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{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.
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{LINKLABEL} ::= "[" @var{NAME} "]"
216 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
217 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
218 @var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
219 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
220 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
223 @section Notes on filtergraph escaping
225 Filtergraph description composition entails several levels of
226 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
227 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
228 information about the employed escaping procedure.
230 A first level escaping affects the content of each filter option
231 value, which may contain the special character @code{:} used to
232 separate values, or one of the escaping characters @code{\'}.
234 A second level escaping affects the whole filter description, which
235 may contain the escaping characters @code{\'} or the special
236 characters @code{[],;} used by the filtergraph description.
238 Finally, when you specify a filtergraph on a shell commandline, you
239 need to perform a third level escaping for the shell special
240 characters contained within it.
242 For example, consider the following string to be embedded in
243 the @ref{drawtext} filter description @option{text} value:
245 this is a 'string': may contain one, or more, special characters
248 This string contains the @code{'} special escaping character, and the
249 @code{:} special character, so it needs to be escaped in this way:
251 text=this is a \'string\'\: may contain one, or more, special characters
254 A second level of escaping is required when embedding the filter
255 description in a filtergraph description, in order to escape all the
256 filtergraph special characters. Thus the example above becomes:
258 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
260 (note that in addition to the @code{\'} escaping special characters,
261 also @code{,} needs to be escaped).
263 Finally an additional level of escaping is needed when writing the
264 filtergraph description in a shell command, which depends on the
265 escaping rules of the adopted shell. For example, assuming that
266 @code{\} is special and needs to be escaped with another @code{\}, the
267 previous string will finally result in:
269 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
272 @chapter Timeline editing
274 Some filters support a generic @option{enable} option. For the filters
275 supporting timeline editing, this option can be set to an expression which is
276 evaluated before sending a frame to the filter. If the evaluation is non-zero,
277 the filter will be enabled, otherwise the frame will be sent unchanged to the
278 next filter in the filtergraph.
280 The expression accepts the following values:
283 timestamp expressed in seconds, NAN if the input timestamp is unknown
286 sequential number of the input frame, starting from 0
289 the position in the file of the input frame, NAN if unknown
293 width and height of the input frame if video
296 Additionally, these filters support an @option{enable} command that can be used
297 to re-define the expression.
299 Like any other filtering option, the @option{enable} option follows the same
302 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
303 minutes, and a @ref{curves} filter starting at 3 seconds:
305 smartblur = enable='between(t,10,3*60)',
306 curves = enable='gte(t,3)' : preset=cross_process
309 @c man end FILTERGRAPH DESCRIPTION
311 @chapter Audio Filters
312 @c man begin AUDIO FILTERS
314 When you configure your FFmpeg build, you can disable any of the
315 existing filters using @code{--disable-filters}.
316 The configure output will show the audio filters included in your
319 Below is a description of the currently available audio filters.
323 A compressor is mainly used to reduce the dynamic range of a signal.
324 Especially modern music is mostly compressed at a high ratio to
325 improve the overall loudness. It's done to get the highest attention
326 of a listener, "fatten" the sound and bring more "power" to the track.
327 If a signal is compressed too much it may sound dull or "dead"
328 afterwards or it may start to "pump" (which could be a powerful effect
329 but can also destroy a track completely).
330 The right compression is the key to reach a professional sound and is
331 the high art of mixing and mastering. Because of its complex settings
332 it may take a long time to get the right feeling for this kind of effect.
334 Compression is done by detecting the volume above a chosen level
335 @code{threshold} and dividing it by the factor set with @code{ratio}.
336 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
337 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
338 the signal would cause distortion of the waveform the reduction can be
339 levelled over the time. This is done by setting "Attack" and "Release".
340 @code{attack} determines how long the signal has to rise above the threshold
341 before any reduction will occur and @code{release} sets the time the signal
342 has to fall below the threshold to reduce the reduction again. Shorter signals
343 than the chosen attack time will be left untouched.
344 The overall reduction of the signal can be made up afterwards with the
345 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
346 raising the makeup to this level results in a signal twice as loud than the
347 source. To gain a softer entry in the compression the @code{knee} flattens the
348 hard edge at the threshold in the range of the chosen decibels.
350 The filter accepts the following options:
354 Set input gain. Default is 1. Range is between 0.015625 and 64.
357 If a signal of second stream rises above this level it will affect the gain
358 reduction of the first stream.
359 By default it is 0.125. Range is between 0.00097563 and 1.
362 Set a ratio by which the signal is reduced. 1:2 means that if the level
363 rose 4dB above the threshold, it will be only 2dB above after the reduction.
364 Default is 2. Range is between 1 and 20.
367 Amount of milliseconds the signal has to rise above the threshold before gain
368 reduction starts. Default is 20. Range is between 0.01 and 2000.
371 Amount of milliseconds the signal has to fall below the threshold before
372 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
375 Set the amount by how much signal will be amplified after processing.
376 Default is 2. Range is from 1 and 64.
379 Curve the sharp knee around the threshold to enter gain reduction more softly.
380 Default is 2.82843. Range is between 1 and 8.
383 Choose if the @code{average} level between all channels of input stream
384 or the louder(@code{maximum}) channel of input stream affects the
385 reduction. Default is @code{average}.
388 Should the exact signal be taken in case of @code{peak} or an RMS one in case
389 of @code{rms}. Default is @code{rms} which is mostly smoother.
392 How much to use compressed signal in output. Default is 1.
393 Range is between 0 and 1.
398 Apply cross fade from one input audio stream to another input audio stream.
399 The cross fade is applied for specified duration near the end of first stream.
401 The filter accepts the following options:
405 Specify the number of samples for which the cross fade effect has to last.
406 At the end of the cross fade effect the first input audio will be completely
407 silent. Default is 44100.
410 Specify the duration of the cross fade effect. See
411 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
412 for the accepted syntax.
413 By default the duration is determined by @var{nb_samples}.
414 If set this option is used instead of @var{nb_samples}.
417 Should first stream end overlap with second stream start. Default is enabled.
420 Set curve for cross fade transition for first stream.
423 Set curve for cross fade transition for second stream.
425 For description of available curve types see @ref{afade} filter description.
432 Cross fade from one input to another:
434 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
438 Cross fade from one input to another but without overlapping:
440 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
446 Reduce audio bit resolution.
448 This filter is bit crusher with enhanced functionality. A bit crusher
449 is used to audibly reduce number of bits an audio signal is sampled
450 with. This doesn't change the bit depth at all, it just produces the
451 effect. Material reduced in bit depth sounds more harsh and "digital".
452 This filter is able to even round to continous values instead of discrete
454 Additionally it has a D/C offset which results in different crushing of
455 the lower and the upper half of the signal.
456 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
458 Another feature of this filter is the logarithmic mode.
459 This setting switches from linear distances between bits to logarithmic ones.
460 The result is a much more "natural" sounding crusher which doesn't gate low
461 signals for example. The human ear has a logarithmic perception, too
462 so this kind of crushing is much more pleasant.
463 Logarithmic crushing is also able to get anti-aliased.
465 The filter accepts the following options:
481 Can be linear: @code{lin} or logarithmic: @code{log}.
490 Set sample reduction.
493 Enable LFO. By default disabled.
504 Delay one or more audio channels.
506 Samples in delayed channel are filled with silence.
508 The filter accepts the following option:
512 Set list of delays in milliseconds for each channel separated by '|'.
513 At least one delay greater than 0 should be provided.
514 Unused delays will be silently ignored. If number of given delays is
515 smaller than number of channels all remaining channels will not be delayed.
516 If you want to delay exact number of samples, append 'S' to number.
523 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
524 the second channel (and any other channels that may be present) unchanged.
530 Delay second channel by 500 samples, the third channel by 700 samples and leave
531 the first channel (and any other channels that may be present) unchanged.
539 Apply echoing to the input audio.
541 Echoes are reflected sound and can occur naturally amongst mountains
542 (and sometimes large buildings) when talking or shouting; digital echo
543 effects emulate this behaviour and are often used to help fill out the
544 sound of a single instrument or vocal. The time difference between the
545 original signal and the reflection is the @code{delay}, and the
546 loudness of the reflected signal is the @code{decay}.
547 Multiple echoes can have different delays and decays.
549 A description of the accepted parameters follows.
553 Set input gain of reflected signal. Default is @code{0.6}.
556 Set output gain of reflected signal. Default is @code{0.3}.
559 Set list of time intervals in milliseconds between original signal and reflections
560 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
561 Default is @code{1000}.
564 Set list of loudnesses of reflected signals separated by '|'.
565 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
566 Default is @code{0.5}.
573 Make it sound as if there are twice as many instruments as are actually playing:
575 aecho=0.8:0.88:60:0.4
579 If delay is very short, then it sound like a (metallic) robot playing music:
585 A longer delay will sound like an open air concert in the mountains:
587 aecho=0.8:0.9:1000:0.3
591 Same as above but with one more mountain:
593 aecho=0.8:0.9:1000|1800:0.3|0.25
598 Audio emphasis filter creates or restores material directly taken from LPs or
599 emphased CDs with different filter curves. E.g. to store music on vinyl the
600 signal has to be altered by a filter first to even out the disadvantages of
601 this recording medium.
602 Once the material is played back the inverse filter has to be applied to
603 restore the distortion of the frequency response.
605 The filter accepts the following options:
615 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
616 use @code{production} mode. Default is @code{reproduction} mode.
619 Set filter type. Selects medium. Can be one of the following:
631 select Compact Disc (CD).
637 select 50µs (FM-KF).
639 select 75µs (FM-KF).
645 Modify an audio signal according to the specified expressions.
647 This filter accepts one or more expressions (one for each channel),
648 which are evaluated and used to modify a corresponding audio signal.
650 It accepts the following parameters:
654 Set the '|'-separated expressions list for each separate channel. If
655 the number of input channels is greater than the number of
656 expressions, the last specified expression is used for the remaining
659 @item channel_layout, c
660 Set output channel layout. If not specified, the channel layout is
661 specified by the number of expressions. If set to @samp{same}, it will
662 use by default the same input channel layout.
665 Each expression in @var{exprs} can contain the following constants and functions:
669 channel number of the current expression
672 number of the evaluated sample, starting from 0
678 time of the evaluated sample expressed in seconds
681 @item nb_out_channels
682 input and output number of channels
685 the value of input channel with number @var{CH}
688 Note: this filter is slow. For faster processing you should use a
697 aeval=val(ch)/2:c=same
701 Invert phase of the second channel:
710 Apply fade-in/out effect to input audio.
712 A description of the accepted parameters follows.
716 Specify the effect type, can be either @code{in} for fade-in, or
717 @code{out} for a fade-out effect. Default is @code{in}.
719 @item start_sample, ss
720 Specify the number of the start sample for starting to apply the fade
721 effect. Default is 0.
724 Specify the number of samples for which the fade effect has to last. At
725 the end of the fade-in effect the output audio will have the same
726 volume as the input audio, at the end of the fade-out transition
727 the output audio will be silence. Default is 44100.
730 Specify the start time of the fade effect. Default is 0.
731 The value must be specified as a time duration; see
732 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
733 for the accepted syntax.
734 If set this option is used instead of @var{start_sample}.
737 Specify the duration of the fade effect. See
738 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
739 for the accepted syntax.
740 At the end of the fade-in effect the output audio will have the same
741 volume as the input audio, at the end of the fade-out transition
742 the output audio will be silence.
743 By default the duration is determined by @var{nb_samples}.
744 If set this option is used instead of @var{nb_samples}.
747 Set curve for fade transition.
749 It accepts the following values:
752 select triangular, linear slope (default)
754 select quarter of sine wave
756 select half of sine wave
758 select exponential sine wave
762 select inverted parabola
776 select inverted quarter of sine wave
778 select inverted half of sine wave
780 select double-exponential seat
782 select double-exponential sigmoid
790 Fade in first 15 seconds of audio:
796 Fade out last 25 seconds of a 900 seconds audio:
798 afade=t=out:st=875:d=25
803 Apply arbitrary expressions to samples in frequency domain.
807 Set frequency domain real expression for each separate channel separated
808 by '|'. Default is "1".
809 If the number of input channels is greater than the number of
810 expressions, the last specified expression is used for the remaining
814 Set frequency domain imaginary expression for each separate channel
815 separated by '|'. If not set, @var{real} option is used.
817 Each expression in @var{real} and @var{imag} can contain the following
825 current frequency bin number
828 number of available bins
831 channel number of the current expression
843 It accepts the following values:
859 Default is @code{w4096}
862 Set window function. Default is @code{hann}.
865 Set window overlap. If set to 1, the recommended overlap for selected
866 window function will be picked. Default is @code{0.75}.
873 Leave almost only low frequencies in audio:
875 afftfilt="1-clip((b/nb)*b,0,1)"
882 Set output format constraints for the input audio. The framework will
883 negotiate the most appropriate format to minimize conversions.
885 It accepts the following parameters:
889 A '|'-separated list of requested sample formats.
892 A '|'-separated list of requested sample rates.
894 @item channel_layouts
895 A '|'-separated list of requested channel layouts.
897 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
898 for the required syntax.
901 If a parameter is omitted, all values are allowed.
903 Force the output to either unsigned 8-bit or signed 16-bit stereo
905 aformat=sample_fmts=u8|s16:channel_layouts=stereo
910 A gate is mainly used to reduce lower parts of a signal. This kind of signal
911 processing reduces disturbing noise between useful signals.
913 Gating is done by detecting the volume below a chosen level @var{threshold}
914 and divide it by the factor set with @var{ratio}. The bottom of the noise
915 floor is set via @var{range}. Because an exact manipulation of the signal
916 would cause distortion of the waveform the reduction can be levelled over
917 time. This is done by setting @var{attack} and @var{release}.
919 @var{attack} determines how long the signal has to fall below the threshold
920 before any reduction will occur and @var{release} sets the time the signal
921 has to raise above the threshold to reduce the reduction again.
922 Shorter signals than the chosen attack time will be left untouched.
926 Set input level before filtering.
927 Default is 1. Allowed range is from 0.015625 to 64.
930 Set the level of gain reduction when the signal is below the threshold.
931 Default is 0.06125. Allowed range is from 0 to 1.
934 If a signal rises above this level the gain reduction is released.
935 Default is 0.125. Allowed range is from 0 to 1.
938 Set a ratio about which the signal is reduced.
939 Default is 2. Allowed range is from 1 to 9000.
942 Amount of milliseconds the signal has to rise above the threshold before gain
944 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
947 Amount of milliseconds the signal has to fall below the threshold before the
948 reduction is increased again. Default is 250 milliseconds.
949 Allowed range is from 0.01 to 9000.
952 Set amount of amplification of signal after processing.
953 Default is 1. Allowed range is from 1 to 64.
956 Curve the sharp knee around the threshold to enter gain reduction more softly.
957 Default is 2.828427125. Allowed range is from 1 to 8.
960 Choose if exact signal should be taken for detection or an RMS like one.
961 Default is rms. Can be peak or rms.
964 Choose if the average level between all channels or the louder channel affects
966 Default is average. Can be average or maximum.
971 The limiter prevents input signal from raising over a desired threshold.
972 This limiter uses lookahead technology to prevent your signal from distorting.
973 It means that there is a small delay after signal is processed. Keep in mind
974 that the delay it produces is the attack time you set.
976 The filter accepts the following options:
980 Set input gain. Default is 1.
983 Set output gain. Default is 1.
986 Don't let signals above this level pass the limiter. Default is 1.
989 The limiter will reach its attenuation level in this amount of time in
990 milliseconds. Default is 5 milliseconds.
993 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
994 Default is 50 milliseconds.
997 When gain reduction is always needed ASC takes care of releasing to an
998 average reduction level rather than reaching a reduction of 0 in the release
1002 Select how much the release time is affected by ASC, 0 means nearly no changes
1003 in release time while 1 produces higher release times.
1006 Auto level output signal. Default is enabled.
1007 This normalizes audio back to 0dB if enabled.
1010 Depending on picked setting it is recommended to upsample input 2x or 4x times
1011 with @ref{aresample} before applying this filter.
1015 Apply a two-pole all-pass filter with central frequency (in Hz)
1016 @var{frequency}, and filter-width @var{width}.
1017 An all-pass filter changes the audio's frequency to phase relationship
1018 without changing its frequency to amplitude relationship.
1020 The filter accepts the following options:
1024 Set frequency in Hz.
1027 Set method to specify band-width of filter.
1040 Specify the band-width of a filter in width_type units.
1047 The filter accepts the following options:
1051 Set the number of loops.
1054 Set maximal number of samples.
1057 Set first sample of loop.
1063 Merge two or more audio streams into a single multi-channel stream.
1065 The filter accepts the following options:
1070 Set the number of inputs. Default is 2.
1074 If the channel layouts of the inputs are disjoint, and therefore compatible,
1075 the channel layout of the output will be set accordingly and the channels
1076 will be reordered as necessary. If the channel layouts of the inputs are not
1077 disjoint, the output will have all the channels of the first input then all
1078 the channels of the second input, in that order, and the channel layout of
1079 the output will be the default value corresponding to the total number of
1082 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1083 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1084 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1085 first input, b1 is the first channel of the second input).
1087 On the other hand, if both input are in stereo, the output channels will be
1088 in the default order: a1, a2, b1, b2, and the channel layout will be
1089 arbitrarily set to 4.0, which may or may not be the expected value.
1091 All inputs must have the same sample rate, and format.
1093 If inputs do not have the same duration, the output will stop with the
1096 @subsection Examples
1100 Merge two mono files into a stereo stream:
1102 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1106 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1108 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
1114 Mixes multiple audio inputs into a single output.
1116 Note that this filter only supports float samples (the @var{amerge}
1117 and @var{pan} audio filters support many formats). If the @var{amix}
1118 input has integer samples then @ref{aresample} will be automatically
1119 inserted to perform the conversion to float samples.
1123 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1125 will mix 3 input audio streams to a single output with the same duration as the
1126 first input and a dropout transition time of 3 seconds.
1128 It accepts the following parameters:
1132 The number of inputs. If unspecified, it defaults to 2.
1135 How to determine the end-of-stream.
1139 The duration of the longest input. (default)
1142 The duration of the shortest input.
1145 The duration of the first input.
1149 @item dropout_transition
1150 The transition time, in seconds, for volume renormalization when an input
1151 stream ends. The default value is 2 seconds.
1155 @section anequalizer
1157 High-order parametric multiband equalizer for each channel.
1159 It accepts the following parameters:
1163 This option string is in format:
1164 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1165 Each equalizer band is separated by '|'.
1169 Set channel number to which equalization will be applied.
1170 If input doesn't have that channel the entry is ignored.
1173 Set central frequency for band.
1174 If input doesn't have that frequency the entry is ignored.
1177 Set band width in hertz.
1180 Set band gain in dB.
1183 Set filter type for band, optional, can be:
1187 Butterworth, this is default.
1198 With this option activated frequency response of anequalizer is displayed
1202 Set video stream size. Only useful if curves option is activated.
1205 Set max gain that will be displayed. Only useful if curves option is activated.
1206 Setting this to reasonable value allows to display gain which is derived from
1207 neighbour bands which are too close to each other and thus produce higher gain
1208 when both are activated.
1211 Set frequency scale used to draw frequency response in video output.
1212 Can be linear or logarithmic. Default is logarithmic.
1215 Set color for each channel curve which is going to be displayed in video stream.
1216 This is list of color names separated by space or by '|'.
1217 Unrecognised or missing colors will be replaced by white color.
1220 @subsection Examples
1224 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1225 for first 2 channels using Chebyshev type 1 filter:
1227 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1231 @subsection Commands
1233 This filter supports the following commands:
1236 Alter existing filter parameters.
1237 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1239 @var{fN} is existing filter number, starting from 0, if no such filter is available
1241 @var{freq} set new frequency parameter.
1242 @var{width} set new width parameter in herz.
1243 @var{gain} set new gain parameter in dB.
1245 Full filter invocation with asendcmd may look like this:
1246 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1251 Pass the audio source unchanged to the output.
1255 Pad the end of an audio stream with silence.
1257 This can be used together with @command{ffmpeg} @option{-shortest} to
1258 extend audio streams to the same length as the video stream.
1260 A description of the accepted options follows.
1264 Set silence packet size. Default value is 4096.
1267 Set the number of samples of silence to add to the end. After the
1268 value is reached, the stream is terminated. This option is mutually
1269 exclusive with @option{whole_len}.
1272 Set the minimum total number of samples in the output audio stream. If
1273 the value is longer than the input audio length, silence is added to
1274 the end, until the value is reached. This option is mutually exclusive
1275 with @option{pad_len}.
1278 If neither the @option{pad_len} nor the @option{whole_len} option is
1279 set, the filter will add silence to the end of the input stream
1282 @subsection Examples
1286 Add 1024 samples of silence to the end of the input:
1292 Make sure the audio output will contain at least 10000 samples, pad
1293 the input with silence if required:
1295 apad=whole_len=10000
1299 Use @command{ffmpeg} to pad the audio input with silence, so that the
1300 video stream will always result the shortest and will be converted
1301 until the end in the output file when using the @option{shortest}
1304 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1309 Add a phasing effect to the input audio.
1311 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1312 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1314 A description of the accepted parameters follows.
1318 Set input gain. Default is 0.4.
1321 Set output gain. Default is 0.74
1324 Set delay in milliseconds. Default is 3.0.
1327 Set decay. Default is 0.4.
1330 Set modulation speed in Hz. Default is 0.5.
1333 Set modulation type. Default is triangular.
1335 It accepts the following values:
1344 Audio pulsator is something between an autopanner and a tremolo.
1345 But it can produce funny stereo effects as well. Pulsator changes the volume
1346 of the left and right channel based on a LFO (low frequency oscillator) with
1347 different waveforms and shifted phases.
1348 This filter have the ability to define an offset between left and right
1349 channel. An offset of 0 means that both LFO shapes match each other.
1350 The left and right channel are altered equally - a conventional tremolo.
1351 An offset of 50% means that the shape of the right channel is exactly shifted
1352 in phase (or moved backwards about half of the frequency) - pulsator acts as
1353 an autopanner. At 1 both curves match again. Every setting in between moves the
1354 phase shift gapless between all stages and produces some "bypassing" sounds with
1355 sine and triangle waveforms. The more you set the offset near 1 (starting from
1356 the 0.5) the faster the signal passes from the left to the right speaker.
1358 The filter accepts the following options:
1362 Set input gain. By default it is 1. Range is [0.015625 - 64].
1365 Set output gain. By default it is 1. Range is [0.015625 - 64].
1368 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1369 sawup or sawdown. Default is sine.
1372 Set modulation. Define how much of original signal is affected by the LFO.
1375 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1378 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1381 Set pulse width. Default is 1. Allowed range is [0 - 2].
1384 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1387 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1391 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1395 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1396 if timing is set to hz.
1402 Resample the input audio to the specified parameters, using the
1403 libswresample library. If none are specified then the filter will
1404 automatically convert between its input and output.
1406 This filter is also able to stretch/squeeze the audio data to make it match
1407 the timestamps or to inject silence / cut out audio to make it match the
1408 timestamps, do a combination of both or do neither.
1410 The filter accepts the syntax
1411 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1412 expresses a sample rate and @var{resampler_options} is a list of
1413 @var{key}=@var{value} pairs, separated by ":". See the
1414 ffmpeg-resampler manual for the complete list of supported options.
1416 @subsection Examples
1420 Resample the input audio to 44100Hz:
1426 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1427 samples per second compensation:
1429 aresample=async=1000
1435 Reverse an audio clip.
1437 Warning: This filter requires memory to buffer the entire clip, so trimming
1440 @subsection Examples
1444 Take the first 5 seconds of a clip, and reverse it.
1446 atrim=end=5,areverse
1450 @section asetnsamples
1452 Set the number of samples per each output audio frame.
1454 The last output packet may contain a different number of samples, as
1455 the filter will flush all the remaining samples when the input audio
1458 The filter accepts the following options:
1462 @item nb_out_samples, n
1463 Set the number of frames per each output audio frame. The number is
1464 intended as the number of samples @emph{per each channel}.
1465 Default value is 1024.
1468 If set to 1, the filter will pad the last audio frame with zeroes, so
1469 that the last frame will contain the same number of samples as the
1470 previous ones. Default value is 1.
1473 For example, to set the number of per-frame samples to 1234 and
1474 disable padding for the last frame, use:
1476 asetnsamples=n=1234:p=0
1481 Set the sample rate without altering the PCM data.
1482 This will result in a change of speed and pitch.
1484 The filter accepts the following options:
1487 @item sample_rate, r
1488 Set the output sample rate. Default is 44100 Hz.
1493 Show a line containing various information for each input audio frame.
1494 The input audio is not modified.
1496 The shown line contains a sequence of key/value pairs of the form
1497 @var{key}:@var{value}.
1499 The following values are shown in the output:
1503 The (sequential) number of the input frame, starting from 0.
1506 The presentation timestamp of the input frame, in time base units; the time base
1507 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1510 The presentation timestamp of the input frame in seconds.
1513 position of the frame in the input stream, -1 if this information in
1514 unavailable and/or meaningless (for example in case of synthetic audio)
1523 The sample rate for the audio frame.
1526 The number of samples (per channel) in the frame.
1529 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1530 audio, the data is treated as if all the planes were concatenated.
1532 @item plane_checksums
1533 A list of Adler-32 checksums for each data plane.
1539 Display time domain statistical information about the audio channels.
1540 Statistics are calculated and displayed for each audio channel and,
1541 where applicable, an overall figure is also given.
1543 It accepts the following option:
1546 Short window length in seconds, used for peak and trough RMS measurement.
1547 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1551 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1552 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1555 Available keys for each channel are:
1586 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1587 this @code{lavfi.astats.Overall.Peak_count}.
1589 For description what each key means read below.
1592 Set number of frame after which stats are going to be recalculated.
1593 Default is disabled.
1596 A description of each shown parameter follows:
1600 Mean amplitude displacement from zero.
1603 Minimal sample level.
1606 Maximal sample level.
1608 @item Min difference
1609 Minimal difference between two consecutive samples.
1611 @item Max difference
1612 Maximal difference between two consecutive samples.
1614 @item Mean difference
1615 Mean difference between two consecutive samples.
1616 The average of each difference between two consecutive samples.
1620 Standard peak and RMS level measured in dBFS.
1624 Peak and trough values for RMS level measured over a short window.
1627 Standard ratio of peak to RMS level (note: not in dB).
1630 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1631 (i.e. either @var{Min level} or @var{Max level}).
1634 Number of occasions (not the number of samples) that the signal attained either
1635 @var{Min level} or @var{Max level}.
1638 Overall bit depth of audio. Number of bits used for each sample.
1643 Synchronize audio data with timestamps by squeezing/stretching it and/or
1644 dropping samples/adding silence when needed.
1646 This filter is not built by default, please use @ref{aresample} to do squeezing/stretching.
1648 It accepts the following parameters:
1652 Enable stretching/squeezing the data to make it match the timestamps. Disabled
1653 by default. When disabled, time gaps are covered with silence.
1656 The minimum difference between timestamps and audio data (in seconds) to trigger
1657 adding/dropping samples. The default value is 0.1. If you get an imperfect
1658 sync with this filter, try setting this parameter to 0.
1661 The maximum compensation in samples per second. Only relevant with compensate=1.
1662 The default value is 500.
1665 Assume that the first PTS should be this value. The time base is 1 / sample
1666 rate. This allows for padding/trimming at the start of the stream. By default,
1667 no assumption is made about the first frame's expected PTS, so no padding or
1668 trimming is done. For example, this could be set to 0 to pad the beginning with
1669 silence if an audio stream starts after the video stream or to trim any samples
1670 with a negative PTS due to encoder delay.
1678 The filter accepts exactly one parameter, the audio tempo. If not
1679 specified then the filter will assume nominal 1.0 tempo. Tempo must
1680 be in the [0.5, 2.0] range.
1682 @subsection Examples
1686 Slow down audio to 80% tempo:
1692 To speed up audio to 125% tempo:
1700 Trim the input so that the output contains one continuous subpart of the input.
1702 It accepts the following parameters:
1705 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1706 sample with the timestamp @var{start} will be the first sample in the output.
1709 Specify time of the first audio sample that will be dropped, i.e. the
1710 audio sample immediately preceding the one with the timestamp @var{end} will be
1711 the last sample in the output.
1714 Same as @var{start}, except this option sets the start timestamp in samples
1718 Same as @var{end}, except this option sets the end timestamp in samples instead
1722 The maximum duration of the output in seconds.
1725 The number of the first sample that should be output.
1728 The number of the first sample that should be dropped.
1731 @option{start}, @option{end}, and @option{duration} are expressed as time
1732 duration specifications; see
1733 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1735 Note that the first two sets of the start/end options and the @option{duration}
1736 option look at the frame timestamp, while the _sample options simply count the
1737 samples that pass through the filter. So start/end_pts and start/end_sample will
1738 give different results when the timestamps are wrong, inexact or do not start at
1739 zero. Also note that this filter does not modify the timestamps. If you wish
1740 to have the output timestamps start at zero, insert the asetpts filter after the
1743 If multiple start or end options are set, this filter tries to be greedy and
1744 keep all samples that match at least one of the specified constraints. To keep
1745 only the part that matches all the constraints at once, chain multiple atrim
1748 The defaults are such that all the input is kept. So it is possible to set e.g.
1749 just the end values to keep everything before the specified time.
1754 Drop everything except the second minute of input:
1756 ffmpeg -i INPUT -af atrim=60:120
1760 Keep only the first 1000 samples:
1762 ffmpeg -i INPUT -af atrim=end_sample=1000
1769 Apply a two-pole Butterworth band-pass filter with central
1770 frequency @var{frequency}, and (3dB-point) band-width width.
1771 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1772 instead of the default: constant 0dB peak gain.
1773 The filter roll off at 6dB per octave (20dB per decade).
1775 The filter accepts the following options:
1779 Set the filter's central frequency. Default is @code{3000}.
1782 Constant skirt gain if set to 1. Defaults to 0.
1785 Set method to specify band-width of filter.
1798 Specify the band-width of a filter in width_type units.
1803 Apply a two-pole Butterworth band-reject filter with central
1804 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1805 The filter roll off at 6dB per octave (20dB per decade).
1807 The filter accepts the following options:
1811 Set the filter's central frequency. Default is @code{3000}.
1814 Set method to specify band-width of filter.
1827 Specify the band-width of a filter in width_type units.
1832 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1833 shelving filter with a response similar to that of a standard
1834 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1836 The filter accepts the following options:
1840 Give the gain at 0 Hz. Its useful range is about -20
1841 (for a large cut) to +20 (for a large boost).
1842 Beware of clipping when using a positive gain.
1845 Set the filter's central frequency and so can be used
1846 to extend or reduce the frequency range to be boosted or cut.
1847 The default value is @code{100} Hz.
1850 Set method to specify band-width of filter.
1863 Determine how steep is the filter's shelf transition.
1868 Apply a biquad IIR filter with the given coefficients.
1869 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1870 are the numerator and denominator coefficients respectively.
1873 Bauer stereo to binaural transformation, which improves headphone listening of
1874 stereo audio records.
1876 It accepts the following parameters:
1880 Pre-defined crossfeed level.
1884 Default level (fcut=700, feed=50).
1887 Chu Moy circuit (fcut=700, feed=60).
1890 Jan Meier circuit (fcut=650, feed=95).
1895 Cut frequency (in Hz).
1904 Remap input channels to new locations.
1906 It accepts the following parameters:
1908 @item channel_layout
1909 The channel layout of the output stream.
1912 Map channels from input to output. The argument is a '|'-separated list of
1913 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1914 @var{in_channel} form. @var{in_channel} can be either the name of the input
1915 channel (e.g. FL for front left) or its index in the input channel layout.
1916 @var{out_channel} is the name of the output channel or its index in the output
1917 channel layout. If @var{out_channel} is not given then it is implicitly an
1918 index, starting with zero and increasing by one for each mapping.
1921 If no mapping is present, the filter will implicitly map input channels to
1922 output channels, preserving indices.
1924 For example, assuming a 5.1+downmix input MOV file,
1926 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1928 will create an output WAV file tagged as stereo from the downmix channels of
1931 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1933 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
1936 @section channelsplit
1938 Split each channel from an input audio stream into a separate output stream.
1940 It accepts the following parameters:
1942 @item channel_layout
1943 The channel layout of the input stream. The default is "stereo".
1946 For example, assuming a stereo input MP3 file,
1948 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1950 will create an output Matroska file with two audio streams, one containing only
1951 the left channel and the other the right channel.
1953 Split a 5.1 WAV file into per-channel files:
1955 ffmpeg -i in.wav -filter_complex
1956 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1957 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1958 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
1963 Add a chorus effect to the audio.
1965 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
1967 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
1968 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
1969 The modulation depth defines the range the modulated delay is played before or after
1970 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
1971 sound tuned around the original one, like in a chorus where some vocals are slightly
1974 It accepts the following parameters:
1977 Set input gain. Default is 0.4.
1980 Set output gain. Default is 0.4.
1983 Set delays. A typical delay is around 40ms to 60ms.
1995 @subsection Examples
2001 chorus=0.7:0.9:55:0.4:0.25:2
2007 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2011 Fuller sounding chorus with three delays:
2013 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
2018 Compress or expand the audio's dynamic range.
2020 It accepts the following parameters:
2026 A list of times in seconds for each channel over which the instantaneous level
2027 of the input signal is averaged to determine its volume. @var{attacks} refers to
2028 increase of volume and @var{decays} refers to decrease of volume. For most
2029 situations, the attack time (response to the audio getting louder) should be
2030 shorter than the decay time, because the human ear is more sensitive to sudden
2031 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2032 a typical value for decay is 0.8 seconds.
2033 If specified number of attacks & decays is lower than number of channels, the last
2034 set attack/decay will be used for all remaining channels.
2037 A list of points for the transfer function, specified in dB relative to the
2038 maximum possible signal amplitude. Each key points list must be defined using
2039 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2040 @code{x0/y0 x1/y1 x2/y2 ....}
2042 The input values must be in strictly increasing order but the transfer function
2043 does not have to be monotonically rising. The point @code{0/0} is assumed but
2044 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2045 function are @code{-70/-70|-60/-20}.
2048 Set the curve radius in dB for all joints. It defaults to 0.01.
2051 Set the additional gain in dB to be applied at all points on the transfer
2052 function. This allows for easy adjustment of the overall gain.
2056 Set an initial volume, in dB, to be assumed for each channel when filtering
2057 starts. This permits the user to supply a nominal level initially, so that, for
2058 example, a very large gain is not applied to initial signal levels before the
2059 companding has begun to operate. A typical value for audio which is initially
2060 quiet is -90 dB. It defaults to 0.
2063 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2064 delayed before being fed to the volume adjuster. Specifying a delay
2065 approximately equal to the attack/decay times allows the filter to effectively
2066 operate in predictive rather than reactive mode. It defaults to 0.
2070 @subsection Examples
2074 Make music with both quiet and loud passages suitable for listening to in a
2077 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2080 Another example for audio with whisper and explosion parts:
2082 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2086 A noise gate for when the noise is at a lower level than the signal:
2088 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2092 Here is another noise gate, this time for when the noise is at a higher level
2093 than the signal (making it, in some ways, similar to squelch):
2095 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2099 2:1 compression starting at -6dB:
2101 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2105 2:1 compression starting at -9dB:
2107 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2111 2:1 compression starting at -12dB:
2113 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2117 2:1 compression starting at -18dB:
2119 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2123 3:1 compression starting at -15dB:
2125 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2131 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2137 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
2141 Hard limiter at -6dB:
2143 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2147 Hard limiter at -12dB:
2149 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2153 Hard noise gate at -35 dB:
2155 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2161 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2165 @section compensationdelay
2167 Compensation Delay Line is a metric based delay to compensate differing
2168 positions of microphones or speakers.
2170 For example, you have recorded guitar with two microphones placed in
2171 different location. Because the front of sound wave has fixed speed in
2172 normal conditions, the phasing of microphones can vary and depends on
2173 their location and interposition. The best sound mix can be achieved when
2174 these microphones are in phase (synchronized). Note that distance of
2175 ~30 cm between microphones makes one microphone to capture signal in
2176 antiphase to another microphone. That makes the final mix sounding moody.
2177 This filter helps to solve phasing problems by adding different delays
2178 to each microphone track and make them synchronized.
2180 The best result can be reached when you take one track as base and
2181 synchronize other tracks one by one with it.
2182 Remember that synchronization/delay tolerance depends on sample rate, too.
2183 Higher sample rates will give more tolerance.
2185 It accepts the following parameters:
2189 Set millimeters distance. This is compensation distance for fine tuning.
2193 Set cm distance. This is compensation distance for tightening distance setup.
2197 Set meters distance. This is compensation distance for hard distance setup.
2201 Set dry amount. Amount of unprocessed (dry) signal.
2205 Set wet amount. Amount of processed (wet) signal.
2209 Set temperature degree in Celsius. This is the temperature of the environment.
2213 @section crystalizer
2214 Simple algorithm to expand audio dynamic range.
2216 The filter accepts the following options:
2220 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2221 (unchanged sound) to 10.0 (maximum effect).
2224 Enable clipping. By default is enabled.
2228 Apply a DC shift to the audio.
2230 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2231 in the recording chain) from the audio. The effect of a DC offset is reduced
2232 headroom and hence volume. The @ref{astats} filter can be used to determine if
2233 a signal has a DC offset.
2237 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2241 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2242 used to prevent clipping.
2246 Dynamic Audio Normalizer.
2248 This filter applies a certain amount of gain to the input audio in order
2249 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2250 contrast to more "simple" normalization algorithms, the Dynamic Audio
2251 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2252 This allows for applying extra gain to the "quiet" sections of the audio
2253 while avoiding distortions or clipping the "loud" sections. In other words:
2254 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2255 sections, in the sense that the volume of each section is brought to the
2256 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2257 this goal *without* applying "dynamic range compressing". It will retain 100%
2258 of the dynamic range *within* each section of the audio file.
2262 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2263 Default is 500 milliseconds.
2264 The Dynamic Audio Normalizer processes the input audio in small chunks,
2265 referred to as frames. This is required, because a peak magnitude has no
2266 meaning for just a single sample value. Instead, we need to determine the
2267 peak magnitude for a contiguous sequence of sample values. While a "standard"
2268 normalizer would simply use the peak magnitude of the complete file, the
2269 Dynamic Audio Normalizer determines the peak magnitude individually for each
2270 frame. The length of a frame is specified in milliseconds. By default, the
2271 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2272 been found to give good results with most files.
2273 Note that the exact frame length, in number of samples, will be determined
2274 automatically, based on the sampling rate of the individual input audio file.
2277 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2278 number. Default is 31.
2279 Probably the most important parameter of the Dynamic Audio Normalizer is the
2280 @code{window size} of the Gaussian smoothing filter. The filter's window size
2281 is specified in frames, centered around the current frame. For the sake of
2282 simplicity, this must be an odd number. Consequently, the default value of 31
2283 takes into account the current frame, as well as the 15 preceding frames and
2284 the 15 subsequent frames. Using a larger window results in a stronger
2285 smoothing effect and thus in less gain variation, i.e. slower gain
2286 adaptation. Conversely, using a smaller window results in a weaker smoothing
2287 effect and thus in more gain variation, i.e. faster gain adaptation.
2288 In other words, the more you increase this value, the more the Dynamic Audio
2289 Normalizer will behave like a "traditional" normalization filter. On the
2290 contrary, the more you decrease this value, the more the Dynamic Audio
2291 Normalizer will behave like a dynamic range compressor.
2294 Set the target peak value. This specifies the highest permissible magnitude
2295 level for the normalized audio input. This filter will try to approach the
2296 target peak magnitude as closely as possible, but at the same time it also
2297 makes sure that the normalized signal will never exceed the peak magnitude.
2298 A frame's maximum local gain factor is imposed directly by the target peak
2299 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2300 It is not recommended to go above this value.
2303 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2304 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2305 factor for each input frame, i.e. the maximum gain factor that does not
2306 result in clipping or distortion. The maximum gain factor is determined by
2307 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2308 additionally bounds the frame's maximum gain factor by a predetermined
2309 (global) maximum gain factor. This is done in order to avoid excessive gain
2310 factors in "silent" or almost silent frames. By default, the maximum gain
2311 factor is 10.0, For most inputs the default value should be sufficient and
2312 it usually is not recommended to increase this value. Though, for input
2313 with an extremely low overall volume level, it may be necessary to allow even
2314 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2315 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2316 Instead, a "sigmoid" threshold function will be applied. This way, the
2317 gain factors will smoothly approach the threshold value, but never exceed that
2321 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2322 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2323 This means that the maximum local gain factor for each frame is defined
2324 (only) by the frame's highest magnitude sample. This way, the samples can
2325 be amplified as much as possible without exceeding the maximum signal
2326 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2327 Normalizer can also take into account the frame's root mean square,
2328 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2329 determine the power of a time-varying signal. It is therefore considered
2330 that the RMS is a better approximation of the "perceived loudness" than
2331 just looking at the signal's peak magnitude. Consequently, by adjusting all
2332 frames to a constant RMS value, a uniform "perceived loudness" can be
2333 established. If a target RMS value has been specified, a frame's local gain
2334 factor is defined as the factor that would result in exactly that RMS value.
2335 Note, however, that the maximum local gain factor is still restricted by the
2336 frame's highest magnitude sample, in order to prevent clipping.
2339 Enable channels coupling. By default is enabled.
2340 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2341 amount. This means the same gain factor will be applied to all channels, i.e.
2342 the maximum possible gain factor is determined by the "loudest" channel.
2343 However, in some recordings, it may happen that the volume of the different
2344 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2345 In this case, this option can be used to disable the channel coupling. This way,
2346 the gain factor will be determined independently for each channel, depending
2347 only on the individual channel's highest magnitude sample. This allows for
2348 harmonizing the volume of the different channels.
2351 Enable DC bias correction. By default is disabled.
2352 An audio signal (in the time domain) is a sequence of sample values.
2353 In the Dynamic Audio Normalizer these sample values are represented in the
2354 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2355 audio signal, or "waveform", should be centered around the zero point.
2356 That means if we calculate the mean value of all samples in a file, or in a
2357 single frame, then the result should be 0.0 or at least very close to that
2358 value. If, however, there is a significant deviation of the mean value from
2359 0.0, in either positive or negative direction, this is referred to as a
2360 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2361 Audio Normalizer provides optional DC bias correction.
2362 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2363 the mean value, or "DC correction" offset, of each input frame and subtract
2364 that value from all of the frame's sample values which ensures those samples
2365 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2366 boundaries, the DC correction offset values will be interpolated smoothly
2367 between neighbouring frames.
2370 Enable alternative boundary mode. By default is disabled.
2371 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2372 around each frame. This includes the preceding frames as well as the
2373 subsequent frames. However, for the "boundary" frames, located at the very
2374 beginning and at the very end of the audio file, not all neighbouring
2375 frames are available. In particular, for the first few frames in the audio
2376 file, the preceding frames are not known. And, similarly, for the last few
2377 frames in the audio file, the subsequent frames are not known. Thus, the
2378 question arises which gain factors should be assumed for the missing frames
2379 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2380 to deal with this situation. The default boundary mode assumes a gain factor
2381 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2382 "fade out" at the beginning and at the end of the input, respectively.
2385 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2386 By default, the Dynamic Audio Normalizer does not apply "traditional"
2387 compression. This means that signal peaks will not be pruned and thus the
2388 full dynamic range will be retained within each local neighbourhood. However,
2389 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2390 normalization algorithm with a more "traditional" compression.
2391 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2392 (thresholding) function. If (and only if) the compression feature is enabled,
2393 all input frames will be processed by a soft knee thresholding function prior
2394 to the actual normalization process. Put simply, the thresholding function is
2395 going to prune all samples whose magnitude exceeds a certain threshold value.
2396 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2397 value. Instead, the threshold value will be adjusted for each individual
2399 In general, smaller parameters result in stronger compression, and vice versa.
2400 Values below 3.0 are not recommended, because audible distortion may appear.
2405 Make audio easier to listen to on headphones.
2407 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2408 so that when listened to on headphones the stereo image is moved from
2409 inside your head (standard for headphones) to outside and in front of
2410 the listener (standard for speakers).
2416 Apply a two-pole peaking equalisation (EQ) filter. With this
2417 filter, the signal-level at and around a selected frequency can
2418 be increased or decreased, whilst (unlike bandpass and bandreject
2419 filters) that at all other frequencies is unchanged.
2421 In order to produce complex equalisation curves, this filter can
2422 be given several times, each with a different central frequency.
2424 The filter accepts the following options:
2428 Set the filter's central frequency in Hz.
2431 Set method to specify band-width of filter.
2444 Specify the band-width of a filter in width_type units.
2447 Set the required gain or attenuation in dB.
2448 Beware of clipping when using a positive gain.
2451 @subsection Examples
2454 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2456 equalizer=f=1000:width_type=h:width=200:g=-10
2460 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2462 equalizer=f=1000:width_type=q:width=1:g=2,equalizer=f=100:width_type=q:width=2:g=-5
2466 @section extrastereo
2468 Linearly increases the difference between left and right channels which
2469 adds some sort of "live" effect to playback.
2471 The filter accepts the following options:
2475 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2476 (average of both channels), with 1.0 sound will be unchanged, with
2477 -1.0 left and right channels will be swapped.
2480 Enable clipping. By default is enabled.
2483 @section firequalizer
2484 Apply FIR Equalization using arbitrary frequency response.
2486 The filter accepts the following option:
2490 Set gain curve equation (in dB). The expression can contain variables:
2493 the evaluated frequency
2497 channel number, set to 0 when multichannels evaluation is disabled
2499 channel id, see libavutil/channel_layout.h, set to the first channel id when
2500 multichannels evaluation is disabled
2504 channel_layout, see libavutil/channel_layout.h
2509 @item gain_interpolate(f)
2510 interpolate gain on frequency f based on gain_entry
2512 This option is also available as command. Default is @code{gain_interpolate(f)}.
2515 Set gain entry for gain_interpolate function. The expression can
2519 store gain entry at frequency f with value g
2521 This option is also available as command.
2524 Set filter delay in seconds. Higher value means more accurate.
2525 Default is @code{0.01}.
2528 Set filter accuracy in Hz. Lower value means more accurate.
2529 Default is @code{5}.
2532 Set window function. Acceptable values are:
2535 rectangular window, useful when gain curve is already smooth
2537 hann window (default)
2543 3-terms continuous 1st derivative nuttall window
2545 minimum 3-terms discontinuous nuttall window
2547 4-terms continuous 1st derivative nuttall window
2549 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2551 blackman-harris window
2555 If enabled, use fixed number of audio samples. This improves speed when
2556 filtering with large delay. Default is disabled.
2559 Enable multichannels evaluation on gain. Default is disabled.
2562 Enable zero phase mode by substracting timestamp to compensate delay.
2563 Default is disabled.
2566 @subsection Examples
2571 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2574 lowpass at 1000 Hz with gain_entry:
2576 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2579 custom equalization:
2581 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2584 higher delay with zero phase to compensate delay:
2586 firequalizer=delay=0.1:fixed=on:zero_phase=on
2589 lowpass on left channel, highpass on right channel:
2591 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2592 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2597 Apply a flanging effect to the audio.
2599 The filter accepts the following options:
2603 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2606 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2609 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2613 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2614 Default value is 71.
2617 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2620 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2621 Default value is @var{sinusoidal}.
2624 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2625 Default value is 25.
2628 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2629 Default is @var{linear}.
2634 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2635 embedded HDCD codes is expanded into a 20-bit PCM stream.
2637 The filter supports the Peak Extend and Low-level Gain Adjustment features
2638 of HDCD, and detects the Transient Filter flag.
2641 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2644 When using the filter with wav, note the default encoding for wav is 16-bit,
2645 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2646 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2648 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2649 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2652 The filter accepts the following options:
2655 @item disable_autoconvert
2656 Disable any automatic format conversion or resampling in the filter graph.
2658 @item process_stereo
2659 Process the stereo channels together. If target_gain does not match between
2660 channels, consider it invalid and use the last valid target_gain.
2663 Set the code detect timer period in ms.
2666 Always extend peaks above -3dBFS even if PE isn't signaled.
2669 Replace audio with a solid tone and adjust the amplitude to signal some
2670 specific aspect of the decoding process. The output file can be loaded in
2671 an audio editor alongside the original to aid analysis.
2673 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2680 Gain adjustment level at each sample
2682 Samples where peak extend occurs
2684 Samples where the code detect timer is active
2686 Samples where the target gain does not match between channels
2692 Apply a high-pass filter with 3dB point frequency.
2693 The filter can be either single-pole, or double-pole (the default).
2694 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2696 The filter accepts the following options:
2700 Set frequency in Hz. Default is 3000.
2703 Set number of poles. Default is 2.
2706 Set method to specify band-width of filter.
2719 Specify the band-width of a filter in width_type units.
2720 Applies only to double-pole filter.
2721 The default is 0.707q and gives a Butterworth response.
2726 Join multiple input streams into one multi-channel stream.
2728 It accepts the following parameters:
2732 The number of input streams. It defaults to 2.
2734 @item channel_layout
2735 The desired output channel layout. It defaults to stereo.
2738 Map channels from inputs to output. The argument is a '|'-separated list of
2739 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2740 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2741 can be either the name of the input channel (e.g. FL for front left) or its
2742 index in the specified input stream. @var{out_channel} is the name of the output
2746 The filter will attempt to guess the mappings when they are not specified
2747 explicitly. It does so by first trying to find an unused matching input channel
2748 and if that fails it picks the first unused input channel.
2750 Join 3 inputs (with properly set channel layouts):
2752 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2755 Build a 5.1 output from 6 single-channel streams:
2757 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2758 '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'
2764 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2766 To enable compilation of this filter you need to configure FFmpeg with
2767 @code{--enable-ladspa}.
2771 Specifies the name of LADSPA plugin library to load. If the environment
2772 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2773 each one of the directories specified by the colon separated list in
2774 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2775 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2776 @file{/usr/lib/ladspa/}.
2779 Specifies the plugin within the library. Some libraries contain only
2780 one plugin, but others contain many of them. If this is not set filter
2781 will list all available plugins within the specified library.
2784 Set the '|' separated list of controls which are zero or more floating point
2785 values that determine the behavior of the loaded plugin (for example delay,
2787 Controls need to be defined using the following syntax:
2788 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2789 @var{valuei} is the value set on the @var{i}-th control.
2790 Alternatively they can be also defined using the following syntax:
2791 @var{value0}|@var{value1}|@var{value2}|..., where
2792 @var{valuei} is the value set on the @var{i}-th control.
2793 If @option{controls} is set to @code{help}, all available controls and
2794 their valid ranges are printed.
2796 @item sample_rate, s
2797 Specify the sample rate, default to 44100. Only used if plugin have
2801 Set the number of samples per channel per each output frame, default
2802 is 1024. Only used if plugin have zero inputs.
2805 Set the minimum duration of the sourced audio. See
2806 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2807 for the accepted syntax.
2808 Note that the resulting duration may be greater than the specified duration,
2809 as the generated audio is always cut at the end of a complete frame.
2810 If not specified, or the expressed duration is negative, the audio is
2811 supposed to be generated forever.
2812 Only used if plugin have zero inputs.
2816 @subsection Examples
2820 List all available plugins within amp (LADSPA example plugin) library:
2826 List all available controls and their valid ranges for @code{vcf_notch}
2827 plugin from @code{VCF} library:
2829 ladspa=f=vcf:p=vcf_notch:c=help
2833 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2836 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2840 Add reverberation to the audio using TAP-plugins
2841 (Tom's Audio Processing plugins):
2843 ladspa=file=tap_reverb:tap_reverb
2847 Generate white noise, with 0.2 amplitude:
2849 ladspa=file=cmt:noise_source_white:c=c0=.2
2853 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2854 @code{C* Audio Plugin Suite} (CAPS) library:
2856 ladspa=file=caps:Click:c=c1=20'
2860 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2862 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2866 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2867 @code{SWH Plugins} collection:
2869 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2873 Attenuate low frequencies using Multiband EQ from Steve Harris
2874 @code{SWH Plugins} collection:
2876 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2880 @subsection Commands
2882 This filter supports the following commands:
2885 Modify the @var{N}-th control value.
2887 If the specified value is not valid, it is ignored and prior one is kept.
2892 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2893 Support for both single pass (livestreams, files) and double pass (files) modes.
2894 This algorithm can target IL, LRA, and maximum true peak.
2896 To enable compilation of this filter you need to configure FFmpeg with
2897 @code{--enable-libebur128}.
2899 The filter accepts the following options:
2903 Set integrated loudness target.
2904 Range is -70.0 - -5.0. Default value is -24.0.
2907 Set loudness range target.
2908 Range is 1.0 - 20.0. Default value is 7.0.
2911 Set maximum true peak.
2912 Range is -9.0 - +0.0. Default value is -2.0.
2914 @item measured_I, measured_i
2915 Measured IL of input file.
2916 Range is -99.0 - +0.0.
2918 @item measured_LRA, measured_lra
2919 Measured LRA of input file.
2920 Range is 0.0 - 99.0.
2922 @item measured_TP, measured_tp
2923 Measured true peak of input file.
2924 Range is -99.0 - +99.0.
2926 @item measured_thresh
2927 Measured threshold of input file.
2928 Range is -99.0 - +0.0.
2931 Set offset gain. Gain is applied before the true-peak limiter.
2932 Range is -99.0 - +99.0. Default is +0.0.
2935 Normalize linearly if possible.
2936 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2937 to be specified in order to use this mode.
2938 Options are true or false. Default is true.
2941 Treat mono input files as "dual-mono". If a mono file is intended for playback
2942 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2943 If set to @code{true}, this option will compensate for this effect.
2944 Multi-channel input files are not affected by this option.
2945 Options are true or false. Default is false.
2948 Set print format for stats. Options are summary, json, or none.
2949 Default value is none.
2954 Apply a low-pass filter with 3dB point frequency.
2955 The filter can be either single-pole or double-pole (the default).
2956 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2958 The filter accepts the following options:
2962 Set frequency in Hz. Default is 500.
2965 Set number of poles. Default is 2.
2968 Set method to specify band-width of filter.
2981 Specify the band-width of a filter in width_type units.
2982 Applies only to double-pole filter.
2983 The default is 0.707q and gives a Butterworth response.
2989 Mix channels with specific gain levels. The filter accepts the output
2990 channel layout followed by a set of channels definitions.
2992 This filter is also designed to efficiently remap the channels of an audio
2995 The filter accepts parameters of the form:
2996 "@var{l}|@var{outdef}|@var{outdef}|..."
3000 output channel layout or number of channels
3003 output channel specification, of the form:
3004 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
3007 output channel to define, either a channel name (FL, FR, etc.) or a channel
3008 number (c0, c1, etc.)
3011 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3014 input channel to use, see out_name for details; it is not possible to mix
3015 named and numbered input channels
3018 If the `=' in a channel specification is replaced by `<', then the gains for
3019 that specification will be renormalized so that the total is 1, thus
3020 avoiding clipping noise.
3022 @subsection Mixing examples
3024 For example, if you want to down-mix from stereo to mono, but with a bigger
3025 factor for the left channel:
3027 pan=1c|c0=0.9*c0+0.1*c1
3030 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3031 7-channels surround:
3033 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3036 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3037 that should be preferred (see "-ac" option) unless you have very specific
3040 @subsection Remapping examples
3042 The channel remapping will be effective if, and only if:
3045 @item gain coefficients are zeroes or ones,
3046 @item only one input per channel output,
3049 If all these conditions are satisfied, the filter will notify the user ("Pure
3050 channel mapping detected"), and use an optimized and lossless method to do the
3053 For example, if you have a 5.1 source and want a stereo audio stream by
3054 dropping the extra channels:
3056 pan="stereo| c0=FL | c1=FR"
3059 Given the same source, you can also switch front left and front right channels
3060 and keep the input channel layout:
3062 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3065 If the input is a stereo audio stream, you can mute the front left channel (and
3066 still keep the stereo channel layout) with:
3071 Still with a stereo audio stream input, you can copy the right channel in both
3072 front left and right:
3074 pan="stereo| c0=FR | c1=FR"
3079 ReplayGain scanner filter. This filter takes an audio stream as an input and
3080 outputs it unchanged.
3081 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3085 Convert the audio sample format, sample rate and channel layout. It is
3086 not meant to be used directly.
3089 Apply time-stretching and pitch-shifting with librubberband.
3091 The filter accepts the following options:
3095 Set tempo scale factor.
3098 Set pitch scale factor.
3101 Set transients detector.
3102 Possible values are:
3111 Possible values are:
3120 Possible values are:
3127 Set processing window size.
3128 Possible values are:
3137 Possible values are:
3144 Enable formant preservation when shift pitching.
3145 Possible values are:
3153 Possible values are:
3162 Possible values are:
3169 @section sidechaincompress
3171 This filter acts like normal compressor but has the ability to compress
3172 detected signal using second input signal.
3173 It needs two input streams and returns one output stream.
3174 First input stream will be processed depending on second stream signal.
3175 The filtered signal then can be filtered with other filters in later stages of
3176 processing. See @ref{pan} and @ref{amerge} filter.
3178 The filter accepts the following options:
3182 Set input gain. Default is 1. Range is between 0.015625 and 64.
3185 If a signal of second stream raises above this level it will affect the gain
3186 reduction of first stream.
3187 By default is 0.125. Range is between 0.00097563 and 1.
3190 Set a ratio about which the signal is reduced. 1:2 means that if the level
3191 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3192 Default is 2. Range is between 1 and 20.
3195 Amount of milliseconds the signal has to rise above the threshold before gain
3196 reduction starts. Default is 20. Range is between 0.01 and 2000.
3199 Amount of milliseconds the signal has to fall below the threshold before
3200 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3203 Set the amount by how much signal will be amplified after processing.
3204 Default is 2. Range is from 1 and 64.
3207 Curve the sharp knee around the threshold to enter gain reduction more softly.
3208 Default is 2.82843. Range is between 1 and 8.
3211 Choose if the @code{average} level between all channels of side-chain stream
3212 or the louder(@code{maximum}) channel of side-chain stream affects the
3213 reduction. Default is @code{average}.
3216 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3217 of @code{rms}. Default is @code{rms} which is mainly smoother.
3220 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3223 How much to use compressed signal in output. Default is 1.
3224 Range is between 0 and 1.
3227 @subsection Examples
3231 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3232 depending on the signal of 2nd input and later compressed signal to be
3233 merged with 2nd input:
3235 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3239 @section sidechaingate
3241 A sidechain gate acts like a normal (wideband) gate but has the ability to
3242 filter the detected signal before sending it to the gain reduction stage.
3243 Normally a gate uses the full range signal to detect a level above the
3245 For example: If you cut all lower frequencies from your sidechain signal
3246 the gate will decrease the volume of your track only if not enough highs
3247 appear. With this technique you are able to reduce the resonation of a
3248 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3250 It needs two input streams and returns one output stream.
3251 First input stream will be processed depending on second stream signal.
3253 The filter accepts the following options:
3257 Set input level before filtering.
3258 Default is 1. Allowed range is from 0.015625 to 64.
3261 Set the level of gain reduction when the signal is below the threshold.
3262 Default is 0.06125. Allowed range is from 0 to 1.
3265 If a signal rises above this level the gain reduction is released.
3266 Default is 0.125. Allowed range is from 0 to 1.
3269 Set a ratio about which the signal is reduced.
3270 Default is 2. Allowed range is from 1 to 9000.
3273 Amount of milliseconds the signal has to rise above the threshold before gain
3275 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3278 Amount of milliseconds the signal has to fall below the threshold before the
3279 reduction is increased again. Default is 250 milliseconds.
3280 Allowed range is from 0.01 to 9000.
3283 Set amount of amplification of signal after processing.
3284 Default is 1. Allowed range is from 1 to 64.
3287 Curve the sharp knee around the threshold to enter gain reduction more softly.
3288 Default is 2.828427125. Allowed range is from 1 to 8.
3291 Choose if exact signal should be taken for detection or an RMS like one.
3292 Default is rms. Can be peak or rms.
3295 Choose if the average level between all channels or the louder channel affects
3297 Default is average. Can be average or maximum.
3300 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3303 @section silencedetect
3305 Detect silence in an audio stream.
3307 This filter logs a message when it detects that the input audio volume is less
3308 or equal to a noise tolerance value for a duration greater or equal to the
3309 minimum detected noise duration.
3311 The printed times and duration are expressed in seconds.
3313 The filter accepts the following options:
3317 Set silence duration until notification (default is 2 seconds).
3320 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3321 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3324 @subsection Examples
3328 Detect 5 seconds of silence with -50dB noise tolerance:
3330 silencedetect=n=-50dB:d=5
3334 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3335 tolerance in @file{silence.mp3}:
3337 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3341 @section silenceremove
3343 Remove silence from the beginning, middle or end of the audio.
3345 The filter accepts the following options:
3349 This value is used to indicate if audio should be trimmed at beginning of
3350 the audio. A value of zero indicates no silence should be trimmed from the
3351 beginning. When specifying a non-zero value, it trims audio up until it
3352 finds non-silence. Normally, when trimming silence from beginning of audio
3353 the @var{start_periods} will be @code{1} but it can be increased to higher
3354 values to trim all audio up to specific count of non-silence periods.
3355 Default value is @code{0}.
3357 @item start_duration
3358 Specify the amount of time that non-silence must be detected before it stops
3359 trimming audio. By increasing the duration, bursts of noises can be treated
3360 as silence and trimmed off. Default value is @code{0}.
3362 @item start_threshold
3363 This indicates what sample value should be treated as silence. For digital
3364 audio, a value of @code{0} may be fine but for audio recorded from analog,
3365 you may wish to increase the value to account for background noise.
3366 Can be specified in dB (in case "dB" is appended to the specified value)
3367 or amplitude ratio. Default value is @code{0}.
3370 Set the count for trimming silence from the end of audio.
3371 To remove silence from the middle of a file, specify a @var{stop_periods}
3372 that is negative. This value is then treated as a positive value and is
3373 used to indicate the effect should restart processing as specified by
3374 @var{start_periods}, making it suitable for removing periods of silence
3375 in the middle of the audio.
3376 Default value is @code{0}.
3379 Specify a duration of silence that must exist before audio is not copied any
3380 more. By specifying a higher duration, silence that is wanted can be left in
3382 Default value is @code{0}.
3384 @item stop_threshold
3385 This is the same as @option{start_threshold} but for trimming silence from
3387 Can be specified in dB (in case "dB" is appended to the specified value)
3388 or amplitude ratio. Default value is @code{0}.
3391 This indicate that @var{stop_duration} length of audio should be left intact
3392 at the beginning of each period of silence.
3393 For example, if you want to remove long pauses between words but do not want
3394 to remove the pauses completely. Default value is @code{0}.
3397 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3398 and works better with digital silence which is exactly 0.
3399 Default value is @code{rms}.
3402 Set ratio used to calculate size of window for detecting silence.
3403 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3406 @subsection Examples
3410 The following example shows how this filter can be used to start a recording
3411 that does not contain the delay at the start which usually occurs between
3412 pressing the record button and the start of the performance:
3414 silenceremove=1:5:0.02
3418 Trim all silence encountered from beginning to end where there is more than 1
3419 second of silence in audio:
3421 silenceremove=0:0:0:-1:1:-90dB
3427 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3428 loudspeakers around the user for binaural listening via headphones (audio
3429 formats up to 9 channels supported).
3430 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3431 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3432 Austrian Academy of Sciences.
3434 To enable compilation of this filter you need to configure FFmpeg with
3435 @code{--enable-netcdf}.
3437 The filter accepts the following options:
3441 Set the SOFA file used for rendering.
3444 Set gain applied to audio. Value is in dB. Default is 0.
3447 Set rotation of virtual loudspeakers in deg. Default is 0.
3450 Set elevation of virtual speakers in deg. Default is 0.
3453 Set distance in meters between loudspeakers and the listener with near-field
3454 HRTFs. Default is 1.
3457 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3458 processing audio in time domain which is slow.
3459 @var{freq} is processing audio in frequency domain which is fast.
3460 Default is @var{freq}.
3463 Set custom positions of virtual loudspeakers. Syntax for this option is:
3464 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3465 Each virtual loudspeaker is described with short channel name following with
3466 azimuth and elevation in degreees.
3467 Each virtual loudspeaker description is separated by '|'.
3468 For example to override front left and front right channel positions use:
3469 'speakers=FL 45 15|FR 345 15'.
3470 Descriptions with unrecognised channel names are ignored.
3473 @subsection Examples
3477 Using ClubFritz6 sofa file:
3479 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3483 Using ClubFritz12 sofa file and bigger radius with small rotation:
3485 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3489 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3490 and also with custom gain:
3492 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3496 @section stereotools
3498 This filter has some handy utilities to manage stereo signals, for converting
3499 M/S stereo recordings to L/R signal while having control over the parameters
3500 or spreading the stereo image of master track.
3502 The filter accepts the following options:
3506 Set input level before filtering for both channels. Defaults is 1.
3507 Allowed range is from 0.015625 to 64.
3510 Set output level after filtering for both channels. Defaults is 1.
3511 Allowed range is from 0.015625 to 64.
3514 Set input balance between both channels. Default is 0.
3515 Allowed range is from -1 to 1.
3518 Set output balance between both channels. Default is 0.
3519 Allowed range is from -1 to 1.
3522 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3523 clipping. Disabled by default.
3526 Mute the left channel. Disabled by default.
3529 Mute the right channel. Disabled by default.
3532 Change the phase of the left channel. Disabled by default.
3535 Change the phase of the right channel. Disabled by default.
3538 Set stereo mode. Available values are:
3542 Left/Right to Left/Right, this is default.
3545 Left/Right to Mid/Side.
3548 Mid/Side to Left/Right.
3551 Left/Right to Left/Left.
3554 Left/Right to Right/Right.
3557 Left/Right to Left + Right.
3560 Left/Right to Right/Left.
3564 Set level of side signal. Default is 1.
3565 Allowed range is from 0.015625 to 64.
3568 Set balance of side signal. Default is 0.
3569 Allowed range is from -1 to 1.
3572 Set level of the middle signal. Default is 1.
3573 Allowed range is from 0.015625 to 64.
3576 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3579 Set stereo base between mono and inversed channels. Default is 0.
3580 Allowed range is from -1 to 1.
3583 Set delay in milliseconds how much to delay left from right channel and
3584 vice versa. Default is 0. Allowed range is from -20 to 20.
3587 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3590 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3593 @subsection Examples
3597 Apply karaoke like effect:
3599 stereotools=mlev=0.015625
3603 Convert M/S signal to L/R:
3605 "stereotools=mode=ms>lr"
3609 @section stereowiden
3611 This filter enhance the stereo effect by suppressing signal common to both
3612 channels and by delaying the signal of left into right and vice versa,
3613 thereby widening the stereo effect.
3615 The filter accepts the following options:
3619 Time in milliseconds of the delay of left signal into right and vice versa.
3620 Default is 20 milliseconds.
3623 Amount of gain in delayed signal into right and vice versa. Gives a delay
3624 effect of left signal in right output and vice versa which gives widening
3625 effect. Default is 0.3.
3628 Cross feed of left into right with inverted phase. This helps in suppressing
3629 the mono. If the value is 1 it will cancel all the signal common to both
3630 channels. Default is 0.3.
3633 Set level of input signal of original channel. Default is 0.8.
3638 Boost or cut treble (upper) frequencies of the audio using a two-pole
3639 shelving filter with a response similar to that of a standard
3640 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3642 The filter accepts the following options:
3646 Give the gain at whichever is the lower of ~22 kHz and the
3647 Nyquist frequency. Its useful range is about -20 (for a large cut)
3648 to +20 (for a large boost). Beware of clipping when using a positive gain.
3651 Set the filter's central frequency and so can be used
3652 to extend or reduce the frequency range to be boosted or cut.
3653 The default value is @code{3000} Hz.
3656 Set method to specify band-width of filter.
3669 Determine how steep is the filter's shelf transition.
3674 Sinusoidal amplitude modulation.
3676 The filter accepts the following options:
3680 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3681 (20 Hz or lower) will result in a tremolo effect.
3682 This filter may also be used as a ring modulator by specifying
3683 a modulation frequency higher than 20 Hz.
3684 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3687 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3688 Default value is 0.5.
3693 Sinusoidal phase modulation.
3695 The filter accepts the following options:
3699 Modulation frequency in Hertz.
3700 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3703 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3704 Default value is 0.5.
3709 Adjust the input audio volume.
3711 It accepts the following parameters:
3715 Set audio volume expression.
3717 Output values are clipped to the maximum value.
3719 The output audio volume is given by the relation:
3721 @var{output_volume} = @var{volume} * @var{input_volume}
3724 The default value for @var{volume} is "1.0".
3727 This parameter represents the mathematical precision.
3729 It determines which input sample formats will be allowed, which affects the
3730 precision of the volume scaling.
3734 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3736 32-bit floating-point; this limits input sample format to FLT. (default)
3738 64-bit floating-point; this limits input sample format to DBL.
3742 Choose the behaviour on encountering ReplayGain side data in input frames.
3746 Remove ReplayGain side data, ignoring its contents (the default).
3749 Ignore ReplayGain side data, but leave it in the frame.
3752 Prefer the track gain, if present.
3755 Prefer the album gain, if present.
3758 @item replaygain_preamp
3759 Pre-amplification gain in dB to apply to the selected replaygain gain.
3761 Default value for @var{replaygain_preamp} is 0.0.
3764 Set when the volume expression is evaluated.
3766 It accepts the following values:
3769 only evaluate expression once during the filter initialization, or
3770 when the @samp{volume} command is sent
3773 evaluate expression for each incoming frame
3776 Default value is @samp{once}.
3779 The volume expression can contain the following parameters.
3783 frame number (starting at zero)
3786 @item nb_consumed_samples
3787 number of samples consumed by the filter
3789 number of samples in the current frame
3791 original frame position in the file
3797 PTS at start of stream
3799 time at start of stream
3805 last set volume value
3808 Note that when @option{eval} is set to @samp{once} only the
3809 @var{sample_rate} and @var{tb} variables are available, all other
3810 variables will evaluate to NAN.
3812 @subsection Commands
3814 This filter supports the following commands:
3817 Modify the volume expression.
3818 The command accepts the same syntax of the corresponding option.
3820 If the specified expression is not valid, it is kept at its current
3822 @item replaygain_noclip
3823 Prevent clipping by limiting the gain applied.
3825 Default value for @var{replaygain_noclip} is 1.
3829 @subsection Examples
3833 Halve the input audio volume:
3837 volume=volume=-6.0206dB
3840 In all the above example the named key for @option{volume} can be
3841 omitted, for example like in:
3847 Increase input audio power by 6 decibels using fixed-point precision:
3849 volume=volume=6dB:precision=fixed
3853 Fade volume after time 10 with an annihilation period of 5 seconds:
3855 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3859 @section volumedetect
3861 Detect the volume of the input video.
3863 The filter has no parameters. The input is not modified. Statistics about
3864 the volume will be printed in the log when the input stream end is reached.
3866 In particular it will show the mean volume (root mean square), maximum
3867 volume (on a per-sample basis), and the beginning of a histogram of the
3868 registered volume values (from the maximum value to a cumulated 1/1000 of
3871 All volumes are in decibels relative to the maximum PCM value.
3873 @subsection Examples
3875 Here is an excerpt of the output:
3877 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3878 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3879 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3880 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3881 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3882 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3883 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3884 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3885 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3891 The mean square energy is approximately -27 dB, or 10^-2.7.
3893 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3895 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3898 In other words, raising the volume by +4 dB does not cause any clipping,
3899 raising it by +5 dB causes clipping for 6 samples, etc.
3901 @c man end AUDIO FILTERS
3903 @chapter Audio Sources
3904 @c man begin AUDIO SOURCES
3906 Below is a description of the currently available audio sources.
3910 Buffer audio frames, and make them available to the filter chain.
3912 This source is mainly intended for a programmatic use, in particular
3913 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3915 It accepts the following parameters:
3919 The timebase which will be used for timestamps of submitted frames. It must be
3920 either a floating-point number or in @var{numerator}/@var{denominator} form.
3923 The sample rate of the incoming audio buffers.
3926 The sample format of the incoming audio buffers.
3927 Either a sample format name or its corresponding integer representation from
3928 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3930 @item channel_layout
3931 The channel layout of the incoming audio buffers.
3932 Either a channel layout name from channel_layout_map in
3933 @file{libavutil/channel_layout.c} or its corresponding integer representation
3934 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3937 The number of channels of the incoming audio buffers.
3938 If both @var{channels} and @var{channel_layout} are specified, then they
3943 @subsection Examples
3946 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3949 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3950 Since the sample format with name "s16p" corresponds to the number
3951 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3954 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3959 Generate an audio signal specified by an expression.
3961 This source accepts in input one or more expressions (one for each
3962 channel), which are evaluated and used to generate a corresponding
3965 This source accepts the following options:
3969 Set the '|'-separated expressions list for each separate channel. In case the
3970 @option{channel_layout} option is not specified, the selected channel layout
3971 depends on the number of provided expressions. Otherwise the last
3972 specified expression is applied to the remaining output channels.
3974 @item channel_layout, c
3975 Set the channel layout. The number of channels in the specified layout
3976 must be equal to the number of specified expressions.
3979 Set the minimum duration of the sourced audio. See
3980 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3981 for the accepted syntax.
3982 Note that the resulting duration may be greater than the specified
3983 duration, as the generated audio is always cut at the end of a
3986 If not specified, or the expressed duration is negative, the audio is
3987 supposed to be generated forever.
3990 Set the number of samples per channel per each output frame,
3993 @item sample_rate, s
3994 Specify the sample rate, default to 44100.
3997 Each expression in @var{exprs} can contain the following constants:
4001 number of the evaluated sample, starting from 0
4004 time of the evaluated sample expressed in seconds, starting from 0
4011 @subsection Examples
4021 Generate a sin signal with frequency of 440 Hz, set sample rate to
4024 aevalsrc="sin(440*2*PI*t):s=8000"
4028 Generate a two channels signal, specify the channel layout (Front
4029 Center + Back Center) explicitly:
4031 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4035 Generate white noise:
4037 aevalsrc="-2+random(0)"
4041 Generate an amplitude modulated signal:
4043 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4047 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4049 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4056 The null audio source, return unprocessed audio frames. It is mainly useful
4057 as a template and to be employed in analysis / debugging tools, or as
4058 the source for filters which ignore the input data (for example the sox
4061 This source accepts the following options:
4065 @item channel_layout, cl
4067 Specifies the channel layout, and can be either an integer or a string
4068 representing a channel layout. The default value of @var{channel_layout}
4071 Check the channel_layout_map definition in
4072 @file{libavutil/channel_layout.c} for the mapping between strings and
4073 channel layout values.
4075 @item sample_rate, r
4076 Specifies the sample rate, and defaults to 44100.
4079 Set the number of samples per requested frames.
4083 @subsection Examples
4087 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4089 anullsrc=r=48000:cl=4
4093 Do the same operation with a more obvious syntax:
4095 anullsrc=r=48000:cl=mono
4099 All the parameters need to be explicitly defined.
4103 Synthesize a voice utterance using the libflite library.
4105 To enable compilation of this filter you need to configure FFmpeg with
4106 @code{--enable-libflite}.
4108 Note that the flite library is not thread-safe.
4110 The filter accepts the following options:
4115 If set to 1, list the names of the available voices and exit
4116 immediately. Default value is 0.
4119 Set the maximum number of samples per frame. Default value is 512.
4122 Set the filename containing the text to speak.
4125 Set the text to speak.
4128 Set the voice to use for the speech synthesis. Default value is
4129 @code{kal}. See also the @var{list_voices} option.
4132 @subsection Examples
4136 Read from file @file{speech.txt}, and synthesize the text using the
4137 standard flite voice:
4139 flite=textfile=speech.txt
4143 Read the specified text selecting the @code{slt} voice:
4145 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4149 Input text to ffmpeg:
4151 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4155 Make @file{ffplay} speak the specified text, using @code{flite} and
4156 the @code{lavfi} device:
4158 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4162 For more information about libflite, check:
4163 @url{http://www.speech.cs.cmu.edu/flite/}
4167 Generate a noise audio signal.
4169 The filter accepts the following options:
4172 @item sample_rate, r
4173 Specify the sample rate. Default value is 48000 Hz.
4176 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4180 Specify the duration of the generated audio stream. Not specifying this option
4181 results in noise with an infinite length.
4183 @item color, colour, c
4184 Specify the color of noise. Available noise colors are white, pink, and brown.
4185 Default color is white.
4188 Specify a value used to seed the PRNG.
4191 Set the number of samples per each output frame, default is 1024.
4194 @subsection Examples
4199 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4201 anoisesrc=d=60:c=pink:r=44100:a=0.5
4207 Generate an audio signal made of a sine wave with amplitude 1/8.
4209 The audio signal is bit-exact.
4211 The filter accepts the following options:
4216 Set the carrier frequency. Default is 440 Hz.
4218 @item beep_factor, b
4219 Enable a periodic beep every second with frequency @var{beep_factor} times
4220 the carrier frequency. Default is 0, meaning the beep is disabled.
4222 @item sample_rate, r
4223 Specify the sample rate, default is 44100.
4226 Specify the duration of the generated audio stream.
4228 @item samples_per_frame
4229 Set the number of samples per output frame.
4231 The expression can contain the following constants:
4235 The (sequential) number of the output audio frame, starting from 0.
4238 The PTS (Presentation TimeStamp) of the output audio frame,
4239 expressed in @var{TB} units.
4242 The PTS of the output audio frame, expressed in seconds.
4245 The timebase of the output audio frames.
4248 Default is @code{1024}.
4251 @subsection Examples
4256 Generate a simple 440 Hz sine wave:
4262 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4266 sine=frequency=220:beep_factor=4:duration=5
4270 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4273 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4277 @c man end AUDIO SOURCES
4279 @chapter Audio Sinks
4280 @c man begin AUDIO SINKS
4282 Below is a description of the currently available audio sinks.
4284 @section abuffersink
4286 Buffer audio frames, and make them available to the end of filter chain.
4288 This sink is mainly intended for programmatic use, in particular
4289 through the interface defined in @file{libavfilter/buffersink.h}
4290 or the options system.
4292 It accepts a pointer to an AVABufferSinkContext structure, which
4293 defines the incoming buffers' formats, to be passed as the opaque
4294 parameter to @code{avfilter_init_filter} for initialization.
4297 Null audio sink; do absolutely nothing with the input audio. It is
4298 mainly useful as a template and for use in analysis / debugging
4301 @c man end AUDIO SINKS
4303 @chapter Video Filters
4304 @c man begin VIDEO FILTERS
4306 When you configure your FFmpeg build, you can disable any of the
4307 existing filters using @code{--disable-filters}.
4308 The configure output will show the video filters included in your
4311 Below is a description of the currently available video filters.
4313 @section alphaextract
4315 Extract the alpha component from the input as a grayscale video. This
4316 is especially useful with the @var{alphamerge} filter.
4320 Add or replace the alpha component of the primary input with the
4321 grayscale value of a second input. This is intended for use with
4322 @var{alphaextract} to allow the transmission or storage of frame
4323 sequences that have alpha in a format that doesn't support an alpha
4326 For example, to reconstruct full frames from a normal YUV-encoded video
4327 and a separate video created with @var{alphaextract}, you might use:
4329 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4332 Since this filter is designed for reconstruction, it operates on frame
4333 sequences without considering timestamps, and terminates when either
4334 input reaches end of stream. This will cause problems if your encoding
4335 pipeline drops frames. If you're trying to apply an image as an
4336 overlay to a video stream, consider the @var{overlay} filter instead.
4340 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4341 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4342 Substation Alpha) subtitles files.
4344 This filter accepts the following option in addition to the common options from
4345 the @ref{subtitles} filter:
4349 Set the shaping engine
4351 Available values are:
4354 The default libass shaping engine, which is the best available.
4356 Fast, font-agnostic shaper that can do only substitutions
4358 Slower shaper using OpenType for substitutions and positioning
4361 The default is @code{auto}.
4365 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4367 The filter accepts the following options:
4371 Set threshold A for 1st plane. Default is 0.02.
4372 Valid range is 0 to 0.3.
4375 Set threshold B for 1st plane. Default is 0.04.
4376 Valid range is 0 to 5.
4379 Set threshold A for 2nd plane. Default is 0.02.
4380 Valid range is 0 to 0.3.
4383 Set threshold B for 2nd plane. Default is 0.04.
4384 Valid range is 0 to 5.
4387 Set threshold A for 3rd plane. Default is 0.02.
4388 Valid range is 0 to 0.3.
4391 Set threshold B for 3rd plane. Default is 0.04.
4392 Valid range is 0 to 5.
4394 Threshold A is designed to react on abrupt changes in the input signal and
4395 threshold B is designed to react on continuous changes in the input signal.
4398 Set number of frames filter will use for averaging. Default is 33. Must be odd
4399 number in range [5, 129].
4404 Compute the bounding box for the non-black pixels in the input frame
4407 This filter computes the bounding box containing all the pixels with a
4408 luminance value greater than the minimum allowed value.
4409 The parameters describing the bounding box are printed on the filter
4412 The filter accepts the following option:
4416 Set the minimal luminance value. Default is @code{16}.
4419 @section bitplanenoise
4421 Show and measure bit plane noise.
4423 The filter accepts the following options:
4427 Set which plane to analyze. Default is @code{1}.
4430 Filter out noisy pixels from @code{bitplane} set above.
4431 Default is disabled.
4434 @section blackdetect
4436 Detect video intervals that are (almost) completely black. Can be
4437 useful to detect chapter transitions, commercials, or invalid
4438 recordings. Output lines contains the time for the start, end and
4439 duration of the detected black interval expressed in seconds.
4441 In order to display the output lines, you need to set the loglevel at
4442 least to the AV_LOG_INFO value.
4444 The filter accepts the following options:
4447 @item black_min_duration, d
4448 Set the minimum detected black duration expressed in seconds. It must
4449 be a non-negative floating point number.
4451 Default value is 2.0.
4453 @item picture_black_ratio_th, pic_th
4454 Set the threshold for considering a picture "black".
4455 Express the minimum value for the ratio:
4457 @var{nb_black_pixels} / @var{nb_pixels}
4460 for which a picture is considered black.
4461 Default value is 0.98.
4463 @item pixel_black_th, pix_th
4464 Set the threshold for considering a pixel "black".
4466 The threshold expresses the maximum pixel luminance value for which a
4467 pixel is considered "black". The provided value is scaled according to
4468 the following equation:
4470 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4473 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4474 the input video format, the range is [0-255] for YUV full-range
4475 formats and [16-235] for YUV non full-range formats.
4477 Default value is 0.10.
4480 The following example sets the maximum pixel threshold to the minimum
4481 value, and detects only black intervals of 2 or more seconds:
4483 blackdetect=d=2:pix_th=0.00
4488 Detect frames that are (almost) completely black. Can be useful to
4489 detect chapter transitions or commercials. Output lines consist of
4490 the frame number of the detected frame, the percentage of blackness,
4491 the position in the file if known or -1 and the timestamp in seconds.
4493 In order to display the output lines, you need to set the loglevel at
4494 least to the AV_LOG_INFO value.
4496 It accepts the following parameters:
4501 The percentage of the pixels that have to be below the threshold; it defaults to
4504 @item threshold, thresh
4505 The threshold below which a pixel value is considered black; it defaults to
4510 @section blend, tblend
4512 Blend two video frames into each other.
4514 The @code{blend} filter takes two input streams and outputs one
4515 stream, the first input is the "top" layer and second input is
4516 "bottom" layer. Output terminates when shortest input terminates.
4518 The @code{tblend} (time blend) filter takes two consecutive frames
4519 from one single stream, and outputs the result obtained by blending
4520 the new frame on top of the old frame.
4522 A description of the accepted options follows.
4530 Set blend mode for specific pixel component or all pixel components in case
4531 of @var{all_mode}. Default value is @code{normal}.
4533 Available values for component modes are:
4574 Set blend opacity for specific pixel component or all pixel components in case
4575 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4582 Set blend expression for specific pixel component or all pixel components in case
4583 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4585 The expressions can use the following variables:
4589 The sequential number of the filtered frame, starting from @code{0}.
4593 the coordinates of the current sample
4597 the width and height of currently filtered plane
4601 Width and height scale depending on the currently filtered plane. It is the
4602 ratio between the corresponding luma plane number of pixels and the current
4603 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4604 @code{0.5,0.5} for chroma planes.
4607 Time of the current frame, expressed in seconds.
4610 Value of pixel component at current location for first video frame (top layer).
4613 Value of pixel component at current location for second video frame (bottom layer).
4617 Force termination when the shortest input terminates. Default is
4618 @code{0}. This option is only defined for the @code{blend} filter.
4621 Continue applying the last bottom frame after the end of the stream. A value of
4622 @code{0} disable the filter after the last frame of the bottom layer is reached.
4623 Default is @code{1}. This option is only defined for the @code{blend} filter.
4626 @subsection Examples
4630 Apply transition from bottom layer to top layer in first 10 seconds:
4632 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4636 Apply 1x1 checkerboard effect:
4638 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4642 Apply uncover left effect:
4644 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4648 Apply uncover down effect:
4650 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4654 Apply uncover up-left effect:
4656 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4660 Split diagonally video and shows top and bottom layer on each side:
4662 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4666 Display differences between the current and the previous frame:
4668 tblend=all_mode=difference128
4674 Apply a boxblur algorithm to the input video.
4676 It accepts the following parameters:
4680 @item luma_radius, lr
4681 @item luma_power, lp
4682 @item chroma_radius, cr
4683 @item chroma_power, cp
4684 @item alpha_radius, ar
4685 @item alpha_power, ap
4689 A description of the accepted options follows.
4692 @item luma_radius, lr
4693 @item chroma_radius, cr
4694 @item alpha_radius, ar
4695 Set an expression for the box radius in pixels used for blurring the
4696 corresponding input plane.
4698 The radius value must be a non-negative number, and must not be
4699 greater than the value of the expression @code{min(w,h)/2} for the
4700 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4703 Default value for @option{luma_radius} is "2". If not specified,
4704 @option{chroma_radius} and @option{alpha_radius} default to the
4705 corresponding value set for @option{luma_radius}.
4707 The expressions can contain the following constants:
4711 The input width and height in pixels.
4715 The input chroma image width and height in pixels.
4719 The horizontal and vertical chroma subsample values. For example, for the
4720 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4723 @item luma_power, lp
4724 @item chroma_power, cp
4725 @item alpha_power, ap
4726 Specify how many times the boxblur filter is applied to the
4727 corresponding plane.
4729 Default value for @option{luma_power} is 2. If not specified,
4730 @option{chroma_power} and @option{alpha_power} default to the
4731 corresponding value set for @option{luma_power}.
4733 A value of 0 will disable the effect.
4736 @subsection Examples
4740 Apply a boxblur filter with the luma, chroma, and alpha radii
4743 boxblur=luma_radius=2:luma_power=1
4748 Set the luma radius to 2, and alpha and chroma radius to 0:
4750 boxblur=2:1:cr=0:ar=0
4754 Set the luma and chroma radii to a fraction of the video dimension:
4756 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4762 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4763 Deinterlacing Filter").
4765 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4766 interpolation algorithms.
4767 It accepts the following parameters:
4771 The interlacing mode to adopt. It accepts one of the following values:
4775 Output one frame for each frame.
4777 Output one frame for each field.
4780 The default value is @code{send_field}.
4783 The picture field parity assumed for the input interlaced video. It accepts one
4784 of the following values:
4788 Assume the top field is first.
4790 Assume the bottom field is first.
4792 Enable automatic detection of field parity.
4795 The default value is @code{auto}.
4796 If the interlacing is unknown or the decoder does not export this information,
4797 top field first will be assumed.
4800 Specify which frames to deinterlace. Accept one of the following
4805 Deinterlace all frames.
4807 Only deinterlace frames marked as interlaced.
4810 The default value is @code{all}.
4814 YUV colorspace color/chroma keying.
4816 The filter accepts the following options:
4820 The color which will be replaced with transparency.
4823 Similarity percentage with the key color.
4825 0.01 matches only the exact key color, while 1.0 matches everything.
4830 0.0 makes pixels either fully transparent, or not transparent at all.
4832 Higher values result in semi-transparent pixels, with a higher transparency
4833 the more similar the pixels color is to the key color.
4836 Signals that the color passed is already in YUV instead of RGB.
4838 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4839 This can be used to pass exact YUV values as hexadecimal numbers.
4842 @subsection Examples
4846 Make every green pixel in the input image transparent:
4848 ffmpeg -i input.png -vf chromakey=green out.png
4852 Overlay a greenscreen-video on top of a static black background.
4854 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
4860 Display CIE color diagram with pixels overlaid onto it.
4862 The filter accepts the following options:
4877 @item uhdtv, rec2020
4890 Set what gamuts to draw.
4892 See @code{system} option for available values.
4895 Set ciescope size, by default set to 512.
4898 Set intensity used to map input pixel values to CIE diagram.
4901 Set contrast used to draw tongue colors that are out of active color system gamut.
4904 Correct gamma displayed on scope, by default enabled.
4907 Show white point on CIE diagram, by default disabled.
4910 Set input gamma. Used only with XYZ input color space.
4915 Visualize information exported by some codecs.
4917 Some codecs can export information through frames using side-data or other
4918 means. For example, some MPEG based codecs export motion vectors through the
4919 @var{export_mvs} flag in the codec @option{flags2} option.
4921 The filter accepts the following option:
4925 Set motion vectors to visualize.
4927 Available flags for @var{mv} are:
4931 forward predicted MVs of P-frames
4933 forward predicted MVs of B-frames
4935 backward predicted MVs of B-frames
4939 Display quantization parameters using the chroma planes.
4942 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4944 Available flags for @var{mv_type} are:
4948 forward predicted MVs
4950 backward predicted MVs
4953 @item frame_type, ft
4954 Set frame type to visualize motion vectors of.
4956 Available flags for @var{frame_type} are:
4960 intra-coded frames (I-frames)
4962 predicted frames (P-frames)
4964 bi-directionally predicted frames (B-frames)
4968 @subsection Examples
4972 Visualize forward predicted MVs of all frames using @command{ffplay}:
4974 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
4978 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
4980 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
4984 @section colorbalance
4985 Modify intensity of primary colors (red, green and blue) of input frames.
4987 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
4988 regions for the red-cyan, green-magenta or blue-yellow balance.
4990 A positive adjustment value shifts the balance towards the primary color, a negative
4991 value towards the complementary color.
4993 The filter accepts the following options:
4999 Adjust red, green and blue shadows (darkest pixels).
5004 Adjust red, green and blue midtones (medium pixels).
5009 Adjust red, green and blue highlights (brightest pixels).
5011 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5014 @subsection Examples
5018 Add red color cast to shadows:
5025 RGB colorspace color keying.
5027 The filter accepts the following options:
5031 The color which will be replaced with transparency.
5034 Similarity percentage with the key color.
5036 0.01 matches only the exact key color, while 1.0 matches everything.
5041 0.0 makes pixels either fully transparent, or not transparent at all.
5043 Higher values result in semi-transparent pixels, with a higher transparency
5044 the more similar the pixels color is to the key color.
5047 @subsection Examples
5051 Make every green pixel in the input image transparent:
5053 ffmpeg -i input.png -vf colorkey=green out.png
5057 Overlay a greenscreen-video on top of a static background image.
5059 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
5063 @section colorlevels
5065 Adjust video input frames using levels.
5067 The filter accepts the following options:
5074 Adjust red, green, blue and alpha input black point.
5075 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5081 Adjust red, green, blue and alpha input white point.
5082 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5084 Input levels are used to lighten highlights (bright tones), darken shadows
5085 (dark tones), change the balance of bright and dark tones.
5091 Adjust red, green, blue and alpha output black point.
5092 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5098 Adjust red, green, blue and alpha output white point.
5099 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5101 Output levels allows manual selection of a constrained output level range.
5104 @subsection Examples
5108 Make video output darker:
5110 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5116 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5120 Make video output lighter:
5122 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5126 Increase brightness:
5128 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5132 @section colorchannelmixer
5134 Adjust video input frames by re-mixing color channels.
5136 This filter modifies a color channel by adding the values associated to
5137 the other channels of the same pixels. For example if the value to
5138 modify is red, the output value will be:
5140 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5143 The filter accepts the following options:
5150 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5151 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5157 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5158 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5164 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5165 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5171 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5172 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5174 Allowed ranges for options are @code{[-2.0, 2.0]}.
5177 @subsection Examples
5181 Convert source to grayscale:
5183 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5186 Simulate sepia tones:
5188 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5192 @section colormatrix
5194 Convert color matrix.
5196 The filter accepts the following options:
5201 Specify the source and destination color matrix. Both values must be
5204 The accepted values are:
5223 For example to convert from BT.601 to SMPTE-240M, use the command:
5225 colormatrix=bt601:smpte240m
5230 Convert colorspace, transfer characteristics or color primaries.
5232 The filter accepts the following options:
5236 Specify all color properties at once.
5238 The accepted values are:
5267 Specify output colorspace.
5269 The accepted values are:
5278 BT.470BG or BT.601-6 625
5281 SMPTE-170M or BT.601-6 525
5287 BT.2020 with non-constant luminance
5292 Specify output transfer characteristics.
5294 The accepted values are:
5300 Constant gamma of 2.2
5303 Constant gamma of 2.8
5306 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5312 BT.2020 for 10-bits content
5315 BT.2020 for 12-bits content
5320 Specify output color primaries.
5322 The accepted values are:
5331 BT.470BG or BT.601-6 625
5334 SMPTE-170M or BT.601-6 525
5345 Specify output color range.
5347 The accepted values are:
5350 MPEG (restricted) range
5358 Specify output color format.
5360 The accepted values are:
5363 YUV 4:2:0 planar 8-bits
5366 YUV 4:2:0 planar 10-bits
5369 YUV 4:2:0 planar 12-bits
5372 YUV 4:2:2 planar 8-bits
5375 YUV 4:2:2 planar 10-bits
5378 YUV 4:2:2 planar 12-bits
5381 YUV 4:4:4 planar 8-bits
5384 YUV 4:4:4 planar 10-bits
5387 YUV 4:4:4 planar 12-bits
5392 Do a fast conversion, which skips gamma/primary correction. This will take
5393 significantly less CPU, but will be mathematically incorrect. To get output
5394 compatible with that produced by the colormatrix filter, use fast=1.
5397 Specify dithering mode.
5399 The accepted values are:
5405 Floyd-Steinberg dithering
5409 Whitepoint adaptation mode.
5411 The accepted values are:
5414 Bradford whitepoint adaptation
5417 von Kries whitepoint adaptation
5420 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5425 The filter converts the transfer characteristics, color space and color
5426 primaries to the specified user values. The output value, if not specified,
5427 is set to a default value based on the "all" property. If that property is
5428 also not specified, the filter will log an error. The output color range and
5429 format default to the same value as the input color range and format. The
5430 input transfer characteristics, color space, color primaries and color range
5431 should be set on the input data. If any of these are missing, the filter will
5432 log an error and no conversion will take place.
5434 For example to convert the input to SMPTE-240M, use the command:
5436 colorspace=smpte240m
5439 @section convolution
5441 Apply convolution 3x3 or 5x5 filter.
5443 The filter accepts the following options:
5450 Set matrix for each plane.
5451 Matrix is sequence of 9 or 25 signed integers.
5457 Set multiplier for calculated value for each plane.
5463 Set bias for each plane. This value is added to the result of the multiplication.
5464 Useful for making the overall image brighter or darker. Default is 0.0.
5467 @subsection Examples
5473 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"
5479 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"
5485 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"
5491 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"
5497 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"
5503 Copy the input source unchanged to the output. This is mainly useful for
5508 Video filtering on GPU using Apple's CoreImage API on OSX.
5510 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5511 processed by video hardware. However, software-based OpenGL implementations
5512 exist which means there is no guarantee for hardware processing. It depends on
5515 There are many filters and image generators provided by Apple that come with a
5516 large variety of options. The filter has to be referenced by its name along
5519 The coreimage filter accepts the following options:
5522 List all available filters and generators along with all their respective
5523 options as well as possible minimum and maximum values along with the default
5530 Specify all filters by their respective name and options.
5531 Use @var{list_filters} to determine all valid filter names and options.
5532 Numerical options are specified by a float value and are automatically clamped
5533 to their respective value range. Vector and color options have to be specified
5534 by a list of space separated float values. Character escaping has to be done.
5535 A special option name @code{default} is available to use default options for a
5538 It is required to specify either @code{default} or at least one of the filter options.
5539 All omitted options are used with their default values.
5540 The syntax of the filter string is as follows:
5542 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5546 Specify a rectangle where the output of the filter chain is copied into the
5547 input image. It is given by a list of space separated float values:
5549 output_rect=x\ y\ width\ height
5551 If not given, the output rectangle equals the dimensions of the input image.
5552 The output rectangle is automatically cropped at the borders of the input
5553 image. Negative values are valid for each component.
5555 output_rect=25\ 25\ 100\ 100
5559 Several filters can be chained for successive processing without GPU-HOST
5560 transfers allowing for fast processing of complex filter chains.
5561 Currently, only filters with zero (generators) or exactly one (filters) input
5562 image and one output image are supported. Also, transition filters are not yet
5565 Some filters generate output images with additional padding depending on the
5566 respective filter kernel. The padding is automatically removed to ensure the
5567 filter output has the same size as the input image.
5569 For image generators, the size of the output image is determined by the
5570 previous output image of the filter chain or the input image of the whole
5571 filterchain, respectively. The generators do not use the pixel information of
5572 this image to generate their output. However, the generated output is
5573 blended onto this image, resulting in partial or complete coverage of the
5576 The @ref{coreimagesrc} video source can be used for generating input images
5577 which are directly fed into the filter chain. By using it, providing input
5578 images by another video source or an input video is not required.
5580 @subsection Examples
5585 List all filters available:
5587 coreimage=list_filters=true
5591 Use the CIBoxBlur filter with default options to blur an image:
5593 coreimage=filter=CIBoxBlur@@default
5597 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5598 its center at 100x100 and a radius of 50 pixels:
5600 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5604 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5605 given as complete and escaped command-line for Apple's standard bash shell:
5607 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5613 Crop the input video to given dimensions.
5615 It accepts the following parameters:
5619 The width of the output video. It defaults to @code{iw}.
5620 This expression is evaluated only once during the filter
5621 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5624 The height of the output video. It defaults to @code{ih}.
5625 This expression is evaluated only once during the filter
5626 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5629 The horizontal position, in the input video, of the left edge of the output
5630 video. It defaults to @code{(in_w-out_w)/2}.
5631 This expression is evaluated per-frame.
5634 The vertical position, in the input video, of the top edge of the output video.
5635 It defaults to @code{(in_h-out_h)/2}.
5636 This expression is evaluated per-frame.
5639 If set to 1 will force the output display aspect ratio
5640 to be the same of the input, by changing the output sample aspect
5641 ratio. It defaults to 0.
5644 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5645 width/height/x/y as specified and will not be rounded to nearest smaller value.
5649 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5650 expressions containing the following constants:
5655 The computed values for @var{x} and @var{y}. They are evaluated for
5660 The input width and height.
5664 These are the same as @var{in_w} and @var{in_h}.
5668 The output (cropped) width and height.
5672 These are the same as @var{out_w} and @var{out_h}.
5675 same as @var{iw} / @var{ih}
5678 input sample aspect ratio
5681 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5685 horizontal and vertical chroma subsample values. For example for the
5686 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5689 The number of the input frame, starting from 0.
5692 the position in the file of the input frame, NAN if unknown
5695 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5699 The expression for @var{out_w} may depend on the value of @var{out_h},
5700 and the expression for @var{out_h} may depend on @var{out_w}, but they
5701 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5702 evaluated after @var{out_w} and @var{out_h}.
5704 The @var{x} and @var{y} parameters specify the expressions for the
5705 position of the top-left corner of the output (non-cropped) area. They
5706 are evaluated for each frame. If the evaluated value is not valid, it
5707 is approximated to the nearest valid value.
5709 The expression for @var{x} may depend on @var{y}, and the expression
5710 for @var{y} may depend on @var{x}.
5712 @subsection Examples
5716 Crop area with size 100x100 at position (12,34).
5721 Using named options, the example above becomes:
5723 crop=w=100:h=100:x=12:y=34
5727 Crop the central input area with size 100x100:
5733 Crop the central input area with size 2/3 of the input video:
5735 crop=2/3*in_w:2/3*in_h
5739 Crop the input video central square:
5746 Delimit the rectangle with the top-left corner placed at position
5747 100:100 and the right-bottom corner corresponding to the right-bottom
5748 corner of the input image.
5750 crop=in_w-100:in_h-100:100:100
5754 Crop 10 pixels from the left and right borders, and 20 pixels from
5755 the top and bottom borders
5757 crop=in_w-2*10:in_h-2*20
5761 Keep only the bottom right quarter of the input image:
5763 crop=in_w/2:in_h/2:in_w/2:in_h/2
5767 Crop height for getting Greek harmony:
5769 crop=in_w:1/PHI*in_w
5773 Apply trembling effect:
5775 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)
5779 Apply erratic camera effect depending on timestamp:
5781 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)"
5785 Set x depending on the value of y:
5787 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5791 @subsection Commands
5793 This filter supports the following commands:
5799 Set width/height of the output video and the horizontal/vertical position
5801 The command accepts the same syntax of the corresponding option.
5803 If the specified expression is not valid, it is kept at its current
5809 Auto-detect the crop size.
5811 It calculates the necessary cropping parameters and prints the
5812 recommended parameters via the logging system. The detected dimensions
5813 correspond to the non-black area of the input video.
5815 It accepts the following parameters:
5820 Set higher black value threshold, which can be optionally specified
5821 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5822 value greater to the set value is considered non-black. It defaults to 24.
5823 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5824 on the bitdepth of the pixel format.
5827 The value which the width/height should be divisible by. It defaults to
5828 16. The offset is automatically adjusted to center the video. Use 2 to
5829 get only even dimensions (needed for 4:2:2 video). 16 is best when
5830 encoding to most video codecs.
5832 @item reset_count, reset
5833 Set the counter that determines after how many frames cropdetect will
5834 reset the previously detected largest video area and start over to
5835 detect the current optimal crop area. Default value is 0.
5837 This can be useful when channel logos distort the video area. 0
5838 indicates 'never reset', and returns the largest area encountered during
5845 Apply color adjustments using curves.
5847 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5848 component (red, green and blue) has its values defined by @var{N} key points
5849 tied from each other using a smooth curve. The x-axis represents the pixel
5850 values from the input frame, and the y-axis the new pixel values to be set for
5853 By default, a component curve is defined by the two points @var{(0;0)} and
5854 @var{(1;1)}. This creates a straight line where each original pixel value is
5855 "adjusted" to its own value, which means no change to the image.
5857 The filter allows you to redefine these two points and add some more. A new
5858 curve (using a natural cubic spline interpolation) will be define to pass
5859 smoothly through all these new coordinates. The new defined points needs to be
5860 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5861 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5862 the vector spaces, the values will be clipped accordingly.
5864 The filter accepts the following options:
5868 Select one of the available color presets. This option can be used in addition
5869 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5870 options takes priority on the preset values.
5871 Available presets are:
5874 @item color_negative
5877 @item increase_contrast
5879 @item linear_contrast
5880 @item medium_contrast
5882 @item strong_contrast
5885 Default is @code{none}.
5887 Set the master key points. These points will define a second pass mapping. It
5888 is sometimes called a "luminance" or "value" mapping. It can be used with
5889 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5890 post-processing LUT.
5892 Set the key points for the red component.
5894 Set the key points for the green component.
5896 Set the key points for the blue component.
5898 Set the key points for all components (not including master).
5899 Can be used in addition to the other key points component
5900 options. In this case, the unset component(s) will fallback on this
5901 @option{all} setting.
5903 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5905 Save Gnuplot script of the curves in specified file.
5908 To avoid some filtergraph syntax conflicts, each key points list need to be
5909 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5911 @subsection Examples
5915 Increase slightly the middle level of blue:
5917 curves=blue='0/0 0.5/0.58 1/1'
5923 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'
5925 Here we obtain the following coordinates for each components:
5928 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5930 @code{(0;0) (0.50;0.48) (1;1)}
5932 @code{(0;0.22) (0.49;0.44) (1;0.80)}
5936 The previous example can also be achieved with the associated built-in preset:
5938 curves=preset=vintage
5948 Use a Photoshop preset and redefine the points of the green component:
5950 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
5954 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
5955 and @command{gnuplot}:
5957 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
5958 gnuplot -p /tmp/curves.plt
5964 Video data analysis filter.
5966 This filter shows hexadecimal pixel values of part of video.
5968 The filter accepts the following options:
5972 Set output video size.
5975 Set x offset from where to pick pixels.
5978 Set y offset from where to pick pixels.
5981 Set scope mode, can be one of the following:
5984 Draw hexadecimal pixel values with white color on black background.
5987 Draw hexadecimal pixel values with input video pixel color on black
5991 Draw hexadecimal pixel values on color background picked from input video,
5992 the text color is picked in such way so its always visible.
5996 Draw rows and columns numbers on left and top of video.
6001 Denoise frames using 2D DCT (frequency domain filtering).
6003 This filter is not designed for real time.
6005 The filter accepts the following options:
6009 Set the noise sigma constant.
6011 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6012 coefficient (absolute value) below this threshold with be dropped.
6014 If you need a more advanced filtering, see @option{expr}.
6016 Default is @code{0}.
6019 Set number overlapping pixels for each block. Since the filter can be slow, you
6020 may want to reduce this value, at the cost of a less effective filter and the
6021 risk of various artefacts.
6023 If the overlapping value doesn't permit processing the whole input width or
6024 height, a warning will be displayed and according borders won't be denoised.
6026 Default value is @var{blocksize}-1, which is the best possible setting.
6029 Set the coefficient factor expression.
6031 For each coefficient of a DCT block, this expression will be evaluated as a
6032 multiplier value for the coefficient.
6034 If this is option is set, the @option{sigma} option will be ignored.
6036 The absolute value of the coefficient can be accessed through the @var{c}
6040 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6041 @var{blocksize}, which is the width and height of the processed blocks.
6043 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6044 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6045 on the speed processing. Also, a larger block size does not necessarily means a
6049 @subsection Examples
6051 Apply a denoise with a @option{sigma} of @code{4.5}:
6056 The same operation can be achieved using the expression system:
6058 dctdnoiz=e='gte(c, 4.5*3)'
6061 Violent denoise using a block size of @code{16x16}:
6068 Remove banding artifacts from input video.
6069 It works by replacing banded pixels with average value of referenced pixels.
6071 The filter accepts the following options:
6078 Set banding detection threshold for each plane. Default is 0.02.
6079 Valid range is 0.00003 to 0.5.
6080 If difference between current pixel and reference pixel is less than threshold,
6081 it will be considered as banded.
6084 Banding detection range in pixels. Default is 16. If positive, random number
6085 in range 0 to set value will be used. If negative, exact absolute value
6087 The range defines square of four pixels around current pixel.
6090 Set direction in radians from which four pixel will be compared. If positive,
6091 random direction from 0 to set direction will be picked. If negative, exact of
6092 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6093 will pick only pixels on same row and -PI/2 will pick only pixels on same
6097 If enabled, current pixel is compared with average value of all four
6098 surrounding pixels. The default is enabled. If disabled current pixel is
6099 compared with all four surrounding pixels. The pixel is considered banded
6100 if only all four differences with surrounding pixels are less than threshold.
6106 Drop duplicated frames at regular intervals.
6108 The filter accepts the following options:
6112 Set the number of frames from which one will be dropped. Setting this to
6113 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6114 Default is @code{5}.
6117 Set the threshold for duplicate detection. If the difference metric for a frame
6118 is less than or equal to this value, then it is declared as duplicate. Default
6122 Set scene change threshold. Default is @code{15}.
6126 Set the size of the x and y-axis blocks used during metric calculations.
6127 Larger blocks give better noise suppression, but also give worse detection of
6128 small movements. Must be a power of two. Default is @code{32}.
6131 Mark main input as a pre-processed input and activate clean source input
6132 stream. This allows the input to be pre-processed with various filters to help
6133 the metrics calculation while keeping the frame selection lossless. When set to
6134 @code{1}, the first stream is for the pre-processed input, and the second
6135 stream is the clean source from where the kept frames are chosen. Default is
6139 Set whether or not chroma is considered in the metric calculations. Default is
6145 Apply deflate effect to the video.
6147 This filter replaces the pixel by the local(3x3) average by taking into account
6148 only values lower than the pixel.
6150 It accepts the following options:
6157 Limit the maximum change for each plane, default is 65535.
6158 If 0, plane will remain unchanged.
6163 Remove judder produced by partially interlaced telecined content.
6165 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6166 source was partially telecined content then the output of @code{pullup,dejudder}
6167 will have a variable frame rate. May change the recorded frame rate of the
6168 container. Aside from that change, this filter will not affect constant frame
6171 The option available in this filter is:
6175 Specify the length of the window over which the judder repeats.
6177 Accepts any integer greater than 1. Useful values are:
6181 If the original was telecined from 24 to 30 fps (Film to NTSC).
6184 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6187 If a mixture of the two.
6190 The default is @samp{4}.
6195 Suppress a TV station logo by a simple interpolation of the surrounding
6196 pixels. Just set a rectangle covering the logo and watch it disappear
6197 (and sometimes something even uglier appear - your mileage may vary).
6199 It accepts the following parameters:
6204 Specify the top left corner coordinates of the logo. They must be
6209 Specify the width and height of the logo to clear. They must be
6213 Specify the thickness of the fuzzy edge of the rectangle (added to
6214 @var{w} and @var{h}). The default value is 1. This option is
6215 deprecated, setting higher values should no longer be necessary and
6219 When set to 1, a green rectangle is drawn on the screen to simplify
6220 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6221 The default value is 0.
6223 The rectangle is drawn on the outermost pixels which will be (partly)
6224 replaced with interpolated values. The values of the next pixels
6225 immediately outside this rectangle in each direction will be used to
6226 compute the interpolated pixel values inside the rectangle.
6230 @subsection Examples
6234 Set a rectangle covering the area with top left corner coordinates 0,0
6235 and size 100x77, and a band of size 10:
6237 delogo=x=0:y=0:w=100:h=77:band=10
6244 Attempt to fix small changes in horizontal and/or vertical shift. This
6245 filter helps remove camera shake from hand-holding a camera, bumping a
6246 tripod, moving on a vehicle, etc.
6248 The filter accepts the following options:
6256 Specify a rectangular area where to limit the search for motion
6258 If desired the search for motion vectors can be limited to a
6259 rectangular area of the frame defined by its top left corner, width
6260 and height. These parameters have the same meaning as the drawbox
6261 filter which can be used to visualise the position of the bounding
6264 This is useful when simultaneous movement of subjects within the frame
6265 might be confused for camera motion by the motion vector search.
6267 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6268 then the full frame is used. This allows later options to be set
6269 without specifying the bounding box for the motion vector search.
6271 Default - search the whole frame.
6275 Specify the maximum extent of movement in x and y directions in the
6276 range 0-64 pixels. Default 16.
6279 Specify how to generate pixels to fill blanks at the edge of the
6280 frame. Available values are:
6283 Fill zeroes at blank locations
6285 Original image at blank locations
6287 Extruded edge value at blank locations
6289 Mirrored edge at blank locations
6291 Default value is @samp{mirror}.
6294 Specify the blocksize to use for motion search. Range 4-128 pixels,
6298 Specify the contrast threshold for blocks. Only blocks with more than
6299 the specified contrast (difference between darkest and lightest
6300 pixels) will be considered. Range 1-255, default 125.
6303 Specify the search strategy. Available values are:
6306 Set exhaustive search
6308 Set less exhaustive search.
6310 Default value is @samp{exhaustive}.
6313 If set then a detailed log of the motion search is written to the
6317 If set to 1, specify using OpenCL capabilities, only available if
6318 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6324 Apply an exact inverse of the telecine operation. It requires a predefined
6325 pattern specified using the pattern option which must be the same as that passed
6326 to the telecine filter.
6328 This filter accepts the following options:
6337 The default value is @code{top}.
6341 A string of numbers representing the pulldown pattern you wish to apply.
6342 The default value is @code{23}.
6345 A number representing position of the first frame with respect to the telecine
6346 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6351 Apply dilation effect to the video.
6353 This filter replaces the pixel by the local(3x3) maximum.
6355 It accepts the following options:
6362 Limit the maximum change for each plane, default is 65535.
6363 If 0, plane will remain unchanged.
6366 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6369 Flags to local 3x3 coordinates maps like this:
6378 Displace pixels as indicated by second and third input stream.
6380 It takes three input streams and outputs one stream, the first input is the
6381 source, and second and third input are displacement maps.
6383 The second input specifies how much to displace pixels along the
6384 x-axis, while the third input specifies how much to displace pixels
6386 If one of displacement map streams terminates, last frame from that
6387 displacement map will be used.
6389 Note that once generated, displacements maps can be reused over and over again.
6391 A description of the accepted options follows.
6395 Set displace behavior for pixels that are out of range.
6397 Available values are:
6400 Missing pixels are replaced by black pixels.
6403 Adjacent pixels will spread out to replace missing pixels.
6406 Out of range pixels are wrapped so they point to pixels of other side.
6408 Default is @samp{smear}.
6412 @subsection Examples
6416 Add ripple effect to rgb input of video size hd720:
6418 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
6422 Add wave effect to rgb input of video size hd720:
6424 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
6430 Draw a colored box on the input image.
6432 It accepts the following parameters:
6437 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6441 The expressions which specify the width and height of the box; if 0 they are interpreted as
6442 the input width and height. It defaults to 0.
6445 Specify the color of the box to write. For the general syntax of this option,
6446 check the "Color" section in the ffmpeg-utils manual. If the special
6447 value @code{invert} is used, the box edge color is the same as the
6448 video with inverted luma.
6451 The expression which sets the thickness of the box edge. Default value is @code{3}.
6453 See below for the list of accepted constants.
6456 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6457 following constants:
6461 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6465 horizontal and vertical chroma subsample values. For example for the
6466 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6470 The input width and height.
6473 The input sample aspect ratio.
6477 The x and y offset coordinates where the box is drawn.
6481 The width and height of the drawn box.
6484 The thickness of the drawn box.
6486 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6487 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6491 @subsection Examples
6495 Draw a black box around the edge of the input image:
6501 Draw a box with color red and an opacity of 50%:
6503 drawbox=10:20:200:60:red@@0.5
6506 The previous example can be specified as:
6508 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6512 Fill the box with pink color:
6514 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6518 Draw a 2-pixel red 2.40:1 mask:
6520 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
6526 Draw a grid on the input image.
6528 It accepts the following parameters:
6533 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6537 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6538 input width and height, respectively, minus @code{thickness}, so image gets
6539 framed. Default to 0.
6542 Specify the color of the grid. For the general syntax of this option,
6543 check the "Color" section in the ffmpeg-utils manual. If the special
6544 value @code{invert} is used, the grid color is the same as the
6545 video with inverted luma.
6548 The expression which sets the thickness of the grid line. Default value is @code{1}.
6550 See below for the list of accepted constants.
6553 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6554 following constants:
6558 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6562 horizontal and vertical chroma subsample values. For example for the
6563 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6567 The input grid cell width and height.
6570 The input sample aspect ratio.
6574 The x and y coordinates of some point of grid intersection (meant to configure offset).
6578 The width and height of the drawn cell.
6581 The thickness of the drawn cell.
6583 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6584 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6588 @subsection Examples
6592 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6594 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6598 Draw a white 3x3 grid with an opacity of 50%:
6600 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6607 Draw a text string or text from a specified file on top of a video, using the
6608 libfreetype library.
6610 To enable compilation of this filter, you need to configure FFmpeg with
6611 @code{--enable-libfreetype}.
6612 To enable default font fallback and the @var{font} option you need to
6613 configure FFmpeg with @code{--enable-libfontconfig}.
6614 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6615 @code{--enable-libfribidi}.
6619 It accepts the following parameters:
6624 Used to draw a box around text using the background color.
6625 The value must be either 1 (enable) or 0 (disable).
6626 The default value of @var{box} is 0.
6629 Set the width of the border to be drawn around the box using @var{boxcolor}.
6630 The default value of @var{boxborderw} is 0.
6633 The color to be used for drawing box around text. For the syntax of this
6634 option, check the "Color" section in the ffmpeg-utils manual.
6636 The default value of @var{boxcolor} is "white".
6639 Set the width of the border to be drawn around the text using @var{bordercolor}.
6640 The default value of @var{borderw} is 0.
6643 Set the color to be used for drawing border around text. For the syntax of this
6644 option, check the "Color" section in the ffmpeg-utils manual.
6646 The default value of @var{bordercolor} is "black".
6649 Select how the @var{text} is expanded. Can be either @code{none},
6650 @code{strftime} (deprecated) or
6651 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6655 If true, check and fix text coords to avoid clipping.
6658 The color to be used for drawing fonts. For the syntax of this option, check
6659 the "Color" section in the ffmpeg-utils manual.
6661 The default value of @var{fontcolor} is "black".
6663 @item fontcolor_expr
6664 String which is expanded the same way as @var{text} to obtain dynamic
6665 @var{fontcolor} value. By default this option has empty value and is not
6666 processed. When this option is set, it overrides @var{fontcolor} option.
6669 The font family to be used for drawing text. By default Sans.
6672 The font file to be used for drawing text. The path must be included.
6673 This parameter is mandatory if the fontconfig support is disabled.
6676 This option does not exist, please see the timeline system
6679 Draw the text applying alpha blending. The value can
6680 be either a number between 0.0 and 1.0
6681 The expression accepts the same variables @var{x, y} do.
6682 The default value is 1.
6683 Please see fontcolor_expr
6686 The font size to be used for drawing text.
6687 The default value of @var{fontsize} is 16.
6690 If set to 1, attempt to shape the text (for example, reverse the order of
6691 right-to-left text and join Arabic characters) before drawing it.
6692 Otherwise, just draw the text exactly as given.
6693 By default 1 (if supported).
6696 The flags to be used for loading the fonts.
6698 The flags map the corresponding flags supported by libfreetype, and are
6699 a combination of the following values:
6706 @item vertical_layout
6707 @item force_autohint
6710 @item ignore_global_advance_width
6712 @item ignore_transform
6718 Default value is "default".
6720 For more information consult the documentation for the FT_LOAD_*
6724 The color to be used for drawing a shadow behind the drawn text. For the
6725 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6727 The default value of @var{shadowcolor} is "black".
6731 The x and y offsets for the text shadow position with respect to the
6732 position of the text. They can be either positive or negative
6733 values. The default value for both is "0".
6736 The starting frame number for the n/frame_num variable. The default value
6740 The size in number of spaces to use for rendering the tab.
6744 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6745 format. It can be used with or without text parameter. @var{timecode_rate}
6746 option must be specified.
6748 @item timecode_rate, rate, r
6749 Set the timecode frame rate (timecode only).
6752 The text string to be drawn. The text must be a sequence of UTF-8
6754 This parameter is mandatory if no file is specified with the parameter
6758 A text file containing text to be drawn. The text must be a sequence
6759 of UTF-8 encoded characters.
6761 This parameter is mandatory if no text string is specified with the
6762 parameter @var{text}.
6764 If both @var{text} and @var{textfile} are specified, an error is thrown.
6767 If set to 1, the @var{textfile} will be reloaded before each frame.
6768 Be sure to update it atomically, or it may be read partially, or even fail.
6772 The expressions which specify the offsets where text will be drawn
6773 within the video frame. They are relative to the top/left border of the
6776 The default value of @var{x} and @var{y} is "0".
6778 See below for the list of accepted constants and functions.
6781 The parameters for @var{x} and @var{y} are expressions containing the
6782 following constants and functions:
6786 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6790 horizontal and vertical chroma subsample values. For example for the
6791 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6794 the height of each text line
6802 @item max_glyph_a, ascent
6803 the maximum distance from the baseline to the highest/upper grid
6804 coordinate used to place a glyph outline point, for all the rendered
6806 It is a positive value, due to the grid's orientation with the Y axis
6809 @item max_glyph_d, descent
6810 the maximum distance from the baseline to the lowest grid coordinate
6811 used to place a glyph outline point, for all the rendered glyphs.
6812 This is a negative value, due to the grid's orientation, with the Y axis
6816 maximum glyph height, that is the maximum height for all the glyphs
6817 contained in the rendered text, it is equivalent to @var{ascent} -
6821 maximum glyph width, that is the maximum width for all the glyphs
6822 contained in the rendered text
6825 the number of input frame, starting from 0
6827 @item rand(min, max)
6828 return a random number included between @var{min} and @var{max}
6831 The input sample aspect ratio.
6834 timestamp expressed in seconds, NAN if the input timestamp is unknown
6837 the height of the rendered text
6840 the width of the rendered text
6844 the x and y offset coordinates where the text is drawn.
6846 These parameters allow the @var{x} and @var{y} expressions to refer
6847 each other, so you can for example specify @code{y=x/dar}.
6850 @anchor{drawtext_expansion}
6851 @subsection Text expansion
6853 If @option{expansion} is set to @code{strftime},
6854 the filter recognizes strftime() sequences in the provided text and
6855 expands them accordingly. Check the documentation of strftime(). This
6856 feature is deprecated.
6858 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6860 If @option{expansion} is set to @code{normal} (which is the default),
6861 the following expansion mechanism is used.
6863 The backslash character @samp{\}, followed by any character, always expands to
6864 the second character.
6866 Sequence of the form @code{%@{...@}} are expanded. The text between the
6867 braces is a function name, possibly followed by arguments separated by ':'.
6868 If the arguments contain special characters or delimiters (':' or '@}'),
6869 they should be escaped.
6871 Note that they probably must also be escaped as the value for the
6872 @option{text} option in the filter argument string and as the filter
6873 argument in the filtergraph description, and possibly also for the shell,
6874 that makes up to four levels of escaping; using a text file avoids these
6877 The following functions are available:
6882 The expression evaluation result.
6884 It must take one argument specifying the expression to be evaluated,
6885 which accepts the same constants and functions as the @var{x} and
6886 @var{y} values. Note that not all constants should be used, for
6887 example the text size is not known when evaluating the expression, so
6888 the constants @var{text_w} and @var{text_h} will have an undefined
6891 @item expr_int_format, eif
6892 Evaluate the expression's value and output as formatted integer.
6894 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6895 The second argument specifies the output format. Allowed values are @samp{x},
6896 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6897 @code{printf} function.
6898 The third parameter is optional and sets the number of positions taken by the output.
6899 It can be used to add padding with zeros from the left.
6902 The time at which the filter is running, expressed in UTC.
6903 It can accept an argument: a strftime() format string.
6906 The time at which the filter is running, expressed in the local time zone.
6907 It can accept an argument: a strftime() format string.
6910 Frame metadata. Takes one or two arguments.
6912 The first argument is mandatory and specifies the metadata key.
6914 The second argument is optional and specifies a default value, used when the
6915 metadata key is not found or empty.
6918 The frame number, starting from 0.
6921 A 1 character description of the current picture type.
6924 The timestamp of the current frame.
6925 It can take up to three arguments.
6927 The first argument is the format of the timestamp; it defaults to @code{flt}
6928 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6929 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6930 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
6931 @code{localtime} stands for the timestamp of the frame formatted as
6932 local time zone time.
6934 The second argument is an offset added to the timestamp.
6936 If the format is set to @code{localtime} or @code{gmtime},
6937 a third argument may be supplied: a strftime() format string.
6938 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
6941 @subsection Examples
6945 Draw "Test Text" with font FreeSerif, using the default values for the
6946 optional parameters.
6949 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
6953 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
6954 and y=50 (counting from the top-left corner of the screen), text is
6955 yellow with a red box around it. Both the text and the box have an
6959 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
6960 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
6963 Note that the double quotes are not necessary if spaces are not used
6964 within the parameter list.
6967 Show the text at the center of the video frame:
6969 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
6973 Show the text at a random position, switching to a new position every 30 seconds:
6975 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)"
6979 Show a text line sliding from right to left in the last row of the video
6980 frame. The file @file{LONG_LINE} is assumed to contain a single line
6983 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
6987 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
6989 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
6993 Draw a single green letter "g", at the center of the input video.
6994 The glyph baseline is placed at half screen height.
6996 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7000 Show text for 1 second every 3 seconds:
7002 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7006 Use fontconfig to set the font. Note that the colons need to be escaped.
7008 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7012 Print the date of a real-time encoding (see strftime(3)):
7014 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7018 Show text fading in and out (appearing/disappearing):
7021 DS=1.0 # display start
7022 DE=10.0 # display end
7023 FID=1.5 # fade in duration
7024 FOD=5 # fade out duration
7025 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 @}"
7030 For more information about libfreetype, check:
7031 @url{http://www.freetype.org/}.
7033 For more information about fontconfig, check:
7034 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7036 For more information about libfribidi, check:
7037 @url{http://fribidi.org/}.
7041 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7043 The filter accepts the following options:
7048 Set low and high threshold values used by the Canny thresholding
7051 The high threshold selects the "strong" edge pixels, which are then
7052 connected through 8-connectivity with the "weak" edge pixels selected
7053 by the low threshold.
7055 @var{low} and @var{high} threshold values must be chosen in the range
7056 [0,1], and @var{low} should be lesser or equal to @var{high}.
7058 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7062 Define the drawing mode.
7066 Draw white/gray wires on black background.
7069 Mix the colors to create a paint/cartoon effect.
7072 Default value is @var{wires}.
7075 @subsection Examples
7079 Standard edge detection with custom values for the hysteresis thresholding:
7081 edgedetect=low=0.1:high=0.4
7085 Painting effect without thresholding:
7087 edgedetect=mode=colormix:high=0
7092 Set brightness, contrast, saturation and approximate gamma adjustment.
7094 The filter accepts the following options:
7098 Set the contrast expression. The value must be a float value in range
7099 @code{-2.0} to @code{2.0}. The default value is "1".
7102 Set the brightness expression. The value must be a float value in
7103 range @code{-1.0} to @code{1.0}. The default value is "0".
7106 Set the saturation expression. The value must be a float in
7107 range @code{0.0} to @code{3.0}. The default value is "1".
7110 Set the gamma expression. The value must be a float in range
7111 @code{0.1} to @code{10.0}. The default value is "1".
7114 Set the gamma expression for red. The value must be a float in
7115 range @code{0.1} to @code{10.0}. The default value is "1".
7118 Set the gamma expression for green. The value must be a float in range
7119 @code{0.1} to @code{10.0}. The default value is "1".
7122 Set the gamma expression for blue. The value must be a float in range
7123 @code{0.1} to @code{10.0}. The default value is "1".
7126 Set the gamma weight expression. It can be used to reduce the effect
7127 of a high gamma value on bright image areas, e.g. keep them from
7128 getting overamplified and just plain white. The value must be a float
7129 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7130 gamma correction all the way down while @code{1.0} leaves it at its
7131 full strength. Default is "1".
7134 Set when the expressions for brightness, contrast, saturation and
7135 gamma expressions are evaluated.
7137 It accepts the following values:
7140 only evaluate expressions once during the filter initialization or
7141 when a command is processed
7144 evaluate expressions for each incoming frame
7147 Default value is @samp{init}.
7150 The expressions accept the following parameters:
7153 frame count of the input frame starting from 0
7156 byte position of the corresponding packet in the input file, NAN if
7160 frame rate of the input video, NAN if the input frame rate is unknown
7163 timestamp expressed in seconds, NAN if the input timestamp is unknown
7166 @subsection Commands
7167 The filter supports the following commands:
7171 Set the contrast expression.
7174 Set the brightness expression.
7177 Set the saturation expression.
7180 Set the gamma expression.
7183 Set the gamma_r expression.
7186 Set gamma_g expression.
7189 Set gamma_b expression.
7192 Set gamma_weight expression.
7194 The command accepts the same syntax of the corresponding option.
7196 If the specified expression is not valid, it is kept at its current
7203 Apply erosion effect to the video.
7205 This filter replaces the pixel by the local(3x3) minimum.
7207 It accepts the following options:
7214 Limit the maximum change for each plane, default is 65535.
7215 If 0, plane will remain unchanged.
7218 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7221 Flags to local 3x3 coordinates maps like this:
7228 @section extractplanes
7230 Extract color channel components from input video stream into
7231 separate grayscale video streams.
7233 The filter accepts the following option:
7237 Set plane(s) to extract.
7239 Available values for planes are:
7250 Choosing planes not available in the input will result in an error.
7251 That means you cannot select @code{r}, @code{g}, @code{b} planes
7252 with @code{y}, @code{u}, @code{v} planes at same time.
7255 @subsection Examples
7259 Extract luma, u and v color channel component from input video frame
7260 into 3 grayscale outputs:
7262 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
7268 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7270 For each input image, the filter will compute the optimal mapping from
7271 the input to the output given the codebook length, that is the number
7272 of distinct output colors.
7274 This filter accepts the following options.
7277 @item codebook_length, l
7278 Set codebook length. The value must be a positive integer, and
7279 represents the number of distinct output colors. Default value is 256.
7282 Set the maximum number of iterations to apply for computing the optimal
7283 mapping. The higher the value the better the result and the higher the
7284 computation time. Default value is 1.
7287 Set a random seed, must be an integer included between 0 and
7288 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7289 will try to use a good random seed on a best effort basis.
7292 Set pal8 output pixel format. This option does not work with codebook
7293 length greater than 256.
7298 Apply a fade-in/out effect to the input video.
7300 It accepts the following parameters:
7304 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7306 Default is @code{in}.
7308 @item start_frame, s
7309 Specify the number of the frame to start applying the fade
7310 effect at. Default is 0.
7313 The number of frames that the fade effect lasts. At the end of the
7314 fade-in effect, the output video will have the same intensity as the input video.
7315 At the end of the fade-out transition, the output video will be filled with the
7316 selected @option{color}.
7320 If set to 1, fade only alpha channel, if one exists on the input.
7323 @item start_time, st
7324 Specify the timestamp (in seconds) of the frame to start to apply the fade
7325 effect. If both start_frame and start_time are specified, the fade will start at
7326 whichever comes last. Default is 0.
7329 The number of seconds for which the fade effect has to last. At the end of the
7330 fade-in effect the output video will have the same intensity as the input video,
7331 at the end of the fade-out transition the output video will be filled with the
7332 selected @option{color}.
7333 If both duration and nb_frames are specified, duration is used. Default is 0
7334 (nb_frames is used by default).
7337 Specify the color of the fade. Default is "black".
7340 @subsection Examples
7344 Fade in the first 30 frames of video:
7349 The command above is equivalent to:
7355 Fade out the last 45 frames of a 200-frame video:
7358 fade=type=out:start_frame=155:nb_frames=45
7362 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7364 fade=in:0:25, fade=out:975:25
7368 Make the first 5 frames yellow, then fade in from frame 5-24:
7370 fade=in:5:20:color=yellow
7374 Fade in alpha over first 25 frames of video:
7376 fade=in:0:25:alpha=1
7380 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7382 fade=t=in:st=5.5:d=0.5
7388 Apply arbitrary expressions to samples in frequency domain
7392 Adjust the dc value (gain) of the luma plane of the image. The filter
7393 accepts an integer value in range @code{0} to @code{1000}. The default
7394 value is set to @code{0}.
7397 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7398 filter accepts an integer value in range @code{0} to @code{1000}. The
7399 default value is set to @code{0}.
7402 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7403 filter accepts an integer value in range @code{0} to @code{1000}. The
7404 default value is set to @code{0}.
7407 Set the frequency domain weight expression for the luma plane.
7410 Set the frequency domain weight expression for the 1st chroma plane.
7413 Set the frequency domain weight expression for the 2nd chroma plane.
7415 The filter accepts the following variables:
7418 The coordinates of the current sample.
7422 The width and height of the image.
7425 @subsection Examples
7431 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7437 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7443 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7449 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7456 Extract a single field from an interlaced image using stride
7457 arithmetic to avoid wasting CPU time. The output frames are marked as
7460 The filter accepts the following options:
7464 Specify whether to extract the top (if the value is @code{0} or
7465 @code{top}) or the bottom field (if the value is @code{1} or
7471 Create new frames by copying the top and bottom fields from surrounding frames
7472 supplied as numbers by the hint file.
7476 Set file containing hints: absolute/relative frame numbers.
7478 There must be one line for each frame in a clip. Each line must contain two
7479 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7480 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7481 is current frame number for @code{absolute} mode or out of [-1, 1] range
7482 for @code{relative} mode. First number tells from which frame to pick up top
7483 field and second number tells from which frame to pick up bottom field.
7485 If optionally followed by @code{+} output frame will be marked as interlaced,
7486 else if followed by @code{-} output frame will be marked as progressive, else
7487 it will be marked same as input frame.
7488 If line starts with @code{#} or @code{;} that line is skipped.
7491 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7494 Example of first several lines of @code{hint} file for @code{relative} mode:
7497 1,0 - # second frame, use third's frame top field and second's frame bottom field
7498 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7515 Field matching filter for inverse telecine. It is meant to reconstruct the
7516 progressive frames from a telecined stream. The filter does not drop duplicated
7517 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7518 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7520 The separation of the field matching and the decimation is notably motivated by
7521 the possibility of inserting a de-interlacing filter fallback between the two.
7522 If the source has mixed telecined and real interlaced content,
7523 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7524 But these remaining combed frames will be marked as interlaced, and thus can be
7525 de-interlaced by a later filter such as @ref{yadif} before decimation.
7527 In addition to the various configuration options, @code{fieldmatch} can take an
7528 optional second stream, activated through the @option{ppsrc} option. If
7529 enabled, the frames reconstruction will be based on the fields and frames from
7530 this second stream. This allows the first input to be pre-processed in order to
7531 help the various algorithms of the filter, while keeping the output lossless
7532 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7533 or brightness/contrast adjustments can help.
7535 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7536 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7537 which @code{fieldmatch} is based on. While the semantic and usage are very
7538 close, some behaviour and options names can differ.
7540 The @ref{decimate} filter currently only works for constant frame rate input.
7541 If your input has mixed telecined (30fps) and progressive content with a lower
7542 framerate like 24fps use the following filterchain to produce the necessary cfr
7543 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7545 The filter accepts the following options:
7549 Specify the assumed field order of the input stream. Available values are:
7553 Auto detect parity (use FFmpeg's internal parity value).
7555 Assume bottom field first.
7557 Assume top field first.
7560 Note that it is sometimes recommended not to trust the parity announced by the
7563 Default value is @var{auto}.
7566 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7567 sense that it won't risk creating jerkiness due to duplicate frames when
7568 possible, but if there are bad edits or blended fields it will end up
7569 outputting combed frames when a good match might actually exist. On the other
7570 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7571 but will almost always find a good frame if there is one. The other values are
7572 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7573 jerkiness and creating duplicate frames versus finding good matches in sections
7574 with bad edits, orphaned fields, blended fields, etc.
7576 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7578 Available values are:
7582 2-way matching (p/c)
7584 2-way matching, and trying 3rd match if still combed (p/c + n)
7586 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7588 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7589 still combed (p/c + n + u/b)
7591 3-way matching (p/c/n)
7593 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7594 detected as combed (p/c/n + u/b)
7597 The parenthesis at the end indicate the matches that would be used for that
7598 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7601 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7604 Default value is @var{pc_n}.
7607 Mark the main input stream as a pre-processed input, and enable the secondary
7608 input stream as the clean source to pick the fields from. See the filter
7609 introduction for more details. It is similar to the @option{clip2} feature from
7612 Default value is @code{0} (disabled).
7615 Set the field to match from. It is recommended to set this to the same value as
7616 @option{order} unless you experience matching failures with that setting. In
7617 certain circumstances changing the field that is used to match from can have a
7618 large impact on matching performance. Available values are:
7622 Automatic (same value as @option{order}).
7624 Match from the bottom field.
7626 Match from the top field.
7629 Default value is @var{auto}.
7632 Set whether or not chroma is included during the match comparisons. In most
7633 cases it is recommended to leave this enabled. You should set this to @code{0}
7634 only if your clip has bad chroma problems such as heavy rainbowing or other
7635 artifacts. Setting this to @code{0} could also be used to speed things up at
7636 the cost of some accuracy.
7638 Default value is @code{1}.
7642 These define an exclusion band which excludes the lines between @option{y0} and
7643 @option{y1} from being included in the field matching decision. An exclusion
7644 band can be used to ignore subtitles, a logo, or other things that may
7645 interfere with the matching. @option{y0} sets the starting scan line and
7646 @option{y1} sets the ending line; all lines in between @option{y0} and
7647 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7648 @option{y0} and @option{y1} to the same value will disable the feature.
7649 @option{y0} and @option{y1} defaults to @code{0}.
7652 Set the scene change detection threshold as a percentage of maximum change on
7653 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7654 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7655 @option{scthresh} is @code{[0.0, 100.0]}.
7657 Default value is @code{12.0}.
7660 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7661 account the combed scores of matches when deciding what match to use as the
7662 final match. Available values are:
7666 No final matching based on combed scores.
7668 Combed scores are only used when a scene change is detected.
7670 Use combed scores all the time.
7673 Default is @var{sc}.
7676 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7677 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7678 Available values are:
7682 No forced calculation.
7684 Force p/c/n calculations.
7686 Force p/c/n/u/b calculations.
7689 Default value is @var{none}.
7692 This is the area combing threshold used for combed frame detection. This
7693 essentially controls how "strong" or "visible" combing must be to be detected.
7694 Larger values mean combing must be more visible and smaller values mean combing
7695 can be less visible or strong and still be detected. Valid settings are from
7696 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7697 be detected as combed). This is basically a pixel difference value. A good
7698 range is @code{[8, 12]}.
7700 Default value is @code{9}.
7703 Sets whether or not chroma is considered in the combed frame decision. Only
7704 disable this if your source has chroma problems (rainbowing, etc.) that are
7705 causing problems for the combed frame detection with chroma enabled. Actually,
7706 using @option{chroma}=@var{0} is usually more reliable, except for the case
7707 where there is chroma only combing in the source.
7709 Default value is @code{0}.
7713 Respectively set the x-axis and y-axis size of the window used during combed
7714 frame detection. This has to do with the size of the area in which
7715 @option{combpel} pixels are required to be detected as combed for a frame to be
7716 declared combed. See the @option{combpel} parameter description for more info.
7717 Possible values are any number that is a power of 2 starting at 4 and going up
7720 Default value is @code{16}.
7723 The number of combed pixels inside any of the @option{blocky} by
7724 @option{blockx} size blocks on the frame for the frame to be detected as
7725 combed. While @option{cthresh} controls how "visible" the combing must be, this
7726 setting controls "how much" combing there must be in any localized area (a
7727 window defined by the @option{blockx} and @option{blocky} settings) on the
7728 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7729 which point no frames will ever be detected as combed). This setting is known
7730 as @option{MI} in TFM/VFM vocabulary.
7732 Default value is @code{80}.
7735 @anchor{p/c/n/u/b meaning}
7736 @subsection p/c/n/u/b meaning
7738 @subsubsection p/c/n
7740 We assume the following telecined stream:
7743 Top fields: 1 2 2 3 4
7744 Bottom fields: 1 2 3 4 4
7747 The numbers correspond to the progressive frame the fields relate to. Here, the
7748 first two frames are progressive, the 3rd and 4th are combed, and so on.
7750 When @code{fieldmatch} is configured to run a matching from bottom
7751 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7756 B 1 2 3 4 4 <-- matching reference
7765 As a result of the field matching, we can see that some frames get duplicated.
7766 To perform a complete inverse telecine, you need to rely on a decimation filter
7767 after this operation. See for instance the @ref{decimate} filter.
7769 The same operation now matching from top fields (@option{field}=@var{top})
7774 T 1 2 2 3 4 <-- matching reference
7784 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7785 basically, they refer to the frame and field of the opposite parity:
7788 @item @var{p} matches the field of the opposite parity in the previous frame
7789 @item @var{c} matches the field of the opposite parity in the current frame
7790 @item @var{n} matches the field of the opposite parity in the next frame
7795 The @var{u} and @var{b} matching are a bit special in the sense that they match
7796 from the opposite parity flag. In the following examples, we assume that we are
7797 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7798 'x' is placed above and below each matched fields.
7800 With bottom matching (@option{field}=@var{bottom}):
7805 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7806 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7814 With top matching (@option{field}=@var{top}):
7819 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7820 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7828 @subsection Examples
7830 Simple IVTC of a top field first telecined stream:
7832 fieldmatch=order=tff:combmatch=none, decimate
7835 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7837 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7842 Transform the field order of the input video.
7844 It accepts the following parameters:
7849 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7850 for bottom field first.
7853 The default value is @samp{tff}.
7855 The transformation is done by shifting the picture content up or down
7856 by one line, and filling the remaining line with appropriate picture content.
7857 This method is consistent with most broadcast field order converters.
7859 If the input video is not flagged as being interlaced, or it is already
7860 flagged as being of the required output field order, then this filter does
7861 not alter the incoming video.
7863 It is very useful when converting to or from PAL DV material,
7864 which is bottom field first.
7868 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7871 @section fifo, afifo
7873 Buffer input images and send them when they are requested.
7875 It is mainly useful when auto-inserted by the libavfilter
7878 It does not take parameters.
7882 Find a rectangular object
7884 It accepts the following options:
7888 Filepath of the object image, needs to be in gray8.
7891 Detection threshold, default is 0.5.
7894 Number of mipmaps, default is 3.
7896 @item xmin, ymin, xmax, ymax
7897 Specifies the rectangle in which to search.
7900 @subsection Examples
7904 Generate a representative palette of a given video using @command{ffmpeg}:
7906 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7912 Cover a rectangular object
7914 It accepts the following options:
7918 Filepath of the optional cover image, needs to be in yuv420.
7923 It accepts the following values:
7926 cover it by the supplied image
7928 cover it by interpolating the surrounding pixels
7931 Default value is @var{blur}.
7934 @subsection Examples
7938 Generate a representative palette of a given video using @command{ffmpeg}:
7940 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7947 Convert the input video to one of the specified pixel formats.
7948 Libavfilter will try to pick one that is suitable as input to
7951 It accepts the following parameters:
7955 A '|'-separated list of pixel format names, such as
7956 "pix_fmts=yuv420p|monow|rgb24".
7960 @subsection Examples
7964 Convert the input video to the @var{yuv420p} format
7966 format=pix_fmts=yuv420p
7969 Convert the input video to any of the formats in the list
7971 format=pix_fmts=yuv420p|yuv444p|yuv410p
7978 Convert the video to specified constant frame rate by duplicating or dropping
7979 frames as necessary.
7981 It accepts the following parameters:
7985 The desired output frame rate. The default is @code{25}.
7990 Possible values are:
7993 zero round towards 0
7997 round towards -infinity
7999 round towards +infinity
8003 The default is @code{near}.
8006 Assume the first PTS should be the given value, in seconds. This allows for
8007 padding/trimming at the start of stream. By default, no assumption is made
8008 about the first frame's expected PTS, so no padding or trimming is done.
8009 For example, this could be set to 0 to pad the beginning with duplicates of
8010 the first frame if a video stream starts after the audio stream or to trim any
8011 frames with a negative PTS.
8015 Alternatively, the options can be specified as a flat string:
8016 @var{fps}[:@var{round}].
8018 See also the @ref{setpts} filter.
8020 @subsection Examples
8024 A typical usage in order to set the fps to 25:
8030 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8032 fps=fps=film:round=near
8038 Pack two different video streams into a stereoscopic video, setting proper
8039 metadata on supported codecs. The two views should have the same size and
8040 framerate and processing will stop when the shorter video ends. Please note
8041 that you may conveniently adjust view properties with the @ref{scale} and
8044 It accepts the following parameters:
8048 The desired packing format. Supported values are:
8053 The views are next to each other (default).
8056 The views are on top of each other.
8059 The views are packed by line.
8062 The views are packed by column.
8065 The views are temporally interleaved.
8074 # Convert left and right views into a frame-sequential video
8075 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8077 # Convert views into a side-by-side video with the same output resolution as the input
8078 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
8083 Change the frame rate by interpolating new video output frames from the source
8086 This filter is not designed to function correctly with interlaced media. If
8087 you wish to change the frame rate of interlaced media then you are required
8088 to deinterlace before this filter and re-interlace after this filter.
8090 A description of the accepted options follows.
8094 Specify the output frames per second. This option can also be specified
8095 as a value alone. The default is @code{50}.
8098 Specify the start of a range where the output frame will be created as a
8099 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8100 the default is @code{15}.
8103 Specify the end of a range where the output frame will be created as a
8104 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8105 the default is @code{240}.
8108 Specify the level at which a scene change is detected as a value between
8109 0 and 100 to indicate a new scene; a low value reflects a low
8110 probability for the current frame to introduce a new scene, while a higher
8111 value means the current frame is more likely to be one.
8112 The default is @code{7}.
8115 Specify flags influencing the filter process.
8117 Available value for @var{flags} is:
8120 @item scene_change_detect, scd
8121 Enable scene change detection using the value of the option @var{scene}.
8122 This flag is enabled by default.
8128 Select one frame every N-th frame.
8130 This filter accepts the following option:
8133 Select frame after every @code{step} frames.
8134 Allowed values are positive integers higher than 0. Default value is @code{1}.
8140 Apply a frei0r effect to the input video.
8142 To enable the compilation of this filter, you need to install the frei0r
8143 header and configure FFmpeg with @code{--enable-frei0r}.
8145 It accepts the following parameters:
8150 The name of the frei0r effect to load. If the environment variable
8151 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8152 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8153 Otherwise, the standard frei0r paths are searched, in this order:
8154 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8155 @file{/usr/lib/frei0r-1/}.
8158 A '|'-separated list of parameters to pass to the frei0r effect.
8162 A frei0r effect parameter can be a boolean (its value is either
8163 "y" or "n"), a double, a color (specified as
8164 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8165 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8166 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8167 @var{X} and @var{Y} are floating point numbers) and/or a string.
8169 The number and types of parameters depend on the loaded effect. If an
8170 effect parameter is not specified, the default value is set.
8172 @subsection Examples
8176 Apply the distort0r effect, setting the first two double parameters:
8178 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8182 Apply the colordistance effect, taking a color as the first parameter:
8184 frei0r=colordistance:0.2/0.3/0.4
8185 frei0r=colordistance:violet
8186 frei0r=colordistance:0x112233
8190 Apply the perspective effect, specifying the top left and top right image
8193 frei0r=perspective:0.2/0.2|0.8/0.2
8197 For more information, see
8198 @url{http://frei0r.dyne.org}
8202 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8204 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8205 processing filter, one of them is performed once per block, not per pixel.
8206 This allows for much higher speed.
8208 The filter accepts the following options:
8212 Set quality. This option defines the number of levels for averaging. It accepts
8213 an integer in the range 4-5. Default value is @code{4}.
8216 Force a constant quantization parameter. It accepts an integer in range 0-63.
8217 If not set, the filter will use the QP from the video stream (if available).
8220 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8221 more details but also more artifacts, while higher values make the image smoother
8222 but also blurrier. Default value is @code{0} − PSNR optimal.
8225 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8226 option may cause flicker since the B-Frames have often larger QP. Default is
8227 @code{0} (not enabled).
8233 The filter accepts the following options:
8237 Set the luminance expression.
8239 Set the chrominance blue expression.
8241 Set the chrominance red expression.
8243 Set the alpha expression.
8245 Set the red expression.
8247 Set the green expression.
8249 Set the blue expression.
8252 The colorspace is selected according to the specified options. If one
8253 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8254 options is specified, the filter will automatically select a YCbCr
8255 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8256 @option{blue_expr} options is specified, it will select an RGB
8259 If one of the chrominance expression is not defined, it falls back on the other
8260 one. If no alpha expression is specified it will evaluate to opaque value.
8261 If none of chrominance expressions are specified, they will evaluate
8262 to the luminance expression.
8264 The expressions can use the following variables and functions:
8268 The sequential number of the filtered frame, starting from @code{0}.
8272 The coordinates of the current sample.
8276 The width and height of the image.
8280 Width and height scale depending on the currently filtered plane. It is the
8281 ratio between the corresponding luma plane number of pixels and the current
8282 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8283 @code{0.5,0.5} for chroma planes.
8286 Time of the current frame, expressed in seconds.
8289 Return the value of the pixel at location (@var{x},@var{y}) of the current
8293 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8297 Return the value of the pixel at location (@var{x},@var{y}) of the
8298 blue-difference chroma plane. Return 0 if there is no such plane.
8301 Return the value of the pixel at location (@var{x},@var{y}) of the
8302 red-difference chroma plane. Return 0 if there is no such plane.
8307 Return the value of the pixel at location (@var{x},@var{y}) of the
8308 red/green/blue component. Return 0 if there is no such component.
8311 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8312 plane. Return 0 if there is no such plane.
8315 For functions, if @var{x} and @var{y} are outside the area, the value will be
8316 automatically clipped to the closer edge.
8318 @subsection Examples
8322 Flip the image horizontally:
8328 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8329 wavelength of 100 pixels:
8331 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8335 Generate a fancy enigmatic moving light:
8337 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
8341 Generate a quick emboss effect:
8343 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8347 Modify RGB components depending on pixel position:
8349 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8353 Create a radial gradient that is the same size as the input (also see
8354 the @ref{vignette} filter):
8356 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8362 Fix the banding artifacts that are sometimes introduced into nearly flat
8363 regions by truncation to 8-bit color depth.
8364 Interpolate the gradients that should go where the bands are, and
8367 It is designed for playback only. Do not use it prior to
8368 lossy compression, because compression tends to lose the dither and
8369 bring back the bands.
8371 It accepts the following parameters:
8376 The maximum amount by which the filter will change any one pixel. This is also
8377 the threshold for detecting nearly flat regions. Acceptable values range from
8378 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8382 The neighborhood to fit the gradient to. A larger radius makes for smoother
8383 gradients, but also prevents the filter from modifying the pixels near detailed
8384 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8385 values will be clipped to the valid range.
8389 Alternatively, the options can be specified as a flat string:
8390 @var{strength}[:@var{radius}]
8392 @subsection Examples
8396 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8402 Specify radius, omitting the strength (which will fall-back to the default
8413 Apply a Hald CLUT to a video stream.
8415 First input is the video stream to process, and second one is the Hald CLUT.
8416 The Hald CLUT input can be a simple picture or a complete video stream.
8418 The filter accepts the following options:
8422 Force termination when the shortest input terminates. Default is @code{0}.
8424 Continue applying the last CLUT after the end of the stream. A value of
8425 @code{0} disable the filter after the last frame of the CLUT is reached.
8426 Default is @code{1}.
8429 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8430 filters share the same internals).
8432 More information about the Hald CLUT can be found on Eskil Steenberg's website
8433 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8435 @subsection Workflow examples
8437 @subsubsection Hald CLUT video stream
8439 Generate an identity Hald CLUT stream altered with various effects:
8441 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
8444 Note: make sure you use a lossless codec.
8446 Then use it with @code{haldclut} to apply it on some random stream:
8448 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8451 The Hald CLUT will be applied to the 10 first seconds (duration of
8452 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8453 to the remaining frames of the @code{mandelbrot} stream.
8455 @subsubsection Hald CLUT with preview
8457 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8458 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8459 biggest possible square starting at the top left of the picture. The remaining
8460 padding pixels (bottom or right) will be ignored. This area can be used to add
8461 a preview of the Hald CLUT.
8463 Typically, the following generated Hald CLUT will be supported by the
8464 @code{haldclut} filter:
8467 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8468 pad=iw+320 [padded_clut];
8469 smptebars=s=320x256, split [a][b];
8470 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8471 [main][b] overlay=W-320" -frames:v 1 clut.png
8474 It contains the original and a preview of the effect of the CLUT: SMPTE color
8475 bars are displayed on the right-top, and below the same color bars processed by
8478 Then, the effect of this Hald CLUT can be visualized with:
8480 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8485 Flip the input video horizontally.
8487 For example, to horizontally flip the input video with @command{ffmpeg}:
8489 ffmpeg -i in.avi -vf "hflip" out.avi
8493 This filter applies a global color histogram equalization on a
8496 It can be used to correct video that has a compressed range of pixel
8497 intensities. The filter redistributes the pixel intensities to
8498 equalize their distribution across the intensity range. It may be
8499 viewed as an "automatically adjusting contrast filter". This filter is
8500 useful only for correcting degraded or poorly captured source
8503 The filter accepts the following options:
8507 Determine the amount of equalization to be applied. As the strength
8508 is reduced, the distribution of pixel intensities more-and-more
8509 approaches that of the input frame. The value must be a float number
8510 in the range [0,1] and defaults to 0.200.
8513 Set the maximum intensity that can generated and scale the output
8514 values appropriately. The strength should be set as desired and then
8515 the intensity can be limited if needed to avoid washing-out. The value
8516 must be a float number in the range [0,1] and defaults to 0.210.
8519 Set the antibanding level. If enabled the filter will randomly vary
8520 the luminance of output pixels by a small amount to avoid banding of
8521 the histogram. Possible values are @code{none}, @code{weak} or
8522 @code{strong}. It defaults to @code{none}.
8527 Compute and draw a color distribution histogram for the input video.
8529 The computed histogram is a representation of the color component
8530 distribution in an image.
8532 Standard histogram displays the color components distribution in an image.
8533 Displays color graph for each color component. Shows distribution of
8534 the Y, U, V, A or R, G, B components, depending on input format, in the
8535 current frame. Below each graph a color component scale meter is shown.
8537 The filter accepts the following options:
8541 Set height of level. Default value is @code{200}.
8542 Allowed range is [50, 2048].
8545 Set height of color scale. Default value is @code{12}.
8546 Allowed range is [0, 40].
8550 It accepts the following values:
8553 Per color component graphs are placed below each other.
8556 Presents information identical to that in the @code{parade}, except
8557 that the graphs representing color components are superimposed directly
8560 Default is @code{parade}.
8563 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8564 Default is @code{linear}.
8567 Set what color components to display.
8568 Default is @code{7}.
8571 Set foreground opacity. Default is @code{0.7}.
8574 Set background opacity. Default is @code{0.5}.
8577 @subsection Examples
8582 Calculate and draw histogram:
8584 ffplay -i input -vf histogram
8592 This is a high precision/quality 3d denoise filter. It aims to reduce
8593 image noise, producing smooth images and making still images really
8594 still. It should enhance compressibility.
8596 It accepts the following optional parameters:
8600 A non-negative floating point number which specifies spatial luma strength.
8603 @item chroma_spatial
8604 A non-negative floating point number which specifies spatial chroma strength.
8605 It defaults to 3.0*@var{luma_spatial}/4.0.
8608 A floating point number which specifies luma temporal strength. It defaults to
8609 6.0*@var{luma_spatial}/4.0.
8612 A floating point number which specifies chroma temporal strength. It defaults to
8613 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8616 @anchor{hwupload_cuda}
8617 @section hwupload_cuda
8619 Upload system memory frames to a CUDA device.
8621 It accepts the following optional parameters:
8625 The number of the CUDA device to use
8630 Apply a high-quality magnification filter designed for pixel art. This filter
8631 was originally created by Maxim Stepin.
8633 It accepts the following option:
8637 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8638 @code{hq3x} and @code{4} for @code{hq4x}.
8639 Default is @code{3}.
8643 Stack input videos horizontally.
8645 All streams must be of same pixel format and of same height.
8647 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8648 to create same output.
8650 The filter accept the following option:
8654 Set number of input streams. Default is 2.
8657 If set to 1, force the output to terminate when the shortest input
8658 terminates. Default value is 0.
8663 Modify the hue and/or the saturation of the input.
8665 It accepts the following parameters:
8669 Specify the hue angle as a number of degrees. It accepts an expression,
8670 and defaults to "0".
8673 Specify the saturation in the [-10,10] range. It accepts an expression and
8677 Specify the hue angle as a number of radians. It accepts an
8678 expression, and defaults to "0".
8681 Specify the brightness in the [-10,10] range. It accepts an expression and
8685 @option{h} and @option{H} are mutually exclusive, and can't be
8686 specified at the same time.
8688 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8689 expressions containing the following constants:
8693 frame count of the input frame starting from 0
8696 presentation timestamp of the input frame expressed in time base units
8699 frame rate of the input video, NAN if the input frame rate is unknown
8702 timestamp expressed in seconds, NAN if the input timestamp is unknown
8705 time base of the input video
8708 @subsection Examples
8712 Set the hue to 90 degrees and the saturation to 1.0:
8718 Same command but expressing the hue in radians:
8724 Rotate hue and make the saturation swing between 0
8725 and 2 over a period of 1 second:
8727 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8731 Apply a 3 seconds saturation fade-in effect starting at 0:
8736 The general fade-in expression can be written as:
8738 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8742 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8744 hue="s=max(0\, min(1\, (8-t)/3))"
8747 The general fade-out expression can be written as:
8749 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8754 @subsection Commands
8756 This filter supports the following commands:
8762 Modify the hue and/or the saturation and/or brightness of the input video.
8763 The command accepts the same syntax of the corresponding option.
8765 If the specified expression is not valid, it is kept at its current
8771 Grow first stream into second stream by connecting components.
8772 This allows to build more robust edge masks.
8774 This filter accepts the following options:
8778 Set which planes will be processed as bitmap, unprocessed planes will be
8779 copied from first stream.
8780 By default value 0xf, all planes will be processed.
8783 Set threshold which is used in filtering. If pixel component value is higher than
8784 this value filter algorithm for connecting components is activated.
8785 By default value is 0.
8790 Detect video interlacing type.
8792 This filter tries to detect if the input frames as interlaced, progressive,
8793 top or bottom field first. It will also try and detect fields that are
8794 repeated between adjacent frames (a sign of telecine).
8796 Single frame detection considers only immediately adjacent frames when classifying each frame.
8797 Multiple frame detection incorporates the classification history of previous frames.
8799 The filter will log these metadata values:
8802 @item single.current_frame
8803 Detected type of current frame using single-frame detection. One of:
8804 ``tff'' (top field first), ``bff'' (bottom field first),
8805 ``progressive'', or ``undetermined''
8808 Cumulative number of frames detected as top field first using single-frame detection.
8811 Cumulative number of frames detected as top field first using multiple-frame detection.
8814 Cumulative number of frames detected as bottom field first using single-frame detection.
8816 @item multiple.current_frame
8817 Detected type of current frame using multiple-frame detection. One of:
8818 ``tff'' (top field first), ``bff'' (bottom field first),
8819 ``progressive'', or ``undetermined''
8822 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8824 @item single.progressive
8825 Cumulative number of frames detected as progressive using single-frame detection.
8827 @item multiple.progressive
8828 Cumulative number of frames detected as progressive using multiple-frame detection.
8830 @item single.undetermined
8831 Cumulative number of frames that could not be classified using single-frame detection.
8833 @item multiple.undetermined
8834 Cumulative number of frames that could not be classified using multiple-frame detection.
8836 @item repeated.current_frame
8837 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8839 @item repeated.neither
8840 Cumulative number of frames with no repeated field.
8843 Cumulative number of frames with the top field repeated from the previous frame's top field.
8845 @item repeated.bottom
8846 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8849 The filter accepts the following options:
8853 Set interlacing threshold.
8855 Set progressive threshold.
8857 Threshold for repeated field detection.
8859 Number of frames after which a given frame's contribution to the
8860 statistics is halved (i.e., it contributes only 0.5 to it's
8861 classification). The default of 0 means that all frames seen are given
8862 full weight of 1.0 forever.
8863 @item analyze_interlaced_flag
8864 When this is not 0 then idet will use the specified number of frames to determine
8865 if the interlaced flag is accurate, it will not count undetermined frames.
8866 If the flag is found to be accurate it will be used without any further
8867 computations, if it is found to be inaccurate it will be cleared without any
8868 further computations. This allows inserting the idet filter as a low computational
8869 method to clean up the interlaced flag
8874 Deinterleave or interleave fields.
8876 This filter allows one to process interlaced images fields without
8877 deinterlacing them. Deinterleaving splits the input frame into 2
8878 fields (so called half pictures). Odd lines are moved to the top
8879 half of the output image, even lines to the bottom half.
8880 You can process (filter) them independently and then re-interleave them.
8882 The filter accepts the following options:
8886 @item chroma_mode, c
8888 Available values for @var{luma_mode}, @var{chroma_mode} and
8889 @var{alpha_mode} are:
8895 @item deinterleave, d
8896 Deinterleave fields, placing one above the other.
8899 Interleave fields. Reverse the effect of deinterleaving.
8901 Default value is @code{none}.
8904 @item chroma_swap, cs
8905 @item alpha_swap, as
8906 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
8911 Apply inflate effect to the video.
8913 This filter replaces the pixel by the local(3x3) average by taking into account
8914 only values higher than the pixel.
8916 It accepts the following options:
8923 Limit the maximum change for each plane, default is 65535.
8924 If 0, plane will remain unchanged.
8929 Simple interlacing filter from progressive contents. This interleaves upper (or
8930 lower) lines from odd frames with lower (or upper) lines from even frames,
8931 halving the frame rate and preserving image height.
8934 Original Original New Frame
8935 Frame 'j' Frame 'j+1' (tff)
8936 ========== =========== ==================
8937 Line 0 --------------------> Frame 'j' Line 0
8938 Line 1 Line 1 ----> Frame 'j+1' Line 1
8939 Line 2 ---------------------> Frame 'j' Line 2
8940 Line 3 Line 3 ----> Frame 'j+1' Line 3
8942 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
8945 It accepts the following optional parameters:
8949 This determines whether the interlaced frame is taken from the even
8950 (tff - default) or odd (bff) lines of the progressive frame.
8953 Enable (default) or disable the vertical lowpass filter to avoid twitter
8954 interlacing and reduce moire patterns.
8959 Deinterlace input video by applying Donald Graft's adaptive kernel
8960 deinterling. Work on interlaced parts of a video to produce
8963 The description of the accepted parameters follows.
8967 Set the threshold which affects the filter's tolerance when
8968 determining if a pixel line must be processed. It must be an integer
8969 in the range [0,255] and defaults to 10. A value of 0 will result in
8970 applying the process on every pixels.
8973 Paint pixels exceeding the threshold value to white if set to 1.
8977 Set the fields order. Swap fields if set to 1, leave fields alone if
8981 Enable additional sharpening if set to 1. Default is 0.
8984 Enable twoway sharpening if set to 1. Default is 0.
8987 @subsection Examples
8991 Apply default values:
8993 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
8997 Enable additional sharpening:
9003 Paint processed pixels in white:
9009 @section lenscorrection
9011 Correct radial lens distortion
9013 This filter can be used to correct for radial distortion as can result from the use
9014 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9015 one can use tools available for example as part of opencv or simply trial-and-error.
9016 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9017 and extract the k1 and k2 coefficients from the resulting matrix.
9019 Note that effectively the same filter is available in the open-source tools Krita and
9020 Digikam from the KDE project.
9022 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9023 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9024 brightness distribution, so you may want to use both filters together in certain
9025 cases, though you will have to take care of ordering, i.e. whether vignetting should
9026 be applied before or after lens correction.
9030 The filter accepts the following options:
9034 Relative x-coordinate of the focal point of the image, and thereby the center of the
9035 distortion. This value has a range [0,1] and is expressed as fractions of the image
9038 Relative y-coordinate of the focal point of the image, and thereby the center of the
9039 distortion. This value has a range [0,1] and is expressed as fractions of the image
9042 Coefficient of the quadratic correction term. 0.5 means no correction.
9044 Coefficient of the double quadratic correction term. 0.5 means no correction.
9047 The formula that generates the correction is:
9049 @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)
9051 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9052 distances from the focal point in the source and target images, respectively.
9058 The filter accepts the following options:
9062 Set the number of loops.
9065 Set maximal size in number of frames.
9068 Set first frame of loop.
9074 Apply a 3D LUT to an input video.
9076 The filter accepts the following options:
9080 Set the 3D LUT file name.
9082 Currently supported formats:
9094 Select interpolation mode.
9096 Available values are:
9100 Use values from the nearest defined point.
9102 Interpolate values using the 8 points defining a cube.
9104 Interpolate values using a tetrahedron.
9108 @section lut, lutrgb, lutyuv
9110 Compute a look-up table for binding each pixel component input value
9111 to an output value, and apply it to the input video.
9113 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9114 to an RGB input video.
9116 These filters accept the following parameters:
9119 set first pixel component expression
9121 set second pixel component expression
9123 set third pixel component expression
9125 set fourth pixel component expression, corresponds to the alpha component
9128 set red component expression
9130 set green component expression
9132 set blue component expression
9134 alpha component expression
9137 set Y/luminance component expression
9139 set U/Cb component expression
9141 set V/Cr component expression
9144 Each of them specifies the expression to use for computing the lookup table for
9145 the corresponding pixel component values.
9147 The exact component associated to each of the @var{c*} options depends on the
9150 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9151 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9153 The expressions can contain the following constants and functions:
9158 The input width and height.
9161 The input value for the pixel component.
9164 The input value, clipped to the @var{minval}-@var{maxval} range.
9167 The maximum value for the pixel component.
9170 The minimum value for the pixel component.
9173 The negated value for the pixel component value, clipped to the
9174 @var{minval}-@var{maxval} range; it corresponds to the expression
9175 "maxval-clipval+minval".
9178 The computed value in @var{val}, clipped to the
9179 @var{minval}-@var{maxval} range.
9181 @item gammaval(gamma)
9182 The computed gamma correction value of the pixel component value,
9183 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9185 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9189 All expressions default to "val".
9191 @subsection Examples
9197 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9198 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9201 The above is the same as:
9203 lutrgb="r=negval:g=negval:b=negval"
9204 lutyuv="y=negval:u=negval:v=negval"
9214 Remove chroma components, turning the video into a graytone image:
9216 lutyuv="u=128:v=128"
9220 Apply a luma burning effect:
9226 Remove green and blue components:
9232 Set a constant alpha channel value on input:
9234 format=rgba,lutrgb=a="maxval-minval/2"
9238 Correct luminance gamma by a factor of 0.5:
9240 lutyuv=y=gammaval(0.5)
9244 Discard least significant bits of luma:
9246 lutyuv=y='bitand(val, 128+64+32)'
9250 Technicolor like effect:
9252 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9256 @section maskedclamp
9258 Clamp the first input stream with the second input and third input stream.
9260 Returns the value of first stream to be between second input
9261 stream - @code{undershoot} and third input stream + @code{overshoot}.
9263 This filter accepts the following options:
9266 Default value is @code{0}.
9269 Default value is @code{0}.
9272 Set which planes will be processed as bitmap, unprocessed planes will be
9273 copied from first stream.
9274 By default value 0xf, all planes will be processed.
9277 @section maskedmerge
9279 Merge the first input stream with the second input stream using per pixel
9280 weights in the third input stream.
9282 A value of 0 in the third stream pixel component means that pixel component
9283 from first stream is returned unchanged, while maximum value (eg. 255 for
9284 8-bit videos) means that pixel component from second stream is returned
9285 unchanged. Intermediate values define the amount of merging between both
9286 input stream's pixel components.
9288 This filter accepts the following options:
9291 Set which planes will be processed as bitmap, unprocessed planes will be
9292 copied from first stream.
9293 By default value 0xf, all planes will be processed.
9298 Apply motion-compensation deinterlacing.
9300 It needs one field per frame as input and must thus be used together
9301 with yadif=1/3 or equivalent.
9303 This filter accepts the following options:
9306 Set the deinterlacing mode.
9308 It accepts one of the following values:
9313 use iterative motion estimation
9315 like @samp{slow}, but use multiple reference frames.
9317 Default value is @samp{fast}.
9320 Set the picture field parity assumed for the input video. It must be
9321 one of the following values:
9325 assume top field first
9327 assume bottom field first
9330 Default value is @samp{bff}.
9333 Set per-block quantization parameter (QP) used by the internal
9336 Higher values should result in a smoother motion vector field but less
9337 optimal individual vectors. Default value is 1.
9340 @section mergeplanes
9342 Merge color channel components from several video streams.
9344 The filter accepts up to 4 input streams, and merge selected input
9345 planes to the output video.
9347 This filter accepts the following options:
9350 Set input to output plane mapping. Default is @code{0}.
9352 The mappings is specified as a bitmap. It should be specified as a
9353 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9354 mapping for the first plane of the output stream. 'A' sets the number of
9355 the input stream to use (from 0 to 3), and 'a' the plane number of the
9356 corresponding input to use (from 0 to 3). The rest of the mappings is
9357 similar, 'Bb' describes the mapping for the output stream second
9358 plane, 'Cc' describes the mapping for the output stream third plane and
9359 'Dd' describes the mapping for the output stream fourth plane.
9362 Set output pixel format. Default is @code{yuva444p}.
9365 @subsection Examples
9369 Merge three gray video streams of same width and height into single video stream:
9371 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9375 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9377 [a0][a1]mergeplanes=0x00010210:yuva444p
9381 Swap Y and A plane in yuva444p stream:
9383 format=yuva444p,mergeplanes=0x03010200:yuva444p
9387 Swap U and V plane in yuv420p stream:
9389 format=yuv420p,mergeplanes=0x000201:yuv420p
9393 Cast a rgb24 clip to yuv444p:
9395 format=rgb24,mergeplanes=0x000102:yuv444p
9401 Drop frames that do not differ greatly from the previous frame in
9402 order to reduce frame rate.
9404 The main use of this filter is for very-low-bitrate encoding
9405 (e.g. streaming over dialup modem), but it could in theory be used for
9406 fixing movies that were inverse-telecined incorrectly.
9408 A description of the accepted options follows.
9412 Set the maximum number of consecutive frames which can be dropped (if
9413 positive), or the minimum interval between dropped frames (if
9414 negative). If the value is 0, the frame is dropped unregarding the
9415 number of previous sequentially dropped frames.
9422 Set the dropping threshold values.
9424 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9425 represent actual pixel value differences, so a threshold of 64
9426 corresponds to 1 unit of difference for each pixel, or the same spread
9427 out differently over the block.
9429 A frame is a candidate for dropping if no 8x8 blocks differ by more
9430 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9431 meaning the whole image) differ by more than a threshold of @option{lo}.
9433 Default value for @option{hi} is 64*12, default value for @option{lo} is
9434 64*5, and default value for @option{frac} is 0.33.
9442 It accepts an integer in input; if non-zero it negates the
9443 alpha component (if available). The default value in input is 0.
9447 Deinterlace video using neural network edge directed interpolation.
9449 This filter accepts the following options:
9453 Mandatory option, without binary file filter can not work.
9454 Currently file can be found here:
9455 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9458 Set which frames to deinterlace, by default it is @code{all}.
9459 Can be @code{all} or @code{interlaced}.
9462 Set mode of operation.
9464 Can be one of the following:
9468 Use frame flags, both fields.
9470 Use frame flags, single field.
9474 Use bottom field only.
9476 Use both fields, top first.
9478 Use both fields, bottom first.
9482 Set which planes to process, by default filter process all frames.
9485 Set size of local neighborhood around each pixel, used by the predictor neural
9488 Can be one of the following:
9501 Set the number of neurons in predicctor neural network.
9502 Can be one of the following:
9513 Controls the number of different neural network predictions that are blended
9514 together to compute the final output value. Can be @code{fast}, default or
9518 Set which set of weights to use in the predictor.
9519 Can be one of the following:
9523 weights trained to minimize absolute error
9525 weights trained to minimize squared error
9529 Controls whether or not the prescreener neural network is used to decide
9530 which pixels should be processed by the predictor neural network and which
9531 can be handled by simple cubic interpolation.
9532 The prescreener is trained to know whether cubic interpolation will be
9533 sufficient for a pixel or whether it should be predicted by the predictor nn.
9534 The computational complexity of the prescreener nn is much less than that of
9535 the predictor nn. Since most pixels can be handled by cubic interpolation,
9536 using the prescreener generally results in much faster processing.
9537 The prescreener is pretty accurate, so the difference between using it and not
9538 using it is almost always unnoticeable.
9540 Can be one of the following:
9548 Default is @code{new}.
9551 Set various debugging flags.
9556 Force libavfilter not to use any of the specified pixel formats for the
9557 input to the next filter.
9559 It accepts the following parameters:
9563 A '|'-separated list of pixel format names, such as
9564 apix_fmts=yuv420p|monow|rgb24".
9568 @subsection Examples
9572 Force libavfilter to use a format different from @var{yuv420p} for the
9573 input to the vflip filter:
9575 noformat=pix_fmts=yuv420p,vflip
9579 Convert the input video to any of the formats not contained in the list:
9581 noformat=yuv420p|yuv444p|yuv410p
9587 Add noise on video input frame.
9589 The filter accepts the following options:
9597 Set noise seed for specific pixel component or all pixel components in case
9598 of @var{all_seed}. Default value is @code{123457}.
9600 @item all_strength, alls
9601 @item c0_strength, c0s
9602 @item c1_strength, c1s
9603 @item c2_strength, c2s
9604 @item c3_strength, c3s
9605 Set noise strength for specific pixel component or all pixel components in case
9606 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9608 @item all_flags, allf
9613 Set pixel component flags or set flags for all components if @var{all_flags}.
9614 Available values for component flags are:
9617 averaged temporal noise (smoother)
9619 mix random noise with a (semi)regular pattern
9621 temporal noise (noise pattern changes between frames)
9623 uniform noise (gaussian otherwise)
9627 @subsection Examples
9629 Add temporal and uniform noise to input video:
9631 noise=alls=20:allf=t+u
9636 Pass the video source unchanged to the output.
9639 Optical Character Recognition
9641 This filter uses Tesseract for optical character recognition.
9643 It accepts the following options:
9647 Set datapath to tesseract data. Default is to use whatever was
9648 set at installation.
9651 Set language, default is "eng".
9654 Set character whitelist.
9657 Set character blacklist.
9660 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9664 Apply a video transform using libopencv.
9666 To enable this filter, install the libopencv library and headers and
9667 configure FFmpeg with @code{--enable-libopencv}.
9669 It accepts the following parameters:
9674 The name of the libopencv filter to apply.
9677 The parameters to pass to the libopencv filter. If not specified, the default
9682 Refer to the official libopencv documentation for more precise
9684 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9686 Several libopencv filters are supported; see the following subsections.
9691 Dilate an image by using a specific structuring element.
9692 It corresponds to the libopencv function @code{cvDilate}.
9694 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
9696 @var{struct_el} represents a structuring element, and has the syntax:
9697 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
9699 @var{cols} and @var{rows} represent the number of columns and rows of
9700 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
9701 point, and @var{shape} the shape for the structuring element. @var{shape}
9702 must be "rect", "cross", "ellipse", or "custom".
9704 If the value for @var{shape} is "custom", it must be followed by a
9705 string of the form "=@var{filename}". The file with name
9706 @var{filename} is assumed to represent a binary image, with each
9707 printable character corresponding to a bright pixel. When a custom
9708 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
9709 or columns and rows of the read file are assumed instead.
9711 The default value for @var{struct_el} is "3x3+0x0/rect".
9713 @var{nb_iterations} specifies the number of times the transform is
9714 applied to the image, and defaults to 1.
9718 # Use the default values
9721 # Dilate using a structuring element with a 5x5 cross, iterating two times
9722 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
9724 # Read the shape from the file diamond.shape, iterating two times.
9725 # The file diamond.shape may contain a pattern of characters like this
9731 # The specified columns and rows are ignored
9732 # but the anchor point coordinates are not
9733 ocv=dilate:0x0+2x2/custom=diamond.shape|2
9738 Erode an image by using a specific structuring element.
9739 It corresponds to the libopencv function @code{cvErode}.
9741 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
9742 with the same syntax and semantics as the @ref{dilate} filter.
9746 Smooth the input video.
9748 The filter takes the following parameters:
9749 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
9751 @var{type} is the type of smooth filter to apply, and must be one of
9752 the following values: "blur", "blur_no_scale", "median", "gaussian",
9753 or "bilateral". The default value is "gaussian".
9755 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
9756 depend on the smooth type. @var{param1} and
9757 @var{param2} accept integer positive values or 0. @var{param3} and
9758 @var{param4} accept floating point values.
9760 The default value for @var{param1} is 3. The default value for the
9761 other parameters is 0.
9763 These parameters correspond to the parameters assigned to the
9764 libopencv function @code{cvSmooth}.
9769 Overlay one video on top of another.
9771 It takes two inputs and has one output. The first input is the "main"
9772 video on which the second input is overlaid.
9774 It accepts the following parameters:
9776 A description of the accepted options follows.
9781 Set the expression for the x and y coordinates of the overlaid video
9782 on the main video. Default value is "0" for both expressions. In case
9783 the expression is invalid, it is set to a huge value (meaning that the
9784 overlay will not be displayed within the output visible area).
9787 The action to take when EOF is encountered on the secondary input; it accepts
9788 one of the following values:
9792 Repeat the last frame (the default).
9796 Pass the main input through.
9800 Set when the expressions for @option{x}, and @option{y} are evaluated.
9802 It accepts the following values:
9805 only evaluate expressions once during the filter initialization or
9806 when a command is processed
9809 evaluate expressions for each incoming frame
9812 Default value is @samp{frame}.
9815 If set to 1, force the output to terminate when the shortest input
9816 terminates. Default value is 0.
9819 Set the format for the output video.
9821 It accepts the following values:
9836 Default value is @samp{yuv420}.
9838 @item rgb @emph{(deprecated)}
9839 If set to 1, force the filter to accept inputs in the RGB
9840 color space. Default value is 0. This option is deprecated, use
9841 @option{format} instead.
9844 If set to 1, force the filter to draw the last overlay frame over the
9845 main input until the end of the stream. A value of 0 disables this
9846 behavior. Default value is 1.
9849 The @option{x}, and @option{y} expressions can contain the following
9855 The main input width and height.
9859 The overlay input width and height.
9863 The computed values for @var{x} and @var{y}. They are evaluated for
9868 horizontal and vertical chroma subsample values of the output
9869 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
9873 the number of input frame, starting from 0
9876 the position in the file of the input frame, NAN if unknown
9879 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
9883 Note that the @var{n}, @var{pos}, @var{t} variables are available only
9884 when evaluation is done @emph{per frame}, and will evaluate to NAN
9885 when @option{eval} is set to @samp{init}.
9887 Be aware that frames are taken from each input video in timestamp
9888 order, hence, if their initial timestamps differ, it is a good idea
9889 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
9890 have them begin in the same zero timestamp, as the example for
9891 the @var{movie} filter does.
9893 You can chain together more overlays but you should test the
9894 efficiency of such approach.
9896 @subsection Commands
9898 This filter supports the following commands:
9902 Modify the x and y of the overlay input.
9903 The command accepts the same syntax of the corresponding option.
9905 If the specified expression is not valid, it is kept at its current
9909 @subsection Examples
9913 Draw the overlay at 10 pixels from the bottom right corner of the main
9916 overlay=main_w-overlay_w-10:main_h-overlay_h-10
9919 Using named options the example above becomes:
9921 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
9925 Insert a transparent PNG logo in the bottom left corner of the input,
9926 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
9928 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
9932 Insert 2 different transparent PNG logos (second logo on bottom
9933 right corner) using the @command{ffmpeg} tool:
9935 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
9939 Add a transparent color layer on top of the main video; @code{WxH}
9940 must specify the size of the main input to the overlay filter:
9942 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
9946 Play an original video and a filtered version (here with the deshake
9947 filter) side by side using the @command{ffplay} tool:
9949 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
9952 The above command is the same as:
9954 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
9958 Make a sliding overlay appearing from the left to the right top part of the
9959 screen starting since time 2:
9961 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
9965 Compose output by putting two input videos side to side:
9967 ffmpeg -i left.avi -i right.avi -filter_complex "
9968 nullsrc=size=200x100 [background];
9969 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
9970 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
9971 [background][left] overlay=shortest=1 [background+left];
9972 [background+left][right] overlay=shortest=1:x=100 [left+right]
9977 Mask 10-20 seconds of a video by applying the delogo filter to a section
9979 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
9980 -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]'
9985 Chain several overlays in cascade:
9987 nullsrc=s=200x200 [bg];
9988 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
9989 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
9990 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
9991 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
9992 [in3] null, [mid2] overlay=100:100 [out0]
9999 Apply Overcomplete Wavelet denoiser.
10001 The filter accepts the following options:
10007 Larger depth values will denoise lower frequency components more, but
10008 slow down filtering.
10010 Must be an int in the range 8-16, default is @code{8}.
10012 @item luma_strength, ls
10015 Must be a double value in the range 0-1000, default is @code{1.0}.
10017 @item chroma_strength, cs
10018 Set chroma strength.
10020 Must be a double value in the range 0-1000, default is @code{1.0}.
10026 Add paddings to the input image, and place the original input at the
10027 provided @var{x}, @var{y} coordinates.
10029 It accepts the following parameters:
10034 Specify an expression for the size of the output image with the
10035 paddings added. If the value for @var{width} or @var{height} is 0, the
10036 corresponding input size is used for the output.
10038 The @var{width} expression can reference the value set by the
10039 @var{height} expression, and vice versa.
10041 The default value of @var{width} and @var{height} is 0.
10045 Specify the offsets to place the input image at within the padded area,
10046 with respect to the top/left border of the output image.
10048 The @var{x} expression can reference the value set by the @var{y}
10049 expression, and vice versa.
10051 The default value of @var{x} and @var{y} is 0.
10054 Specify the color of the padded area. For the syntax of this option,
10055 check the "Color" section in the ffmpeg-utils manual.
10057 The default value of @var{color} is "black".
10060 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10061 options are expressions containing the following constants:
10066 The input video width and height.
10070 These are the same as @var{in_w} and @var{in_h}.
10074 The output width and height (the size of the padded area), as
10075 specified by the @var{width} and @var{height} expressions.
10079 These are the same as @var{out_w} and @var{out_h}.
10083 The x and y offsets as specified by the @var{x} and @var{y}
10084 expressions, or NAN if not yet specified.
10087 same as @var{iw} / @var{ih}
10090 input sample aspect ratio
10093 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10097 The horizontal and vertical chroma subsample values. For example for the
10098 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10101 @subsection Examples
10105 Add paddings with the color "violet" to the input video. The output video
10106 size is 640x480, and the top-left corner of the input video is placed at
10109 pad=640:480:0:40:violet
10112 The example above is equivalent to the following command:
10114 pad=width=640:height=480:x=0:y=40:color=violet
10118 Pad the input to get an output with dimensions increased by 3/2,
10119 and put the input video at the center of the padded area:
10121 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10125 Pad the input to get a squared output with size equal to the maximum
10126 value between the input width and height, and put the input video at
10127 the center of the padded area:
10129 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10133 Pad the input to get a final w/h ratio of 16:9:
10135 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10139 In case of anamorphic video, in order to set the output display aspect
10140 correctly, it is necessary to use @var{sar} in the expression,
10141 according to the relation:
10143 (ih * X / ih) * sar = output_dar
10144 X = output_dar / sar
10147 Thus the previous example needs to be modified to:
10149 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10153 Double the output size and put the input video in the bottom-right
10154 corner of the output padded area:
10156 pad="2*iw:2*ih:ow-iw:oh-ih"
10160 @anchor{palettegen}
10161 @section palettegen
10163 Generate one palette for a whole video stream.
10165 It accepts the following options:
10169 Set the maximum number of colors to quantize in the palette.
10170 Note: the palette will still contain 256 colors; the unused palette entries
10173 @item reserve_transparent
10174 Create a palette of 255 colors maximum and reserve the last one for
10175 transparency. Reserving the transparency color is useful for GIF optimization.
10176 If not set, the maximum of colors in the palette will be 256. You probably want
10177 to disable this option for a standalone image.
10181 Set statistics mode.
10183 It accepts the following values:
10186 Compute full frame histograms.
10188 Compute histograms only for the part that differs from previous frame. This
10189 might be relevant to give more importance to the moving part of your input if
10190 the background is static.
10193 Default value is @var{full}.
10196 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10197 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10198 color quantization of the palette. This information is also visible at
10199 @var{info} logging level.
10201 @subsection Examples
10205 Generate a representative palette of a given video using @command{ffmpeg}:
10207 ffmpeg -i input.mkv -vf palettegen palette.png
10211 @section paletteuse
10213 Use a palette to downsample an input video stream.
10215 The filter takes two inputs: one video stream and a palette. The palette must
10216 be a 256 pixels image.
10218 It accepts the following options:
10222 Select dithering mode. Available algorithms are:
10225 Ordered 8x8 bayer dithering (deterministic)
10227 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10228 Note: this dithering is sometimes considered "wrong" and is included as a
10230 @item floyd_steinberg
10231 Floyd and Steingberg dithering (error diffusion)
10233 Frankie Sierra dithering v2 (error diffusion)
10235 Frankie Sierra dithering v2 "Lite" (error diffusion)
10238 Default is @var{sierra2_4a}.
10241 When @var{bayer} dithering is selected, this option defines the scale of the
10242 pattern (how much the crosshatch pattern is visible). A low value means more
10243 visible pattern for less banding, and higher value means less visible pattern
10244 at the cost of more banding.
10246 The option must be an integer value in the range [0,5]. Default is @var{2}.
10249 If set, define the zone to process
10253 Only the changing rectangle will be reprocessed. This is similar to GIF
10254 cropping/offsetting compression mechanism. This option can be useful for speed
10255 if only a part of the image is changing, and has use cases such as limiting the
10256 scope of the error diffusal @option{dither} to the rectangle that bounds the
10257 moving scene (it leads to more deterministic output if the scene doesn't change
10258 much, and as a result less moving noise and better GIF compression).
10261 Default is @var{none}.
10264 @subsection Examples
10268 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10269 using @command{ffmpeg}:
10271 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10275 @section perspective
10277 Correct perspective of video not recorded perpendicular to the screen.
10279 A description of the accepted parameters follows.
10290 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10291 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10292 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10293 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10294 then the corners of the source will be sent to the specified coordinates.
10296 The expressions can use the following variables:
10301 the width and height of video frame.
10305 Output frame count.
10308 @item interpolation
10309 Set interpolation for perspective correction.
10311 It accepts the following values:
10317 Default value is @samp{linear}.
10320 Set interpretation of coordinate options.
10322 It accepts the following values:
10326 Send point in the source specified by the given coordinates to
10327 the corners of the destination.
10329 @item 1, destination
10331 Send the corners of the source to the point in the destination specified
10332 by the given coordinates.
10334 Default value is @samp{source}.
10338 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10340 It accepts the following values:
10343 only evaluate expressions once during the filter initialization or
10344 when a command is processed
10347 evaluate expressions for each incoming frame
10350 Default value is @samp{init}.
10355 Delay interlaced video by one field time so that the field order changes.
10357 The intended use is to fix PAL movies that have been captured with the
10358 opposite field order to the film-to-video transfer.
10360 A description of the accepted parameters follows.
10366 It accepts the following values:
10369 Capture field order top-first, transfer bottom-first.
10370 Filter will delay the bottom field.
10373 Capture field order bottom-first, transfer top-first.
10374 Filter will delay the top field.
10377 Capture and transfer with the same field order. This mode only exists
10378 for the documentation of the other options to refer to, but if you
10379 actually select it, the filter will faithfully do nothing.
10382 Capture field order determined automatically by field flags, transfer
10384 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10385 basis using field flags. If no field information is available,
10386 then this works just like @samp{u}.
10389 Capture unknown or varying, transfer opposite.
10390 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10391 analyzing the images and selecting the alternative that produces best
10392 match between the fields.
10395 Capture top-first, transfer unknown or varying.
10396 Filter selects among @samp{t} and @samp{p} using image analysis.
10399 Capture bottom-first, transfer unknown or varying.
10400 Filter selects among @samp{b} and @samp{p} using image analysis.
10403 Capture determined by field flags, transfer unknown or varying.
10404 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10405 image analysis. If no field information is available, then this works just
10406 like @samp{U}. This is the default mode.
10409 Both capture and transfer unknown or varying.
10410 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10414 @section pixdesctest
10416 Pixel format descriptor test filter, mainly useful for internal
10417 testing. The output video should be equal to the input video.
10421 format=monow, pixdesctest
10424 can be used to test the monowhite pixel format descriptor definition.
10428 Enable the specified chain of postprocessing subfilters using libpostproc. This
10429 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10430 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10431 Each subfilter and some options have a short and a long name that can be used
10432 interchangeably, i.e. dr/dering are the same.
10434 The filters accept the following options:
10438 Set postprocessing subfilters string.
10441 All subfilters share common options to determine their scope:
10445 Honor the quality commands for this subfilter.
10448 Do chrominance filtering, too (default).
10451 Do luminance filtering only (no chrominance).
10454 Do chrominance filtering only (no luminance).
10457 These options can be appended after the subfilter name, separated by a '|'.
10459 Available subfilters are:
10462 @item hb/hdeblock[|difference[|flatness]]
10463 Horizontal deblocking filter
10466 Difference factor where higher values mean more deblocking (default: @code{32}).
10468 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10471 @item vb/vdeblock[|difference[|flatness]]
10472 Vertical deblocking filter
10475 Difference factor where higher values mean more deblocking (default: @code{32}).
10477 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10480 @item ha/hadeblock[|difference[|flatness]]
10481 Accurate horizontal deblocking filter
10484 Difference factor where higher values mean more deblocking (default: @code{32}).
10486 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10489 @item va/vadeblock[|difference[|flatness]]
10490 Accurate vertical deblocking filter
10493 Difference factor where higher values mean more deblocking (default: @code{32}).
10495 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10499 The horizontal and vertical deblocking filters share the difference and
10500 flatness values so you cannot set different horizontal and vertical
10504 @item h1/x1hdeblock
10505 Experimental horizontal deblocking filter
10507 @item v1/x1vdeblock
10508 Experimental vertical deblocking filter
10513 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10516 larger -> stronger filtering
10518 larger -> stronger filtering
10520 larger -> stronger filtering
10523 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10526 Stretch luminance to @code{0-255}.
10529 @item lb/linblenddeint
10530 Linear blend deinterlacing filter that deinterlaces the given block by
10531 filtering all lines with a @code{(1 2 1)} filter.
10533 @item li/linipoldeint
10534 Linear interpolating deinterlacing filter that deinterlaces the given block by
10535 linearly interpolating every second line.
10537 @item ci/cubicipoldeint
10538 Cubic interpolating deinterlacing filter deinterlaces the given block by
10539 cubically interpolating every second line.
10541 @item md/mediandeint
10542 Median deinterlacing filter that deinterlaces the given block by applying a
10543 median filter to every second line.
10545 @item fd/ffmpegdeint
10546 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10547 second line with a @code{(-1 4 2 4 -1)} filter.
10550 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10551 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10553 @item fq/forceQuant[|quantizer]
10554 Overrides the quantizer table from the input with the constant quantizer you
10562 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10565 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10568 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10571 @subsection Examples
10575 Apply horizontal and vertical deblocking, deringing and automatic
10576 brightness/contrast:
10582 Apply default filters without brightness/contrast correction:
10588 Apply default filters and temporal denoiser:
10590 pp=default/tmpnoise|1|2|3
10594 Apply deblocking on luminance only, and switch vertical deblocking on or off
10595 automatically depending on available CPU time:
10602 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10603 similar to spp = 6 with 7 point DCT, where only the center sample is
10606 The filter accepts the following options:
10610 Force a constant quantization parameter. It accepts an integer in range
10611 0 to 63. If not set, the filter will use the QP from the video stream
10615 Set thresholding mode. Available modes are:
10619 Set hard thresholding.
10621 Set soft thresholding (better de-ringing effect, but likely blurrier).
10623 Set medium thresholding (good results, default).
10629 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10630 Ratio) between two input videos.
10632 This filter takes in input two input videos, the first input is
10633 considered the "main" source and is passed unchanged to the
10634 output. The second input is used as a "reference" video for computing
10637 Both video inputs must have the same resolution and pixel format for
10638 this filter to work correctly. Also it assumes that both inputs
10639 have the same number of frames, which are compared one by one.
10641 The obtained average PSNR is printed through the logging system.
10643 The filter stores the accumulated MSE (mean squared error) of each
10644 frame, and at the end of the processing it is averaged across all frames
10645 equally, and the following formula is applied to obtain the PSNR:
10648 PSNR = 10*log10(MAX^2/MSE)
10651 Where MAX is the average of the maximum values of each component of the
10654 The description of the accepted parameters follows.
10657 @item stats_file, f
10658 If specified the filter will use the named file to save the PSNR of
10659 each individual frame. When filename equals "-" the data is sent to
10662 @item stats_version
10663 Specifies which version of the stats file format to use. Details of
10664 each format are written below.
10665 Default value is 1.
10668 The file printed if @var{stats_file} is selected, contains a sequence of
10669 key/value pairs of the form @var{key}:@var{value} for each compared
10672 If a @var{stats_version} greater than 1 is specified, a header line precedes
10673 the list of per-frame-pair stats, with key value pairs following the frame
10674 format with the following parameters:
10677 @item psnr_log_version
10678 The version of the log file format. Will match @var{stats_version}.
10681 A comma separated list of the per-frame-pair parameters included in
10685 A description of each shown per-frame-pair parameter follows:
10689 sequential number of the input frame, starting from 1
10692 Mean Square Error pixel-by-pixel average difference of the compared
10693 frames, averaged over all the image components.
10695 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
10696 Mean Square Error pixel-by-pixel average difference of the compared
10697 frames for the component specified by the suffix.
10699 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
10700 Peak Signal to Noise ratio of the compared frames for the component
10701 specified by the suffix.
10706 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
10707 [main][ref] psnr="stats_file=stats.log" [out]
10710 On this example the input file being processed is compared with the
10711 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
10712 is stored in @file{stats.log}.
10717 Pulldown reversal (inverse telecine) filter, capable of handling mixed
10718 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
10721 The pullup filter is designed to take advantage of future context in making
10722 its decisions. This filter is stateless in the sense that it does not lock
10723 onto a pattern to follow, but it instead looks forward to the following
10724 fields in order to identify matches and rebuild progressive frames.
10726 To produce content with an even framerate, insert the fps filter after
10727 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
10728 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
10730 The filter accepts the following options:
10737 These options set the amount of "junk" to ignore at the left, right, top, and
10738 bottom of the image, respectively. Left and right are in units of 8 pixels,
10739 while top and bottom are in units of 2 lines.
10740 The default is 8 pixels on each side.
10743 Set the strict breaks. Setting this option to 1 will reduce the chances of
10744 filter generating an occasional mismatched frame, but it may also cause an
10745 excessive number of frames to be dropped during high motion sequences.
10746 Conversely, setting it to -1 will make filter match fields more easily.
10747 This may help processing of video where there is slight blurring between
10748 the fields, but may also cause there to be interlaced frames in the output.
10749 Default value is @code{0}.
10752 Set the metric plane to use. It accepts the following values:
10758 Use chroma blue plane.
10761 Use chroma red plane.
10764 This option may be set to use chroma plane instead of the default luma plane
10765 for doing filter's computations. This may improve accuracy on very clean
10766 source material, but more likely will decrease accuracy, especially if there
10767 is chroma noise (rainbow effect) or any grayscale video.
10768 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
10769 load and make pullup usable in realtime on slow machines.
10772 For best results (without duplicated frames in the output file) it is
10773 necessary to change the output frame rate. For example, to inverse
10774 telecine NTSC input:
10776 ffmpeg -i input -vf pullup -r 24000/1001 ...
10781 Change video quantization parameters (QP).
10783 The filter accepts the following option:
10787 Set expression for quantization parameter.
10790 The expression is evaluated through the eval API and can contain, among others,
10791 the following constants:
10795 1 if index is not 129, 0 otherwise.
10798 Sequentional index starting from -129 to 128.
10801 @subsection Examples
10805 Some equation like:
10813 Flush video frames from internal cache of frames into a random order.
10814 No frame is discarded.
10815 Inspired by @ref{frei0r} nervous filter.
10819 Set size in number of frames of internal cache, in range from @code{2} to
10820 @code{512}. Default is @code{30}.
10823 Set seed for random number generator, must be an integer included between
10824 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
10825 less than @code{0}, the filter will try to use a good random seed on a
10831 Read vertical interval timecode (VITC) information from the top lines of a
10834 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
10835 timecode value, if a valid timecode has been detected. Further metadata key
10836 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
10837 timecode data has been found or not.
10839 This filter accepts the following options:
10843 Set the maximum number of lines to scan for VITC data. If the value is set to
10844 @code{-1} the full video frame is scanned. Default is @code{45}.
10847 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
10848 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
10851 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
10852 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
10855 @subsection Examples
10859 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
10860 draw @code{--:--:--:--} as a placeholder:
10862 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
10868 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
10870 Destination pixel at position (X, Y) will be picked from source (x, y) position
10871 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
10872 value for pixel will be used for destination pixel.
10874 Xmap and Ymap input video streams must be of same dimensions. Output video stream
10875 will have Xmap/Ymap video stream dimensions.
10876 Xmap and Ymap input video streams are 16bit depth, single channel.
10878 @section removegrain
10880 The removegrain filter is a spatial denoiser for progressive video.
10884 Set mode for the first plane.
10887 Set mode for the second plane.
10890 Set mode for the third plane.
10893 Set mode for the fourth plane.
10896 Range of mode is from 0 to 24. Description of each mode follows:
10900 Leave input plane unchanged. Default.
10903 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
10906 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
10909 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
10912 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
10913 This is equivalent to a median filter.
10916 Line-sensitive clipping giving the minimal change.
10919 Line-sensitive clipping, intermediate.
10922 Line-sensitive clipping, intermediate.
10925 Line-sensitive clipping, intermediate.
10928 Line-sensitive clipping on a line where the neighbours pixels are the closest.
10931 Replaces the target pixel with the closest neighbour.
10934 [1 2 1] horizontal and vertical kernel blur.
10940 Bob mode, interpolates top field from the line where the neighbours
10941 pixels are the closest.
10944 Bob mode, interpolates bottom field from the line where the neighbours
10945 pixels are the closest.
10948 Bob mode, interpolates top field. Same as 13 but with a more complicated
10949 interpolation formula.
10952 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
10953 interpolation formula.
10956 Clips the pixel with the minimum and maximum of respectively the maximum and
10957 minimum of each pair of opposite neighbour pixels.
10960 Line-sensitive clipping using opposite neighbours whose greatest distance from
10961 the current pixel is minimal.
10964 Replaces the pixel with the average of its 8 neighbours.
10967 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
10970 Clips pixels using the averages of opposite neighbour.
10973 Same as mode 21 but simpler and faster.
10976 Small edge and halo removal, but reputed useless.
10982 @section removelogo
10984 Suppress a TV station logo, using an image file to determine which
10985 pixels comprise the logo. It works by filling in the pixels that
10986 comprise the logo with neighboring pixels.
10988 The filter accepts the following options:
10992 Set the filter bitmap file, which can be any image format supported by
10993 libavformat. The width and height of the image file must match those of the
10994 video stream being processed.
10997 Pixels in the provided bitmap image with a value of zero are not
10998 considered part of the logo, non-zero pixels are considered part of
10999 the logo. If you use white (255) for the logo and black (0) for the
11000 rest, you will be safe. For making the filter bitmap, it is
11001 recommended to take a screen capture of a black frame with the logo
11002 visible, and then using a threshold filter followed by the erode
11003 filter once or twice.
11005 If needed, little splotches can be fixed manually. Remember that if
11006 logo pixels are not covered, the filter quality will be much
11007 reduced. Marking too many pixels as part of the logo does not hurt as
11008 much, but it will increase the amount of blurring needed to cover over
11009 the image and will destroy more information than necessary, and extra
11010 pixels will slow things down on a large logo.
11012 @section repeatfields
11014 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11015 fields based on its value.
11019 Reverse a video clip.
11021 Warning: This filter requires memory to buffer the entire clip, so trimming
11024 @subsection Examples
11028 Take the first 5 seconds of a clip, and reverse it.
11036 Rotate video by an arbitrary angle expressed in radians.
11038 The filter accepts the following options:
11040 A description of the optional parameters follows.
11043 Set an expression for the angle by which to rotate the input video
11044 clockwise, expressed as a number of radians. A negative value will
11045 result in a counter-clockwise rotation. By default it is set to "0".
11047 This expression is evaluated for each frame.
11050 Set the output width expression, default value is "iw".
11051 This expression is evaluated just once during configuration.
11054 Set the output height expression, default value is "ih".
11055 This expression is evaluated just once during configuration.
11058 Enable bilinear interpolation if set to 1, a value of 0 disables
11059 it. Default value is 1.
11062 Set the color used to fill the output area not covered by the rotated
11063 image. For the general syntax of this option, check the "Color" section in the
11064 ffmpeg-utils manual. If the special value "none" is selected then no
11065 background is printed (useful for example if the background is never shown).
11067 Default value is "black".
11070 The expressions for the angle and the output size can contain the
11071 following constants and functions:
11075 sequential number of the input frame, starting from 0. It is always NAN
11076 before the first frame is filtered.
11079 time in seconds of the input frame, it is set to 0 when the filter is
11080 configured. It is always NAN before the first frame is filtered.
11084 horizontal and vertical chroma subsample values. For example for the
11085 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11089 the input video width and height
11093 the output width and height, that is the size of the padded area as
11094 specified by the @var{width} and @var{height} expressions
11098 the minimal width/height required for completely containing the input
11099 video rotated by @var{a} radians.
11101 These are only available when computing the @option{out_w} and
11102 @option{out_h} expressions.
11105 @subsection Examples
11109 Rotate the input by PI/6 radians clockwise:
11115 Rotate the input by PI/6 radians counter-clockwise:
11121 Rotate the input by 45 degrees clockwise:
11127 Apply a constant rotation with period T, starting from an angle of PI/3:
11129 rotate=PI/3+2*PI*t/T
11133 Make the input video rotation oscillating with a period of T
11134 seconds and an amplitude of A radians:
11136 rotate=A*sin(2*PI/T*t)
11140 Rotate the video, output size is chosen so that the whole rotating
11141 input video is always completely contained in the output:
11143 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11147 Rotate the video, reduce the output size so that no background is ever
11150 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11154 @subsection Commands
11156 The filter supports the following commands:
11160 Set the angle expression.
11161 The command accepts the same syntax of the corresponding option.
11163 If the specified expression is not valid, it is kept at its current
11169 Apply Shape Adaptive Blur.
11171 The filter accepts the following options:
11174 @item luma_radius, lr
11175 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11176 value is 1.0. A greater value will result in a more blurred image, and
11177 in slower processing.
11179 @item luma_pre_filter_radius, lpfr
11180 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11183 @item luma_strength, ls
11184 Set luma maximum difference between pixels to still be considered, must
11185 be a value in the 0.1-100.0 range, default value is 1.0.
11187 @item chroma_radius, cr
11188 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11189 greater value will result in a more blurred image, and in slower
11192 @item chroma_pre_filter_radius, cpfr
11193 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11195 @item chroma_strength, cs
11196 Set chroma maximum difference between pixels to still be considered,
11197 must be a value in the -0.9-100.0 range.
11200 Each chroma option value, if not explicitly specified, is set to the
11201 corresponding luma option value.
11206 Scale (resize) the input video, using the libswscale library.
11208 The scale filter forces the output display aspect ratio to be the same
11209 of the input, by changing the output sample aspect ratio.
11211 If the input image format is different from the format requested by
11212 the next filter, the scale filter will convert the input to the
11215 @subsection Options
11216 The filter accepts the following options, or any of the options
11217 supported by the libswscale scaler.
11219 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11220 the complete list of scaler options.
11225 Set the output video dimension expression. Default value is the input
11228 If the value is 0, the input width is used for the output.
11230 If one of the values is -1, the scale filter will use a value that
11231 maintains the aspect ratio of the input image, calculated from the
11232 other specified dimension. If both of them are -1, the input size is
11235 If one of the values is -n with n > 1, the scale filter will also use a value
11236 that maintains the aspect ratio of the input image, calculated from the other
11237 specified dimension. After that it will, however, make sure that the calculated
11238 dimension is divisible by n and adjust the value if necessary.
11240 See below for the list of accepted constants for use in the dimension
11244 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11248 Only evaluate expressions once during the filter initialization or when a command is processed.
11251 Evaluate expressions for each incoming frame.
11255 Default value is @samp{init}.
11259 Set the interlacing mode. It accepts the following values:
11263 Force interlaced aware scaling.
11266 Do not apply interlaced scaling.
11269 Select interlaced aware scaling depending on whether the source frames
11270 are flagged as interlaced or not.
11273 Default value is @samp{0}.
11276 Set libswscale scaling flags. See
11277 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11278 complete list of values. If not explicitly specified the filter applies
11282 @item param0, param1
11283 Set libswscale input parameters for scaling algorithms that need them. See
11284 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11285 complete documentation. If not explicitly specified the filter applies
11291 Set the video size. For the syntax of this option, check the
11292 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11294 @item in_color_matrix
11295 @item out_color_matrix
11296 Set in/output YCbCr color space type.
11298 This allows the autodetected value to be overridden as well as allows forcing
11299 a specific value used for the output and encoder.
11301 If not specified, the color space type depends on the pixel format.
11307 Choose automatically.
11310 Format conforming to International Telecommunication Union (ITU)
11311 Recommendation BT.709.
11314 Set color space conforming to the United States Federal Communications
11315 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11318 Set color space conforming to:
11322 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11325 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11328 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11333 Set color space conforming to SMPTE ST 240:1999.
11338 Set in/output YCbCr sample range.
11340 This allows the autodetected value to be overridden as well as allows forcing
11341 a specific value used for the output and encoder. If not specified, the
11342 range depends on the pixel format. Possible values:
11346 Choose automatically.
11349 Set full range (0-255 in case of 8-bit luma).
11352 Set "MPEG" range (16-235 in case of 8-bit luma).
11355 @item force_original_aspect_ratio
11356 Enable decreasing or increasing output video width or height if necessary to
11357 keep the original aspect ratio. Possible values:
11361 Scale the video as specified and disable this feature.
11364 The output video dimensions will automatically be decreased if needed.
11367 The output video dimensions will automatically be increased if needed.
11371 One useful instance of this option is that when you know a specific device's
11372 maximum allowed resolution, you can use this to limit the output video to
11373 that, while retaining the aspect ratio. For example, device A allows
11374 1280x720 playback, and your video is 1920x800. Using this option (set it to
11375 decrease) and specifying 1280x720 to the command line makes the output
11378 Please note that this is a different thing than specifying -1 for @option{w}
11379 or @option{h}, you still need to specify the output resolution for this option
11384 The values of the @option{w} and @option{h} options are expressions
11385 containing the following constants:
11390 The input width and height
11394 These are the same as @var{in_w} and @var{in_h}.
11398 The output (scaled) width and height
11402 These are the same as @var{out_w} and @var{out_h}
11405 The same as @var{iw} / @var{ih}
11408 input sample aspect ratio
11411 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11415 horizontal and vertical input chroma subsample values. For example for the
11416 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11420 horizontal and vertical output chroma subsample values. For example for the
11421 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11424 @subsection Examples
11428 Scale the input video to a size of 200x100
11433 This is equivalent to:
11444 Specify a size abbreviation for the output size:
11449 which can also be written as:
11455 Scale the input to 2x:
11457 scale=w=2*iw:h=2*ih
11461 The above is the same as:
11463 scale=2*in_w:2*in_h
11467 Scale the input to 2x with forced interlaced scaling:
11469 scale=2*iw:2*ih:interl=1
11473 Scale the input to half size:
11475 scale=w=iw/2:h=ih/2
11479 Increase the width, and set the height to the same size:
11485 Seek Greek harmony:
11492 Increase the height, and set the width to 3/2 of the height:
11494 scale=w=3/2*oh:h=3/5*ih
11498 Increase the size, making the size a multiple of the chroma
11501 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11505 Increase the width to a maximum of 500 pixels,
11506 keeping the same aspect ratio as the input:
11508 scale=w='min(500\, iw*3/2):h=-1'
11512 @subsection Commands
11514 This filter supports the following commands:
11518 Set the output video dimension expression.
11519 The command accepts the same syntax of the corresponding option.
11521 If the specified expression is not valid, it is kept at its current
11527 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11528 format conversion on CUDA video frames. Setting the output width and height
11529 works in the same way as for the @var{scale} filter.
11531 The following additional options are accepted:
11534 The pixel format of the output CUDA frames. If set to the string "same" (the
11535 default), the input format will be kept. Note that automatic format negotiation
11536 and conversion is not yet supported for hardware frames
11539 The interpolation algorithm used for resizing. One of the following:
11546 @item cubic2p_bspline
11547 2-parameter cubic (B=1, C=0)
11549 @item cubic2p_catmullrom
11550 2-parameter cubic (B=0, C=1/2)
11552 @item cubic2p_b05c03
11553 2-parameter cubic (B=1/2, C=3/10)
11565 Scale (resize) the input video, based on a reference video.
11567 See the scale filter for available options, scale2ref supports the same but
11568 uses the reference video instead of the main input as basis.
11570 @subsection Examples
11574 Scale a subtitle stream to match the main video in size before overlaying
11576 'scale2ref[b][a];[a][b]overlay'
11580 @anchor{selectivecolor}
11581 @section selectivecolor
11583 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11584 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11585 by the "purity" of the color (that is, how saturated it already is).
11587 This filter is similar to the Adobe Photoshop Selective Color tool.
11589 The filter accepts the following options:
11592 @item correction_method
11593 Select color correction method.
11595 Available values are:
11598 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11601 Specified adjustments are relative to the original component value.
11603 Default is @code{absolute}.
11605 Adjustments for red pixels (pixels where the red component is the maximum)
11607 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11609 Adjustments for green pixels (pixels where the green component is the maximum)
11611 Adjustments for cyan pixels (pixels where the red component is the minimum)
11613 Adjustments for blue pixels (pixels where the blue component is the maximum)
11615 Adjustments for magenta pixels (pixels where the green component is the minimum)
11617 Adjustments for white pixels (pixels where all components are greater than 128)
11619 Adjustments for all pixels except pure black and pure white
11621 Adjustments for black pixels (pixels where all components are lesser than 128)
11623 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11626 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11627 4 space separated floating point adjustment values in the [-1,1] range,
11628 respectively to adjust the amount of cyan, magenta, yellow and black for the
11629 pixels of its range.
11631 @subsection Examples
11635 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11636 increase magenta by 27% in blue areas:
11638 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11642 Use a Photoshop selective color preset:
11644 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11648 @section separatefields
11650 The @code{separatefields} takes a frame-based video input and splits
11651 each frame into its components fields, producing a new half height clip
11652 with twice the frame rate and twice the frame count.
11654 This filter use field-dominance information in frame to decide which
11655 of each pair of fields to place first in the output.
11656 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
11658 @section setdar, setsar
11660 The @code{setdar} filter sets the Display Aspect Ratio for the filter
11663 This is done by changing the specified Sample (aka Pixel) Aspect
11664 Ratio, according to the following equation:
11666 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
11669 Keep in mind that the @code{setdar} filter does not modify the pixel
11670 dimensions of the video frame. Also, the display aspect ratio set by
11671 this filter may be changed by later filters in the filterchain,
11672 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
11675 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
11676 the filter output video.
11678 Note that as a consequence of the application of this filter, the
11679 output display aspect ratio will change according to the equation
11682 Keep in mind that the sample aspect ratio set by the @code{setsar}
11683 filter may be changed by later filters in the filterchain, e.g. if
11684 another "setsar" or a "setdar" filter is applied.
11686 It accepts the following parameters:
11689 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
11690 Set the aspect ratio used by the filter.
11692 The parameter can be a floating point number string, an expression, or
11693 a string of the form @var{num}:@var{den}, where @var{num} and
11694 @var{den} are the numerator and denominator of the aspect ratio. If
11695 the parameter is not specified, it is assumed the value "0".
11696 In case the form "@var{num}:@var{den}" is used, the @code{:} character
11700 Set the maximum integer value to use for expressing numerator and
11701 denominator when reducing the expressed aspect ratio to a rational.
11702 Default value is @code{100}.
11706 The parameter @var{sar} is an expression containing
11707 the following constants:
11711 These are approximated values for the mathematical constants e
11712 (Euler's number), pi (Greek pi), and phi (the golden ratio).
11715 The input width and height.
11718 These are the same as @var{w} / @var{h}.
11721 The input sample aspect ratio.
11724 The input display aspect ratio. It is the same as
11725 (@var{w} / @var{h}) * @var{sar}.
11728 Horizontal and vertical chroma subsample values. For example, for the
11729 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11732 @subsection Examples
11737 To change the display aspect ratio to 16:9, specify one of the following:
11744 To change the sample aspect ratio to 10:11, specify:
11750 To set a display aspect ratio of 16:9, and specify a maximum integer value of
11751 1000 in the aspect ratio reduction, use the command:
11753 setdar=ratio=16/9:max=1000
11761 Force field for the output video frame.
11763 The @code{setfield} filter marks the interlace type field for the
11764 output frames. It does not change the input frame, but only sets the
11765 corresponding property, which affects how the frame is treated by
11766 following filters (e.g. @code{fieldorder} or @code{yadif}).
11768 The filter accepts the following options:
11773 Available values are:
11777 Keep the same field property.
11780 Mark the frame as bottom-field-first.
11783 Mark the frame as top-field-first.
11786 Mark the frame as progressive.
11792 Show a line containing various information for each input video frame.
11793 The input video is not modified.
11795 The shown line contains a sequence of key/value pairs of the form
11796 @var{key}:@var{value}.
11798 The following values are shown in the output:
11802 The (sequential) number of the input frame, starting from 0.
11805 The Presentation TimeStamp of the input frame, expressed as a number of
11806 time base units. The time base unit depends on the filter input pad.
11809 The Presentation TimeStamp of the input frame, expressed as a number of
11813 The position of the frame in the input stream, or -1 if this information is
11814 unavailable and/or meaningless (for example in case of synthetic video).
11817 The pixel format name.
11820 The sample aspect ratio of the input frame, expressed in the form
11821 @var{num}/@var{den}.
11824 The size of the input frame. For the syntax of this option, check the
11825 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11828 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
11829 for bottom field first).
11832 This is 1 if the frame is a key frame, 0 otherwise.
11835 The picture type of the input frame ("I" for an I-frame, "P" for a
11836 P-frame, "B" for a B-frame, or "?" for an unknown type).
11837 Also refer to the documentation of the @code{AVPictureType} enum and of
11838 the @code{av_get_picture_type_char} function defined in
11839 @file{libavutil/avutil.h}.
11842 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
11844 @item plane_checksum
11845 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
11846 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
11849 @section showpalette
11851 Displays the 256 colors palette of each frame. This filter is only relevant for
11852 @var{pal8} pixel format frames.
11854 It accepts the following option:
11858 Set the size of the box used to represent one palette color entry. Default is
11859 @code{30} (for a @code{30x30} pixel box).
11862 @section shuffleframes
11864 Reorder and/or duplicate video frames.
11866 It accepts the following parameters:
11870 Set the destination indexes of input frames.
11871 This is space or '|' separated list of indexes that maps input frames to output
11872 frames. Number of indexes also sets maximal value that each index may have.
11875 The first frame has the index 0. The default is to keep the input unchanged.
11877 Swap second and third frame of every three frames of the input:
11879 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
11882 @section shuffleplanes
11884 Reorder and/or duplicate video planes.
11886 It accepts the following parameters:
11891 The index of the input plane to be used as the first output plane.
11894 The index of the input plane to be used as the second output plane.
11897 The index of the input plane to be used as the third output plane.
11900 The index of the input plane to be used as the fourth output plane.
11904 The first plane has the index 0. The default is to keep the input unchanged.
11906 Swap the second and third planes of the input:
11908 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
11911 @anchor{signalstats}
11912 @section signalstats
11913 Evaluate various visual metrics that assist in determining issues associated
11914 with the digitization of analog video media.
11916 By default the filter will log these metadata values:
11920 Display the minimal Y value contained within the input frame. Expressed in
11924 Display the Y value at the 10% percentile within the input frame. Expressed in
11928 Display the average Y value within the input frame. Expressed in range of
11932 Display the Y value at the 90% percentile within the input frame. Expressed in
11936 Display the maximum Y value contained within the input frame. Expressed in
11940 Display the minimal U value contained within the input frame. Expressed in
11944 Display the U value at the 10% percentile within the input frame. Expressed in
11948 Display the average U value within the input frame. Expressed in range of
11952 Display the U value at the 90% percentile within the input frame. Expressed in
11956 Display the maximum U value contained within the input frame. Expressed in
11960 Display the minimal V value contained within the input frame. Expressed in
11964 Display the V value at the 10% percentile within the input frame. Expressed in
11968 Display the average V value within the input frame. Expressed in range of
11972 Display the V value at the 90% percentile within the input frame. Expressed in
11976 Display the maximum V value contained within the input frame. Expressed in
11980 Display the minimal saturation value contained within the input frame.
11981 Expressed in range of [0-~181.02].
11984 Display the saturation value at the 10% percentile within the input frame.
11985 Expressed in range of [0-~181.02].
11988 Display the average saturation value within the input frame. Expressed in range
11992 Display the saturation value at the 90% percentile within the input frame.
11993 Expressed in range of [0-~181.02].
11996 Display the maximum saturation value contained within the input frame.
11997 Expressed in range of [0-~181.02].
12000 Display the median value for hue within the input frame. Expressed in range of
12004 Display the average value for hue within the input frame. Expressed in range of
12008 Display the average of sample value difference between all values of the Y
12009 plane in the current frame and corresponding values of the previous input frame.
12010 Expressed in range of [0-255].
12013 Display the average of sample value difference between all values of the U
12014 plane in the current frame and corresponding values of the previous input frame.
12015 Expressed in range of [0-255].
12018 Display the average of sample value difference between all values of the V
12019 plane in the current frame and corresponding values of the previous input frame.
12020 Expressed in range of [0-255].
12023 Display bit depth of Y plane in current frame.
12024 Expressed in range of [0-16].
12027 Display bit depth of U plane in current frame.
12028 Expressed in range of [0-16].
12031 Display bit depth of V plane in current frame.
12032 Expressed in range of [0-16].
12035 The filter accepts the following options:
12041 @option{stat} specify an additional form of image analysis.
12042 @option{out} output video with the specified type of pixel highlighted.
12044 Both options accept the following values:
12048 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12049 unlike the neighboring pixels of the same field. Examples of temporal outliers
12050 include the results of video dropouts, head clogs, or tape tracking issues.
12053 Identify @var{vertical line repetition}. Vertical line repetition includes
12054 similar rows of pixels within a frame. In born-digital video vertical line
12055 repetition is common, but this pattern is uncommon in video digitized from an
12056 analog source. When it occurs in video that results from the digitization of an
12057 analog source it can indicate concealment from a dropout compensator.
12060 Identify pixels that fall outside of legal broadcast range.
12064 Set the highlight color for the @option{out} option. The default color is
12068 @subsection Examples
12072 Output data of various video metrics:
12074 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12078 Output specific data about the minimum and maximum values of the Y plane per frame:
12080 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12084 Playback video while highlighting pixels that are outside of broadcast range in red.
12086 ffplay example.mov -vf signalstats="out=brng:color=red"
12090 Playback video with signalstats metadata drawn over the frame.
12092 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12095 The contents of signalstat_drawtext.txt used in the command are:
12098 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12099 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12100 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12101 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12109 Blur the input video without impacting the outlines.
12111 It accepts the following options:
12114 @item luma_radius, lr
12115 Set the luma radius. The option value must be a float number in
12116 the range [0.1,5.0] that specifies the variance of the gaussian filter
12117 used to blur the image (slower if larger). Default value is 1.0.
12119 @item luma_strength, ls
12120 Set the luma strength. The option value must be a float number
12121 in the range [-1.0,1.0] that configures the blurring. A value included
12122 in [0.0,1.0] will blur the image whereas a value included in
12123 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12125 @item luma_threshold, lt
12126 Set the luma threshold used as a coefficient to determine
12127 whether a pixel should be blurred or not. The option value must be an
12128 integer in the range [-30,30]. A value of 0 will filter all the image,
12129 a value included in [0,30] will filter flat areas and a value included
12130 in [-30,0] will filter edges. Default value is 0.
12132 @item chroma_radius, cr
12133 Set the chroma radius. The option value must be a float number in
12134 the range [0.1,5.0] that specifies the variance of the gaussian filter
12135 used to blur the image (slower if larger). Default value is 1.0.
12137 @item chroma_strength, cs
12138 Set the chroma strength. The option value must be a float number
12139 in the range [-1.0,1.0] that configures the blurring. A value included
12140 in [0.0,1.0] will blur the image whereas a value included in
12141 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12143 @item chroma_threshold, ct
12144 Set the chroma threshold used as a coefficient to determine
12145 whether a pixel should be blurred or not. The option value must be an
12146 integer in the range [-30,30]. A value of 0 will filter all the image,
12147 a value included in [0,30] will filter flat areas and a value included
12148 in [-30,0] will filter edges. Default value is 0.
12151 If a chroma option is not explicitly set, the corresponding luma value
12156 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12158 This filter takes in input two input videos, the first input is
12159 considered the "main" source and is passed unchanged to the
12160 output. The second input is used as a "reference" video for computing
12163 Both video inputs must have the same resolution and pixel format for
12164 this filter to work correctly. Also it assumes that both inputs
12165 have the same number of frames, which are compared one by one.
12167 The filter stores the calculated SSIM of each frame.
12169 The description of the accepted parameters follows.
12172 @item stats_file, f
12173 If specified the filter will use the named file to save the SSIM of
12174 each individual frame. When filename equals "-" the data is sent to
12178 The file printed if @var{stats_file} is selected, contains a sequence of
12179 key/value pairs of the form @var{key}:@var{value} for each compared
12182 A description of each shown parameter follows:
12186 sequential number of the input frame, starting from 1
12188 @item Y, U, V, R, G, B
12189 SSIM of the compared frames for the component specified by the suffix.
12192 SSIM of the compared frames for the whole frame.
12195 Same as above but in dB representation.
12200 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12201 [main][ref] ssim="stats_file=stats.log" [out]
12204 On this example the input file being processed is compared with the
12205 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12206 is stored in @file{stats.log}.
12208 Another example with both psnr and ssim at same time:
12210 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12215 Convert between different stereoscopic image formats.
12217 The filters accept the following options:
12221 Set stereoscopic image format of input.
12223 Available values for input image formats are:
12226 side by side parallel (left eye left, right eye right)
12229 side by side crosseye (right eye left, left eye right)
12232 side by side parallel with half width resolution
12233 (left eye left, right eye right)
12236 side by side crosseye with half width resolution
12237 (right eye left, left eye right)
12240 above-below (left eye above, right eye below)
12243 above-below (right eye above, left eye below)
12246 above-below with half height resolution
12247 (left eye above, right eye below)
12250 above-below with half height resolution
12251 (right eye above, left eye below)
12254 alternating frames (left eye first, right eye second)
12257 alternating frames (right eye first, left eye second)
12260 interleaved rows (left eye has top row, right eye starts on next row)
12263 interleaved rows (right eye has top row, left eye starts on next row)
12266 interleaved columns, left eye first
12269 interleaved columns, right eye first
12271 Default value is @samp{sbsl}.
12275 Set stereoscopic image format of output.
12279 side by side parallel (left eye left, right eye right)
12282 side by side crosseye (right eye left, left eye right)
12285 side by side parallel with half width resolution
12286 (left eye left, right eye right)
12289 side by side crosseye with half width resolution
12290 (right eye left, left eye right)
12293 above-below (left eye above, right eye below)
12296 above-below (right eye above, left eye below)
12299 above-below with half height resolution
12300 (left eye above, right eye below)
12303 above-below with half height resolution
12304 (right eye above, left eye below)
12307 alternating frames (left eye first, right eye second)
12310 alternating frames (right eye first, left eye second)
12313 interleaved rows (left eye has top row, right eye starts on next row)
12316 interleaved rows (right eye has top row, left eye starts on next row)
12319 anaglyph red/blue gray
12320 (red filter on left eye, blue filter on right eye)
12323 anaglyph red/green gray
12324 (red filter on left eye, green filter on right eye)
12327 anaglyph red/cyan gray
12328 (red filter on left eye, cyan filter on right eye)
12331 anaglyph red/cyan half colored
12332 (red filter on left eye, cyan filter on right eye)
12335 anaglyph red/cyan color
12336 (red filter on left eye, cyan filter on right eye)
12339 anaglyph red/cyan color optimized with the least squares projection of dubois
12340 (red filter on left eye, cyan filter on right eye)
12343 anaglyph green/magenta gray
12344 (green filter on left eye, magenta filter on right eye)
12347 anaglyph green/magenta half colored
12348 (green filter on left eye, magenta filter on right eye)
12351 anaglyph green/magenta colored
12352 (green filter on left eye, magenta filter on right eye)
12355 anaglyph green/magenta color optimized with the least squares projection of dubois
12356 (green filter on left eye, magenta filter on right eye)
12359 anaglyph yellow/blue gray
12360 (yellow filter on left eye, blue filter on right eye)
12363 anaglyph yellow/blue half colored
12364 (yellow filter on left eye, blue filter on right eye)
12367 anaglyph yellow/blue colored
12368 (yellow filter on left eye, blue filter on right eye)
12371 anaglyph yellow/blue color optimized with the least squares projection of dubois
12372 (yellow filter on left eye, blue filter on right eye)
12375 mono output (left eye only)
12378 mono output (right eye only)
12381 checkerboard, left eye first
12384 checkerboard, right eye first
12387 interleaved columns, left eye first
12390 interleaved columns, right eye first
12396 Default value is @samp{arcd}.
12399 @subsection Examples
12403 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12409 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12415 @section streamselect, astreamselect
12416 Select video or audio streams.
12418 The filter accepts the following options:
12422 Set number of inputs. Default is 2.
12425 Set input indexes to remap to outputs.
12428 @subsection Commands
12430 The @code{streamselect} and @code{astreamselect} filter supports the following
12435 Set input indexes to remap to outputs.
12438 @subsection Examples
12442 Select first 5 seconds 1st stream and rest of time 2nd stream:
12444 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12448 Same as above, but for audio:
12450 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12457 Apply a simple postprocessing filter that compresses and decompresses the image
12458 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12459 and average the results.
12461 The filter accepts the following options:
12465 Set quality. This option defines the number of levels for averaging. It accepts
12466 an integer in the range 0-6. If set to @code{0}, the filter will have no
12467 effect. A value of @code{6} means the higher quality. For each increment of
12468 that value the speed drops by a factor of approximately 2. Default value is
12472 Force a constant quantization parameter. If not set, the filter will use the QP
12473 from the video stream (if available).
12476 Set thresholding mode. Available modes are:
12480 Set hard thresholding (default).
12482 Set soft thresholding (better de-ringing effect, but likely blurrier).
12485 @item use_bframe_qp
12486 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12487 option may cause flicker since the B-Frames have often larger QP. Default is
12488 @code{0} (not enabled).
12494 Draw subtitles on top of input video using the libass library.
12496 To enable compilation of this filter you need to configure FFmpeg with
12497 @code{--enable-libass}. This filter also requires a build with libavcodec and
12498 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12499 Alpha) subtitles format.
12501 The filter accepts the following options:
12505 Set the filename of the subtitle file to read. It must be specified.
12507 @item original_size
12508 Specify the size of the original video, the video for which the ASS file
12509 was composed. For the syntax of this option, check the
12510 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12511 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12512 correctly scale the fonts if the aspect ratio has been changed.
12515 Set a directory path containing fonts that can be used by the filter.
12516 These fonts will be used in addition to whatever the font provider uses.
12519 Set subtitles input character encoding. @code{subtitles} filter only. Only
12520 useful if not UTF-8.
12522 @item stream_index, si
12523 Set subtitles stream index. @code{subtitles} filter only.
12526 Override default style or script info parameters of the subtitles. It accepts a
12527 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12530 If the first key is not specified, it is assumed that the first value
12531 specifies the @option{filename}.
12533 For example, to render the file @file{sub.srt} on top of the input
12534 video, use the command:
12539 which is equivalent to:
12541 subtitles=filename=sub.srt
12544 To render the default subtitles stream from file @file{video.mkv}, use:
12546 subtitles=video.mkv
12549 To render the second subtitles stream from that file, use:
12551 subtitles=video.mkv:si=1
12554 To make the subtitles stream from @file{sub.srt} appear in transparent green
12555 @code{DejaVu Serif}, use:
12557 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12560 @section super2xsai
12562 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12563 Interpolate) pixel art scaling algorithm.
12565 Useful for enlarging pixel art images without reducing sharpness.
12569 Swap two rectangular objects in video.
12571 This filter accepts the following options:
12581 Set 1st rect x coordinate.
12584 Set 1st rect y coordinate.
12587 Set 2nd rect x coordinate.
12590 Set 2nd rect y coordinate.
12592 All expressions are evaluated once for each frame.
12595 The all options are expressions containing the following constants:
12600 The input width and height.
12603 same as @var{w} / @var{h}
12606 input sample aspect ratio
12609 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12612 The number of the input frame, starting from 0.
12615 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12618 the position in the file of the input frame, NAN if unknown
12626 Apply telecine process to the video.
12628 This filter accepts the following options:
12637 The default value is @code{top}.
12641 A string of numbers representing the pulldown pattern you wish to apply.
12642 The default value is @code{23}.
12646 Some typical patterns:
12651 24p: 2332 (preferred)
12658 24p: 222222222223 ("Euro pulldown")
12664 Select the most representative frame in a given sequence of consecutive frames.
12666 The filter accepts the following options:
12670 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
12671 will pick one of them, and then handle the next batch of @var{n} frames until
12672 the end. Default is @code{100}.
12675 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
12676 value will result in a higher memory usage, so a high value is not recommended.
12678 @subsection Examples
12682 Extract one picture each 50 frames:
12688 Complete example of a thumbnail creation with @command{ffmpeg}:
12690 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
12696 Tile several successive frames together.
12698 The filter accepts the following options:
12703 Set the grid size (i.e. the number of lines and columns). For the syntax of
12704 this option, check the
12705 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12708 Set the maximum number of frames to render in the given area. It must be less
12709 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
12710 the area will be used.
12713 Set the outer border margin in pixels.
12716 Set the inner border thickness (i.e. the number of pixels between frames). For
12717 more advanced padding options (such as having different values for the edges),
12718 refer to the pad video filter.
12721 Specify the color of the unused area. For the syntax of this option, check the
12722 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
12726 @subsection Examples
12730 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
12732 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
12734 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
12735 duplicating each output frame to accommodate the originally detected frame
12739 Display @code{5} pictures in an area of @code{3x2} frames,
12740 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
12741 mixed flat and named options:
12743 tile=3x2:nb_frames=5:padding=7:margin=2
12747 @section tinterlace
12749 Perform various types of temporal field interlacing.
12751 Frames are counted starting from 1, so the first input frame is
12754 The filter accepts the following options:
12759 Specify the mode of the interlacing. This option can also be specified
12760 as a value alone. See below for a list of values for this option.
12762 Available values are:
12766 Move odd frames into the upper field, even into the lower field,
12767 generating a double height frame at half frame rate.
12771 Frame 1 Frame 2 Frame 3 Frame 4
12773 11111 22222 33333 44444
12774 11111 22222 33333 44444
12775 11111 22222 33333 44444
12776 11111 22222 33333 44444
12790 Only output odd frames, even frames are dropped, generating a frame with
12791 unchanged height at half frame rate.
12796 Frame 1 Frame 2 Frame 3 Frame 4
12798 11111 22222 33333 44444
12799 11111 22222 33333 44444
12800 11111 22222 33333 44444
12801 11111 22222 33333 44444
12811 Only output even frames, odd frames are dropped, generating a frame with
12812 unchanged height at half frame rate.
12817 Frame 1 Frame 2 Frame 3 Frame 4
12819 11111 22222 33333 44444
12820 11111 22222 33333 44444
12821 11111 22222 33333 44444
12822 11111 22222 33333 44444
12832 Expand each frame to full height, but pad alternate lines with black,
12833 generating a frame with double height at the same input frame rate.
12838 Frame 1 Frame 2 Frame 3 Frame 4
12840 11111 22222 33333 44444
12841 11111 22222 33333 44444
12842 11111 22222 33333 44444
12843 11111 22222 33333 44444
12846 11111 ..... 33333 .....
12847 ..... 22222 ..... 44444
12848 11111 ..... 33333 .....
12849 ..... 22222 ..... 44444
12850 11111 ..... 33333 .....
12851 ..... 22222 ..... 44444
12852 11111 ..... 33333 .....
12853 ..... 22222 ..... 44444
12857 @item interleave_top, 4
12858 Interleave the upper field from odd frames with the lower field from
12859 even frames, generating a frame with unchanged height at half frame rate.
12864 Frame 1 Frame 2 Frame 3 Frame 4
12866 11111<- 22222 33333<- 44444
12867 11111 22222<- 33333 44444<-
12868 11111<- 22222 33333<- 44444
12869 11111 22222<- 33333 44444<-
12879 @item interleave_bottom, 5
12880 Interleave the lower field from odd frames with the upper field from
12881 even frames, generating a frame with unchanged height at half frame rate.
12886 Frame 1 Frame 2 Frame 3 Frame 4
12888 11111 22222<- 33333 44444<-
12889 11111<- 22222 33333<- 44444
12890 11111 22222<- 33333 44444<-
12891 11111<- 22222 33333<- 44444
12901 @item interlacex2, 6
12902 Double frame rate with unchanged height. Frames are inserted each
12903 containing the second temporal field from the previous input frame and
12904 the first temporal field from the next input frame. This mode relies on
12905 the top_field_first flag. Useful for interlaced video displays with no
12906 field synchronisation.
12911 Frame 1 Frame 2 Frame 3 Frame 4
12913 11111 22222 33333 44444
12914 11111 22222 33333 44444
12915 11111 22222 33333 44444
12916 11111 22222 33333 44444
12919 11111 22222 22222 33333 33333 44444 44444
12920 11111 11111 22222 22222 33333 33333 44444
12921 11111 22222 22222 33333 33333 44444 44444
12922 11111 11111 22222 22222 33333 33333 44444
12927 Move odd frames into the upper field, even into the lower field,
12928 generating a double height frame at same frame rate.
12933 Frame 1 Frame 2 Frame 3 Frame 4
12935 11111 22222 33333 44444
12936 11111 22222 33333 44444
12937 11111 22222 33333 44444
12938 11111 22222 33333 44444
12941 11111 33333 33333 55555
12942 22222 22222 44444 44444
12943 11111 33333 33333 55555
12944 22222 22222 44444 44444
12945 11111 33333 33333 55555
12946 22222 22222 44444 44444
12947 11111 33333 33333 55555
12948 22222 22222 44444 44444
12953 Numeric values are deprecated but are accepted for backward
12954 compatibility reasons.
12956 Default mode is @code{merge}.
12959 Specify flags influencing the filter process.
12961 Available value for @var{flags} is:
12964 @item low_pass_filter, vlfp
12965 Enable vertical low-pass filtering in the filter.
12966 Vertical low-pass filtering is required when creating an interlaced
12967 destination from a progressive source which contains high-frequency
12968 vertical detail. Filtering will reduce interlace 'twitter' and Moire
12971 Vertical low-pass filtering can only be enabled for @option{mode}
12972 @var{interleave_top} and @var{interleave_bottom}.
12979 Transpose rows with columns in the input video and optionally flip it.
12981 It accepts the following parameters:
12986 Specify the transposition direction.
12988 Can assume the following values:
12990 @item 0, 4, cclock_flip
12991 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
12999 Rotate by 90 degrees clockwise, that is:
13007 Rotate by 90 degrees counterclockwise, that is:
13014 @item 3, 7, clock_flip
13015 Rotate by 90 degrees clockwise and vertically flip, that is:
13023 For values between 4-7, the transposition is only done if the input
13024 video geometry is portrait and not landscape. These values are
13025 deprecated, the @code{passthrough} option should be used instead.
13027 Numerical values are deprecated, and should be dropped in favor of
13028 symbolic constants.
13031 Do not apply the transposition if the input geometry matches the one
13032 specified by the specified value. It accepts the following values:
13035 Always apply transposition.
13037 Preserve portrait geometry (when @var{height} >= @var{width}).
13039 Preserve landscape geometry (when @var{width} >= @var{height}).
13042 Default value is @code{none}.
13045 For example to rotate by 90 degrees clockwise and preserve portrait
13048 transpose=dir=1:passthrough=portrait
13051 The command above can also be specified as:
13053 transpose=1:portrait
13057 Trim the input so that the output contains one continuous subpart of the input.
13059 It accepts the following parameters:
13062 Specify the time of the start of the kept section, i.e. the frame with the
13063 timestamp @var{start} will be the first frame in the output.
13066 Specify the time of the first frame that will be dropped, i.e. the frame
13067 immediately preceding the one with the timestamp @var{end} will be the last
13068 frame in the output.
13071 This is the same as @var{start}, except this option sets the start timestamp
13072 in timebase units instead of seconds.
13075 This is the same as @var{end}, except this option sets the end timestamp
13076 in timebase units instead of seconds.
13079 The maximum duration of the output in seconds.
13082 The number of the first frame that should be passed to the output.
13085 The number of the first frame that should be dropped.
13088 @option{start}, @option{end}, and @option{duration} are expressed as time
13089 duration specifications; see
13090 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13091 for the accepted syntax.
13093 Note that the first two sets of the start/end options and the @option{duration}
13094 option look at the frame timestamp, while the _frame variants simply count the
13095 frames that pass through the filter. Also note that this filter does not modify
13096 the timestamps. If you wish for the output timestamps to start at zero, insert a
13097 setpts filter after the trim filter.
13099 If multiple start or end options are set, this filter tries to be greedy and
13100 keep all the frames that match at least one of the specified constraints. To keep
13101 only the part that matches all the constraints at once, chain multiple trim
13104 The defaults are such that all the input is kept. So it is possible to set e.g.
13105 just the end values to keep everything before the specified time.
13110 Drop everything except the second minute of input:
13112 ffmpeg -i INPUT -vf trim=60:120
13116 Keep only the first second:
13118 ffmpeg -i INPUT -vf trim=duration=1
13127 Sharpen or blur the input video.
13129 It accepts the following parameters:
13132 @item luma_msize_x, lx
13133 Set the luma matrix horizontal size. It must be an odd integer between
13134 3 and 63. The default value is 5.
13136 @item luma_msize_y, ly
13137 Set the luma matrix vertical size. It must be an odd integer between 3
13138 and 63. The default value is 5.
13140 @item luma_amount, la
13141 Set the luma effect strength. It must be a floating point number, reasonable
13142 values lay between -1.5 and 1.5.
13144 Negative values will blur the input video, while positive values will
13145 sharpen it, a value of zero will disable the effect.
13147 Default value is 1.0.
13149 @item chroma_msize_x, cx
13150 Set the chroma matrix horizontal size. It must be an odd integer
13151 between 3 and 63. The default value is 5.
13153 @item chroma_msize_y, cy
13154 Set the chroma matrix vertical size. It must be an odd integer
13155 between 3 and 63. The default value is 5.
13157 @item chroma_amount, ca
13158 Set the chroma effect strength. It must be a floating point number, reasonable
13159 values lay between -1.5 and 1.5.
13161 Negative values will blur the input video, while positive values will
13162 sharpen it, a value of zero will disable the effect.
13164 Default value is 0.0.
13167 If set to 1, specify using OpenCL capabilities, only available if
13168 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13172 All parameters are optional and default to the equivalent of the
13173 string '5:5:1.0:5:5:0.0'.
13175 @subsection Examples
13179 Apply strong luma sharpen effect:
13181 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13185 Apply a strong blur of both luma and chroma parameters:
13187 unsharp=7:7:-2:7:7:-2
13193 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13194 the image at several (or - in the case of @option{quality} level @code{8} - all)
13195 shifts and average the results.
13197 The way this differs from the behavior of spp is that uspp actually encodes &
13198 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13199 DCT similar to MJPEG.
13201 The filter accepts the following options:
13205 Set quality. This option defines the number of levels for averaging. It accepts
13206 an integer in the range 0-8. If set to @code{0}, the filter will have no
13207 effect. A value of @code{8} means the higher quality. For each increment of
13208 that value the speed drops by a factor of approximately 2. Default value is
13212 Force a constant quantization parameter. If not set, the filter will use the QP
13213 from the video stream (if available).
13216 @section vectorscope
13218 Display 2 color component values in the two dimensional graph (which is called
13221 This filter accepts the following options:
13225 Set vectorscope mode.
13227 It accepts the following values:
13230 Gray values are displayed on graph, higher brightness means more pixels have
13231 same component color value on location in graph. This is the default mode.
13234 Gray values are displayed on graph. Surrounding pixels values which are not
13235 present in video frame are drawn in gradient of 2 color components which are
13236 set by option @code{x} and @code{y}. The 3rd color component is static.
13239 Actual color components values present in video frame are displayed on graph.
13242 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13243 on graph increases value of another color component, which is luminance by
13244 default values of @code{x} and @code{y}.
13247 Actual colors present in video frame are displayed on graph. If two different
13248 colors map to same position on graph then color with higher value of component
13249 not present in graph is picked.
13252 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13253 component picked from radial gradient.
13257 Set which color component will be represented on X-axis. Default is @code{1}.
13260 Set which color component will be represented on Y-axis. Default is @code{2}.
13263 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13264 of color component which represents frequency of (X, Y) location in graph.
13269 No envelope, this is default.
13272 Instant envelope, even darkest single pixel will be clearly highlighted.
13275 Hold maximum and minimum values presented in graph over time. This way you
13276 can still spot out of range values without constantly looking at vectorscope.
13279 Peak and instant envelope combined together.
13283 Set what kind of graticule to draw.
13291 Set graticule opacity.
13294 Set graticule flags.
13298 Draw graticule for white point.
13301 Draw graticule for black point.
13304 Draw color points short names.
13308 Set background opacity.
13310 @item lthreshold, l
13311 Set low threshold for color component not represented on X or Y axis.
13312 Values lower than this value will be ignored. Default is 0.
13313 Note this value is multiplied with actual max possible value one pixel component
13314 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13317 @item hthreshold, h
13318 Set high threshold for color component not represented on X or Y axis.
13319 Values higher than this value will be ignored. Default is 1.
13320 Note this value is multiplied with actual max possible value one pixel component
13321 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13322 is 0.9 * 255 = 230.
13324 @item colorspace, c
13325 Set what kind of colorspace to use when drawing graticule.
13334 @anchor{vidstabdetect}
13335 @section vidstabdetect
13337 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13338 @ref{vidstabtransform} for pass 2.
13340 This filter generates a file with relative translation and rotation
13341 transform information about subsequent frames, which is then used by
13342 the @ref{vidstabtransform} filter.
13344 To enable compilation of this filter you need to configure FFmpeg with
13345 @code{--enable-libvidstab}.
13347 This filter accepts the following options:
13351 Set the path to the file used to write the transforms information.
13352 Default value is @file{transforms.trf}.
13355 Set how shaky the video is and how quick the camera is. It accepts an
13356 integer in the range 1-10, a value of 1 means little shakiness, a
13357 value of 10 means strong shakiness. Default value is 5.
13360 Set the accuracy of the detection process. It must be a value in the
13361 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13362 accuracy. Default value is 15.
13365 Set stepsize of the search process. The region around minimum is
13366 scanned with 1 pixel resolution. Default value is 6.
13369 Set minimum contrast. Below this value a local measurement field is
13370 discarded. Must be a floating point value in the range 0-1. Default
13374 Set reference frame number for tripod mode.
13376 If enabled, the motion of the frames is compared to a reference frame
13377 in the filtered stream, identified by the specified number. The idea
13378 is to compensate all movements in a more-or-less static scene and keep
13379 the camera view absolutely still.
13381 If set to 0, it is disabled. The frames are counted starting from 1.
13384 Show fields and transforms in the resulting frames. It accepts an
13385 integer in the range 0-2. Default value is 0, which disables any
13389 @subsection Examples
13393 Use default values:
13399 Analyze strongly shaky movie and put the results in file
13400 @file{mytransforms.trf}:
13402 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13406 Visualize the result of internal transformations in the resulting
13409 vidstabdetect=show=1
13413 Analyze a video with medium shakiness using @command{ffmpeg}:
13415 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13419 @anchor{vidstabtransform}
13420 @section vidstabtransform
13422 Video stabilization/deshaking: pass 2 of 2,
13423 see @ref{vidstabdetect} for pass 1.
13425 Read a file with transform information for each frame and
13426 apply/compensate them. Together with the @ref{vidstabdetect}
13427 filter this can be used to deshake videos. See also
13428 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13429 the @ref{unsharp} filter, see below.
13431 To enable compilation of this filter you need to configure FFmpeg with
13432 @code{--enable-libvidstab}.
13434 @subsection Options
13438 Set path to the file used to read the transforms. Default value is
13439 @file{transforms.trf}.
13442 Set the number of frames (value*2 + 1) used for lowpass filtering the
13443 camera movements. Default value is 10.
13445 For example a number of 10 means that 21 frames are used (10 in the
13446 past and 10 in the future) to smoothen the motion in the video. A
13447 larger value leads to a smoother video, but limits the acceleration of
13448 the camera (pan/tilt movements). 0 is a special case where a static
13449 camera is simulated.
13452 Set the camera path optimization algorithm.
13454 Accepted values are:
13457 gaussian kernel low-pass filter on camera motion (default)
13459 averaging on transformations
13463 Set maximal number of pixels to translate frames. Default value is -1,
13467 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13468 value is -1, meaning no limit.
13471 Specify how to deal with borders that may be visible due to movement
13474 Available values are:
13477 keep image information from previous frame (default)
13479 fill the border black
13483 Invert transforms if set to 1. Default value is 0.
13486 Consider transforms as relative to previous frame if set to 1,
13487 absolute if set to 0. Default value is 0.
13490 Set percentage to zoom. A positive value will result in a zoom-in
13491 effect, a negative value in a zoom-out effect. Default value is 0 (no
13495 Set optimal zooming to avoid borders.
13497 Accepted values are:
13502 optimal static zoom value is determined (only very strong movements
13503 will lead to visible borders) (default)
13505 optimal adaptive zoom value is determined (no borders will be
13506 visible), see @option{zoomspeed}
13509 Note that the value given at zoom is added to the one calculated here.
13512 Set percent to zoom maximally each frame (enabled when
13513 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13517 Specify type of interpolation.
13519 Available values are:
13524 linear only horizontal
13526 linear in both directions (default)
13528 cubic in both directions (slow)
13532 Enable virtual tripod mode if set to 1, which is equivalent to
13533 @code{relative=0:smoothing=0}. Default value is 0.
13535 Use also @code{tripod} option of @ref{vidstabdetect}.
13538 Increase log verbosity if set to 1. Also the detected global motions
13539 are written to the temporary file @file{global_motions.trf}. Default
13543 @subsection Examples
13547 Use @command{ffmpeg} for a typical stabilization with default values:
13549 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13552 Note the use of the @ref{unsharp} filter which is always recommended.
13555 Zoom in a bit more and load transform data from a given file:
13557 vidstabtransform=zoom=5:input="mytransforms.trf"
13561 Smoothen the video even more:
13563 vidstabtransform=smoothing=30
13569 Flip the input video vertically.
13571 For example, to vertically flip a video with @command{ffmpeg}:
13573 ffmpeg -i in.avi -vf "vflip" out.avi
13579 Make or reverse a natural vignetting effect.
13581 The filter accepts the following options:
13585 Set lens angle expression as a number of radians.
13587 The value is clipped in the @code{[0,PI/2]} range.
13589 Default value: @code{"PI/5"}
13593 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
13597 Set forward/backward mode.
13599 Available modes are:
13602 The larger the distance from the central point, the darker the image becomes.
13605 The larger the distance from the central point, the brighter the image becomes.
13606 This can be used to reverse a vignette effect, though there is no automatic
13607 detection to extract the lens @option{angle} and other settings (yet). It can
13608 also be used to create a burning effect.
13611 Default value is @samp{forward}.
13614 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
13616 It accepts the following values:
13619 Evaluate expressions only once during the filter initialization.
13622 Evaluate expressions for each incoming frame. This is way slower than the
13623 @samp{init} mode since it requires all the scalers to be re-computed, but it
13624 allows advanced dynamic expressions.
13627 Default value is @samp{init}.
13630 Set dithering to reduce the circular banding effects. Default is @code{1}
13634 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
13635 Setting this value to the SAR of the input will make a rectangular vignetting
13636 following the dimensions of the video.
13638 Default is @code{1/1}.
13641 @subsection Expressions
13643 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
13644 following parameters.
13649 input width and height
13652 the number of input frame, starting from 0
13655 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
13656 @var{TB} units, NAN if undefined
13659 frame rate of the input video, NAN if the input frame rate is unknown
13662 the PTS (Presentation TimeStamp) of the filtered video frame,
13663 expressed in seconds, NAN if undefined
13666 time base of the input video
13670 @subsection Examples
13674 Apply simple strong vignetting effect:
13680 Make a flickering vignetting:
13682 vignette='PI/4+random(1)*PI/50':eval=frame
13688 Stack input videos vertically.
13690 All streams must be of same pixel format and of same width.
13692 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13693 to create same output.
13695 The filter accept the following option:
13699 Set number of input streams. Default is 2.
13702 If set to 1, force the output to terminate when the shortest input
13703 terminates. Default value is 0.
13708 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
13709 Deinterlacing Filter").
13711 Based on the process described by Martin Weston for BBC R&D, and
13712 implemented based on the de-interlace algorithm written by Jim
13713 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
13714 uses filter coefficients calculated by BBC R&D.
13716 There are two sets of filter coefficients, so called "simple":
13717 and "complex". Which set of filter coefficients is used can
13718 be set by passing an optional parameter:
13722 Set the interlacing filter coefficients. Accepts one of the following values:
13726 Simple filter coefficient set.
13728 More-complex filter coefficient set.
13730 Default value is @samp{complex}.
13733 Specify which frames to deinterlace. Accept one of the following values:
13737 Deinterlace all frames,
13739 Only deinterlace frames marked as interlaced.
13742 Default value is @samp{all}.
13746 Video waveform monitor.
13748 The waveform monitor plots color component intensity. By default luminance
13749 only. Each column of the waveform corresponds to a column of pixels in the
13752 It accepts the following options:
13756 Can be either @code{row}, or @code{column}. Default is @code{column}.
13757 In row mode, the graph on the left side represents color component value 0 and
13758 the right side represents value = 255. In column mode, the top side represents
13759 color component value = 0 and bottom side represents value = 255.
13762 Set intensity. Smaller values are useful to find out how many values of the same
13763 luminance are distributed across input rows/columns.
13764 Default value is @code{0.04}. Allowed range is [0, 1].
13767 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
13768 In mirrored mode, higher values will be represented on the left
13769 side for @code{row} mode and at the top for @code{column} mode. Default is
13770 @code{1} (mirrored).
13774 It accepts the following values:
13777 Presents information identical to that in the @code{parade}, except
13778 that the graphs representing color components are superimposed directly
13781 This display mode makes it easier to spot relative differences or similarities
13782 in overlapping areas of the color components that are supposed to be identical,
13783 such as neutral whites, grays, or blacks.
13786 Display separate graph for the color components side by side in
13787 @code{row} mode or one below the other in @code{column} mode.
13790 Display separate graph for the color components side by side in
13791 @code{column} mode or one below the other in @code{row} mode.
13793 Using this display mode makes it easy to spot color casts in the highlights
13794 and shadows of an image, by comparing the contours of the top and the bottom
13795 graphs of each waveform. Since whites, grays, and blacks are characterized
13796 by exactly equal amounts of red, green, and blue, neutral areas of the picture
13797 should display three waveforms of roughly equal width/height. If not, the
13798 correction is easy to perform by making level adjustments the three waveforms.
13800 Default is @code{stack}.
13802 @item components, c
13803 Set which color components to display. Default is 1, which means only luminance
13804 or red color component if input is in RGB colorspace. If is set for example to
13805 7 it will display all 3 (if) available color components.
13810 No envelope, this is default.
13813 Instant envelope, minimum and maximum values presented in graph will be easily
13814 visible even with small @code{step} value.
13817 Hold minimum and maximum values presented in graph across time. This way you
13818 can still spot out of range values without constantly looking at waveforms.
13821 Peak and instant envelope combined together.
13827 No filtering, this is default.
13830 Luma and chroma combined together.
13833 Similar as above, but shows difference between blue and red chroma.
13836 Displays only chroma.
13839 Displays actual color value on waveform.
13842 Similar as above, but with luma showing frequency of chroma values.
13846 Set which graticule to display.
13850 Do not display graticule.
13853 Display green graticule showing legal broadcast ranges.
13857 Set graticule opacity.
13860 Set graticule flags.
13864 Draw numbers above lines. By default enabled.
13867 Draw dots instead of lines.
13871 Set scale used for displaying graticule.
13878 Default is digital.
13882 Apply the xBR high-quality magnification filter which is designed for pixel
13883 art. It follows a set of edge-detection rules, see
13884 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
13886 It accepts the following option:
13890 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
13891 @code{3xBR} and @code{4} for @code{4xBR}.
13892 Default is @code{3}.
13898 Deinterlace the input video ("yadif" means "yet another deinterlacing
13901 It accepts the following parameters:
13907 The interlacing mode to adopt. It accepts one of the following values:
13910 @item 0, send_frame
13911 Output one frame for each frame.
13912 @item 1, send_field
13913 Output one frame for each field.
13914 @item 2, send_frame_nospatial
13915 Like @code{send_frame}, but it skips the spatial interlacing check.
13916 @item 3, send_field_nospatial
13917 Like @code{send_field}, but it skips the spatial interlacing check.
13920 The default value is @code{send_frame}.
13923 The picture field parity assumed for the input interlaced video. It accepts one
13924 of the following values:
13928 Assume the top field is first.
13930 Assume the bottom field is first.
13932 Enable automatic detection of field parity.
13935 The default value is @code{auto}.
13936 If the interlacing is unknown or the decoder does not export this information,
13937 top field first will be assumed.
13940 Specify which frames to deinterlace. Accept one of the following
13945 Deinterlace all frames.
13946 @item 1, interlaced
13947 Only deinterlace frames marked as interlaced.
13950 The default value is @code{all}.
13955 Apply Zoom & Pan effect.
13957 This filter accepts the following options:
13961 Set the zoom expression. Default is 1.
13965 Set the x and y expression. Default is 0.
13968 Set the duration expression in number of frames.
13969 This sets for how many number of frames effect will last for
13970 single input image.
13973 Set the output image size, default is 'hd720'.
13976 Set the output frame rate, default is '25'.
13979 Each expression can contain the following constants:
13998 Output frame count.
14002 Last calculated 'x' and 'y' position from 'x' and 'y' expression
14003 for current input frame.
14007 'x' and 'y' of last output frame of previous input frame or 0 when there was
14008 not yet such frame (first input frame).
14011 Last calculated zoom from 'z' expression for current input frame.
14014 Last calculated zoom of last output frame of previous input frame.
14017 Number of output frames for current input frame. Calculated from 'd' expression
14018 for each input frame.
14021 number of output frames created for previous input frame
14024 Rational number: input width / input height
14027 sample aspect ratio
14030 display aspect ratio
14034 @subsection Examples
14038 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
14040 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
14044 Zoom-in up to 1.5 and pan always at center of picture:
14046 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14050 Same as above but without pausing:
14052 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14057 Scale (resize) the input video, using the z.lib library:
14058 https://github.com/sekrit-twc/zimg.
14060 The zscale filter forces the output display aspect ratio to be the same
14061 as the input, by changing the output sample aspect ratio.
14063 If the input image format is different from the format requested by
14064 the next filter, the zscale filter will convert the input to the
14067 @subsection Options
14068 The filter accepts the following options.
14073 Set the output video dimension expression. Default value is the input
14076 If the @var{width} or @var{w} is 0, the input width is used for the output.
14077 If the @var{height} or @var{h} is 0, the input height is used for the output.
14079 If one of the values is -1, the zscale filter will use a value that
14080 maintains the aspect ratio of the input image, calculated from the
14081 other specified dimension. If both of them are -1, the input size is
14084 If one of the values is -n with n > 1, the zscale filter will also use a value
14085 that maintains the aspect ratio of the input image, calculated from the other
14086 specified dimension. After that it will, however, make sure that the calculated
14087 dimension is divisible by n and adjust the value if necessary.
14089 See below for the list of accepted constants for use in the dimension
14093 Set the video size. For the syntax of this option, check the
14094 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14097 Set the dither type.
14099 Possible values are:
14104 @item error_diffusion
14110 Set the resize filter type.
14112 Possible values are:
14122 Default is bilinear.
14125 Set the color range.
14127 Possible values are:
14134 Default is same as input.
14137 Set the color primaries.
14139 Possible values are:
14149 Default is same as input.
14152 Set the transfer characteristics.
14154 Possible values are:
14165 Default is same as input.
14168 Set the colorspace matrix.
14170 Possible value are:
14181 Default is same as input.
14184 Set the input color range.
14186 Possible values are:
14193 Default is same as input.
14195 @item primariesin, pin
14196 Set the input color primaries.
14198 Possible values are:
14208 Default is same as input.
14210 @item transferin, tin
14211 Set the input transfer characteristics.
14213 Possible values are:
14224 Default is same as input.
14226 @item matrixin, min
14227 Set the input colorspace matrix.
14229 Possible value are:
14241 The values of the @option{w} and @option{h} options are expressions
14242 containing the following constants:
14247 The input width and height
14251 These are the same as @var{in_w} and @var{in_h}.
14255 The output (scaled) width and height
14259 These are the same as @var{out_w} and @var{out_h}
14262 The same as @var{iw} / @var{ih}
14265 input sample aspect ratio
14268 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14272 horizontal and vertical input chroma subsample values. For example for the
14273 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14277 horizontal and vertical output chroma subsample values. For example for the
14278 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14284 @c man end VIDEO FILTERS
14286 @chapter Video Sources
14287 @c man begin VIDEO SOURCES
14289 Below is a description of the currently available video sources.
14293 Buffer video frames, and make them available to the filter chain.
14295 This source is mainly intended for a programmatic use, in particular
14296 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14298 It accepts the following parameters:
14303 Specify the size (width and height) of the buffered video frames. For the
14304 syntax of this option, check the
14305 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14308 The input video width.
14311 The input video height.
14314 A string representing the pixel format of the buffered video frames.
14315 It may be a number corresponding to a pixel format, or a pixel format
14319 Specify the timebase assumed by the timestamps of the buffered frames.
14322 Specify the frame rate expected for the video stream.
14324 @item pixel_aspect, sar
14325 The sample (pixel) aspect ratio of the input video.
14328 Specify the optional parameters to be used for the scale filter which
14329 is automatically inserted when an input change is detected in the
14330 input size or format.
14332 @item hw_frames_ctx
14333 When using a hardware pixel format, this should be a reference to an
14334 AVHWFramesContext describing input frames.
14339 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14342 will instruct the source to accept video frames with size 320x240 and
14343 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14344 square pixels (1:1 sample aspect ratio).
14345 Since the pixel format with name "yuv410p" corresponds to the number 6
14346 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14347 this example corresponds to:
14349 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14352 Alternatively, the options can be specified as a flat string, but this
14353 syntax is deprecated:
14355 @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}]
14359 Create a pattern generated by an elementary cellular automaton.
14361 The initial state of the cellular automaton can be defined through the
14362 @option{filename}, and @option{pattern} options. If such options are
14363 not specified an initial state is created randomly.
14365 At each new frame a new row in the video is filled with the result of
14366 the cellular automaton next generation. The behavior when the whole
14367 frame is filled is defined by the @option{scroll} option.
14369 This source accepts the following options:
14373 Read the initial cellular automaton state, i.e. the starting row, from
14374 the specified file.
14375 In the file, each non-whitespace character is considered an alive
14376 cell, a newline will terminate the row, and further characters in the
14377 file will be ignored.
14380 Read the initial cellular automaton state, i.e. the starting row, from
14381 the specified string.
14383 Each non-whitespace character in the string is considered an alive
14384 cell, a newline will terminate the row, and further characters in the
14385 string will be ignored.
14388 Set the video rate, that is the number of frames generated per second.
14391 @item random_fill_ratio, ratio
14392 Set the random fill ratio for the initial cellular automaton row. It
14393 is a floating point number value ranging from 0 to 1, defaults to
14396 This option is ignored when a file or a pattern is specified.
14398 @item random_seed, seed
14399 Set the seed for filling randomly the initial row, must be an integer
14400 included between 0 and UINT32_MAX. If not specified, or if explicitly
14401 set to -1, the filter will try to use a good random seed on a best
14405 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14406 Default value is 110.
14409 Set the size of the output video. For the syntax of this option, check the
14410 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14412 If @option{filename} or @option{pattern} is specified, the size is set
14413 by default to the width of the specified initial state row, and the
14414 height is set to @var{width} * PHI.
14416 If @option{size} is set, it must contain the width of the specified
14417 pattern string, and the specified pattern will be centered in the
14420 If a filename or a pattern string is not specified, the size value
14421 defaults to "320x518" (used for a randomly generated initial state).
14424 If set to 1, scroll the output upward when all the rows in the output
14425 have been already filled. If set to 0, the new generated row will be
14426 written over the top row just after the bottom row is filled.
14429 @item start_full, full
14430 If set to 1, completely fill the output with generated rows before
14431 outputting the first frame.
14432 This is the default behavior, for disabling set the value to 0.
14435 If set to 1, stitch the left and right row edges together.
14436 This is the default behavior, for disabling set the value to 0.
14439 @subsection Examples
14443 Read the initial state from @file{pattern}, and specify an output of
14446 cellauto=f=pattern:s=200x400
14450 Generate a random initial row with a width of 200 cells, with a fill
14453 cellauto=ratio=2/3:s=200x200
14457 Create a pattern generated by rule 18 starting by a single alive cell
14458 centered on an initial row with width 100:
14460 cellauto=p=@@:s=100x400:full=0:rule=18
14464 Specify a more elaborated initial pattern:
14466 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14471 @anchor{coreimagesrc}
14472 @section coreimagesrc
14473 Video source generated on GPU using Apple's CoreImage API on OSX.
14475 This video source is a specialized version of the @ref{coreimage} video filter.
14476 Use a core image generator at the beginning of the applied filterchain to
14477 generate the content.
14479 The coreimagesrc video source accepts the following options:
14481 @item list_generators
14482 List all available generators along with all their respective options as well as
14483 possible minimum and maximum values along with the default values.
14485 list_generators=true
14489 Specify the size of the sourced video. For the syntax of this option, check the
14490 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14491 The default value is @code{320x240}.
14494 Specify the frame rate of the sourced video, as the number of frames
14495 generated per second. It has to be a string in the format
14496 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14497 number or a valid video frame rate abbreviation. The default value is
14501 Set the sample aspect ratio of the sourced video.
14504 Set the duration of the sourced video. See
14505 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14506 for the accepted syntax.
14508 If not specified, or the expressed duration is negative, the video is
14509 supposed to be generated forever.
14512 Additionally, all options of the @ref{coreimage} video filter are accepted.
14513 A complete filterchain can be used for further processing of the
14514 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
14515 and examples for details.
14517 @subsection Examples
14522 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
14523 given as complete and escaped command-line for Apple's standard bash shell:
14525 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
14527 This example is equivalent to the QRCode example of @ref{coreimage} without the
14528 need for a nullsrc video source.
14532 @section mandelbrot
14534 Generate a Mandelbrot set fractal, and progressively zoom towards the
14535 point specified with @var{start_x} and @var{start_y}.
14537 This source accepts the following options:
14542 Set the terminal pts value. Default value is 400.
14545 Set the terminal scale value.
14546 Must be a floating point value. Default value is 0.3.
14549 Set the inner coloring mode, that is the algorithm used to draw the
14550 Mandelbrot fractal internal region.
14552 It shall assume one of the following values:
14557 Show time until convergence.
14559 Set color based on point closest to the origin of the iterations.
14564 Default value is @var{mincol}.
14567 Set the bailout value. Default value is 10.0.
14570 Set the maximum of iterations performed by the rendering
14571 algorithm. Default value is 7189.
14574 Set outer coloring mode.
14575 It shall assume one of following values:
14577 @item iteration_count
14578 Set iteration cound mode.
14579 @item normalized_iteration_count
14580 set normalized iteration count mode.
14582 Default value is @var{normalized_iteration_count}.
14585 Set frame rate, expressed as number of frames per second. Default
14589 Set frame size. For the syntax of this option, check the "Video
14590 size" section in the ffmpeg-utils manual. Default value is "640x480".
14593 Set the initial scale value. Default value is 3.0.
14596 Set the initial x position. Must be a floating point value between
14597 -100 and 100. Default value is -0.743643887037158704752191506114774.
14600 Set the initial y position. Must be a floating point value between
14601 -100 and 100. Default value is -0.131825904205311970493132056385139.
14606 Generate various test patterns, as generated by the MPlayer test filter.
14608 The size of the generated video is fixed, and is 256x256.
14609 This source is useful in particular for testing encoding features.
14611 This source accepts the following options:
14616 Specify the frame rate of the sourced video, as the number of frames
14617 generated per second. It has to be a string in the format
14618 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14619 number or a valid video frame rate abbreviation. The default value is
14623 Set the duration of the sourced video. See
14624 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14625 for the accepted syntax.
14627 If not specified, or the expressed duration is negative, the video is
14628 supposed to be generated forever.
14632 Set the number or the name of the test to perform. Supported tests are:
14648 Default value is "all", which will cycle through the list of all tests.
14653 mptestsrc=t=dc_luma
14656 will generate a "dc_luma" test pattern.
14658 @section frei0r_src
14660 Provide a frei0r source.
14662 To enable compilation of this filter you need to install the frei0r
14663 header and configure FFmpeg with @code{--enable-frei0r}.
14665 This source accepts the following parameters:
14670 The size of the video to generate. For the syntax of this option, check the
14671 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14674 The framerate of the generated video. It may be a string of the form
14675 @var{num}/@var{den} or a frame rate abbreviation.
14678 The name to the frei0r source to load. For more information regarding frei0r and
14679 how to set the parameters, read the @ref{frei0r} section in the video filters
14682 @item filter_params
14683 A '|'-separated list of parameters to pass to the frei0r source.
14687 For example, to generate a frei0r partik0l source with size 200x200
14688 and frame rate 10 which is overlaid on the overlay filter main input:
14690 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
14695 Generate a life pattern.
14697 This source is based on a generalization of John Conway's life game.
14699 The sourced input represents a life grid, each pixel represents a cell
14700 which can be in one of two possible states, alive or dead. Every cell
14701 interacts with its eight neighbours, which are the cells that are
14702 horizontally, vertically, or diagonally adjacent.
14704 At each interaction the grid evolves according to the adopted rule,
14705 which specifies the number of neighbor alive cells which will make a
14706 cell stay alive or born. The @option{rule} option allows one to specify
14709 This source accepts the following options:
14713 Set the file from which to read the initial grid state. In the file,
14714 each non-whitespace character is considered an alive cell, and newline
14715 is used to delimit the end of each row.
14717 If this option is not specified, the initial grid is generated
14721 Set the video rate, that is the number of frames generated per second.
14724 @item random_fill_ratio, ratio
14725 Set the random fill ratio for the initial random grid. It is a
14726 floating point number value ranging from 0 to 1, defaults to 1/PHI.
14727 It is ignored when a file is specified.
14729 @item random_seed, seed
14730 Set the seed for filling the initial random grid, must be an integer
14731 included between 0 and UINT32_MAX. If not specified, or if explicitly
14732 set to -1, the filter will try to use a good random seed on a best
14738 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
14739 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
14740 @var{NS} specifies the number of alive neighbor cells which make a
14741 live cell stay alive, and @var{NB} the number of alive neighbor cells
14742 which make a dead cell to become alive (i.e. to "born").
14743 "s" and "b" can be used in place of "S" and "B", respectively.
14745 Alternatively a rule can be specified by an 18-bits integer. The 9
14746 high order bits are used to encode the next cell state if it is alive
14747 for each number of neighbor alive cells, the low order bits specify
14748 the rule for "borning" new cells. Higher order bits encode for an
14749 higher number of neighbor cells.
14750 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
14751 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
14753 Default value is "S23/B3", which is the original Conway's game of life
14754 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
14755 cells, and will born a new cell if there are three alive cells around
14759 Set the size of the output video. For the syntax of this option, check the
14760 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14762 If @option{filename} is specified, the size is set by default to the
14763 same size of the input file. If @option{size} is set, it must contain
14764 the size specified in the input file, and the initial grid defined in
14765 that file is centered in the larger resulting area.
14767 If a filename is not specified, the size value defaults to "320x240"
14768 (used for a randomly generated initial grid).
14771 If set to 1, stitch the left and right grid edges together, and the
14772 top and bottom edges also. Defaults to 1.
14775 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
14776 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
14777 value from 0 to 255.
14780 Set the color of living (or new born) cells.
14783 Set the color of dead cells. If @option{mold} is set, this is the first color
14784 used to represent a dead cell.
14787 Set mold color, for definitely dead and moldy cells.
14789 For the syntax of these 3 color options, check the "Color" section in the
14790 ffmpeg-utils manual.
14793 @subsection Examples
14797 Read a grid from @file{pattern}, and center it on a grid of size
14800 life=f=pattern:s=300x300
14804 Generate a random grid of size 200x200, with a fill ratio of 2/3:
14806 life=ratio=2/3:s=200x200
14810 Specify a custom rule for evolving a randomly generated grid:
14816 Full example with slow death effect (mold) using @command{ffplay}:
14818 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
14825 @anchor{haldclutsrc}
14827 @anchor{rgbtestsrc}
14829 @anchor{smptehdbars}
14832 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2
14834 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
14836 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
14838 The @code{color} source provides an uniformly colored input.
14840 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
14841 @ref{haldclut} filter.
14843 The @code{nullsrc} source returns unprocessed video frames. It is
14844 mainly useful to be employed in analysis / debugging tools, or as the
14845 source for filters which ignore the input data.
14847 The @code{rgbtestsrc} source generates an RGB test pattern useful for
14848 detecting RGB vs BGR issues. You should see a red, green and blue
14849 stripe from top to bottom.
14851 The @code{smptebars} source generates a color bars pattern, based on
14852 the SMPTE Engineering Guideline EG 1-1990.
14854 The @code{smptehdbars} source generates a color bars pattern, based on
14855 the SMPTE RP 219-2002.
14857 The @code{testsrc} source generates a test video pattern, showing a
14858 color pattern, a scrolling gradient and a timestamp. This is mainly
14859 intended for testing purposes.
14861 The @code{testsrc2} source is similar to testsrc, but supports more
14862 pixel formats instead of just @code{rgb24}. This allows using it as an
14863 input for other tests without requiring a format conversion.
14865 The sources accept the following parameters:
14870 Specify the color of the source, only available in the @code{color}
14871 source. For the syntax of this option, check the "Color" section in the
14872 ffmpeg-utils manual.
14875 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
14876 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
14877 pixels to be used as identity matrix for 3D lookup tables. Each component is
14878 coded on a @code{1/(N*N)} scale.
14881 Specify the size of the sourced video. For the syntax of this option, check the
14882 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14883 The default value is @code{320x240}.
14885 This option is not available with the @code{haldclutsrc} filter.
14888 Specify the frame rate of the sourced video, as the number of frames
14889 generated per second. It has to be a string in the format
14890 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14891 number or a valid video frame rate abbreviation. The default value is
14895 Set the sample aspect ratio of the sourced video.
14898 Set the duration of the sourced video. See
14899 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14900 for the accepted syntax.
14902 If not specified, or the expressed duration is negative, the video is
14903 supposed to be generated forever.
14906 Set the number of decimals to show in the timestamp, only available in the
14907 @code{testsrc} source.
14909 The displayed timestamp value will correspond to the original
14910 timestamp value multiplied by the power of 10 of the specified
14911 value. Default value is 0.
14914 For example the following:
14916 testsrc=duration=5.3:size=qcif:rate=10
14919 will generate a video with a duration of 5.3 seconds, with size
14920 176x144 and a frame rate of 10 frames per second.
14922 The following graph description will generate a red source
14923 with an opacity of 0.2, with size "qcif" and a frame rate of 10
14926 color=c=red@@0.2:s=qcif:r=10
14929 If the input content is to be ignored, @code{nullsrc} can be used. The
14930 following command generates noise in the luminance plane by employing
14931 the @code{geq} filter:
14933 nullsrc=s=256x256, geq=random(1)*255:128:128
14936 @subsection Commands
14938 The @code{color} source supports the following commands:
14942 Set the color of the created image. Accepts the same syntax of the
14943 corresponding @option{color} option.
14946 @c man end VIDEO SOURCES
14948 @chapter Video Sinks
14949 @c man begin VIDEO SINKS
14951 Below is a description of the currently available video sinks.
14953 @section buffersink
14955 Buffer video frames, and make them available to the end of the filter
14958 This sink is mainly intended for programmatic use, in particular
14959 through the interface defined in @file{libavfilter/buffersink.h}
14960 or the options system.
14962 It accepts a pointer to an AVBufferSinkContext structure, which
14963 defines the incoming buffers' formats, to be passed as the opaque
14964 parameter to @code{avfilter_init_filter} for initialization.
14968 Null video sink: do absolutely nothing with the input video. It is
14969 mainly useful as a template and for use in analysis / debugging
14972 @c man end VIDEO SINKS
14974 @chapter Multimedia Filters
14975 @c man begin MULTIMEDIA FILTERS
14977 Below is a description of the currently available multimedia filters.
14979 @section ahistogram
14981 Convert input audio to a video output, displaying the volume histogram.
14983 The filter accepts the following options:
14987 Specify how histogram is calculated.
14989 It accepts the following values:
14992 Use single histogram for all channels.
14994 Use separate histogram for each channel.
14996 Default is @code{single}.
14999 Set frame rate, expressed as number of frames per second. Default
15003 Specify the video size for the output. For the syntax of this option, check the
15004 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15005 Default value is @code{hd720}.
15010 It accepts the following values:
15021 reverse logarithmic
15023 Default is @code{log}.
15026 Set amplitude scale.
15028 It accepts the following values:
15035 Default is @code{log}.
15038 Set how much frames to accumulate in histogram.
15039 Defauls is 1. Setting this to -1 accumulates all frames.
15042 Set histogram ratio of window height.
15045 Set sonogram sliding.
15047 It accepts the following values:
15050 replace old rows with new ones.
15052 scroll from top to bottom.
15054 Default is @code{replace}.
15057 @section aphasemeter
15059 Convert input audio to a video output, displaying the audio phase.
15061 The filter accepts the following options:
15065 Set the output frame rate. Default value is @code{25}.
15068 Set the video size for the output. For the syntax of this option, check the
15069 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15070 Default value is @code{800x400}.
15075 Specify the red, green, blue contrast. Default values are @code{2},
15076 @code{7} and @code{1}.
15077 Allowed range is @code{[0, 255]}.
15080 Set color which will be used for drawing median phase. If color is
15081 @code{none} which is default, no median phase value will be drawn.
15084 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15085 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15086 The @code{-1} means left and right channels are completely out of phase and
15087 @code{1} means channels are in phase.
15089 @section avectorscope
15091 Convert input audio to a video output, representing the audio vector
15094 The filter is used to measure the difference between channels of stereo
15095 audio stream. A monoaural signal, consisting of identical left and right
15096 signal, results in straight vertical line. Any stereo separation is visible
15097 as a deviation from this line, creating a Lissajous figure.
15098 If the straight (or deviation from it) but horizontal line appears this
15099 indicates that the left and right channels are out of phase.
15101 The filter accepts the following options:
15105 Set the vectorscope mode.
15107 Available values are:
15110 Lissajous rotated by 45 degrees.
15113 Same as above but not rotated.
15116 Shape resembling half of circle.
15119 Default value is @samp{lissajous}.
15122 Set the video size for the output. For the syntax of this option, check the
15123 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15124 Default value is @code{400x400}.
15127 Set the output frame rate. Default value is @code{25}.
15133 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15134 @code{160}, @code{80} and @code{255}.
15135 Allowed range is @code{[0, 255]}.
15141 Specify the red, green, blue and alpha fade. Default values are @code{15},
15142 @code{10}, @code{5} and @code{5}.
15143 Allowed range is @code{[0, 255]}.
15146 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15149 Set the vectorscope drawing mode.
15151 Available values are:
15154 Draw dot for each sample.
15157 Draw line between previous and current sample.
15160 Default value is @samp{dot}.
15163 Specify amplitude scale of audio samples.
15165 Available values are:
15182 @subsection Examples
15186 Complete example using @command{ffplay}:
15188 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15189 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15193 @section bench, abench
15195 Benchmark part of a filtergraph.
15197 The filter accepts the following options:
15201 Start or stop a timer.
15203 Available values are:
15206 Get the current time, set it as frame metadata (using the key
15207 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15210 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15211 the input frame metadata to get the time difference. Time difference, average,
15212 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15213 @code{min}) are then printed. The timestamps are expressed in seconds.
15217 @subsection Examples
15221 Benchmark @ref{selectivecolor} filter:
15223 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15229 Concatenate audio and video streams, joining them together one after the
15232 The filter works on segments of synchronized video and audio streams. All
15233 segments must have the same number of streams of each type, and that will
15234 also be the number of streams at output.
15236 The filter accepts the following options:
15241 Set the number of segments. Default is 2.
15244 Set the number of output video streams, that is also the number of video
15245 streams in each segment. Default is 1.
15248 Set the number of output audio streams, that is also the number of audio
15249 streams in each segment. Default is 0.
15252 Activate unsafe mode: do not fail if segments have a different format.
15256 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15257 @var{a} audio outputs.
15259 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15260 segment, in the same order as the outputs, then the inputs for the second
15263 Related streams do not always have exactly the same duration, for various
15264 reasons including codec frame size or sloppy authoring. For that reason,
15265 related synchronized streams (e.g. a video and its audio track) should be
15266 concatenated at once. The concat filter will use the duration of the longest
15267 stream in each segment (except the last one), and if necessary pad shorter
15268 audio streams with silence.
15270 For this filter to work correctly, all segments must start at timestamp 0.
15272 All corresponding streams must have the same parameters in all segments; the
15273 filtering system will automatically select a common pixel format for video
15274 streams, and a common sample format, sample rate and channel layout for
15275 audio streams, but other settings, such as resolution, must be converted
15276 explicitly by the user.
15278 Different frame rates are acceptable but will result in variable frame rate
15279 at output; be sure to configure the output file to handle it.
15281 @subsection Examples
15285 Concatenate an opening, an episode and an ending, all in bilingual version
15286 (video in stream 0, audio in streams 1 and 2):
15288 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15289 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15290 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15291 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15295 Concatenate two parts, handling audio and video separately, using the
15296 (a)movie sources, and adjusting the resolution:
15298 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15299 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15300 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15302 Note that a desync will happen at the stitch if the audio and video streams
15303 do not have exactly the same duration in the first file.
15307 @section drawgraph, adrawgraph
15309 Draw a graph using input video or audio metadata.
15311 It accepts the following parameters:
15315 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15318 Set 1st foreground color expression.
15321 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15324 Set 2nd foreground color expression.
15327 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15330 Set 3rd foreground color expression.
15333 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15336 Set 4th foreground color expression.
15339 Set minimal value of metadata value.
15342 Set maximal value of metadata value.
15345 Set graph background color. Default is white.
15350 Available values for mode is:
15357 Default is @code{line}.
15362 Available values for slide is:
15365 Draw new frame when right border is reached.
15368 Replace old columns with new ones.
15371 Scroll from right to left.
15374 Scroll from left to right.
15377 Draw single picture.
15380 Default is @code{frame}.
15383 Set size of graph video. For the syntax of this option, check the
15384 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15385 The default value is @code{900x256}.
15387 The foreground color expressions can use the following variables:
15390 Minimal value of metadata value.
15393 Maximal value of metadata value.
15396 Current metadata key value.
15399 The color is defined as 0xAABBGGRR.
15402 Example using metadata from @ref{signalstats} filter:
15404 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15407 Example using metadata from @ref{ebur128} filter:
15409 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15415 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15416 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15417 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15418 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15420 The filter also has a video output (see the @var{video} option) with a real
15421 time graph to observe the loudness evolution. The graphic contains the logged
15422 message mentioned above, so it is not printed anymore when this option is set,
15423 unless the verbose logging is set. The main graphing area contains the
15424 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15425 the momentary loudness (400 milliseconds).
15427 More information about the Loudness Recommendation EBU R128 on
15428 @url{http://tech.ebu.ch/loudness}.
15430 The filter accepts the following options:
15435 Activate the video output. The audio stream is passed unchanged whether this
15436 option is set or no. The video stream will be the first output stream if
15437 activated. Default is @code{0}.
15440 Set the video size. This option is for video only. For the syntax of this
15442 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15443 Default and minimum resolution is @code{640x480}.
15446 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15447 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15448 other integer value between this range is allowed.
15451 Set metadata injection. If set to @code{1}, the audio input will be segmented
15452 into 100ms output frames, each of them containing various loudness information
15453 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15455 Default is @code{0}.
15458 Force the frame logging level.
15460 Available values are:
15463 information logging level
15465 verbose logging level
15468 By default, the logging level is set to @var{info}. If the @option{video} or
15469 the @option{metadata} options are set, it switches to @var{verbose}.
15474 Available modes can be cumulated (the option is a @code{flag} type). Possible
15478 Disable any peak mode (default).
15480 Enable sample-peak mode.
15482 Simple peak mode looking for the higher sample value. It logs a message
15483 for sample-peak (identified by @code{SPK}).
15485 Enable true-peak mode.
15487 If enabled, the peak lookup is done on an over-sampled version of the input
15488 stream for better peak accuracy. It logs a message for true-peak.
15489 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15490 This mode requires a build with @code{libswresample}.
15494 Treat mono input files as "dual mono". If a mono file is intended for playback
15495 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15496 If set to @code{true}, this option will compensate for this effect.
15497 Multi-channel input files are not affected by this option.
15500 Set a specific pan law to be used for the measurement of dual mono files.
15501 This parameter is optional, and has a default value of -3.01dB.
15504 @subsection Examples
15508 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
15510 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
15514 Run an analysis with @command{ffmpeg}:
15516 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
15520 @section interleave, ainterleave
15522 Temporally interleave frames from several inputs.
15524 @code{interleave} works with video inputs, @code{ainterleave} with audio.
15526 These filters read frames from several inputs and send the oldest
15527 queued frame to the output.
15529 Input streams must have a well defined, monotonically increasing frame
15532 In order to submit one frame to output, these filters need to enqueue
15533 at least one frame for each input, so they cannot work in case one
15534 input is not yet terminated and will not receive incoming frames.
15536 For example consider the case when one input is a @code{select} filter
15537 which always drop input frames. The @code{interleave} filter will keep
15538 reading from that input, but it will never be able to send new frames
15539 to output until the input will send an end-of-stream signal.
15541 Also, depending on inputs synchronization, the filters will drop
15542 frames in case one input receives more frames than the other ones, and
15543 the queue is already filled.
15545 These filters accept the following options:
15549 Set the number of different inputs, it is 2 by default.
15552 @subsection Examples
15556 Interleave frames belonging to different streams using @command{ffmpeg}:
15558 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
15562 Add flickering blur effect:
15564 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
15568 @section metadata, ametadata
15570 Manipulate frame metadata.
15572 This filter accepts the following options:
15576 Set mode of operation of the filter.
15578 Can be one of the following:
15582 If both @code{value} and @code{key} is set, select frames
15583 which have such metadata. If only @code{key} is set, select
15584 every frame that has such key in metadata.
15587 Add new metadata @code{key} and @code{value}. If key is already available
15591 Modify value of already present key.
15594 If @code{value} is set, delete only keys that have such value.
15595 Otherwise, delete key.
15598 Print key and its value if metadata was found. If @code{key} is not set print all
15599 metadata values available in frame.
15603 Set key used with all modes. Must be set for all modes except @code{print}.
15606 Set metadata value which will be used. This option is mandatory for
15607 @code{modify} and @code{add} mode.
15610 Which function to use when comparing metadata value and @code{value}.
15612 Can be one of following:
15616 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
15619 Values are interpreted as strings, returns true if metadata value starts with
15620 the @code{value} option string.
15623 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
15626 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
15629 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
15632 Values are interpreted as floats, returns true if expression from option @code{expr}
15637 Set expression which is used when @code{function} is set to @code{expr}.
15638 The expression is evaluated through the eval API and can contain the following
15643 Float representation of @code{value} from metadata key.
15646 Float representation of @code{value} as supplied by user in @code{value} option.
15649 If specified in @code{print} mode, output is written to the named file. Instead of
15650 plain filename any writable url can be specified. Filename ``-'' is a shorthand
15651 for standard output. If @code{file} option is not set, output is written to the log
15652 with AV_LOG_INFO loglevel.
15657 @subsection Examples
15661 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
15664 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
15667 Print silencedetect output to file @file{metadata.txt}.
15669 silencedetect,ametadata=mode=print:file=metadata.txt
15672 Direct all metadata to a pipe with file descriptor 4.
15674 metadata=mode=print:file='pipe\:4'
15678 @section perms, aperms
15680 Set read/write permissions for the output frames.
15682 These filters are mainly aimed at developers to test direct path in the
15683 following filter in the filtergraph.
15685 The filters accept the following options:
15689 Select the permissions mode.
15691 It accepts the following values:
15694 Do nothing. This is the default.
15696 Set all the output frames read-only.
15698 Set all the output frames directly writable.
15700 Make the frame read-only if writable, and writable if read-only.
15702 Set each output frame read-only or writable randomly.
15706 Set the seed for the @var{random} mode, must be an integer included between
15707 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
15708 @code{-1}, the filter will try to use a good random seed on a best effort
15712 Note: in case of auto-inserted filter between the permission filter and the
15713 following one, the permission might not be received as expected in that
15714 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
15715 perms/aperms filter can avoid this problem.
15717 @section realtime, arealtime
15719 Slow down filtering to match real time approximatively.
15721 These filters will pause the filtering for a variable amount of time to
15722 match the output rate with the input timestamps.
15723 They are similar to the @option{re} option to @code{ffmpeg}.
15725 They accept the following options:
15729 Time limit for the pauses. Any pause longer than that will be considered
15730 a timestamp discontinuity and reset the timer. Default is 2 seconds.
15733 @section select, aselect
15735 Select frames to pass in output.
15737 This filter accepts the following options:
15742 Set expression, which is evaluated for each input frame.
15744 If the expression is evaluated to zero, the frame is discarded.
15746 If the evaluation result is negative or NaN, the frame is sent to the
15747 first output; otherwise it is sent to the output with index
15748 @code{ceil(val)-1}, assuming that the input index starts from 0.
15750 For example a value of @code{1.2} corresponds to the output with index
15751 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
15754 Set the number of outputs. The output to which to send the selected
15755 frame is based on the result of the evaluation. Default value is 1.
15758 The expression can contain the following constants:
15762 The (sequential) number of the filtered frame, starting from 0.
15765 The (sequential) number of the selected frame, starting from 0.
15767 @item prev_selected_n
15768 The sequential number of the last selected frame. It's NAN if undefined.
15771 The timebase of the input timestamps.
15774 The PTS (Presentation TimeStamp) of the filtered video frame,
15775 expressed in @var{TB} units. It's NAN if undefined.
15778 The PTS of the filtered video frame,
15779 expressed in seconds. It's NAN if undefined.
15782 The PTS of the previously filtered video frame. It's NAN if undefined.
15784 @item prev_selected_pts
15785 The PTS of the last previously filtered video frame. It's NAN if undefined.
15787 @item prev_selected_t
15788 The PTS of the last previously selected video frame. It's NAN if undefined.
15791 The PTS of the first video frame in the video. It's NAN if undefined.
15794 The time of the first video frame in the video. It's NAN if undefined.
15796 @item pict_type @emph{(video only)}
15797 The type of the filtered frame. It can assume one of the following
15809 @item interlace_type @emph{(video only)}
15810 The frame interlace type. It can assume one of the following values:
15813 The frame is progressive (not interlaced).
15815 The frame is top-field-first.
15817 The frame is bottom-field-first.
15820 @item consumed_sample_n @emph{(audio only)}
15821 the number of selected samples before the current frame
15823 @item samples_n @emph{(audio only)}
15824 the number of samples in the current frame
15826 @item sample_rate @emph{(audio only)}
15827 the input sample rate
15830 This is 1 if the filtered frame is a key-frame, 0 otherwise.
15833 the position in the file of the filtered frame, -1 if the information
15834 is not available (e.g. for synthetic video)
15836 @item scene @emph{(video only)}
15837 value between 0 and 1 to indicate a new scene; a low value reflects a low
15838 probability for the current frame to introduce a new scene, while a higher
15839 value means the current frame is more likely to be one (see the example below)
15841 @item concatdec_select
15842 The concat demuxer can select only part of a concat input file by setting an
15843 inpoint and an outpoint, but the output packets may not be entirely contained
15844 in the selected interval. By using this variable, it is possible to skip frames
15845 generated by the concat demuxer which are not exactly contained in the selected
15848 This works by comparing the frame pts against the @var{lavf.concat.start_time}
15849 and the @var{lavf.concat.duration} packet metadata values which are also
15850 present in the decoded frames.
15852 The @var{concatdec_select} variable is -1 if the frame pts is at least
15853 start_time and either the duration metadata is missing or the frame pts is less
15854 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
15857 That basically means that an input frame is selected if its pts is within the
15858 interval set by the concat demuxer.
15862 The default value of the select expression is "1".
15864 @subsection Examples
15868 Select all frames in input:
15873 The example above is the same as:
15885 Select only I-frames:
15887 select='eq(pict_type\,I)'
15891 Select one frame every 100:
15893 select='not(mod(n\,100))'
15897 Select only frames contained in the 10-20 time interval:
15899 select=between(t\,10\,20)
15903 Select only I-frames contained in the 10-20 time interval:
15905 select=between(t\,10\,20)*eq(pict_type\,I)
15909 Select frames with a minimum distance of 10 seconds:
15911 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
15915 Use aselect to select only audio frames with samples number > 100:
15917 aselect='gt(samples_n\,100)'
15921 Create a mosaic of the first scenes:
15923 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
15926 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
15930 Send even and odd frames to separate outputs, and compose them:
15932 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
15936 Select useful frames from an ffconcat file which is using inpoints and
15937 outpoints but where the source files are not intra frame only.
15939 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
15943 @section sendcmd, asendcmd
15945 Send commands to filters in the filtergraph.
15947 These filters read commands to be sent to other filters in the
15950 @code{sendcmd} must be inserted between two video filters,
15951 @code{asendcmd} must be inserted between two audio filters, but apart
15952 from that they act the same way.
15954 The specification of commands can be provided in the filter arguments
15955 with the @var{commands} option, or in a file specified by the
15956 @var{filename} option.
15958 These filters accept the following options:
15961 Set the commands to be read and sent to the other filters.
15963 Set the filename of the commands to be read and sent to the other
15967 @subsection Commands syntax
15969 A commands description consists of a sequence of interval
15970 specifications, comprising a list of commands to be executed when a
15971 particular event related to that interval occurs. The occurring event
15972 is typically the current frame time entering or leaving a given time
15975 An interval is specified by the following syntax:
15977 @var{START}[-@var{END}] @var{COMMANDS};
15980 The time interval is specified by the @var{START} and @var{END} times.
15981 @var{END} is optional and defaults to the maximum time.
15983 The current frame time is considered within the specified interval if
15984 it is included in the interval [@var{START}, @var{END}), that is when
15985 the time is greater or equal to @var{START} and is lesser than
15988 @var{COMMANDS} consists of a sequence of one or more command
15989 specifications, separated by ",", relating to that interval. The
15990 syntax of a command specification is given by:
15992 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
15995 @var{FLAGS} is optional and specifies the type of events relating to
15996 the time interval which enable sending the specified command, and must
15997 be a non-null sequence of identifier flags separated by "+" or "|" and
15998 enclosed between "[" and "]".
16000 The following flags are recognized:
16003 The command is sent when the current frame timestamp enters the
16004 specified interval. In other words, the command is sent when the
16005 previous frame timestamp was not in the given interval, and the
16009 The command is sent when the current frame timestamp leaves the
16010 specified interval. In other words, the command is sent when the
16011 previous frame timestamp was in the given interval, and the
16015 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
16018 @var{TARGET} specifies the target of the command, usually the name of
16019 the filter class or a specific filter instance name.
16021 @var{COMMAND} specifies the name of the command for the target filter.
16023 @var{ARG} is optional and specifies the optional list of argument for
16024 the given @var{COMMAND}.
16026 Between one interval specification and another, whitespaces, or
16027 sequences of characters starting with @code{#} until the end of line,
16028 are ignored and can be used to annotate comments.
16030 A simplified BNF description of the commands specification syntax
16033 @var{COMMAND_FLAG} ::= "enter" | "leave"
16034 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
16035 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
16036 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
16037 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
16038 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
16041 @subsection Examples
16045 Specify audio tempo change at second 4:
16047 asendcmd=c='4.0 atempo tempo 1.5',atempo
16051 Specify a list of drawtext and hue commands in a file.
16053 # show text in the interval 5-10
16054 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
16055 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16057 # desaturate the image in the interval 15-20
16058 15.0-20.0 [enter] hue s 0,
16059 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16061 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16063 # apply an exponential saturation fade-out effect, starting from time 25
16064 25 [enter] hue s exp(25-t)
16067 A filtergraph allowing to read and process the above command list
16068 stored in a file @file{test.cmd}, can be specified with:
16070 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16075 @section setpts, asetpts
16077 Change the PTS (presentation timestamp) of the input frames.
16079 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16081 This filter accepts the following options:
16086 The expression which is evaluated for each frame to construct its timestamp.
16090 The expression is evaluated through the eval API and can contain the following
16095 frame rate, only defined for constant frame-rate video
16098 The presentation timestamp in input
16101 The count of the input frame for video or the number of consumed samples,
16102 not including the current frame for audio, starting from 0.
16104 @item NB_CONSUMED_SAMPLES
16105 The number of consumed samples, not including the current frame (only
16108 @item NB_SAMPLES, S
16109 The number of samples in the current frame (only audio)
16111 @item SAMPLE_RATE, SR
16112 The audio sample rate.
16115 The PTS of the first frame.
16118 the time in seconds of the first frame
16121 State whether the current frame is interlaced.
16124 the time in seconds of the current frame
16127 original position in the file of the frame, or undefined if undefined
16128 for the current frame
16131 The previous input PTS.
16134 previous input time in seconds
16137 The previous output PTS.
16140 previous output time in seconds
16143 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16147 The wallclock (RTC) time at the start of the movie in microseconds.
16150 The timebase of the input timestamps.
16154 @subsection Examples
16158 Start counting PTS from zero
16160 setpts=PTS-STARTPTS
16164 Apply fast motion effect:
16170 Apply slow motion effect:
16176 Set fixed rate of 25 frames per second:
16182 Set fixed rate 25 fps with some jitter:
16184 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16188 Apply an offset of 10 seconds to the input PTS:
16194 Generate timestamps from a "live source" and rebase onto the current timebase:
16196 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16200 Generate timestamps by counting samples:
16207 @section settb, asettb
16209 Set the timebase to use for the output frames timestamps.
16210 It is mainly useful for testing timebase configuration.
16212 It accepts the following parameters:
16217 The expression which is evaluated into the output timebase.
16221 The value for @option{tb} is an arithmetic expression representing a
16222 rational. The expression can contain the constants "AVTB" (the default
16223 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16224 audio only). Default value is "intb".
16226 @subsection Examples
16230 Set the timebase to 1/25:
16236 Set the timebase to 1/10:
16242 Set the timebase to 1001/1000:
16248 Set the timebase to 2*intb:
16254 Set the default timebase value:
16261 Convert input audio to a video output representing frequency spectrum
16262 logarithmically using Brown-Puckette constant Q transform algorithm with
16263 direct frequency domain coefficient calculation (but the transform itself
16264 is not really constant Q, instead the Q factor is actually variable/clamped),
16265 with musical tone scale, from E0 to D#10.
16267 The filter accepts the following options:
16271 Specify the video size for the output. It must be even. For the syntax of this option,
16272 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16273 Default value is @code{1920x1080}.
16276 Set the output frame rate. Default value is @code{25}.
16279 Set the bargraph height. It must be even. Default value is @code{-1} which
16280 computes the bargraph height automatically.
16283 Set the axis height. It must be even. Default value is @code{-1} which computes
16284 the axis height automatically.
16287 Set the sonogram height. It must be even. Default value is @code{-1} which
16288 computes the sonogram height automatically.
16291 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16292 instead. Default value is @code{1}.
16294 @item sono_v, volume
16295 Specify the sonogram volume expression. It can contain variables:
16298 the @var{bar_v} evaluated expression
16299 @item frequency, freq, f
16300 the frequency where it is evaluated
16301 @item timeclamp, tc
16302 the value of @var{timeclamp} option
16306 @item a_weighting(f)
16307 A-weighting of equal loudness
16308 @item b_weighting(f)
16309 B-weighting of equal loudness
16310 @item c_weighting(f)
16311 C-weighting of equal loudness.
16313 Default value is @code{16}.
16315 @item bar_v, volume2
16316 Specify the bargraph volume expression. It can contain variables:
16319 the @var{sono_v} evaluated expression
16320 @item frequency, freq, f
16321 the frequency where it is evaluated
16322 @item timeclamp, tc
16323 the value of @var{timeclamp} option
16327 @item a_weighting(f)
16328 A-weighting of equal loudness
16329 @item b_weighting(f)
16330 B-weighting of equal loudness
16331 @item c_weighting(f)
16332 C-weighting of equal loudness.
16334 Default value is @code{sono_v}.
16336 @item sono_g, gamma
16337 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16338 higher gamma makes the spectrum having more range. Default value is @code{3}.
16339 Acceptable range is @code{[1, 7]}.
16341 @item bar_g, gamma2
16342 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16345 @item timeclamp, tc
16346 Specify the transform timeclamp. At low frequency, there is trade-off between
16347 accuracy in time domain and frequency domain. If timeclamp is lower,
16348 event in time domain is represented more accurately (such as fast bass drum),
16349 otherwise event in frequency domain is represented more accurately
16350 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16353 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16354 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16357 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16358 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16361 This option is deprecated and ignored.
16364 Specify the transform length in time domain. Use this option to control accuracy
16365 trade-off between time domain and frequency domain at every frequency sample.
16366 It can contain variables:
16368 @item frequency, freq, f
16369 the frequency where it is evaluated
16370 @item timeclamp, tc
16371 the value of @var{timeclamp} option.
16373 Default value is @code{384*tc/(384+tc*f)}.
16376 Specify the transform count for every video frame. Default value is @code{6}.
16377 Acceptable range is @code{[1, 30]}.
16380 Specify the transform count for every single pixel. Default value is @code{0},
16381 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16384 Specify font file for use with freetype to draw the axis. If not specified,
16385 use embedded font. Note that drawing with font file or embedded font is not
16386 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16390 Specify font color expression. This is arithmetic expression that should return
16391 integer value 0xRRGGBB. It can contain variables:
16393 @item frequency, freq, f
16394 the frequency where it is evaluated
16395 @item timeclamp, tc
16396 the value of @var{timeclamp} option
16401 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16402 @item r(x), g(x), b(x)
16403 red, green, and blue value of intensity x.
16405 Default value is @code{st(0, (midi(f)-59.5)/12);
16406 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16407 r(1-ld(1)) + b(ld(1))}.
16410 Specify image file to draw the axis. This option override @var{fontfile} and
16411 @var{fontcolor} option.
16414 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16415 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16416 Default value is @code{1}.
16420 @subsection Examples
16424 Playing audio while showing the spectrum:
16426 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16430 Same as above, but with frame rate 30 fps:
16432 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16436 Playing at 1280x720:
16438 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16442 Disable sonogram display:
16448 A1 and its harmonics: A1, A2, (near)E3, A3:
16450 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),
16451 asplit[a][out1]; [a] showcqt [out0]'
16455 Same as above, but with more accuracy in frequency domain:
16457 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),
16458 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16464 bar_v=10:sono_v=bar_v*a_weighting(f)
16468 Custom gamma, now spectrum is linear to the amplitude.
16474 Custom tlength equation:
16476 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)))'
16480 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
16482 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
16486 Custom frequency range with custom axis using image file:
16488 axisfile=myaxis.png:basefreq=40:endfreq=10000
16494 Convert input audio to video output representing the audio power spectrum.
16495 Audio amplitude is on Y-axis while frequency is on X-axis.
16497 The filter accepts the following options:
16501 Specify size of video. For the syntax of this option, check the
16502 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16503 Default is @code{1024x512}.
16507 This set how each frequency bin will be represented.
16509 It accepts the following values:
16515 Default is @code{bar}.
16518 Set amplitude scale.
16520 It accepts the following values:
16534 Default is @code{log}.
16537 Set frequency scale.
16539 It accepts the following values:
16548 Reverse logarithmic scale.
16550 Default is @code{lin}.
16555 It accepts the following values:
16571 Default is @code{w2048}
16574 Set windowing function.
16576 It accepts the following values:
16598 Default is @code{hanning}.
16601 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
16602 which means optimal overlap for selected window function will be picked.
16605 Set time averaging. Setting this to 0 will display current maximal peaks.
16606 Default is @code{1}, which means time averaging is disabled.
16609 Specify list of colors separated by space or by '|' which will be used to
16610 draw channel frequencies. Unrecognized or missing colors will be replaced
16614 Set channel display mode.
16616 It accepts the following values:
16621 Default is @code{combined}.
16624 Set minimum amplitude used in @code{log} amplitude scaler.
16628 @anchor{showspectrum}
16629 @section showspectrum
16631 Convert input audio to a video output, representing the audio frequency
16634 The filter accepts the following options:
16638 Specify the video size for the output. For the syntax of this option, check the
16639 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16640 Default value is @code{640x512}.
16643 Specify how the spectrum should slide along the window.
16645 It accepts the following values:
16648 the samples start again on the left when they reach the right
16650 the samples scroll from right to left
16652 the samples scroll from left to right
16654 frames are only produced when the samples reach the right
16657 Default value is @code{replace}.
16660 Specify display mode.
16662 It accepts the following values:
16665 all channels are displayed in the same row
16667 all channels are displayed in separate rows
16670 Default value is @samp{combined}.
16673 Specify display color mode.
16675 It accepts the following values:
16678 each channel is displayed in a separate color
16680 each channel is displayed using the same color scheme
16682 each channel is displayed using the rainbow color scheme
16684 each channel is displayed using the moreland color scheme
16686 each channel is displayed using the nebulae color scheme
16688 each channel is displayed using the fire color scheme
16690 each channel is displayed using the fiery color scheme
16692 each channel is displayed using the fruit color scheme
16694 each channel is displayed using the cool color scheme
16697 Default value is @samp{channel}.
16700 Specify scale used for calculating intensity color values.
16702 It accepts the following values:
16707 square root, default
16718 Default value is @samp{sqrt}.
16721 Set saturation modifier for displayed colors. Negative values provide
16722 alternative color scheme. @code{0} is no saturation at all.
16723 Saturation must be in [-10.0, 10.0] range.
16724 Default value is @code{1}.
16727 Set window function.
16729 It accepts the following values:
16753 Default value is @code{hann}.
16756 Set orientation of time vs frequency axis. Can be @code{vertical} or
16757 @code{horizontal}. Default is @code{vertical}.
16760 Set ratio of overlap window. Default value is @code{0}.
16761 When value is @code{1} overlap is set to recommended size for specific
16762 window function currently used.
16765 Set scale gain for calculating intensity color values.
16766 Default value is @code{1}.
16769 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
16772 Set color rotation, must be in [-1.0, 1.0] range.
16773 Default value is @code{0}.
16776 The usage is very similar to the showwaves filter; see the examples in that
16779 @subsection Examples
16783 Large window with logarithmic color scaling:
16785 showspectrum=s=1280x480:scale=log
16789 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
16791 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
16792 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
16796 @section showspectrumpic
16798 Convert input audio to a single video frame, representing the audio frequency
16801 The filter accepts the following options:
16805 Specify the video size for the output. For the syntax of this option, check the
16806 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16807 Default value is @code{4096x2048}.
16810 Specify display mode.
16812 It accepts the following values:
16815 all channels are displayed in the same row
16817 all channels are displayed in separate rows
16819 Default value is @samp{combined}.
16822 Specify display color mode.
16824 It accepts the following values:
16827 each channel is displayed in a separate color
16829 each channel is displayed using the same color scheme
16831 each channel is displayed using the rainbow color scheme
16833 each channel is displayed using the moreland color scheme
16835 each channel is displayed using the nebulae color scheme
16837 each channel is displayed using the fire color scheme
16839 each channel is displayed using the fiery color scheme
16841 each channel is displayed using the fruit color scheme
16843 each channel is displayed using the cool color scheme
16845 Default value is @samp{intensity}.
16848 Specify scale used for calculating intensity color values.
16850 It accepts the following values:
16855 square root, default
16865 Default value is @samp{log}.
16868 Set saturation modifier for displayed colors. Negative values provide
16869 alternative color scheme. @code{0} is no saturation at all.
16870 Saturation must be in [-10.0, 10.0] range.
16871 Default value is @code{1}.
16874 Set window function.
16876 It accepts the following values:
16899 Default value is @code{hann}.
16902 Set orientation of time vs frequency axis. Can be @code{vertical} or
16903 @code{horizontal}. Default is @code{vertical}.
16906 Set scale gain for calculating intensity color values.
16907 Default value is @code{1}.
16910 Draw time and frequency axes and legends. Default is enabled.
16913 Set color rotation, must be in [-1.0, 1.0] range.
16914 Default value is @code{0}.
16917 @subsection Examples
16921 Extract an audio spectrogram of a whole audio track
16922 in a 1024x1024 picture using @command{ffmpeg}:
16924 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
16928 @section showvolume
16930 Convert input audio volume to a video output.
16932 The filter accepts the following options:
16939 Set border width, allowed range is [0, 5]. Default is 1.
16942 Set channel width, allowed range is [80, 8192]. Default is 400.
16945 Set channel height, allowed range is [1, 900]. Default is 20.
16948 Set fade, allowed range is [0.001, 1]. Default is 0.95.
16951 Set volume color expression.
16953 The expression can use the following variables:
16957 Current max volume of channel in dB.
16963 Current channel number, starting from 0.
16967 If set, displays channel names. Default is enabled.
16970 If set, displays volume values. Default is enabled.
16973 Set orientation, can be @code{horizontal} or @code{vertical},
16974 default is @code{horizontal}.
16977 Set step size, allowed range s [0, 5]. Default is 0, which means
16983 Convert input audio to a video output, representing the samples waves.
16985 The filter accepts the following options:
16989 Specify the video size for the output. For the syntax of this option, check the
16990 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16991 Default value is @code{600x240}.
16996 Available values are:
16999 Draw a point for each sample.
17002 Draw a vertical line for each sample.
17005 Draw a point for each sample and a line between them.
17008 Draw a centered vertical line for each sample.
17011 Default value is @code{point}.
17014 Set the number of samples which are printed on the same column. A
17015 larger value will decrease the frame rate. Must be a positive
17016 integer. This option can be set only if the value for @var{rate}
17017 is not explicitly specified.
17020 Set the (approximate) output frame rate. This is done by setting the
17021 option @var{n}. Default value is "25".
17023 @item split_channels
17024 Set if channels should be drawn separately or overlap. Default value is 0.
17027 Set colors separated by '|' which are going to be used for drawing of each channel.
17030 Set amplitude scale.
17032 Available values are:
17050 @subsection Examples
17054 Output the input file audio and the corresponding video representation
17057 amovie=a.mp3,asplit[out0],showwaves[out1]
17061 Create a synthetic signal and show it with showwaves, forcing a
17062 frame rate of 30 frames per second:
17064 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17068 @section showwavespic
17070 Convert input audio to a single video frame, representing the samples waves.
17072 The filter accepts the following options:
17076 Specify the video size for the output. For the syntax of this option, check the
17077 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17078 Default value is @code{600x240}.
17080 @item split_channels
17081 Set if channels should be drawn separately or overlap. Default value is 0.
17084 Set colors separated by '|' which are going to be used for drawing of each channel.
17087 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17091 @subsection Examples
17095 Extract a channel split representation of the wave form of a whole audio track
17096 in a 1024x800 picture using @command{ffmpeg}:
17098 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17102 @section spectrumsynth
17104 Sythesize audio from 2 input video spectrums, first input stream represents
17105 magnitude across time and second represents phase across time.
17106 The filter will transform from frequency domain as displayed in videos back
17107 to time domain as presented in audio output.
17109 This filter is primarly created for reversing processed @ref{showspectrum}
17110 filter outputs, but can synthesize sound from other spectrograms too.
17111 But in such case results are going to be poor if the phase data is not
17112 available, because in such cases phase data need to be recreated, usually
17113 its just recreated from random noise.
17114 For best results use gray only output (@code{channel} color mode in
17115 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17116 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17117 @code{data} option. Inputs videos should generally use @code{fullframe}
17118 slide mode as that saves resources needed for decoding video.
17120 The filter accepts the following options:
17124 Specify sample rate of output audio, the sample rate of audio from which
17125 spectrum was generated may differ.
17128 Set number of channels represented in input video spectrums.
17131 Set scale which was used when generating magnitude input spectrum.
17132 Can be @code{lin} or @code{log}. Default is @code{log}.
17135 Set slide which was used when generating inputs spectrums.
17136 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17137 Default is @code{fullframe}.
17140 Set window function used for resynthesis.
17143 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17144 which means optimal overlap for selected window function will be picked.
17147 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17148 Default is @code{vertical}.
17151 @subsection Examples
17155 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17156 then resynthesize videos back to audio with spectrumsynth:
17158 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
17159 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
17160 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17164 @section split, asplit
17166 Split input into several identical outputs.
17168 @code{asplit} works with audio input, @code{split} with video.
17170 The filter accepts a single parameter which specifies the number of outputs. If
17171 unspecified, it defaults to 2.
17173 @subsection Examples
17177 Create two separate outputs from the same input:
17179 [in] split [out0][out1]
17183 To create 3 or more outputs, you need to specify the number of
17186 [in] asplit=3 [out0][out1][out2]
17190 Create two separate outputs from the same input, one cropped and
17193 [in] split [splitout1][splitout2];
17194 [splitout1] crop=100:100:0:0 [cropout];
17195 [splitout2] pad=200:200:100:100 [padout];
17199 Create 5 copies of the input audio with @command{ffmpeg}:
17201 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17207 Receive commands sent through a libzmq client, and forward them to
17208 filters in the filtergraph.
17210 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17211 must be inserted between two video filters, @code{azmq} between two
17214 To enable these filters you need to install the libzmq library and
17215 headers and configure FFmpeg with @code{--enable-libzmq}.
17217 For more information about libzmq see:
17218 @url{http://www.zeromq.org/}
17220 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17221 receives messages sent through a network interface defined by the
17222 @option{bind_address} option.
17224 The received message must be in the form:
17226 @var{TARGET} @var{COMMAND} [@var{ARG}]
17229 @var{TARGET} specifies the target of the command, usually the name of
17230 the filter class or a specific filter instance name.
17232 @var{COMMAND} specifies the name of the command for the target filter.
17234 @var{ARG} is optional and specifies the optional argument list for the
17235 given @var{COMMAND}.
17237 Upon reception, the message is processed and the corresponding command
17238 is injected into the filtergraph. Depending on the result, the filter
17239 will send a reply to the client, adopting the format:
17241 @var{ERROR_CODE} @var{ERROR_REASON}
17245 @var{MESSAGE} is optional.
17247 @subsection Examples
17249 Look at @file{tools/zmqsend} for an example of a zmq client which can
17250 be used to send commands processed by these filters.
17252 Consider the following filtergraph generated by @command{ffplay}
17254 ffplay -dumpgraph 1 -f lavfi "
17255 color=s=100x100:c=red [l];
17256 color=s=100x100:c=blue [r];
17257 nullsrc=s=200x100, zmq [bg];
17258 [bg][l] overlay [bg+l];
17259 [bg+l][r] overlay=x=100 "
17262 To change the color of the left side of the video, the following
17263 command can be used:
17265 echo Parsed_color_0 c yellow | tools/zmqsend
17268 To change the right side:
17270 echo Parsed_color_1 c pink | tools/zmqsend
17273 @c man end MULTIMEDIA FILTERS
17275 @chapter Multimedia Sources
17276 @c man begin MULTIMEDIA SOURCES
17278 Below is a description of the currently available multimedia sources.
17282 This is the same as @ref{movie} source, except it selects an audio
17288 Read audio and/or video stream(s) from a movie container.
17290 It accepts the following parameters:
17294 The name of the resource to read (not necessarily a file; it can also be a
17295 device or a stream accessed through some protocol).
17297 @item format_name, f
17298 Specifies the format assumed for the movie to read, and can be either
17299 the name of a container or an input device. If not specified, the
17300 format is guessed from @var{movie_name} or by probing.
17302 @item seek_point, sp
17303 Specifies the seek point in seconds. The frames will be output
17304 starting from this seek point. The parameter is evaluated with
17305 @code{av_strtod}, so the numerical value may be suffixed by an IS
17306 postfix. The default value is "0".
17309 Specifies the streams to read. Several streams can be specified,
17310 separated by "+". The source will then have as many outputs, in the
17311 same order. The syntax is explained in the ``Stream specifiers''
17312 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17313 respectively the default (best suited) video and audio stream. Default
17314 is "dv", or "da" if the filter is called as "amovie".
17316 @item stream_index, si
17317 Specifies the index of the video stream to read. If the value is -1,
17318 the most suitable video stream will be automatically selected. The default
17319 value is "-1". Deprecated. If the filter is called "amovie", it will select
17320 audio instead of video.
17323 Specifies how many times to read the stream in sequence.
17324 If the value is less than 1, the stream will be read again and again.
17325 Default value is "1".
17327 Note that when the movie is looped the source timestamps are not
17328 changed, so it will generate non monotonically increasing timestamps.
17330 @item discontinuity
17331 Specifies the time difference between frames above which the point is
17332 considered a timestamp discontinuity which is removed by adjusting the later
17336 It allows overlaying a second video on top of the main input of
17337 a filtergraph, as shown in this graph:
17339 input -----------> deltapts0 --> overlay --> output
17342 movie --> scale--> deltapts1 -------+
17344 @subsection Examples
17348 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17349 on top of the input labelled "in":
17351 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17352 [in] setpts=PTS-STARTPTS [main];
17353 [main][over] overlay=16:16 [out]
17357 Read from a video4linux2 device, and overlay it on top of the input
17360 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17361 [in] setpts=PTS-STARTPTS [main];
17362 [main][over] overlay=16:16 [out]
17366 Read the first video stream and the audio stream with id 0x81 from
17367 dvd.vob; the video is connected to the pad named "video" and the audio is
17368 connected to the pad named "audio":
17370 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17374 @subsection Commands
17376 Both movie and amovie support the following commands:
17379 Perform seek using "av_seek_frame".
17380 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17383 @var{stream_index}: If stream_index is -1, a default
17384 stream is selected, and @var{timestamp} is automatically converted
17385 from AV_TIME_BASE units to the stream specific time_base.
17387 @var{timestamp}: Timestamp in AVStream.time_base units
17388 or, if no stream is specified, in AV_TIME_BASE units.
17390 @var{flags}: Flags which select direction and seeking mode.
17394 Get movie duration in AV_TIME_BASE units.
17398 @c man end MULTIMEDIA SOURCES