audio Normalizer does not apply "traditional"
2730 compression. This means that signal peaks will not be pruned and thus the
2731 full dynamic range will be retained within each local neighbourhood. However,
2732 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2733 normalization algorithm with a more "traditional" compression.
2734 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2735 (thresholding) function. If (and only if) the compression feature is enabled,
2736 all input frames will be processed by a soft knee thresholding function prior
2737 to the actual normalization process. Put simply, the thresholding function is
2738 going to prune all samples whose magnitude exceeds a certain threshold value.
2739 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2740 value. Instead, the threshold value will be adjusted for each individual
2742 In general, smaller parameters result in stronger compression, and vice versa.
2743 Values below 3.0 are not recommended, because audible distortion may appear.
2748 Make audio easier to listen to on headphones.
2750 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2751 so that when listened to on headphones the stereo image is moved from
2752 inside your head (standard for headphones) to outside and in front of
2753 the listener (standard for speakers).
2759 Apply a two-pole peaking equalisation (EQ) filter. With this
2760 filter, the signal-level at and around a selected frequency can
2761 be increased or decreased, whilst (unlike bandpass and bandreject
2762 filters) that at all other frequencies is unchanged.
2764 In order to produce complex equalisation curves, this filter can
2765 be given several times, each with a different central frequency.
2767 The filter accepts the following options:
2771 Set the filter's central frequency in Hz.
2774 Set method to specify band-width of filter.
2789 Specify the band-width of a filter in width_type units.
2792 Set the required gain or attenuation in dB.
2793 Beware of clipping when using a positive gain.
2796 Specify which channels to filter, by default all available are filtered.
2799 @subsection Examples
2802 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2804 equalizer=f=1000:t=h:width=200:g=-10
2808 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2810 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
2814 @subsection Commands
2816 This filter supports the following commands:
2819 Change equalizer frequency.
2820 Syntax for the command is : "@var{frequency}"
2823 Change equalizer width_type.
2824 Syntax for the command is : "@var{width_type}"
2827 Change equalizer width.
2828 Syntax for the command is : "@var{width}"
2831 Change equalizer gain.
2832 Syntax for the command is : "@var{gain}"
2835 @section extrastereo
2837 Linearly increases the difference between left and right channels which
2838 adds some sort of "live" effect to playback.
2840 The filter accepts the following options:
2844 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2845 (average of both channels), with 1.0 sound will be unchanged, with
2846 -1.0 left and right channels will be swapped.
2849 Enable clipping. By default is enabled.
2852 @section firequalizer
2853 Apply FIR Equalization using arbitrary frequency response.
2855 The filter accepts the following option:
2859 Set gain curve equation (in dB). The expression can contain variables:
2862 the evaluated frequency
2866 channel number, set to 0 when multichannels evaluation is disabled
2868 channel id, see libavutil/channel_layout.h, set to the first channel id when
2869 multichannels evaluation is disabled
2873 channel_layout, see libavutil/channel_layout.h
2878 @item gain_interpolate(f)
2879 interpolate gain on frequency f based on gain_entry
2880 @item cubic_interpolate(f)
2881 same as gain_interpolate, but smoother
2883 This option is also available as command. Default is @code{gain_interpolate(f)}.
2886 Set gain entry for gain_interpolate function. The expression can
2890 store gain entry at frequency f with value g
2892 This option is also available as command.
2895 Set filter delay in seconds. Higher value means more accurate.
2896 Default is @code{0.01}.
2899 Set filter accuracy in Hz. Lower value means more accurate.
2900 Default is @code{5}.
2903 Set window function. Acceptable values are:
2906 rectangular window, useful when gain curve is already smooth
2908 hann window (default)
2914 3-terms continuous 1st derivative nuttall window
2916 minimum 3-terms discontinuous nuttall window
2918 4-terms continuous 1st derivative nuttall window
2920 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2922 blackman-harris window
2928 If enabled, use fixed number of audio samples. This improves speed when
2929 filtering with large delay. Default is disabled.
2932 Enable multichannels evaluation on gain. Default is disabled.
2935 Enable zero phase mode by subtracting timestamp to compensate delay.
2936 Default is disabled.
2939 Set scale used by gain. Acceptable values are:
2942 linear frequency, linear gain
2944 linear frequency, logarithmic (in dB) gain (default)
2946 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2948 logarithmic frequency, logarithmic gain
2952 Set file for dumping, suitable for gnuplot.
2955 Set scale for dumpfile. Acceptable values are same with scale option.
2959 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2960 Default is disabled.
2963 Enable minimum phase impulse response. Default is disabled.
2966 @subsection Examples
2971 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2974 lowpass at 1000 Hz with gain_entry:
2976 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2979 custom equalization:
2981 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2984 higher delay with zero phase to compensate delay:
2986 firequalizer=delay=0.1:fixed=on:zero_phase=on
2989 lowpass on left channel, highpass on right channel:
2991 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2992 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2997 Apply a flanging effect to the audio.
2999 The filter accepts the following options:
3003 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3006 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3009 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3013 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3014 Default value is 71.
3017 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3020 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3021 Default value is @var{sinusoidal}.
3024 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3025 Default value is 25.
3028 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3029 Default is @var{linear}.
3033 Apply Haas effect to audio.
3035 Note that this makes most sense to apply on mono signals.
3036 With this filter applied to mono signals it give some directionality and
3037 stretches its stereo image.
3039 The filter accepts the following options:
3043 Set input level. By default is @var{1}, or 0dB
3046 Set output level. By default is @var{1}, or 0dB.
3049 Set gain applied to side part of signal. By default is @var{1}.
3052 Set kind of middle source. Can be one of the following:
3062 Pick middle part signal of stereo image.
3065 Pick side part signal of stereo image.
3069 Change middle phase. By default is disabled.
3072 Set left channel delay. By default is @var{2.05} milliseconds.
3075 Set left channel balance. By default is @var{-1}.
3078 Set left channel gain. By default is @var{1}.
3081 Change left phase. By default is disabled.
3084 Set right channel delay. By defaults is @var{2.12} milliseconds.
3087 Set right channel balance. By default is @var{1}.
3090 Set right channel gain. By default is @var{1}.
3093 Change right phase. By default is enabled.
3098 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3099 embedded HDCD codes is expanded into a 20-bit PCM stream.
3101 The filter supports the Peak Extend and Low-level Gain Adjustment features
3102 of HDCD, and detects the Transient Filter flag.
3105 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3108 When using the filter with wav, note the default encoding for wav is 16-bit,
3109 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3110 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3112 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3113 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3116 The filter accepts the following options:
3119 @item disable_autoconvert
3120 Disable any automatic format conversion or resampling in the filter graph.
3122 @item process_stereo
3123 Process the stereo channels together. If target_gain does not match between
3124 channels, consider it invalid and use the last valid target_gain.
3127 Set the code detect timer period in ms.
3130 Always extend peaks above -3dBFS even if PE isn't signaled.
3133 Replace audio with a solid tone and adjust the amplitude to signal some
3134 specific aspect of the decoding process. The output file can be loaded in
3135 an audio editor alongside the original to aid analysis.
3137 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3144 Gain adjustment level at each sample
3146 Samples where peak extend occurs
3148 Samples where the code detect timer is active
3150 Samples where the target gain does not match between channels
3156 Apply head-related transfer functions (HRTFs) to create virtual
3157 loudspeakers around the user for binaural listening via headphones.
3158 The HRIRs are provided via additional streams, for each channel
3159 one stereo input stream is needed.
3161 The filter accepts the following options:
3165 Set mapping of input streams for convolution.
3166 The argument is a '|'-separated list of channel names in order as they
3167 are given as additional stream inputs for filter.
3168 This also specify number of input streams. Number of input streams
3169 must be not less than number of channels in first stream plus one.
3172 Set gain applied to audio. Value is in dB. Default is 0.
3175 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3176 processing audio in time domain which is slow.
3177 @var{freq} is processing audio in frequency domain which is fast.
3178 Default is @var{freq}.
3181 Set custom gain for LFE channels. Value is in dB. Default is 0.
3184 Set size of frame in number of samples which will be processed at once.
3185 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3188 Set format of hrir stream.
3189 Default value is @var{stereo}. Alternative value is @var{multich}.
3190 If value is set to @var{stereo}, number of additional streams should
3191 be greater or equal to number of input channels in first input stream.
3192 Also each additional stream should have stereo number of channels.
3193 If value is set to @var{multich}, number of additional streams should
3194 be exactly one. Also number of input channels of additional stream
3195 should be equal or greater than twice number of channels of first input
3199 @subsection Examples
3203 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3204 each amovie filter use stereo file with IR coefficients as input.
3205 The files give coefficients for each position of virtual loudspeaker:
3207 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"
3212 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3213 but now in @var{multich} @var{hrir} format.
3215 ffmpeg -i input.wav -lavfi-complex "amovie=minp.wav[hrirs],[a:0][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3222 Apply a high-pass filter with 3dB point frequency.
3223 The filter can be either single-pole, or double-pole (the default).
3224 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3226 The filter accepts the following options:
3230 Set frequency in Hz. Default is 3000.
3233 Set number of poles. Default is 2.
3236 Set method to specify band-width of filter.
3251 Specify the band-width of a filter in width_type units.
3252 Applies only to double-pole filter.
3253 The default is 0.707q and gives a Butterworth response.
3256 Specify which channels to filter, by default all available are filtered.
3259 @subsection Commands
3261 This filter supports the following commands:
3264 Change highpass frequency.
3265 Syntax for the command is : "@var{frequency}"
3268 Change highpass width_type.
3269 Syntax for the command is : "@var{width_type}"
3272 Change highpass width.
3273 Syntax for the command is : "@var{width}"
3278 Join multiple input streams into one multi-channel stream.
3280 It accepts the following parameters:
3284 The number of input streams. It defaults to 2.
3286 @item channel_layout
3287 The desired output channel layout. It defaults to stereo.
3290 Map channels from inputs to output. The argument is a '|'-separated list of
3291 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3292 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3293 can be either the name of the input channel (e.g. FL for front left) or its
3294 index in the specified input stream. @var{out_channel} is the name of the output
3298 The filter will attempt to guess the mappings when they are not specified
3299 explicitly. It does so by first trying to find an unused matching input channel
3300 and if that fails it picks the first unused input channel.
3302 Join 3 inputs (with properly set channel layouts):
3304 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3307 Build a 5.1 output from 6 single-channel streams:
3309 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3310 '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'
3316 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3318 To enable compilation of this filter you need to configure FFmpeg with
3319 @code{--enable-ladspa}.
3323 Specifies the name of LADSPA plugin library to load. If the environment
3324 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3325 each one of the directories specified by the colon separated list in
3326 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3327 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3328 @file{/usr/lib/ladspa/}.
3331 Specifies the plugin within the library. Some libraries contain only
3332 one plugin, but others contain many of them. If this is not set filter
3333 will list all available plugins within the specified library.
3336 Set the '|' separated list of controls which are zero or more floating point
3337 values that determine the behavior of the loaded plugin (for example delay,
3339 Controls need to be defined using the following syntax:
3340 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3341 @var{valuei} is the value set on the @var{i}-th control.
3342 Alternatively they can be also defined using the following syntax:
3343 @var{value0}|@var{value1}|@var{value2}|..., where
3344 @var{valuei} is the value set on the @var{i}-th control.
3345 If @option{controls} is set to @code{help}, all available controls and
3346 their valid ranges are printed.
3348 @item sample_rate, s
3349 Specify the sample rate, default to 44100. Only used if plugin have
3353 Set the number of samples per channel per each output frame, default
3354 is 1024. Only used if plugin have zero inputs.
3357 Set the minimum duration of the sourced audio. See
3358 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3359 for the accepted syntax.
3360 Note that the resulting duration may be greater than the specified duration,
3361 as the generated audio is always cut at the end of a complete frame.
3362 If not specified, or the expressed duration is negative, the audio is
3363 supposed to be generated forever.
3364 Only used if plugin have zero inputs.
3368 @subsection Examples
3372 List all available plugins within amp (LADSPA example plugin) library:
3378 List all available controls and their valid ranges for @code{vcf_notch}
3379 plugin from @code{VCF} library:
3381 ladspa=f=vcf:p=vcf_notch:c=help
3385 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3388 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3392 Add reverberation to the audio using TAP-plugins
3393 (Tom's Audio Processing plugins):
3395 ladspa=file=tap_reverb:tap_reverb
3399 Generate white noise, with 0.2 amplitude:
3401 ladspa=file=cmt:noise_source_white:c=c0=.2
3405 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3406 @code{C* Audio Plugin Suite} (CAPS) library:
3408 ladspa=file=caps:Click:c=c1=20'
3412 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3414 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3418 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3419 @code{SWH Plugins} collection:
3421 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3425 Attenuate low frequencies using Multiband EQ from Steve Harris
3426 @code{SWH Plugins} collection:
3428 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3432 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3435 ladspa=caps:Narrower
3439 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3441 ladspa=caps:White:.2
3445 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3447 ladspa=caps:Fractal:c=c1=1
3451 Dynamic volume normalization using @code{VLevel} plugin:
3453 ladspa=vlevel-ladspa:vlevel_mono
3457 @subsection Commands
3459 This filter supports the following commands:
3462 Modify the @var{N}-th control value.
3464 If the specified value is not valid, it is ignored and prior one is kept.
3469 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3470 Support for both single pass (livestreams, files) and double pass (files) modes.
3471 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3472 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3473 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3475 The filter accepts the following options:
3479 Set integrated loudness target.
3480 Range is -70.0 - -5.0. Default value is -24.0.
3483 Set loudness range target.
3484 Range is 1.0 - 20.0. Default value is 7.0.
3487 Set maximum true peak.
3488 Range is -9.0 - +0.0. Default value is -2.0.
3490 @item measured_I, measured_i
3491 Measured IL of input file.
3492 Range is -99.0 - +0.0.
3494 @item measured_LRA, measured_lra
3495 Measured LRA of input file.
3496 Range is 0.0 - 99.0.
3498 @item measured_TP, measured_tp
3499 Measured true peak of input file.
3500 Range is -99.0 - +99.0.
3502 @item measured_thresh
3503 Measured threshold of input file.
3504 Range is -99.0 - +0.0.
3507 Set offset gain. Gain is applied before the true-peak limiter.
3508 Range is -99.0 - +99.0. Default is +0.0.
3511 Normalize linearly if possible.
3512 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3513 to be specified in order to use this mode.
3514 Options are true or false. Default is true.
3517 Treat mono input files as "dual-mono". If a mono file is intended for playback
3518 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3519 If set to @code{true}, this option will compensate for this effect.
3520 Multi-channel input files are not affected by this option.
3521 Options are true or false. Default is false.
3524 Set print format for stats. Options are summary, json, or none.
3525 Default value is none.
3530 Apply a low-pass filter with 3dB point frequency.
3531 The filter can be either single-pole or double-pole (the default).
3532 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3534 The filter accepts the following options:
3538 Set frequency in Hz. Default is 500.
3541 Set number of poles. Default is 2.
3544 Set method to specify band-width of filter.
3559 Specify the band-width of a filter in width_type units.
3560 Applies only to double-pole filter.
3561 The default is 0.707q and gives a Butterworth response.
3564 Specify which channels to filter, by default all available are filtered.
3567 @subsection Examples
3570 Lowpass only LFE channel, it LFE is not present it does nothing:
3576 @subsection Commands
3578 This filter supports the following commands:
3581 Change lowpass frequency.
3582 Syntax for the command is : "@var{frequency}"
3585 Change lowpass width_type.
3586 Syntax for the command is : "@var{width_type}"
3589 Change lowpass width.
3590 Syntax for the command is : "@var{width}"
3595 Load a LV2 (LADSPA Version 2) plugin.
3597 To enable compilation of this filter you need to configure FFmpeg with
3598 @code{--enable-lv2}.
3602 Specifies the plugin URI. You may need to escape ':'.
3605 Set the '|' separated list of controls which are zero or more floating point
3606 values that determine the behavior of the loaded plugin (for example delay,
3608 If @option{controls} is set to @code{help}, all available controls and
3609 their valid ranges are printed.
3611 @item sample_rate, s
3612 Specify the sample rate, default to 44100. Only used if plugin have
3616 Set the number of samples per channel per each output frame, default
3617 is 1024. Only used if plugin have zero inputs.
3620 Set the minimum duration of the sourced audio. See
3621 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3622 for the accepted syntax.
3623 Note that the resulting duration may be greater than the specified duration,
3624 as the generated audio is always cut at the end of a complete frame.
3625 If not specified, or the expressed duration is negative, the audio is
3626 supposed to be generated forever.
3627 Only used if plugin have zero inputs.
3630 @subsection Examples
3634 Apply bass enhancer plugin from Calf:
3636 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3640 Apply vinyl plugin from Calf:
3642 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3646 Apply bit crusher plugin from ArtyFX:
3648 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3653 Multiband Compress or expand the audio's dynamic range.
3655 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3656 This is akin to the crossover of a loudspeaker, and results in flat frequency
3657 response when absent compander action.
3659 It accepts the following parameters:
3663 This option syntax is:
3664 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3665 For explanation of each item refer to compand filter documentation.
3671 Mix channels with specific gain levels. The filter accepts the output
3672 channel layout followed by a set of channels definitions.
3674 This filter is also designed to efficiently remap the channels of an audio
3677 The filter accepts parameters of the form:
3678 "@var{l}|@var{outdef}|@var{outdef}|..."
3682 output channel layout or number of channels
3685 output channel specification, of the form:
3686 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3689 output channel to define, either a channel name (FL, FR, etc.) or a channel
3690 number (c0, c1, etc.)
3693 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3696 input channel to use, see out_name for details; it is not possible to mix
3697 named and numbered input channels
3700 If the `=' in a channel specification is replaced by `<', then the gains for
3701 that specification will be renormalized so that the total is 1, thus
3702 avoiding clipping noise.
3704 @subsection Mixing examples
3706 For example, if you want to down-mix from stereo to mono, but with a bigger
3707 factor for the left channel:
3709 pan=1c|c0=0.9*c0+0.1*c1
3712 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3713 7-channels surround:
3715 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3718 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3719 that should be preferred (see "-ac" option) unless you have very specific
3722 @subsection Remapping examples
3724 The channel remapping will be effective if, and only if:
3727 @item gain coefficients are zeroes or ones,
3728 @item only one input per channel output,
3731 If all these conditions are satisfied, the filter will notify the user ("Pure
3732 channel mapping detected"), and use an optimized and lossless method to do the
3735 For example, if you have a 5.1 source and want a stereo audio stream by
3736 dropping the extra channels:
3738 pan="stereo| c0=FL | c1=FR"
3741 Given the same source, you can also switch front left and front right channels
3742 and keep the input channel layout:
3744 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3747 If the input is a stereo audio stream, you can mute the front left channel (and
3748 still keep the stereo channel layout) with:
3753 Still with a stereo audio stream input, you can copy the right channel in both
3754 front left and right:
3756 pan="stereo| c0=FR | c1=FR"
3761 ReplayGain scanner filter. This filter takes an audio stream as an input and
3762 outputs it unchanged.
3763 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3767 Convert the audio sample format, sample rate and channel layout. It is
3768 not meant to be used directly.
3771 Apply time-stretching and pitch-shifting with librubberband.
3773 The filter accepts the following options:
3777 Set tempo scale factor.
3780 Set pitch scale factor.
3783 Set transients detector.
3784 Possible values are:
3793 Possible values are:
3802 Possible values are:
3809 Set processing window size.
3810 Possible values are:
3819 Possible values are:
3826 Enable formant preservation when shift pitching.
3827 Possible values are:
3835 Possible values are:
3844 Possible values are:
3851 @section sidechaincompress
3853 This filter acts like normal compressor but has the ability to compress
3854 detected signal using second input signal.
3855 It needs two input streams and returns one output stream.
3856 First input stream will be processed depending on second stream signal.
3857 The filtered signal then can be filtered with other filters in later stages of
3858 processing. See @ref{pan} and @ref{amerge} filter.
3860 The filter accepts the following options:
3864 Set input gain. Default is 1. Range is between 0.015625 and 64.
3867 If a signal of second stream raises above this level it will affect the gain
3868 reduction of first stream.
3869 By default is 0.125. Range is between 0.00097563 and 1.
3872 Set a ratio about which the signal is reduced. 1:2 means that if the level
3873 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3874 Default is 2. Range is between 1 and 20.
3877 Amount of milliseconds the signal has to rise above the threshold before gain
3878 reduction starts. Default is 20. Range is between 0.01 and 2000.
3881 Amount of milliseconds the signal has to fall below the threshold before
3882 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3885 Set the amount by how much signal will be amplified after processing.
3886 Default is 1. Range is from 1 to 64.
3889 Curve the sharp knee around the threshold to enter gain reduction more softly.
3890 Default is 2.82843. Range is between 1 and 8.
3893 Choose if the @code{average} level between all channels of side-chain stream
3894 or the louder(@code{maximum}) channel of side-chain stream affects the
3895 reduction. Default is @code{average}.
3898 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3899 of @code{rms}. Default is @code{rms} which is mainly smoother.
3902 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3905 How much to use compressed signal in output. Default is 1.
3906 Range is between 0 and 1.
3909 @subsection Examples
3913 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3914 depending on the signal of 2nd input and later compressed signal to be
3915 merged with 2nd input:
3917 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3921 @section sidechaingate
3923 A sidechain gate acts like a normal (wideband) gate but has the ability to
3924 filter the detected signal before sending it to the gain reduction stage.
3925 Normally a gate uses the full range signal to detect a level above the
3927 For example: If you cut all lower frequencies from your sidechain signal
3928 the gate will decrease the volume of your track only if not enough highs
3929 appear. With this technique you are able to reduce the resonation of a
3930 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3932 It needs two input streams and returns one output stream.
3933 First input stream will be processed depending on second stream signal.
3935 The filter accepts the following options:
3939 Set input level before filtering.
3940 Default is 1. Allowed range is from 0.015625 to 64.
3943 Set the level of gain reduction when the signal is below the threshold.
3944 Default is 0.06125. Allowed range is from 0 to 1.
3947 If a signal rises above this level the gain reduction is released.
3948 Default is 0.125. Allowed range is from 0 to 1.
3951 Set a ratio about which the signal is reduced.
3952 Default is 2. Allowed range is from 1 to 9000.
3955 Amount of milliseconds the signal has to rise above the threshold before gain
3957 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3960 Amount of milliseconds the signal has to fall below the threshold before the
3961 reduction is increased again. Default is 250 milliseconds.
3962 Allowed range is from 0.01 to 9000.
3965 Set amount of amplification of signal after processing.
3966 Default is 1. Allowed range is from 1 to 64.
3969 Curve the sharp knee around the threshold to enter gain reduction more softly.
3970 Default is 2.828427125. Allowed range is from 1 to 8.
3973 Choose if exact signal should be taken for detection or an RMS like one.
3974 Default is rms. Can be peak or rms.
3977 Choose if the average level between all channels or the louder channel affects
3979 Default is average. Can be average or maximum.
3982 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3985 @section silencedetect
3987 Detect silence in an audio stream.
3989 This filter logs a message when it detects that the input audio volume is less
3990 or equal to a noise tolerance value for a duration greater or equal to the
3991 minimum detected noise duration.
3993 The printed times and duration are expressed in seconds.
3995 The filter accepts the following options:
3999 Set silence duration until notification (default is 2 seconds).
4002 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4003 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4006 @subsection Examples
4010 Detect 5 seconds of silence with -50dB noise tolerance:
4012 silencedetect=n=-50dB:d=5
4016 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4017 tolerance in @file{silence.mp3}:
4019 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4023 @section silenceremove
4025 Remove silence from the beginning, middle or end of the audio.
4027 The filter accepts the following options:
4031 This value is used to indicate if audio should be trimmed at beginning of
4032 the audio. A value of zero indicates no silence should be trimmed from the
4033 beginning. When specifying a non-zero value, it trims audio up until it
4034 finds non-silence. Normally, when trimming silence from beginning of audio
4035 the @var{start_periods} will be @code{1} but it can be increased to higher
4036 values to trim all audio up to specific count of non-silence periods.
4037 Default value is @code{0}.
4039 @item start_duration
4040 Specify the amount of time that non-silence must be detected before it stops
4041 trimming audio. By increasing the duration, bursts of noises can be treated
4042 as silence and trimmed off. Default value is @code{0}.
4044 @item start_threshold
4045 This indicates what sample value should be treated as silence. For digital
4046 audio, a value of @code{0} may be fine but for audio recorded from analog,
4047 you may wish to increase the value to account for background noise.
4048 Can be specified in dB (in case "dB" is appended to the specified value)
4049 or amplitude ratio. Default value is @code{0}.
4052 Set the count for trimming silence from the end of audio.
4053 To remove silence from the middle of a file, specify a @var{stop_periods}
4054 that is negative. This value is then treated as a positive value and is
4055 used to indicate the effect should restart processing as specified by
4056 @var{start_periods}, making it suitable for removing periods of silence
4057 in the middle of the audio.
4058 Default value is @code{0}.
4061 Specify a duration of silence that must exist before audio is not copied any
4062 more. By specifying a higher duration, silence that is wanted can be left in
4064 Default value is @code{0}.
4066 @item stop_threshold
4067 This is the same as @option{start_threshold} but for trimming silence from
4069 Can be specified in dB (in case "dB" is appended to the specified value)
4070 or amplitude ratio. Default value is @code{0}.
4073 This indicates that @var{stop_duration} length of audio should be left intact
4074 at the beginning of each period of silence.
4075 For example, if you want to remove long pauses between words but do not want
4076 to remove the pauses completely. Default value is @code{0}.
4079 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4080 and works better with digital silence which is exactly 0.
4081 Default value is @code{rms}.
4084 Set ratio used to calculate size of window for detecting silence.
4085 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4088 @subsection Examples
4092 The following example shows how this filter can be used to start a recording
4093 that does not contain the delay at the start which usually occurs between
4094 pressing the record button and the start of the performance:
4096 silenceremove=1:5:0.02
4100 Trim all silence encountered from beginning to end where there is more than 1
4101 second of silence in audio:
4103 silenceremove=0:0:0:-1:1:-90dB
4109 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4110 loudspeakers around the user for binaural listening via headphones (audio
4111 formats up to 9 channels supported).
4112 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4113 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4114 Austrian Academy of Sciences.
4116 To enable compilation of this filter you need to configure FFmpeg with
4117 @code{--enable-libmysofa}.
4119 The filter accepts the following options:
4123 Set the SOFA file used for rendering.
4126 Set gain applied to audio. Value is in dB. Default is 0.
4129 Set rotation of virtual loudspeakers in deg. Default is 0.
4132 Set elevation of virtual speakers in deg. Default is 0.
4135 Set distance in meters between loudspeakers and the listener with near-field
4136 HRTFs. Default is 1.
4139 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4140 processing audio in time domain which is slow.
4141 @var{freq} is processing audio in frequency domain which is fast.
4142 Default is @var{freq}.
4145 Set custom positions of virtual loudspeakers. Syntax for this option is:
4146 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4147 Each virtual loudspeaker is described with short channel name following with
4148 azimuth and elevation in degrees.
4149 Each virtual loudspeaker description is separated by '|'.
4150 For example to override front left and front right channel positions use:
4151 'speakers=FL 45 15|FR 345 15'.
4152 Descriptions with unrecognised channel names are ignored.
4155 Set custom gain for LFE channels. Value is in dB. Default is 0.
4158 @subsection Examples
4162 Using ClubFritz6 sofa file:
4164 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4168 Using ClubFritz12 sofa file and bigger radius with small rotation:
4170 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4174 Similar as above but with custom speaker positions for front left, front right, back left and back right
4175 and also with custom gain:
4177 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4181 @section stereotools
4183 This filter has some handy utilities to manage stereo signals, for converting
4184 M/S stereo recordings to L/R signal while having control over the parameters
4185 or spreading the stereo image of master track.
4187 The filter accepts the following options:
4191 Set input level before filtering for both channels. Defaults is 1.
4192 Allowed range is from 0.015625 to 64.
4195 Set output level after filtering for both channels. Defaults is 1.
4196 Allowed range is from 0.015625 to 64.
4199 Set input balance between both channels. Default is 0.
4200 Allowed range is from -1 to 1.
4203 Set output balance between both channels. Default is 0.
4204 Allowed range is from -1 to 1.
4207 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4208 clipping. Disabled by default.
4211 Mute the left channel. Disabled by default.
4214 Mute the right channel. Disabled by default.
4217 Change the phase of the left channel. Disabled by default.
4220 Change the phase of the right channel. Disabled by default.
4223 Set stereo mode. Available values are:
4227 Left/Right to Left/Right, this is default.
4230 Left/Right to Mid/Side.
4233 Mid/Side to Left/Right.
4236 Left/Right to Left/Left.
4239 Left/Right to Right/Right.
4242 Left/Right to Left + Right.
4245 Left/Right to Right/Left.
4248 Mid/Side to Left/Left.
4251 Mid/Side to Right/Right.
4255 Set level of side signal. Default is 1.
4256 Allowed range is from 0.015625 to 64.
4259 Set balance of side signal. Default is 0.
4260 Allowed range is from -1 to 1.
4263 Set level of the middle signal. Default is 1.
4264 Allowed range is from 0.015625 to 64.
4267 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4270 Set stereo base between mono and inversed channels. Default is 0.
4271 Allowed range is from -1 to 1.
4274 Set delay in milliseconds how much to delay left from right channel and
4275 vice versa. Default is 0. Allowed range is from -20 to 20.
4278 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4281 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4283 @item bmode_in, bmode_out
4284 Set balance mode for balance_in/balance_out option.
4286 Can be one of the following:
4290 Classic balance mode. Attenuate one channel at time.
4291 Gain is raised up to 1.
4294 Similar as classic mode above but gain is raised up to 2.
4297 Equal power distribution, from -6dB to +6dB range.
4301 @subsection Examples
4305 Apply karaoke like effect:
4307 stereotools=mlev=0.015625
4311 Convert M/S signal to L/R:
4313 "stereotools=mode=ms>lr"
4317 @section stereowiden
4319 This filter enhance the stereo effect by suppressing signal common to both
4320 channels and by delaying the signal of left into right and vice versa,
4321 thereby widening the stereo effect.
4323 The filter accepts the following options:
4327 Time in milliseconds of the delay of left signal into right and vice versa.
4328 Default is 20 milliseconds.
4331 Amount of gain in delayed signal into right and vice versa. Gives a delay
4332 effect of left signal in right output and vice versa which gives widening
4333 effect. Default is 0.3.
4336 Cross feed of left into right with inverted phase. This helps in suppressing
4337 the mono. If the value is 1 it will cancel all the signal common to both
4338 channels. Default is 0.3.
4341 Set level of input signal of original channel. Default is 0.8.
4344 @section superequalizer
4345 Apply 18 band equalizer.
4347 The filter accepts the following options:
4354 Set 131Hz band gain.
4356 Set 185Hz band gain.
4358 Set 262Hz band gain.
4360 Set 370Hz band gain.
4362 Set 523Hz band gain.
4364 Set 740Hz band gain.
4366 Set 1047Hz band gain.
4368 Set 1480Hz band gain.
4370 Set 2093Hz band gain.
4372 Set 2960Hz band gain.
4374 Set 4186Hz band gain.
4376 Set 5920Hz band gain.
4378 Set 8372Hz band gain.
4380 Set 11840Hz band gain.
4382 Set 16744Hz band gain.
4384 Set 20000Hz band gain.
4388 Apply audio surround upmix filter.
4390 This filter allows to produce multichannel output from audio stream.
4392 The filter accepts the following options:
4396 Set output channel layout. By default, this is @var{5.1}.
4398 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4399 for the required syntax.
4402 Set input channel layout. By default, this is @var{stereo}.
4404 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4405 for the required syntax.
4408 Set input volume level. By default, this is @var{1}.
4411 Set output volume level. By default, this is @var{1}.
4414 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4417 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4420 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4423 Set front center input volume. By default, this is @var{1}.
4426 Set front center output volume. By default, this is @var{1}.
4429 Set LFE input volume. By default, this is @var{1}.
4432 Set LFE output volume. By default, this is @var{1}.
4435 @section treble, highshelf
4437 Boost or cut treble (upper) frequencies of the audio using a two-pole
4438 shelving filter with a response similar to that of a standard
4439 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4441 The filter accepts the following options:
4445 Give the gain at whichever is the lower of ~22 kHz and the
4446 Nyquist frequency. Its useful range is about -20 (for a large cut)
4447 to +20 (for a large boost). Beware of clipping when using a positive gain.
4450 Set the filter's central frequency and so can be used
4451 to extend or reduce the frequency range to be boosted or cut.
4452 The default value is @code{3000} Hz.
4455 Set method to specify band-width of filter.
4470 Determine how steep is the filter's shelf transition.
4473 Specify which channels to filter, by default all available are filtered.
4476 @subsection Commands
4478 This filter supports the following commands:
4481 Change treble frequency.
4482 Syntax for the command is : "@var{frequency}"
4485 Change treble width_type.
4486 Syntax for the command is : "@var{width_type}"
4489 Change treble width.
4490 Syntax for the command is : "@var{width}"
4494 Syntax for the command is : "@var{gain}"
4499 Sinusoidal amplitude modulation.
4501 The filter accepts the following options:
4505 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4506 (20 Hz or lower) will result in a tremolo effect.
4507 This filter may also be used as a ring modulator by specifying
4508 a modulation frequency higher than 20 Hz.
4509 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4512 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4513 Default value is 0.5.
4518 Sinusoidal phase modulation.
4520 The filter accepts the following options:
4524 Modulation frequency in Hertz.
4525 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4528 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4529 Default value is 0.5.
4534 Adjust the input audio volume.
4536 It accepts the following parameters:
4540 Set audio volume expression.
4542 Output values are clipped to the maximum value.
4544 The output audio volume is given by the relation:
4546 @var{output_volume} = @var{volume} * @var{input_volume}
4549 The default value for @var{volume} is "1.0".
4552 This parameter represents the mathematical precision.
4554 It determines which input sample formats will be allowed, which affects the
4555 precision of the volume scaling.
4559 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4561 32-bit floating-point; this limits input sample format to FLT. (default)
4563 64-bit floating-point; this limits input sample format to DBL.
4567 Choose the behaviour on encountering ReplayGain side data in input frames.
4571 Remove ReplayGain side data, ignoring its contents (the default).
4574 Ignore ReplayGain side data, but leave it in the frame.
4577 Prefer the track gain, if present.
4580 Prefer the album gain, if present.
4583 @item replaygain_preamp
4584 Pre-amplification gain in dB to apply to the selected replaygain gain.
4586 Default value for @var{replaygain_preamp} is 0.0.
4589 Set when the volume expression is evaluated.
4591 It accepts the following values:
4594 only evaluate expression once during the filter initialization, or
4595 when the @samp{volume} command is sent
4598 evaluate expression for each incoming frame
4601 Default value is @samp{once}.
4604 The volume expression can contain the following parameters.
4608 frame number (starting at zero)
4611 @item nb_consumed_samples
4612 number of samples consumed by the filter
4614 number of samples in the current frame
4616 original frame position in the file
4622 PTS at start of stream
4624 time at start of stream
4630 last set volume value
4633 Note that when @option{eval} is set to @samp{once} only the
4634 @var{sample_rate} and @var{tb} variables are available, all other
4635 variables will evaluate to NAN.
4637 @subsection Commands
4639 This filter supports the following commands:
4642 Modify the volume expression.
4643 The command accepts the same syntax of the corresponding option.
4645 If the specified expression is not valid, it is kept at its current
4647 @item replaygain_noclip
4648 Prevent clipping by limiting the gain applied.
4650 Default value for @var{replaygain_noclip} is 1.
4654 @subsection Examples
4658 Halve the input audio volume:
4662 volume=volume=-6.0206dB
4665 In all the above example the named key for @option{volume} can be
4666 omitted, for example like in:
4672 Increase input audio power by 6 decibels using fixed-point precision:
4674 volume=volume=6dB:precision=fixed
4678 Fade volume after time 10 with an annihilation period of 5 seconds:
4680 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
4684 @section volumedetect
4686 Detect the volume of the input video.
4688 The filter has no parameters. The input is not modified. Statistics about
4689 the volume will be printed in the log when the input stream end is reached.
4691 In particular it will show the mean volume (root mean square), maximum
4692 volume (on a per-sample basis), and the beginning of a histogram of the
4693 registered volume values (from the maximum value to a cumulated 1/1000 of
4696 All volumes are in decibels relative to the maximum PCM value.
4698 @subsection Examples
4700 Here is an excerpt of the output:
4702 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
4703 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
4704 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
4705 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
4706 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
4707 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
4708 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
4709 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
4710 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
4716 The mean square energy is approximately -27 dB, or 10^-2.7.
4718 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
4720 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
4723 In other words, raising the volume by +4 dB does not cause any clipping,
4724 raising it by +5 dB causes clipping for 6 samples, etc.
4726 @c man end AUDIO FILTERS
4728 @chapter Audio Sources
4729 @c man begin AUDIO SOURCES
4731 Below is a description of the currently available audio sources.
4735 Buffer audio frames, and make them available to the filter chain.
4737 This source is mainly intended for a programmatic use, in particular
4738 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
4740 It accepts the following parameters:
4744 The timebase which will be used for timestamps of submitted frames. It must be
4745 either a floating-point number or in @var{numerator}/@var{denominator} form.
4748 The sample rate of the incoming audio buffers.
4751 The sample format of the incoming audio buffers.
4752 Either a sample format name or its corresponding integer representation from
4753 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4755 @item channel_layout
4756 The channel layout of the incoming audio buffers.
4757 Either a channel layout name from channel_layout_map in
4758 @file{libavutil/channel_layout.c} or its corresponding integer representation
4759 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4762 The number of channels of the incoming audio buffers.
4763 If both @var{channels} and @var{channel_layout} are specified, then they
4768 @subsection Examples
4771 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4774 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4775 Since the sample format with name "s16p" corresponds to the number
4776 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4779 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4784 Generate an audio signal specified by an expression.
4786 This source accepts in input one or more expressions (one for each
4787 channel), which are evaluated and used to generate a corresponding
4790 This source accepts the following options:
4794 Set the '|'-separated expressions list for each separate channel. In case the
4795 @option{channel_layout} option is not specified, the selected channel layout
4796 depends on the number of provided expressions. Otherwise the last
4797 specified expression is applied to the remaining output channels.
4799 @item channel_layout, c
4800 Set the channel layout. The number of channels in the specified layout
4801 must be equal to the number of specified expressions.
4804 Set the minimum duration of the sourced audio. See
4805 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4806 for the accepted syntax.
4807 Note that the resulting duration may be greater than the specified
4808 duration, as the generated audio is always cut at the end of a
4811 If not specified, or the expressed duration is negative, the audio is
4812 supposed to be generated forever.
4815 Set the number of samples per channel per each output frame,
4818 @item sample_rate, s
4819 Specify the sample rate, default to 44100.
4822 Each expression in @var{exprs} can contain the following constants:
4826 number of the evaluated sample, starting from 0
4829 time of the evaluated sample expressed in seconds, starting from 0
4836 @subsection Examples
4846 Generate a sin signal with frequency of 440 Hz, set sample rate to
4849 aevalsrc="sin(440*2*PI*t):s=8000"
4853 Generate a two channels signal, specify the channel layout (Front
4854 Center + Back Center) explicitly:
4856 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4860 Generate white noise:
4862 aevalsrc="-2+random(0)"
4866 Generate an amplitude modulated signal:
4868 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4872 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4874 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4881 The null audio source, return unprocessed audio frames. It is mainly useful
4882 as a template and to be employed in analysis / debugging tools, or as
4883 the source for filters which ignore the input data (for example the sox
4886 This source accepts the following options:
4890 @item channel_layout, cl
4892 Specifies the channel layout, and can be either an integer or a string
4893 representing a channel layout. The default value of @var{channel_layout}
4896 Check the channel_layout_map definition in
4897 @file{libavutil/channel_layout.c} for the mapping between strings and
4898 channel layout values.
4900 @item sample_rate, r
4901 Specifies the sample rate, and defaults to 44100.
4904 Set the number of samples per requested frames.
4908 @subsection Examples
4912 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4914 anullsrc=r=48000:cl=4
4918 Do the same operation with a more obvious syntax:
4920 anullsrc=r=48000:cl=mono
4924 All the parameters need to be explicitly defined.
4928 Synthesize a voice utterance using the libflite library.
4930 To enable compilation of this filter you need to configure FFmpeg with
4931 @code{--enable-libflite}.
4933 Note that versions of the flite library prior to 2.0 are not thread-safe.
4935 The filter accepts the following options:
4940 If set to 1, list the names of the available voices and exit
4941 immediately. Default value is 0.
4944 Set the maximum number of samples per frame. Default value is 512.
4947 Set the filename containing the text to speak.
4950 Set the text to speak.
4953 Set the voice to use for the speech synthesis. Default value is
4954 @code{kal}. See also the @var{list_voices} option.
4957 @subsection Examples
4961 Read from file @file{speech.txt}, and synthesize the text using the
4962 standard flite voice:
4964 flite=textfile=speech.txt
4968 Read the specified text selecting the @code{slt} voice:
4970 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4974 Input text to ffmpeg:
4976 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4980 Make @file{ffplay} speak the specified text, using @code{flite} and
4981 the @code{lavfi} device:
4983 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4987 For more information about libflite, check:
4988 @url{http://www.festvox.org/flite/}
4992 Generate a noise audio signal.
4994 The filter accepts the following options:
4997 @item sample_rate, r
4998 Specify the sample rate. Default value is 48000 Hz.
5001 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5005 Specify the duration of the generated audio stream. Not specifying this option
5006 results in noise with an infinite length.
5008 @item color, colour, c
5009 Specify the color of noise. Available noise colors are white, pink, brown,
5010 blue and violet. Default color is white.
5013 Specify a value used to seed the PRNG.
5016 Set the number of samples per each output frame, default is 1024.
5019 @subsection Examples
5024 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5026 anoisesrc=d=60:c=pink:r=44100:a=0.5
5032 Generate odd-tap Hilbert transform FIR coefficients.
5034 The resulting stream can be used with @ref{afir} filter for phase-shifting
5035 the signal by 90 degrees.
5037 This is used in many matrix coding schemes and for analytic signal generation.
5038 The process is often written as a multiplication by i (or j), the imaginary unit.
5040 The filter accepts the following options:
5044 @item sample_rate, s
5045 Set sample rate, default is 44100.
5048 Set length of FIR filter, default is 22051.
5051 Set number of samples per each frame.
5054 Set window function to be used when generating FIR coefficients.
5059 Generate an audio signal made of a sine wave with amplitude 1/8.
5061 The audio signal is bit-exact.
5063 The filter accepts the following options:
5068 Set the carrier frequency. Default is 440 Hz.
5070 @item beep_factor, b
5071 Enable a periodic beep every second with frequency @var{beep_factor} times
5072 the carrier frequency. Default is 0, meaning the beep is disabled.
5074 @item sample_rate, r
5075 Specify the sample rate, default is 44100.
5078 Specify the duration of the generated audio stream.
5080 @item samples_per_frame
5081 Set the number of samples per output frame.
5083 The expression can contain the following constants:
5087 The (sequential) number of the output audio frame, starting from 0.
5090 The PTS (Presentation TimeStamp) of the output audio frame,
5091 expressed in @var{TB} units.
5094 The PTS of the output audio frame, expressed in seconds.
5097 The timebase of the output audio frames.
5100 Default is @code{1024}.
5103 @subsection Examples
5108 Generate a simple 440 Hz sine wave:
5114 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5118 sine=frequency=220:beep_factor=4:duration=5
5122 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5125 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5129 @c man end AUDIO SOURCES
5131 @chapter Audio Sinks
5132 @c man begin AUDIO SINKS
5134 Below is a description of the currently available audio sinks.
5136 @section abuffersink
5138 Buffer audio frames, and make them available to the end of filter chain.
5140 This sink is mainly intended for programmatic use, in particular
5141 through the interface defined in @file{libavfilter/buffersink.h}
5142 or the options system.
5144 It accepts a pointer to an AVABufferSinkContext structure, which
5145 defines the incoming buffers' formats, to be passed as the opaque
5146 parameter to @code{avfilter_init_filter} for initialization.
5149 Null audio sink; do absolutely nothing with the input audio. It is
5150 mainly useful as a template and for use in analysis / debugging
5153 @c man end AUDIO SINKS
5155 @chapter Video Filters
5156 @c man begin VIDEO FILTERS
5158 When you configure your FFmpeg build, you can disable any of the
5159 existing filters using @code{--disable-filters}.
5160 The configure output will show the video filters included in your
5163 Below is a description of the currently available video filters.
5165 @section alphaextract
5167 Extract the alpha component from the input as a grayscale video. This
5168 is especially useful with the @var{alphamerge} filter.
5172 Add or replace the alpha component of the primary input with the
5173 grayscale value of a second input. This is intended for use with
5174 @var{alphaextract} to allow the transmission or storage of frame
5175 sequences that have alpha in a format that doesn't support an alpha
5178 For example, to reconstruct full frames from a normal YUV-encoded video
5179 and a separate video created with @var{alphaextract}, you might use:
5181 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5184 Since this filter is designed for reconstruction, it operates on frame
5185 sequences without considering timestamps, and terminates when either
5186 input reaches end of stream. This will cause problems if your encoding
5187 pipeline drops frames. If you're trying to apply an image as an
5188 overlay to a video stream, consider the @var{overlay} filter instead.
5192 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5193 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5194 Substation Alpha) subtitles files.
5196 This filter accepts the following option in addition to the common options from
5197 the @ref{subtitles} filter:
5201 Set the shaping engine
5203 Available values are:
5206 The default libass shaping engine, which is the best available.
5208 Fast, font-agnostic shaper that can do only substitutions
5210 Slower shaper using OpenType for substitutions and positioning
5213 The default is @code{auto}.
5217 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5219 The filter accepts the following options:
5223 Set threshold A for 1st plane. Default is 0.02.
5224 Valid range is 0 to 0.3.
5227 Set threshold B for 1st plane. Default is 0.04.
5228 Valid range is 0 to 5.
5231 Set threshold A for 2nd plane. Default is 0.02.
5232 Valid range is 0 to 0.3.
5235 Set threshold B for 2nd plane. Default is 0.04.
5236 Valid range is 0 to 5.
5239 Set threshold A for 3rd plane. Default is 0.02.
5240 Valid range is 0 to 0.3.
5243 Set threshold B for 3rd plane. Default is 0.04.
5244 Valid range is 0 to 5.
5246 Threshold A is designed to react on abrupt changes in the input signal and
5247 threshold B is designed to react on continuous changes in the input signal.
5250 Set number of frames filter will use for averaging. Default is 9. Must be odd
5251 number in range [5, 129].
5254 Set what planes of frame filter will use for averaging. Default is all.
5259 Apply average blur filter.
5261 The filter accepts the following options:
5265 Set horizontal kernel size.
5268 Set which planes to filter. By default all planes are filtered.
5271 Set vertical kernel size, if zero it will be same as @code{sizeX}.
5272 Default is @code{0}.
5277 Compute the bounding box for the non-black pixels in the input frame
5280 This filter computes the bounding box containing all the pixels with a
5281 luminance value greater than the minimum allowed value.
5282 The parameters describing the bounding box are printed on the filter
5285 The filter accepts the following option:
5289 Set the minimal luminance value. Default is @code{16}.
5292 @section bitplanenoise
5294 Show and measure bit plane noise.
5296 The filter accepts the following options:
5300 Set which plane to analyze. Default is @code{1}.
5303 Filter out noisy pixels from @code{bitplane} set above.
5304 Default is disabled.
5307 @section blackdetect
5309 Detect video intervals that are (almost) completely black. Can be
5310 useful to detect chapter transitions, commercials, or invalid
5311 recordings. Output lines contains the time for the start, end and
5312 duration of the detected black interval expressed in seconds.
5314 In order to display the output lines, you need to set the loglevel at
5315 least to the AV_LOG_INFO value.
5317 The filter accepts the following options:
5320 @item black_min_duration, d
5321 Set the minimum detected black duration expressed in seconds. It must
5322 be a non-negative floating point number.
5324 Default value is 2.0.
5326 @item picture_black_ratio_th, pic_th
5327 Set the threshold for considering a picture "black".
5328 Express the minimum value for the ratio:
5330 @var{nb_black_pixels} / @var{nb_pixels}
5333 for which a picture is considered black.
5334 Default value is 0.98.
5336 @item pixel_black_th, pix_th
5337 Set the threshold for considering a pixel "black".
5339 The threshold expresses the maximum pixel luminance value for which a
5340 pixel is considered "black". The provided value is scaled according to
5341 the following equation:
5343 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5346 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5347 the input video format, the range is [0-255] for YUV full-range
5348 formats and [16-235] for YUV non full-range formats.
5350 Default value is 0.10.
5353 The following example sets the maximum pixel threshold to the minimum
5354 value, and detects only black intervals of 2 or more seconds:
5356 blackdetect=d=2:pix_th=0.00
5361 Detect frames that are (almost) completely black. Can be useful to
5362 detect chapter transitions or commercials. Output lines consist of
5363 the frame number of the detected frame, the percentage of blackness,
5364 the position in the file if known or -1 and the timestamp in seconds.
5366 In order to display the output lines, you need to set the loglevel at
5367 least to the AV_LOG_INFO value.
5369 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5370 The value represents the percentage of pixels in the picture that
5371 are below the threshold value.
5373 It accepts the following parameters:
5378 The percentage of the pixels that have to be below the threshold; it defaults to
5381 @item threshold, thresh
5382 The threshold below which a pixel value is considered black; it defaults to
5387 @section blend, tblend
5389 Blend two video frames into each other.
5391 The @code{blend} filter takes two input streams and outputs one
5392 stream, the first input is the "top" layer and second input is
5393 "bottom" layer. By default, the output terminates when the longest input terminates.
5395 The @code{tblend} (time blend) filter takes two consecutive frames
5396 from one single stream, and outputs the result obtained by blending
5397 the new frame on top of the old frame.
5399 A description of the accepted options follows.
5407 Set blend mode for specific pixel component or all pixel components in case
5408 of @var{all_mode}. Default value is @code{normal}.
5410 Available values for component modes are:
5452 Set blend opacity for specific pixel component or all pixel components in case
5453 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5460 Set blend expression for specific pixel component or all pixel components in case
5461 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5463 The expressions can use the following variables:
5467 The sequential number of the filtered frame, starting from @code{0}.
5471 the coordinates of the current sample
5475 the width and height of currently filtered plane
5479 Width and height scale depending on the currently filtered plane. It is the
5480 ratio between the corresponding luma plane number of pixels and the current
5481 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
5482 @code{0.5,0.5} for chroma planes.
5485 Time of the current frame, expressed in seconds.
5488 Value of pixel component at current location for first video frame (top layer).
5491 Value of pixel component at current location for second video frame (bottom layer).
5495 The @code{blend} filter also supports the @ref{framesync} options.
5497 @subsection Examples
5501 Apply transition from bottom layer to top layer in first 10 seconds:
5503 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5507 Apply linear horizontal transition from top layer to bottom layer:
5509 blend=all_expr='A*(X/W)+B*(1-X/W)'
5513 Apply 1x1 checkerboard effect:
5515 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5519 Apply uncover left effect:
5521 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5525 Apply uncover down effect:
5527 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5531 Apply uncover up-left effect:
5533 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5537 Split diagonally video and shows top and bottom layer on each side:
5539 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5543 Display differences between the current and the previous frame:
5545 tblend=all_mode=grainextract
5551 Apply a boxblur algorithm to the input video.
5553 It accepts the following parameters:
5557 @item luma_radius, lr
5558 @item luma_power, lp
5559 @item chroma_radius, cr
5560 @item chroma_power, cp
5561 @item alpha_radius, ar
5562 @item alpha_power, ap
5566 A description of the accepted options follows.
5569 @item luma_radius, lr
5570 @item chroma_radius, cr
5571 @item alpha_radius, ar
5572 Set an expression for the box radius in pixels used for blurring the
5573 corresponding input plane.
5575 The radius value must be a non-negative number, and must not be
5576 greater than the value of the expression @code{min(w,h)/2} for the
5577 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
5580 Default value for @option{luma_radius} is "2". If not specified,
5581 @option{chroma_radius} and @option{alpha_radius} default to the
5582 corresponding value set for @option{luma_radius}.
5584 The expressions can contain the following constants:
5588 The input width and height in pixels.
5592 The input chroma image width and height in pixels.
5596 The horizontal and vertical chroma subsample values. For example, for the
5597 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
5600 @item luma_power, lp
5601 @item chroma_power, cp
5602 @item alpha_power, ap
5603 Specify how many times the boxblur filter is applied to the
5604 corresponding plane.
5606 Default value for @option{luma_power} is 2. If not specified,
5607 @option{chroma_power} and @option{alpha_power} default to the
5608 corresponding value set for @option{luma_power}.
5610 A value of 0 will disable the effect.
5613 @subsection Examples
5617 Apply a boxblur filter with the luma, chroma, and alpha radii
5620 boxblur=luma_radius=2:luma_power=1
5625 Set the luma radius to 2, and alpha and chroma radius to 0:
5627 boxblur=2:1:cr=0:ar=0
5631 Set the luma and chroma radii to a fraction of the video dimension:
5633 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
5639 Deinterlace the input video ("bwdif" stands for "Bob Weaver
5640 Deinterlacing Filter").
5642 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
5643 interpolation algorithms.
5644 It accepts the following parameters:
5648 The interlacing mode to adopt. It accepts one of the following values:
5652 Output one frame for each frame.
5654 Output one frame for each field.
5657 The default value is @code{send_field}.
5660 The picture field parity assumed for the input interlaced video. It accepts one
5661 of the following values:
5665 Assume the top field is first.
5667 Assume the bottom field is first.
5669 Enable automatic detection of field parity.
5672 The default value is @code{auto}.
5673 If the interlacing is unknown or the decoder does not export this information,
5674 top field first will be assumed.
5677 Specify which frames to deinterlace. Accept one of the following
5682 Deinterlace all frames.
5684 Only deinterlace frames marked as interlaced.
5687 The default value is @code{all}.
5691 YUV colorspace color/chroma keying.
5693 The filter accepts the following options:
5697 The color which will be replaced with transparency.
5700 Similarity percentage with the key color.
5702 0.01 matches only the exact key color, while 1.0 matches everything.
5707 0.0 makes pixels either fully transparent, or not transparent at all.
5709 Higher values result in semi-transparent pixels, with a higher transparency
5710 the more similar the pixels color is to the key color.
5713 Signals that the color passed is already in YUV instead of RGB.
5715 Literal colors like "green" or "red" don't make sense with this enabled anymore.
5716 This can be used to pass exact YUV values as hexadecimal numbers.
5719 @subsection Examples
5723 Make every green pixel in the input image transparent:
5725 ffmpeg -i input.png -vf chromakey=green out.png
5729 Overlay a greenscreen-video on top of a static black background.
5731 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
5737 Display CIE color diagram with pixels overlaid onto it.
5739 The filter accepts the following options:
5754 @item uhdtv, rec2020
5767 Set what gamuts to draw.
5769 See @code{system} option for available values.
5772 Set ciescope size, by default set to 512.
5775 Set intensity used to map input pixel values to CIE diagram.
5778 Set contrast used to draw tongue colors that are out of active color system gamut.
5781 Correct gamma displayed on scope, by default enabled.
5784 Show white point on CIE diagram, by default disabled.
5787 Set input gamma. Used only with XYZ input color space.
5792 Visualize information exported by some codecs.
5794 Some codecs can export information through frames using side-data or other
5795 means. For example, some MPEG based codecs export motion vectors through the
5796 @var{export_mvs} flag in the codec @option{flags2} option.
5798 The filter accepts the following option:
5802 Set motion vectors to visualize.
5804 Available flags for @var{mv} are:
5808 forward predicted MVs of P-frames
5810 forward predicted MVs of B-frames
5812 backward predicted MVs of B-frames
5816 Display quantization parameters using the chroma planes.
5819 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5821 Available flags for @var{mv_type} are:
5825 forward predicted MVs
5827 backward predicted MVs
5830 @item frame_type, ft
5831 Set frame type to visualize motion vectors of.
5833 Available flags for @var{frame_type} are:
5837 intra-coded frames (I-frames)
5839 predicted frames (P-frames)
5841 bi-directionally predicted frames (B-frames)
5845 @subsection Examples
5849 Visualize forward predicted MVs of all frames using @command{ffplay}:
5851 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5855 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5857 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5861 @section colorbalance
5862 Modify intensity of primary colors (red, green and blue) of input frames.
5864 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5865 regions for the red-cyan, green-magenta or blue-yellow balance.
5867 A positive adjustment value shifts the balance towards the primary color, a negative
5868 value towards the complementary color.
5870 The filter accepts the following options:
5876 Adjust red, green and blue shadows (darkest pixels).
5881 Adjust red, green and blue midtones (medium pixels).
5886 Adjust red, green and blue highlights (brightest pixels).
5888 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5891 @subsection Examples
5895 Add red color cast to shadows:
5902 RGB colorspace color keying.
5904 The filter accepts the following options:
5908 The color which will be replaced with transparency.
5911 Similarity percentage with the key color.
5913 0.01 matches only the exact key color, while 1.0 matches everything.
5918 0.0 makes pixels either fully transparent, or not transparent at all.
5920 Higher values result in semi-transparent pixels, with a higher transparency
5921 the more similar the pixels color is to the key color.
5924 @subsection Examples
5928 Make every green pixel in the input image transparent:
5930 ffmpeg -i input.png -vf colorkey=green out.png
5934 Overlay a greenscreen-video on top of a static background image.
5936 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
5940 @section colorlevels
5942 Adjust video input frames using levels.
5944 The filter accepts the following options:
5951 Adjust red, green, blue and alpha input black point.
5952 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5958 Adjust red, green, blue and alpha input white point.
5959 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5961 Input levels are used to lighten highlights (bright tones), darken shadows
5962 (dark tones), change the balance of bright and dark tones.
5968 Adjust red, green, blue and alpha output black point.
5969 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5975 Adjust red, green, blue and alpha output white point.
5976 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5978 Output levels allows manual selection of a constrained output level range.
5981 @subsection Examples
5985 Make video output darker:
5987 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5993 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5997 Make video output lighter:
5999 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6003 Increase brightness:
6005 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6009 @section colorchannelmixer
6011 Adjust video input frames by re-mixing color channels.
6013 This filter modifies a color channel by adding the values associated to
6014 the other channels of the same pixels. For example if the value to
6015 modify is red, the output value will be:
6017 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6020 The filter accepts the following options:
6027 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6028 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6034 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6035 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6041 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6042 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6048 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6049 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6051 Allowed ranges for options are @code{[-2.0, 2.0]}.
6054 @subsection Examples
6058 Convert source to grayscale:
6060 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6063 Simulate sepia tones:
6065 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6069 @section colormatrix
6071 Convert color matrix.
6073 The filter accepts the following options:
6078 Specify the source and destination color matrix. Both values must be
6081 The accepted values are:
6109 For example to convert from BT.601 to SMPTE-240M, use the command:
6111 colormatrix=bt601:smpte240m
6116 Convert colorspace, transfer characteristics or color primaries.
6117 Input video needs to have an even size.
6119 The filter accepts the following options:
6124 Specify all color properties at once.
6126 The accepted values are:
6156 Specify output colorspace.
6158 The accepted values are:
6167 BT.470BG or BT.601-6 625
6170 SMPTE-170M or BT.601-6 525
6179 BT.2020 with non-constant luminance
6185 Specify output transfer characteristics.
6187 The accepted values are:
6199 Constant gamma of 2.2
6202 Constant gamma of 2.8
6205 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6223 BT.2020 for 10-bits content
6226 BT.2020 for 12-bits content
6232 Specify output color primaries.
6234 The accepted values are:
6243 BT.470BG or BT.601-6 625
6246 SMPTE-170M or BT.601-6 525
6270 Specify output color range.
6272 The accepted values are:
6275 TV (restricted) range
6278 MPEG (restricted) range
6289 Specify output color format.
6291 The accepted values are:
6294 YUV 4:2:0 planar 8-bits
6297 YUV 4:2:0 planar 10-bits
6300 YUV 4:2:0 planar 12-bits
6303 YUV 4:2:2 planar 8-bits
6306 YUV 4:2:2 planar 10-bits
6309 YUV 4:2:2 planar 12-bits
6312 YUV 4:4:4 planar 8-bits
6315 YUV 4:4:4 planar 10-bits
6318 YUV 4:4:4 planar 12-bits
6323 Do a fast conversion, which skips gamma/primary correction. This will take
6324 significantly less CPU, but will be mathematically incorrect. To get output
6325 compatible with that produced by the colormatrix filter, use fast=1.
6328 Specify dithering mode.
6330 The accepted values are:
6336 Floyd-Steinberg dithering
6340 Whitepoint adaptation mode.
6342 The accepted values are:
6345 Bradford whitepoint adaptation
6348 von Kries whitepoint adaptation
6351 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6355 Override all input properties at once. Same accepted values as @ref{all}.
6358 Override input colorspace. Same accepted values as @ref{space}.
6361 Override input color primaries. Same accepted values as @ref{primaries}.
6364 Override input transfer characteristics. Same accepted values as @ref{trc}.
6367 Override input color range. Same accepted values as @ref{range}.
6371 The filter converts the transfer characteristics, color space and color
6372 primaries to the specified user values. The output value, if not specified,
6373 is set to a default value based on the "all" property. If that property is
6374 also not specified, the filter will log an error. The output color range and
6375 format default to the same value as the input color range and format. The
6376 input transfer characteristics, color space, color primaries and color range
6377 should be set on the input data. If any of these are missing, the filter will
6378 log an error and no conversion will take place.
6380 For example to convert the input to SMPTE-240M, use the command:
6382 colorspace=smpte240m
6385 @section convolution
6387 Apply convolution 3x3, 5x5 or 7x7 filter.
6389 The filter accepts the following options:
6396 Set matrix for each plane.
6397 Matrix is sequence of 9, 25 or 49 signed integers.
6403 Set multiplier for calculated value for each plane.
6404 If unset or 0, it will be sum of all matrix elements.
6410 Set bias for each plane. This value is added to the result of the multiplication.
6411 Useful for making the overall image brighter or darker. Default is 0.0.
6414 @subsection Examples
6420 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"
6426 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"
6432 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"
6438 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"
6442 Apply laplacian edge detector which includes diagonals:
6444 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"
6450 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"
6456 Apply 2D convolution of video stream in frequency domain using second stream
6459 The filter accepts the following options:
6463 Set which planes to process.
6466 Set which impulse video frames will be processed, can be @var{first}
6467 or @var{all}. Default is @var{all}.
6470 The @code{convolve} filter also supports the @ref{framesync} options.
6474 Copy the input video source unchanged to the output. This is mainly useful for
6479 Video filtering on GPU using Apple's CoreImage API on OSX.
6481 Hardware acceleration is based on an OpenGL context. Usually, this means it is
6482 processed by video hardware. However, software-based OpenGL implementations
6483 exist which means there is no guarantee for hardware processing. It depends on
6486 There are many filters and image generators provided by Apple that come with a
6487 large variety of options. The filter has to be referenced by its name along
6490 The coreimage filter accepts the following options:
6493 List all available filters and generators along with all their respective
6494 options as well as possible minimum and maximum values along with the default
6501 Specify all filters by their respective name and options.
6502 Use @var{list_filters} to determine all valid filter names and options.
6503 Numerical options are specified by a float value and are automatically clamped
6504 to their respective value range. Vector and color options have to be specified
6505 by a list of space separated float values. Character escaping has to be done.
6506 A special option name @code{default} is available to use default options for a
6509 It is required to specify either @code{default} or at least one of the filter options.
6510 All omitted options are used with their default values.
6511 The syntax of the filter string is as follows:
6513 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
6517 Specify a rectangle where the output of the filter chain is copied into the
6518 input image. It is given by a list of space separated float values:
6520 output_rect=x\ y\ width\ height
6522 If not given, the output rectangle equals the dimensions of the input image.
6523 The output rectangle is automatically cropped at the borders of the input
6524 image. Negative values are valid for each component.
6526 output_rect=25\ 25\ 100\ 100
6530 Several filters can be chained for successive processing without GPU-HOST
6531 transfers allowing for fast processing of complex filter chains.
6532 Currently, only filters with zero (generators) or exactly one (filters) input
6533 image and one output image are supported. Also, transition filters are not yet
6536 Some filters generate output images with additional padding depending on the
6537 respective filter kernel. The padding is automatically removed to ensure the
6538 filter output has the same size as the input image.
6540 For image generators, the size of the output image is determined by the
6541 previous output image of the filter chain or the input image of the whole
6542 filterchain, respectively. The generators do not use the pixel information of
6543 this image to generate their output. However, the generated output is
6544 blended onto this image, resulting in partial or complete coverage of the
6547 The @ref{coreimagesrc} video source can be used for generating input images
6548 which are directly fed into the filter chain. By using it, providing input
6549 images by another video source or an input video is not required.
6551 @subsection Examples
6556 List all filters available:
6558 coreimage=list_filters=true
6562 Use the CIBoxBlur filter with default options to blur an image:
6564 coreimage=filter=CIBoxBlur@@default
6568 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
6569 its center at 100x100 and a radius of 50 pixels:
6571 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
6575 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
6576 given as complete and escaped command-line for Apple's standard bash shell:
6578 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
6584 Crop the input video to given dimensions.
6586 It accepts the following parameters:
6590 The width of the output video. It defaults to @code{iw}.
6591 This expression is evaluated only once during the filter
6592 configuration, or when the @samp{w} or @samp{out_w} command is sent.
6595 The height of the output video. It defaults to @code{ih}.
6596 This expression is evaluated only once during the filter
6597 configuration, or when the @samp{h} or @samp{out_h} command is sent.
6600 The horizontal position, in the input video, of the left edge of the output
6601 video. It defaults to @code{(in_w-out_w)/2}.
6602 This expression is evaluated per-frame.
6605 The vertical position, in the input video, of the top edge of the output video.
6606 It defaults to @code{(in_h-out_h)/2}.
6607 This expression is evaluated per-frame.
6610 If set to 1 will force the output display aspect ratio
6611 to be the same of the input, by changing the output sample aspect
6612 ratio. It defaults to 0.
6615 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
6616 width/height/x/y as specified and will not be rounded to nearest smaller value.
6620 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
6621 expressions containing the following constants:
6626 The computed values for @var{x} and @var{y}. They are evaluated for
6631 The input width and height.
6635 These are the same as @var{in_w} and @var{in_h}.
6639 The output (cropped) width and height.
6643 These are the same as @var{out_w} and @var{out_h}.
6646 same as @var{iw} / @var{ih}
6649 input sample aspect ratio
6652 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
6656 horizontal and vertical chroma subsample values. For example for the
6657 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6660 The number of the input frame, starting from 0.
6663 the position in the file of the input frame, NAN if unknown
6666 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
6670 The expression for @var{out_w} may depend on the value of @var{out_h},
6671 and the expression for @var{out_h} may depend on @var{out_w}, but they
6672 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
6673 evaluated after @var{out_w} and @var{out_h}.
6675 The @var{x} and @var{y} parameters specify the expressions for the
6676 position of the top-left corner of the output (non-cropped) area. They
6677 are evaluated for each frame. If the evaluated value is not valid, it
6678 is approximated to the nearest valid value.
6680 The expression for @var{x} may depend on @var{y}, and the expression
6681 for @var{y} may depend on @var{x}.
6683 @subsection Examples
6687 Crop area with size 100x100 at position (12,34).
6692 Using named options, the example above becomes:
6694 crop=w=100:h=100:x=12:y=34
6698 Crop the central input area with size 100x100:
6704 Crop the central input area with size 2/3 of the input video:
6706 crop=2/3*in_w:2/3*in_h
6710 Crop the input video central square:
6717 Delimit the rectangle with the top-left corner placed at position
6718 100:100 and the right-bottom corner corresponding to the right-bottom
6719 corner of the input image.
6721 crop=in_w-100:in_h-100:100:100
6725 Crop 10 pixels from the left and right borders, and 20 pixels from
6726 the top and bottom borders
6728 crop=in_w-2*10:in_h-2*20
6732 Keep only the bottom right quarter of the input image:
6734 crop=in_w/2:in_h/2:in_w/2:in_h/2
6738 Crop height for getting Greek harmony:
6740 crop=in_w:1/PHI*in_w
6744 Apply trembling effect:
6746 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)
6750 Apply erratic camera effect depending on timestamp:
6752 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)"
6756 Set x depending on the value of y:
6758 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
6762 @subsection Commands
6764 This filter supports the following commands:
6770 Set width/height of the output video and the horizontal/vertical position
6772 The command accepts the same syntax of the corresponding option.
6774 If the specified expression is not valid, it is kept at its current
6780 Auto-detect the crop size.
6782 It calculates the necessary cropping parameters and prints the
6783 recommended parameters via the logging system. The detected dimensions
6784 correspond to the non-black area of the input video.
6786 It accepts the following parameters:
6791 Set higher black value threshold, which can be optionally specified
6792 from nothing (0) to everything (255 for 8-bit based formats). An intensity
6793 value greater to the set value is considered non-black. It defaults to 24.
6794 You can also specify a value between 0.0 and 1.0 which will be scaled depending
6795 on the bitdepth of the pixel format.
6798 The value which the width/height should be divisible by. It defaults to
6799 16. The offset is automatically adjusted to center the video. Use 2 to
6800 get only even dimensions (needed for 4:2:2 video). 16 is best when
6801 encoding to most video codecs.
6803 @item reset_count, reset
6804 Set the counter that determines after how many frames cropdetect will
6805 reset the previously detected largest video area and start over to
6806 detect the current optimal crop area. Default value is 0.
6808 This can be useful when channel logos distort the video area. 0
6809 indicates 'never reset', and returns the largest area encountered during
6816 Apply color adjustments using curves.
6818 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6819 component (red, green and blue) has its values defined by @var{N} key points
6820 tied from each other using a smooth curve. The x-axis represents the pixel
6821 values from the input frame, and the y-axis the new pixel values to be set for
6824 By default, a component curve is defined by the two points @var{(0;0)} and
6825 @var{(1;1)}. This creates a straight line where each original pixel value is
6826 "adjusted" to its own value, which means no change to the image.
6828 The filter allows you to redefine these two points and add some more. A new
6829 curve (using a natural cubic spline interpolation) will be define to pass
6830 smoothly through all these new coordinates. The new defined points needs to be
6831 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6832 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6833 the vector spaces, the values will be clipped accordingly.
6835 The filter accepts the following options:
6839 Select one of the available color presets. This option can be used in addition
6840 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6841 options takes priority on the preset values.
6842 Available presets are:
6845 @item color_negative
6848 @item increase_contrast
6850 @item linear_contrast
6851 @item medium_contrast
6853 @item strong_contrast
6856 Default is @code{none}.
6858 Set the master key points. These points will define a second pass mapping. It
6859 is sometimes called a "luminance" or "value" mapping. It can be used with
6860 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6861 post-processing LUT.
6863 Set the key points for the red component.
6865 Set the key points for the green component.
6867 Set the key points for the blue component.
6869 Set the key points for all components (not including master).
6870 Can be used in addition to the other key points component
6871 options. In this case, the unset component(s) will fallback on this
6872 @option{all} setting.
6874 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6876 Save Gnuplot script of the curves in specified file.
6879 To avoid some filtergraph syntax conflicts, each key points list need to be
6880 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6882 @subsection Examples
6886 Increase slightly the middle level of blue:
6888 curves=blue='0/0 0.5/0.58 1/1'
6894 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'
6896 Here we obtain the following coordinates for each components:
6899 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6901 @code{(0;0) (0.50;0.48) (1;1)}
6903 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6907 The previous example can also be achieved with the associated built-in preset:
6909 curves=preset=vintage
6919 Use a Photoshop preset and redefine the points of the green component:
6921 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6925 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6926 and @command{gnuplot}:
6928 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6929 gnuplot -p /tmp/curves.plt
6935 Video data analysis filter.
6937 This filter shows hexadecimal pixel values of part of video.
6939 The filter accepts the following options:
6943 Set output video size.
6946 Set x offset from where to pick pixels.
6949 Set y offset from where to pick pixels.
6952 Set scope mode, can be one of the following:
6955 Draw hexadecimal pixel values with white color on black background.
6958 Draw hexadecimal pixel values with input video pixel color on black
6962 Draw hexadecimal pixel values on color background picked from input video,
6963 the text color is picked in such way so its always visible.
6967 Draw rows and columns numbers on left and top of video.
6970 Set background opacity.
6975 Denoise frames using 2D DCT (frequency domain filtering).
6977 This filter is not designed for real time.
6979 The filter accepts the following options:
6983 Set the noise sigma constant.
6985 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6986 coefficient (absolute value) below this threshold with be dropped.
6988 If you need a more advanced filtering, see @option{expr}.
6990 Default is @code{0}.
6993 Set number overlapping pixels for each block. Since the filter can be slow, you
6994 may want to reduce this value, at the cost of a less effective filter and the
6995 risk of various artefacts.
6997 If the overlapping value doesn't permit processing the whole input width or
6998 height, a warning will be displayed and according borders won't be denoised.
7000 Default value is @var{blocksize}-1, which is the best possible setting.
7003 Set the coefficient factor expression.
7005 For each coefficient of a DCT block, this expression will be evaluated as a
7006 multiplier value for the coefficient.
7008 If this is option is set, the @option{sigma} option will be ignored.
7010 The absolute value of the coefficient can be accessed through the @var{c}
7014 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7015 @var{blocksize}, which is the width and height of the processed blocks.
7017 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7018 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7019 on the speed processing. Also, a larger block size does not necessarily means a
7023 @subsection Examples
7025 Apply a denoise with a @option{sigma} of @code{4.5}:
7030 The same operation can be achieved using the expression system:
7032 dctdnoiz=e='gte(c, 4.5*3)'
7035 Violent denoise using a block size of @code{16x16}:
7042 Remove banding artifacts from input video.
7043 It works by replacing banded pixels with average value of referenced pixels.
7045 The filter accepts the following options:
7052 Set banding detection threshold for each plane. Default is 0.02.
7053 Valid range is 0.00003 to 0.5.
7054 If difference between current pixel and reference pixel is less than threshold,
7055 it will be considered as banded.
7058 Banding detection range in pixels. Default is 16. If positive, random number
7059 in range 0 to set value will be used. If negative, exact absolute value
7061 The range defines square of four pixels around current pixel.
7064 Set direction in radians from which four pixel will be compared. If positive,
7065 random direction from 0 to set direction will be picked. If negative, exact of
7066 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7067 will pick only pixels on same row and -PI/2 will pick only pixels on same
7071 If enabled, current pixel is compared with average value of all four
7072 surrounding pixels. The default is enabled. If disabled current pixel is
7073 compared with all four surrounding pixels. The pixel is considered banded
7074 if only all four differences with surrounding pixels are less than threshold.
7077 If enabled, current pixel is changed if and only if all pixel components are banded,
7078 e.g. banding detection threshold is triggered for all color components.
7079 The default is disabled.
7084 Remove blocking artifacts from input video.
7086 The filter accepts the following options:
7090 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7091 This controls what kind of deblocking is applied.
7094 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7100 Set blocking detection thresholds. Allowed range is 0 to 1.
7101 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7102 Using higher threshold gives more deblocking strength.
7103 Setting @var{alpha} controls threshold detection at exact edge of block.
7104 Remaining options controls threshold detection near the edge. Each one for
7105 below/above or left/right. Setting any of those to @var{0} disables
7109 Set planes to filter. Default is to filter all available planes.
7112 @subsection Examples
7116 Deblock using weak filter and block size of 4 pixels.
7118 deblock=filter=weak:block=4
7122 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7123 deblocking more edges.
7125 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7129 Similar as above, but filter only first plane.
7131 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7135 Similar as above, but filter only second and third plane.
7137 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7144 Drop duplicated frames at regular intervals.
7146 The filter accepts the following options:
7150 Set the number of frames from which one will be dropped. Setting this to
7151 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7152 Default is @code{5}.
7155 Set the threshold for duplicate detection. If the difference metric for a frame
7156 is less than or equal to this value, then it is declared as duplicate. Default
7160 Set scene change threshold. Default is @code{15}.
7164 Set the size of the x and y-axis blocks used during metric calculations.
7165 Larger blocks give better noise suppression, but also give worse detection of
7166 small movements. Must be a power of two. Default is @code{32}.
7169 Mark main input as a pre-processed input and activate clean source input
7170 stream. This allows the input to be pre-processed with various filters to help
7171 the metrics calculation while keeping the frame selection lossless. When set to
7172 @code{1}, the first stream is for the pre-processed input, and the second
7173 stream is the clean source from where the kept frames are chosen. Default is
7177 Set whether or not chroma is considered in the metric calculations. Default is
7183 Apply 2D deconvolution of video stream in frequency domain using second stream
7186 The filter accepts the following options:
7190 Set which planes to process.
7193 Set which impulse video frames will be processed, can be @var{first}
7194 or @var{all}. Default is @var{all}.
7197 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7198 and height are not same and not power of 2 or if stream prior to convolving
7202 The @code{deconvolve} filter also supports the @ref{framesync} options.
7206 Apply deflate effect to the video.
7208 This filter replaces the pixel by the local(3x3) average by taking into account
7209 only values lower than the pixel.
7211 It accepts the following options:
7218 Limit the maximum change for each plane, default is 65535.
7219 If 0, plane will remain unchanged.
7224 Remove temporal frame luminance variations.
7226 It accepts the following options:
7230 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7233 Set averaging mode to smooth temporal luminance variations.
7235 Available values are:
7260 Do not actually modify frame. Useful when one only wants metadata.
7265 Remove judder produced by partially interlaced telecined content.
7267 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7268 source was partially telecined content then the output of @code{pullup,dejudder}
7269 will have a variable frame rate. May change the recorded frame rate of the
7270 container. Aside from that change, this filter will not affect constant frame
7273 The option available in this filter is:
7277 Specify the length of the window over which the judder repeats.
7279 Accepts any integer greater than 1. Useful values are:
7283 If the original was telecined from 24 to 30 fps (Film to NTSC).
7286 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7289 If a mixture of the two.
7292 The default is @samp{4}.
7297 Suppress a TV station logo by a simple interpolation of the surrounding
7298 pixels. Just set a rectangle covering the logo and watch it disappear
7299 (and sometimes something even uglier appear - your mileage may vary).
7301 It accepts the following parameters:
7306 Specify the top left corner coordinates of the logo. They must be
7311 Specify the width and height of the logo to clear. They must be
7315 Specify the thickness of the fuzzy edge of the rectangle (added to
7316 @var{w} and @var{h}). The default value is 1. This option is
7317 deprecated, setting higher values should no longer be necessary and
7321 When set to 1, a green rectangle is drawn on the screen to simplify
7322 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7323 The default value is 0.
7325 The rectangle is drawn on the outermost pixels which will be (partly)
7326 replaced with interpolated values. The values of the next pixels
7327 immediately outside this rectangle in each direction will be used to
7328 compute the interpolated pixel values inside the rectangle.
7332 @subsection Examples
7336 Set a rectangle covering the area with top left corner coordinates 0,0
7337 and size 100x77, and a band of size 10:
7339 delogo=x=0:y=0:w=100:h=77:band=10
7346 Attempt to fix small changes in horizontal and/or vertical shift. This
7347 filter helps remove camera shake from hand-holding a camera, bumping a
7348 tripod, moving on a vehicle, etc.
7350 The filter accepts the following options:
7358 Specify a rectangular area where to limit the search for motion
7360 If desired the search for motion vectors can be limited to a
7361 rectangular area of the frame defined by its top left corner, width
7362 and height. These parameters have the same meaning as the drawbox
7363 filter which can be used to visualise the position of the bounding
7366 This is useful when simultaneous movement of subjects within the frame
7367 might be confused for camera motion by the motion vector search.
7369 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
7370 then the full frame is used. This allows later options to be set
7371 without specifying the bounding box for the motion vector search.
7373 Default - search the whole frame.
7377 Specify the maximum extent of movement in x and y directions in the
7378 range 0-64 pixels. Default 16.
7381 Specify how to generate pixels to fill blanks at the edge of the
7382 frame. Available values are:
7385 Fill zeroes at blank locations
7387 Original image at blank locations
7389 Extruded edge value at blank locations
7391 Mirrored edge at blank locations
7393 Default value is @samp{mirror}.
7396 Specify the blocksize to use for motion search. Range 4-128 pixels,
7400 Specify the contrast threshold for blocks. Only blocks with more than
7401 the specified contrast (difference between darkest and lightest
7402 pixels) will be considered. Range 1-255, default 125.
7405 Specify the search strategy. Available values are:
7408 Set exhaustive search
7410 Set less exhaustive search.
7412 Default value is @samp{exhaustive}.
7415 If set then a detailed log of the motion search is written to the
7422 Remove unwanted contamination of foreground colors, caused by reflected color of
7423 greenscreen or bluescreen.
7425 This filter accepts the following options:
7429 Set what type of despill to use.
7432 Set how spillmap will be generated.
7435 Set how much to get rid of still remaining spill.
7438 Controls amount of red in spill area.
7441 Controls amount of green in spill area.
7442 Should be -1 for greenscreen.
7445 Controls amount of blue in spill area.
7446 Should be -1 for bluescreen.
7449 Controls brightness of spill area, preserving colors.
7452 Modify alpha from generated spillmap.
7457 Apply an exact inverse of the telecine operation. It requires a predefined
7458 pattern specified using the pattern option which must be the same as that passed
7459 to the telecine filter.
7461 This filter accepts the following options:
7470 The default value is @code{top}.
7474 A string of numbers representing the pulldown pattern you wish to apply.
7475 The default value is @code{23}.
7478 A number representing position of the first frame with respect to the telecine
7479 pattern. This is to be used if the stream is cut. The default value is @code{0}.
7484 Apply dilation effect to the video.
7486 This filter replaces the pixel by the local(3x3) maximum.
7488 It accepts the following options:
7495 Limit the maximum change for each plane, default is 65535.
7496 If 0, plane will remain unchanged.
7499 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7502 Flags to local 3x3 coordinates maps like this:
7511 Displace pixels as indicated by second and third input stream.
7513 It takes three input streams and outputs one stream, the first input is the
7514 source, and second and third input are displacement maps.
7516 The second input specifies how much to displace pixels along the
7517 x-axis, while the third input specifies how much to displace pixels
7519 If one of displacement map streams terminates, last frame from that
7520 displacement map will be used.
7522 Note that once generated, displacements maps can be reused over and over again.
7524 A description of the accepted options follows.
7528 Set displace behavior for pixels that are out of range.
7530 Available values are:
7533 Missing pixels are replaced by black pixels.
7536 Adjacent pixels will spread out to replace missing pixels.
7539 Out of range pixels are wrapped so they point to pixels of other side.
7542 Out of range pixels will be replaced with mirrored pixels.
7544 Default is @samp{smear}.
7548 @subsection Examples
7552 Add ripple effect to rgb input of video size hd720:
7554 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
7558 Add wave effect to rgb input of video size hd720:
7560 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
7566 Draw a colored box on the input image.
7568 It accepts the following parameters:
7573 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
7577 The expressions which specify the width and height of the box; if 0 they are interpreted as
7578 the input width and height. It defaults to 0.
7581 Specify the color of the box to write. For the general syntax of this option,
7582 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7583 value @code{invert} is used, the box edge color is the same as the
7584 video with inverted luma.
7587 The expression which sets the thickness of the box edge.
7588 A value of @code{fill} will create a filled box. Default value is @code{3}.
7590 See below for the list of accepted constants.
7593 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
7594 will overwrite the video's color and alpha pixels.
7595 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
7598 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7599 following constants:
7603 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7607 horizontal and vertical chroma subsample values. For example for the
7608 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7612 The input width and height.
7615 The input sample aspect ratio.
7619 The x and y offset coordinates where the box is drawn.
7623 The width and height of the drawn box.
7626 The thickness of the drawn box.
7628 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7629 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7633 @subsection Examples
7637 Draw a black box around the edge of the input image:
7643 Draw a box with color red and an opacity of 50%:
7645 drawbox=10:20:200:60:red@@0.5
7648 The previous example can be specified as:
7650 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
7654 Fill the box with pink color:
7656 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
7660 Draw a 2-pixel red 2.40:1 mask:
7662 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
7668 Draw a grid on the input image.
7670 It accepts the following parameters:
7675 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
7679 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
7680 input width and height, respectively, minus @code{thickness}, so image gets
7681 framed. Default to 0.
7684 Specify the color of the grid. For the general syntax of this option,
7685 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
7686 value @code{invert} is used, the grid color is the same as the
7687 video with inverted luma.
7690 The expression which sets the thickness of the grid line. Default value is @code{1}.
7692 See below for the list of accepted constants.
7695 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
7696 will overwrite the video's color and alpha pixels.
7697 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
7700 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
7701 following constants:
7705 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
7709 horizontal and vertical chroma subsample values. For example for the
7710 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7714 The input grid cell width and height.
7717 The input sample aspect ratio.
7721 The x and y coordinates of some point of grid intersection (meant to configure offset).
7725 The width and height of the drawn cell.
7728 The thickness of the drawn cell.
7730 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
7731 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
7735 @subsection Examples
7739 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
7741 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
7745 Draw a white 3x3 grid with an opacity of 50%:
7747 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
7754 Draw a text string or text from a specified file on top of a video, using the
7755 libfreetype library.
7757 To enable compilation of this filter, you need to configure FFmpeg with
7758 @code{--enable-libfreetype}.
7759 To enable default font fallback and the @var{font} option you need to
7760 configure FFmpeg with @code{--enable-libfontconfig}.
7761 To enable the @var{text_shaping} option, you need to configure FFmpeg with
7762 @code{--enable-libfribidi}.
7766 It accepts the following parameters:
7771 Used to draw a box around text using the background color.
7772 The value must be either 1 (enable) or 0 (disable).
7773 The default value of @var{box} is 0.
7776 Set the width of the border to be drawn around the box using @var{boxcolor}.
7777 The default value of @var{boxborderw} is 0.
7780 The color to be used for drawing box around text. For the syntax of this
7781 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7783 The default value of @var{boxcolor} is "white".
7786 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
7787 The default value of @var{line_spacing} is 0.
7790 Set the width of the border to be drawn around the text using @var{bordercolor}.
7791 The default value of @var{borderw} is 0.
7794 Set the color to be used for drawing border around text. For the syntax of this
7795 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7797 The default value of @var{bordercolor} is "black".
7800 Select how the @var{text} is expanded. Can be either @code{none},
7801 @code{strftime} (deprecated) or
7802 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
7806 Set a start time for the count. Value is in microseconds. Only applied
7807 in the deprecated strftime expansion mode. To emulate in normal expansion
7808 mode use the @code{pts} function, supplying the start time (in seconds)
7809 as the second argument.
7812 If true, check and fix text coords to avoid clipping.
7815 The color to be used for drawing fonts. For the syntax of this option, check
7816 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
7818 The default value of @var{fontcolor} is "black".
7820 @item fontcolor_expr
7821 String which is expanded the same way as @var{text} to obtain dynamic
7822 @var{fontcolor} value. By default this option has empty value and is not
7823 processed. When this option is set, it overrides @var{fontcolor} option.
7826 The font family to be used for drawing text. By default Sans.
7829 The font file to be used for drawing text. The path must be included.
7830 This parameter is mandatory if the fontconfig support is disabled.
7833 Draw the text applying alpha blending. The value can
7834 be a number between 0.0 and 1.0.
7835 The expression accepts the same variables @var{x, y} as well.
7836 The default value is 1.
7837 Please see @var{fontcolor_expr}.
7840 The font size to be used for drawing text.
7841 The default value of @var{fontsize} is 16.
7844 If set to 1, attempt to shape the text (for example, reverse the order of
7845 right-to-left text and join Arabic characters) before drawing it.
7846 Otherwise, just draw the text exactly as given.
7847 By default 1 (if supported).
7850 The flags to be used for loading the fonts.
7852 The flags map the corresponding flags supported by libfreetype, and are
7853 a combination of the following values:
7860 @item vertical_layout
7861 @item force_autohint
7864 @item ignore_global_advance_width
7866 @item ignore_transform
7872 Default value is "default".
7874 For more information consult the documentation for the FT_LOAD_*
7878 The color to be used for drawing a shadow behind the drawn text. For the
7879 syntax of this option, check the @ref{color syntax,,"Color" section in the
7880 ffmpeg-utils manual,ffmpeg-utils}.
7882 The default value of @var{shadowcolor} is "black".
7886 The x and y offsets for the text shadow position with respect to the
7887 position of the text. They can be either positive or negative
7888 values. The default value for both is "0".
7891 The starting frame number for the n/frame_num variable. The default value
7895 The size in number of spaces to use for rendering the tab.
7899 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
7900 format. It can be used with or without text parameter. @var{timecode_rate}
7901 option must be specified.
7903 @item timecode_rate, rate, r
7904 Set the timecode frame rate (timecode only). Value will be rounded to nearest
7905 integer. Minimum value is "1".
7906 Drop-frame timecode is supported for frame rates 30 & 60.
7909 If set to 1, the output of the timecode option will wrap around at 24 hours.
7910 Default is 0 (disabled).
7913 The text string to be drawn. The text must be a sequence of UTF-8
7915 This parameter is mandatory if no file is specified with the parameter
7919 A text file containing text to be drawn. The text must be a sequence
7920 of UTF-8 encoded characters.
7922 This parameter is mandatory if no text string is specified with the
7923 parameter @var{text}.
7925 If both @var{text} and @var{textfile} are specified, an error is thrown.
7928 If set to 1, the @var{textfile} will be reloaded before each frame.
7929 Be sure to update it atomically, or it may be read partially, or even fail.
7933 The expressions which specify the offsets where text will be drawn
7934 within the video frame. They are relative to the top/left border of the
7937 The default value of @var{x} and @var{y} is "0".
7939 See below for the list of accepted constants and functions.
7942 The parameters for @var{x} and @var{y} are expressions containing the
7943 following constants and functions:
7947 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7951 horizontal and vertical chroma subsample values. For example for the
7952 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7955 the height of each text line
7963 @item max_glyph_a, ascent
7964 the maximum distance from the baseline to the highest/upper grid
7965 coordinate used to place a glyph outline point, for all the rendered
7967 It is a positive value, due to the grid's orientation with the Y axis
7970 @item max_glyph_d, descent
7971 the maximum distance from the baseline to the lowest grid coordinate
7972 used to place a glyph outline point, for all the rendered glyphs.
7973 This is a negative value, due to the grid's orientation, with the Y axis
7977 maximum glyph height, that is the maximum height for all the glyphs
7978 contained in the rendered text, it is equivalent to @var{ascent} -
7982 maximum glyph width, that is the maximum width for all the glyphs
7983 contained in the rendered text
7986 the number of input frame, starting from 0
7988 @item rand(min, max)
7989 return a random number included between @var{min} and @var{max}
7992 The input sample aspect ratio.
7995 timestamp expressed in seconds, NAN if the input timestamp is unknown
7998 the height of the rendered text
8001 the width of the rendered text
8005 the x and y offset coordinates where the text is drawn.
8007 These parameters allow the @var{x} and @var{y} expressions to refer
8008 each other, so you can for example specify @code{y=x/dar}.
8011 @anchor{drawtext_expansion}
8012 @subsection Text expansion
8014 If @option{expansion} is set to @code{strftime},
8015 the filter recognizes strftime() sequences in the provided text and
8016 expands them accordingly. Check the documentation of strftime(). This
8017 feature is deprecated.
8019 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8021 If @option{expansion} is set to @code{normal} (which is the default),
8022 the following expansion mechanism is used.
8024 The backslash character @samp{\}, followed by any character, always expands to
8025 the second character.
8027 Sequences of the form @code{%@{...@}} are expanded. The text between the
8028 braces is a function name, possibly followed by arguments separated by ':'.
8029 If the arguments contain special characters or delimiters (':' or '@}'),
8030 they should be escaped.
8032 Note that they probably must also be escaped as the value for the
8033 @option{text} option in the filter argument string and as the filter
8034 argument in the filtergraph description, and possibly also for the shell,
8035 that makes up to four levels of escaping; using a text file avoids these
8038 The following functions are available:
8043 The expression evaluation result.
8045 It must take one argument specifying the expression to be evaluated,
8046 which accepts the same constants and functions as the @var{x} and
8047 @var{y} values. Note that not all constants should be used, for
8048 example the text size is not known when evaluating the expression, so
8049 the constants @var{text_w} and @var{text_h} will have an undefined
8052 @item expr_int_format, eif
8053 Evaluate the expression's value and output as formatted integer.
8055 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8056 The second argument specifies the output format. Allowed values are @samp{x},
8057 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8058 @code{printf} function.
8059 The third parameter is optional and sets the number of positions taken by the output.
8060 It can be used to add padding with zeros from the left.
8063 The time at which the filter is running, expressed in UTC.
8064 It can accept an argument: a strftime() format string.
8067 The time at which the filter is running, expressed in the local time zone.
8068 It can accept an argument: a strftime() format string.
8071 Frame metadata. Takes one or two arguments.
8073 The first argument is mandatory and specifies the metadata key.
8075 The second argument is optional and specifies a default value, used when the
8076 metadata key is not found or empty.
8079 The frame number, starting from 0.
8082 A 1 character description of the current picture type.
8085 The timestamp of the current frame.
8086 It can take up to three arguments.
8088 The first argument is the format of the timestamp; it defaults to @code{flt}
8089 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8090 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8091 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8092 @code{localtime} stands for the timestamp of the frame formatted as
8093 local time zone time.
8095 The second argument is an offset added to the timestamp.
8097 If the format is set to @code{localtime} or @code{gmtime},
8098 a third argument may be supplied: a strftime() format string.
8099 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8102 @subsection Examples
8106 Draw "Test Text" with font FreeSerif, using the default values for the
8107 optional parameters.
8110 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8114 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8115 and y=50 (counting from the top-left corner of the screen), text is
8116 yellow with a red box around it. Both the text and the box have an
8120 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8121 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8124 Note that the double quotes are not necessary if spaces are not used
8125 within the parameter list.
8128 Show the text at the center of the video frame:
8130 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8134 Show the text at a random position, switching to a new position every 30 seconds:
8136 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)"
8140 Show a text line sliding from right to left in the last row of the video
8141 frame. The file @file{LONG_LINE} is assumed to contain a single line
8144 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8148 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8150 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8154 Draw a single green letter "g", at the center of the input video.
8155 The glyph baseline is placed at half screen height.
8157 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8161 Show text for 1 second every 3 seconds:
8163 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8167 Use fontconfig to set the font. Note that the colons need to be escaped.
8169 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8173 Print the date of a real-time encoding (see strftime(3)):
8175 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8179 Show text fading in and out (appearing/disappearing):
8182 DS=1.0 # display start
8183 DE=10.0 # display end
8184 FID=1.5 # fade in duration
8185 FOD=5 # fade out duration
8186 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 @}"
8190 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8191 and the @option{fontsize} value are included in the @option{y} offset.
8193 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8194 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8199 For more information about libfreetype, check:
8200 @url{http://www.freetype.org/}.
8202 For more information about fontconfig, check:
8203 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8205 For more information about libfribidi, check:
8206 @url{http://fribidi.org/}.
8210 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8212 The filter accepts the following options:
8217 Set low and high threshold values used by the Canny thresholding
8220 The high threshold selects the "strong" edge pixels, which are then
8221 connected through 8-connectivity with the "weak" edge pixels selected
8222 by the low threshold.
8224 @var{low} and @var{high} threshold values must be chosen in the range
8225 [0,1], and @var{low} should be lesser or equal to @var{high}.
8227 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8231 Define the drawing mode.
8235 Draw white/gray wires on black background.
8238 Mix the colors to create a paint/cartoon effect.
8241 Default value is @var{wires}.
8244 @subsection Examples
8248 Standard edge detection with custom values for the hysteresis thresholding:
8250 edgedetect=low=0.1:high=0.4
8254 Painting effect without thresholding:
8256 edgedetect=mode=colormix:high=0
8261 Set brightness, contrast, saturation and approximate gamma adjustment.
8263 The filter accepts the following options:
8267 Set the contrast expression. The value must be a float value in range
8268 @code{-2.0} to @code{2.0}. The default value is "1".
8271 Set the brightness expression. The value must be a float value in
8272 range @code{-1.0} to @code{1.0}. The default value is "0".
8275 Set the saturation expression. The value must be a float in
8276 range @code{0.0} to @code{3.0}. The default value is "1".
8279 Set the gamma expression. The value must be a float in range
8280 @code{0.1} to @code{10.0}. The default value is "1".
8283 Set the gamma expression for red. The value must be a float in
8284 range @code{0.1} to @code{10.0}. The default value is "1".
8287 Set the gamma expression for green. The value must be a float in range
8288 @code{0.1} to @code{10.0}. The default value is "1".
8291 Set the gamma expression for blue. The value must be a float in range
8292 @code{0.1} to @code{10.0}. The default value is "1".
8295 Set the gamma weight expression. It can be used to reduce the effect
8296 of a high gamma value on bright image areas, e.g. keep them from
8297 getting overamplified and just plain white. The value must be a float
8298 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8299 gamma correction all the way down while @code{1.0} leaves it at its
8300 full strength. Default is "1".
8303 Set when the expressions for brightness, contrast, saturation and
8304 gamma expressions are evaluated.
8306 It accepts the following values:
8309 only evaluate expressions once during the filter initialization or
8310 when a command is processed
8313 evaluate expressions for each incoming frame
8316 Default value is @samp{init}.
8319 The expressions accept the following parameters:
8322 frame count of the input frame starting from 0
8325 byte position of the corresponding packet in the input file, NAN if
8329 frame rate of the input video, NAN if the input frame rate is unknown
8332 timestamp expressed in seconds, NAN if the input timestamp is unknown
8335 @subsection Commands
8336 The filter supports the following commands:
8340 Set the contrast expression.
8343 Set the brightness expression.
8346 Set the saturation expression.
8349 Set the gamma expression.
8352 Set the gamma_r expression.
8355 Set gamma_g expression.
8358 Set gamma_b expression.
8361 Set gamma_weight expression.
8363 The command accepts the same syntax of the corresponding option.
8365 If the specified expression is not valid, it is kept at its current
8372 Apply erosion effect to the video.
8374 This filter replaces the pixel by the local(3x3) minimum.
8376 It accepts the following options:
8383 Limit the maximum change for each plane, default is 65535.
8384 If 0, plane will remain unchanged.
8387 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8390 Flags to local 3x3 coordinates maps like this:
8397 @section extractplanes
8399 Extract color channel components from input video stream into
8400 separate grayscale video streams.
8402 The filter accepts the following option:
8406 Set plane(s) to extract.
8408 Available values for planes are:
8419 Choosing planes not available in the input will result in an error.
8420 That means you cannot select @code{r}, @code{g}, @code{b} planes
8421 with @code{y}, @code{u}, @code{v} planes at same time.
8424 @subsection Examples
8428 Extract luma, u and v color channel component from input video frame
8429 into 3 grayscale outputs:
8431 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
8437 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
8439 For each input image, the filter will compute the optimal mapping from
8440 the input to the output given the codebook length, that is the number
8441 of distinct output colors.
8443 This filter accepts the following options.
8446 @item codebook_length, l
8447 Set codebook length. The value must be a positive integer, and
8448 represents the number of distinct output colors. Default value is 256.
8451 Set the maximum number of iterations to apply for computing the optimal
8452 mapping. The higher the value the better the result and the higher the
8453 computation time. Default value is 1.
8456 Set a random seed, must be an integer included between 0 and
8457 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
8458 will try to use a good random seed on a best effort basis.
8461 Set pal8 output pixel format. This option does not work with codebook
8462 length greater than 256.
8467 Measure graylevel entropy in histogram of color channels of video frames.
8469 It accepts the following parameters:
8473 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
8475 @var{diff} mode measures entropy of histogram delta values, absolute differences
8476 between neighbour histogram values.
8481 Apply a fade-in/out effect to the input video.
8483 It accepts the following parameters:
8487 The effect type can be either "in" for a fade-in, or "out" for a fade-out
8489 Default is @code{in}.
8491 @item start_frame, s
8492 Specify the number of the frame to start applying the fade
8493 effect at. Default is 0.
8496 The number of frames that the fade effect lasts. At the end of the
8497 fade-in effect, the output video will have the same intensity as the input video.
8498 At the end of the fade-out transition, the output video will be filled with the
8499 selected @option{color}.
8503 If set to 1, fade only alpha channel, if one exists on the input.
8506 @item start_time, st
8507 Specify the timestamp (in seconds) of the frame to start to apply the fade
8508 effect. If both start_frame and start_time are specified, the fade will start at
8509 whichever comes last. Default is 0.
8512 The number of seconds for which the fade effect has to last. At the end of the
8513 fade-in effect the output video will have the same intensity as the input video,
8514 at the end of the fade-out transition the output video will be filled with the
8515 selected @option{color}.
8516 If both duration and nb_frames are specified, duration is used. Default is 0
8517 (nb_frames is used by default).
8520 Specify the color of the fade. Default is "black".
8523 @subsection Examples
8527 Fade in the first 30 frames of video:
8532 The command above is equivalent to:
8538 Fade out the last 45 frames of a 200-frame video:
8541 fade=type=out:start_frame=155:nb_frames=45
8545 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
8547 fade=in:0:25, fade=out:975:25
8551 Make the first 5 frames yellow, then fade in from frame 5-24:
8553 fade=in:5:20:color=yellow
8557 Fade in alpha over first 25 frames of video:
8559 fade=in:0:25:alpha=1
8563 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
8565 fade=t=in:st=5.5:d=0.5
8571 Apply arbitrary expressions to samples in frequency domain
8575 Adjust the dc value (gain) of the luma plane of the image. The filter
8576 accepts an integer value in range @code{0} to @code{1000}. The default
8577 value is set to @code{0}.
8580 Adjust the dc value (gain) of the 1st chroma plane of the image. The
8581 filter accepts an integer value in range @code{0} to @code{1000}. The
8582 default value is set to @code{0}.
8585 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
8586 filter accepts an integer value in range @code{0} to @code{1000}. The
8587 default value is set to @code{0}.
8590 Set the frequency domain weight expression for the luma plane.
8593 Set the frequency domain weight expression for the 1st chroma plane.
8596 Set the frequency domain weight expression for the 2nd chroma plane.
8599 Set when the expressions are evaluated.
8601 It accepts the following values:
8604 Only evaluate expressions once during the filter initialization.
8607 Evaluate expressions for each incoming frame.
8610 Default value is @samp{init}.
8612 The filter accepts the following variables:
8615 The coordinates of the current sample.
8619 The width and height of the image.
8622 The number of input frame, starting from 0.
8625 @subsection Examples
8631 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
8637 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
8643 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
8649 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
8656 Extract a single field from an interlaced image using stride
8657 arithmetic to avoid wasting CPU time. The output frames are marked as
8660 The filter accepts the following options:
8664 Specify whether to extract the top (if the value is @code{0} or
8665 @code{top}) or the bottom field (if the value is @code{1} or
8671 Create new frames by copying the top and bottom fields from surrounding frames
8672 supplied as numbers by the hint file.
8676 Set file containing hints: absolute/relative frame numbers.
8678 There must be one line for each frame in a clip. Each line must contain two
8679 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
8680 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
8681 is current frame number for @code{absolute} mode or out of [-1, 1] range
8682 for @code{relative} mode. First number tells from which frame to pick up top
8683 field and second number tells from which frame to pick up bottom field.
8685 If optionally followed by @code{+} output frame will be marked as interlaced,
8686 else if followed by @code{-} output frame will be marked as progressive, else
8687 it will be marked same as input frame.
8688 If line starts with @code{#} or @code{;} that line is skipped.
8691 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
8694 Example of first several lines of @code{hint} file for @code{relative} mode:
8697 1,0 - # second frame, use third's frame top field and second's frame bottom field
8698 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
8715 Field matching filter for inverse telecine. It is meant to reconstruct the
8716 progressive frames from a telecined stream. The filter does not drop duplicated
8717 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
8718 followed by a decimation filter such as @ref{decimate} in the filtergraph.
8720 The separation of the field matching and the decimation is notably motivated by
8721 the possibility of inserting a de-interlacing filter fallback between the two.
8722 If the source has mixed telecined and real interlaced content,
8723 @code{fieldmatch} will not be able to match fields for the interlaced parts.
8724 But these remaining combed frames will be marked as interlaced, and thus can be
8725 de-interlaced by a later filter such as @ref{yadif} before decimation.
8727 In addition to the various configuration options, @code{fieldmatch} can take an
8728 optional second stream, activated through the @option{ppsrc} option. If
8729 enabled, the frames reconstruction will be based on the fields and frames from
8730 this second stream. This allows the first input to be pre-processed in order to
8731 help the various algorithms of the filter, while keeping the output lossless
8732 (assuming the fields are matched properly). Typically, a field-aware denoiser,
8733 or brightness/contrast adjustments can help.
8735 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
8736 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
8737 which @code{fieldmatch} is based on. While the semantic and usage are very
8738 close, some behaviour and options names can differ.
8740 The @ref{decimate} filter currently only works for constant frame rate input.
8741 If your input has mixed telecined (30fps) and progressive content with a lower
8742 framerate like 24fps use the following filterchain to produce the necessary cfr
8743 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
8745 The filter accepts the following options:
8749 Specify the assumed field order of the input stream. Available values are:
8753 Auto detect parity (use FFmpeg's internal parity value).
8755 Assume bottom field first.
8757 Assume top field first.
8760 Note that it is sometimes recommended not to trust the parity announced by the
8763 Default value is @var{auto}.
8766 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
8767 sense that it won't risk creating jerkiness due to duplicate frames when
8768 possible, but if there are bad edits or blended fields it will end up
8769 outputting combed frames when a good match might actually exist. On the other
8770 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
8771 but will almost always find a good frame if there is one. The other values are
8772 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
8773 jerkiness and creating duplicate frames versus finding good matches in sections
8774 with bad edits, orphaned fields, blended fields, etc.
8776 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
8778 Available values are:
8782 2-way matching (p/c)
8784 2-way matching, and trying 3rd match if still combed (p/c + n)
8786 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
8788 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
8789 still combed (p/c + n + u/b)
8791 3-way matching (p/c/n)
8793 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
8794 detected as combed (p/c/n + u/b)
8797 The parenthesis at the end indicate the matches that would be used for that
8798 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
8801 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
8804 Default value is @var{pc_n}.
8807 Mark the main input stream as a pre-processed input, and enable the secondary
8808 input stream as the clean source to pick the fields from. See the filter
8809 introduction for more details. It is similar to the @option{clip2} feature from
8812 Default value is @code{0} (disabled).
8815 Set the field to match from. It is recommended to set this to the same value as
8816 @option{order} unless you experience matching failures with that setting. In
8817 certain circumstances changing the field that is used to match from can have a
8818 large impact on matching performance. Available values are:
8822 Automatic (same value as @option{order}).
8824 Match from the bottom field.
8826 Match from the top field.
8829 Default value is @var{auto}.
8832 Set whether or not chroma is included during the match comparisons. In most
8833 cases it is recommended to leave this enabled. You should set this to @code{0}
8834 only if your clip has bad chroma problems such as heavy rainbowing or other
8835 artifacts. Setting this to @code{0} could also be used to speed things up at
8836 the cost of some accuracy.
8838 Default value is @code{1}.
8842 These define an exclusion band which excludes the lines between @option{y0} and
8843 @option{y1} from being included in the field matching decision. An exclusion
8844 band can be used to ignore subtitles, a logo, or other things that may
8845 interfere with the matching. @option{y0} sets the starting scan line and
8846 @option{y1} sets the ending line; all lines in between @option{y0} and
8847 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
8848 @option{y0} and @option{y1} to the same value will disable the feature.
8849 @option{y0} and @option{y1} defaults to @code{0}.
8852 Set the scene change detection threshold as a percentage of maximum change on
8853 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
8854 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
8855 @option{scthresh} is @code{[0.0, 100.0]}.
8857 Default value is @code{12.0}.
8860 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
8861 account the combed scores of matches when deciding what match to use as the
8862 final match. Available values are:
8866 No final matching based on combed scores.
8868 Combed scores are only used when a scene change is detected.
8870 Use combed scores all the time.
8873 Default is @var{sc}.
8876 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
8877 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
8878 Available values are:
8882 No forced calculation.
8884 Force p/c/n calculations.
8886 Force p/c/n/u/b calculations.
8889 Default value is @var{none}.
8892 This is the area combing threshold used for combed frame detection. This
8893 essentially controls how "strong" or "visible" combing must be to be detected.
8894 Larger values mean combing must be more visible and smaller values mean combing
8895 can be less visible or strong and still be detected. Valid settings are from
8896 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
8897 be detected as combed). This is basically a pixel difference value. A good
8898 range is @code{[8, 12]}.
8900 Default value is @code{9}.
8903 Sets whether or not chroma is considered in the combed frame decision. Only
8904 disable this if your source has chroma problems (rainbowing, etc.) that are
8905 causing problems for the combed frame detection with chroma enabled. Actually,
8906 using @option{chroma}=@var{0} is usually more reliable, except for the case
8907 where there is chroma only combing in the source.
8909 Default value is @code{0}.
8913 Respectively set the x-axis and y-axis size of the window used during combed
8914 frame detection. This has to do with the size of the area in which
8915 @option{combpel} pixels are required to be detected as combed for a frame to be
8916 declared combed. See the @option{combpel} parameter description for more info.
8917 Possible values are any number that is a power of 2 starting at 4 and going up
8920 Default value is @code{16}.
8923 The number of combed pixels inside any of the @option{blocky} by
8924 @option{blockx} size blocks on the frame for the frame to be detected as
8925 combed. While @option{cthresh} controls how "visible" the combing must be, this
8926 setting controls "how much" combing there must be in any localized area (a
8927 window defined by the @option{blockx} and @option{blocky} settings) on the
8928 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
8929 which point no frames will ever be detected as combed). This setting is known
8930 as @option{MI} in TFM/VFM vocabulary.
8932 Default value is @code{80}.
8935 @anchor{p/c/n/u/b meaning}
8936 @subsection p/c/n/u/b meaning
8938 @subsubsection p/c/n
8940 We assume the following telecined stream:
8943 Top fields: 1 2 2 3 4
8944 Bottom fields: 1 2 3 4 4
8947 The numbers correspond to the progressive frame the fields relate to. Here, the
8948 first two frames are progressive, the 3rd and 4th are combed, and so on.
8950 When @code{fieldmatch} is configured to run a matching from bottom
8951 (@option{field}=@var{bottom}) this is how this input stream get transformed:
8956 B 1 2 3 4 4 <-- matching reference
8965 As a result of the field matching, we can see that some frames get duplicated.
8966 To perform a complete inverse telecine, you need to rely on a decimation filter
8967 after this operation. See for instance the @ref{decimate} filter.
8969 The same operation now matching from top fields (@option{field}=@var{top})
8974 T 1 2 2 3 4 <-- matching reference
8984 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8985 basically, they refer to the frame and field of the opposite parity:
8988 @item @var{p} matches the field of the opposite parity in the previous frame
8989 @item @var{c} matches the field of the opposite parity in the current frame
8990 @item @var{n} matches the field of the opposite parity in the next frame
8995 The @var{u} and @var{b} matching are a bit special in the sense that they match
8996 from the opposite parity flag. In the following examples, we assume that we are
8997 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8998 'x' is placed above and below each matched fields.
9000 With bottom matching (@option{field}=@var{bottom}):
9005 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9006 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9014 With top matching (@option{field}=@var{top}):
9019 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9020 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9028 @subsection Examples
9030 Simple IVTC of a top field first telecined stream:
9032 fieldmatch=order=tff:combmatch=none, decimate
9035 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9037 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9042 Transform the field order of the input video.
9044 It accepts the following parameters:
9049 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9050 for bottom field first.
9053 The default value is @samp{tff}.
9055 The transformation is done by shifting the picture content up or down
9056 by one line, and filling the remaining line with appropriate picture content.
9057 This method is consistent with most broadcast field order converters.
9059 If the input video is not flagged as being interlaced, or it is already
9060 flagged as being of the required output field order, then this filter does
9061 not alter the incoming video.
9063 It is very useful when converting to or from PAL DV material,
9064 which is bottom field first.
9068 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9071 @section fifo, afifo
9073 Buffer input images and send them when they are requested.
9075 It is mainly useful when auto-inserted by the libavfilter
9078 It does not take parameters.
9080 @section fillborders
9082 Fill borders of the input video, without changing video stream dimensions.
9083 Sometimes video can have garbage at the four edges and you may not want to
9084 crop video input to keep size multiple of some number.
9086 This filter accepts the following options:
9090 Number of pixels to fill from left border.
9093 Number of pixels to fill from right border.
9096 Number of pixels to fill from top border.
9099 Number of pixels to fill from bottom border.
9104 It accepts the following values:
9107 fill pixels using outermost pixels
9110 fill pixels using mirroring
9113 fill pixels with constant value
9116 Default is @var{smear}.
9119 Set color for pixels in fixed mode. Default is @var{black}.
9124 Find a rectangular object
9126 It accepts the following options:
9130 Filepath of the object image, needs to be in gray8.
9133 Detection threshold, default is 0.5.
9136 Number of mipmaps, default is 3.
9138 @item xmin, ymin, xmax, ymax
9139 Specifies the rectangle in which to search.
9142 @subsection Examples
9146 Generate a representative palette of a given video using @command{ffmpeg}:
9148 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9154 Cover a rectangular object
9156 It accepts the following options:
9160 Filepath of the optional cover image, needs to be in yuv420.
9165 It accepts the following values:
9168 cover it by the supplied image
9170 cover it by interpolating the surrounding pixels
9173 Default value is @var{blur}.
9176 @subsection Examples
9180 Generate a representative palette of a given video using @command{ffmpeg}:
9182 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9188 Flood area with values of same pixel components with another values.
9190 It accepts the following options:
9193 Set pixel x coordinate.
9196 Set pixel y coordinate.
9199 Set source #0 component value.
9202 Set source #1 component value.
9205 Set source #2 component value.
9208 Set source #3 component value.
9211 Set destination #0 component value.
9214 Set destination #1 component value.
9217 Set destination #2 component value.
9220 Set destination #3 component value.
9226 Convert the input video to one of the specified pixel formats.
9227 Libavfilter will try to pick one that is suitable as input to
9230 It accepts the following parameters:
9234 A '|'-separated list of pixel format names, such as
9235 "pix_fmts=yuv420p|monow|rgb24".
9239 @subsection Examples
9243 Convert the input video to the @var{yuv420p} format
9245 format=pix_fmts=yuv420p
9248 Convert the input video to any of the formats in the list
9250 format=pix_fmts=yuv420p|yuv444p|yuv410p
9257 Convert the video to specified constant frame rate by duplicating or dropping
9258 frames as necessary.
9260 It accepts the following parameters:
9264 The desired output frame rate. The default is @code{25}.
9267 Assume the first PTS should be the given value, in seconds. This allows for
9268 padding/trimming at the start of stream. By default, no assumption is made
9269 about the first frame's expected PTS, so no padding or trimming is done.
9270 For example, this could be set to 0 to pad the beginning with duplicates of
9271 the first frame if a video stream starts after the audio stream or to trim any
9272 frames with a negative PTS.
9275 Timestamp (PTS) rounding method.
9277 Possible values are:
9284 round towards -infinity
9286 round towards +infinity
9290 The default is @code{near}.
9293 Action performed when reading the last frame.
9295 Possible values are:
9298 Use same timestamp rounding method as used for other frames.
9300 Pass through last frame if input duration has not been reached yet.
9302 The default is @code{round}.
9306 Alternatively, the options can be specified as a flat string:
9307 @var{fps}[:@var{start_time}[:@var{round}]].
9309 See also the @ref{setpts} filter.
9311 @subsection Examples
9315 A typical usage in order to set the fps to 25:
9321 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
9323 fps=fps=film:round=near
9329 Pack two different video streams into a stereoscopic video, setting proper
9330 metadata on supported codecs. The two views should have the same size and
9331 framerate and processing will stop when the shorter video ends. Please note
9332 that you may conveniently adjust view properties with the @ref{scale} and
9335 It accepts the following parameters:
9339 The desired packing format. Supported values are:
9344 The views are next to each other (default).
9347 The views are on top of each other.
9350 The views are packed by line.
9353 The views are packed by column.
9356 The views are temporally interleaved.
9365 # Convert left and right views into a frame-sequential video
9366 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
9368 # Convert views into a side-by-side video with the same output resolution as the input
9369 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
9374 Change the frame rate by interpolating new video output frames from the source
9377 This filter is not designed to function correctly with interlaced media. If
9378 you wish to change the frame rate of interlaced media then you are required
9379 to deinterlace before this filter and re-interlace after this filter.
9381 A description of the accepted options follows.
9385 Specify the output frames per second. This option can also be specified
9386 as a value alone. The default is @code{50}.
9389 Specify the start of a range where the output frame will be created as a
9390 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9391 the default is @code{15}.
9394 Specify the end of a range where the output frame will be created as a
9395 linear interpolation of two frames. The range is [@code{0}-@code{255}],
9396 the default is @code{240}.
9399 Specify the level at which a scene change is detected as a value between
9400 0 and 100 to indicate a new scene; a low value reflects a low
9401 probability for the current frame to introduce a new scene, while a higher
9402 value means the current frame is more likely to be one.
9403 The default is @code{8.2}.
9406 Specify flags influencing the filter process.
9408 Available value for @var{flags} is:
9411 @item scene_change_detect, scd
9412 Enable scene change detection using the value of the option @var{scene}.
9413 This flag is enabled by default.
9419 Select one frame every N-th frame.
9421 This filter accepts the following option:
9424 Select frame after every @code{step} frames.
9425 Allowed values are positive integers higher than 0. Default value is @code{1}.
9431 Apply a frei0r effect to the input video.
9433 To enable the compilation of this filter, you need to install the frei0r
9434 header and configure FFmpeg with @code{--enable-frei0r}.
9436 It accepts the following parameters:
9441 The name of the frei0r effect to load. If the environment variable
9442 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
9443 directories specified by the colon-separated list in @env{FREI0R_PATH}.
9444 Otherwise, the standard frei0r paths are searched, in this order:
9445 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
9446 @file{/usr/lib/frei0r-1/}.
9449 A '|'-separated list of parameters to pass to the frei0r effect.
9453 A frei0r effect parameter can be a boolean (its value is either
9454 "y" or "n"), a double, a color (specified as
9455 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
9456 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
9457 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
9458 a position (specified as @var{X}/@var{Y}, where
9459 @var{X} and @var{Y} are floating point numbers) and/or a string.
9461 The number and types of parameters depend on the loaded effect. If an
9462 effect parameter is not specified, the default value is set.
9464 @subsection Examples
9468 Apply the distort0r effect, setting the first two double parameters:
9470 frei0r=filter_name=distort0r:filter_params=0.5|0.01
9474 Apply the colordistance effect, taking a color as the first parameter:
9476 frei0r=colordistance:0.2/0.3/0.4
9477 frei0r=colordistance:violet
9478 frei0r=colordistance:0x112233
9482 Apply the perspective effect, specifying the top left and top right image
9485 frei0r=perspective:0.2/0.2|0.8/0.2
9489 For more information, see
9490 @url{http://frei0r.dyne.org}
9494 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
9496 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
9497 processing filter, one of them is performed once per block, not per pixel.
9498 This allows for much higher speed.
9500 The filter accepts the following options:
9504 Set quality. This option defines the number of levels for averaging. It accepts
9505 an integer in the range 4-5. Default value is @code{4}.
9508 Force a constant quantization parameter. It accepts an integer in range 0-63.
9509 If not set, the filter will use the QP from the video stream (if available).
9512 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
9513 more details but also more artifacts, while higher values make the image smoother
9514 but also blurrier. Default value is @code{0} − PSNR optimal.
9517 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
9518 option may cause flicker since the B-Frames have often larger QP. Default is
9519 @code{0} (not enabled).
9525 Apply Gaussian blur filter.
9527 The filter accepts the following options:
9531 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
9534 Set number of steps for Gaussian approximation. Defauls is @code{1}.
9537 Set which planes to filter. By default all planes are filtered.
9540 Set vertical sigma, if negative it will be same as @code{sigma}.
9541 Default is @code{-1}.
9546 The filter accepts the following options:
9550 Set the luminance expression.
9552 Set the chrominance blue expression.
9554 Set the chrominance red expression.
9556 Set the alpha expression.
9558 Set the red expression.
9560 Set the green expression.
9562 Set the blue expression.
9565 The colorspace is selected according to the specified options. If one
9566 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
9567 options is specified, the filter will automatically select a YCbCr
9568 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
9569 @option{blue_expr} options is specified, it will select an RGB
9572 If one of the chrominance expression is not defined, it falls back on the other
9573 one. If no alpha expression is specified it will evaluate to opaque value.
9574 If none of chrominance expressions are specified, they will evaluate
9575 to the luminance expression.
9577 The expressions can use the following variables and functions:
9581 The sequential number of the filtered frame, starting from @code{0}.
9585 The coordinates of the current sample.
9589 The width and height of the image.
9593 Width and height scale depending on the currently filtered plane. It is the
9594 ratio between the corresponding luma plane number of pixels and the current
9595 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
9596 @code{0.5,0.5} for chroma planes.
9599 Time of the current frame, expressed in seconds.
9602 Return the value of the pixel at location (@var{x},@var{y}) of the current
9606 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
9610 Return the value of the pixel at location (@var{x},@var{y}) of the
9611 blue-difference chroma plane. Return 0 if there is no such plane.
9614 Return the value of the pixel at location (@var{x},@var{y}) of the
9615 red-difference chroma plane. Return 0 if there is no such plane.
9620 Return the value of the pixel at location (@var{x},@var{y}) of the
9621 red/green/blue component. Return 0 if there is no such component.
9624 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
9625 plane. Return 0 if there is no such plane.
9628 For functions, if @var{x} and @var{y} are outside the area, the value will be
9629 automatically clipped to the closer edge.
9631 @subsection Examples
9635 Flip the image horizontally:
9641 Generate a bidimensional sine wave, with angle @code{PI/3} and a
9642 wavelength of 100 pixels:
9644 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
9648 Generate a fancy enigmatic moving light:
9650 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
9654 Generate a quick emboss effect:
9656 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
9660 Modify RGB components depending on pixel position:
9662 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
9666 Create a radial gradient that is the same size as the input (also see
9667 the @ref{vignette} filter):
9669 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
9675 Fix the banding artifacts that are sometimes introduced into nearly flat
9676 regions by truncation to 8-bit color depth.
9677 Interpolate the gradients that should go where the bands are, and
9680 It is designed for playback only. Do not use it prior to
9681 lossy compression, because compression tends to lose the dither and
9682 bring back the bands.
9684 It accepts the following parameters:
9689 The maximum amount by which the filter will change any one pixel. This is also
9690 the threshold for detecting nearly flat regions. Acceptable values range from
9691 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
9695 The neighborhood to fit the gradient to. A larger radius makes for smoother
9696 gradients, but also prevents the filter from modifying the pixels near detailed
9697 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
9698 values will be clipped to the valid range.
9702 Alternatively, the options can be specified as a flat string:
9703 @var{strength}[:@var{radius}]
9705 @subsection Examples
9709 Apply the filter with a @code{3.5} strength and radius of @code{8}:
9715 Specify radius, omitting the strength (which will fall-back to the default
9726 Apply a Hald CLUT to a video stream.
9728 First input is the video stream to process, and second one is the Hald CLUT.
9729 The Hald CLUT input can be a simple picture or a complete video stream.
9731 The filter accepts the following options:
9735 Force termination when the shortest input terminates. Default is @code{0}.
9737 Continue applying the last CLUT after the end of the stream. A value of
9738 @code{0} disable the filter after the last frame of the CLUT is reached.
9739 Default is @code{1}.
9742 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
9743 filters share the same internals).
9745 More information about the Hald CLUT can be found on Eskil Steenberg's website
9746 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
9748 @subsection Workflow examples
9750 @subsubsection Hald CLUT video stream
9752 Generate an identity Hald CLUT stream altered with various effects:
9754 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
9757 Note: make sure you use a lossless codec.
9759 Then use it with @code{haldclut} to apply it on some random stream:
9761 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
9764 The Hald CLUT will be applied to the 10 first seconds (duration of
9765 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
9766 to the remaining frames of the @code{mandelbrot} stream.
9768 @subsubsection Hald CLUT with preview
9770 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
9771 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
9772 biggest possible square starting at the top left of the picture. The remaining
9773 padding pixels (bottom or right) will be ignored. This area can be used to add
9774 a preview of the Hald CLUT.
9776 Typically, the following generated Hald CLUT will be supported by the
9777 @code{haldclut} filter:
9780 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
9781 pad=iw+320 [padded_clut];
9782 smptebars=s=320x256, split [a][b];
9783 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
9784 [main][b] overlay=W-320" -frames:v 1 clut.png
9787 It contains the original and a preview of the effect of the CLUT: SMPTE color
9788 bars are displayed on the right-top, and below the same color bars processed by
9791 Then, the effect of this Hald CLUT can be visualized with:
9793 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
9798 Flip the input video horizontally.
9800 For example, to horizontally flip the input video with @command{ffmpeg}:
9802 ffmpeg -i in.avi -vf "hflip" out.avi
9806 This filter applies a global color histogram equalization on a
9809 It can be used to correct video that has a compressed range of pixel
9810 intensities. The filter redistributes the pixel intensities to
9811 equalize their distribution across the intensity range. It may be
9812 viewed as an "automatically adjusting contrast filter". This filter is
9813 useful only for correcting degraded or poorly captured source
9816 The filter accepts the following options:
9820 Determine the amount of equalization to be applied. As the strength
9821 is reduced, the distribution of pixel intensities more-and-more
9822 approaches that of the input frame. The value must be a float number
9823 in the range [0,1] and defaults to 0.200.
9826 Set the maximum intensity that can generated and scale the output
9827 values appropriately. The strength should be set as desired and then
9828 the intensity can be limited if needed to avoid washing-out. The value
9829 must be a float number in the range [0,1] and defaults to 0.210.
9832 Set the antibanding level. If enabled the filter will randomly vary
9833 the luminance of output pixels by a small amount to avoid banding of
9834 the histogram. Possible values are @code{none}, @code{weak} or
9835 @code{strong}. It defaults to @code{none}.
9840 Compute and draw a color distribution histogram for the input video.
9842 The computed histogram is a representation of the color component
9843 distribution in an image.
9845 Standard histogram displays the color components distribution in an image.
9846 Displays color graph for each color component. Shows distribution of
9847 the Y, U, V, A or R, G, B components, depending on input format, in the
9848 current frame. Below each graph a color component scale meter is shown.
9850 The filter accepts the following options:
9854 Set height of level. Default value is @code{200}.
9855 Allowed range is [50, 2048].
9858 Set height of color scale. Default value is @code{12}.
9859 Allowed range is [0, 40].
9863 It accepts the following values:
9866 Per color component graphs are placed below each other.
9869 Per color component graphs are placed side by side.
9872 Presents information identical to that in the @code{parade}, except
9873 that the graphs representing color components are superimposed directly
9876 Default is @code{stack}.
9879 Set mode. Can be either @code{linear}, or @code{logarithmic}.
9880 Default is @code{linear}.
9883 Set what color components to display.
9884 Default is @code{7}.
9887 Set foreground opacity. Default is @code{0.7}.
9890 Set background opacity. Default is @code{0.5}.
9893 @subsection Examples
9898 Calculate and draw histogram:
9900 ffplay -i input -vf histogram
9908 This is a high precision/quality 3d denoise filter. It aims to reduce
9909 image noise, producing smooth images and making still images really
9910 still. It should enhance compressibility.
9912 It accepts the following optional parameters:
9916 A non-negative floating point number which specifies spatial luma strength.
9919 @item chroma_spatial
9920 A non-negative floating point number which specifies spatial chroma strength.
9921 It defaults to 3.0*@var{luma_spatial}/4.0.
9924 A floating point number which specifies luma temporal strength. It defaults to
9925 6.0*@var{luma_spatial}/4.0.
9928 A floating point number which specifies chroma temporal strength. It defaults to
9929 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
9934 Download hardware frames to system memory.
9936 The input must be in hardware frames, and the output a non-hardware format.
9937 Not all formats will be supported on the output - it may be necessary to insert
9938 an additional @option{format} filter immediately following in the graph to get
9939 the output in a supported format.
9943 Map hardware frames to system memory or to another device.
9945 This filter has several different modes of operation; which one is used depends
9946 on the input and output formats:
9949 Hardware frame input, normal frame output
9951 Map the input frames to system memory and pass them to the output. If the
9952 original hardware frame is later required (for example, after overlaying
9953 something else on part of it), the @option{hwmap} filter can be used again
9954 in the next mode to retrieve it.
9956 Normal frame input, hardware frame output
9958 If the input is actually a software-mapped hardware frame, then unmap it -
9959 that is, return the original hardware frame.
9961 Otherwise, a device must be provided. Create new hardware surfaces on that
9962 device for the output, then map them back to the software format at the input
9963 and give those frames to the preceding filter. This will then act like the
9964 @option{hwupload} filter, but may be able to avoid an additional copy when
9965 the input is already in a compatible format.
9967 Hardware frame input and output
9969 A device must be supplied for the output, either directly or with the
9970 @option{derive_device} option. The input and output devices must be of
9971 different types and compatible - the exact meaning of this is
9972 system-dependent, but typically it means that they must refer to the same
9973 underlying hardware context (for example, refer to the same graphics card).
9975 If the input frames were originally created on the output device, then unmap
9976 to retrieve the original frames.
9978 Otherwise, map the frames to the output device - create new hardware frames
9979 on the output corresponding to the frames on the input.
9982 The following additional parameters are accepted:
9986 Set the frame mapping mode. Some combination of:
9989 The mapped frame should be readable.
9991 The mapped frame should be writeable.
9993 The mapping will always overwrite the entire frame.
9995 This may improve performance in some cases, as the original contents of the
9996 frame need not be loaded.
9998 The mapping must not involve any copying.
10000 Indirect mappings to copies of frames are created in some cases where either
10001 direct mapping is not possible or it would have unexpected properties.
10002 Setting this flag ensures that the mapping is direct and will fail if that is
10005 Defaults to @var{read+write} if not specified.
10007 @item derive_device @var{type}
10008 Rather than using the device supplied at initialisation, instead derive a new
10009 device of type @var{type} from the device the input frames exist on.
10012 In a hardware to hardware mapping, map in reverse - create frames in the sink
10013 and map them back to the source. This may be necessary in some cases where
10014 a mapping in one direction is required but only the opposite direction is
10015 supported by the devices being used.
10017 This option is dangerous - it may break the preceding filter in undefined
10018 ways if there are any additional constraints on that filter's output.
10019 Do not use it without fully understanding the implications of its use.
10024 Upload system memory frames to hardware surfaces.
10026 The device to upload to must be supplied when the filter is initialised. If
10027 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10030 @anchor{hwupload_cuda}
10031 @section hwupload_cuda
10033 Upload system memory frames to a CUDA device.
10035 It accepts the following optional parameters:
10039 The number of the CUDA device to use
10044 Apply a high-quality magnification filter designed for pixel art. This filter
10045 was originally created by Maxim Stepin.
10047 It accepts the following option:
10051 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10052 @code{hq3x} and @code{4} for @code{hq4x}.
10053 Default is @code{3}.
10057 Stack input videos horizontally.
10059 All streams must be of same pixel format and of same height.
10061 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10062 to create same output.
10064 The filter accept the following option:
10068 Set number of input streams. Default is 2.
10071 If set to 1, force the output to terminate when the shortest input
10072 terminates. Default value is 0.
10077 Modify the hue and/or the saturation of the input.
10079 It accepts the following parameters:
10083 Specify the hue angle as a number of degrees. It accepts an expression,
10084 and defaults to "0".
10087 Specify the saturation in the [-10,10] range. It accepts an expression and
10091 Specify the hue angle as a number of radians. It accepts an
10092 expression, and defaults to "0".
10095 Specify the brightness in the [-10,10] range. It accepts an expression and
10099 @option{h} and @option{H} are mutually exclusive, and can't be
10100 specified at the same time.
10102 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10103 expressions containing the following constants:
10107 frame count of the input frame starting from 0
10110 presentation timestamp of the input frame expressed in time base units
10113 frame rate of the input video, NAN if the input frame rate is unknown
10116 timestamp expressed in seconds, NAN if the input timestamp is unknown
10119 time base of the input video
10122 @subsection Examples
10126 Set the hue to 90 degrees and the saturation to 1.0:
10132 Same command but expressing the hue in radians:
10138 Rotate hue and make the saturation swing between 0
10139 and 2 over a period of 1 second:
10141 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10145 Apply a 3 seconds saturation fade-in effect starting at 0:
10147 hue="s=min(t/3\,1)"
10150 The general fade-in expression can be written as:
10152 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10156 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10158 hue="s=max(0\, min(1\, (8-t)/3))"
10161 The general fade-out expression can be written as:
10163 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10168 @subsection Commands
10170 This filter supports the following commands:
10176 Modify the hue and/or the saturation and/or brightness of the input video.
10177 The command accepts the same syntax of the corresponding option.
10179 If the specified expression is not valid, it is kept at its current
10183 @section hysteresis
10185 Grow first stream into second stream by connecting components.
10186 This makes it possible to build more robust edge masks.
10188 This filter accepts the following options:
10192 Set which planes will be processed as bitmap, unprocessed planes will be
10193 copied from first stream.
10194 By default value 0xf, all planes will be processed.
10197 Set threshold which is used in filtering. If pixel component value is higher than
10198 this value filter algorithm for connecting components is activated.
10199 By default value is 0.
10204 Detect video interlacing type.
10206 This filter tries to detect if the input frames are interlaced, progressive,
10207 top or bottom field first. It will also try to detect fields that are
10208 repeated between adjacent frames (a sign of telecine).
10210 Single frame detection considers only immediately adjacent frames when classifying each frame.
10211 Multiple frame detection incorporates the classification history of previous frames.
10213 The filter will log these metadata values:
10216 @item single.current_frame
10217 Detected type of current frame using single-frame detection. One of:
10218 ``tff'' (top field first), ``bff'' (bottom field first),
10219 ``progressive'', or ``undetermined''
10222 Cumulative number of frames detected as top field first using single-frame detection.
10225 Cumulative number of frames detected as top field first using multiple-frame detection.
10228 Cumulative number of frames detected as bottom field first using single-frame detection.
10230 @item multiple.current_frame
10231 Detected type of current frame using multiple-frame detection. One of:
10232 ``tff'' (top field first), ``bff'' (bottom field first),
10233 ``progressive'', or ``undetermined''
10236 Cumulative number of frames detected as bottom field first using multiple-frame detection.
10238 @item single.progressive
10239 Cumulative number of frames detected as progressive using single-frame detection.
10241 @item multiple.progressive
10242 Cumulative number of frames detected as progressive using multiple-frame detection.
10244 @item single.undetermined
10245 Cumulative number of frames that could not be classified using single-frame detection.
10247 @item multiple.undetermined
10248 Cumulative number of frames that could not be classified using multiple-frame detection.
10250 @item repeated.current_frame
10251 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
10253 @item repeated.neither
10254 Cumulative number of frames with no repeated field.
10257 Cumulative number of frames with the top field repeated from the previous frame's top field.
10259 @item repeated.bottom
10260 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
10263 The filter accepts the following options:
10267 Set interlacing threshold.
10269 Set progressive threshold.
10271 Threshold for repeated field detection.
10273 Number of frames after which a given frame's contribution to the
10274 statistics is halved (i.e., it contributes only 0.5 to its
10275 classification). The default of 0 means that all frames seen are given
10276 full weight of 1.0 forever.
10277 @item analyze_interlaced_flag
10278 When this is not 0 then idet will use the specified number of frames to determine
10279 if the interlaced flag is accurate, it will not count undetermined frames.
10280 If the flag is found to be accurate it will be used without any further
10281 computations, if it is found to be inaccurate it will be cleared without any
10282 further computations. This allows inserting the idet filter as a low computational
10283 method to clean up the interlaced flag
10288 Deinterleave or interleave fields.
10290 This filter allows one to process interlaced images fields without
10291 deinterlacing them. Deinterleaving splits the input frame into 2
10292 fields (so called half pictures). Odd lines are moved to the top
10293 half of the output image, even lines to the bottom half.
10294 You can process (filter) them independently and then re-interleave them.
10296 The filter accepts the following options:
10300 @item chroma_mode, c
10301 @item alpha_mode, a
10302 Available values for @var{luma_mode}, @var{chroma_mode} and
10303 @var{alpha_mode} are:
10309 @item deinterleave, d
10310 Deinterleave fields, placing one above the other.
10312 @item interleave, i
10313 Interleave fields. Reverse the effect of deinterleaving.
10315 Default value is @code{none}.
10317 @item luma_swap, ls
10318 @item chroma_swap, cs
10319 @item alpha_swap, as
10320 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
10325 Apply inflate effect to the video.
10327 This filter replaces the pixel by the local(3x3) average by taking into account
10328 only values higher than the pixel.
10330 It accepts the following options:
10337 Limit the maximum change for each plane, default is 65535.
10338 If 0, plane will remain unchanged.
10343 Simple interlacing filter from progressive contents. This interleaves upper (or
10344 lower) lines from odd frames with lower (or upper) lines from even frames,
10345 halving the frame rate and preserving image height.
10348 Original Original New Frame
10349 Frame 'j' Frame 'j+1' (tff)
10350 ========== =========== ==================
10351 Line 0 --------------------> Frame 'j' Line 0
10352 Line 1 Line 1 ----> Frame 'j+1' Line 1
10353 Line 2 ---------------------> Frame 'j' Line 2
10354 Line 3 Line 3 ----> Frame 'j+1' Line 3
10356 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
10359 It accepts the following optional parameters:
10363 This determines whether the interlaced frame is taken from the even
10364 (tff - default) or odd (bff) lines of the progressive frame.
10367 Vertical lowpass filter to avoid twitter interlacing and
10368 reduce moire patterns.
10372 Disable vertical lowpass filter
10375 Enable linear filter (default)
10378 Enable complex filter. This will slightly less reduce twitter and moire
10379 but better retain detail and subjective sharpness impression.
10386 Deinterlace input video by applying Donald Graft's adaptive kernel
10387 deinterling. Work on interlaced parts of a video to produce
10388 progressive frames.
10390 The description of the accepted parameters follows.
10394 Set the threshold which affects the filter's tolerance when
10395 determining if a pixel line must be processed. It must be an integer
10396 in the range [0,255] and defaults to 10. A value of 0 will result in
10397 applying the process on every pixels.
10400 Paint pixels exceeding the threshold value to white if set to 1.
10404 Set the fields order. Swap fields if set to 1, leave fields alone if
10408 Enable additional sharpening if set to 1. Default is 0.
10411 Enable twoway sharpening if set to 1. Default is 0.
10414 @subsection Examples
10418 Apply default values:
10420 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
10424 Enable additional sharpening:
10430 Paint processed pixels in white:
10436 @section lenscorrection
10438 Correct radial lens distortion
10440 This filter can be used to correct for radial distortion as can result from the use
10441 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
10442 one can use tools available for example as part of opencv or simply trial-and-error.
10443 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
10444 and extract the k1 and k2 coefficients from the resulting matrix.
10446 Note that effectively the same filter is available in the open-source tools Krita and
10447 Digikam from the KDE project.
10449 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
10450 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
10451 brightness distribution, so you may want to use both filters together in certain
10452 cases, though you will have to take care of ordering, i.e. whether vignetting should
10453 be applied before or after lens correction.
10455 @subsection Options
10457 The filter accepts the following options:
10461 Relative x-coordinate of the focal point of the image, and thereby the center of the
10462 distortion. This value has a range [0,1] and is expressed as fractions of the image
10463 width. Default is 0.5.
10465 Relative y-coordinate of the focal point of the image, and thereby the center of the
10466 distortion. This value has a range [0,1] and is expressed as fractions of the image
10467 height. Default is 0.5.
10469 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
10470 no correction. Default is 0.
10472 Coefficient of the double quadratic correction term. This value has a range [-1,1].
10473 0 means no correction. Default is 0.
10476 The formula that generates the correction is:
10478 @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)
10480 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
10481 distances from the focal point in the source and target images, respectively.
10485 Obtain the VMAF (Video Multi-Method Assessment Fusion)
10486 score between two input videos.
10488 The obtained VMAF score is printed through the logging system.
10490 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
10491 After installing the library it can be enabled using:
10492 @code{./configure --enable-libvmaf}.
10493 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
10495 The filter has following options:
10499 Set the model path which is to be used for SVM.
10500 Default value: @code{"vmaf_v0.6.1.pkl"}
10503 Set the file path to be used to store logs.
10506 Set the format of the log file (xml or json).
10508 @item enable_transform
10509 Enables transform for computing vmaf.
10512 Invokes the phone model which will generate VMAF scores higher than in the
10513 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
10516 Enables computing psnr along with vmaf.
10519 Enables computing ssim along with vmaf.
10522 Enables computing ms_ssim along with vmaf.
10525 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
10528 This filter also supports the @ref{framesync} options.
10530 On the below examples the input file @file{main.mpg} being processed is
10531 compared with the reference file @file{ref.mpg}.
10534 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
10537 Example with options:
10539 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:enable-transform=1" -f null -
10544 Limits the pixel components values to the specified range [min, max].
10546 The filter accepts the following options:
10550 Lower bound. Defaults to the lowest allowed value for the input.
10553 Upper bound. Defaults to the highest allowed value for the input.
10556 Specify which planes will be processed. Defaults to all available.
10563 The filter accepts the following options:
10567 Set the number of loops. Setting this value to -1 will result in infinite loops.
10571 Set maximal size in number of frames. Default is 0.
10574 Set first frame of loop. Default is 0.
10580 Apply a 3D LUT to an input video.
10582 The filter accepts the following options:
10586 Set the 3D LUT file name.
10588 Currently supported formats:
10600 Select interpolation mode.
10602 Available values are:
10606 Use values from the nearest defined point.
10608 Interpolate values using the 8 points defining a cube.
10610 Interpolate values using a tetrahedron.
10614 This filter also supports the @ref{framesync} options.
10618 Turn certain luma values into transparency.
10620 The filter accepts the following options:
10624 Set the luma which will be used as base for transparency.
10625 Default value is @code{0}.
10628 Set the range of luma values to be keyed out.
10629 Default value is @code{0}.
10632 Set the range of softness. Default value is @code{0}.
10633 Use this to control gradual transition from zero to full transparency.
10636 @section lut, lutrgb, lutyuv
10638 Compute a look-up table for binding each pixel component input value
10639 to an output value, and apply it to the input video.
10641 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
10642 to an RGB input video.
10644 These filters accept the following parameters:
10647 set first pixel component expression
10649 set second pixel component expression
10651 set third pixel component expression
10653 set fourth pixel component expression, corresponds to the alpha component
10656 set red component expression
10658 set green component expression
10660 set blue component expression
10662 alpha component expression
10665 set Y/luminance component expression
10667 set U/Cb component expression
10669 set V/Cr component expression
10672 Each of them specifies the expression to use for computing the lookup table for
10673 the corresponding pixel component values.
10675 The exact component associated to each of the @var{c*} options depends on the
10678 The @var{lut} filter requires either YUV or RGB pixel formats in input,
10679 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
10681 The expressions can contain the following constants and functions:
10686 The input width and height.
10689 The input value for the pixel component.
10692 The input value, clipped to the @var{minval}-@var{maxval} range.
10695 The maximum value for the pixel component.
10698 The minimum value for the pixel component.
10701 The negated value for the pixel component value, clipped to the
10702 @var{minval}-@var{maxval} range; it corresponds to the expression
10703 "maxval-clipval+minval".
10706 The computed value in @var{val}, clipped to the
10707 @var{minval}-@var{maxval} range.
10709 @item gammaval(gamma)
10710 The computed gamma correction value of the pixel component value,
10711 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
10713 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
10717 All expressions default to "val".
10719 @subsection Examples
10723 Negate input video:
10725 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
10726 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
10729 The above is the same as:
10731 lutrgb="r=negval:g=negval:b=negval"
10732 lutyuv="y=negval:u=negval:v=negval"
10742 Remove chroma components, turning the video into a graytone image:
10744 lutyuv="u=128:v=128"
10748 Apply a luma burning effect:
10754 Remove green and blue components:
10760 Set a constant alpha channel value on input:
10762 format=rgba,lutrgb=a="maxval-minval/2"
10766 Correct luminance gamma by a factor of 0.5:
10768 lutyuv=y=gammaval(0.5)
10772 Discard least significant bits of luma:
10774 lutyuv=y='bitand(val, 128+64+32)'
10778 Technicolor like effect:
10780 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
10784 @section lut2, tlut2
10786 The @code{lut2} filter takes two input streams and outputs one
10789 The @code{tlut2} (time lut2) filter takes two consecutive frames
10790 from one single stream.
10792 This filter accepts the following parameters:
10795 set first pixel component expression
10797 set second pixel component expression
10799 set third pixel component expression
10801 set fourth pixel component expression, corresponds to the alpha component
10804 Each of them specifies the expression to use for computing the lookup table for
10805 the corresponding pixel component values.
10807 The exact component associated to each of the @var{c*} options depends on the
10810 The expressions can contain the following constants:
10815 The input width and height.
10818 The first input value for the pixel component.
10821 The second input value for the pixel component.
10824 The first input video bit depth.
10827 The second input video bit depth.
10830 All expressions default to "x".
10832 @subsection Examples
10836 Highlight differences between two RGB video streams:
10838 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)'
10842 Highlight differences between two YUV video streams:
10844 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)'
10848 Show max difference between two video streams:
10850 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)))'
10854 @section maskedclamp
10856 Clamp the first input stream with the second input and third input stream.
10858 Returns the value of first stream to be between second input
10859 stream - @code{undershoot} and third input stream + @code{overshoot}.
10861 This filter accepts the following options:
10864 Default value is @code{0}.
10867 Default value is @code{0}.
10870 Set which planes will be processed as bitmap, unprocessed planes will be
10871 copied from first stream.
10872 By default value 0xf, all planes will be processed.
10875 @section maskedmerge
10877 Merge the first input stream with the second input stream using per pixel
10878 weights in the third input stream.
10880 A value of 0 in the third stream pixel component means that pixel component
10881 from first stream is returned unchanged, while maximum value (eg. 255 for
10882 8-bit videos) means that pixel component from second stream is returned
10883 unchanged. Intermediate values define the amount of merging between both
10884 input stream's pixel components.
10886 This filter accepts the following options:
10889 Set which planes will be processed as bitmap, unprocessed planes will be
10890 copied from first stream.
10891 By default value 0xf, all planes will be processed.
10896 Apply motion-compensation deinterlacing.
10898 It needs one field per frame as input and must thus be used together
10899 with yadif=1/3 or equivalent.
10901 This filter accepts the following options:
10904 Set the deinterlacing mode.
10906 It accepts one of the following values:
10911 use iterative motion estimation
10913 like @samp{slow}, but use multiple reference frames.
10915 Default value is @samp{fast}.
10918 Set the picture field parity assumed for the input video. It must be
10919 one of the following values:
10923 assume top field first
10925 assume bottom field first
10928 Default value is @samp{bff}.
10931 Set per-block quantization parameter (QP) used by the internal
10934 Higher values should result in a smoother motion vector field but less
10935 optimal individual vectors. Default value is 1.
10938 @section mergeplanes
10940 Merge color channel components from several video streams.
10942 The filter accepts up to 4 input streams, and merge selected input
10943 planes to the output video.
10945 This filter accepts the following options:
10948 Set input to output plane mapping. Default is @code{0}.
10950 The mappings is specified as a bitmap. It should be specified as a
10951 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
10952 mapping for the first plane of the output stream. 'A' sets the number of
10953 the input stream to use (from 0 to 3), and 'a' the plane number of the
10954 corresponding input to use (from 0 to 3). The rest of the mappings is
10955 similar, 'Bb' describes the mapping for the output stream second
10956 plane, 'Cc' describes the mapping for the output stream third plane and
10957 'Dd' describes the mapping for the output stream fourth plane.
10960 Set output pixel format. Default is @code{yuva444p}.
10963 @subsection Examples
10967 Merge three gray video streams of same width and height into single video stream:
10969 [a0][a1][a2]mergeplanes=0x001020:yuv444p
10973 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
10975 [a0][a1]mergeplanes=0x00010210:yuva444p
10979 Swap Y and A plane in yuva444p stream:
10981 format=yuva444p,mergeplanes=0x03010200:yuva444p
10985 Swap U and V plane in yuv420p stream:
10987 format=yuv420p,mergeplanes=0x000201:yuv420p
10991 Cast a rgb24 clip to yuv444p:
10993 format=rgb24,mergeplanes=0x000102:yuv444p
10999 Estimate and export motion vectors using block matching algorithms.
11000 Motion vectors are stored in frame side data to be used by other filters.
11002 This filter accepts the following options:
11005 Specify the motion estimation method. Accepts one of the following values:
11009 Exhaustive search algorithm.
11011 Three step search algorithm.
11013 Two dimensional logarithmic search algorithm.
11015 New three step search algorithm.
11017 Four step search algorithm.
11019 Diamond search algorithm.
11021 Hexagon-based search algorithm.
11023 Enhanced predictive zonal search algorithm.
11025 Uneven multi-hexagon search algorithm.
11027 Default value is @samp{esa}.
11030 Macroblock size. Default @code{16}.
11033 Search parameter. Default @code{7}.
11036 @section midequalizer
11038 Apply Midway Image Equalization effect using two video streams.
11040 Midway Image Equalization adjusts a pair of images to have the same
11041 histogram, while maintaining their dynamics as much as possible. It's
11042 useful for e.g. matching exposures from a pair of stereo cameras.
11044 This filter has two inputs and one output, which must be of same pixel format, but
11045 may be of different sizes. The output of filter is first input adjusted with
11046 midway histogram of both inputs.
11048 This filter accepts the following option:
11052 Set which planes to process. Default is @code{15}, which is all available planes.
11055 @section minterpolate
11057 Convert the video to specified frame rate using motion interpolation.
11059 This filter accepts the following options:
11062 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}.
11065 Motion interpolation mode. Following values are accepted:
11068 Duplicate previous or next frame for interpolating new ones.
11070 Blend source frames. Interpolated frame is mean of previous and next frames.
11072 Motion compensated interpolation. Following options are effective when this mode is selected:
11076 Motion compensation mode. Following values are accepted:
11079 Overlapped block motion compensation.
11081 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
11083 Default mode is @samp{obmc}.
11086 Motion estimation mode. Following values are accepted:
11089 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
11091 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
11093 Default mode is @samp{bilat}.
11096 The algorithm to be used for motion estimation. Following values are accepted:
11099 Exhaustive search algorithm.
11101 Three step search algorithm.
11103 Two dimensional logarithmic search algorithm.
11105 New three step search algorithm.
11107 Four step search algorithm.
11109 Diamond search algorithm.
11111 Hexagon-based search algorithm.
11113 Enhanced predictive zonal search algorithm.
11115 Uneven multi-hexagon search algorithm.
11117 Default algorithm is @samp{epzs}.
11120 Macroblock size. Default @code{16}.
11123 Motion estimation search parameter. Default @code{32}.
11126 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).
11131 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:
11134 Disable scene change detection.
11136 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
11138 Default method is @samp{fdiff}.
11140 @item scd_threshold
11141 Scene change detection threshold. Default is @code{5.0}.
11146 Mix several video input streams into one video stream.
11148 A description of the accepted options follows.
11152 The number of inputs. If unspecified, it defaults to 2.
11155 Specify weight of each input video stream as sequence.
11156 Each weight is separated by space.
11159 Specify scale, if it is set it will be multiplied with sum
11160 of each weight multiplied with pixel values to give final destination
11161 pixel value. By default @var{scale} is auto scaled to sum of weights.
11164 Specify how end of stream is determined.
11167 The duration of the longest input. (default)
11170 The duration of the shortest input.
11173 The duration of the first input.
11177 @section mpdecimate
11179 Drop frames that do not differ greatly from the previous frame in
11180 order to reduce frame rate.
11182 The main use of this filter is for very-low-bitrate encoding
11183 (e.g. streaming over dialup modem), but it could in theory be used for
11184 fixing movies that were inverse-telecined incorrectly.
11186 A description of the accepted options follows.
11190 Set the maximum number of consecutive frames which can be dropped (if
11191 positive), or the minimum interval between dropped frames (if
11192 negative). If the value is 0, the frame is dropped disregarding the
11193 number of previous sequentially dropped frames.
11195 Default value is 0.
11200 Set the dropping threshold values.
11202 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
11203 represent actual pixel value differences, so a threshold of 64
11204 corresponds to 1 unit of difference for each pixel, or the same spread
11205 out differently over the block.
11207 A frame is a candidate for dropping if no 8x8 blocks differ by more
11208 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
11209 meaning the whole image) differ by more than a threshold of @option{lo}.
11211 Default value for @option{hi} is 64*12, default value for @option{lo} is
11212 64*5, and default value for @option{frac} is 0.33.
11218 Negate input video.
11220 It accepts an integer in input; if non-zero it negates the
11221 alpha component (if available). The default value in input is 0.
11225 Denoise frames using Non-Local Means algorithm.
11227 Each pixel is adjusted by looking for other pixels with similar contexts. This
11228 context similarity is defined by comparing their surrounding patches of size
11229 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
11232 Note that the research area defines centers for patches, which means some
11233 patches will be made of pixels outside that research area.
11235 The filter accepts the following options.
11239 Set denoising strength.
11245 Same as @option{p} but for chroma planes.
11247 The default value is @var{0} and means automatic.
11253 Same as @option{r} but for chroma planes.
11255 The default value is @var{0} and means automatic.
11260 Deinterlace video using neural network edge directed interpolation.
11262 This filter accepts the following options:
11266 Mandatory option, without binary file filter can not work.
11267 Currently file can be found here:
11268 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
11271 Set which frames to deinterlace, by default it is @code{all}.
11272 Can be @code{all} or @code{interlaced}.
11275 Set mode of operation.
11277 Can be one of the following:
11281 Use frame flags, both fields.
11283 Use frame flags, single field.
11285 Use top field only.
11287 Use bottom field only.
11289 Use both fields, top first.
11291 Use both fields, bottom first.
11295 Set which planes to process, by default filter process all frames.
11298 Set size of local neighborhood around each pixel, used by the predictor neural
11301 Can be one of the following:
11314 Set the number of neurons in predictor neural network.
11315 Can be one of the following:
11326 Controls the number of different neural network predictions that are blended
11327 together to compute the final output value. Can be @code{fast}, default or
11331 Set which set of weights to use in the predictor.
11332 Can be one of the following:
11336 weights trained to minimize absolute error
11338 weights trained to minimize squared error
11342 Controls whether or not the prescreener neural network is used to decide
11343 which pixels should be processed by the predictor neural network and which
11344 can be handled by simple cubic interpolation.
11345 The prescreener is trained to know whether cubic interpolation will be
11346 sufficient for a pixel or whether it should be predicted by the predictor nn.
11347 The computational complexity of the prescreener nn is much less than that of
11348 the predictor nn. Since most pixels can be handled by cubic interpolation,
11349 using the prescreener generally results in much faster processing.
11350 The prescreener is pretty accurate, so the difference between using it and not
11351 using it is almost always unnoticeable.
11353 Can be one of the following:
11361 Default is @code{new}.
11364 Set various debugging flags.
11369 Force libavfilter not to use any of the specified pixel formats for the
11370 input to the next filter.
11372 It accepts the following parameters:
11376 A '|'-separated list of pixel format names, such as
11377 pix_fmts=yuv420p|monow|rgb24".
11381 @subsection Examples
11385 Force libavfilter to use a format different from @var{yuv420p} for the
11386 input to the vflip filter:
11388 noformat=pix_fmts=yuv420p,vflip
11392 Convert the input video to any of the formats not contained in the list:
11394 noformat=yuv420p|yuv444p|yuv410p
11400 Add noise on video input frame.
11402 The filter accepts the following options:
11410 Set noise seed for specific pixel component or all pixel components in case
11411 of @var{all_seed}. Default value is @code{123457}.
11413 @item all_strength, alls
11414 @item c0_strength, c0s
11415 @item c1_strength, c1s
11416 @item c2_strength, c2s
11417 @item c3_strength, c3s
11418 Set noise strength for specific pixel component or all pixel components in case
11419 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
11421 @item all_flags, allf
11422 @item c0_flags, c0f
11423 @item c1_flags, c1f
11424 @item c2_flags, c2f
11425 @item c3_flags, c3f
11426 Set pixel component flags or set flags for all components if @var{all_flags}.
11427 Available values for component flags are:
11430 averaged temporal noise (smoother)
11432 mix random noise with a (semi)regular pattern
11434 temporal noise (noise pattern changes between frames)
11436 uniform noise (gaussian otherwise)
11440 @subsection Examples
11442 Add temporal and uniform noise to input video:
11444 noise=alls=20:allf=t+u
11449 Normalize RGB video (aka histogram stretching, contrast stretching).
11450 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
11452 For each channel of each frame, the filter computes the input range and maps
11453 it linearly to the user-specified output range. The output range defaults
11454 to the full dynamic range from pure black to pure white.
11456 Temporal smoothing can be used on the input range to reduce flickering (rapid
11457 changes in brightness) caused when small dark or bright objects enter or leave
11458 the scene. This is similar to the auto-exposure (automatic gain control) on a
11459 video camera, and, like a video camera, it may cause a period of over- or
11460 under-exposure of the video.
11462 The R,G,B channels can be normalized independently, which may cause some
11463 color shifting, or linked together as a single channel, which prevents
11464 color shifting. Linked normalization preserves hue. Independent normalization
11465 does not, so it can be used to remove some color casts. Independent and linked
11466 normalization can be combined in any ratio.
11468 The normalize filter accepts the following options:
11473 Colors which define the output range. The minimum input value is mapped to
11474 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
11475 The defaults are black and white respectively. Specifying white for
11476 @var{blackpt} and black for @var{whitept} will give color-inverted,
11477 normalized video. Shades of grey can be used to reduce the dynamic range
11478 (contrast). Specifying saturated colors here can create some interesting
11482 The number of previous frames to use for temporal smoothing. The input range
11483 of each channel is smoothed using a rolling average over the current frame
11484 and the @var{smoothing} previous frames. The default is 0 (no temporal
11488 Controls the ratio of independent (color shifting) channel normalization to
11489 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
11490 independent. Defaults to 1.0 (fully independent).
11493 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
11494 expensive no-op. Defaults to 1.0 (full strength).
11498 @subsection Examples
11500 Stretch video contrast to use the full dynamic range, with no temporal
11501 smoothing; may flicker depending on the source content:
11503 normalize=blackpt=black:whitept=white:smoothing=0
11506 As above, but with 50 frames of temporal smoothing; flicker should be
11507 reduced, depending on the source content:
11509 normalize=blackpt=black:whitept=white:smoothing=50
11512 As above, but with hue-preserving linked channel normalization:
11514 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
11517 As above, but with half strength:
11519 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
11522 Map the darkest input color to red, the brightest input color to cyan:
11524 normalize=blackpt=red:whitept=cyan
11529 Pass the video source unchanged to the output.
11532 Optical Character Recognition
11534 This filter uses Tesseract for optical character recognition.
11536 It accepts the following options:
11540 Set datapath to tesseract data. Default is to use whatever was
11541 set at installation.
11544 Set language, default is "eng".
11547 Set character whitelist.
11550 Set character blacklist.
11553 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
11557 Apply a video transform using libopencv.
11559 To enable this filter, install the libopencv library and headers and
11560 configure FFmpeg with @code{--enable-libopencv}.
11562 It accepts the following parameters:
11567 The name of the libopencv filter to apply.
11569 @item filter_params
11570 The parameters to pass to the libopencv filter. If not specified, the default
11571 values are assumed.
11575 Refer to the official libopencv documentation for more precise
11577 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
11579 Several libopencv filters are supported; see the following subsections.
11584 Dilate an image by using a specific structuring element.
11585 It corresponds to the libopencv function @code{cvDilate}.
11587 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
11589 @var{struct_el} represents a structuring element, and has the syntax:
11590 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
11592 @var{cols} and @var{rows} represent the number of columns and rows of
11593 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
11594 point, and @var{shape} the shape for the structuring element. @var{shape}
11595 must be "rect", "cross", "ellipse", or "custom".
11597 If the value for @var{shape} is "custom", it must be followed by a
11598 string of the form "=@var{filename}". The file with name
11599 @var{filename} is assumed to represent a binary image, with each
11600 printable character corresponding to a bright pixel. When a custom
11601 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
11602 or columns and rows of the read file are assumed instead.
11604 The default value for @var{struct_el} is "3x3+0x0/rect".
11606 @var{nb_iterations} specifies the number of times the transform is
11607 applied to the image, and defaults to 1.
11611 # Use the default values
11614 # Dilate using a structuring element with a 5x5 cross, iterating two times
11615 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
11617 # Read the shape from the file diamond.shape, iterating two times.
11618 # The file diamond.shape may contain a pattern of characters like this
11624 # The specified columns and rows are ignored
11625 # but the anchor point coordinates are not
11626 ocv=dilate:0x0+2x2/custom=diamond.shape|2
11631 Erode an image by using a specific structuring element.
11632 It corresponds to the libopencv function @code{cvErode}.
11634 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
11635 with the same syntax and semantics as the @ref{dilate} filter.
11639 Smooth the input video.
11641 The filter takes the following parameters:
11642 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
11644 @var{type} is the type of smooth filter to apply, and must be one of
11645 the following values: "blur", "blur_no_scale", "median", "gaussian",
11646 or "bilateral". The default value is "gaussian".
11648 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
11649 depend on the smooth type. @var{param1} and
11650 @var{param2} accept integer positive values or 0. @var{param3} and
11651 @var{param4} accept floating point values.
11653 The default value for @var{param1} is 3. The default value for the
11654 other parameters is 0.
11656 These parameters correspond to the parameters assigned to the
11657 libopencv function @code{cvSmooth}.
11659 @section oscilloscope
11661 2D Video Oscilloscope.
11663 Useful to measure spatial impulse, step responses, chroma delays, etc.
11665 It accepts the following parameters:
11669 Set scope center x position.
11672 Set scope center y position.
11675 Set scope size, relative to frame diagonal.
11678 Set scope tilt/rotation.
11684 Set trace center x position.
11687 Set trace center y position.
11690 Set trace width, relative to width of frame.
11693 Set trace height, relative to height of frame.
11696 Set which components to trace. By default it traces first three components.
11699 Draw trace grid. By default is enabled.
11702 Draw some statistics. By default is enabled.
11705 Draw scope. By default is enabled.
11708 @subsection Examples
11712 Inspect full first row of video frame.
11714 oscilloscope=x=0.5:y=0:s=1
11718 Inspect full last row of video frame.
11720 oscilloscope=x=0.5:y=1:s=1
11724 Inspect full 5th line of video frame of height 1080.
11726 oscilloscope=x=0.5:y=5/1080:s=1
11730 Inspect full last column of video frame.
11732 oscilloscope=x=1:y=0.5:s=1:t=1
11740 Overlay one video on top of another.
11742 It takes two inputs and has one output. The first input is the "main"
11743 video on which the second input is overlaid.
11745 It accepts the following parameters:
11747 A description of the accepted options follows.
11752 Set the expression for the x and y coordinates of the overlaid video
11753 on the main video. Default value is "0" for both expressions. In case
11754 the expression is invalid, it is set to a huge value (meaning that the
11755 overlay will not be displayed within the output visible area).
11758 See @ref{framesync}.
11761 Set when the expressions for @option{x}, and @option{y} are evaluated.
11763 It accepts the following values:
11766 only evaluate expressions once during the filter initialization or
11767 when a command is processed
11770 evaluate expressions for each incoming frame
11773 Default value is @samp{frame}.
11776 See @ref{framesync}.
11779 Set the format for the output video.
11781 It accepts the following values:
11784 force YUV420 output
11787 force YUV422 output
11790 force YUV444 output
11793 force packed RGB output
11796 force planar RGB output
11799 automatically pick format
11802 Default value is @samp{yuv420}.
11805 See @ref{framesync}.
11808 Set format of alpha of the overlaid video, it can be @var{straight} or
11809 @var{premultiplied}. Default is @var{straight}.
11812 The @option{x}, and @option{y} expressions can contain the following
11818 The main input width and height.
11822 The overlay input width and height.
11826 The computed values for @var{x} and @var{y}. They are evaluated for
11831 horizontal and vertical chroma subsample values of the output
11832 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
11836 the number of input frame, starting from 0
11839 the position in the file of the input frame, NAN if unknown
11842 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
11846 This filter also supports the @ref{framesync} options.
11848 Note that the @var{n}, @var{pos}, @var{t} variables are available only
11849 when evaluation is done @emph{per frame}, and will evaluate to NAN
11850 when @option{eval} is set to @samp{init}.
11852 Be aware that frames are taken from each input video in timestamp
11853 order, hence, if their initial timestamps differ, it is a good idea
11854 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
11855 have them begin in the same zero timestamp, as the example for
11856 the @var{movie} filter does.
11858 You can chain together more overlays but you should test the
11859 efficiency of such approach.
11861 @subsection Commands
11863 This filter supports the following commands:
11867 Modify the x and y of the overlay input.
11868 The command accepts the same syntax of the corresponding option.
11870 If the specified expression is not valid, it is kept at its current
11874 @subsection Examples
11878 Draw the overlay at 10 pixels from the bottom right corner of the main
11881 overlay=main_w-overlay_w-10:main_h-overlay_h-10
11884 Using named options the example above becomes:
11886 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
11890 Insert a transparent PNG logo in the bottom left corner of the input,
11891 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
11893 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
11897 Insert 2 different transparent PNG logos (second logo on bottom
11898 right corner) using the @command{ffmpeg} tool:
11900 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
11904 Add a transparent color layer on top of the main video; @code{WxH}
11905 must specify the size of the main input to the overlay filter:
11907 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
11911 Play an original video and a filtered version (here with the deshake
11912 filter) side by side using the @command{ffplay} tool:
11914 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
11917 The above command is the same as:
11919 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
11923 Make a sliding overlay appearing from the left to the right top part of the
11924 screen starting since time 2:
11926 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
11930 Compose output by putting two input videos side to side:
11932 ffmpeg -i left.avi -i right.avi -filter_complex "
11933 nullsrc=size=200x100 [background];
11934 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
11935 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
11936 [background][left] overlay=shortest=1 [background+left];
11937 [background+left][right] overlay=shortest=1:x=100 [left+right]
11942 Mask 10-20 seconds of a video by applying the delogo filter to a section
11944 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
11945 -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]'
11950 Chain several overlays in cascade:
11952 nullsrc=s=200x200 [bg];
11953 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
11954 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
11955 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
11956 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
11957 [in3] null, [mid2] overlay=100:100 [out0]
11964 Apply Overcomplete Wavelet denoiser.
11966 The filter accepts the following options:
11972 Larger depth values will denoise lower frequency components more, but
11973 slow down filtering.
11975 Must be an int in the range 8-16, default is @code{8}.
11977 @item luma_strength, ls
11980 Must be a double value in the range 0-1000, default is @code{1.0}.
11982 @item chroma_strength, cs
11983 Set chroma strength.
11985 Must be a double value in the range 0-1000, default is @code{1.0}.
11991 Add paddings to the input image, and place the original input at the
11992 provided @var{x}, @var{y} coordinates.
11994 It accepts the following parameters:
11999 Specify an expression for the size of the output image with the
12000 paddings added. If the value for @var{width} or @var{height} is 0, the
12001 corresponding input size is used for the output.
12003 The @var{width} expression can reference the value set by the
12004 @var{height} expression, and vice versa.
12006 The default value of @var{width} and @var{height} is 0.
12010 Specify the offsets to place the input image at within the padded area,
12011 with respect to the top/left border of the output image.
12013 The @var{x} expression can reference the value set by the @var{y}
12014 expression, and vice versa.
12016 The default value of @var{x} and @var{y} is 0.
12018 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
12019 so the input image is centered on the padded area.
12022 Specify the color of the padded area. For the syntax of this option,
12023 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
12024 manual,ffmpeg-utils}.
12026 The default value of @var{color} is "black".
12029 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
12031 It accepts the following values:
12035 Only evaluate expressions once during the filter initialization or when
12036 a command is processed.
12039 Evaluate expressions for each incoming frame.
12043 Default value is @samp{init}.
12046 Pad to aspect instead to a resolution.
12050 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
12051 options are expressions containing the following constants:
12056 The input video width and height.
12060 These are the same as @var{in_w} and @var{in_h}.
12064 The output width and height (the size of the padded area), as
12065 specified by the @var{width} and @var{height} expressions.
12069 These are the same as @var{out_w} and @var{out_h}.
12073 The x and y offsets as specified by the @var{x} and @var{y}
12074 expressions, or NAN if not yet specified.
12077 same as @var{iw} / @var{ih}
12080 input sample aspect ratio
12083 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
12087 The horizontal and vertical chroma subsample values. For example for the
12088 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12091 @subsection Examples
12095 Add paddings with the color "violet" to the input video. The output video
12096 size is 640x480, and the top-left corner of the input video is placed at
12099 pad=640:480:0:40:violet
12102 The example above is equivalent to the following command:
12104 pad=width=640:height=480:x=0:y=40:color=violet
12108 Pad the input to get an output with dimensions increased by 3/2,
12109 and put the input video at the center of the padded area:
12111 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
12115 Pad the input to get a squared output with size equal to the maximum
12116 value between the input width and height, and put the input video at
12117 the center of the padded area:
12119 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
12123 Pad the input to get a final w/h ratio of 16:9:
12125 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
12129 In case of anamorphic video, in order to set the output display aspect
12130 correctly, it is necessary to use @var{sar} in the expression,
12131 according to the relation:
12133 (ih * X / ih) * sar = output_dar
12134 X = output_dar / sar
12137 Thus the previous example needs to be modified to:
12139 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
12143 Double the output size and put the input video in the bottom-right
12144 corner of the output padded area:
12146 pad="2*iw:2*ih:ow-iw:oh-ih"
12150 @anchor{palettegen}
12151 @section palettegen
12153 Generate one palette for a whole video stream.
12155 It accepts the following options:
12159 Set the maximum number of colors to quantize in the palette.
12160 Note: the palette will still contain 256 colors; the unused palette entries
12163 @item reserve_transparent
12164 Create a palette of 255 colors maximum and reserve the last one for
12165 transparency. Reserving the transparency color is useful for GIF optimization.
12166 If not set, the maximum of colors in the palette will be 256. You probably want
12167 to disable this option for a standalone image.
12170 @item transparency_color
12171 Set the color that will be used as background for transparency.
12174 Set statistics mode.
12176 It accepts the following values:
12179 Compute full frame histograms.
12181 Compute histograms only for the part that differs from previous frame. This
12182 might be relevant to give more importance to the moving part of your input if
12183 the background is static.
12185 Compute new histogram for each frame.
12188 Default value is @var{full}.
12191 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
12192 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
12193 color quantization of the palette. This information is also visible at
12194 @var{info} logging level.
12196 @subsection Examples
12200 Generate a representative palette of a given video using @command{ffmpeg}:
12202 ffmpeg -i input.mkv -vf palettegen palette.png
12206 @section paletteuse
12208 Use a palette to downsample an input video stream.
12210 The filter takes two inputs: one video stream and a palette. The palette must
12211 be a 256 pixels image.
12213 It accepts the following options:
12217 Select dithering mode. Available algorithms are:
12220 Ordered 8x8 bayer dithering (deterministic)
12222 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
12223 Note: this dithering is sometimes considered "wrong" and is included as a
12225 @item floyd_steinberg
12226 Floyd and Steingberg dithering (error diffusion)
12228 Frankie Sierra dithering v2 (error diffusion)
12230 Frankie Sierra dithering v2 "Lite" (error diffusion)
12233 Default is @var{sierra2_4a}.
12236 When @var{bayer} dithering is selected, this option defines the scale of the
12237 pattern (how much the crosshatch pattern is visible). A low value means more
12238 visible pattern for less banding, and higher value means less visible pattern
12239 at the cost of more banding.
12241 The option must be an integer value in the range [0,5]. Default is @var{2}.
12244 If set, define the zone to process
12248 Only the changing rectangle will be reprocessed. This is similar to GIF
12249 cropping/offsetting compression mechanism. This option can be useful for speed
12250 if only a part of the image is changing, and has use cases such as limiting the
12251 scope of the error diffusal @option{dither} to the rectangle that bounds the
12252 moving scene (it leads to more deterministic output if the scene doesn't change
12253 much, and as a result less moving noise and better GIF compression).
12256 Default is @var{none}.
12259 Take new palette for each output frame.
12261 @item alpha_threshold
12262 Sets the alpha threshold for transparency. Alpha values above this threshold
12263 will be treated as completely opaque, and values below this threshold will be
12264 treated as completely transparent.
12266 The option must be an integer value in the range [0,255]. Default is @var{128}.
12269 @subsection Examples
12273 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
12274 using @command{ffmpeg}:
12276 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
12280 @section perspective
12282 Correct perspective of video not recorded perpendicular to the screen.
12284 A description of the accepted parameters follows.
12295 Set coordinates expression for top left, top right, bottom left and bottom right corners.
12296 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
12297 If the @code{sense} option is set to @code{source}, then the specified points will be sent
12298 to the corners of the destination. If the @code{sense} option is set to @code{destination},
12299 then the corners of the source will be sent to the specified coordinates.
12301 The expressions can use the following variables:
12306 the width and height of video frame.
12310 Output frame count.
12313 @item interpolation
12314 Set interpolation for perspective correction.
12316 It accepts the following values:
12322 Default value is @samp{linear}.
12325 Set interpretation of coordinate options.
12327 It accepts the following values:
12331 Send point in the source specified by the given coordinates to
12332 the corners of the destination.
12334 @item 1, destination
12336 Send the corners of the source to the point in the destination specified
12337 by the given coordinates.
12339 Default value is @samp{source}.
12343 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
12345 It accepts the following values:
12348 only evaluate expressions once during the filter initialization or
12349 when a command is processed
12352 evaluate expressions for each incoming frame
12355 Default value is @samp{init}.
12360 Delay interlaced video by one field time so that the field order changes.
12362 The intended use is to fix PAL movies that have been captured with the
12363 opposite field order to the film-to-video transfer.
12365 A description of the accepted parameters follows.
12371 It accepts the following values:
12374 Capture field order top-first, transfer bottom-first.
12375 Filter will delay the bottom field.
12378 Capture field order bottom-first, transfer top-first.
12379 Filter will delay the top field.
12382 Capture and transfer with the same field order. This mode only exists
12383 for the documentation of the other options to refer to, but if you
12384 actually select it, the filter will faithfully do nothing.
12387 Capture field order determined automatically by field flags, transfer
12389 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
12390 basis using field flags. If no field information is available,
12391 then this works just like @samp{u}.
12394 Capture unknown or varying, transfer opposite.
12395 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
12396 analyzing the images and selecting the alternative that produces best
12397 match between the fields.
12400 Capture top-first, transfer unknown or varying.
12401 Filter selects among @samp{t} and @samp{p} using image analysis.
12404 Capture bottom-first, transfer unknown or varying.
12405 Filter selects among @samp{b} and @samp{p} using image analysis.
12408 Capture determined by field flags, transfer unknown or varying.
12409 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
12410 image analysis. If no field information is available, then this works just
12411 like @samp{U}. This is the default mode.
12414 Both capture and transfer unknown or varying.
12415 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
12419 @section pixdesctest
12421 Pixel format descriptor test filter, mainly useful for internal
12422 testing. The output video should be equal to the input video.
12426 format=monow, pixdesctest
12429 can be used to test the monowhite pixel format descriptor definition.
12433 Display sample values of color channels. Mainly useful for checking color
12434 and levels. Minimum supported resolution is 640x480.
12436 The filters accept the following options:
12440 Set scope X position, relative offset on X axis.
12443 Set scope Y position, relative offset on Y axis.
12452 Set window opacity. This window also holds statistics about pixel area.
12455 Set window X position, relative offset on X axis.
12458 Set window Y position, relative offset on Y axis.
12463 Enable the specified chain of postprocessing subfilters using libpostproc. This
12464 library should be automatically selected with a GPL build (@code{--enable-gpl}).
12465 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
12466 Each subfilter and some options have a short and a long name that can be used
12467 interchangeably, i.e. dr/dering are the same.
12469 The filters accept the following options:
12473 Set postprocessing subfilters string.
12476 All subfilters share common options to determine their scope:
12480 Honor the quality commands for this subfilter.
12483 Do chrominance filtering, too (default).
12486 Do luminance filtering only (no chrominance).
12489 Do chrominance filtering only (no luminance).
12492 These options can be appended after the subfilter name, separated by a '|'.
12494 Available subfilters are:
12497 @item hb/hdeblock[|difference[|flatness]]
12498 Horizontal deblocking filter
12501 Difference factor where higher values mean more deblocking (default: @code{32}).
12503 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12506 @item vb/vdeblock[|difference[|flatness]]
12507 Vertical deblocking filter
12510 Difference factor where higher values mean more deblocking (default: @code{32}).
12512 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12515 @item ha/hadeblock[|difference[|flatness]]
12516 Accurate horizontal deblocking filter
12519 Difference factor where higher values mean more deblocking (default: @code{32}).
12521 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12524 @item va/vadeblock[|difference[|flatness]]
12525 Accurate vertical deblocking filter
12528 Difference factor where higher values mean more deblocking (default: @code{32}).
12530 Flatness threshold where lower values mean more deblocking (default: @code{39}).
12534 The horizontal and vertical deblocking filters share the difference and
12535 flatness values so you cannot set different horizontal and vertical
12539 @item h1/x1hdeblock
12540 Experimental horizontal deblocking filter
12542 @item v1/x1vdeblock
12543 Experimental vertical deblocking filter
12548 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
12551 larger -> stronger filtering
12553 larger -> stronger filtering
12555 larger -> stronger filtering
12558 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
12561 Stretch luminance to @code{0-255}.
12564 @item lb/linblenddeint
12565 Linear blend deinterlacing filter that deinterlaces the given block by
12566 filtering all lines with a @code{(1 2 1)} filter.
12568 @item li/linipoldeint
12569 Linear interpolating deinterlacing filter that deinterlaces the given block by
12570 linearly interpolating every second line.
12572 @item ci/cubicipoldeint
12573 Cubic interpolating deinterlacing filter deinterlaces the given block by
12574 cubically interpolating every second line.
12576 @item md/mediandeint
12577 Median deinterlacing filter that deinterlaces the given block by applying a
12578 median filter to every second line.
12580 @item fd/ffmpegdeint
12581 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
12582 second line with a @code{(-1 4 2 4 -1)} filter.
12585 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
12586 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
12588 @item fq/forceQuant[|quantizer]
12589 Overrides the quantizer table from the input with the constant quantizer you
12597 Default pp filter combination (@code{hb|a,vb|a,dr|a})
12600 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
12603 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
12606 @subsection Examples
12610 Apply horizontal and vertical deblocking, deringing and automatic
12611 brightness/contrast:
12617 Apply default filters without brightness/contrast correction:
12623 Apply default filters and temporal denoiser:
12625 pp=default/tmpnoise|1|2|3
12629 Apply deblocking on luminance only, and switch vertical deblocking on or off
12630 automatically depending on available CPU time:
12637 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
12638 similar to spp = 6 with 7 point DCT, where only the center sample is
12641 The filter accepts the following options:
12645 Force a constant quantization parameter. It accepts an integer in range
12646 0 to 63. If not set, the filter will use the QP from the video stream
12650 Set thresholding mode. Available modes are:
12654 Set hard thresholding.
12656 Set soft thresholding (better de-ringing effect, but likely blurrier).
12658 Set medium thresholding (good results, default).
12662 @section premultiply
12663 Apply alpha premultiply effect to input video stream using first plane
12664 of second stream as alpha.
12666 Both streams must have same dimensions and same pixel format.
12668 The filter accepts the following option:
12672 Set which planes will be processed, unprocessed planes will be copied.
12673 By default value 0xf, all planes will be processed.
12676 Do not require 2nd input for processing, instead use alpha plane from input stream.
12680 Apply prewitt operator to input video stream.
12682 The filter accepts the following option:
12686 Set which planes will be processed, unprocessed planes will be copied.
12687 By default value 0xf, all planes will be processed.
12690 Set value which will be multiplied with filtered result.
12693 Set value which will be added to filtered result.
12696 @anchor{program_opencl}
12697 @section program_opencl
12699 Filter video using an OpenCL program.
12704 OpenCL program source file.
12707 Kernel name in program.
12710 Number of inputs to the filter. Defaults to 1.
12713 Size of output frames. Defaults to the same as the first input.
12717 The program source file must contain a kernel function with the given name,
12718 which will be run once for each plane of the output. Each run on a plane
12719 gets enqueued as a separate 2D global NDRange with one work-item for each
12720 pixel to be generated. The global ID offset for each work-item is therefore
12721 the coordinates of a pixel in the destination image.
12723 The kernel function needs to take the following arguments:
12726 Destination image, @var{__write_only image2d_t}.
12728 This image will become the output; the kernel should write all of it.
12730 Frame index, @var{unsigned int}.
12732 This is a counter starting from zero and increasing by one for each frame.
12734 Source images, @var{__read_only image2d_t}.
12736 These are the most recent images on each input. The kernel may read from
12737 them to generate the output, but they can't be written to.
12744 Copy the input to the output (output must be the same size as the input).
12746 __kernel void copy(__write_only image2d_t destination,
12747 unsigned int index,
12748 __read_only image2d_t source)
12750 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
12752 int2 location = (int2)(get_global_id(0), get_global_id(1));
12754 float4 value = read_imagef(source, sampler, location);
12756 write_imagef(destination, location, value);
12761 Apply a simple transformation, rotating the input by an amount increasing
12762 with the index counter. Pixel values are linearly interpolated by the
12763 sampler, and the output need not have the same dimensions as the input.
12765 __kernel void rotate_image(__write_only image2d_t dst,
12766 unsigned int index,
12767 __read_only image2d_t src)
12769 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12770 CLK_FILTER_LINEAR);
12772 float angle = (float)index / 100.0f;
12774 float2 dst_dim = convert_float2(get_image_dim(dst));
12775 float2 src_dim = convert_float2(get_image_dim(src));
12777 float2 dst_cen = dst_dim / 2.0f;
12778 float2 src_cen = src_dim / 2.0f;
12780 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12782 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
12784 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
12785 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
12787 src_pos = src_pos * src_dim / dst_dim;
12789 float2 src_loc = src_pos + src_cen;
12791 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
12792 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
12793 write_imagef(dst, dst_loc, 0.5f);
12795 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
12800 Blend two inputs together, with the amount of each input used varying
12801 with the index counter.
12803 __kernel void blend_images(__write_only image2d_t dst,
12804 unsigned int index,
12805 __read_only image2d_t src1,
12806 __read_only image2d_t src2)
12808 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
12809 CLK_FILTER_LINEAR);
12811 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
12813 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
12814 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
12815 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
12817 float4 val1 = read_imagef(src1, sampler, src1_loc);
12818 float4 val2 = read_imagef(src2, sampler, src2_loc);
12820 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
12826 @section pseudocolor
12828 Alter frame colors in video with pseudocolors.
12830 This filter accept the following options:
12834 set pixel first component expression
12837 set pixel second component expression
12840 set pixel third component expression
12843 set pixel fourth component expression, corresponds to the alpha component
12846 set component to use as base for altering colors
12849 Each of them specifies the expression to use for computing the lookup table for
12850 the corresponding pixel component values.
12852 The expressions can contain the following constants and functions:
12857 The input width and height.
12860 The input value for the pixel component.
12862 @item ymin, umin, vmin, amin
12863 The minimum allowed component value.
12865 @item ymax, umax, vmax, amax
12866 The maximum allowed component value.
12869 All expressions default to "val".
12871 @subsection Examples
12875 Change too high luma values to gradient:
12877 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'"
12883 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
12884 Ratio) between two input videos.
12886 This filter takes in input two input videos, the first input is
12887 considered the "main" source and is passed unchanged to the
12888 output. The second input is used as a "reference" video for computing
12891 Both video inputs must have the same resolution and pixel format for
12892 this filter to work correctly. Also it assumes that both inputs
12893 have the same number of frames, which are compared one by one.
12895 The obtained average PSNR is printed through the logging system.
12897 The filter stores the accumulated MSE (mean squared error) of each
12898 frame, and at the end of the processing it is averaged across all frames
12899 equally, and the following formula is applied to obtain the PSNR:
12902 PSNR = 10*log10(MAX^2/MSE)
12905 Where MAX is the average of the maximum values of each component of the
12908 The description of the accepted parameters follows.
12911 @item stats_file, f
12912 If specified the filter will use the named file to save the PSNR of
12913 each individual frame. When filename equals "-" the data is sent to
12916 @item stats_version
12917 Specifies which version of the stats file format to use. Details of
12918 each format are written below.
12919 Default value is 1.
12921 @item stats_add_max
12922 Determines whether the max value is output to the stats log.
12923 Default value is 0.
12924 Requires stats_version >= 2. If this is set and stats_version < 2,
12925 the filter will return an error.
12928 This filter also supports the @ref{framesync} options.
12930 The file printed if @var{stats_file} is selected, contains a sequence of
12931 key/value pairs of the form @var{key}:@var{value} for each compared
12934 If a @var{stats_version} greater than 1 is specified, a header line precedes
12935 the list of per-frame-pair stats, with key value pairs following the frame
12936 format with the following parameters:
12939 @item psnr_log_version
12940 The version of the log file format. Will match @var{stats_version}.
12943 A comma separated list of the per-frame-pair parameters included in
12947 A description of each shown per-frame-pair parameter follows:
12951 sequential number of the input frame, starting from 1
12954 Mean Square Error pixel-by-pixel average difference of the compared
12955 frames, averaged over all the image components.
12957 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
12958 Mean Square Error pixel-by-pixel average difference of the compared
12959 frames for the component specified by the suffix.
12961 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
12962 Peak Signal to Noise ratio of the compared frames for the component
12963 specified by the suffix.
12965 @item max_avg, max_y, max_u, max_v
12966 Maximum allowed value for each channel, and average over all
12972 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12973 [main][ref] psnr="stats_file=stats.log" [out]
12976 On this example the input file being processed is compared with the
12977 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
12978 is stored in @file{stats.log}.
12983 Pulldown reversal (inverse telecine) filter, capable of handling mixed
12984 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
12987 The pullup filter is designed to take advantage of future context in making
12988 its decisions. This filter is stateless in the sense that it does not lock
12989 onto a pattern to follow, but it instead looks forward to the following
12990 fields in order to identify matches and rebuild progressive frames.
12992 To produce content with an even framerate, insert the fps filter after
12993 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
12994 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
12996 The filter accepts the following options:
13003 These options set the amount of "junk" to ignore at the left, right, top, and
13004 bottom of the image, respectively. Left and right are in units of 8 pixels,
13005 while top and bottom are in units of 2 lines.
13006 The default is 8 pixels on each side.
13009 Set the strict breaks. Setting this option to 1 will reduce the chances of
13010 filter generating an occasional mismatched frame, but it may also cause an
13011 excessive number of frames to be dropped during high motion sequences.
13012 Conversely, setting it to -1 will make filter match fields more easily.
13013 This may help processing of video where there is slight blurring between
13014 the fields, but may also cause there to be interlaced frames in the output.
13015 Default value is @code{0}.
13018 Set the metric plane to use. It accepts the following values:
13024 Use chroma blue plane.
13027 Use chroma red plane.
13030 This option may be set to use chroma plane instead of the default luma plane
13031 for doing filter's computations. This may improve accuracy on very clean
13032 source material, but more likely will decrease accuracy, especially if there
13033 is chroma noise (rainbow effect) or any grayscale video.
13034 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
13035 load and make pullup usable in realtime on slow machines.
13038 For best results (without duplicated frames in the output file) it is
13039 necessary to change the output frame rate. For example, to inverse
13040 telecine NTSC input:
13042 ffmpeg -i input -vf pullup -r 24000/1001 ...
13047 Change video quantization parameters (QP).
13049 The filter accepts the following option:
13053 Set expression for quantization parameter.
13056 The expression is evaluated through the eval API and can contain, among others,
13057 the following constants:
13061 1 if index is not 129, 0 otherwise.
13064 Sequential index starting from -129 to 128.
13067 @subsection Examples
13071 Some equation like:
13079 Flush video frames from internal cache of frames into a random order.
13080 No frame is discarded.
13081 Inspired by @ref{frei0r} nervous filter.
13085 Set size in number of frames of internal cache, in range from @code{2} to
13086 @code{512}. Default is @code{30}.
13089 Set seed for random number generator, must be an integer included between
13090 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
13091 less than @code{0}, the filter will try to use a good random seed on a
13095 @section readeia608
13097 Read closed captioning (EIA-608) information from the top lines of a video frame.
13099 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
13100 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
13101 with EIA-608 data (starting from 0). A description of each metadata value follows:
13104 @item lavfi.readeia608.X.cc
13105 The two bytes stored as EIA-608 data (printed in hexadecimal).
13107 @item lavfi.readeia608.X.line
13108 The number of the line on which the EIA-608 data was identified and read.
13111 This filter accepts the following options:
13115 Set the line to start scanning for EIA-608 data. Default is @code{0}.
13118 Set the line to end scanning for EIA-608 data. Default is @code{29}.
13121 Set minimal acceptable amplitude change for sync codes detection.
13122 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
13125 Set the ratio of width reserved for sync code detection.
13126 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
13129 Set the max peaks height difference for sync code detection.
13130 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13133 Set max peaks period difference for sync code detection.
13134 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
13137 Set the first two max start code bits differences.
13138 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
13141 Set the minimum ratio of bits height compared to 3rd start code bit.
13142 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
13145 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
13148 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
13151 Enable checking the parity bit. In the event of a parity error, the filter will output
13152 @code{0x00} for that character. Default is false.
13155 @subsection Examples
13159 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
13161 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
13167 Read vertical interval timecode (VITC) information from the top lines of a
13170 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
13171 timecode value, if a valid timecode has been detected. Further metadata key
13172 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
13173 timecode data has been found or not.
13175 This filter accepts the following options:
13179 Set the maximum number of lines to scan for VITC data. If the value is set to
13180 @code{-1} the full video frame is scanned. Default is @code{45}.
13183 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
13184 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
13187 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
13188 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
13191 @subsection Examples
13195 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
13196 draw @code{--:--:--:--} as a placeholder:
13198 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
13204 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
13206 Destination pixel at position (X, Y) will be picked from source (x, y) position
13207 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
13208 value for pixel will be used for destination pixel.
13210 Xmap and Ymap input video streams must be of same dimensions. Output video stream
13211 will have Xmap/Ymap video stream dimensions.
13212 Xmap and Ymap input video streams are 16bit depth, single channel.
13214 @section removegrain
13216 The removegrain filter is a spatial denoiser for progressive video.
13220 Set mode for the first plane.
13223 Set mode for the second plane.
13226 Set mode for the third plane.
13229 Set mode for the fourth plane.
13232 Range of mode is from 0 to 24. Description of each mode follows:
13236 Leave input plane unchanged. Default.
13239 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
13242 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
13245 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
13248 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
13249 This is equivalent to a median filter.
13252 Line-sensitive clipping giving the minimal change.
13255 Line-sensitive clipping, intermediate.
13258 Line-sensitive clipping, intermediate.
13261 Line-sensitive clipping, intermediate.
13264 Line-sensitive clipping on a line where the neighbours pixels are the closest.
13267 Replaces the target pixel with the closest neighbour.
13270 [1 2 1] horizontal and vertical kernel blur.
13276 Bob mode, interpolates top field from the line where the neighbours
13277 pixels are the closest.
13280 Bob mode, interpolates bottom field from the line where the neighbours
13281 pixels are the closest.
13284 Bob mode, interpolates top field. Same as 13 but with a more complicated
13285 interpolation formula.
13288 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
13289 interpolation formula.
13292 Clips the pixel with the minimum and maximum of respectively the maximum and
13293 minimum of each pair of opposite neighbour pixels.
13296 Line-sensitive clipping using opposite neighbours whose greatest distance from
13297 the current pixel is minimal.
13300 Replaces the pixel with the average of its 8 neighbours.
13303 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
13306 Clips pixels using the averages of opposite neighbour.
13309 Same as mode 21 but simpler and faster.
13312 Small edge and halo removal, but reputed useless.
13318 @section removelogo
13320 Suppress a TV station logo, using an image file to determine which
13321 pixels comprise the logo. It works by filling in the pixels that
13322 comprise the logo with neighboring pixels.
13324 The filter accepts the following options:
13328 Set the filter bitmap file, which can be any image format supported by
13329 libavformat. The width and height of the image file must match those of the
13330 video stream being processed.
13333 Pixels in the provided bitmap image with a value of zero are not
13334 considered part of the logo, non-zero pixels are considered part of
13335 the logo. If you use white (255) for the logo and black (0) for the
13336 rest, you will be safe. For making the filter bitmap, it is
13337 recommended to take a screen capture of a black frame with the logo
13338 visible, and then using a threshold filter followed by the erode
13339 filter once or twice.
13341 If needed, little splotches can be fixed manually. Remember that if
13342 logo pixels are not covered, the filter quality will be much
13343 reduced. Marking too many pixels as part of the logo does not hurt as
13344 much, but it will increase the amount of blurring needed to cover over
13345 the image and will destroy more information than necessary, and extra
13346 pixels will slow things down on a large logo.
13348 @section repeatfields
13350 This filter uses the repeat_field flag from the Video ES headers and hard repeats
13351 fields based on its value.
13355 Reverse a video clip.
13357 Warning: This filter requires memory to buffer the entire clip, so trimming
13360 @subsection Examples
13364 Take the first 5 seconds of a clip, and reverse it.
13371 Apply roberts cross operator to input video stream.
13373 The filter accepts the following option:
13377 Set which planes will be processed, unprocessed planes will be copied.
13378 By default value 0xf, all planes will be processed.
13381 Set value which will be multiplied with filtered result.
13384 Set value which will be added to filtered result.
13389 Rotate video by an arbitrary angle expressed in radians.
13391 The filter accepts the following options:
13393 A description of the optional parameters follows.
13396 Set an expression for the angle by which to rotate the input video
13397 clockwise, expressed as a number of radians. A negative value will
13398 result in a counter-clockwise rotation. By default it is set to "0".
13400 This expression is evaluated for each frame.
13403 Set the output width expression, default value is "iw".
13404 This expression is evaluated just once during configuration.
13407 Set the output height expression, default value is "ih".
13408 This expression is evaluated just once during configuration.
13411 Enable bilinear interpolation if set to 1, a value of 0 disables
13412 it. Default value is 1.
13415 Set the color used to fill the output area not covered by the rotated
13416 image. For the general syntax of this option, check the
13417 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
13418 If the special value "none" is selected then no
13419 background is printed (useful for example if the background is never shown).
13421 Default value is "black".
13424 The expressions for the angle and the output size can contain the
13425 following constants and functions:
13429 sequential number of the input frame, starting from 0. It is always NAN
13430 before the first frame is filtered.
13433 time in seconds of the input frame, it is set to 0 when the filter is
13434 configured. It is always NAN before the first frame is filtered.
13438 horizontal and vertical chroma subsample values. For example for the
13439 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13443 the input video width and height
13447 the output width and height, that is the size of the padded area as
13448 specified by the @var{width} and @var{height} expressions
13452 the minimal width/height required for completely containing the input
13453 video rotated by @var{a} radians.
13455 These are only available when computing the @option{out_w} and
13456 @option{out_h} expressions.
13459 @subsection Examples
13463 Rotate the input by PI/6 radians clockwise:
13469 Rotate the input by PI/6 radians counter-clockwise:
13475 Rotate the input by 45 degrees clockwise:
13481 Apply a constant rotation with period T, starting from an angle of PI/3:
13483 rotate=PI/3+2*PI*t/T
13487 Make the input video rotation oscillating with a period of T
13488 seconds and an amplitude of A radians:
13490 rotate=A*sin(2*PI/T*t)
13494 Rotate the video, output size is chosen so that the whole rotating
13495 input video is always completely contained in the output:
13497 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
13501 Rotate the video, reduce the output size so that no background is ever
13504 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
13508 @subsection Commands
13510 The filter supports the following commands:
13514 Set the angle expression.
13515 The command accepts the same syntax of the corresponding option.
13517 If the specified expression is not valid, it is kept at its current
13523 Apply Shape Adaptive Blur.
13525 The filter accepts the following options:
13528 @item luma_radius, lr
13529 Set luma blur filter strength, must be a value in range 0.1-4.0, default
13530 value is 1.0. A greater value will result in a more blurred image, and
13531 in slower processing.
13533 @item luma_pre_filter_radius, lpfr
13534 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
13537 @item luma_strength, ls
13538 Set luma maximum difference between pixels to still be considered, must
13539 be a value in the 0.1-100.0 range, default value is 1.0.
13541 @item chroma_radius, cr
13542 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
13543 greater value will result in a more blurred image, and in slower
13546 @item chroma_pre_filter_radius, cpfr
13547 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
13549 @item chroma_strength, cs
13550 Set chroma maximum difference between pixels to still be considered,
13551 must be a value in the -0.9-100.0 range.
13554 Each chroma option value, if not explicitly specified, is set to the
13555 corresponding luma option value.
13560 Scale (resize) the input video, using the libswscale library.
13562 The scale filter forces the output display aspect ratio to be the same
13563 of the input, by changing the output sample aspect ratio.
13565 If the input image format is different from the format requested by
13566 the next filter, the scale filter will convert the input to the
13569 @subsection Options
13570 The filter accepts the following options, or any of the options
13571 supported by the libswscale scaler.
13573 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
13574 the complete list of scaler options.
13579 Set the output video dimension expression. Default value is the input
13582 If the @var{width} or @var{w} value is 0, the input width is used for
13583 the output. If the @var{height} or @var{h} value is 0, the input height
13584 is used for the output.
13586 If one and only one of the values is -n with n >= 1, the scale filter
13587 will use a value that maintains the aspect ratio of the input image,
13588 calculated from the other specified dimension. After that it will,
13589 however, make sure that the calculated dimension is divisible by n and
13590 adjust the value if necessary.
13592 If both values are -n with n >= 1, the behavior will be identical to
13593 both values being set to 0 as previously detailed.
13595 See below for the list of accepted constants for use in the dimension
13599 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
13603 Only evaluate expressions once during the filter initialization or when a command is processed.
13606 Evaluate expressions for each incoming frame.
13610 Default value is @samp{init}.
13614 Set the interlacing mode. It accepts the following values:
13618 Force interlaced aware scaling.
13621 Do not apply interlaced scaling.
13624 Select interlaced aware scaling depending on whether the source frames
13625 are flagged as interlaced or not.
13628 Default value is @samp{0}.
13631 Set libswscale scaling flags. See
13632 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13633 complete list of values. If not explicitly specified the filter applies
13637 @item param0, param1
13638 Set libswscale input parameters for scaling algorithms that need them. See
13639 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
13640 complete documentation. If not explicitly specified the filter applies
13646 Set the video size. For the syntax of this option, check the
13647 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13649 @item in_color_matrix
13650 @item out_color_matrix
13651 Set in/output YCbCr color space type.
13653 This allows the autodetected value to be overridden as well as allows forcing
13654 a specific value used for the output and encoder.
13656 If not specified, the color space type depends on the pixel format.
13662 Choose automatically.
13665 Format conforming to International Telecommunication Union (ITU)
13666 Recommendation BT.709.
13669 Set color space conforming to the United States Federal Communications
13670 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
13673 Set color space conforming to:
13677 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
13680 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
13683 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
13688 Set color space conforming to SMPTE ST 240:1999.
13693 Set in/output YCbCr sample range.
13695 This allows the autodetected value to be overridden as well as allows forcing
13696 a specific value used for the output and encoder. If not specified, the
13697 range depends on the pixel format. Possible values:
13701 Choose automatically.
13704 Set full range (0-255 in case of 8-bit luma).
13706 @item mpeg/limited/tv
13707 Set "MPEG" range (16-235 in case of 8-bit luma).
13710 @item force_original_aspect_ratio
13711 Enable decreasing or increasing output video width or height if necessary to
13712 keep the original aspect ratio. Possible values:
13716 Scale the video as specified and disable this feature.
13719 The output video dimensions will automatically be decreased if needed.
13722 The output video dimensions will automatically be increased if needed.
13726 One useful instance of this option is that when you know a specific device's
13727 maximum allowed resolution, you can use this to limit the output video to
13728 that, while retaining the aspect ratio. For example, device A allows
13729 1280x720 playback, and your video is 1920x800. Using this option (set it to
13730 decrease) and specifying 1280x720 to the command line makes the output
13733 Please note that this is a different thing than specifying -1 for @option{w}
13734 or @option{h}, you still need to specify the output resolution for this option
13739 The values of the @option{w} and @option{h} options are expressions
13740 containing the following constants:
13745 The input width and height
13749 These are the same as @var{in_w} and @var{in_h}.
13753 The output (scaled) width and height
13757 These are the same as @var{out_w} and @var{out_h}
13760 The same as @var{iw} / @var{ih}
13763 input sample aspect ratio
13766 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
13770 horizontal and vertical input chroma subsample values. For example for the
13771 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13775 horizontal and vertical output chroma subsample values. For example for the
13776 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13779 @subsection Examples
13783 Scale the input video to a size of 200x100
13788 This is equivalent to:
13799 Specify a size abbreviation for the output size:
13804 which can also be written as:
13810 Scale the input to 2x:
13812 scale=w=2*iw:h=2*ih
13816 The above is the same as:
13818 scale=2*in_w:2*in_h
13822 Scale the input to 2x with forced interlaced scaling:
13824 scale=2*iw:2*ih:interl=1
13828 Scale the input to half size:
13830 scale=w=iw/2:h=ih/2
13834 Increase the width, and set the height to the same size:
13840 Seek Greek harmony:
13847 Increase the height, and set the width to 3/2 of the height:
13849 scale=w=3/2*oh:h=3/5*ih
13853 Increase the size, making the size a multiple of the chroma
13856 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
13860 Increase the width to a maximum of 500 pixels,
13861 keeping the same aspect ratio as the input:
13863 scale=w='min(500\, iw*3/2):h=-1'
13867 Make pixels square by combining scale and setsar:
13869 scale='trunc(ih*dar):ih',setsar=1/1
13873 Make pixels square by combining scale and setsar,
13874 making sure the resulting resolution is even (required by some codecs):
13876 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
13880 @subsection Commands
13882 This filter supports the following commands:
13886 Set the output video dimension expression.
13887 The command accepts the same syntax of the corresponding option.
13889 If the specified expression is not valid, it is kept at its current
13895 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
13896 format conversion on CUDA video frames. Setting the output width and height
13897 works in the same way as for the @var{scale} filter.
13899 The following additional options are accepted:
13902 The pixel format of the output CUDA frames. If set to the string "same" (the
13903 default), the input format will be kept. Note that automatic format negotiation
13904 and conversion is not yet supported for hardware frames
13907 The interpolation algorithm used for resizing. One of the following:
13914 @item cubic2p_bspline
13915 2-parameter cubic (B=1, C=0)
13917 @item cubic2p_catmullrom
13918 2-parameter cubic (B=0, C=1/2)
13920 @item cubic2p_b05c03
13921 2-parameter cubic (B=1/2, C=3/10)
13933 Scale (resize) the input video, based on a reference video.
13935 See the scale filter for available options, scale2ref supports the same but
13936 uses the reference video instead of the main input as basis. scale2ref also
13937 supports the following additional constants for the @option{w} and
13938 @option{h} options:
13943 The main input video's width and height
13946 The same as @var{main_w} / @var{main_h}
13949 The main input video's sample aspect ratio
13951 @item main_dar, mdar
13952 The main input video's display aspect ratio. Calculated from
13953 @code{(main_w / main_h) * main_sar}.
13957 The main input video's horizontal and vertical chroma subsample values.
13958 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
13962 @subsection Examples
13966 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
13968 'scale2ref[b][a];[a][b]overlay'
13972 @anchor{selectivecolor}
13973 @section selectivecolor
13975 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
13976 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
13977 by the "purity" of the color (that is, how saturated it already is).
13979 This filter is similar to the Adobe Photoshop Selective Color tool.
13981 The filter accepts the following options:
13984 @item correction_method
13985 Select color correction method.
13987 Available values are:
13990 Specified adjustments are applied "as-is" (added/subtracted to original pixel
13993 Specified adjustments are relative to the original component value.
13995 Default is @code{absolute}.
13997 Adjustments for red pixels (pixels where the red component is the maximum)
13999 Adjustments for yellow pixels (pixels where the blue component is the minimum)
14001 Adjustments for green pixels (pixels where the green component is the maximum)
14003 Adjustments for cyan pixels (pixels where the red component is the minimum)
14005 Adjustments for blue pixels (pixels where the blue component is the maximum)
14007 Adjustments for magenta pixels (pixels where the green component is the minimum)
14009 Adjustments for white pixels (pixels where all components are greater than 128)
14011 Adjustments for all pixels except pure black and pure white
14013 Adjustments for black pixels (pixels where all components are lesser than 128)
14015 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
14018 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
14019 4 space separated floating point adjustment values in the [-1,1] range,
14020 respectively to adjust the amount of cyan, magenta, yellow and black for the
14021 pixels of its range.
14023 @subsection Examples
14027 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
14028 increase magenta by 27% in blue areas:
14030 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
14034 Use a Photoshop selective color preset:
14036 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
14040 @anchor{separatefields}
14041 @section separatefields
14043 The @code{separatefields} takes a frame-based video input and splits
14044 each frame into its components fields, producing a new half height clip
14045 with twice the frame rate and twice the frame count.
14047 This filter use field-dominance information in frame to decide which
14048 of each pair of fields to place first in the output.
14049 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
14051 @section setdar, setsar
14053 The @code{setdar} filter sets the Display Aspect Ratio for the filter
14056 This is done by changing the specified Sample (aka Pixel) Aspect
14057 Ratio, according to the following equation:
14059 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
14062 Keep in mind that the @code{setdar} filter does not modify the pixel
14063 dimensions of the video frame. Also, the display aspect ratio set by
14064 this filter may be changed by later filters in the filterchain,
14065 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
14068 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
14069 the filter output video.
14071 Note that as a consequence of the application of this filter, the
14072 output display aspect ratio will change according to the equation
14075 Keep in mind that the sample aspect ratio set by the @code{setsar}
14076 filter may be changed by later filters in the filterchain, e.g. if
14077 another "setsar" or a "setdar" filter is applied.
14079 It accepts the following parameters:
14082 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
14083 Set the aspect ratio used by the filter.
14085 The parameter can be a floating point number string, an expression, or
14086 a string of the form @var{num}:@var{den}, where @var{num} and
14087 @var{den} are the numerator and denominator of the aspect ratio. If
14088 the parameter is not specified, it is assumed the value "0".
14089 In case the form "@var{num}:@var{den}" is used, the @code{:} character
14093 Set the maximum integer value to use for expressing numerator and
14094 denominator when reducing the expressed aspect ratio to a rational.
14095 Default value is @code{100}.
14099 The parameter @var{sar} is an expression containing
14100 the following constants:
14104 These are approximated values for the mathematical constants e
14105 (Euler's number), pi (Greek pi), and phi (the golden ratio).
14108 The input width and height.
14111 These are the same as @var{w} / @var{h}.
14114 The input sample aspect ratio.
14117 The input display aspect ratio. It is the same as
14118 (@var{w} / @var{h}) * @var{sar}.
14121 Horizontal and vertical chroma subsample values. For example, for the
14122 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14125 @subsection Examples
14130 To change the display aspect ratio to 16:9, specify one of the following:
14137 To change the sample aspect ratio to 10:11, specify:
14143 To set a display aspect ratio of 16:9, and specify a maximum integer value of
14144 1000 in the aspect ratio reduction, use the command:
14146 setdar=ratio=16/9:max=1000
14154 Force field for the output video frame.
14156 The @code{setfield} filter marks the interlace type field for the
14157 output frames. It does not change the input frame, but only sets the
14158 corresponding property, which affects how the frame is treated by
14159 following filters (e.g. @code{fieldorder} or @code{yadif}).
14161 The filter accepts the following options:
14166 Available values are:
14170 Keep the same field property.
14173 Mark the frame as bottom-field-first.
14176 Mark the frame as top-field-first.
14179 Mark the frame as progressive.
14185 Show a line containing various information for each input video frame.
14186 The input video is not modified.
14188 The shown line contains a sequence of key/value pairs of the form
14189 @var{key}:@var{value}.
14191 The following values are shown in the output:
14195 The (sequential) number of the input frame, starting from 0.
14198 The Presentation TimeStamp of the input frame, expressed as a number of
14199 time base units. The time base unit depends on the filter input pad.
14202 The Presentation TimeStamp of the input frame, expressed as a number of
14206 The position of the frame in the input stream, or -1 if this information is
14207 unavailable and/or meaningless (for example in case of synthetic video).
14210 The pixel format name.
14213 The sample aspect ratio of the input frame, expressed in the form
14214 @var{num}/@var{den}.
14217 The size of the input frame. For the syntax of this option, check the
14218 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14221 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
14222 for bottom field first).
14225 This is 1 if the frame is a key frame, 0 otherwise.
14228 The picture type of the input frame ("I" for an I-frame, "P" for a
14229 P-frame, "B" for a B-frame, or "?" for an unknown type).
14230 Also refer to the documentation of the @code{AVPictureType} enum and of
14231 the @code{av_get_picture_type_char} function defined in
14232 @file{libavutil/avutil.h}.
14235 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
14237 @item plane_checksum
14238 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
14239 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
14242 @section showpalette
14244 Displays the 256 colors palette of each frame. This filter is only relevant for
14245 @var{pal8} pixel format frames.
14247 It accepts the following option:
14251 Set the size of the box used to represent one palette color entry. Default is
14252 @code{30} (for a @code{30x30} pixel box).
14255 @section shuffleframes
14257 Reorder and/or duplicate and/or drop video frames.
14259 It accepts the following parameters:
14263 Set the destination indexes of input frames.
14264 This is space or '|' separated list of indexes that maps input frames to output
14265 frames. Number of indexes also sets maximal value that each index may have.
14266 '-1' index have special meaning and that is to drop frame.
14269 The first frame has the index 0. The default is to keep the input unchanged.
14271 @subsection Examples
14275 Swap second and third frame of every three frames of the input:
14277 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
14281 Swap 10th and 1st frame of every ten frames of the input:
14283 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
14287 @section shuffleplanes
14289 Reorder and/or duplicate video planes.
14291 It accepts the following parameters:
14296 The index of the input plane to be used as the first output plane.
14299 The index of the input plane to be used as the second output plane.
14302 The index of the input plane to be used as the third output plane.
14305 The index of the input plane to be used as the fourth output plane.
14309 The first plane has the index 0. The default is to keep the input unchanged.
14311 @subsection Examples
14315 Swap the second and third planes of the input:
14317 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
14321 @anchor{signalstats}
14322 @section signalstats
14323 Evaluate various visual metrics that assist in determining issues associated
14324 with the digitization of analog video media.
14326 By default the filter will log these metadata values:
14330 Display the minimal Y value contained within the input frame. Expressed in
14334 Display the Y value at the 10% percentile within the input frame. Expressed in
14338 Display the average Y value within the input frame. Expressed in range of
14342 Display the Y value at the 90% percentile within the input frame. Expressed in
14346 Display the maximum Y value contained within the input frame. Expressed in
14350 Display the minimal U value contained within the input frame. Expressed in
14354 Display the U value at the 10% percentile within the input frame. Expressed in
14358 Display the average U value within the input frame. Expressed in range of
14362 Display the U value at the 90% percentile within the input frame. Expressed in
14366 Display the maximum U value contained within the input frame. Expressed in
14370 Display the minimal V value contained within the input frame. Expressed in
14374 Display the V value at the 10% percentile within the input frame. Expressed in
14378 Display the average V value within the input frame. Expressed in range of
14382 Display the V value at the 90% percentile within the input frame. Expressed in
14386 Display the maximum V value contained within the input frame. Expressed in
14390 Display the minimal saturation value contained within the input frame.
14391 Expressed in range of [0-~181.02].
14394 Display the saturation value at the 10% percentile within the input frame.
14395 Expressed in range of [0-~181.02].
14398 Display the average saturation value within the input frame. Expressed in range
14402 Display the saturation value at the 90% percentile within the input frame.
14403 Expressed in range of [0-~181.02].
14406 Display the maximum saturation value contained within the input frame.
14407 Expressed in range of [0-~181.02].
14410 Display the median value for hue within the input frame. Expressed in range of
14414 Display the average value for hue within the input frame. Expressed in range of
14418 Display the average of sample value difference between all values of the Y
14419 plane in the current frame and corresponding values of the previous input frame.
14420 Expressed in range of [0-255].
14423 Display the average of sample value difference between all values of the U
14424 plane in the current frame and corresponding values of the previous input frame.
14425 Expressed in range of [0-255].
14428 Display the average of sample value difference between all values of the V
14429 plane in the current frame and corresponding values of the previous input frame.
14430 Expressed in range of [0-255].
14433 Display bit depth of Y plane in current frame.
14434 Expressed in range of [0-16].
14437 Display bit depth of U plane in current frame.
14438 Expressed in range of [0-16].
14441 Display bit depth of V plane in current frame.
14442 Expressed in range of [0-16].
14445 The filter accepts the following options:
14451 @option{stat} specify an additional form of image analysis.
14452 @option{out} output video with the specified type of pixel highlighted.
14454 Both options accept the following values:
14458 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
14459 unlike the neighboring pixels of the same field. Examples of temporal outliers
14460 include the results of video dropouts, head clogs, or tape tracking issues.
14463 Identify @var{vertical line repetition}. Vertical line repetition includes
14464 similar rows of pixels within a frame. In born-digital video vertical line
14465 repetition is common, but this pattern is uncommon in video digitized from an
14466 analog source. When it occurs in video that results from the digitization of an
14467 analog source it can indicate concealment from a dropout compensator.
14470 Identify pixels that fall outside of legal broadcast range.
14474 Set the highlight color for the @option{out} option. The default color is
14478 @subsection Examples
14482 Output data of various video metrics:
14484 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
14488 Output specific data about the minimum and maximum values of the Y plane per frame:
14490 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
14494 Playback video while highlighting pixels that are outside of broadcast range in red.
14496 ffplay example.mov -vf signalstats="out=brng:color=red"
14500 Playback video with signalstats metadata drawn over the frame.
14502 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
14505 The contents of signalstat_drawtext.txt used in the command are:
14508 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
14509 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
14510 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
14511 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
14519 Calculates the MPEG-7 Video Signature. The filter can handle more than one
14520 input. In this case the matching between the inputs can be calculated additionally.
14521 The filter always passes through the first input. The signature of each stream can
14522 be written into a file.
14524 It accepts the following options:
14528 Enable or disable the matching process.
14530 Available values are:
14534 Disable the calculation of a matching (default).
14536 Calculate the matching for the whole video and output whether the whole video
14537 matches or only parts.
14539 Calculate only until a matching is found or the video ends. Should be faster in
14544 Set the number of inputs. The option value must be a non negative integer.
14545 Default value is 1.
14548 Set the path to which the output is written. If there is more than one input,
14549 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
14550 integer), that will be replaced with the input number. If no filename is
14551 specified, no output will be written. This is the default.
14554 Choose the output format.
14556 Available values are:
14560 Use the specified binary representation (default).
14562 Use the specified xml representation.
14566 Set threshold to detect one word as similar. The option value must be an integer
14567 greater than zero. The default value is 9000.
14570 Set threshold to detect all words as similar. The option value must be an integer
14571 greater than zero. The default value is 60000.
14574 Set threshold to detect frames as similar. The option value must be an integer
14575 greater than zero. The default value is 116.
14578 Set the minimum length of a sequence in frames to recognize it as matching
14579 sequence. The option value must be a non negative integer value.
14580 The default value is 0.
14583 Set the minimum relation, that matching frames to all frames must have.
14584 The option value must be a double value between 0 and 1. The default value is 0.5.
14587 @subsection Examples
14591 To calculate the signature of an input video and store it in signature.bin:
14593 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
14597 To detect whether two videos match and store the signatures in XML format in
14598 signature0.xml and signature1.xml:
14600 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 -
14608 Blur the input video without impacting the outlines.
14610 It accepts the following options:
14613 @item luma_radius, lr
14614 Set the luma radius. The option value must be a float number in
14615 the range [0.1,5.0] that specifies the variance of the gaussian filter
14616 used to blur the image (slower if larger). Default value is 1.0.
14618 @item luma_strength, ls
14619 Set the luma strength. The option value must be a float number
14620 in the range [-1.0,1.0] that configures the blurring. A value included
14621 in [0.0,1.0] will blur the image whereas a value included in
14622 [-1.0,0.0] will sharpen the image. Default value is 1.0.
14624 @item luma_threshold, lt
14625 Set the luma threshold used as a coefficient to determine
14626 whether a pixel should be blurred or not. The option value must be an
14627 integer in the range [-30,30]. A value of 0 will filter all the image,
14628 a value included in [0,30] will filter flat areas and a value included
14629 in [-30,0] will filter edges. Default value is 0.
14631 @item chroma_radius, cr
14632 Set the chroma radius. The option value must be a float number in
14633 the range [0.1,5.0] that specifies the variance of the gaussian filter
14634 used to blur the image (slower if larger). Default value is @option{luma_radius}.
14636 @item chroma_strength, cs
14637 Set the chroma strength. The option value must be a float number
14638 in the range [-1.0,1.0] that configures the blurring. A value included
14639 in [0.0,1.0] will blur the image whereas a value included in
14640 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
14642 @item chroma_threshold, ct
14643 Set the chroma threshold used as a coefficient to determine
14644 whether a pixel should be blurred or not. The option value must be an
14645 integer in the range [-30,30]. A value of 0 will filter all the image,
14646 a value included in [0,30] will filter flat areas and a value included
14647 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
14650 If a chroma option is not explicitly set, the corresponding luma value
14655 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
14657 This filter takes in input two input videos, the first input is
14658 considered the "main" source and is passed unchanged to the
14659 output. The second input is used as a "reference" video for computing
14662 Both video inputs must have the same resolution and pixel format for
14663 this filter to work correctly. Also it assumes that both inputs
14664 have the same number of frames, which are compared one by one.
14666 The filter stores the calculated SSIM of each frame.
14668 The description of the accepted parameters follows.
14671 @item stats_file, f
14672 If specified the filter will use the named file to save the SSIM of
14673 each individual frame. When filename equals "-" the data is sent to
14677 The file printed if @var{stats_file} is selected, contains a sequence of
14678 key/value pairs of the form @var{key}:@var{value} for each compared
14681 A description of each shown parameter follows:
14685 sequential number of the input frame, starting from 1
14687 @item Y, U, V, R, G, B
14688 SSIM of the compared frames for the component specified by the suffix.
14691 SSIM of the compared frames for the whole frame.
14694 Same as above but in dB representation.
14697 This filter also supports the @ref{framesync} options.
14701 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14702 [main][ref] ssim="stats_file=stats.log" [out]
14705 On this example the input file being processed is compared with the
14706 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
14707 is stored in @file{stats.log}.
14709 Another example with both psnr and ssim at same time:
14711 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
14716 Convert between different stereoscopic image formats.
14718 The filters accept the following options:
14722 Set stereoscopic image format of input.
14724 Available values for input image formats are:
14727 side by side parallel (left eye left, right eye right)
14730 side by side crosseye (right eye left, left eye right)
14733 side by side parallel with half width resolution
14734 (left eye left, right eye right)
14737 side by side crosseye with half width resolution
14738 (right eye left, left eye right)
14741 above-below (left eye above, right eye below)
14744 above-below (right eye above, left eye below)
14747 above-below with half height resolution
14748 (left eye above, right eye below)
14751 above-below with half height resolution
14752 (right eye above, left eye below)
14755 alternating frames (left eye first, right eye second)
14758 alternating frames (right eye first, left eye second)
14761 interleaved rows (left eye has top row, right eye starts on next row)
14764 interleaved rows (right eye has top row, left eye starts on next row)
14767 interleaved columns, left eye first
14770 interleaved columns, right eye first
14772 Default value is @samp{sbsl}.
14776 Set stereoscopic image format of output.
14780 side by side parallel (left eye left, right eye right)
14783 side by side crosseye (right eye left, left eye right)
14786 side by side parallel with half width resolution
14787 (left eye left, right eye right)
14790 side by side crosseye with half width resolution
14791 (right eye left, left eye right)
14794 above-below (left eye above, right eye below)
14797 above-below (right eye above, left eye below)
14800 above-below with half height resolution
14801 (left eye above, right eye below)
14804 above-below with half height resolution
14805 (right eye above, left eye below)
14808 alternating frames (left eye first, right eye second)
14811 alternating frames (right eye first, left eye second)
14814 interleaved rows (left eye has top row, right eye starts on next row)
14817 interleaved rows (right eye has top row, left eye starts on next row)
14820 anaglyph red/blue gray
14821 (red filter on left eye, blue filter on right eye)
14824 anaglyph red/green gray
14825 (red filter on left eye, green filter on right eye)
14828 anaglyph red/cyan gray
14829 (red filter on left eye, cyan filter on right eye)
14832 anaglyph red/cyan half colored
14833 (red filter on left eye, cyan filter on right eye)
14836 anaglyph red/cyan color
14837 (red filter on left eye, cyan filter on right eye)
14840 anaglyph red/cyan color optimized with the least squares projection of dubois
14841 (red filter on left eye, cyan filter on right eye)
14844 anaglyph green/magenta gray
14845 (green filter on left eye, magenta filter on right eye)
14848 anaglyph green/magenta half colored
14849 (green filter on left eye, magenta filter on right eye)
14852 anaglyph green/magenta colored
14853 (green filter on left eye, magenta filter on right eye)
14856 anaglyph green/magenta color optimized with the least squares projection of dubois
14857 (green filter on left eye, magenta filter on right eye)
14860 anaglyph yellow/blue gray
14861 (yellow filter on left eye, blue filter on right eye)
14864 anaglyph yellow/blue half colored
14865 (yellow filter on left eye, blue filter on right eye)
14868 anaglyph yellow/blue colored
14869 (yellow filter on left eye, blue filter on right eye)
14872 anaglyph yellow/blue color optimized with the least squares projection of dubois
14873 (yellow filter on left eye, blue filter on right eye)
14876 mono output (left eye only)
14879 mono output (right eye only)
14882 checkerboard, left eye first
14885 checkerboard, right eye first
14888 interleaved columns, left eye first
14891 interleaved columns, right eye first
14897 Default value is @samp{arcd}.
14900 @subsection Examples
14904 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
14910 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
14916 @section streamselect, astreamselect
14917 Select video or audio streams.
14919 The filter accepts the following options:
14923 Set number of inputs. Default is 2.
14926 Set input indexes to remap to outputs.
14929 @subsection Commands
14931 The @code{streamselect} and @code{astreamselect} filter supports the following
14936 Set input indexes to remap to outputs.
14939 @subsection Examples
14943 Select first 5 seconds 1st stream and rest of time 2nd stream:
14945 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
14949 Same as above, but for audio:
14951 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
14956 Apply sobel operator to input video stream.
14958 The filter accepts the following option:
14962 Set which planes will be processed, unprocessed planes will be copied.
14963 By default value 0xf, all planes will be processed.
14966 Set value which will be multiplied with filtered result.
14969 Set value which will be added to filtered result.
14975 Apply a simple postprocessing filter that compresses and decompresses the image
14976 at several (or - in the case of @option{quality} level @code{6} - all) shifts
14977 and average the results.
14979 The filter accepts the following options:
14983 Set quality. This option defines the number of levels for averaging. It accepts
14984 an integer in the range 0-6. If set to @code{0}, the filter will have no
14985 effect. A value of @code{6} means the higher quality. For each increment of
14986 that value the speed drops by a factor of approximately 2. Default value is
14990 Force a constant quantization parameter. If not set, the filter will use the QP
14991 from the video stream (if available).
14994 Set thresholding mode. Available modes are:
14998 Set hard thresholding (default).
15000 Set soft thresholding (better de-ringing effect, but likely blurrier).
15003 @item use_bframe_qp
15004 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
15005 option may cause flicker since the B-Frames have often larger QP. Default is
15006 @code{0} (not enabled).
15012 Draw subtitles on top of input video using the libass library.
15014 To enable compilation of this filter you need to configure FFmpeg with
15015 @code{--enable-libass}. This filter also requires a build with libavcodec and
15016 libavformat to convert the passed subtitles file to ASS (Advanced Substation
15017 Alpha) subtitles format.
15019 The filter accepts the following options:
15023 Set the filename of the subtitle file to read. It must be specified.
15025 @item original_size
15026 Specify the size of the original video, the video for which the ASS file
15027 was composed. For the syntax of this option, check the
15028 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15029 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
15030 correctly scale the fonts if the aspect ratio has been changed.
15033 Set a directory path containing fonts that can be used by the filter.
15034 These fonts will be used in addition to whatever the font provider uses.
15037 Process alpha channel, by default alpha channel is untouched.
15040 Set subtitles input character encoding. @code{subtitles} filter only. Only
15041 useful if not UTF-8.
15043 @item stream_index, si
15044 Set subtitles stream index. @code{subtitles} filter only.
15047 Override default style or script info parameters of the subtitles. It accepts a
15048 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
15051 If the first key is not specified, it is assumed that the first value
15052 specifies the @option{filename}.
15054 For example, to render the file @file{sub.srt} on top of the input
15055 video, use the command:
15060 which is equivalent to:
15062 subtitles=filename=sub.srt
15065 To render the default subtitles stream from file @file{video.mkv}, use:
15067 subtitles=video.mkv
15070 To render the second subtitles stream from that file, use:
15072 subtitles=video.mkv:si=1
15075 To make the subtitles stream from @file{sub.srt} appear in transparent green
15076 @code{DejaVu Serif}, use:
15078 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
15081 @section super2xsai
15083 Scale the input by 2x and smooth using the Super2xSaI (Scale and
15084 Interpolate) pixel art scaling algorithm.
15086 Useful for enlarging pixel art images without reducing sharpness.
15090 Swap two rectangular objects in video.
15092 This filter accepts the following options:
15102 Set 1st rect x coordinate.
15105 Set 1st rect y coordinate.
15108 Set 2nd rect x coordinate.
15111 Set 2nd rect y coordinate.
15113 All expressions are evaluated once for each frame.
15116 The all options are expressions containing the following constants:
15121 The input width and height.
15124 same as @var{w} / @var{h}
15127 input sample aspect ratio
15130 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
15133 The number of the input frame, starting from 0.
15136 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
15139 the position in the file of the input frame, NAN if unknown
15147 Apply telecine process to the video.
15149 This filter accepts the following options:
15158 The default value is @code{top}.
15162 A string of numbers representing the pulldown pattern you wish to apply.
15163 The default value is @code{23}.
15167 Some typical patterns:
15172 24p: 2332 (preferred)
15179 24p: 222222222223 ("Euro pulldown")
15186 Apply threshold effect to video stream.
15188 This filter needs four video streams to perform thresholding.
15189 First stream is stream we are filtering.
15190 Second stream is holding threshold values, third stream is holding min values,
15191 and last, fourth stream is holding max values.
15193 The filter accepts the following option:
15197 Set which planes will be processed, unprocessed planes will be copied.
15198 By default value 0xf, all planes will be processed.
15201 For example if first stream pixel's component value is less then threshold value
15202 of pixel component from 2nd threshold stream, third stream value will picked,
15203 otherwise fourth stream pixel component value will be picked.
15205 Using color source filter one can perform various types of thresholding:
15207 @subsection Examples
15211 Binary threshold, using gray color as threshold:
15213 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
15217 Inverted binary threshold, using gray color as threshold:
15219 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
15223 Truncate binary threshold, using gray color as threshold:
15225 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
15229 Threshold to zero, using gray color as threshold:
15231 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
15235 Inverted threshold to zero, using gray color as threshold:
15237 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
15242 Select the most representative frame in a given sequence of consecutive frames.
15244 The filter accepts the following options:
15248 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
15249 will pick one of them, and then handle the next batch of @var{n} frames until
15250 the end. Default is @code{100}.
15253 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
15254 value will result in a higher memory usage, so a high value is not recommended.
15256 @subsection Examples
15260 Extract one picture each 50 frames:
15266 Complete example of a thumbnail creation with @command{ffmpeg}:
15268 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
15274 Tile several successive frames together.
15276 The filter accepts the following options:
15281 Set the grid size (i.e. the number of lines and columns). For the syntax of
15282 this option, check the
15283 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15286 Set the maximum number of frames to render in the given area. It must be less
15287 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
15288 the area will be used.
15291 Set the outer border margin in pixels.
15294 Set the inner border thickness (i.e. the number of pixels between frames). For
15295 more advanced padding options (such as having different values for the edges),
15296 refer to the pad video filter.
15299 Specify the color of the unused area. For the syntax of this option, check the
15300 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15301 The default value of @var{color} is "black".
15304 Set the number of frames to overlap when tiling several successive frames together.
15305 The value must be between @code{0} and @var{nb_frames - 1}.
15308 Set the number of frames to initially be empty before displaying first output frame.
15309 This controls how soon will one get first output frame.
15310 The value must be between @code{0} and @var{nb_frames - 1}.
15313 @subsection Examples
15317 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
15319 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
15321 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
15322 duplicating each output frame to accommodate the originally detected frame
15326 Display @code{5} pictures in an area of @code{3x2} frames,
15327 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
15328 mixed flat and named options:
15330 tile=3x2:nb_frames=5:padding=7:margin=2
15334 @section tinterlace
15336 Perform various types of temporal field interlacing.
15338 Frames are counted starting from 1, so the first input frame is
15341 The filter accepts the following options:
15346 Specify the mode of the interlacing. This option can also be specified
15347 as a value alone. See below for a list of values for this option.
15349 Available values are:
15353 Move odd frames into the upper field, even into the lower field,
15354 generating a double height frame at half frame rate.
15358 Frame 1 Frame 2 Frame 3 Frame 4
15360 11111 22222 33333 44444
15361 11111 22222 33333 44444
15362 11111 22222 33333 44444
15363 11111 22222 33333 44444
15377 Only output odd frames, even frames are dropped, generating a frame with
15378 unchanged height at half frame rate.
15383 Frame 1 Frame 2 Frame 3 Frame 4
15385 11111 22222 33333 44444
15386 11111 22222 33333 44444
15387 11111 22222 33333 44444
15388 11111 22222 33333 44444
15398 Only output even frames, odd frames are dropped, generating a frame with
15399 unchanged height at half frame rate.
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
15419 Expand each frame to full height, but pad alternate lines with black,
15420 generating a frame with double height at the same input frame rate.
15425 Frame 1 Frame 2 Frame 3 Frame 4
15427 11111 22222 33333 44444
15428 11111 22222 33333 44444
15429 11111 22222 33333 44444
15430 11111 22222 33333 44444
15433 11111 ..... 33333 .....
15434 ..... 22222 ..... 44444
15435 11111 ..... 33333 .....
15436 ..... 22222 ..... 44444
15437 11111 ..... 33333 .....
15438 ..... 22222 ..... 44444
15439 11111 ..... 33333 .....
15440 ..... 22222 ..... 44444
15444 @item interleave_top, 4
15445 Interleave the upper field from odd frames with the lower field from
15446 even frames, generating a frame with unchanged height at half frame rate.
15451 Frame 1 Frame 2 Frame 3 Frame 4
15453 11111<- 22222 33333<- 44444
15454 11111 22222<- 33333 44444<-
15455 11111<- 22222 33333<- 44444
15456 11111 22222<- 33333 44444<-
15466 @item interleave_bottom, 5
15467 Interleave the lower field from odd frames with the upper field from
15468 even frames, generating a frame with unchanged height at half frame rate.
15473 Frame 1 Frame 2 Frame 3 Frame 4
15475 11111 22222<- 33333 44444<-
15476 11111<- 22222 33333<- 44444
15477 11111 22222<- 33333 44444<-
15478 11111<- 22222 33333<- 44444
15488 @item interlacex2, 6
15489 Double frame rate with unchanged height. Frames are inserted each
15490 containing the second temporal field from the previous input frame and
15491 the first temporal field from the next input frame. This mode relies on
15492 the top_field_first flag. Useful for interlaced video displays with no
15493 field synchronisation.
15498 Frame 1 Frame 2 Frame 3 Frame 4
15500 11111 22222 33333 44444
15501 11111 22222 33333 44444
15502 11111 22222 33333 44444
15503 11111 22222 33333 44444
15506 11111 22222 22222 33333 33333 44444 44444
15507 11111 11111 22222 22222 33333 33333 44444
15508 11111 22222 22222 33333 33333 44444 44444
15509 11111 11111 22222 22222 33333 33333 44444
15514 Move odd frames into the upper field, even into the lower field,
15515 generating a double height frame at same frame rate.
15520 Frame 1 Frame 2 Frame 3 Frame 4
15522 11111 22222 33333 44444
15523 11111 22222 33333 44444
15524 11111 22222 33333 44444
15525 11111 22222 33333 44444
15528 11111 33333 33333 55555
15529 22222 22222 44444 44444
15530 11111 33333 33333 55555
15531 22222 22222 44444 44444
15532 11111 33333 33333 55555
15533 22222 22222 44444 44444
15534 11111 33333 33333 55555
15535 22222 22222 44444 44444
15540 Numeric values are deprecated but are accepted for backward
15541 compatibility reasons.
15543 Default mode is @code{merge}.
15546 Specify flags influencing the filter process.
15548 Available value for @var{flags} is:
15551 @item low_pass_filter, vlfp
15552 Enable linear vertical low-pass filtering in the filter.
15553 Vertical low-pass filtering is required when creating an interlaced
15554 destination from a progressive source which contains high-frequency
15555 vertical detail. Filtering will reduce interlace 'twitter' and Moire
15558 @item complex_filter, cvlfp
15559 Enable complex vertical low-pass filtering.
15560 This will slightly less reduce interlace 'twitter' and Moire
15561 patterning but better retain detail and subjective sharpness impression.
15565 Vertical low-pass filtering can only be enabled for @option{mode}
15566 @var{interleave_top} and @var{interleave_bottom}.
15572 Mix successive video frames.
15574 A description of the accepted options follows.
15578 The number of successive frames to mix. If unspecified, it defaults to 3.
15581 Specify weight of each input video frame.
15582 Each weight is separated by space.
15585 Specify scale, if it is set it will be multiplied with sum
15586 of each weight multiplied with pixel values to give final destination
15587 pixel value. By default @var{scale} is auto scaled to sum of weights.
15591 Tone map colors from different dynamic ranges.
15593 This filter expects data in single precision floating point, as it needs to
15594 operate on (and can output) out-of-range values. Another filter, such as
15595 @ref{zscale}, is needed to convert the resulting frame to a usable format.
15597 The tonemapping algorithms implemented only work on linear light, so input
15598 data should be linearized beforehand (and possibly correctly tagged).
15601 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
15604 @subsection Options
15605 The filter accepts the following options.
15609 Set the tone map algorithm to use.
15611 Possible values are:
15614 Do not apply any tone map, only desaturate overbright pixels.
15617 Hard-clip any out-of-range values. Use it for perfect color accuracy for
15618 in-range values, while distorting out-of-range values.
15621 Stretch the entire reference gamut to a linear multiple of the display.
15624 Fit a logarithmic transfer between the tone curves.
15627 Preserve overall image brightness with a simple curve, using nonlinear
15628 contrast, which results in flattening details and degrading color accuracy.
15631 Preserve both dark and bright details better than @var{reinhard}, at the cost
15632 of slightly darkening everything. Use it when detail preservation is more
15633 important than color and brightness accuracy.
15636 Smoothly map out-of-range values, while retaining contrast and colors for
15637 in-range material as much as possible. Use it when color accuracy is more
15638 important than detail preservation.
15644 Tune the tone mapping algorithm.
15646 This affects the following algorithms:
15652 Specifies the scale factor to use while stretching.
15656 Specifies the exponent of the function.
15660 Specify an extra linear coefficient to multiply into the signal before clipping.
15664 Specify the local contrast coefficient at the display peak.
15665 Default to 0.5, which means that in-gamut values will be about half as bright
15672 Specify the transition point from linear to mobius transform. Every value
15673 below this point is guaranteed to be mapped 1:1. The higher the value, the
15674 more accurate the result will be, at the cost of losing bright details.
15675 Default to 0.3, which due to the steep initial slope still preserves in-range
15676 colors fairly accurately.
15680 Apply desaturation for highlights that exceed this level of brightness. The
15681 higher the parameter, the more color information will be preserved. This
15682 setting helps prevent unnaturally blown-out colors for super-highlights, by
15683 (smoothly) turning into white instead. This makes images feel more natural,
15684 at the cost of reducing information about out-of-range colors.
15686 The default of 2.0 is somewhat conservative and will mostly just apply to
15687 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
15689 This option works only if the input frame has a supported color tag.
15692 Override signal/nominal/reference peak with this value. Useful when the
15693 embedded peak information in display metadata is not reliable or when tone
15694 mapping from a lower range to a higher range.
15699 Transpose rows with columns in the input video and optionally flip it.
15701 It accepts the following parameters:
15706 Specify the transposition direction.
15708 Can assume the following values:
15710 @item 0, 4, cclock_flip
15711 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
15719 Rotate by 90 degrees clockwise, that is:
15727 Rotate by 90 degrees counterclockwise, that is:
15734 @item 3, 7, clock_flip
15735 Rotate by 90 degrees clockwise and vertically flip, that is:
15743 For values between 4-7, the transposition is only done if the input
15744 video geometry is portrait and not landscape. These values are
15745 deprecated, the @code{passthrough} option should be used instead.
15747 Numerical values are deprecated, and should be dropped in favor of
15748 symbolic constants.
15751 Do not apply the transposition if the input geometry matches the one
15752 specified by the specified value. It accepts the following values:
15755 Always apply transposition.
15757 Preserve portrait geometry (when @var{height} >= @var{width}).
15759 Preserve landscape geometry (when @var{width} >= @var{height}).
15762 Default value is @code{none}.
15765 For example to rotate by 90 degrees clockwise and preserve portrait
15768 transpose=dir=1:passthrough=portrait
15771 The command above can also be specified as:
15773 transpose=1:portrait
15777 Trim the input so that the output contains one continuous subpart of the input.
15779 It accepts the following parameters:
15782 Specify the time of the start of the kept section, i.e. the frame with the
15783 timestamp @var{start} will be the first frame in the output.
15786 Specify the time of the first frame that will be dropped, i.e. the frame
15787 immediately preceding the one with the timestamp @var{end} will be the last
15788 frame in the output.
15791 This is the same as @var{start}, except this option sets the start timestamp
15792 in timebase units instead of seconds.
15795 This is the same as @var{end}, except this option sets the end timestamp
15796 in timebase units instead of seconds.
15799 The maximum duration of the output in seconds.
15802 The number of the first frame that should be passed to the output.
15805 The number of the first frame that should be dropped.
15808 @option{start}, @option{end}, and @option{duration} are expressed as time
15809 duration specifications; see
15810 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15811 for the accepted syntax.
15813 Note that the first two sets of the start/end options and the @option{duration}
15814 option look at the frame timestamp, while the _frame variants simply count the
15815 frames that pass through the filter. Also note that this filter does not modify
15816 the timestamps. If you wish for the output timestamps to start at zero, insert a
15817 setpts filter after the trim filter.
15819 If multiple start or end options are set, this filter tries to be greedy and
15820 keep all the frames that match at least one of the specified constraints. To keep
15821 only the part that matches all the constraints at once, chain multiple trim
15824 The defaults are such that all the input is kept. So it is possible to set e.g.
15825 just the end values to keep everything before the specified time.
15830 Drop everything except the second minute of input:
15832 ffmpeg -i INPUT -vf trim=60:120
15836 Keep only the first second:
15838 ffmpeg -i INPUT -vf trim=duration=1
15843 @section unpremultiply
15844 Apply alpha unpremultiply effect to input video stream using first plane
15845 of second stream as alpha.
15847 Both streams must have same dimensions and same pixel format.
15849 The filter accepts the following option:
15853 Set which planes will be processed, unprocessed planes will be copied.
15854 By default value 0xf, all planes will be processed.
15856 If the format has 1 or 2 components, then luma is bit 0.
15857 If the format has 3 or 4 components:
15858 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
15859 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
15860 If present, the alpha channel is always the last bit.
15863 Do not require 2nd input for processing, instead use alpha plane from input stream.
15869 Sharpen or blur the input video.
15871 It accepts the following parameters:
15874 @item luma_msize_x, lx
15875 Set the luma matrix horizontal size. It must be an odd integer between
15876 3 and 23. The default value is 5.
15878 @item luma_msize_y, ly
15879 Set the luma matrix vertical size. It must be an odd integer between 3
15880 and 23. The default value is 5.
15882 @item luma_amount, la
15883 Set the luma effect strength. It must be a floating point number, reasonable
15884 values lay between -1.5 and 1.5.
15886 Negative values will blur the input video, while positive values will
15887 sharpen it, a value of zero will disable the effect.
15889 Default value is 1.0.
15891 @item chroma_msize_x, cx
15892 Set the chroma matrix horizontal size. It must be an odd integer
15893 between 3 and 23. The default value is 5.
15895 @item chroma_msize_y, cy
15896 Set the chroma matrix vertical size. It must be an odd integer
15897 between 3 and 23. The default value is 5.
15899 @item chroma_amount, ca
15900 Set the chroma effect strength. It must be a floating point number, reasonable
15901 values lay between -1.5 and 1.5.
15903 Negative values will blur the input video, while positive values will
15904 sharpen it, a value of zero will disable the effect.
15906 Default value is 0.0.
15910 All parameters are optional and default to the equivalent of the
15911 string '5:5:1.0:5:5:0.0'.
15913 @subsection Examples
15917 Apply strong luma sharpen effect:
15919 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
15923 Apply a strong blur of both luma and chroma parameters:
15925 unsharp=7:7:-2:7:7:-2
15931 Apply ultra slow/simple postprocessing filter that compresses and decompresses
15932 the image at several (or - in the case of @option{quality} level @code{8} - all)
15933 shifts and average the results.
15935 The way this differs from the behavior of spp is that uspp actually encodes &
15936 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
15937 DCT similar to MJPEG.
15939 The filter accepts the following options:
15943 Set quality. This option defines the number of levels for averaging. It accepts
15944 an integer in the range 0-8. If set to @code{0}, the filter will have no
15945 effect. A value of @code{8} means the higher quality. For each increment of
15946 that value the speed drops by a factor of approximately 2. Default value is
15950 Force a constant quantization parameter. If not set, the filter will use the QP
15951 from the video stream (if available).
15954 @section vaguedenoiser
15956 Apply a wavelet based denoiser.
15958 It transforms each frame from the video input into the wavelet domain,
15959 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
15960 the obtained coefficients. It does an inverse wavelet transform after.
15961 Due to wavelet properties, it should give a nice smoothed result, and
15962 reduced noise, without blurring picture features.
15964 This filter accepts the following options:
15968 The filtering strength. The higher, the more filtered the video will be.
15969 Hard thresholding can use a higher threshold than soft thresholding
15970 before the video looks overfiltered. Default value is 2.
15973 The filtering method the filter will use.
15975 It accepts the following values:
15978 All values under the threshold will be zeroed.
15981 All values under the threshold will be zeroed. All values above will be
15982 reduced by the threshold.
15985 Scales or nullifies coefficients - intermediary between (more) soft and
15986 (less) hard thresholding.
15989 Default is garrote.
15992 Number of times, the wavelet will decompose the picture. Picture can't
15993 be decomposed beyond a particular point (typically, 8 for a 640x480
15994 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
15997 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
16000 A list of the planes to process. By default all planes are processed.
16003 @section vectorscope
16005 Display 2 color component values in the two dimensional graph (which is called
16008 This filter accepts the following options:
16012 Set vectorscope mode.
16014 It accepts the following values:
16017 Gray values are displayed on graph, higher brightness means more pixels have
16018 same component color value on location in graph. This is the default mode.
16021 Gray values are displayed on graph. Surrounding pixels values which are not
16022 present in video frame are drawn in gradient of 2 color components which are
16023 set by option @code{x} and @code{y}. The 3rd color component is static.
16026 Actual color components values present in video frame are displayed on graph.
16029 Similar as color2 but higher frequency of same values @code{x} and @code{y}
16030 on graph increases value of another color component, which is luminance by
16031 default values of @code{x} and @code{y}.
16034 Actual colors present in video frame are displayed on graph. If two different
16035 colors map to same position on graph then color with higher value of component
16036 not present in graph is picked.
16039 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
16040 component picked from radial gradient.
16044 Set which color component will be represented on X-axis. Default is @code{1}.
16047 Set which color component will be represented on Y-axis. Default is @code{2}.
16050 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
16051 of color component which represents frequency of (X, Y) location in graph.
16056 No envelope, this is default.
16059 Instant envelope, even darkest single pixel will be clearly highlighted.
16062 Hold maximum and minimum values presented in graph over time. This way you
16063 can still spot out of range values without constantly looking at vectorscope.
16066 Peak and instant envelope combined together.
16070 Set what kind of graticule to draw.
16078 Set graticule opacity.
16081 Set graticule flags.
16085 Draw graticule for white point.
16088 Draw graticule for black point.
16091 Draw color points short names.
16095 Set background opacity.
16097 @item lthreshold, l
16098 Set low threshold for color component not represented on X or Y axis.
16099 Values lower than this value will be ignored. Default is 0.
16100 Note this value is multiplied with actual max possible value one pixel component
16101 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
16104 @item hthreshold, h
16105 Set high threshold for color component not represented on X or Y axis.
16106 Values higher than this value will be ignored. Default is 1.
16107 Note this value is multiplied with actual max possible value one pixel component
16108 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
16109 is 0.9 * 255 = 230.
16111 @item colorspace, c
16112 Set what kind of colorspace to use when drawing graticule.
16121 @anchor{vidstabdetect}
16122 @section vidstabdetect
16124 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
16125 @ref{vidstabtransform} for pass 2.
16127 This filter generates a file with relative translation and rotation
16128 transform information about subsequent frames, which is then used by
16129 the @ref{vidstabtransform} filter.
16131 To enable compilation of this filter you need to configure FFmpeg with
16132 @code{--enable-libvidstab}.
16134 This filter accepts the following options:
16138 Set the path to the file used to write the transforms information.
16139 Default value is @file{transforms.trf}.
16142 Set how shaky the video is and how quick the camera is. It accepts an
16143 integer in the range 1-10, a value of 1 means little shakiness, a
16144 value of 10 means strong shakiness. Default value is 5.
16147 Set the accuracy of the detection process. It must be a value in the
16148 range 1-15. A value of 1 means low accuracy, a value of 15 means high
16149 accuracy. Default value is 15.
16152 Set stepsize of the search process. The region around minimum is
16153 scanned with 1 pixel resolution. Default value is 6.
16156 Set minimum contrast. Below this value a local measurement field is
16157 discarded. Must be a floating point value in the range 0-1. Default
16161 Set reference frame number for tripod mode.
16163 If enabled, the motion of the frames is compared to a reference frame
16164 in the filtered stream, identified by the specified number. The idea
16165 is to compensate all movements in a more-or-less static scene and keep
16166 the camera view absolutely still.
16168 If set to 0, it is disabled. The frames are counted starting from 1.
16171 Show fields and transforms in the resulting frames. It accepts an
16172 integer in the range 0-2. Default value is 0, which disables any
16176 @subsection Examples
16180 Use default values:
16186 Analyze strongly shaky movie and put the results in file
16187 @file{mytransforms.trf}:
16189 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
16193 Visualize the result of internal transformations in the resulting
16196 vidstabdetect=show=1
16200 Analyze a video with medium shakiness using @command{ffmpeg}:
16202 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
16206 @anchor{vidstabtransform}
16207 @section vidstabtransform
16209 Video stabilization/deshaking: pass 2 of 2,
16210 see @ref{vidstabdetect} for pass 1.
16212 Read a file with transform information for each frame and
16213 apply/compensate them. Together with the @ref{vidstabdetect}
16214 filter this can be used to deshake videos. See also
16215 @url{http://public.hronopik.de/vid.stab}. It is important to also use
16216 the @ref{unsharp} filter, see below.
16218 To enable compilation of this filter you need to configure FFmpeg with
16219 @code{--enable-libvidstab}.
16221 @subsection Options
16225 Set path to the file used to read the transforms. Default value is
16226 @file{transforms.trf}.
16229 Set the number of frames (value*2 + 1) used for lowpass filtering the
16230 camera movements. Default value is 10.
16232 For example a number of 10 means that 21 frames are used (10 in the
16233 past and 10 in the future) to smoothen the motion in the video. A
16234 larger value leads to a smoother video, but limits the acceleration of
16235 the camera (pan/tilt movements). 0 is a special case where a static
16236 camera is simulated.
16239 Set the camera path optimization algorithm.
16241 Accepted values are:
16244 gaussian kernel low-pass filter on camera motion (default)
16246 averaging on transformations
16250 Set maximal number of pixels to translate frames. Default value is -1,
16254 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
16255 value is -1, meaning no limit.
16258 Specify how to deal with borders that may be visible due to movement
16261 Available values are:
16264 keep image information from previous frame (default)
16266 fill the border black
16270 Invert transforms if set to 1. Default value is 0.
16273 Consider transforms as relative to previous frame if set to 1,
16274 absolute if set to 0. Default value is 0.
16277 Set percentage to zoom. A positive value will result in a zoom-in
16278 effect, a negative value in a zoom-out effect. Default value is 0 (no
16282 Set optimal zooming to avoid borders.
16284 Accepted values are:
16289 optimal static zoom value is determined (only very strong movements
16290 will lead to visible borders) (default)
16292 optimal adaptive zoom value is determined (no borders will be
16293 visible), see @option{zoomspeed}
16296 Note that the value given at zoom is added to the one calculated here.
16299 Set percent to zoom maximally each frame (enabled when
16300 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
16304 Specify type of interpolation.
16306 Available values are:
16311 linear only horizontal
16313 linear in both directions (default)
16315 cubic in both directions (slow)
16319 Enable virtual tripod mode if set to 1, which is equivalent to
16320 @code{relative=0:smoothing=0}. Default value is 0.
16322 Use also @code{tripod} option of @ref{vidstabdetect}.
16325 Increase log verbosity if set to 1. Also the detected global motions
16326 are written to the temporary file @file{global_motions.trf}. Default
16330 @subsection Examples
16334 Use @command{ffmpeg} for a typical stabilization with default values:
16336 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
16339 Note the use of the @ref{unsharp} filter which is always recommended.
16342 Zoom in a bit more and load transform data from a given file:
16344 vidstabtransform=zoom=5:input="mytransforms.trf"
16348 Smoothen the video even more:
16350 vidstabtransform=smoothing=30
16356 Flip the input video vertically.
16358 For example, to vertically flip a video with @command{ffmpeg}:
16360 ffmpeg -i in.avi -vf "vflip" out.avi
16365 Detect variable frame rate video.
16367 This filter tries to detect if the input is variable or constant frame rate.
16369 At end it will output number of frames detected as having variable delta pts,
16370 and ones with constant delta pts.
16371 If there was frames with variable delta, than it will also show min and max delta
16377 Make or reverse a natural vignetting effect.
16379 The filter accepts the following options:
16383 Set lens angle expression as a number of radians.
16385 The value is clipped in the @code{[0,PI/2]} range.
16387 Default value: @code{"PI/5"}
16391 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
16395 Set forward/backward mode.
16397 Available modes are:
16400 The larger the distance from the central point, the darker the image becomes.
16403 The larger the distance from the central point, the brighter the image becomes.
16404 This can be used to reverse a vignette effect, though there is no automatic
16405 detection to extract the lens @option{angle} and other settings (yet). It can
16406 also be used to create a burning effect.
16409 Default value is @samp{forward}.
16412 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
16414 It accepts the following values:
16417 Evaluate expressions only once during the filter initialization.
16420 Evaluate expressions for each incoming frame. This is way slower than the
16421 @samp{init} mode since it requires all the scalers to be re-computed, but it
16422 allows advanced dynamic expressions.
16425 Default value is @samp{init}.
16428 Set dithering to reduce the circular banding effects. Default is @code{1}
16432 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
16433 Setting this value to the SAR of the input will make a rectangular vignetting
16434 following the dimensions of the video.
16436 Default is @code{1/1}.
16439 @subsection Expressions
16441 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
16442 following parameters.
16447 input width and height
16450 the number of input frame, starting from 0
16453 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
16454 @var{TB} units, NAN if undefined
16457 frame rate of the input video, NAN if the input frame rate is unknown
16460 the PTS (Presentation TimeStamp) of the filtered video frame,
16461 expressed in seconds, NAN if undefined
16464 time base of the input video
16468 @subsection Examples
16472 Apply simple strong vignetting effect:
16478 Make a flickering vignetting:
16480 vignette='PI/4+random(1)*PI/50':eval=frame
16485 @section vmafmotion
16487 Obtain the average vmaf motion score of a video.
16488 It is one of the component filters of VMAF.
16490 The obtained average motion score is printed through the logging system.
16492 In the below example the input file @file{ref.mpg} is being processed and score
16496 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
16500 Stack input videos vertically.
16502 All streams must be of same pixel format and of same width.
16504 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
16505 to create same output.
16507 The filter accept the following option:
16511 Set number of input streams. Default is 2.
16514 If set to 1, force the output to terminate when the shortest input
16515 terminates. Default value is 0.
16520 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
16521 Deinterlacing Filter").
16523 Based on the process described by Martin Weston for BBC R&D, and
16524 implemented based on the de-interlace algorithm written by Jim
16525 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
16526 uses filter coefficients calculated by BBC R&D.
16528 There are two sets of filter coefficients, so called "simple":
16529 and "complex". Which set of filter coefficients is used can
16530 be set by passing an optional parameter:
16534 Set the interlacing filter coefficients. Accepts one of the following values:
16538 Simple filter coefficient set.
16540 More-complex filter coefficient set.
16542 Default value is @samp{complex}.
16545 Specify which frames to deinterlace. Accept one of the following values:
16549 Deinterlace all frames,
16551 Only deinterlace frames marked as interlaced.
16554 Default value is @samp{all}.
16558 Video waveform monitor.
16560 The waveform monitor plots color component intensity. By default luminance
16561 only. Each column of the waveform corresponds to a column of pixels in the
16564 It accepts the following options:
16568 Can be either @code{row}, or @code{column}. Default is @code{column}.
16569 In row mode, the graph on the left side represents color component value 0 and
16570 the right side represents value = 255. In column mode, the top side represents
16571 color component value = 0 and bottom side represents value = 255.
16574 Set intensity. Smaller values are useful to find out how many values of the same
16575 luminance are distributed across input rows/columns.
16576 Default value is @code{0.04}. Allowed range is [0, 1].
16579 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
16580 In mirrored mode, higher values will be represented on the left
16581 side for @code{row} mode and at the top for @code{column} mode. Default is
16582 @code{1} (mirrored).
16586 It accepts the following values:
16589 Presents information identical to that in the @code{parade}, except
16590 that the graphs representing color components are superimposed directly
16593 This display mode makes it easier to spot relative differences or similarities
16594 in overlapping areas of the color components that are supposed to be identical,
16595 such as neutral whites, grays, or blacks.
16598 Display separate graph for the color components side by side in
16599 @code{row} mode or one below the other in @code{column} mode.
16602 Display separate graph for the color components side by side in
16603 @code{column} mode or one below the other in @code{row} mode.
16605 Using this display mode makes it easy to spot color casts in the highlights
16606 and shadows of an image, by comparing the contours of the top and the bottom
16607 graphs of each waveform. Since whites, grays, and blacks are characterized
16608 by exactly equal amounts of red, green, and blue, neutral areas of the picture
16609 should display three waveforms of roughly equal width/height. If not, the
16610 correction is easy to perform by making level adjustments the three waveforms.
16612 Default is @code{stack}.
16614 @item components, c
16615 Set which color components to display. Default is 1, which means only luminance
16616 or red color component if input is in RGB colorspace. If is set for example to
16617 7 it will display all 3 (if) available color components.
16622 No envelope, this is default.
16625 Instant envelope, minimum and maximum values presented in graph will be easily
16626 visible even with small @code{step} value.
16629 Hold minimum and maximum values presented in graph across time. This way you
16630 can still spot out of range values without constantly looking at waveforms.
16633 Peak and instant envelope combined together.
16639 No filtering, this is default.
16642 Luma and chroma combined together.
16645 Similar as above, but shows difference between blue and red chroma.
16648 Similar as above, but use different colors.
16651 Displays only chroma.
16654 Displays actual color value on waveform.
16657 Similar as above, but with luma showing frequency of chroma values.
16661 Set which graticule to display.
16665 Do not display graticule.
16668 Display green graticule showing legal broadcast ranges.
16671 Display orange graticule showing legal broadcast ranges.
16675 Set graticule opacity.
16678 Set graticule flags.
16682 Draw numbers above lines. By default enabled.
16685 Draw dots instead of lines.
16689 Set scale used for displaying graticule.
16696 Default is digital.
16699 Set background opacity.
16702 @section weave, doubleweave
16704 The @code{weave} takes a field-based video input and join
16705 each two sequential fields into single frame, producing a new double
16706 height clip with half the frame rate and half the frame count.
16708 The @code{doubleweave} works same as @code{weave} but without
16709 halving frame rate and frame count.
16711 It accepts the following option:
16715 Set first field. Available values are:
16719 Set the frame as top-field-first.
16722 Set the frame as bottom-field-first.
16726 @subsection Examples
16730 Interlace video using @ref{select} and @ref{separatefields} filter:
16732 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
16737 Apply the xBR high-quality magnification filter which is designed for pixel
16738 art. It follows a set of edge-detection rules, see
16739 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
16741 It accepts the following option:
16745 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
16746 @code{3xBR} and @code{4} for @code{4xBR}.
16747 Default is @code{3}.
16753 Deinterlace the input video ("yadif" means "yet another deinterlacing
16756 It accepts the following parameters:
16762 The interlacing mode to adopt. It accepts one of the following values:
16765 @item 0, send_frame
16766 Output one frame for each frame.
16767 @item 1, send_field
16768 Output one frame for each field.
16769 @item 2, send_frame_nospatial
16770 Like @code{send_frame}, but it skips the spatial interlacing check.
16771 @item 3, send_field_nospatial
16772 Like @code{send_field}, but it skips the spatial interlacing check.
16775 The default value is @code{send_frame}.
16778 The picture field parity assumed for the input interlaced video. It accepts one
16779 of the following values:
16783 Assume the top field is first.
16785 Assume the bottom field is first.
16787 Enable automatic detection of field parity.
16790 The default value is @code{auto}.
16791 If the interlacing is unknown or the decoder does not export this information,
16792 top field first will be assumed.
16795 Specify which frames to deinterlace. Accept one of the following
16800 Deinterlace all frames.
16801 @item 1, interlaced
16802 Only deinterlace frames marked as interlaced.
16805 The default value is @code{all}.
16810 Apply Zoom & Pan effect.
16812 This filter accepts the following options:
16816 Set the zoom expression. Default is 1.
16820 Set the x and y expression. Default is 0.
16823 Set the duration expression in number of frames.
16824 This sets for how many number of frames effect will last for
16825 single input image.
16828 Set the output image size, default is 'hd720'.
16831 Set the output frame rate, default is '25'.
16834 Each expression can contain the following constants:
16853 Output frame count.
16857 Last calculated 'x' and 'y' position from 'x' and 'y' expression
16858 for current input frame.
16862 'x' and 'y' of last output frame of previous input frame or 0 when there was
16863 not yet such frame (first input frame).
16866 Last calculated zoom from 'z' expression for current input frame.
16869 Last calculated zoom of last output frame of previous input frame.
16872 Number of output frames for current input frame. Calculated from 'd' expression
16873 for each input frame.
16876 number of output frames created for previous input frame
16879 Rational number: input width / input height
16882 sample aspect ratio
16885 display aspect ratio
16889 @subsection Examples
16893 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
16895 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
16899 Zoom-in up to 1.5 and pan always at center of picture:
16901 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16905 Same as above but without pausing:
16907 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
16913 Scale (resize) the input video, using the z.lib library:
16914 https://github.com/sekrit-twc/zimg.
16916 The zscale filter forces the output display aspect ratio to be the same
16917 as the input, by changing the output sample aspect ratio.
16919 If the input image format is different from the format requested by
16920 the next filter, the zscale filter will convert the input to the
16923 @subsection Options
16924 The filter accepts the following options.
16929 Set the output video dimension expression. Default value is the input
16932 If the @var{width} or @var{w} value is 0, the input width is used for
16933 the output. If the @var{height} or @var{h} value is 0, the input height
16934 is used for the output.
16936 If one and only one of the values is -n with n >= 1, the zscale filter
16937 will use a value that maintains the aspect ratio of the input image,
16938 calculated from the other specified dimension. After that it will,
16939 however, make sure that the calculated dimension is divisible by n and
16940 adjust the value if necessary.
16942 If both values are -n with n >= 1, the behavior will be identical to
16943 both values being set to 0 as previously detailed.
16945 See below for the list of accepted constants for use in the dimension
16949 Set the video size. For the syntax of this option, check the
16950 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16953 Set the dither type.
16955 Possible values are:
16960 @item error_diffusion
16966 Set the resize filter type.
16968 Possible values are:
16978 Default is bilinear.
16981 Set the color range.
16983 Possible values are:
16990 Default is same as input.
16993 Set the color primaries.
16995 Possible values are:
17005 Default is same as input.
17008 Set the transfer characteristics.
17010 Possible values are:
17024 Default is same as input.
17027 Set the colorspace matrix.
17029 Possible value are:
17040 Default is same as input.
17043 Set the input color range.
17045 Possible values are:
17052 Default is same as input.
17054 @item primariesin, pin
17055 Set the input color primaries.
17057 Possible values are:
17067 Default is same as input.
17069 @item transferin, tin
17070 Set the input transfer characteristics.
17072 Possible values are:
17083 Default is same as input.
17085 @item matrixin, min
17086 Set the input colorspace matrix.
17088 Possible value are:
17100 Set the output chroma location.
17102 Possible values are:
17113 @item chromalin, cin
17114 Set the input chroma location.
17116 Possible values are:
17128 Set the nominal peak luminance.
17131 The values of the @option{w} and @option{h} options are expressions
17132 containing the following constants:
17137 The input width and height
17141 These are the same as @var{in_w} and @var{in_h}.
17145 The output (scaled) width and height
17149 These are the same as @var{out_w} and @var{out_h}
17152 The same as @var{iw} / @var{ih}
17155 input sample aspect ratio
17158 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
17162 horizontal and vertical input chroma subsample values. For example for the
17163 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17167 horizontal and vertical output chroma subsample values. For example for the
17168 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
17174 @c man end VIDEO FILTERS
17176 @chapter Video Sources
17177 @c man begin VIDEO SOURCES
17179 Below is a description of the currently available video sources.
17183 Buffer video frames, and make them available to the filter chain.
17185 This source is mainly intended for a programmatic use, in particular
17186 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
17188 It accepts the following parameters:
17193 Specify the size (width and height) of the buffered video frames. For the
17194 syntax of this option, check the
17195 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17198 The input video width.
17201 The input video height.
17204 A string representing the pixel format of the buffered video frames.
17205 It may be a number corresponding to a pixel format, or a pixel format
17209 Specify the timebase assumed by the timestamps of the buffered frames.
17212 Specify the frame rate expected for the video stream.
17214 @item pixel_aspect, sar
17215 The sample (pixel) aspect ratio of the input video.
17218 Specify the optional parameters to be used for the scale filter which
17219 is automatically inserted when an input change is detected in the
17220 input size or format.
17222 @item hw_frames_ctx
17223 When using a hardware pixel format, this should be a reference to an
17224 AVHWFramesContext describing input frames.
17229 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
17232 will instruct the source to accept video frames with size 320x240 and
17233 with format "yuv410p", assuming 1/24 as the timestamps timebase and
17234 square pixels (1:1 sample aspect ratio).
17235 Since the pixel format with name "yuv410p" corresponds to the number 6
17236 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
17237 this example corresponds to:
17239 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
17242 Alternatively, the options can be specified as a flat string, but this
17243 syntax is deprecated:
17245 @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}]
17249 Create a pattern generated by an elementary cellular automaton.
17251 The initial state of the cellular automaton can be defined through the
17252 @option{filename} and @option{pattern} options. If such options are
17253 not specified an initial state is created randomly.
17255 At each new frame a new row in the video is filled with the result of
17256 the cellular automaton next generation. The behavior when the whole
17257 frame is filled is defined by the @option{scroll} option.
17259 This source accepts the following options:
17263 Read the initial cellular automaton state, i.e. the starting row, from
17264 the specified file.
17265 In the file, each non-whitespace character is considered an alive
17266 cell, a newline will terminate the row, and further characters in the
17267 file will be ignored.
17270 Read the initial cellular automaton state, i.e. the starting row, from
17271 the specified string.
17273 Each non-whitespace character in the string is considered an alive
17274 cell, a newline will terminate the row, and further characters in the
17275 string will be ignored.
17278 Set the video rate, that is the number of frames generated per second.
17281 @item random_fill_ratio, ratio
17282 Set the random fill ratio for the initial cellular automaton row. It
17283 is a floating point number value ranging from 0 to 1, defaults to
17286 This option is ignored when a file or a pattern is specified.
17288 @item random_seed, seed
17289 Set the seed for filling randomly the initial row, must be an integer
17290 included between 0 and UINT32_MAX. If not specified, or if explicitly
17291 set to -1, the filter will try to use a good random seed on a best
17295 Set the cellular automaton rule, it is a number ranging from 0 to 255.
17296 Default value is 110.
17299 Set the size of the output video. For the syntax of this option, check the
17300 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17302 If @option{filename} or @option{pattern} is specified, the size is set
17303 by default to the width of the specified initial state row, and the
17304 height is set to @var{width} * PHI.
17306 If @option{size} is set, it must contain the width of the specified
17307 pattern string, and the specified pattern will be centered in the
17310 If a filename or a pattern string is not specified, the size value
17311 defaults to "320x518" (used for a randomly generated initial state).
17314 If set to 1, scroll the output upward when all the rows in the output
17315 have been already filled. If set to 0, the new generated row will be
17316 written over the top row just after the bottom row is filled.
17319 @item start_full, full
17320 If set to 1, completely fill the output with generated rows before
17321 outputting the first frame.
17322 This is the default behavior, for disabling set the value to 0.
17325 If set to 1, stitch the left and right row edges together.
17326 This is the default behavior, for disabling set the value to 0.
17329 @subsection Examples
17333 Read the initial state from @file{pattern}, and specify an output of
17336 cellauto=f=pattern:s=200x400
17340 Generate a random initial row with a width of 200 cells, with a fill
17343 cellauto=ratio=2/3:s=200x200
17347 Create a pattern generated by rule 18 starting by a single alive cell
17348 centered on an initial row with width 100:
17350 cellauto=p=@@:s=100x400:full=0:rule=18
17354 Specify a more elaborated initial pattern:
17356 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
17361 @anchor{coreimagesrc}
17362 @section coreimagesrc
17363 Video source generated on GPU using Apple's CoreImage API on OSX.
17365 This video source is a specialized version of the @ref{coreimage} video filter.
17366 Use a core image generator at the beginning of the applied filterchain to
17367 generate the content.
17369 The coreimagesrc video source accepts the following options:
17371 @item list_generators
17372 List all available generators along with all their respective options as well as
17373 possible minimum and maximum values along with the default values.
17375 list_generators=true
17379 Specify the size of the sourced video. For the syntax of this option, check the
17380 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17381 The default value is @code{320x240}.
17384 Specify the frame rate of the sourced video, as the number of frames
17385 generated per second. It has to be a string in the format
17386 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17387 number or a valid video frame rate abbreviation. The default value is
17391 Set the sample aspect ratio of the sourced video.
17394 Set the duration of the sourced video. See
17395 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17396 for the accepted syntax.
17398 If not specified, or the expressed duration is negative, the video is
17399 supposed to be generated forever.
17402 Additionally, all options of the @ref{coreimage} video filter are accepted.
17403 A complete filterchain can be used for further processing of the
17404 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
17405 and examples for details.
17407 @subsection Examples
17412 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
17413 given as complete and escaped command-line for Apple's standard bash shell:
17415 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
17417 This example is equivalent to the QRCode example of @ref{coreimage} without the
17418 need for a nullsrc video source.
17422 @section mandelbrot
17424 Generate a Mandelbrot set fractal, and progressively zoom towards the
17425 point specified with @var{start_x} and @var{start_y}.
17427 This source accepts the following options:
17432 Set the terminal pts value. Default value is 400.
17435 Set the terminal scale value.
17436 Must be a floating point value. Default value is 0.3.
17439 Set the inner coloring mode, that is the algorithm used to draw the
17440 Mandelbrot fractal internal region.
17442 It shall assume one of the following values:
17447 Show time until convergence.
17449 Set color based on point closest to the origin of the iterations.
17454 Default value is @var{mincol}.
17457 Set the bailout value. Default value is 10.0.
17460 Set the maximum of iterations performed by the rendering
17461 algorithm. Default value is 7189.
17464 Set outer coloring mode.
17465 It shall assume one of following values:
17467 @item iteration_count
17468 Set iteration cound mode.
17469 @item normalized_iteration_count
17470 set normalized iteration count mode.
17472 Default value is @var{normalized_iteration_count}.
17475 Set frame rate, expressed as number of frames per second. Default
17479 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
17480 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
17483 Set the initial scale value. Default value is 3.0.
17486 Set the initial x position. Must be a floating point value between
17487 -100 and 100. Default value is -0.743643887037158704752191506114774.
17490 Set the initial y position. Must be a floating point value between
17491 -100 and 100. Default value is -0.131825904205311970493132056385139.
17496 Generate various test patterns, as generated by the MPlayer test filter.
17498 The size of the generated video is fixed, and is 256x256.
17499 This source is useful in particular for testing encoding features.
17501 This source accepts the following options:
17506 Specify the frame rate of the sourced video, as the number of frames
17507 generated per second. It has to be a string in the format
17508 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17509 number or a valid video frame rate abbreviation. The default value is
17513 Set the duration of the sourced video. See
17514 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17515 for the accepted syntax.
17517 If not specified, or the expressed duration is negative, the video is
17518 supposed to be generated forever.
17522 Set the number or the name of the test to perform. Supported tests are:
17538 Default value is "all", which will cycle through the list of all tests.
17543 mptestsrc=t=dc_luma
17546 will generate a "dc_luma" test pattern.
17548 @section frei0r_src
17550 Provide a frei0r source.
17552 To enable compilation of this filter you need to install the frei0r
17553 header and configure FFmpeg with @code{--enable-frei0r}.
17555 This source accepts the following parameters:
17560 The size of the video to generate. For the syntax of this option, check the
17561 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17564 The framerate of the generated video. It may be a string of the form
17565 @var{num}/@var{den} or a frame rate abbreviation.
17568 The name to the frei0r source to load. For more information regarding frei0r and
17569 how to set the parameters, read the @ref{frei0r} section in the video filters
17572 @item filter_params
17573 A '|'-separated list of parameters to pass to the frei0r source.
17577 For example, to generate a frei0r partik0l source with size 200x200
17578 and frame rate 10 which is overlaid on the overlay filter main input:
17580 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
17585 Generate a life pattern.
17587 This source is based on a generalization of John Conway's life game.
17589 The sourced input represents a life grid, each pixel represents a cell
17590 which can be in one of two possible states, alive or dead. Every cell
17591 interacts with its eight neighbours, which are the cells that are
17592 horizontally, vertically, or diagonally adjacent.
17594 At each interaction the grid evolves according to the adopted rule,
17595 which specifies the number of neighbor alive cells which will make a
17596 cell stay alive or born. The @option{rule} option allows one to specify
17599 This source accepts the following options:
17603 Set the file from which to read the initial grid state. In the file,
17604 each non-whitespace character is considered an alive cell, and newline
17605 is used to delimit the end of each row.
17607 If this option is not specified, the initial grid is generated
17611 Set the video rate, that is the number of frames generated per second.
17614 @item random_fill_ratio, ratio
17615 Set the random fill ratio for the initial random grid. It is a
17616 floating point number value ranging from 0 to 1, defaults to 1/PHI.
17617 It is ignored when a file is specified.
17619 @item random_seed, seed
17620 Set the seed for filling the initial random grid, must be an integer
17621 included between 0 and UINT32_MAX. If not specified, or if explicitly
17622 set to -1, the filter will try to use a good random seed on a best
17628 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
17629 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
17630 @var{NS} specifies the number of alive neighbor cells which make a
17631 live cell stay alive, and @var{NB} the number of alive neighbor cells
17632 which make a dead cell to become alive (i.e. to "born").
17633 "s" and "b" can be used in place of "S" and "B", respectively.
17635 Alternatively a rule can be specified by an 18-bits integer. The 9
17636 high order bits are used to encode the next cell state if it is alive
17637 for each number of neighbor alive cells, the low order bits specify
17638 the rule for "borning" new cells. Higher order bits encode for an
17639 higher number of neighbor cells.
17640 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
17641 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
17643 Default value is "S23/B3", which is the original Conway's game of life
17644 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
17645 cells, and will born a new cell if there are three alive cells around
17649 Set the size of the output video. For the syntax of this option, check the
17650 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17652 If @option{filename} is specified, the size is set by default to the
17653 same size of the input file. If @option{size} is set, it must contain
17654 the size specified in the input file, and the initial grid defined in
17655 that file is centered in the larger resulting area.
17657 If a filename is not specified, the size value defaults to "320x240"
17658 (used for a randomly generated initial grid).
17661 If set to 1, stitch the left and right grid edges together, and the
17662 top and bottom edges also. Defaults to 1.
17665 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
17666 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
17667 value from 0 to 255.
17670 Set the color of living (or new born) cells.
17673 Set the color of dead cells. If @option{mold} is set, this is the first color
17674 used to represent a dead cell.
17677 Set mold color, for definitely dead and moldy cells.
17679 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
17680 ffmpeg-utils manual,ffmpeg-utils}.
17683 @subsection Examples
17687 Read a grid from @file{pattern}, and center it on a grid of size
17690 life=f=pattern:s=300x300
17694 Generate a random grid of size 200x200, with a fill ratio of 2/3:
17696 life=ratio=2/3:s=200x200
17700 Specify a custom rule for evolving a randomly generated grid:
17706 Full example with slow death effect (mold) using @command{ffplay}:
17708 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
17715 @anchor{haldclutsrc}
17717 @anchor{rgbtestsrc}
17719 @anchor{smptehdbars}
17722 @anchor{yuvtestsrc}
17723 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
17725 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
17727 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
17729 The @code{color} source provides an uniformly colored input.
17731 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
17732 @ref{haldclut} filter.
17734 The @code{nullsrc} source returns unprocessed video frames. It is
17735 mainly useful to be employed in analysis / debugging tools, or as the
17736 source for filters which ignore the input data.
17738 The @code{rgbtestsrc} source generates an RGB test pattern useful for
17739 detecting RGB vs BGR issues. You should see a red, green and blue
17740 stripe from top to bottom.
17742 The @code{smptebars} source generates a color bars pattern, based on
17743 the SMPTE Engineering Guideline EG 1-1990.
17745 The @code{smptehdbars} source generates a color bars pattern, based on
17746 the SMPTE RP 219-2002.
17748 The @code{testsrc} source generates a test video pattern, showing a
17749 color pattern, a scrolling gradient and a timestamp. This is mainly
17750 intended for testing purposes.
17752 The @code{testsrc2} source is similar to testsrc, but supports more
17753 pixel formats instead of just @code{rgb24}. This allows using it as an
17754 input for other tests without requiring a format conversion.
17756 The @code{yuvtestsrc} source generates an YUV test pattern. You should
17757 see a y, cb and cr stripe from top to bottom.
17759 The sources accept the following parameters:
17764 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
17765 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
17766 pixels to be used as identity matrix for 3D lookup tables. Each component is
17767 coded on a @code{1/(N*N)} scale.
17770 Specify the color of the source, only available in the @code{color}
17771 source. For the syntax of this option, check the
17772 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17775 Specify the size of the sourced video. For the syntax of this option, check the
17776 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17777 The default value is @code{320x240}.
17779 This option is not available with the @code{allrgb}, @code{allyuv}, and
17780 @code{haldclutsrc} filters.
17783 Specify the frame rate of the sourced video, as the number of frames
17784 generated per second. It has to be a string in the format
17785 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
17786 number or a valid video frame rate abbreviation. The default value is
17790 Set the duration of the sourced video. See
17791 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17792 for the accepted syntax.
17794 If not specified, or the expressed duration is negative, the video is
17795 supposed to be generated forever.
17798 Set the sample aspect ratio of the sourced video.
17801 Specify the alpha (opacity) of the background, only available in the
17802 @code{testsrc2} source. The value must be between 0 (fully transparent) and
17803 255 (fully opaque, the default).
17806 Set the number of decimals to show in the timestamp, only available in the
17807 @code{testsrc} source.
17809 The displayed timestamp value will correspond to the original
17810 timestamp value multiplied by the power of 10 of the specified
17811 value. Default value is 0.
17814 @subsection Examples
17818 Generate a video with a duration of 5.3 seconds, with size
17819 176x144 and a frame rate of 10 frames per second:
17821 testsrc=duration=5.3:size=qcif:rate=10
17825 The following graph description will generate a red source
17826 with an opacity of 0.2, with size "qcif" and a frame rate of 10
17829 color=c=red@@0.2:s=qcif:r=10
17833 If the input content is to be ignored, @code{nullsrc} can be used. The
17834 following command generates noise in the luminance plane by employing
17835 the @code{geq} filter:
17837 nullsrc=s=256x256, geq=random(1)*255:128:128
17841 @subsection Commands
17843 The @code{color} source supports the following commands:
17847 Set the color of the created image. Accepts the same syntax of the
17848 corresponding @option{color} option.
17853 Generate video using an OpenCL program.
17858 OpenCL program source file.
17861 Kernel name in program.
17864 Size of frames to generate. This must be set.
17867 Pixel format to use for the generated frames. This must be set.
17870 Number of frames generated every second. Default value is '25'.
17874 For details of how the program loading works, see the @ref{program_opencl}
17881 Generate a colour ramp by setting pixel values from the position of the pixel
17882 in the output image. (Note that this will work with all pixel formats, but
17883 the generated output will not be the same.)
17885 __kernel void ramp(__write_only image2d_t dst,
17886 unsigned int index)
17888 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17891 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
17893 write_imagef(dst, loc, val);
17898 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
17900 __kernel void sierpinski_carpet(__write_only image2d_t dst,
17901 unsigned int index)
17903 int2 loc = (int2)(get_global_id(0), get_global_id(1));
17905 float4 value = 0.0f;
17906 int x = loc.x + index;
17907 int y = loc.y + index;
17908 while (x > 0 || y > 0) {
17909 if (x % 3 == 1 && y % 3 == 1) {
17917 write_imagef(dst, loc, value);
17923 @c man end VIDEO SOURCES
17925 @chapter Video Sinks
17926 @c man begin VIDEO SINKS
17928 Below is a description of the currently available video sinks.
17930 @section buffersink
17932 Buffer video frames, and make them available to the end of the filter
17935 This sink is mainly intended for programmatic use, in particular
17936 through the interface defined in @file{libavfilter/buffersink.h}
17937 or the options system.
17939 It accepts a pointer to an AVBufferSinkContext structure, which
17940 defines the incoming buffers' formats, to be passed as the opaque
17941 parameter to @code{avfilter_init_filter} for initialization.
17945 Null video sink: do absolutely nothing with the input video. It is
17946 mainly useful as a template and for use in analysis / debugging
17949 @c man end VIDEO SINKS
17951 @chapter Multimedia Filters
17952 @c man begin MULTIMEDIA FILTERS
17954 Below is a description of the currently available multimedia filters.
17958 Convert input audio to a video output, displaying the audio bit scope.
17960 The filter accepts the following options:
17964 Set frame rate, expressed as number of frames per second. Default
17968 Specify the video size for the output. For the syntax of this option, check the
17969 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17970 Default value is @code{1024x256}.
17973 Specify list of colors separated by space or by '|' which will be used to
17974 draw channels. Unrecognized or missing colors will be replaced
17978 @section ahistogram
17980 Convert input audio to a video output, displaying the volume histogram.
17982 The filter accepts the following options:
17986 Specify how histogram is calculated.
17988 It accepts the following values:
17991 Use single histogram for all channels.
17993 Use separate histogram for each channel.
17995 Default is @code{single}.
17998 Set frame rate, expressed as number of frames per second. Default
18002 Specify the video size for the output. For the syntax of this option, check the
18003 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18004 Default value is @code{hd720}.
18009 It accepts the following values:
18020 reverse logarithmic
18022 Default is @code{log}.
18025 Set amplitude scale.
18027 It accepts the following values:
18034 Default is @code{log}.
18037 Set how much frames to accumulate in histogram.
18038 Defauls is 1. Setting this to -1 accumulates all frames.
18041 Set histogram ratio of window height.
18044 Set sonogram sliding.
18046 It accepts the following values:
18049 replace old rows with new ones.
18051 scroll from top to bottom.
18053 Default is @code{replace}.
18056 @section aphasemeter
18058 Convert input audio to a video output, displaying the audio phase.
18060 The filter accepts the following options:
18064 Set the output frame rate. Default value is @code{25}.
18067 Set the video size for the output. For the syntax of this option, check the
18068 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18069 Default value is @code{800x400}.
18074 Specify the red, green, blue contrast. Default values are @code{2},
18075 @code{7} and @code{1}.
18076 Allowed range is @code{[0, 255]}.
18079 Set color which will be used for drawing median phase. If color is
18080 @code{none} which is default, no median phase value will be drawn.
18083 Enable video output. Default is enabled.
18086 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
18087 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
18088 The @code{-1} means left and right channels are completely out of phase and
18089 @code{1} means channels are in phase.
18091 @section avectorscope
18093 Convert input audio to a video output, representing the audio vector
18096 The filter is used to measure the difference between channels of stereo
18097 audio stream. A monoaural signal, consisting of identical left and right
18098 signal, results in straight vertical line. Any stereo separation is visible
18099 as a deviation from this line, creating a Lissajous figure.
18100 If the straight (or deviation from it) but horizontal line appears this
18101 indicates that the left and right channels are out of phase.
18103 The filter accepts the following options:
18107 Set the vectorscope mode.
18109 Available values are:
18112 Lissajous rotated by 45 degrees.
18115 Same as above but not rotated.
18118 Shape resembling half of circle.
18121 Default value is @samp{lissajous}.
18124 Set the video size for the output. For the syntax of this option, check the
18125 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18126 Default value is @code{400x400}.
18129 Set the output frame rate. Default value is @code{25}.
18135 Specify the red, green, blue and alpha contrast. Default values are @code{40},
18136 @code{160}, @code{80} and @code{255}.
18137 Allowed range is @code{[0, 255]}.
18143 Specify the red, green, blue and alpha fade. Default values are @code{15},
18144 @code{10}, @code{5} and @code{5}.
18145 Allowed range is @code{[0, 255]}.
18148 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
18149 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
18152 Set the vectorscope drawing mode.
18154 Available values are:
18157 Draw dot for each sample.
18160 Draw line between previous and current sample.
18163 Default value is @samp{dot}.
18166 Specify amplitude scale of audio samples.
18168 Available values are:
18184 Swap left channel axis with right channel axis.
18194 Mirror only x axis.
18197 Mirror only y axis.
18205 @subsection Examples
18209 Complete example using @command{ffplay}:
18211 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
18212 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
18216 @section bench, abench
18218 Benchmark part of a filtergraph.
18220 The filter accepts the following options:
18224 Start or stop a timer.
18226 Available values are:
18229 Get the current time, set it as frame metadata (using the key
18230 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
18233 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
18234 the input frame metadata to get the time difference. Time difference, average,
18235 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
18236 @code{min}) are then printed. The timestamps are expressed in seconds.
18240 @subsection Examples
18244 Benchmark @ref{selectivecolor} filter:
18246 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
18252 Concatenate audio and video streams, joining them together one after the
18255 The filter works on segments of synchronized video and audio streams. All
18256 segments must have the same number of streams of each type, and that will
18257 also be the number of streams at output.
18259 The filter accepts the following options:
18264 Set the number of segments. Default is 2.
18267 Set the number of output video streams, that is also the number of video
18268 streams in each segment. Default is 1.
18271 Set the number of output audio streams, that is also the number of audio
18272 streams in each segment. Default is 0.
18275 Activate unsafe mode: do not fail if segments have a different format.
18279 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
18280 @var{a} audio outputs.
18282 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
18283 segment, in the same order as the outputs, then the inputs for the second
18286 Related streams do not always have exactly the same duration, for various
18287 reasons including codec frame size or sloppy authoring. For that reason,
18288 related synchronized streams (e.g. a video and its audio track) should be
18289 concatenated at once. The concat filter will use the duration of the longest
18290 stream in each segment (except the last one), and if necessary pad shorter
18291 audio streams with silence.
18293 For this filter to work correctly, all segments must start at timestamp 0.
18295 All corresponding streams must have the same parameters in all segments; the
18296 filtering system will automatically select a common pixel format for video
18297 streams, and a common sample format, sample rate and channel layout for
18298 audio streams, but other settings, such as resolution, must be converted
18299 explicitly by the user.
18301 Different frame rates are acceptable but will result in variable frame rate
18302 at output; be sure to configure the output file to handle it.
18304 @subsection Examples
18308 Concatenate an opening, an episode and an ending, all in bilingual version
18309 (video in stream 0, audio in streams 1 and 2):
18311 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
18312 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
18313 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
18314 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
18318 Concatenate two parts, handling audio and video separately, using the
18319 (a)movie sources, and adjusting the resolution:
18321 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
18322 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
18323 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
18325 Note that a desync will happen at the stitch if the audio and video streams
18326 do not have exactly the same duration in the first file.
18330 @subsection Commands
18332 This filter supports the following commands:
18335 Close the current segment and step to the next one
18338 @section drawgraph, adrawgraph
18340 Draw a graph using input video or audio metadata.
18342 It accepts the following parameters:
18346 Set 1st frame metadata key from which metadata values will be used to draw a graph.
18349 Set 1st foreground color expression.
18352 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
18355 Set 2nd foreground color expression.
18358 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
18361 Set 3rd foreground color expression.
18364 Set 4th frame metadata key from which metadata values will be used to draw a graph.
18367 Set 4th foreground color expression.
18370 Set minimal value of metadata value.
18373 Set maximal value of metadata value.
18376 Set graph background color. Default is white.
18381 Available values for mode is:
18388 Default is @code{line}.
18393 Available values for slide is:
18396 Draw new frame when right border is reached.
18399 Replace old columns with new ones.
18402 Scroll from right to left.
18405 Scroll from left to right.
18408 Draw single picture.
18411 Default is @code{frame}.
18414 Set size of graph video. For the syntax of this option, check the
18415 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18416 The default value is @code{900x256}.
18418 The foreground color expressions can use the following variables:
18421 Minimal value of metadata value.
18424 Maximal value of metadata value.
18427 Current metadata key value.
18430 The color is defined as 0xAABBGGRR.
18433 Example using metadata from @ref{signalstats} filter:
18435 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
18438 Example using metadata from @ref{ebur128} filter:
18440 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
18446 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
18447 it unchanged. By default, it logs a message at a frequency of 10Hz with the
18448 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
18449 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
18451 The filter also has a video output (see the @var{video} option) with a real
18452 time graph to observe the loudness evolution. The graphic contains the logged
18453 message mentioned above, so it is not printed anymore when this option is set,
18454 unless the verbose logging is set. The main graphing area contains the
18455 short-term loudness (3 seconds of analysis), and the gauge on the right is for
18456 the momentary loudness (400 milliseconds).
18458 More information about the Loudness Recommendation EBU R128 on
18459 @url{http://tech.ebu.ch/loudness}.
18461 The filter accepts the following options:
18466 Activate the video output. The audio stream is passed unchanged whether this
18467 option is set or no. The video stream will be the first output stream if
18468 activated. Default is @code{0}.
18471 Set the video size. This option is for video only. For the syntax of this
18473 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18474 Default and minimum resolution is @code{640x480}.
18477 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
18478 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
18479 other integer value between this range is allowed.
18482 Set metadata injection. If set to @code{1}, the audio input will be segmented
18483 into 100ms output frames, each of them containing various loudness information
18484 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
18486 Default is @code{0}.
18489 Force the frame logging level.
18491 Available values are:
18494 information logging level
18496 verbose logging level
18499 By default, the logging level is set to @var{info}. If the @option{video} or
18500 the @option{metadata} options are set, it switches to @var{verbose}.
18505 Available modes can be cumulated (the option is a @code{flag} type). Possible
18509 Disable any peak mode (default).
18511 Enable sample-peak mode.
18513 Simple peak mode looking for the higher sample value. It logs a message
18514 for sample-peak (identified by @code{SPK}).
18516 Enable true-peak mode.
18518 If enabled, the peak lookup is done on an over-sampled version of the input
18519 stream for better peak accuracy. It logs a message for true-peak.
18520 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
18521 This mode requires a build with @code{libswresample}.
18525 Treat mono input files as "dual mono". If a mono file is intended for playback
18526 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
18527 If set to @code{true}, this option will compensate for this effect.
18528 Multi-channel input files are not affected by this option.
18531 Set a specific pan law to be used for the measurement of dual mono files.
18532 This parameter is optional, and has a default value of -3.01dB.
18535 @subsection Examples
18539 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
18541 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
18545 Run an analysis with @command{ffmpeg}:
18547 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
18551 @section interleave, ainterleave
18553 Temporally interleave frames from several inputs.
18555 @code{interleave} works with video inputs, @code{ainterleave} with audio.
18557 These filters read frames from several inputs and send the oldest
18558 queued frame to the output.
18560 Input streams must have well defined, monotonically increasing frame
18563 In order to submit one frame to output, these filters need to enqueue
18564 at least one frame for each input, so they cannot work in case one
18565 input is not yet terminated and will not receive incoming frames.
18567 For example consider the case when one input is a @code{select} filter
18568 which always drops input frames. The @code{interleave} filter will keep
18569 reading from that input, but it will never be able to send new frames
18570 to output until the input sends an end-of-stream signal.
18572 Also, depending on inputs synchronization, the filters will drop
18573 frames in case one input receives more frames than the other ones, and
18574 the queue is already filled.
18576 These filters accept the following options:
18580 Set the number of different inputs, it is 2 by default.
18583 @subsection Examples
18587 Interleave frames belonging to different streams using @command{ffmpeg}:
18589 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
18593 Add flickering blur effect:
18595 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
18599 @section metadata, ametadata
18601 Manipulate frame metadata.
18603 This filter accepts the following options:
18607 Set mode of operation of the filter.
18609 Can be one of the following:
18613 If both @code{value} and @code{key} is set, select frames
18614 which have such metadata. If only @code{key} is set, select
18615 every frame that has such key in metadata.
18618 Add new metadata @code{key} and @code{value}. If key is already available
18622 Modify value of already present key.
18625 If @code{value} is set, delete only keys that have such value.
18626 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
18630 Print key and its value if metadata was found. If @code{key} is not set print all
18631 metadata values available in frame.
18635 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
18638 Set metadata value which will be used. This option is mandatory for
18639 @code{modify} and @code{add} mode.
18642 Which function to use when comparing metadata value and @code{value}.
18644 Can be one of following:
18648 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
18651 Values are interpreted as strings, returns true if metadata value starts with
18652 the @code{value} option string.
18655 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
18658 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
18661 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
18664 Values are interpreted as floats, returns true if expression from option @code{expr}
18669 Set expression which is used when @code{function} is set to @code{expr}.
18670 The expression is evaluated through the eval API and can contain the following
18675 Float representation of @code{value} from metadata key.
18678 Float representation of @code{value} as supplied by user in @code{value} option.
18682 If specified in @code{print} mode, output is written to the named file. Instead of
18683 plain filename any writable url can be specified. Filename ``-'' is a shorthand
18684 for standard output. If @code{file} option is not set, output is written to the log
18685 with AV_LOG_INFO loglevel.
18689 @subsection Examples
18693 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
18696 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
18699 Print silencedetect output to file @file{metadata.txt}.
18701 silencedetect,ametadata=mode=print:file=metadata.txt
18704 Direct all metadata to a pipe with file descriptor 4.
18706 metadata=mode=print:file='pipe\:4'
18710 @section perms, aperms
18712 Set read/write permissions for the output frames.
18714 These filters are mainly aimed at developers to test direct path in the
18715 following filter in the filtergraph.
18717 The filters accept the following options:
18721 Select the permissions mode.
18723 It accepts the following values:
18726 Do nothing. This is the default.
18728 Set all the output frames read-only.
18730 Set all the output frames directly writable.
18732 Make the frame read-only if writable, and writable if read-only.
18734 Set each output frame read-only or writable randomly.
18738 Set the seed for the @var{random} mode, must be an integer included between
18739 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
18740 @code{-1}, the filter will try to use a good random seed on a best effort
18744 Note: in case of auto-inserted filter between the permission filter and the
18745 following one, the permission might not be received as expected in that
18746 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
18747 perms/aperms filter can avoid this problem.
18749 @section realtime, arealtime
18751 Slow down filtering to match real time approximately.
18753 These filters will pause the filtering for a variable amount of time to
18754 match the output rate with the input timestamps.
18755 They are similar to the @option{re} option to @code{ffmpeg}.
18757 They accept the following options:
18761 Time limit for the pauses. Any pause longer than that will be considered
18762 a timestamp discontinuity and reset the timer. Default is 2 seconds.
18766 @section select, aselect
18768 Select frames to pass in output.
18770 This filter accepts the following options:
18775 Set expression, which is evaluated for each input frame.
18777 If the expression is evaluated to zero, the frame is discarded.
18779 If the evaluation result is negative or NaN, the frame is sent to the
18780 first output; otherwise it is sent to the output with index
18781 @code{ceil(val)-1}, assuming that the input index starts from 0.
18783 For example a value of @code{1.2} corresponds to the output with index
18784 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
18787 Set the number of outputs. The output to which to send the selected
18788 frame is based on the result of the evaluation. Default value is 1.
18791 The expression can contain the following constants:
18795 The (sequential) number of the filtered frame, starting from 0.
18798 The (sequential) number of the selected frame, starting from 0.
18800 @item prev_selected_n
18801 The sequential number of the last selected frame. It's NAN if undefined.
18804 The timebase of the input timestamps.
18807 The PTS (Presentation TimeStamp) of the filtered video frame,
18808 expressed in @var{TB} units. It's NAN if undefined.
18811 The PTS of the filtered video frame,
18812 expressed in seconds. It's NAN if undefined.
18815 The PTS of the previously filtered video frame. It's NAN if undefined.
18817 @item prev_selected_pts
18818 The PTS of the last previously filtered video frame. It's NAN if undefined.
18820 @item prev_selected_t
18821 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
18824 The PTS of the first video frame in the video. It's NAN if undefined.
18827 The time of the first video frame in the video. It's NAN if undefined.
18829 @item pict_type @emph{(video only)}
18830 The type of the filtered frame. It can assume one of the following
18842 @item interlace_type @emph{(video only)}
18843 The frame interlace type. It can assume one of the following values:
18846 The frame is progressive (not interlaced).
18848 The frame is top-field-first.
18850 The frame is bottom-field-first.
18853 @item consumed_sample_n @emph{(audio only)}
18854 the number of selected samples before the current frame
18856 @item samples_n @emph{(audio only)}
18857 the number of samples in the current frame
18859 @item sample_rate @emph{(audio only)}
18860 the input sample rate
18863 This is 1 if the filtered frame is a key-frame, 0 otherwise.
18866 the position in the file of the filtered frame, -1 if the information
18867 is not available (e.g. for synthetic video)
18869 @item scene @emph{(video only)}
18870 value between 0 and 1 to indicate a new scene; a low value reflects a low
18871 probability for the current frame to introduce a new scene, while a higher
18872 value means the current frame is more likely to be one (see the example below)
18874 @item concatdec_select
18875 The concat demuxer can select only part of a concat input file by setting an
18876 inpoint and an outpoint, but the output packets may not be entirely contained
18877 in the selected interval. By using this variable, it is possible to skip frames
18878 generated by the concat demuxer which are not exactly contained in the selected
18881 This works by comparing the frame pts against the @var{lavf.concat.start_time}
18882 and the @var{lavf.concat.duration} packet metadata values which are also
18883 present in the decoded frames.
18885 The @var{concatdec_select} variable is -1 if the frame pts is at least
18886 start_time and either the duration metadata is missing or the frame pts is less
18887 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
18890 That basically means that an input frame is selected if its pts is within the
18891 interval set by the concat demuxer.
18895 The default value of the select expression is "1".
18897 @subsection Examples
18901 Select all frames in input:
18906 The example above is the same as:
18918 Select only I-frames:
18920 select='eq(pict_type\,I)'
18924 Select one frame every 100:
18926 select='not(mod(n\,100))'
18930 Select only frames contained in the 10-20 time interval:
18932 select=between(t\,10\,20)
18936 Select only I-frames contained in the 10-20 time interval:
18938 select=between(t\,10\,20)*eq(pict_type\,I)
18942 Select frames with a minimum distance of 10 seconds:
18944 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
18948 Use aselect to select only audio frames with samples number > 100:
18950 aselect='gt(samples_n\,100)'
18954 Create a mosaic of the first scenes:
18956 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
18959 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
18963 Send even and odd frames to separate outputs, and compose them:
18965 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
18969 Select useful frames from an ffconcat file which is using inpoints and
18970 outpoints but where the source files are not intra frame only.
18972 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
18976 @section sendcmd, asendcmd
18978 Send commands to filters in the filtergraph.
18980 These filters read commands to be sent to other filters in the
18983 @code{sendcmd} must be inserted between two video filters,
18984 @code{asendcmd} must be inserted between two audio filters, but apart
18985 from that they act the same way.
18987 The specification of commands can be provided in the filter arguments
18988 with the @var{commands} option, or in a file specified by the
18989 @var{filename} option.
18991 These filters accept the following options:
18994 Set the commands to be read and sent to the other filters.
18996 Set the filename of the commands to be read and sent to the other
19000 @subsection Commands syntax
19002 A commands description consists of a sequence of interval
19003 specifications, comprising a list of commands to be executed when a
19004 particular event related to that interval occurs. The occurring event
19005 is typically the current frame time entering or leaving a given time
19008 An interval is specified by the following syntax:
19010 @var{START}[-@var{END}] @var{COMMANDS};
19013 The time interval is specified by the @var{START} and @var{END} times.
19014 @var{END} is optional and defaults to the maximum time.
19016 The current frame time is considered within the specified interval if
19017 it is included in the interval [@var{START}, @var{END}), that is when
19018 the time is greater or equal to @var{START} and is lesser than
19021 @var{COMMANDS} consists of a sequence of one or more command
19022 specifications, separated by ",", relating to that interval. The
19023 syntax of a command specification is given by:
19025 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
19028 @var{FLAGS} is optional and specifies the type of events relating to
19029 the time interval which enable sending the specified command, and must
19030 be a non-null sequence of identifier flags separated by "+" or "|" and
19031 enclosed between "[" and "]".
19033 The following flags are recognized:
19036 The command is sent when the current frame timestamp enters the
19037 specified interval. In other words, the command is sent when the
19038 previous frame timestamp was not in the given interval, and the
19042 The command is sent when the current frame timestamp leaves the
19043 specified interval. In other words, the command is sent when the
19044 previous frame timestamp was in the given interval, and the
19048 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
19051 @var{TARGET} specifies the target of the command, usually the name of
19052 the filter class or a specific filter instance name.
19054 @var{COMMAND} specifies the name of the command for the target filter.
19056 @var{ARG} is optional and specifies the optional list of argument for
19057 the given @var{COMMAND}.
19059 Between one interval specification and another, whitespaces, or
19060 sequences of characters starting with @code{#} until the end of line,
19061 are ignored and can be used to annotate comments.
19063 A simplified BNF description of the commands specification syntax
19066 @var{COMMAND_FLAG} ::= "enter" | "leave"
19067 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
19068 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
19069 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
19070 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
19071 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
19074 @subsection Examples
19078 Specify audio tempo change at second 4:
19080 asendcmd=c='4.0 atempo tempo 1.5',atempo
19084 Target a specific filter instance:
19086 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
19090 Specify a list of drawtext and hue commands in a file.
19092 # show text in the interval 5-10
19093 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
19094 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
19096 # desaturate the image in the interval 15-20
19097 15.0-20.0 [enter] hue s 0,
19098 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
19100 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
19102 # apply an exponential saturation fade-out effect, starting from time 25
19103 25 [enter] hue s exp(25-t)
19106 A filtergraph allowing to read and process the above command list
19107 stored in a file @file{test.cmd}, can be specified with:
19109 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
19114 @section setpts, asetpts
19116 Change the PTS (presentation timestamp) of the input frames.
19118 @code{setpts} works on video frames, @code{asetpts} on audio frames.
19120 This filter accepts the following options:
19125 The expression which is evaluated for each frame to construct its timestamp.
19129 The expression is evaluated through the eval API and can contain the following
19134 frame rate, only defined for constant frame-rate video
19137 The presentation timestamp in input
19140 The count of the input frame for video or the number of consumed samples,
19141 not including the current frame for audio, starting from 0.
19143 @item NB_CONSUMED_SAMPLES
19144 The number of consumed samples, not including the current frame (only
19147 @item NB_SAMPLES, S
19148 The number of samples in the current frame (only audio)
19150 @item SAMPLE_RATE, SR
19151 The audio sample rate.
19154 The PTS of the first frame.
19157 the time in seconds of the first frame
19160 State whether the current frame is interlaced.
19163 the time in seconds of the current frame
19166 original position in the file of the frame, or undefined if undefined
19167 for the current frame
19170 The previous input PTS.
19173 previous input time in seconds
19176 The previous output PTS.
19179 previous output time in seconds
19182 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
19186 The wallclock (RTC) time at the start of the movie in microseconds.
19189 The timebase of the input timestamps.
19193 @subsection Examples
19197 Start counting PTS from zero
19199 setpts=PTS-STARTPTS
19203 Apply fast motion effect:
19209 Apply slow motion effect:
19215 Set fixed rate of 25 frames per second:
19221 Set fixed rate 25 fps with some jitter:
19223 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
19227 Apply an offset of 10 seconds to the input PTS:
19233 Generate timestamps from a "live source" and rebase onto the current timebase:
19235 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
19239 Generate timestamps by counting samples:
19248 Force color range for the output video frame.
19250 The @code{setrange} filter marks the color range property for the
19251 output frames. It does not change the input frame, but only sets the
19252 corresponding property, which affects how the frame is treated by
19255 The filter accepts the following options:
19260 Available values are:
19264 Keep the same color range property.
19266 @item unspecified, unknown
19267 Set the color range as unspecified.
19269 @item limited, tv, mpeg
19270 Set the color range as limited.
19272 @item full, pc, jpeg
19273 Set the color range as full.
19277 @section settb, asettb
19279 Set the timebase to use for the output frames timestamps.
19280 It is mainly useful for testing timebase configuration.
19282 It accepts the following parameters:
19287 The expression which is evaluated into the output timebase.
19291 The value for @option{tb} is an arithmetic expression representing a
19292 rational. The expression can contain the constants "AVTB" (the default
19293 timebase), "intb" (the input timebase) and "sr" (the sample rate,
19294 audio only). Default value is "intb".
19296 @subsection Examples
19300 Set the timebase to 1/25:
19306 Set the timebase to 1/10:
19312 Set the timebase to 1001/1000:
19318 Set the timebase to 2*intb:
19324 Set the default timebase value:
19331 Convert input audio to a video output representing frequency spectrum
19332 logarithmically using Brown-Puckette constant Q transform algorithm with
19333 direct frequency domain coefficient calculation (but the transform itself
19334 is not really constant Q, instead the Q factor is actually variable/clamped),
19335 with musical tone scale, from E0 to D#10.
19337 The filter accepts the following options:
19341 Specify the video size for the output. It must be even. For the syntax of this option,
19342 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19343 Default value is @code{1920x1080}.
19346 Set the output frame rate. Default value is @code{25}.
19349 Set the bargraph height. It must be even. Default value is @code{-1} which
19350 computes the bargraph height automatically.
19353 Set the axis height. It must be even. Default value is @code{-1} which computes
19354 the axis height automatically.
19357 Set the sonogram height. It must be even. Default value is @code{-1} which
19358 computes the sonogram height automatically.
19361 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
19362 instead. Default value is @code{1}.
19364 @item sono_v, volume
19365 Specify the sonogram volume expression. It can contain variables:
19368 the @var{bar_v} evaluated expression
19369 @item frequency, freq, f
19370 the frequency where it is evaluated
19371 @item timeclamp, tc
19372 the value of @var{timeclamp} option
19376 @item a_weighting(f)
19377 A-weighting of equal loudness
19378 @item b_weighting(f)
19379 B-weighting of equal loudness
19380 @item c_weighting(f)
19381 C-weighting of equal loudness.
19383 Default value is @code{16}.
19385 @item bar_v, volume2
19386 Specify the bargraph volume expression. It can contain variables:
19389 the @var{sono_v} evaluated expression
19390 @item frequency, freq, f
19391 the frequency where it is evaluated
19392 @item timeclamp, tc
19393 the value of @var{timeclamp} option
19397 @item a_weighting(f)
19398 A-weighting of equal loudness
19399 @item b_weighting(f)
19400 B-weighting of equal loudness
19401 @item c_weighting(f)
19402 C-weighting of equal loudness.
19404 Default value is @code{sono_v}.
19406 @item sono_g, gamma
19407 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
19408 higher gamma makes the spectrum having more range. Default value is @code{3}.
19409 Acceptable range is @code{[1, 7]}.
19411 @item bar_g, gamma2
19412 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
19416 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
19417 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
19419 @item timeclamp, tc
19420 Specify the transform timeclamp. At low frequency, there is trade-off between
19421 accuracy in time domain and frequency domain. If timeclamp is lower,
19422 event in time domain is represented more accurately (such as fast bass drum),
19423 otherwise event in frequency domain is represented more accurately
19424 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
19427 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
19428 limits future samples by applying asymmetric windowing in time domain, useful
19429 when low latency is required. Accepted range is @code{[0, 1]}.
19432 Specify the transform base frequency. Default value is @code{20.01523126408007475},
19433 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
19436 Specify the transform end frequency. Default value is @code{20495.59681441799654},
19437 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
19440 This option is deprecated and ignored.
19443 Specify the transform length in time domain. Use this option to control accuracy
19444 trade-off between time domain and frequency domain at every frequency sample.
19445 It can contain variables:
19447 @item frequency, freq, f
19448 the frequency where it is evaluated
19449 @item timeclamp, tc
19450 the value of @var{timeclamp} option.
19452 Default value is @code{384*tc/(384+tc*f)}.
19455 Specify the transform count for every video frame. Default value is @code{6}.
19456 Acceptable range is @code{[1, 30]}.
19459 Specify the transform count for every single pixel. Default value is @code{0},
19460 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
19463 Specify font file for use with freetype to draw the axis. If not specified,
19464 use embedded font. Note that drawing with font file or embedded font is not
19465 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
19469 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
19470 The : in the pattern may be replaced by | to avoid unnecessary escaping.
19473 Specify font color expression. This is arithmetic expression that should return
19474 integer value 0xRRGGBB. It can contain variables:
19476 @item frequency, freq, f
19477 the frequency where it is evaluated
19478 @item timeclamp, tc
19479 the value of @var{timeclamp} option
19484 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
19485 @item r(x), g(x), b(x)
19486 red, green, and blue value of intensity x.
19488 Default value is @code{st(0, (midi(f)-59.5)/12);
19489 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
19490 r(1-ld(1)) + b(ld(1))}.
19493 Specify image file to draw the axis. This option override @var{fontfile} and
19494 @var{fontcolor} option.
19497 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
19498 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
19499 Default value is @code{1}.
19502 Set colorspace. The accepted values are:
19505 Unspecified (default)
19514 BT.470BG or BT.601-6 625
19517 SMPTE-170M or BT.601-6 525
19523 BT.2020 with non-constant luminance
19528 Set spectrogram color scheme. This is list of floating point values with format
19529 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
19530 The default is @code{1|0.5|0|0|0.5|1}.
19534 @subsection Examples
19538 Playing audio while showing the spectrum:
19540 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
19544 Same as above, but with frame rate 30 fps:
19546 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
19550 Playing at 1280x720:
19552 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
19556 Disable sonogram display:
19562 A1 and its harmonics: A1, A2, (near)E3, A3:
19564 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),
19565 asplit[a][out1]; [a] showcqt [out0]'
19569 Same as above, but with more accuracy in frequency domain:
19571 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),
19572 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
19578 bar_v=10:sono_v=bar_v*a_weighting(f)
19582 Custom gamma, now spectrum is linear to the amplitude.
19588 Custom tlength equation:
19590 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)))'
19594 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
19596 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
19600 Custom font using fontconfig:
19602 font='Courier New,Monospace,mono|bold'
19606 Custom frequency range with custom axis using image file:
19608 axisfile=myaxis.png:basefreq=40:endfreq=10000
19614 Convert input audio to video output representing the audio power spectrum.
19615 Audio amplitude is on Y-axis while frequency is on X-axis.
19617 The filter accepts the following options:
19621 Specify size of video. For the syntax of this option, check the
19622 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19623 Default is @code{1024x512}.
19627 This set how each frequency bin will be represented.
19629 It accepts the following values:
19635 Default is @code{bar}.
19638 Set amplitude scale.
19640 It accepts the following values:
19654 Default is @code{log}.
19657 Set frequency scale.
19659 It accepts the following values:
19668 Reverse logarithmic scale.
19670 Default is @code{lin}.
19675 It accepts the following values:
19691 Default is @code{w2048}
19694 Set windowing function.
19696 It accepts the following values:
19718 Default is @code{hanning}.
19721 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
19722 which means optimal overlap for selected window function will be picked.
19725 Set time averaging. Setting this to 0 will display current maximal peaks.
19726 Default is @code{1}, which means time averaging is disabled.
19729 Specify list of colors separated by space or by '|' which will be used to
19730 draw channel frequencies. Unrecognized or missing colors will be replaced
19734 Set channel display mode.
19736 It accepts the following values:
19741 Default is @code{combined}.
19744 Set minimum amplitude used in @code{log} amplitude scaler.
19748 @anchor{showspectrum}
19749 @section showspectrum
19751 Convert input audio to a video output, representing the audio frequency
19754 The filter accepts the following options:
19758 Specify the video size for the output. For the syntax of this option, check the
19759 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19760 Default value is @code{640x512}.
19763 Specify how the spectrum should slide along the window.
19765 It accepts the following values:
19768 the samples start again on the left when they reach the right
19770 the samples scroll from right to left
19772 frames are only produced when the samples reach the right
19774 the samples scroll from left to right
19777 Default value is @code{replace}.
19780 Specify display mode.
19782 It accepts the following values:
19785 all channels are displayed in the same row
19787 all channels are displayed in separate rows
19790 Default value is @samp{combined}.
19793 Specify display color mode.
19795 It accepts the following values:
19798 each channel is displayed in a separate color
19800 each channel is displayed using the same color scheme
19802 each channel is displayed using the rainbow color scheme
19804 each channel is displayed using the moreland color scheme
19806 each channel is displayed using the nebulae color scheme
19808 each channel is displayed using the fire color scheme
19810 each channel is displayed using the fiery color scheme
19812 each channel is displayed using the fruit color scheme
19814 each channel is displayed using the cool color scheme
19817 Default value is @samp{channel}.
19820 Specify scale used for calculating intensity color values.
19822 It accepts the following values:
19827 square root, default
19838 Default value is @samp{sqrt}.
19841 Set saturation modifier for displayed colors. Negative values provide
19842 alternative color scheme. @code{0} is no saturation at all.
19843 Saturation must be in [-10.0, 10.0] range.
19844 Default value is @code{1}.
19847 Set window function.
19849 It accepts the following values:
19873 Default value is @code{hann}.
19876 Set orientation of time vs frequency axis. Can be @code{vertical} or
19877 @code{horizontal}. Default is @code{vertical}.
19880 Set ratio of overlap window. Default value is @code{0}.
19881 When value is @code{1} overlap is set to recommended size for specific
19882 window function currently used.
19885 Set scale gain for calculating intensity color values.
19886 Default value is @code{1}.
19889 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
19892 Set color rotation, must be in [-1.0, 1.0] range.
19893 Default value is @code{0}.
19896 The usage is very similar to the showwaves filter; see the examples in that
19899 @subsection Examples
19903 Large window with logarithmic color scaling:
19905 showspectrum=s=1280x480:scale=log
19909 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
19911 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
19912 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
19916 @section showspectrumpic
19918 Convert input audio to a single video frame, representing the audio frequency
19921 The filter accepts the following options:
19925 Specify the video size for the output. For the syntax of this option, check the
19926 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19927 Default value is @code{4096x2048}.
19930 Specify display mode.
19932 It accepts the following values:
19935 all channels are displayed in the same row
19937 all channels are displayed in separate rows
19939 Default value is @samp{combined}.
19942 Specify display color mode.
19944 It accepts the following values:
19947 each channel is displayed in a separate color
19949 each channel is displayed using the same color scheme
19951 each channel is displayed using the rainbow color scheme
19953 each channel is displayed using the moreland color scheme
19955 each channel is displayed using the nebulae color scheme
19957 each channel is displayed using the fire color scheme
19959 each channel is displayed using the fiery color scheme
19961 each channel is displayed using the fruit color scheme
19963 each channel is displayed using the cool color scheme
19965 Default value is @samp{intensity}.
19968 Specify scale used for calculating intensity color values.
19970 It accepts the following values:
19975 square root, default
19985 Default value is @samp{log}.
19988 Set saturation modifier for displayed colors. Negative values provide
19989 alternative color scheme. @code{0} is no saturation at all.
19990 Saturation must be in [-10.0, 10.0] range.
19991 Default value is @code{1}.
19994 Set window function.
19996 It accepts the following values:
20019 Default value is @code{hann}.
20022 Set orientation of time vs frequency axis. Can be @code{vertical} or
20023 @code{horizontal}. Default is @code{vertical}.
20026 Set scale gain for calculating intensity color values.
20027 Default value is @code{1}.
20030 Draw time and frequency axes and legends. Default is enabled.
20033 Set color rotation, must be in [-1.0, 1.0] range.
20034 Default value is @code{0}.
20037 @subsection Examples
20041 Extract an audio spectrogram of a whole audio track
20042 in a 1024x1024 picture using @command{ffmpeg}:
20044 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
20048 @section showvolume
20050 Convert input audio volume to a video output.
20052 The filter accepts the following options:
20059 Set border width, allowed range is [0, 5]. Default is 1.
20062 Set channel width, allowed range is [80, 8192]. Default is 400.
20065 Set channel height, allowed range is [1, 900]. Default is 20.
20068 Set fade, allowed range is [0, 1]. Default is 0.95.
20071 Set volume color expression.
20073 The expression can use the following variables:
20077 Current max volume of channel in dB.
20083 Current channel number, starting from 0.
20087 If set, displays channel names. Default is enabled.
20090 If set, displays volume values. Default is enabled.
20093 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
20094 default is @code{h}.
20097 Set step size, allowed range is [0, 5]. Default is 0, which means
20101 Set background opacity, allowed range is [0, 1]. Default is 0.
20104 Set metering mode, can be peak: @code{p} or rms: @code{r},
20105 default is @code{p}.
20108 Set display scale, can be linear: @code{lin} or log: @code{log},
20109 default is @code{lin}.
20113 If set to > 0., display a line for the max level
20114 in the previous seconds.
20115 default is disabled: @code{0.}
20118 The color of the max line. Use when @code{dm} option is set to > 0.
20119 default is: @code{orange}
20124 Convert input audio to a video output, representing the samples waves.
20126 The filter accepts the following options:
20130 Specify the video size for the output. For the syntax of this option, check the
20131 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20132 Default value is @code{600x240}.
20137 Available values are:
20140 Draw a point for each sample.
20143 Draw a vertical line for each sample.
20146 Draw a point for each sample and a line between them.
20149 Draw a centered vertical line for each sample.
20152 Default value is @code{point}.
20155 Set the number of samples which are printed on the same column. A
20156 larger value will decrease the frame rate. Must be a positive
20157 integer. This option can be set only if the value for @var{rate}
20158 is not explicitly specified.
20161 Set the (approximate) output frame rate. This is done by setting the
20162 option @var{n}. Default value is "25".
20164 @item split_channels
20165 Set if channels should be drawn separately or overlap. Default value is 0.
20168 Set colors separated by '|' which are going to be used for drawing of each channel.
20171 Set amplitude scale.
20173 Available values are:
20191 Set the draw mode. This is mostly useful to set for high @var{n}.
20193 Available values are:
20196 Scale pixel values for each drawn sample.
20199 Draw every sample directly.
20202 Default value is @code{scale}.
20205 @subsection Examples
20209 Output the input file audio and the corresponding video representation
20212 amovie=a.mp3,asplit[out0],showwaves[out1]
20216 Create a synthetic signal and show it with showwaves, forcing a
20217 frame rate of 30 frames per second:
20219 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
20223 @section showwavespic
20225 Convert input audio to a single video frame, representing the samples waves.
20227 The filter accepts the following options:
20231 Specify the video size for the output. For the syntax of this option, check the
20232 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20233 Default value is @code{600x240}.
20235 @item split_channels
20236 Set if channels should be drawn separately or overlap. Default value is 0.
20239 Set colors separated by '|' which are going to be used for drawing of each channel.
20242 Set amplitude scale.
20244 Available values are:
20262 @subsection Examples
20266 Extract a channel split representation of the wave form of a whole audio track
20267 in a 1024x800 picture using @command{ffmpeg}:
20269 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
20273 @section sidedata, asidedata
20275 Delete frame side data, or select frames based on it.
20277 This filter accepts the following options:
20281 Set mode of operation of the filter.
20283 Can be one of the following:
20287 Select every frame with side data of @code{type}.
20290 Delete side data of @code{type}. If @code{type} is not set, delete all side
20296 Set side data type used with all modes. Must be set for @code{select} mode. For
20297 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
20298 in @file{libavutil/frame.h}. For example, to choose
20299 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
20303 @section spectrumsynth
20305 Sythesize audio from 2 input video spectrums, first input stream represents
20306 magnitude across time and second represents phase across time.
20307 The filter will transform from frequency domain as displayed in videos back
20308 to time domain as presented in audio output.
20310 This filter is primarily created for reversing processed @ref{showspectrum}
20311 filter outputs, but can synthesize sound from other spectrograms too.
20312 But in such case results are going to be poor if the phase data is not
20313 available, because in such cases phase data need to be recreated, usually
20314 its just recreated from random noise.
20315 For best results use gray only output (@code{channel} color mode in
20316 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
20317 @code{lin} scale for phase video. To produce phase, for 2nd video, use
20318 @code{data} option. Inputs videos should generally use @code{fullframe}
20319 slide mode as that saves resources needed for decoding video.
20321 The filter accepts the following options:
20325 Specify sample rate of output audio, the sample rate of audio from which
20326 spectrum was generated may differ.
20329 Set number of channels represented in input video spectrums.
20332 Set scale which was used when generating magnitude input spectrum.
20333 Can be @code{lin} or @code{log}. Default is @code{log}.
20336 Set slide which was used when generating inputs spectrums.
20337 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
20338 Default is @code{fullframe}.
20341 Set window function used for resynthesis.
20344 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
20345 which means optimal overlap for selected window function will be picked.
20348 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
20349 Default is @code{vertical}.
20352 @subsection Examples
20356 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
20357 then resynthesize videos back to audio with spectrumsynth:
20359 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
20360 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
20361 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
20365 @section split, asplit
20367 Split input into several identical outputs.
20369 @code{asplit} works with audio input, @code{split} with video.
20371 The filter accepts a single parameter which specifies the number of outputs. If
20372 unspecified, it defaults to 2.
20374 @subsection Examples
20378 Create two separate outputs from the same input:
20380 [in] split [out0][out1]
20384 To create 3 or more outputs, you need to specify the number of
20387 [in] asplit=3 [out0][out1][out2]
20391 Create two separate outputs from the same input, one cropped and
20394 [in] split [splitout1][splitout2];
20395 [splitout1] crop=100:100:0:0 [cropout];
20396 [splitout2] pad=200:200:100:100 [padout];
20400 Create 5 copies of the input audio with @command{ffmpeg}:
20402 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
20408 Receive commands sent through a libzmq client, and forward them to
20409 filters in the filtergraph.
20411 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
20412 must be inserted between two video filters, @code{azmq} between two
20413 audio filters. Both are capable to send messages to any filter type.
20415 To enable these filters you need to install the libzmq library and
20416 headers and configure FFmpeg with @code{--enable-libzmq}.
20418 For more information about libzmq see:
20419 @url{http://www.zeromq.org/}
20421 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
20422 receives messages sent through a network interface defined by the
20423 @option{bind_address} (or the abbreviation "@option{b}") option.
20424 Default value of this option is @file{tcp://localhost:5555}. You may
20425 want to alter this value to your needs, but do not forget to escape any
20426 ':' signs (see @ref{filtergraph escaping}).
20428 The received message must be in the form:
20430 @var{TARGET} @var{COMMAND} [@var{ARG}]
20433 @var{TARGET} specifies the target of the command, usually the name of
20434 the filter class or a specific filter instance name. The default
20435 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
20436 but you can override this by using the @samp{filter_name@@id} syntax
20437 (see @ref{Filtergraph syntax}).
20439 @var{COMMAND} specifies the name of the command for the target filter.
20441 @var{ARG} is optional and specifies the optional argument list for the
20442 given @var{COMMAND}.
20444 Upon reception, the message is processed and the corresponding command
20445 is injected into the filtergraph. Depending on the result, the filter
20446 will send a reply to the client, adopting the format:
20448 @var{ERROR_CODE} @var{ERROR_REASON}
20452 @var{MESSAGE} is optional.
20454 @subsection Examples
20456 Look at @file{tools/zmqsend} for an example of a zmq client which can
20457 be used to send commands processed by these filters.
20459 Consider the following filtergraph generated by @command{ffplay}.
20460 In this example the last overlay filter has an instance name. All other
20461 filters will have default instance names.
20464 ffplay -dumpgraph 1 -f lavfi "
20465 color=s=100x100:c=red [l];
20466 color=s=100x100:c=blue [r];
20467 nullsrc=s=200x100, zmq [bg];
20468 [bg][l] overlay [bg+l];
20469 [bg+l][r] overlay@@my=x=100 "
20472 To change the color of the left side of the video, the following
20473 command can be used:
20475 echo Parsed_color_0 c yellow | tools/zmqsend
20478 To change the right side:
20480 echo Parsed_color_1 c pink | tools/zmqsend
20483 To change the position of the right side:
20485 echo overlay@@my x 150 | tools/zmqsend
20489 @c man end MULTIMEDIA FILTERS
20491 @chapter Multimedia Sources
20492 @c man begin MULTIMEDIA SOURCES
20494 Below is a description of the currently available multimedia sources.
20498 This is the same as @ref{movie} source, except it selects an audio
20504 Read audio and/or video stream(s) from a movie container.
20506 It accepts the following parameters:
20510 The name of the resource to read (not necessarily a file; it can also be a
20511 device or a stream accessed through some protocol).
20513 @item format_name, f
20514 Specifies the format assumed for the movie to read, and can be either
20515 the name of a container or an input device. If not specified, the
20516 format is guessed from @var{movie_name} or by probing.
20518 @item seek_point, sp
20519 Specifies the seek point in seconds. The frames will be output
20520 starting from this seek point. The parameter is evaluated with
20521 @code{av_strtod}, so the numerical value may be suffixed by an IS
20522 postfix. The default value is "0".
20525 Specifies the streams to read. Several streams can be specified,
20526 separated by "+". The source will then have as many outputs, in the
20527 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
20528 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
20529 respectively the default (best suited) video and audio stream. Default
20530 is "dv", or "da" if the filter is called as "amovie".
20532 @item stream_index, si
20533 Specifies the index of the video stream to read. If the value is -1,
20534 the most suitable video stream will be automatically selected. The default
20535 value is "-1". Deprecated. If the filter is called "amovie", it will select
20536 audio instead of video.
20539 Specifies how many times to read the stream in sequence.
20540 If the value is 0, the stream will be looped infinitely.
20541 Default value is "1".
20543 Note that when the movie is looped the source timestamps are not
20544 changed, so it will generate non monotonically increasing timestamps.
20546 @item discontinuity
20547 Specifies the time difference between frames above which the point is
20548 considered a timestamp discontinuity which is removed by adjusting the later
20552 It allows overlaying a second video on top of the main input of
20553 a filtergraph, as shown in this graph:
20555 input -----------> deltapts0 --> overlay --> output
20558 movie --> scale--> deltapts1 -------+
20560 @subsection Examples
20564 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
20565 on top of the input labelled "in":
20567 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
20568 [in] setpts=PTS-STARTPTS [main];
20569 [main][over] overlay=16:16 [out]
20573 Read from a video4linux2 device, and overlay it on top of the input
20576 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
20577 [in] setpts=PTS-STARTPTS [main];
20578 [main][over] overlay=16:16 [out]
20582 Read the first video stream and the audio stream with id 0x81 from
20583 dvd.vob; the video is connected to the pad named "video" and the audio is
20584 connected to the pad named "audio":
20586 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
20590 @subsection Commands
20592 Both movie and amovie support the following commands:
20595 Perform seek using "av_seek_frame".
20596 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
20599 @var{stream_index}: If stream_index is -1, a default
20600 stream is selected, and @var{timestamp} is automatically converted
20601 from AV_TIME_BASE units to the stream specific time_base.
20603 @var{timestamp}: Timestamp in AVStream.time_base units
20604 or, if no stream is specified, in AV_TIME_BASE units.
20606 @var{flags}: Flags which select direction and seeking mode.
20610 Get movie duration in AV_TIME_BASE units.
20614 @c man end MULTIMEDIA SOURCES