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].
4402 Set what planes of frame filter will use for averaging. Default is all.
4407 Compute the bounding box for the non-black pixels in the input frame
4410 This filter computes the bounding box containing all the pixels with a
4411 luminance value greater than the minimum allowed value.
4412 The parameters describing the bounding box are printed on the filter
4415 The filter accepts the following option:
4419 Set the minimal luminance value. Default is @code{16}.
4422 @section bitplanenoise
4424 Show and measure bit plane noise.
4426 The filter accepts the following options:
4430 Set which plane to analyze. Default is @code{1}.
4433 Filter out noisy pixels from @code{bitplane} set above.
4434 Default is disabled.
4437 @section blackdetect
4439 Detect video intervals that are (almost) completely black. Can be
4440 useful to detect chapter transitions, commercials, or invalid
4441 recordings. Output lines contains the time for the start, end and
4442 duration of the detected black interval expressed in seconds.
4444 In order to display the output lines, you need to set the loglevel at
4445 least to the AV_LOG_INFO value.
4447 The filter accepts the following options:
4450 @item black_min_duration, d
4451 Set the minimum detected black duration expressed in seconds. It must
4452 be a non-negative floating point number.
4454 Default value is 2.0.
4456 @item picture_black_ratio_th, pic_th
4457 Set the threshold for considering a picture "black".
4458 Express the minimum value for the ratio:
4460 @var{nb_black_pixels} / @var{nb_pixels}
4463 for which a picture is considered black.
4464 Default value is 0.98.
4466 @item pixel_black_th, pix_th
4467 Set the threshold for considering a pixel "black".
4469 The threshold expresses the maximum pixel luminance value for which a
4470 pixel is considered "black". The provided value is scaled according to
4471 the following equation:
4473 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4476 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4477 the input video format, the range is [0-255] for YUV full-range
4478 formats and [16-235] for YUV non full-range formats.
4480 Default value is 0.10.
4483 The following example sets the maximum pixel threshold to the minimum
4484 value, and detects only black intervals of 2 or more seconds:
4486 blackdetect=d=2:pix_th=0.00
4491 Detect frames that are (almost) completely black. Can be useful to
4492 detect chapter transitions or commercials. Output lines consist of
4493 the frame number of the detected frame, the percentage of blackness,
4494 the position in the file if known or -1 and the timestamp in seconds.
4496 In order to display the output lines, you need to set the loglevel at
4497 least to the AV_LOG_INFO value.
4499 It accepts the following parameters:
4504 The percentage of the pixels that have to be below the threshold; it defaults to
4507 @item threshold, thresh
4508 The threshold below which a pixel value is considered black; it defaults to
4513 @section blend, tblend
4515 Blend two video frames into each other.
4517 The @code{blend} filter takes two input streams and outputs one
4518 stream, the first input is the "top" layer and second input is
4519 "bottom" layer. Output terminates when shortest input terminates.
4521 The @code{tblend} (time blend) filter takes two consecutive frames
4522 from one single stream, and outputs the result obtained by blending
4523 the new frame on top of the old frame.
4525 A description of the accepted options follows.
4533 Set blend mode for specific pixel component or all pixel components in case
4534 of @var{all_mode}. Default value is @code{normal}.
4536 Available values for component modes are:
4577 Set blend opacity for specific pixel component or all pixel components in case
4578 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4585 Set blend expression for specific pixel component or all pixel components in case
4586 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4588 The expressions can use the following variables:
4592 The sequential number of the filtered frame, starting from @code{0}.
4596 the coordinates of the current sample
4600 the width and height of currently filtered plane
4604 Width and height scale depending on the currently filtered plane. It is the
4605 ratio between the corresponding luma plane number of pixels and the current
4606 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4607 @code{0.5,0.5} for chroma planes.
4610 Time of the current frame, expressed in seconds.
4613 Value of pixel component at current location for first video frame (top layer).
4616 Value of pixel component at current location for second video frame (bottom layer).
4620 Force termination when the shortest input terminates. Default is
4621 @code{0}. This option is only defined for the @code{blend} filter.
4624 Continue applying the last bottom frame after the end of the stream. A value of
4625 @code{0} disable the filter after the last frame of the bottom layer is reached.
4626 Default is @code{1}. This option is only defined for the @code{blend} filter.
4629 @subsection Examples
4633 Apply transition from bottom layer to top layer in first 10 seconds:
4635 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4639 Apply 1x1 checkerboard effect:
4641 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4645 Apply uncover left effect:
4647 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4651 Apply uncover down effect:
4653 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4657 Apply uncover up-left effect:
4659 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4663 Split diagonally video and shows top and bottom layer on each side:
4665 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4669 Display differences between the current and the previous frame:
4671 tblend=all_mode=difference128
4677 Apply a boxblur algorithm to the input video.
4679 It accepts the following parameters:
4683 @item luma_radius, lr
4684 @item luma_power, lp
4685 @item chroma_radius, cr
4686 @item chroma_power, cp
4687 @item alpha_radius, ar
4688 @item alpha_power, ap
4692 A description of the accepted options follows.
4695 @item luma_radius, lr
4696 @item chroma_radius, cr
4697 @item alpha_radius, ar
4698 Set an expression for the box radius in pixels used for blurring the
4699 corresponding input plane.
4701 The radius value must be a non-negative number, and must not be
4702 greater than the value of the expression @code{min(w,h)/2} for the
4703 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4706 Default value for @option{luma_radius} is "2". If not specified,
4707 @option{chroma_radius} and @option{alpha_radius} default to the
4708 corresponding value set for @option{luma_radius}.
4710 The expressions can contain the following constants:
4714 The input width and height in pixels.
4718 The input chroma image width and height in pixels.
4722 The horizontal and vertical chroma subsample values. For example, for the
4723 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4726 @item luma_power, lp
4727 @item chroma_power, cp
4728 @item alpha_power, ap
4729 Specify how many times the boxblur filter is applied to the
4730 corresponding plane.
4732 Default value for @option{luma_power} is 2. If not specified,
4733 @option{chroma_power} and @option{alpha_power} default to the
4734 corresponding value set for @option{luma_power}.
4736 A value of 0 will disable the effect.
4739 @subsection Examples
4743 Apply a boxblur filter with the luma, chroma, and alpha radii
4746 boxblur=luma_radius=2:luma_power=1
4751 Set the luma radius to 2, and alpha and chroma radius to 0:
4753 boxblur=2:1:cr=0:ar=0
4757 Set the luma and chroma radii to a fraction of the video dimension:
4759 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4765 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4766 Deinterlacing Filter").
4768 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4769 interpolation algorithms.
4770 It accepts the following parameters:
4774 The interlacing mode to adopt. It accepts one of the following values:
4778 Output one frame for each frame.
4780 Output one frame for each field.
4783 The default value is @code{send_field}.
4786 The picture field parity assumed for the input interlaced video. It accepts one
4787 of the following values:
4791 Assume the top field is first.
4793 Assume the bottom field is first.
4795 Enable automatic detection of field parity.
4798 The default value is @code{auto}.
4799 If the interlacing is unknown or the decoder does not export this information,
4800 top field first will be assumed.
4803 Specify which frames to deinterlace. Accept one of the following
4808 Deinterlace all frames.
4810 Only deinterlace frames marked as interlaced.
4813 The default value is @code{all}.
4817 YUV colorspace color/chroma keying.
4819 The filter accepts the following options:
4823 The color which will be replaced with transparency.
4826 Similarity percentage with the key color.
4828 0.01 matches only the exact key color, while 1.0 matches everything.
4833 0.0 makes pixels either fully transparent, or not transparent at all.
4835 Higher values result in semi-transparent pixels, with a higher transparency
4836 the more similar the pixels color is to the key color.
4839 Signals that the color passed is already in YUV instead of RGB.
4841 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4842 This can be used to pass exact YUV values as hexadecimal numbers.
4845 @subsection Examples
4849 Make every green pixel in the input image transparent:
4851 ffmpeg -i input.png -vf chromakey=green out.png
4855 Overlay a greenscreen-video on top of a static black background.
4857 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
4863 Display CIE color diagram with pixels overlaid onto it.
4865 The filter accepts the following options:
4880 @item uhdtv, rec2020
4893 Set what gamuts to draw.
4895 See @code{system} option for available values.
4898 Set ciescope size, by default set to 512.
4901 Set intensity used to map input pixel values to CIE diagram.
4904 Set contrast used to draw tongue colors that are out of active color system gamut.
4907 Correct gamma displayed on scope, by default enabled.
4910 Show white point on CIE diagram, by default disabled.
4913 Set input gamma. Used only with XYZ input color space.
4918 Visualize information exported by some codecs.
4920 Some codecs can export information through frames using side-data or other
4921 means. For example, some MPEG based codecs export motion vectors through the
4922 @var{export_mvs} flag in the codec @option{flags2} option.
4924 The filter accepts the following option:
4928 Set motion vectors to visualize.
4930 Available flags for @var{mv} are:
4934 forward predicted MVs of P-frames
4936 forward predicted MVs of B-frames
4938 backward predicted MVs of B-frames
4942 Display quantization parameters using the chroma planes.
4945 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4947 Available flags for @var{mv_type} are:
4951 forward predicted MVs
4953 backward predicted MVs
4956 @item frame_type, ft
4957 Set frame type to visualize motion vectors of.
4959 Available flags for @var{frame_type} are:
4963 intra-coded frames (I-frames)
4965 predicted frames (P-frames)
4967 bi-directionally predicted frames (B-frames)
4971 @subsection Examples
4975 Visualize forward predicted MVs of all frames using @command{ffplay}:
4977 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
4981 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
4983 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
4987 @section colorbalance
4988 Modify intensity of primary colors (red, green and blue) of input frames.
4990 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
4991 regions for the red-cyan, green-magenta or blue-yellow balance.
4993 A positive adjustment value shifts the balance towards the primary color, a negative
4994 value towards the complementary color.
4996 The filter accepts the following options:
5002 Adjust red, green and blue shadows (darkest pixels).
5007 Adjust red, green and blue midtones (medium pixels).
5012 Adjust red, green and blue highlights (brightest pixels).
5014 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5017 @subsection Examples
5021 Add red color cast to shadows:
5028 RGB colorspace color keying.
5030 The filter accepts the following options:
5034 The color which will be replaced with transparency.
5037 Similarity percentage with the key color.
5039 0.01 matches only the exact key color, while 1.0 matches everything.
5044 0.0 makes pixels either fully transparent, or not transparent at all.
5046 Higher values result in semi-transparent pixels, with a higher transparency
5047 the more similar the pixels color is to the key color.
5050 @subsection Examples
5054 Make every green pixel in the input image transparent:
5056 ffmpeg -i input.png -vf colorkey=green out.png
5060 Overlay a greenscreen-video on top of a static background image.
5062 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
5066 @section colorlevels
5068 Adjust video input frames using levels.
5070 The filter accepts the following options:
5077 Adjust red, green, blue and alpha input black point.
5078 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5084 Adjust red, green, blue and alpha input white point.
5085 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5087 Input levels are used to lighten highlights (bright tones), darken shadows
5088 (dark tones), change the balance of bright and dark tones.
5094 Adjust red, green, blue and alpha output black point.
5095 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5101 Adjust red, green, blue and alpha output white point.
5102 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5104 Output levels allows manual selection of a constrained output level range.
5107 @subsection Examples
5111 Make video output darker:
5113 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5119 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5123 Make video output lighter:
5125 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5129 Increase brightness:
5131 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5135 @section colorchannelmixer
5137 Adjust video input frames by re-mixing color channels.
5139 This filter modifies a color channel by adding the values associated to
5140 the other channels of the same pixels. For example if the value to
5141 modify is red, the output value will be:
5143 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5146 The filter accepts the following options:
5153 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5154 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5160 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5161 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5167 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5168 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5174 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5175 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5177 Allowed ranges for options are @code{[-2.0, 2.0]}.
5180 @subsection Examples
5184 Convert source to grayscale:
5186 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5189 Simulate sepia tones:
5191 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5195 @section colormatrix
5197 Convert color matrix.
5199 The filter accepts the following options:
5204 Specify the source and destination color matrix. Both values must be
5207 The accepted values are:
5226 For example to convert from BT.601 to SMPTE-240M, use the command:
5228 colormatrix=bt601:smpte240m
5233 Convert colorspace, transfer characteristics or color primaries.
5235 The filter accepts the following options:
5239 Specify all color properties at once.
5241 The accepted values are:
5270 Specify output colorspace.
5272 The accepted values are:
5281 BT.470BG or BT.601-6 625
5284 SMPTE-170M or BT.601-6 525
5290 BT.2020 with non-constant luminance
5295 Specify output transfer characteristics.
5297 The accepted values are:
5303 Constant gamma of 2.2
5306 Constant gamma of 2.8
5309 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5315 BT.2020 for 10-bits content
5318 BT.2020 for 12-bits content
5323 Specify output color primaries.
5325 The accepted values are:
5334 BT.470BG or BT.601-6 625
5337 SMPTE-170M or BT.601-6 525
5348 Specify output color range.
5350 The accepted values are:
5353 MPEG (restricted) range
5361 Specify output color format.
5363 The accepted values are:
5366 YUV 4:2:0 planar 8-bits
5369 YUV 4:2:0 planar 10-bits
5372 YUV 4:2:0 planar 12-bits
5375 YUV 4:2:2 planar 8-bits
5378 YUV 4:2:2 planar 10-bits
5381 YUV 4:2:2 planar 12-bits
5384 YUV 4:4:4 planar 8-bits
5387 YUV 4:4:4 planar 10-bits
5390 YUV 4:4:4 planar 12-bits
5395 Do a fast conversion, which skips gamma/primary correction. This will take
5396 significantly less CPU, but will be mathematically incorrect. To get output
5397 compatible with that produced by the colormatrix filter, use fast=1.
5400 Specify dithering mode.
5402 The accepted values are:
5408 Floyd-Steinberg dithering
5412 Whitepoint adaptation mode.
5414 The accepted values are:
5417 Bradford whitepoint adaptation
5420 von Kries whitepoint adaptation
5423 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5428 The filter converts the transfer characteristics, color space and color
5429 primaries to the specified user values. The output value, if not specified,
5430 is set to a default value based on the "all" property. If that property is
5431 also not specified, the filter will log an error. The output color range and
5432 format default to the same value as the input color range and format. The
5433 input transfer characteristics, color space, color primaries and color range
5434 should be set on the input data. If any of these are missing, the filter will
5435 log an error and no conversion will take place.
5437 For example to convert the input to SMPTE-240M, use the command:
5439 colorspace=smpte240m
5442 @section convolution
5444 Apply convolution 3x3 or 5x5 filter.
5446 The filter accepts the following options:
5453 Set matrix for each plane.
5454 Matrix is sequence of 9 or 25 signed integers.
5460 Set multiplier for calculated value for each plane.
5466 Set bias for each plane. This value is added to the result of the multiplication.
5467 Useful for making the overall image brighter or darker. Default is 0.0.
5470 @subsection Examples
5476 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"
5482 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"
5488 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"
5494 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"
5500 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"
5506 Copy the input source unchanged to the output. This is mainly useful for
5511 Video filtering on GPU using Apple's CoreImage API on OSX.
5513 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5514 processed by video hardware. However, software-based OpenGL implementations
5515 exist which means there is no guarantee for hardware processing. It depends on
5518 There are many filters and image generators provided by Apple that come with a
5519 large variety of options. The filter has to be referenced by its name along
5522 The coreimage filter accepts the following options:
5525 List all available filters and generators along with all their respective
5526 options as well as possible minimum and maximum values along with the default
5533 Specify all filters by their respective name and options.
5534 Use @var{list_filters} to determine all valid filter names and options.
5535 Numerical options are specified by a float value and are automatically clamped
5536 to their respective value range. Vector and color options have to be specified
5537 by a list of space separated float values. Character escaping has to be done.
5538 A special option name @code{default} is available to use default options for a
5541 It is required to specify either @code{default} or at least one of the filter options.
5542 All omitted options are used with their default values.
5543 The syntax of the filter string is as follows:
5545 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5549 Specify a rectangle where the output of the filter chain is copied into the
5550 input image. It is given by a list of space separated float values:
5552 output_rect=x\ y\ width\ height
5554 If not given, the output rectangle equals the dimensions of the input image.
5555 The output rectangle is automatically cropped at the borders of the input
5556 image. Negative values are valid for each component.
5558 output_rect=25\ 25\ 100\ 100
5562 Several filters can be chained for successive processing without GPU-HOST
5563 transfers allowing for fast processing of complex filter chains.
5564 Currently, only filters with zero (generators) or exactly one (filters) input
5565 image and one output image are supported. Also, transition filters are not yet
5568 Some filters generate output images with additional padding depending on the
5569 respective filter kernel. The padding is automatically removed to ensure the
5570 filter output has the same size as the input image.
5572 For image generators, the size of the output image is determined by the
5573 previous output image of the filter chain or the input image of the whole
5574 filterchain, respectively. The generators do not use the pixel information of
5575 this image to generate their output. However, the generated output is
5576 blended onto this image, resulting in partial or complete coverage of the
5579 The @ref{coreimagesrc} video source can be used for generating input images
5580 which are directly fed into the filter chain. By using it, providing input
5581 images by another video source or an input video is not required.
5583 @subsection Examples
5588 List all filters available:
5590 coreimage=list_filters=true
5594 Use the CIBoxBlur filter with default options to blur an image:
5596 coreimage=filter=CIBoxBlur@@default
5600 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5601 its center at 100x100 and a radius of 50 pixels:
5603 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5607 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5608 given as complete and escaped command-line for Apple's standard bash shell:
5610 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5616 Crop the input video to given dimensions.
5618 It accepts the following parameters:
5622 The width of the output video. It defaults to @code{iw}.
5623 This expression is evaluated only once during the filter
5624 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5627 The height of the output video. It defaults to @code{ih}.
5628 This expression is evaluated only once during the filter
5629 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5632 The horizontal position, in the input video, of the left edge of the output
5633 video. It defaults to @code{(in_w-out_w)/2}.
5634 This expression is evaluated per-frame.
5637 The vertical position, in the input video, of the top edge of the output video.
5638 It defaults to @code{(in_h-out_h)/2}.
5639 This expression is evaluated per-frame.
5642 If set to 1 will force the output display aspect ratio
5643 to be the same of the input, by changing the output sample aspect
5644 ratio. It defaults to 0.
5647 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5648 width/height/x/y as specified and will not be rounded to nearest smaller value.
5652 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5653 expressions containing the following constants:
5658 The computed values for @var{x} and @var{y}. They are evaluated for
5663 The input width and height.
5667 These are the same as @var{in_w} and @var{in_h}.
5671 The output (cropped) width and height.
5675 These are the same as @var{out_w} and @var{out_h}.
5678 same as @var{iw} / @var{ih}
5681 input sample aspect ratio
5684 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5688 horizontal and vertical chroma subsample values. For example for the
5689 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5692 The number of the input frame, starting from 0.
5695 the position in the file of the input frame, NAN if unknown
5698 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5702 The expression for @var{out_w} may depend on the value of @var{out_h},
5703 and the expression for @var{out_h} may depend on @var{out_w}, but they
5704 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5705 evaluated after @var{out_w} and @var{out_h}.
5707 The @var{x} and @var{y} parameters specify the expressions for the
5708 position of the top-left corner of the output (non-cropped) area. They
5709 are evaluated for each frame. If the evaluated value is not valid, it
5710 is approximated to the nearest valid value.
5712 The expression for @var{x} may depend on @var{y}, and the expression
5713 for @var{y} may depend on @var{x}.
5715 @subsection Examples
5719 Crop area with size 100x100 at position (12,34).
5724 Using named options, the example above becomes:
5726 crop=w=100:h=100:x=12:y=34
5730 Crop the central input area with size 100x100:
5736 Crop the central input area with size 2/3 of the input video:
5738 crop=2/3*in_w:2/3*in_h
5742 Crop the input video central square:
5749 Delimit the rectangle with the top-left corner placed at position
5750 100:100 and the right-bottom corner corresponding to the right-bottom
5751 corner of the input image.
5753 crop=in_w-100:in_h-100:100:100
5757 Crop 10 pixels from the left and right borders, and 20 pixels from
5758 the top and bottom borders
5760 crop=in_w-2*10:in_h-2*20
5764 Keep only the bottom right quarter of the input image:
5766 crop=in_w/2:in_h/2:in_w/2:in_h/2
5770 Crop height for getting Greek harmony:
5772 crop=in_w:1/PHI*in_w
5776 Apply trembling effect:
5778 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)
5782 Apply erratic camera effect depending on timestamp:
5784 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)"
5788 Set x depending on the value of y:
5790 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5794 @subsection Commands
5796 This filter supports the following commands:
5802 Set width/height of the output video and the horizontal/vertical position
5804 The command accepts the same syntax of the corresponding option.
5806 If the specified expression is not valid, it is kept at its current
5812 Auto-detect the crop size.
5814 It calculates the necessary cropping parameters and prints the
5815 recommended parameters via the logging system. The detected dimensions
5816 correspond to the non-black area of the input video.
5818 It accepts the following parameters:
5823 Set higher black value threshold, which can be optionally specified
5824 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5825 value greater to the set value is considered non-black. It defaults to 24.
5826 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5827 on the bitdepth of the pixel format.
5830 The value which the width/height should be divisible by. It defaults to
5831 16. The offset is automatically adjusted to center the video. Use 2 to
5832 get only even dimensions (needed for 4:2:2 video). 16 is best when
5833 encoding to most video codecs.
5835 @item reset_count, reset
5836 Set the counter that determines after how many frames cropdetect will
5837 reset the previously detected largest video area and start over to
5838 detect the current optimal crop area. Default value is 0.
5840 This can be useful when channel logos distort the video area. 0
5841 indicates 'never reset', and returns the largest area encountered during
5848 Apply color adjustments using curves.
5850 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5851 component (red, green and blue) has its values defined by @var{N} key points
5852 tied from each other using a smooth curve. The x-axis represents the pixel
5853 values from the input frame, and the y-axis the new pixel values to be set for
5856 By default, a component curve is defined by the two points @var{(0;0)} and
5857 @var{(1;1)}. This creates a straight line where each original pixel value is
5858 "adjusted" to its own value, which means no change to the image.
5860 The filter allows you to redefine these two points and add some more. A new
5861 curve (using a natural cubic spline interpolation) will be define to pass
5862 smoothly through all these new coordinates. The new defined points needs to be
5863 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5864 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5865 the vector spaces, the values will be clipped accordingly.
5867 The filter accepts the following options:
5871 Select one of the available color presets. This option can be used in addition
5872 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5873 options takes priority on the preset values.
5874 Available presets are:
5877 @item color_negative
5880 @item increase_contrast
5882 @item linear_contrast
5883 @item medium_contrast
5885 @item strong_contrast
5888 Default is @code{none}.
5890 Set the master key points. These points will define a second pass mapping. It
5891 is sometimes called a "luminance" or "value" mapping. It can be used with
5892 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5893 post-processing LUT.
5895 Set the key points for the red component.
5897 Set the key points for the green component.
5899 Set the key points for the blue component.
5901 Set the key points for all components (not including master).
5902 Can be used in addition to the other key points component
5903 options. In this case, the unset component(s) will fallback on this
5904 @option{all} setting.
5906 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5908 Save Gnuplot script of the curves in specified file.
5911 To avoid some filtergraph syntax conflicts, each key points list need to be
5912 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5914 @subsection Examples
5918 Increase slightly the middle level of blue:
5920 curves=blue='0/0 0.5/0.58 1/1'
5926 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'
5928 Here we obtain the following coordinates for each components:
5931 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5933 @code{(0;0) (0.50;0.48) (1;1)}
5935 @code{(0;0.22) (0.49;0.44) (1;0.80)}
5939 The previous example can also be achieved with the associated built-in preset:
5941 curves=preset=vintage
5951 Use a Photoshop preset and redefine the points of the green component:
5953 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
5957 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
5958 and @command{gnuplot}:
5960 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
5961 gnuplot -p /tmp/curves.plt
5967 Video data analysis filter.
5969 This filter shows hexadecimal pixel values of part of video.
5971 The filter accepts the following options:
5975 Set output video size.
5978 Set x offset from where to pick pixels.
5981 Set y offset from where to pick pixels.
5984 Set scope mode, can be one of the following:
5987 Draw hexadecimal pixel values with white color on black background.
5990 Draw hexadecimal pixel values with input video pixel color on black
5994 Draw hexadecimal pixel values on color background picked from input video,
5995 the text color is picked in such way so its always visible.
5999 Draw rows and columns numbers on left and top of video.
6004 Denoise frames using 2D DCT (frequency domain filtering).
6006 This filter is not designed for real time.
6008 The filter accepts the following options:
6012 Set the noise sigma constant.
6014 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6015 coefficient (absolute value) below this threshold with be dropped.
6017 If you need a more advanced filtering, see @option{expr}.
6019 Default is @code{0}.
6022 Set number overlapping pixels for each block. Since the filter can be slow, you
6023 may want to reduce this value, at the cost of a less effective filter and the
6024 risk of various artefacts.
6026 If the overlapping value doesn't permit processing the whole input width or
6027 height, a warning will be displayed and according borders won't be denoised.
6029 Default value is @var{blocksize}-1, which is the best possible setting.
6032 Set the coefficient factor expression.
6034 For each coefficient of a DCT block, this expression will be evaluated as a
6035 multiplier value for the coefficient.
6037 If this is option is set, the @option{sigma} option will be ignored.
6039 The absolute value of the coefficient can be accessed through the @var{c}
6043 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6044 @var{blocksize}, which is the width and height of the processed blocks.
6046 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6047 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6048 on the speed processing. Also, a larger block size does not necessarily means a
6052 @subsection Examples
6054 Apply a denoise with a @option{sigma} of @code{4.5}:
6059 The same operation can be achieved using the expression system:
6061 dctdnoiz=e='gte(c, 4.5*3)'
6064 Violent denoise using a block size of @code{16x16}:
6071 Remove banding artifacts from input video.
6072 It works by replacing banded pixels with average value of referenced pixels.
6074 The filter accepts the following options:
6081 Set banding detection threshold for each plane. Default is 0.02.
6082 Valid range is 0.00003 to 0.5.
6083 If difference between current pixel and reference pixel is less than threshold,
6084 it will be considered as banded.
6087 Banding detection range in pixels. Default is 16. If positive, random number
6088 in range 0 to set value will be used. If negative, exact absolute value
6090 The range defines square of four pixels around current pixel.
6093 Set direction in radians from which four pixel will be compared. If positive,
6094 random direction from 0 to set direction will be picked. If negative, exact of
6095 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6096 will pick only pixels on same row and -PI/2 will pick only pixels on same
6100 If enabled, current pixel is compared with average value of all four
6101 surrounding pixels. The default is enabled. If disabled current pixel is
6102 compared with all four surrounding pixels. The pixel is considered banded
6103 if only all four differences with surrounding pixels are less than threshold.
6109 Drop duplicated frames at regular intervals.
6111 The filter accepts the following options:
6115 Set the number of frames from which one will be dropped. Setting this to
6116 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6117 Default is @code{5}.
6120 Set the threshold for duplicate detection. If the difference metric for a frame
6121 is less than or equal to this value, then it is declared as duplicate. Default
6125 Set scene change threshold. Default is @code{15}.
6129 Set the size of the x and y-axis blocks used during metric calculations.
6130 Larger blocks give better noise suppression, but also give worse detection of
6131 small movements. Must be a power of two. Default is @code{32}.
6134 Mark main input as a pre-processed input and activate clean source input
6135 stream. This allows the input to be pre-processed with various filters to help
6136 the metrics calculation while keeping the frame selection lossless. When set to
6137 @code{1}, the first stream is for the pre-processed input, and the second
6138 stream is the clean source from where the kept frames are chosen. Default is
6142 Set whether or not chroma is considered in the metric calculations. Default is
6148 Apply deflate effect to the video.
6150 This filter replaces the pixel by the local(3x3) average by taking into account
6151 only values lower than the pixel.
6153 It accepts the following options:
6160 Limit the maximum change for each plane, default is 65535.
6161 If 0, plane will remain unchanged.
6166 Remove judder produced by partially interlaced telecined content.
6168 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6169 source was partially telecined content then the output of @code{pullup,dejudder}
6170 will have a variable frame rate. May change the recorded frame rate of the
6171 container. Aside from that change, this filter will not affect constant frame
6174 The option available in this filter is:
6178 Specify the length of the window over which the judder repeats.
6180 Accepts any integer greater than 1. Useful values are:
6184 If the original was telecined from 24 to 30 fps (Film to NTSC).
6187 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6190 If a mixture of the two.
6193 The default is @samp{4}.
6198 Suppress a TV station logo by a simple interpolation of the surrounding
6199 pixels. Just set a rectangle covering the logo and watch it disappear
6200 (and sometimes something even uglier appear - your mileage may vary).
6202 It accepts the following parameters:
6207 Specify the top left corner coordinates of the logo. They must be
6212 Specify the width and height of the logo to clear. They must be
6216 Specify the thickness of the fuzzy edge of the rectangle (added to
6217 @var{w} and @var{h}). The default value is 1. This option is
6218 deprecated, setting higher values should no longer be necessary and
6222 When set to 1, a green rectangle is drawn on the screen to simplify
6223 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6224 The default value is 0.
6226 The rectangle is drawn on the outermost pixels which will be (partly)
6227 replaced with interpolated values. The values of the next pixels
6228 immediately outside this rectangle in each direction will be used to
6229 compute the interpolated pixel values inside the rectangle.
6233 @subsection Examples
6237 Set a rectangle covering the area with top left corner coordinates 0,0
6238 and size 100x77, and a band of size 10:
6240 delogo=x=0:y=0:w=100:h=77:band=10
6247 Attempt to fix small changes in horizontal and/or vertical shift. This
6248 filter helps remove camera shake from hand-holding a camera, bumping a
6249 tripod, moving on a vehicle, etc.
6251 The filter accepts the following options:
6259 Specify a rectangular area where to limit the search for motion
6261 If desired the search for motion vectors can be limited to a
6262 rectangular area of the frame defined by its top left corner, width
6263 and height. These parameters have the same meaning as the drawbox
6264 filter which can be used to visualise the position of the bounding
6267 This is useful when simultaneous movement of subjects within the frame
6268 might be confused for camera motion by the motion vector search.
6270 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6271 then the full frame is used. This allows later options to be set
6272 without specifying the bounding box for the motion vector search.
6274 Default - search the whole frame.
6278 Specify the maximum extent of movement in x and y directions in the
6279 range 0-64 pixels. Default 16.
6282 Specify how to generate pixels to fill blanks at the edge of the
6283 frame. Available values are:
6286 Fill zeroes at blank locations
6288 Original image at blank locations
6290 Extruded edge value at blank locations
6292 Mirrored edge at blank locations
6294 Default value is @samp{mirror}.
6297 Specify the blocksize to use for motion search. Range 4-128 pixels,
6301 Specify the contrast threshold for blocks. Only blocks with more than
6302 the specified contrast (difference between darkest and lightest
6303 pixels) will be considered. Range 1-255, default 125.
6306 Specify the search strategy. Available values are:
6309 Set exhaustive search
6311 Set less exhaustive search.
6313 Default value is @samp{exhaustive}.
6316 If set then a detailed log of the motion search is written to the
6320 If set to 1, specify using OpenCL capabilities, only available if
6321 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6327 Apply an exact inverse of the telecine operation. It requires a predefined
6328 pattern specified using the pattern option which must be the same as that passed
6329 to the telecine filter.
6331 This filter accepts the following options:
6340 The default value is @code{top}.
6344 A string of numbers representing the pulldown pattern you wish to apply.
6345 The default value is @code{23}.
6348 A number representing position of the first frame with respect to the telecine
6349 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6354 Apply dilation effect to the video.
6356 This filter replaces the pixel by the local(3x3) maximum.
6358 It accepts the following options:
6365 Limit the maximum change for each plane, default is 65535.
6366 If 0, plane will remain unchanged.
6369 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6372 Flags to local 3x3 coordinates maps like this:
6381 Displace pixels as indicated by second and third input stream.
6383 It takes three input streams and outputs one stream, the first input is the
6384 source, and second and third input are displacement maps.
6386 The second input specifies how much to displace pixels along the
6387 x-axis, while the third input specifies how much to displace pixels
6389 If one of displacement map streams terminates, last frame from that
6390 displacement map will be used.
6392 Note that once generated, displacements maps can be reused over and over again.
6394 A description of the accepted options follows.
6398 Set displace behavior for pixels that are out of range.
6400 Available values are:
6403 Missing pixels are replaced by black pixels.
6406 Adjacent pixels will spread out to replace missing pixels.
6409 Out of range pixels are wrapped so they point to pixels of other side.
6411 Default is @samp{smear}.
6415 @subsection Examples
6419 Add ripple effect to rgb input of video size hd720:
6421 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
6425 Add wave effect to rgb input of video size hd720:
6427 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
6433 Draw a colored box on the input image.
6435 It accepts the following parameters:
6440 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6444 The expressions which specify the width and height of the box; if 0 they are interpreted as
6445 the input width and height. It defaults to 0.
6448 Specify the color of the box to write. For the general syntax of this option,
6449 check the "Color" section in the ffmpeg-utils manual. If the special
6450 value @code{invert} is used, the box edge color is the same as the
6451 video with inverted luma.
6454 The expression which sets the thickness of the box edge. Default value is @code{3}.
6456 See below for the list of accepted constants.
6459 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6460 following constants:
6464 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6468 horizontal and vertical chroma subsample values. For example for the
6469 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6473 The input width and height.
6476 The input sample aspect ratio.
6480 The x and y offset coordinates where the box is drawn.
6484 The width and height of the drawn box.
6487 The thickness of the drawn box.
6489 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6490 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6494 @subsection Examples
6498 Draw a black box around the edge of the input image:
6504 Draw a box with color red and an opacity of 50%:
6506 drawbox=10:20:200:60:red@@0.5
6509 The previous example can be specified as:
6511 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6515 Fill the box with pink color:
6517 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6521 Draw a 2-pixel red 2.40:1 mask:
6523 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
6529 Draw a grid on the input image.
6531 It accepts the following parameters:
6536 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6540 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6541 input width and height, respectively, minus @code{thickness}, so image gets
6542 framed. Default to 0.
6545 Specify the color of the grid. For the general syntax of this option,
6546 check the "Color" section in the ffmpeg-utils manual. If the special
6547 value @code{invert} is used, the grid color is the same as the
6548 video with inverted luma.
6551 The expression which sets the thickness of the grid line. Default value is @code{1}.
6553 See below for the list of accepted constants.
6556 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6557 following constants:
6561 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6565 horizontal and vertical chroma subsample values. For example for the
6566 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6570 The input grid cell width and height.
6573 The input sample aspect ratio.
6577 The x and y coordinates of some point of grid intersection (meant to configure offset).
6581 The width and height of the drawn cell.
6584 The thickness of the drawn cell.
6586 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6587 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6591 @subsection Examples
6595 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6597 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6601 Draw a white 3x3 grid with an opacity of 50%:
6603 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6610 Draw a text string or text from a specified file on top of a video, using the
6611 libfreetype library.
6613 To enable compilation of this filter, you need to configure FFmpeg with
6614 @code{--enable-libfreetype}.
6615 To enable default font fallback and the @var{font} option you need to
6616 configure FFmpeg with @code{--enable-libfontconfig}.
6617 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6618 @code{--enable-libfribidi}.
6622 It accepts the following parameters:
6627 Used to draw a box around text using the background color.
6628 The value must be either 1 (enable) or 0 (disable).
6629 The default value of @var{box} is 0.
6632 Set the width of the border to be drawn around the box using @var{boxcolor}.
6633 The default value of @var{boxborderw} is 0.
6636 The color to be used for drawing box around text. For the syntax of this
6637 option, check the "Color" section in the ffmpeg-utils manual.
6639 The default value of @var{boxcolor} is "white".
6642 Set the width of the border to be drawn around the text using @var{bordercolor}.
6643 The default value of @var{borderw} is 0.
6646 Set the color to be used for drawing border around text. For the syntax of this
6647 option, check the "Color" section in the ffmpeg-utils manual.
6649 The default value of @var{bordercolor} is "black".
6652 Select how the @var{text} is expanded. Can be either @code{none},
6653 @code{strftime} (deprecated) or
6654 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6658 If true, check and fix text coords to avoid clipping.
6661 The color to be used for drawing fonts. For the syntax of this option, check
6662 the "Color" section in the ffmpeg-utils manual.
6664 The default value of @var{fontcolor} is "black".
6666 @item fontcolor_expr
6667 String which is expanded the same way as @var{text} to obtain dynamic
6668 @var{fontcolor} value. By default this option has empty value and is not
6669 processed. When this option is set, it overrides @var{fontcolor} option.
6672 The font family to be used for drawing text. By default Sans.
6675 The font file to be used for drawing text. The path must be included.
6676 This parameter is mandatory if the fontconfig support is disabled.
6679 This option does not exist, please see the timeline system
6682 Draw the text applying alpha blending. The value can
6683 be either a number between 0.0 and 1.0
6684 The expression accepts the same variables @var{x, y} do.
6685 The default value is 1.
6686 Please see fontcolor_expr
6689 The font size to be used for drawing text.
6690 The default value of @var{fontsize} is 16.
6693 If set to 1, attempt to shape the text (for example, reverse the order of
6694 right-to-left text and join Arabic characters) before drawing it.
6695 Otherwise, just draw the text exactly as given.
6696 By default 1 (if supported).
6699 The flags to be used for loading the fonts.
6701 The flags map the corresponding flags supported by libfreetype, and are
6702 a combination of the following values:
6709 @item vertical_layout
6710 @item force_autohint
6713 @item ignore_global_advance_width
6715 @item ignore_transform
6721 Default value is "default".
6723 For more information consult the documentation for the FT_LOAD_*
6727 The color to be used for drawing a shadow behind the drawn text. For the
6728 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6730 The default value of @var{shadowcolor} is "black".
6734 The x and y offsets for the text shadow position with respect to the
6735 position of the text. They can be either positive or negative
6736 values. The default value for both is "0".
6739 The starting frame number for the n/frame_num variable. The default value
6743 The size in number of spaces to use for rendering the tab.
6747 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6748 format. It can be used with or without text parameter. @var{timecode_rate}
6749 option must be specified.
6751 @item timecode_rate, rate, r
6752 Set the timecode frame rate (timecode only).
6755 The text string to be drawn. The text must be a sequence of UTF-8
6757 This parameter is mandatory if no file is specified with the parameter
6761 A text file containing text to be drawn. The text must be a sequence
6762 of UTF-8 encoded characters.
6764 This parameter is mandatory if no text string is specified with the
6765 parameter @var{text}.
6767 If both @var{text} and @var{textfile} are specified, an error is thrown.
6770 If set to 1, the @var{textfile} will be reloaded before each frame.
6771 Be sure to update it atomically, or it may be read partially, or even fail.
6775 The expressions which specify the offsets where text will be drawn
6776 within the video frame. They are relative to the top/left border of the
6779 The default value of @var{x} and @var{y} is "0".
6781 See below for the list of accepted constants and functions.
6784 The parameters for @var{x} and @var{y} are expressions containing the
6785 following constants and functions:
6789 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6793 horizontal and vertical chroma subsample values. For example for the
6794 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6797 the height of each text line
6805 @item max_glyph_a, ascent
6806 the maximum distance from the baseline to the highest/upper grid
6807 coordinate used to place a glyph outline point, for all the rendered
6809 It is a positive value, due to the grid's orientation with the Y axis
6812 @item max_glyph_d, descent
6813 the maximum distance from the baseline to the lowest grid coordinate
6814 used to place a glyph outline point, for all the rendered glyphs.
6815 This is a negative value, due to the grid's orientation, with the Y axis
6819 maximum glyph height, that is the maximum height for all the glyphs
6820 contained in the rendered text, it is equivalent to @var{ascent} -
6824 maximum glyph width, that is the maximum width for all the glyphs
6825 contained in the rendered text
6828 the number of input frame, starting from 0
6830 @item rand(min, max)
6831 return a random number included between @var{min} and @var{max}
6834 The input sample aspect ratio.
6837 timestamp expressed in seconds, NAN if the input timestamp is unknown
6840 the height of the rendered text
6843 the width of the rendered text
6847 the x and y offset coordinates where the text is drawn.
6849 These parameters allow the @var{x} and @var{y} expressions to refer
6850 each other, so you can for example specify @code{y=x/dar}.
6853 @anchor{drawtext_expansion}
6854 @subsection Text expansion
6856 If @option{expansion} is set to @code{strftime},
6857 the filter recognizes strftime() sequences in the provided text and
6858 expands them accordingly. Check the documentation of strftime(). This
6859 feature is deprecated.
6861 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6863 If @option{expansion} is set to @code{normal} (which is the default),
6864 the following expansion mechanism is used.
6866 The backslash character @samp{\}, followed by any character, always expands to
6867 the second character.
6869 Sequence of the form @code{%@{...@}} are expanded. The text between the
6870 braces is a function name, possibly followed by arguments separated by ':'.
6871 If the arguments contain special characters or delimiters (':' or '@}'),
6872 they should be escaped.
6874 Note that they probably must also be escaped as the value for the
6875 @option{text} option in the filter argument string and as the filter
6876 argument in the filtergraph description, and possibly also for the shell,
6877 that makes up to four levels of escaping; using a text file avoids these
6880 The following functions are available:
6885 The expression evaluation result.
6887 It must take one argument specifying the expression to be evaluated,
6888 which accepts the same constants and functions as the @var{x} and
6889 @var{y} values. Note that not all constants should be used, for
6890 example the text size is not known when evaluating the expression, so
6891 the constants @var{text_w} and @var{text_h} will have an undefined
6894 @item expr_int_format, eif
6895 Evaluate the expression's value and output as formatted integer.
6897 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6898 The second argument specifies the output format. Allowed values are @samp{x},
6899 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6900 @code{printf} function.
6901 The third parameter is optional and sets the number of positions taken by the output.
6902 It can be used to add padding with zeros from the left.
6905 The time at which the filter is running, expressed in UTC.
6906 It can accept an argument: a strftime() format string.
6909 The time at which the filter is running, expressed in the local time zone.
6910 It can accept an argument: a strftime() format string.
6913 Frame metadata. Takes one or two arguments.
6915 The first argument is mandatory and specifies the metadata key.
6917 The second argument is optional and specifies a default value, used when the
6918 metadata key is not found or empty.
6921 The frame number, starting from 0.
6924 A 1 character description of the current picture type.
6927 The timestamp of the current frame.
6928 It can take up to three arguments.
6930 The first argument is the format of the timestamp; it defaults to @code{flt}
6931 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6932 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6933 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
6934 @code{localtime} stands for the timestamp of the frame formatted as
6935 local time zone time.
6937 The second argument is an offset added to the timestamp.
6939 If the format is set to @code{localtime} or @code{gmtime},
6940 a third argument may be supplied: a strftime() format string.
6941 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
6944 @subsection Examples
6948 Draw "Test Text" with font FreeSerif, using the default values for the
6949 optional parameters.
6952 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
6956 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
6957 and y=50 (counting from the top-left corner of the screen), text is
6958 yellow with a red box around it. Both the text and the box have an
6962 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
6963 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
6966 Note that the double quotes are not necessary if spaces are not used
6967 within the parameter list.
6970 Show the text at the center of the video frame:
6972 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
6976 Show the text at a random position, switching to a new position every 30 seconds:
6978 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)"
6982 Show a text line sliding from right to left in the last row of the video
6983 frame. The file @file{LONG_LINE} is assumed to contain a single line
6986 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
6990 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
6992 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
6996 Draw a single green letter "g", at the center of the input video.
6997 The glyph baseline is placed at half screen height.
6999 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7003 Show text for 1 second every 3 seconds:
7005 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7009 Use fontconfig to set the font. Note that the colons need to be escaped.
7011 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7015 Print the date of a real-time encoding (see strftime(3)):
7017 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7021 Show text fading in and out (appearing/disappearing):
7024 DS=1.0 # display start
7025 DE=10.0 # display end
7026 FID=1.5 # fade in duration
7027 FOD=5 # fade out duration
7028 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 @}"
7033 For more information about libfreetype, check:
7034 @url{http://www.freetype.org/}.
7036 For more information about fontconfig, check:
7037 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7039 For more information about libfribidi, check:
7040 @url{http://fribidi.org/}.
7044 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7046 The filter accepts the following options:
7051 Set low and high threshold values used by the Canny thresholding
7054 The high threshold selects the "strong" edge pixels, which are then
7055 connected through 8-connectivity with the "weak" edge pixels selected
7056 by the low threshold.
7058 @var{low} and @var{high} threshold values must be chosen in the range
7059 [0,1], and @var{low} should be lesser or equal to @var{high}.
7061 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7065 Define the drawing mode.
7069 Draw white/gray wires on black background.
7072 Mix the colors to create a paint/cartoon effect.
7075 Default value is @var{wires}.
7078 @subsection Examples
7082 Standard edge detection with custom values for the hysteresis thresholding:
7084 edgedetect=low=0.1:high=0.4
7088 Painting effect without thresholding:
7090 edgedetect=mode=colormix:high=0
7095 Set brightness, contrast, saturation and approximate gamma adjustment.
7097 The filter accepts the following options:
7101 Set the contrast expression. The value must be a float value in range
7102 @code{-2.0} to @code{2.0}. The default value is "1".
7105 Set the brightness expression. The value must be a float value in
7106 range @code{-1.0} to @code{1.0}. The default value is "0".
7109 Set the saturation expression. The value must be a float in
7110 range @code{0.0} to @code{3.0}. The default value is "1".
7113 Set the gamma expression. The value must be a float in range
7114 @code{0.1} to @code{10.0}. The default value is "1".
7117 Set the gamma expression for red. The value must be a float in
7118 range @code{0.1} to @code{10.0}. The default value is "1".
7121 Set the gamma expression for green. The value must be a float in range
7122 @code{0.1} to @code{10.0}. The default value is "1".
7125 Set the gamma expression for blue. The value must be a float in range
7126 @code{0.1} to @code{10.0}. The default value is "1".
7129 Set the gamma weight expression. It can be used to reduce the effect
7130 of a high gamma value on bright image areas, e.g. keep them from
7131 getting overamplified and just plain white. The value must be a float
7132 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7133 gamma correction all the way down while @code{1.0} leaves it at its
7134 full strength. Default is "1".
7137 Set when the expressions for brightness, contrast, saturation and
7138 gamma expressions are evaluated.
7140 It accepts the following values:
7143 only evaluate expressions once during the filter initialization or
7144 when a command is processed
7147 evaluate expressions for each incoming frame
7150 Default value is @samp{init}.
7153 The expressions accept the following parameters:
7156 frame count of the input frame starting from 0
7159 byte position of the corresponding packet in the input file, NAN if
7163 frame rate of the input video, NAN if the input frame rate is unknown
7166 timestamp expressed in seconds, NAN if the input timestamp is unknown
7169 @subsection Commands
7170 The filter supports the following commands:
7174 Set the contrast expression.
7177 Set the brightness expression.
7180 Set the saturation expression.
7183 Set the gamma expression.
7186 Set the gamma_r expression.
7189 Set gamma_g expression.
7192 Set gamma_b expression.
7195 Set gamma_weight expression.
7197 The command accepts the same syntax of the corresponding option.
7199 If the specified expression is not valid, it is kept at its current
7206 Apply erosion effect to the video.
7208 This filter replaces the pixel by the local(3x3) minimum.
7210 It accepts the following options:
7217 Limit the maximum change for each plane, default is 65535.
7218 If 0, plane will remain unchanged.
7221 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7224 Flags to local 3x3 coordinates maps like this:
7231 @section extractplanes
7233 Extract color channel components from input video stream into
7234 separate grayscale video streams.
7236 The filter accepts the following option:
7240 Set plane(s) to extract.
7242 Available values for planes are:
7253 Choosing planes not available in the input will result in an error.
7254 That means you cannot select @code{r}, @code{g}, @code{b} planes
7255 with @code{y}, @code{u}, @code{v} planes at same time.
7258 @subsection Examples
7262 Extract luma, u and v color channel component from input video frame
7263 into 3 grayscale outputs:
7265 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
7271 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7273 For each input image, the filter will compute the optimal mapping from
7274 the input to the output given the codebook length, that is the number
7275 of distinct output colors.
7277 This filter accepts the following options.
7280 @item codebook_length, l
7281 Set codebook length. The value must be a positive integer, and
7282 represents the number of distinct output colors. Default value is 256.
7285 Set the maximum number of iterations to apply for computing the optimal
7286 mapping. The higher the value the better the result and the higher the
7287 computation time. Default value is 1.
7290 Set a random seed, must be an integer included between 0 and
7291 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7292 will try to use a good random seed on a best effort basis.
7295 Set pal8 output pixel format. This option does not work with codebook
7296 length greater than 256.
7301 Apply a fade-in/out effect to the input video.
7303 It accepts the following parameters:
7307 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7309 Default is @code{in}.
7311 @item start_frame, s
7312 Specify the number of the frame to start applying the fade
7313 effect at. Default is 0.
7316 The number of frames that the fade effect lasts. At the end of the
7317 fade-in effect, the output video will have the same intensity as the input video.
7318 At the end of the fade-out transition, the output video will be filled with the
7319 selected @option{color}.
7323 If set to 1, fade only alpha channel, if one exists on the input.
7326 @item start_time, st
7327 Specify the timestamp (in seconds) of the frame to start to apply the fade
7328 effect. If both start_frame and start_time are specified, the fade will start at
7329 whichever comes last. Default is 0.
7332 The number of seconds for which the fade effect has to last. At the end of the
7333 fade-in effect the output video will have the same intensity as the input video,
7334 at the end of the fade-out transition the output video will be filled with the
7335 selected @option{color}.
7336 If both duration and nb_frames are specified, duration is used. Default is 0
7337 (nb_frames is used by default).
7340 Specify the color of the fade. Default is "black".
7343 @subsection Examples
7347 Fade in the first 30 frames of video:
7352 The command above is equivalent to:
7358 Fade out the last 45 frames of a 200-frame video:
7361 fade=type=out:start_frame=155:nb_frames=45
7365 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7367 fade=in:0:25, fade=out:975:25
7371 Make the first 5 frames yellow, then fade in from frame 5-24:
7373 fade=in:5:20:color=yellow
7377 Fade in alpha over first 25 frames of video:
7379 fade=in:0:25:alpha=1
7383 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7385 fade=t=in:st=5.5:d=0.5
7391 Apply arbitrary expressions to samples in frequency domain
7395 Adjust the dc value (gain) of the luma plane of the image. The filter
7396 accepts an integer value in range @code{0} to @code{1000}. The default
7397 value is set to @code{0}.
7400 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7401 filter accepts an integer value in range @code{0} to @code{1000}. The
7402 default value is set to @code{0}.
7405 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7406 filter accepts an integer value in range @code{0} to @code{1000}. The
7407 default value is set to @code{0}.
7410 Set the frequency domain weight expression for the luma plane.
7413 Set the frequency domain weight expression for the 1st chroma plane.
7416 Set the frequency domain weight expression for the 2nd chroma plane.
7418 The filter accepts the following variables:
7421 The coordinates of the current sample.
7425 The width and height of the image.
7428 @subsection Examples
7434 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7440 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7446 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7452 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7459 Extract a single field from an interlaced image using stride
7460 arithmetic to avoid wasting CPU time. The output frames are marked as
7463 The filter accepts the following options:
7467 Specify whether to extract the top (if the value is @code{0} or
7468 @code{top}) or the bottom field (if the value is @code{1} or
7474 Create new frames by copying the top and bottom fields from surrounding frames
7475 supplied as numbers by the hint file.
7479 Set file containing hints: absolute/relative frame numbers.
7481 There must be one line for each frame in a clip. Each line must contain two
7482 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7483 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7484 is current frame number for @code{absolute} mode or out of [-1, 1] range
7485 for @code{relative} mode. First number tells from which frame to pick up top
7486 field and second number tells from which frame to pick up bottom field.
7488 If optionally followed by @code{+} output frame will be marked as interlaced,
7489 else if followed by @code{-} output frame will be marked as progressive, else
7490 it will be marked same as input frame.
7491 If line starts with @code{#} or @code{;} that line is skipped.
7494 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7497 Example of first several lines of @code{hint} file for @code{relative} mode:
7500 1,0 - # second frame, use third's frame top field and second's frame bottom field
7501 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7518 Field matching filter for inverse telecine. It is meant to reconstruct the
7519 progressive frames from a telecined stream. The filter does not drop duplicated
7520 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7521 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7523 The separation of the field matching and the decimation is notably motivated by
7524 the possibility of inserting a de-interlacing filter fallback between the two.
7525 If the source has mixed telecined and real interlaced content,
7526 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7527 But these remaining combed frames will be marked as interlaced, and thus can be
7528 de-interlaced by a later filter such as @ref{yadif} before decimation.
7530 In addition to the various configuration options, @code{fieldmatch} can take an
7531 optional second stream, activated through the @option{ppsrc} option. If
7532 enabled, the frames reconstruction will be based on the fields and frames from
7533 this second stream. This allows the first input to be pre-processed in order to
7534 help the various algorithms of the filter, while keeping the output lossless
7535 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7536 or brightness/contrast adjustments can help.
7538 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7539 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7540 which @code{fieldmatch} is based on. While the semantic and usage are very
7541 close, some behaviour and options names can differ.
7543 The @ref{decimate} filter currently only works for constant frame rate input.
7544 If your input has mixed telecined (30fps) and progressive content with a lower
7545 framerate like 24fps use the following filterchain to produce the necessary cfr
7546 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7548 The filter accepts the following options:
7552 Specify the assumed field order of the input stream. Available values are:
7556 Auto detect parity (use FFmpeg's internal parity value).
7558 Assume bottom field first.
7560 Assume top field first.
7563 Note that it is sometimes recommended not to trust the parity announced by the
7566 Default value is @var{auto}.
7569 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7570 sense that it won't risk creating jerkiness due to duplicate frames when
7571 possible, but if there are bad edits or blended fields it will end up
7572 outputting combed frames when a good match might actually exist. On the other
7573 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7574 but will almost always find a good frame if there is one. The other values are
7575 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7576 jerkiness and creating duplicate frames versus finding good matches in sections
7577 with bad edits, orphaned fields, blended fields, etc.
7579 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7581 Available values are:
7585 2-way matching (p/c)
7587 2-way matching, and trying 3rd match if still combed (p/c + n)
7589 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7591 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7592 still combed (p/c + n + u/b)
7594 3-way matching (p/c/n)
7596 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7597 detected as combed (p/c/n + u/b)
7600 The parenthesis at the end indicate the matches that would be used for that
7601 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7604 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7607 Default value is @var{pc_n}.
7610 Mark the main input stream as a pre-processed input, and enable the secondary
7611 input stream as the clean source to pick the fields from. See the filter
7612 introduction for more details. It is similar to the @option{clip2} feature from
7615 Default value is @code{0} (disabled).
7618 Set the field to match from. It is recommended to set this to the same value as
7619 @option{order} unless you experience matching failures with that setting. In
7620 certain circumstances changing the field that is used to match from can have a
7621 large impact on matching performance. Available values are:
7625 Automatic (same value as @option{order}).
7627 Match from the bottom field.
7629 Match from the top field.
7632 Default value is @var{auto}.
7635 Set whether or not chroma is included during the match comparisons. In most
7636 cases it is recommended to leave this enabled. You should set this to @code{0}
7637 only if your clip has bad chroma problems such as heavy rainbowing or other
7638 artifacts. Setting this to @code{0} could also be used to speed things up at
7639 the cost of some accuracy.
7641 Default value is @code{1}.
7645 These define an exclusion band which excludes the lines between @option{y0} and
7646 @option{y1} from being included in the field matching decision. An exclusion
7647 band can be used to ignore subtitles, a logo, or other things that may
7648 interfere with the matching. @option{y0} sets the starting scan line and
7649 @option{y1} sets the ending line; all lines in between @option{y0} and
7650 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7651 @option{y0} and @option{y1} to the same value will disable the feature.
7652 @option{y0} and @option{y1} defaults to @code{0}.
7655 Set the scene change detection threshold as a percentage of maximum change on
7656 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7657 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7658 @option{scthresh} is @code{[0.0, 100.0]}.
7660 Default value is @code{12.0}.
7663 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7664 account the combed scores of matches when deciding what match to use as the
7665 final match. Available values are:
7669 No final matching based on combed scores.
7671 Combed scores are only used when a scene change is detected.
7673 Use combed scores all the time.
7676 Default is @var{sc}.
7679 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7680 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7681 Available values are:
7685 No forced calculation.
7687 Force p/c/n calculations.
7689 Force p/c/n/u/b calculations.
7692 Default value is @var{none}.
7695 This is the area combing threshold used for combed frame detection. This
7696 essentially controls how "strong" or "visible" combing must be to be detected.
7697 Larger values mean combing must be more visible and smaller values mean combing
7698 can be less visible or strong and still be detected. Valid settings are from
7699 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7700 be detected as combed). This is basically a pixel difference value. A good
7701 range is @code{[8, 12]}.
7703 Default value is @code{9}.
7706 Sets whether or not chroma is considered in the combed frame decision. Only
7707 disable this if your source has chroma problems (rainbowing, etc.) that are
7708 causing problems for the combed frame detection with chroma enabled. Actually,
7709 using @option{chroma}=@var{0} is usually more reliable, except for the case
7710 where there is chroma only combing in the source.
7712 Default value is @code{0}.
7716 Respectively set the x-axis and y-axis size of the window used during combed
7717 frame detection. This has to do with the size of the area in which
7718 @option{combpel} pixels are required to be detected as combed for a frame to be
7719 declared combed. See the @option{combpel} parameter description for more info.
7720 Possible values are any number that is a power of 2 starting at 4 and going up
7723 Default value is @code{16}.
7726 The number of combed pixels inside any of the @option{blocky} by
7727 @option{blockx} size blocks on the frame for the frame to be detected as
7728 combed. While @option{cthresh} controls how "visible" the combing must be, this
7729 setting controls "how much" combing there must be in any localized area (a
7730 window defined by the @option{blockx} and @option{blocky} settings) on the
7731 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7732 which point no frames will ever be detected as combed). This setting is known
7733 as @option{MI} in TFM/VFM vocabulary.
7735 Default value is @code{80}.
7738 @anchor{p/c/n/u/b meaning}
7739 @subsection p/c/n/u/b meaning
7741 @subsubsection p/c/n
7743 We assume the following telecined stream:
7746 Top fields: 1 2 2 3 4
7747 Bottom fields: 1 2 3 4 4
7750 The numbers correspond to the progressive frame the fields relate to. Here, the
7751 first two frames are progressive, the 3rd and 4th are combed, and so on.
7753 When @code{fieldmatch} is configured to run a matching from bottom
7754 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7759 B 1 2 3 4 4 <-- matching reference
7768 As a result of the field matching, we can see that some frames get duplicated.
7769 To perform a complete inverse telecine, you need to rely on a decimation filter
7770 after this operation. See for instance the @ref{decimate} filter.
7772 The same operation now matching from top fields (@option{field}=@var{top})
7777 T 1 2 2 3 4 <-- matching reference
7787 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7788 basically, they refer to the frame and field of the opposite parity:
7791 @item @var{p} matches the field of the opposite parity in the previous frame
7792 @item @var{c} matches the field of the opposite parity in the current frame
7793 @item @var{n} matches the field of the opposite parity in the next frame
7798 The @var{u} and @var{b} matching are a bit special in the sense that they match
7799 from the opposite parity flag. In the following examples, we assume that we are
7800 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7801 'x' is placed above and below each matched fields.
7803 With bottom matching (@option{field}=@var{bottom}):
7808 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7809 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7817 With top matching (@option{field}=@var{top}):
7822 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7823 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7831 @subsection Examples
7833 Simple IVTC of a top field first telecined stream:
7835 fieldmatch=order=tff:combmatch=none, decimate
7838 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7840 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7845 Transform the field order of the input video.
7847 It accepts the following parameters:
7852 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7853 for bottom field first.
7856 The default value is @samp{tff}.
7858 The transformation is done by shifting the picture content up or down
7859 by one line, and filling the remaining line with appropriate picture content.
7860 This method is consistent with most broadcast field order converters.
7862 If the input video is not flagged as being interlaced, or it is already
7863 flagged as being of the required output field order, then this filter does
7864 not alter the incoming video.
7866 It is very useful when converting to or from PAL DV material,
7867 which is bottom field first.
7871 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7874 @section fifo, afifo
7876 Buffer input images and send them when they are requested.
7878 It is mainly useful when auto-inserted by the libavfilter
7881 It does not take parameters.
7885 Find a rectangular object
7887 It accepts the following options:
7891 Filepath of the object image, needs to be in gray8.
7894 Detection threshold, default is 0.5.
7897 Number of mipmaps, default is 3.
7899 @item xmin, ymin, xmax, ymax
7900 Specifies the rectangle in which to search.
7903 @subsection Examples
7907 Generate a representative palette of a given video using @command{ffmpeg}:
7909 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7915 Cover a rectangular object
7917 It accepts the following options:
7921 Filepath of the optional cover image, needs to be in yuv420.
7926 It accepts the following values:
7929 cover it by the supplied image
7931 cover it by interpolating the surrounding pixels
7934 Default value is @var{blur}.
7937 @subsection Examples
7941 Generate a representative palette of a given video using @command{ffmpeg}:
7943 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7950 Convert the input video to one of the specified pixel formats.
7951 Libavfilter will try to pick one that is suitable as input to
7954 It accepts the following parameters:
7958 A '|'-separated list of pixel format names, such as
7959 "pix_fmts=yuv420p|monow|rgb24".
7963 @subsection Examples
7967 Convert the input video to the @var{yuv420p} format
7969 format=pix_fmts=yuv420p
7972 Convert the input video to any of the formats in the list
7974 format=pix_fmts=yuv420p|yuv444p|yuv410p
7981 Convert the video to specified constant frame rate by duplicating or dropping
7982 frames as necessary.
7984 It accepts the following parameters:
7988 The desired output frame rate. The default is @code{25}.
7993 Possible values are:
7996 zero round towards 0
8000 round towards -infinity
8002 round towards +infinity
8006 The default is @code{near}.
8009 Assume the first PTS should be the given value, in seconds. This allows for
8010 padding/trimming at the start of stream. By default, no assumption is made
8011 about the first frame's expected PTS, so no padding or trimming is done.
8012 For example, this could be set to 0 to pad the beginning with duplicates of
8013 the first frame if a video stream starts after the audio stream or to trim any
8014 frames with a negative PTS.
8018 Alternatively, the options can be specified as a flat string:
8019 @var{fps}[:@var{round}].
8021 See also the @ref{setpts} filter.
8023 @subsection Examples
8027 A typical usage in order to set the fps to 25:
8033 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8035 fps=fps=film:round=near
8041 Pack two different video streams into a stereoscopic video, setting proper
8042 metadata on supported codecs. The two views should have the same size and
8043 framerate and processing will stop when the shorter video ends. Please note
8044 that you may conveniently adjust view properties with the @ref{scale} and
8047 It accepts the following parameters:
8051 The desired packing format. Supported values are:
8056 The views are next to each other (default).
8059 The views are on top of each other.
8062 The views are packed by line.
8065 The views are packed by column.
8068 The views are temporally interleaved.
8077 # Convert left and right views into a frame-sequential video
8078 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8080 # Convert views into a side-by-side video with the same output resolution as the input
8081 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
8086 Change the frame rate by interpolating new video output frames from the source
8089 This filter is not designed to function correctly with interlaced media. If
8090 you wish to change the frame rate of interlaced media then you are required
8091 to deinterlace before this filter and re-interlace after this filter.
8093 A description of the accepted options follows.
8097 Specify the output frames per second. This option can also be specified
8098 as a value alone. The default is @code{50}.
8101 Specify the start of a range where the output frame will be created as a
8102 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8103 the default is @code{15}.
8106 Specify the end of a range where the output frame will be created as a
8107 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8108 the default is @code{240}.
8111 Specify the level at which a scene change is detected as a value between
8112 0 and 100 to indicate a new scene; a low value reflects a low
8113 probability for the current frame to introduce a new scene, while a higher
8114 value means the current frame is more likely to be one.
8115 The default is @code{7}.
8118 Specify flags influencing the filter process.
8120 Available value for @var{flags} is:
8123 @item scene_change_detect, scd
8124 Enable scene change detection using the value of the option @var{scene}.
8125 This flag is enabled by default.
8131 Select one frame every N-th frame.
8133 This filter accepts the following option:
8136 Select frame after every @code{step} frames.
8137 Allowed values are positive integers higher than 0. Default value is @code{1}.
8143 Apply a frei0r effect to the input video.
8145 To enable the compilation of this filter, you need to install the frei0r
8146 header and configure FFmpeg with @code{--enable-frei0r}.
8148 It accepts the following parameters:
8153 The name of the frei0r effect to load. If the environment variable
8154 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8155 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8156 Otherwise, the standard frei0r paths are searched, in this order:
8157 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8158 @file{/usr/lib/frei0r-1/}.
8161 A '|'-separated list of parameters to pass to the frei0r effect.
8165 A frei0r effect parameter can be a boolean (its value is either
8166 "y" or "n"), a double, a color (specified as
8167 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8168 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8169 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8170 @var{X} and @var{Y} are floating point numbers) and/or a string.
8172 The number and types of parameters depend on the loaded effect. If an
8173 effect parameter is not specified, the default value is set.
8175 @subsection Examples
8179 Apply the distort0r effect, setting the first two double parameters:
8181 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8185 Apply the colordistance effect, taking a color as the first parameter:
8187 frei0r=colordistance:0.2/0.3/0.4
8188 frei0r=colordistance:violet
8189 frei0r=colordistance:0x112233
8193 Apply the perspective effect, specifying the top left and top right image
8196 frei0r=perspective:0.2/0.2|0.8/0.2
8200 For more information, see
8201 @url{http://frei0r.dyne.org}
8205 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8207 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8208 processing filter, one of them is performed once per block, not per pixel.
8209 This allows for much higher speed.
8211 The filter accepts the following options:
8215 Set quality. This option defines the number of levels for averaging. It accepts
8216 an integer in the range 4-5. Default value is @code{4}.
8219 Force a constant quantization parameter. It accepts an integer in range 0-63.
8220 If not set, the filter will use the QP from the video stream (if available).
8223 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8224 more details but also more artifacts, while higher values make the image smoother
8225 but also blurrier. Default value is @code{0} − PSNR optimal.
8228 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8229 option may cause flicker since the B-Frames have often larger QP. Default is
8230 @code{0} (not enabled).
8236 The filter accepts the following options:
8240 Set the luminance expression.
8242 Set the chrominance blue expression.
8244 Set the chrominance red expression.
8246 Set the alpha expression.
8248 Set the red expression.
8250 Set the green expression.
8252 Set the blue expression.
8255 The colorspace is selected according to the specified options. If one
8256 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8257 options is specified, the filter will automatically select a YCbCr
8258 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8259 @option{blue_expr} options is specified, it will select an RGB
8262 If one of the chrominance expression is not defined, it falls back on the other
8263 one. If no alpha expression is specified it will evaluate to opaque value.
8264 If none of chrominance expressions are specified, they will evaluate
8265 to the luminance expression.
8267 The expressions can use the following variables and functions:
8271 The sequential number of the filtered frame, starting from @code{0}.
8275 The coordinates of the current sample.
8279 The width and height of the image.
8283 Width and height scale depending on the currently filtered plane. It is the
8284 ratio between the corresponding luma plane number of pixels and the current
8285 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8286 @code{0.5,0.5} for chroma planes.
8289 Time of the current frame, expressed in seconds.
8292 Return the value of the pixel at location (@var{x},@var{y}) of the current
8296 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8300 Return the value of the pixel at location (@var{x},@var{y}) of the
8301 blue-difference chroma plane. Return 0 if there is no such plane.
8304 Return the value of the pixel at location (@var{x},@var{y}) of the
8305 red-difference chroma plane. Return 0 if there is no such plane.
8310 Return the value of the pixel at location (@var{x},@var{y}) of the
8311 red/green/blue component. Return 0 if there is no such component.
8314 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8315 plane. Return 0 if there is no such plane.
8318 For functions, if @var{x} and @var{y} are outside the area, the value will be
8319 automatically clipped to the closer edge.
8321 @subsection Examples
8325 Flip the image horizontally:
8331 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8332 wavelength of 100 pixels:
8334 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8338 Generate a fancy enigmatic moving light:
8340 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
8344 Generate a quick emboss effect:
8346 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8350 Modify RGB components depending on pixel position:
8352 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8356 Create a radial gradient that is the same size as the input (also see
8357 the @ref{vignette} filter):
8359 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8365 Fix the banding artifacts that are sometimes introduced into nearly flat
8366 regions by truncation to 8-bit color depth.
8367 Interpolate the gradients that should go where the bands are, and
8370 It is designed for playback only. Do not use it prior to
8371 lossy compression, because compression tends to lose the dither and
8372 bring back the bands.
8374 It accepts the following parameters:
8379 The maximum amount by which the filter will change any one pixel. This is also
8380 the threshold for detecting nearly flat regions. Acceptable values range from
8381 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8385 The neighborhood to fit the gradient to. A larger radius makes for smoother
8386 gradients, but also prevents the filter from modifying the pixels near detailed
8387 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8388 values will be clipped to the valid range.
8392 Alternatively, the options can be specified as a flat string:
8393 @var{strength}[:@var{radius}]
8395 @subsection Examples
8399 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8405 Specify radius, omitting the strength (which will fall-back to the default
8416 Apply a Hald CLUT to a video stream.
8418 First input is the video stream to process, and second one is the Hald CLUT.
8419 The Hald CLUT input can be a simple picture or a complete video stream.
8421 The filter accepts the following options:
8425 Force termination when the shortest input terminates. Default is @code{0}.
8427 Continue applying the last CLUT after the end of the stream. A value of
8428 @code{0} disable the filter after the last frame of the CLUT is reached.
8429 Default is @code{1}.
8432 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8433 filters share the same internals).
8435 More information about the Hald CLUT can be found on Eskil Steenberg's website
8436 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8438 @subsection Workflow examples
8440 @subsubsection Hald CLUT video stream
8442 Generate an identity Hald CLUT stream altered with various effects:
8444 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
8447 Note: make sure you use a lossless codec.
8449 Then use it with @code{haldclut} to apply it on some random stream:
8451 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8454 The Hald CLUT will be applied to the 10 first seconds (duration of
8455 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8456 to the remaining frames of the @code{mandelbrot} stream.
8458 @subsubsection Hald CLUT with preview
8460 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8461 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8462 biggest possible square starting at the top left of the picture. The remaining
8463 padding pixels (bottom or right) will be ignored. This area can be used to add
8464 a preview of the Hald CLUT.
8466 Typically, the following generated Hald CLUT will be supported by the
8467 @code{haldclut} filter:
8470 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8471 pad=iw+320 [padded_clut];
8472 smptebars=s=320x256, split [a][b];
8473 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8474 [main][b] overlay=W-320" -frames:v 1 clut.png
8477 It contains the original and a preview of the effect of the CLUT: SMPTE color
8478 bars are displayed on the right-top, and below the same color bars processed by
8481 Then, the effect of this Hald CLUT can be visualized with:
8483 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8488 Flip the input video horizontally.
8490 For example, to horizontally flip the input video with @command{ffmpeg}:
8492 ffmpeg -i in.avi -vf "hflip" out.avi
8496 This filter applies a global color histogram equalization on a
8499 It can be used to correct video that has a compressed range of pixel
8500 intensities. The filter redistributes the pixel intensities to
8501 equalize their distribution across the intensity range. It may be
8502 viewed as an "automatically adjusting contrast filter". This filter is
8503 useful only for correcting degraded or poorly captured source
8506 The filter accepts the following options:
8510 Determine the amount of equalization to be applied. As the strength
8511 is reduced, the distribution of pixel intensities more-and-more
8512 approaches that of the input frame. The value must be a float number
8513 in the range [0,1] and defaults to 0.200.
8516 Set the maximum intensity that can generated and scale the output
8517 values appropriately. The strength should be set as desired and then
8518 the intensity can be limited if needed to avoid washing-out. The value
8519 must be a float number in the range [0,1] and defaults to 0.210.
8522 Set the antibanding level. If enabled the filter will randomly vary
8523 the luminance of output pixels by a small amount to avoid banding of
8524 the histogram. Possible values are @code{none}, @code{weak} or
8525 @code{strong}. It defaults to @code{none}.
8530 Compute and draw a color distribution histogram for the input video.
8532 The computed histogram is a representation of the color component
8533 distribution in an image.
8535 Standard histogram displays the color components distribution in an image.
8536 Displays color graph for each color component. Shows distribution of
8537 the Y, U, V, A or R, G, B components, depending on input format, in the
8538 current frame. Below each graph a color component scale meter is shown.
8540 The filter accepts the following options:
8544 Set height of level. Default value is @code{200}.
8545 Allowed range is [50, 2048].
8548 Set height of color scale. Default value is @code{12}.
8549 Allowed range is [0, 40].
8553 It accepts the following values:
8556 Per color component graphs are placed below each other.
8559 Presents information identical to that in the @code{parade}, except
8560 that the graphs representing color components are superimposed directly
8563 Default is @code{parade}.
8566 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8567 Default is @code{linear}.
8570 Set what color components to display.
8571 Default is @code{7}.
8574 Set foreground opacity. Default is @code{0.7}.
8577 Set background opacity. Default is @code{0.5}.
8580 @subsection Examples
8585 Calculate and draw histogram:
8587 ffplay -i input -vf histogram
8595 This is a high precision/quality 3d denoise filter. It aims to reduce
8596 image noise, producing smooth images and making still images really
8597 still. It should enhance compressibility.
8599 It accepts the following optional parameters:
8603 A non-negative floating point number which specifies spatial luma strength.
8606 @item chroma_spatial
8607 A non-negative floating point number which specifies spatial chroma strength.
8608 It defaults to 3.0*@var{luma_spatial}/4.0.
8611 A floating point number which specifies luma temporal strength. It defaults to
8612 6.0*@var{luma_spatial}/4.0.
8615 A floating point number which specifies chroma temporal strength. It defaults to
8616 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8619 @anchor{hwupload_cuda}
8620 @section hwupload_cuda
8622 Upload system memory frames to a CUDA device.
8624 It accepts the following optional parameters:
8628 The number of the CUDA device to use
8633 Apply a high-quality magnification filter designed for pixel art. This filter
8634 was originally created by Maxim Stepin.
8636 It accepts the following option:
8640 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8641 @code{hq3x} and @code{4} for @code{hq4x}.
8642 Default is @code{3}.
8646 Stack input videos horizontally.
8648 All streams must be of same pixel format and of same height.
8650 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8651 to create same output.
8653 The filter accept the following option:
8657 Set number of input streams. Default is 2.
8660 If set to 1, force the output to terminate when the shortest input
8661 terminates. Default value is 0.
8666 Modify the hue and/or the saturation of the input.
8668 It accepts the following parameters:
8672 Specify the hue angle as a number of degrees. It accepts an expression,
8673 and defaults to "0".
8676 Specify the saturation in the [-10,10] range. It accepts an expression and
8680 Specify the hue angle as a number of radians. It accepts an
8681 expression, and defaults to "0".
8684 Specify the brightness in the [-10,10] range. It accepts an expression and
8688 @option{h} and @option{H} are mutually exclusive, and can't be
8689 specified at the same time.
8691 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8692 expressions containing the following constants:
8696 frame count of the input frame starting from 0
8699 presentation timestamp of the input frame expressed in time base units
8702 frame rate of the input video, NAN if the input frame rate is unknown
8705 timestamp expressed in seconds, NAN if the input timestamp is unknown
8708 time base of the input video
8711 @subsection Examples
8715 Set the hue to 90 degrees and the saturation to 1.0:
8721 Same command but expressing the hue in radians:
8727 Rotate hue and make the saturation swing between 0
8728 and 2 over a period of 1 second:
8730 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8734 Apply a 3 seconds saturation fade-in effect starting at 0:
8739 The general fade-in expression can be written as:
8741 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8745 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8747 hue="s=max(0\, min(1\, (8-t)/3))"
8750 The general fade-out expression can be written as:
8752 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8757 @subsection Commands
8759 This filter supports the following commands:
8765 Modify the hue and/or the saturation and/or brightness of the input video.
8766 The command accepts the same syntax of the corresponding option.
8768 If the specified expression is not valid, it is kept at its current
8774 Grow first stream into second stream by connecting components.
8775 This allows to build more robust edge masks.
8777 This filter accepts the following options:
8781 Set which planes will be processed as bitmap, unprocessed planes will be
8782 copied from first stream.
8783 By default value 0xf, all planes will be processed.
8786 Set threshold which is used in filtering. If pixel component value is higher than
8787 this value filter algorithm for connecting components is activated.
8788 By default value is 0.
8793 Detect video interlacing type.
8795 This filter tries to detect if the input frames as interlaced, progressive,
8796 top or bottom field first. It will also try and detect fields that are
8797 repeated between adjacent frames (a sign of telecine).
8799 Single frame detection considers only immediately adjacent frames when classifying each frame.
8800 Multiple frame detection incorporates the classification history of previous frames.
8802 The filter will log these metadata values:
8805 @item single.current_frame
8806 Detected type of current frame using single-frame detection. One of:
8807 ``tff'' (top field first), ``bff'' (bottom field first),
8808 ``progressive'', or ``undetermined''
8811 Cumulative number of frames detected as top field first using single-frame detection.
8814 Cumulative number of frames detected as top field first using multiple-frame detection.
8817 Cumulative number of frames detected as bottom field first using single-frame detection.
8819 @item multiple.current_frame
8820 Detected type of current frame using multiple-frame detection. One of:
8821 ``tff'' (top field first), ``bff'' (bottom field first),
8822 ``progressive'', or ``undetermined''
8825 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8827 @item single.progressive
8828 Cumulative number of frames detected as progressive using single-frame detection.
8830 @item multiple.progressive
8831 Cumulative number of frames detected as progressive using multiple-frame detection.
8833 @item single.undetermined
8834 Cumulative number of frames that could not be classified using single-frame detection.
8836 @item multiple.undetermined
8837 Cumulative number of frames that could not be classified using multiple-frame detection.
8839 @item repeated.current_frame
8840 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8842 @item repeated.neither
8843 Cumulative number of frames with no repeated field.
8846 Cumulative number of frames with the top field repeated from the previous frame's top field.
8848 @item repeated.bottom
8849 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8852 The filter accepts the following options:
8856 Set interlacing threshold.
8858 Set progressive threshold.
8860 Threshold for repeated field detection.
8862 Number of frames after which a given frame's contribution to the
8863 statistics is halved (i.e., it contributes only 0.5 to it's
8864 classification). The default of 0 means that all frames seen are given
8865 full weight of 1.0 forever.
8866 @item analyze_interlaced_flag
8867 When this is not 0 then idet will use the specified number of frames to determine
8868 if the interlaced flag is accurate, it will not count undetermined frames.
8869 If the flag is found to be accurate it will be used without any further
8870 computations, if it is found to be inaccurate it will be cleared without any
8871 further computations. This allows inserting the idet filter as a low computational
8872 method to clean up the interlaced flag
8877 Deinterleave or interleave fields.
8879 This filter allows one to process interlaced images fields without
8880 deinterlacing them. Deinterleaving splits the input frame into 2
8881 fields (so called half pictures). Odd lines are moved to the top
8882 half of the output image, even lines to the bottom half.
8883 You can process (filter) them independently and then re-interleave them.
8885 The filter accepts the following options:
8889 @item chroma_mode, c
8891 Available values for @var{luma_mode}, @var{chroma_mode} and
8892 @var{alpha_mode} are:
8898 @item deinterleave, d
8899 Deinterleave fields, placing one above the other.
8902 Interleave fields. Reverse the effect of deinterleaving.
8904 Default value is @code{none}.
8907 @item chroma_swap, cs
8908 @item alpha_swap, as
8909 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
8914 Apply inflate effect to the video.
8916 This filter replaces the pixel by the local(3x3) average by taking into account
8917 only values higher than the pixel.
8919 It accepts the following options:
8926 Limit the maximum change for each plane, default is 65535.
8927 If 0, plane will remain unchanged.
8932 Simple interlacing filter from progressive contents. This interleaves upper (or
8933 lower) lines from odd frames with lower (or upper) lines from even frames,
8934 halving the frame rate and preserving image height.
8937 Original Original New Frame
8938 Frame 'j' Frame 'j+1' (tff)
8939 ========== =========== ==================
8940 Line 0 --------------------> Frame 'j' Line 0
8941 Line 1 Line 1 ----> Frame 'j+1' Line 1
8942 Line 2 ---------------------> Frame 'j' Line 2
8943 Line 3 Line 3 ----> Frame 'j+1' Line 3
8945 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
8948 It accepts the following optional parameters:
8952 This determines whether the interlaced frame is taken from the even
8953 (tff - default) or odd (bff) lines of the progressive frame.
8956 Enable (default) or disable the vertical lowpass filter to avoid twitter
8957 interlacing and reduce moire patterns.
8962 Deinterlace input video by applying Donald Graft's adaptive kernel
8963 deinterling. Work on interlaced parts of a video to produce
8966 The description of the accepted parameters follows.
8970 Set the threshold which affects the filter's tolerance when
8971 determining if a pixel line must be processed. It must be an integer
8972 in the range [0,255] and defaults to 10. A value of 0 will result in
8973 applying the process on every pixels.
8976 Paint pixels exceeding the threshold value to white if set to 1.
8980 Set the fields order. Swap fields if set to 1, leave fields alone if
8984 Enable additional sharpening if set to 1. Default is 0.
8987 Enable twoway sharpening if set to 1. Default is 0.
8990 @subsection Examples
8994 Apply default values:
8996 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9000 Enable additional sharpening:
9006 Paint processed pixels in white:
9012 @section lenscorrection
9014 Correct radial lens distortion
9016 This filter can be used to correct for radial distortion as can result from the use
9017 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9018 one can use tools available for example as part of opencv or simply trial-and-error.
9019 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9020 and extract the k1 and k2 coefficients from the resulting matrix.
9022 Note that effectively the same filter is available in the open-source tools Krita and
9023 Digikam from the KDE project.
9025 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9026 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9027 brightness distribution, so you may want to use both filters together in certain
9028 cases, though you will have to take care of ordering, i.e. whether vignetting should
9029 be applied before or after lens correction.
9033 The filter accepts the following options:
9037 Relative x-coordinate of the focal point of the image, and thereby the center of the
9038 distortion. This value has a range [0,1] and is expressed as fractions of the image
9041 Relative y-coordinate of the focal point of the image, and thereby the center of the
9042 distortion. This value has a range [0,1] and is expressed as fractions of the image
9045 Coefficient of the quadratic correction term. 0.5 means no correction.
9047 Coefficient of the double quadratic correction term. 0.5 means no correction.
9050 The formula that generates the correction is:
9052 @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)
9054 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9055 distances from the focal point in the source and target images, respectively.
9061 The filter accepts the following options:
9065 Set the number of loops.
9068 Set maximal size in number of frames.
9071 Set first frame of loop.
9077 Apply a 3D LUT to an input video.
9079 The filter accepts the following options:
9083 Set the 3D LUT file name.
9085 Currently supported formats:
9097 Select interpolation mode.
9099 Available values are:
9103 Use values from the nearest defined point.
9105 Interpolate values using the 8 points defining a cube.
9107 Interpolate values using a tetrahedron.
9111 @section lut, lutrgb, lutyuv
9113 Compute a look-up table for binding each pixel component input value
9114 to an output value, and apply it to the input video.
9116 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9117 to an RGB input video.
9119 These filters accept the following parameters:
9122 set first pixel component expression
9124 set second pixel component expression
9126 set third pixel component expression
9128 set fourth pixel component expression, corresponds to the alpha component
9131 set red component expression
9133 set green component expression
9135 set blue component expression
9137 alpha component expression
9140 set Y/luminance component expression
9142 set U/Cb component expression
9144 set V/Cr component expression
9147 Each of them specifies the expression to use for computing the lookup table for
9148 the corresponding pixel component values.
9150 The exact component associated to each of the @var{c*} options depends on the
9153 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9154 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9156 The expressions can contain the following constants and functions:
9161 The input width and height.
9164 The input value for the pixel component.
9167 The input value, clipped to the @var{minval}-@var{maxval} range.
9170 The maximum value for the pixel component.
9173 The minimum value for the pixel component.
9176 The negated value for the pixel component value, clipped to the
9177 @var{minval}-@var{maxval} range; it corresponds to the expression
9178 "maxval-clipval+minval".
9181 The computed value in @var{val}, clipped to the
9182 @var{minval}-@var{maxval} range.
9184 @item gammaval(gamma)
9185 The computed gamma correction value of the pixel component value,
9186 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9188 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9192 All expressions default to "val".
9194 @subsection Examples
9200 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9201 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9204 The above is the same as:
9206 lutrgb="r=negval:g=negval:b=negval"
9207 lutyuv="y=negval:u=negval:v=negval"
9217 Remove chroma components, turning the video into a graytone image:
9219 lutyuv="u=128:v=128"
9223 Apply a luma burning effect:
9229 Remove green and blue components:
9235 Set a constant alpha channel value on input:
9237 format=rgba,lutrgb=a="maxval-minval/2"
9241 Correct luminance gamma by a factor of 0.5:
9243 lutyuv=y=gammaval(0.5)
9247 Discard least significant bits of luma:
9249 lutyuv=y='bitand(val, 128+64+32)'
9253 Technicolor like effect:
9255 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9261 Compute and apply a lookup table from two video inputs.
9263 This filter accepts the following parameters:
9266 set first pixel component expression
9268 set second pixel component expression
9270 set third pixel component expression
9272 set fourth pixel component expression, corresponds to the alpha component
9275 Each of them specifies the expression to use for computing the lookup table for
9276 the corresponding pixel component values.
9278 The exact component associated to each of the @var{c*} options depends on the
9281 The expressions can contain the following constants:
9286 The input width and height.
9289 The first input value for the pixel component.
9292 The second input value for the pixel component.
9295 All expressions default to "x".
9297 @section maskedclamp
9299 Clamp the first input stream with the second input and third input stream.
9301 Returns the value of first stream to be between second input
9302 stream - @code{undershoot} and third input stream + @code{overshoot}.
9304 This filter accepts the following options:
9307 Default value is @code{0}.
9310 Default value is @code{0}.
9313 Set which planes will be processed as bitmap, unprocessed planes will be
9314 copied from first stream.
9315 By default value 0xf, all planes will be processed.
9318 @section maskedmerge
9320 Merge the first input stream with the second input stream using per pixel
9321 weights in the third input stream.
9323 A value of 0 in the third stream pixel component means that pixel component
9324 from first stream is returned unchanged, while maximum value (eg. 255 for
9325 8-bit videos) means that pixel component from second stream is returned
9326 unchanged. Intermediate values define the amount of merging between both
9327 input stream's pixel components.
9329 This filter accepts the following options:
9332 Set which planes will be processed as bitmap, unprocessed planes will be
9333 copied from first stream.
9334 By default value 0xf, all planes will be processed.
9339 Apply motion-compensation deinterlacing.
9341 It needs one field per frame as input and must thus be used together
9342 with yadif=1/3 or equivalent.
9344 This filter accepts the following options:
9347 Set the deinterlacing mode.
9349 It accepts one of the following values:
9354 use iterative motion estimation
9356 like @samp{slow}, but use multiple reference frames.
9358 Default value is @samp{fast}.
9361 Set the picture field parity assumed for the input video. It must be
9362 one of the following values:
9366 assume top field first
9368 assume bottom field first
9371 Default value is @samp{bff}.
9374 Set per-block quantization parameter (QP) used by the internal
9377 Higher values should result in a smoother motion vector field but less
9378 optimal individual vectors. Default value is 1.
9381 @section mergeplanes
9383 Merge color channel components from several video streams.
9385 The filter accepts up to 4 input streams, and merge selected input
9386 planes to the output video.
9388 This filter accepts the following options:
9391 Set input to output plane mapping. Default is @code{0}.
9393 The mappings is specified as a bitmap. It should be specified as a
9394 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9395 mapping for the first plane of the output stream. 'A' sets the number of
9396 the input stream to use (from 0 to 3), and 'a' the plane number of the
9397 corresponding input to use (from 0 to 3). The rest of the mappings is
9398 similar, 'Bb' describes the mapping for the output stream second
9399 plane, 'Cc' describes the mapping for the output stream third plane and
9400 'Dd' describes the mapping for the output stream fourth plane.
9403 Set output pixel format. Default is @code{yuva444p}.
9406 @subsection Examples
9410 Merge three gray video streams of same width and height into single video stream:
9412 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9416 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9418 [a0][a1]mergeplanes=0x00010210:yuva444p
9422 Swap Y and A plane in yuva444p stream:
9424 format=yuva444p,mergeplanes=0x03010200:yuva444p
9428 Swap U and V plane in yuv420p stream:
9430 format=yuv420p,mergeplanes=0x000201:yuv420p
9434 Cast a rgb24 clip to yuv444p:
9436 format=rgb24,mergeplanes=0x000102:yuv444p
9442 Estimate and export motion vectors using block matching algorithms.
9443 Motion vectors are stored in frame side data to be used by other filters.
9445 This filter accepts the following options:
9448 Specify the motion estimation method. Accepts one of the following values:
9452 Exhaustive search algorithm.
9454 Three step search algorithm.
9456 Two dimensional logarithmic search algorithm.
9458 New three step search algorithm.
9460 Four step search algorithm.
9462 Diamond search algorithm.
9464 Hexagon-based search algorithm.
9466 Enhanced predictive zonal search algorithm.
9468 Uneven multi-hexagon search algorithm.
9470 Default value is @samp{esa}.
9473 Macroblock size. Default @code{16}.
9476 Search parameter. Default @code{7}.
9479 @section minterpolate
9481 Convert the video to specified frame rate using motion interpolation.
9483 This filter accepts the following options:
9486 Specify the output frame rate. This can be rational e.g. @code{60000/1001}. Frames are dropped if @var{fps} is lower than source fps. Default @code{60}.
9489 Motion interpolation mode. Following values are accepted:
9492 Duplicate previous or next frame for interpolating new ones.
9494 Blend source frames. Interpolated frame is mean of previous and next frames.
9496 Motion compensated interpolation. Following options are effective when this mode is selected:
9500 Motion compensation mode. Following values are accepted:
9503 Overlapped block motion compensation.
9505 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
9507 Default mode is @samp{obmc}.
9510 Motion estimation mode. Following values are accepted:
9513 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
9515 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
9517 Default mode is @samp{bilat}.
9520 The algorithm to be used for motion estimation. Following values are accepted:
9523 Exhaustive search algorithm.
9525 Three step search algorithm.
9527 Two dimensional logarithmic search algorithm.
9529 New three step search algorithm.
9531 Four step search algorithm.
9533 Diamond search algorithm.
9535 Hexagon-based search algorithm.
9537 Enhanced predictive zonal search algorithm.
9539 Uneven multi-hexagon search algorithm.
9541 Default algorithm is @samp{epzs}.
9544 Macroblock size. Default @code{16}.
9547 Motion estimation search parameter. Default @code{32}.
9550 Enable variable-size block motion compensation. Motion estimation is applied with smaller block sizes at object boundaries in order to make the them less blur. Default is @code{0} (disabled).
9555 Scene change detection method. Scene change leads motion vectors to be in random direction. Scene change detection replace interpolated frames by duplicate ones. May not be needed for other modes. Following values are accepted:
9558 Disable scene change detection.
9560 Frame difference. Corresponding pixel values are compared and if it statisfies @var{scd_threshold} scene change is detected.
9562 Default method is @samp{fdiff}.
9565 Scene change detection threshold. Default is @code{5.0}.
9570 Drop frames that do not differ greatly from the previous frame in
9571 order to reduce frame rate.
9573 The main use of this filter is for very-low-bitrate encoding
9574 (e.g. streaming over dialup modem), but it could in theory be used for
9575 fixing movies that were inverse-telecined incorrectly.
9577 A description of the accepted options follows.
9581 Set the maximum number of consecutive frames which can be dropped (if
9582 positive), or the minimum interval between dropped frames (if
9583 negative). If the value is 0, the frame is dropped unregarding the
9584 number of previous sequentially dropped frames.
9591 Set the dropping threshold values.
9593 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9594 represent actual pixel value differences, so a threshold of 64
9595 corresponds to 1 unit of difference for each pixel, or the same spread
9596 out differently over the block.
9598 A frame is a candidate for dropping if no 8x8 blocks differ by more
9599 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9600 meaning the whole image) differ by more than a threshold of @option{lo}.
9602 Default value for @option{hi} is 64*12, default value for @option{lo} is
9603 64*5, and default value for @option{frac} is 0.33.
9611 It accepts an integer in input; if non-zero it negates the
9612 alpha component (if available). The default value in input is 0.
9616 Deinterlace video using neural network edge directed interpolation.
9618 This filter accepts the following options:
9622 Mandatory option, without binary file filter can not work.
9623 Currently file can be found here:
9624 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9627 Set which frames to deinterlace, by default it is @code{all}.
9628 Can be @code{all} or @code{interlaced}.
9631 Set mode of operation.
9633 Can be one of the following:
9637 Use frame flags, both fields.
9639 Use frame flags, single field.
9643 Use bottom field only.
9645 Use both fields, top first.
9647 Use both fields, bottom first.
9651 Set which planes to process, by default filter process all frames.
9654 Set size of local neighborhood around each pixel, used by the predictor neural
9657 Can be one of the following:
9670 Set the number of neurons in predicctor neural network.
9671 Can be one of the following:
9682 Controls the number of different neural network predictions that are blended
9683 together to compute the final output value. Can be @code{fast}, default or
9687 Set which set of weights to use in the predictor.
9688 Can be one of the following:
9692 weights trained to minimize absolute error
9694 weights trained to minimize squared error
9698 Controls whether or not the prescreener neural network is used to decide
9699 which pixels should be processed by the predictor neural network and which
9700 can be handled by simple cubic interpolation.
9701 The prescreener is trained to know whether cubic interpolation will be
9702 sufficient for a pixel or whether it should be predicted by the predictor nn.
9703 The computational complexity of the prescreener nn is much less than that of
9704 the predictor nn. Since most pixels can be handled by cubic interpolation,
9705 using the prescreener generally results in much faster processing.
9706 The prescreener is pretty accurate, so the difference between using it and not
9707 using it is almost always unnoticeable.
9709 Can be one of the following:
9717 Default is @code{new}.
9720 Set various debugging flags.
9725 Force libavfilter not to use any of the specified pixel formats for the
9726 input to the next filter.
9728 It accepts the following parameters:
9732 A '|'-separated list of pixel format names, such as
9733 apix_fmts=yuv420p|monow|rgb24".
9737 @subsection Examples
9741 Force libavfilter to use a format different from @var{yuv420p} for the
9742 input to the vflip filter:
9744 noformat=pix_fmts=yuv420p,vflip
9748 Convert the input video to any of the formats not contained in the list:
9750 noformat=yuv420p|yuv444p|yuv410p
9756 Add noise on video input frame.
9758 The filter accepts the following options:
9766 Set noise seed for specific pixel component or all pixel components in case
9767 of @var{all_seed}. Default value is @code{123457}.
9769 @item all_strength, alls
9770 @item c0_strength, c0s
9771 @item c1_strength, c1s
9772 @item c2_strength, c2s
9773 @item c3_strength, c3s
9774 Set noise strength for specific pixel component or all pixel components in case
9775 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9777 @item all_flags, allf
9782 Set pixel component flags or set flags for all components if @var{all_flags}.
9783 Available values for component flags are:
9786 averaged temporal noise (smoother)
9788 mix random noise with a (semi)regular pattern
9790 temporal noise (noise pattern changes between frames)
9792 uniform noise (gaussian otherwise)
9796 @subsection Examples
9798 Add temporal and uniform noise to input video:
9800 noise=alls=20:allf=t+u
9805 Pass the video source unchanged to the output.
9808 Optical Character Recognition
9810 This filter uses Tesseract for optical character recognition.
9812 It accepts the following options:
9816 Set datapath to tesseract data. Default is to use whatever was
9817 set at installation.
9820 Set language, default is "eng".
9823 Set character whitelist.
9826 Set character blacklist.
9829 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9833 Apply a video transform using libopencv.
9835 To enable this filter, install the libopencv library and headers and
9836 configure FFmpeg with @code{--enable-libopencv}.
9838 It accepts the following parameters:
9843 The name of the libopencv filter to apply.
9846 The parameters to pass to the libopencv filter. If not specified, the default
9851 Refer to the official libopencv documentation for more precise
9853 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9855 Several libopencv filters are supported; see the following subsections.
9860 Dilate an image by using a specific structuring element.
9861 It corresponds to the libopencv function @code{cvDilate}.
9863 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
9865 @var{struct_el} represents a structuring element, and has the syntax:
9866 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
9868 @var{cols} and @var{rows} represent the number of columns and rows of
9869 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
9870 point, and @var{shape} the shape for the structuring element. @var{shape}
9871 must be "rect", "cross", "ellipse", or "custom".
9873 If the value for @var{shape} is "custom", it must be followed by a
9874 string of the form "=@var{filename}". The file with name
9875 @var{filename} is assumed to represent a binary image, with each
9876 printable character corresponding to a bright pixel. When a custom
9877 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
9878 or columns and rows of the read file are assumed instead.
9880 The default value for @var{struct_el} is "3x3+0x0/rect".
9882 @var{nb_iterations} specifies the number of times the transform is
9883 applied to the image, and defaults to 1.
9887 # Use the default values
9890 # Dilate using a structuring element with a 5x5 cross, iterating two times
9891 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
9893 # Read the shape from the file diamond.shape, iterating two times.
9894 # The file diamond.shape may contain a pattern of characters like this
9900 # The specified columns and rows are ignored
9901 # but the anchor point coordinates are not
9902 ocv=dilate:0x0+2x2/custom=diamond.shape|2
9907 Erode an image by using a specific structuring element.
9908 It corresponds to the libopencv function @code{cvErode}.
9910 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
9911 with the same syntax and semantics as the @ref{dilate} filter.
9915 Smooth the input video.
9917 The filter takes the following parameters:
9918 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
9920 @var{type} is the type of smooth filter to apply, and must be one of
9921 the following values: "blur", "blur_no_scale", "median", "gaussian",
9922 or "bilateral". The default value is "gaussian".
9924 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
9925 depend on the smooth type. @var{param1} and
9926 @var{param2} accept integer positive values or 0. @var{param3} and
9927 @var{param4} accept floating point values.
9929 The default value for @var{param1} is 3. The default value for the
9930 other parameters is 0.
9932 These parameters correspond to the parameters assigned to the
9933 libopencv function @code{cvSmooth}.
9938 Overlay one video on top of another.
9940 It takes two inputs and has one output. The first input is the "main"
9941 video on which the second input is overlaid.
9943 It accepts the following parameters:
9945 A description of the accepted options follows.
9950 Set the expression for the x and y coordinates of the overlaid video
9951 on the main video. Default value is "0" for both expressions. In case
9952 the expression is invalid, it is set to a huge value (meaning that the
9953 overlay will not be displayed within the output visible area).
9956 The action to take when EOF is encountered on the secondary input; it accepts
9957 one of the following values:
9961 Repeat the last frame (the default).
9965 Pass the main input through.
9969 Set when the expressions for @option{x}, and @option{y} are evaluated.
9971 It accepts the following values:
9974 only evaluate expressions once during the filter initialization or
9975 when a command is processed
9978 evaluate expressions for each incoming frame
9981 Default value is @samp{frame}.
9984 If set to 1, force the output to terminate when the shortest input
9985 terminates. Default value is 0.
9988 Set the format for the output video.
9990 It accepts the following values:
10005 Default value is @samp{yuv420}.
10007 @item rgb @emph{(deprecated)}
10008 If set to 1, force the filter to accept inputs in the RGB
10009 color space. Default value is 0. This option is deprecated, use
10010 @option{format} instead.
10013 If set to 1, force the filter to draw the last overlay frame over the
10014 main input until the end of the stream. A value of 0 disables this
10015 behavior. Default value is 1.
10018 The @option{x}, and @option{y} expressions can contain the following
10024 The main input width and height.
10028 The overlay input width and height.
10032 The computed values for @var{x} and @var{y}. They are evaluated for
10037 horizontal and vertical chroma subsample values of the output
10038 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
10042 the number of input frame, starting from 0
10045 the position in the file of the input frame, NAN if unknown
10048 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
10052 Note that the @var{n}, @var{pos}, @var{t} variables are available only
10053 when evaluation is done @emph{per frame}, and will evaluate to NAN
10054 when @option{eval} is set to @samp{init}.
10056 Be aware that frames are taken from each input video in timestamp
10057 order, hence, if their initial timestamps differ, it is a good idea
10058 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
10059 have them begin in the same zero timestamp, as the example for
10060 the @var{movie} filter does.
10062 You can chain together more overlays but you should test the
10063 efficiency of such approach.
10065 @subsection Commands
10067 This filter supports the following commands:
10071 Modify the x and y of the overlay input.
10072 The command accepts the same syntax of the corresponding option.
10074 If the specified expression is not valid, it is kept at its current
10078 @subsection Examples
10082 Draw the overlay at 10 pixels from the bottom right corner of the main
10085 overlay=main_w-overlay_w-10:main_h-overlay_h-10
10088 Using named options the example above becomes:
10090 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
10094 Insert a transparent PNG logo in the bottom left corner of the input,
10095 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
10097 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
10101 Insert 2 different transparent PNG logos (second logo on bottom
10102 right corner) using the @command{ffmpeg} tool:
10104 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
10108 Add a transparent color layer on top of the main video; @code{WxH}
10109 must specify the size of the main input to the overlay filter:
10111 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
10115 Play an original video and a filtered version (here with the deshake
10116 filter) side by side using the @command{ffplay} tool:
10118 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
10121 The above command is the same as:
10123 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
10127 Make a sliding overlay appearing from the left to the right top part of the
10128 screen starting since time 2:
10130 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
10134 Compose output by putting two input videos side to side:
10136 ffmpeg -i left.avi -i right.avi -filter_complex "
10137 nullsrc=size=200x100 [background];
10138 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
10139 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
10140 [background][left] overlay=shortest=1 [background+left];
10141 [background+left][right] overlay=shortest=1:x=100 [left+right]
10146 Mask 10-20 seconds of a video by applying the delogo filter to a section
10148 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
10149 -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]'
10154 Chain several overlays in cascade:
10156 nullsrc=s=200x200 [bg];
10157 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
10158 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
10159 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
10160 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
10161 [in3] null, [mid2] overlay=100:100 [out0]
10168 Apply Overcomplete Wavelet denoiser.
10170 The filter accepts the following options:
10176 Larger depth values will denoise lower frequency components more, but
10177 slow down filtering.
10179 Must be an int in the range 8-16, default is @code{8}.
10181 @item luma_strength, ls
10184 Must be a double value in the range 0-1000, default is @code{1.0}.
10186 @item chroma_strength, cs
10187 Set chroma strength.
10189 Must be a double value in the range 0-1000, default is @code{1.0}.
10195 Add paddings to the input image, and place the original input at the
10196 provided @var{x}, @var{y} coordinates.
10198 It accepts the following parameters:
10203 Specify an expression for the size of the output image with the
10204 paddings added. If the value for @var{width} or @var{height} is 0, the
10205 corresponding input size is used for the output.
10207 The @var{width} expression can reference the value set by the
10208 @var{height} expression, and vice versa.
10210 The default value of @var{width} and @var{height} is 0.
10214 Specify the offsets to place the input image at within the padded area,
10215 with respect to the top/left border of the output image.
10217 The @var{x} expression can reference the value set by the @var{y}
10218 expression, and vice versa.
10220 The default value of @var{x} and @var{y} is 0.
10223 Specify the color of the padded area. For the syntax of this option,
10224 check the "Color" section in the ffmpeg-utils manual.
10226 The default value of @var{color} is "black".
10229 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10230 options are expressions containing the following constants:
10235 The input video width and height.
10239 These are the same as @var{in_w} and @var{in_h}.
10243 The output width and height (the size of the padded area), as
10244 specified by the @var{width} and @var{height} expressions.
10248 These are the same as @var{out_w} and @var{out_h}.
10252 The x and y offsets as specified by the @var{x} and @var{y}
10253 expressions, or NAN if not yet specified.
10256 same as @var{iw} / @var{ih}
10259 input sample aspect ratio
10262 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10266 The horizontal and vertical chroma subsample values. For example for the
10267 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10270 @subsection Examples
10274 Add paddings with the color "violet" to the input video. The output video
10275 size is 640x480, and the top-left corner of the input video is placed at
10278 pad=640:480:0:40:violet
10281 The example above is equivalent to the following command:
10283 pad=width=640:height=480:x=0:y=40:color=violet
10287 Pad the input to get an output with dimensions increased by 3/2,
10288 and put the input video at the center of the padded area:
10290 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10294 Pad the input to get a squared output with size equal to the maximum
10295 value between the input width and height, and put the input video at
10296 the center of the padded area:
10298 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10302 Pad the input to get a final w/h ratio of 16:9:
10304 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10308 In case of anamorphic video, in order to set the output display aspect
10309 correctly, it is necessary to use @var{sar} in the expression,
10310 according to the relation:
10312 (ih * X / ih) * sar = output_dar
10313 X = output_dar / sar
10316 Thus the previous example needs to be modified to:
10318 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10322 Double the output size and put the input video in the bottom-right
10323 corner of the output padded area:
10325 pad="2*iw:2*ih:ow-iw:oh-ih"
10329 @anchor{palettegen}
10330 @section palettegen
10332 Generate one palette for a whole video stream.
10334 It accepts the following options:
10338 Set the maximum number of colors to quantize in the palette.
10339 Note: the palette will still contain 256 colors; the unused palette entries
10342 @item reserve_transparent
10343 Create a palette of 255 colors maximum and reserve the last one for
10344 transparency. Reserving the transparency color is useful for GIF optimization.
10345 If not set, the maximum of colors in the palette will be 256. You probably want
10346 to disable this option for a standalone image.
10350 Set statistics mode.
10352 It accepts the following values:
10355 Compute full frame histograms.
10357 Compute histograms only for the part that differs from previous frame. This
10358 might be relevant to give more importance to the moving part of your input if
10359 the background is static.
10362 Default value is @var{full}.
10365 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10366 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10367 color quantization of the palette. This information is also visible at
10368 @var{info} logging level.
10370 @subsection Examples
10374 Generate a representative palette of a given video using @command{ffmpeg}:
10376 ffmpeg -i input.mkv -vf palettegen palette.png
10380 @section paletteuse
10382 Use a palette to downsample an input video stream.
10384 The filter takes two inputs: one video stream and a palette. The palette must
10385 be a 256 pixels image.
10387 It accepts the following options:
10391 Select dithering mode. Available algorithms are:
10394 Ordered 8x8 bayer dithering (deterministic)
10396 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10397 Note: this dithering is sometimes considered "wrong" and is included as a
10399 @item floyd_steinberg
10400 Floyd and Steingberg dithering (error diffusion)
10402 Frankie Sierra dithering v2 (error diffusion)
10404 Frankie Sierra dithering v2 "Lite" (error diffusion)
10407 Default is @var{sierra2_4a}.
10410 When @var{bayer} dithering is selected, this option defines the scale of the
10411 pattern (how much the crosshatch pattern is visible). A low value means more
10412 visible pattern for less banding, and higher value means less visible pattern
10413 at the cost of more banding.
10415 The option must be an integer value in the range [0,5]. Default is @var{2}.
10418 If set, define the zone to process
10422 Only the changing rectangle will be reprocessed. This is similar to GIF
10423 cropping/offsetting compression mechanism. This option can be useful for speed
10424 if only a part of the image is changing, and has use cases such as limiting the
10425 scope of the error diffusal @option{dither} to the rectangle that bounds the
10426 moving scene (it leads to more deterministic output if the scene doesn't change
10427 much, and as a result less moving noise and better GIF compression).
10430 Default is @var{none}.
10433 @subsection Examples
10437 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10438 using @command{ffmpeg}:
10440 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10444 @section perspective
10446 Correct perspective of video not recorded perpendicular to the screen.
10448 A description of the accepted parameters follows.
10459 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10460 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10461 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10462 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10463 then the corners of the source will be sent to the specified coordinates.
10465 The expressions can use the following variables:
10470 the width and height of video frame.
10474 Output frame count.
10477 @item interpolation
10478 Set interpolation for perspective correction.
10480 It accepts the following values:
10486 Default value is @samp{linear}.
10489 Set interpretation of coordinate options.
10491 It accepts the following values:
10495 Send point in the source specified by the given coordinates to
10496 the corners of the destination.
10498 @item 1, destination
10500 Send the corners of the source to the point in the destination specified
10501 by the given coordinates.
10503 Default value is @samp{source}.
10507 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10509 It accepts the following values:
10512 only evaluate expressions once during the filter initialization or
10513 when a command is processed
10516 evaluate expressions for each incoming frame
10519 Default value is @samp{init}.
10524 Delay interlaced video by one field time so that the field order changes.
10526 The intended use is to fix PAL movies that have been captured with the
10527 opposite field order to the film-to-video transfer.
10529 A description of the accepted parameters follows.
10535 It accepts the following values:
10538 Capture field order top-first, transfer bottom-first.
10539 Filter will delay the bottom field.
10542 Capture field order bottom-first, transfer top-first.
10543 Filter will delay the top field.
10546 Capture and transfer with the same field order. This mode only exists
10547 for the documentation of the other options to refer to, but if you
10548 actually select it, the filter will faithfully do nothing.
10551 Capture field order determined automatically by field flags, transfer
10553 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10554 basis using field flags. If no field information is available,
10555 then this works just like @samp{u}.
10558 Capture unknown or varying, transfer opposite.
10559 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10560 analyzing the images and selecting the alternative that produces best
10561 match between the fields.
10564 Capture top-first, transfer unknown or varying.
10565 Filter selects among @samp{t} and @samp{p} using image analysis.
10568 Capture bottom-first, transfer unknown or varying.
10569 Filter selects among @samp{b} and @samp{p} using image analysis.
10572 Capture determined by field flags, transfer unknown or varying.
10573 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10574 image analysis. If no field information is available, then this works just
10575 like @samp{U}. This is the default mode.
10578 Both capture and transfer unknown or varying.
10579 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10583 @section pixdesctest
10585 Pixel format descriptor test filter, mainly useful for internal
10586 testing. The output video should be equal to the input video.
10590 format=monow, pixdesctest
10593 can be used to test the monowhite pixel format descriptor definition.
10597 Enable the specified chain of postprocessing subfilters using libpostproc. This
10598 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10599 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10600 Each subfilter and some options have a short and a long name that can be used
10601 interchangeably, i.e. dr/dering are the same.
10603 The filters accept the following options:
10607 Set postprocessing subfilters string.
10610 All subfilters share common options to determine their scope:
10614 Honor the quality commands for this subfilter.
10617 Do chrominance filtering, too (default).
10620 Do luminance filtering only (no chrominance).
10623 Do chrominance filtering only (no luminance).
10626 These options can be appended after the subfilter name, separated by a '|'.
10628 Available subfilters are:
10631 @item hb/hdeblock[|difference[|flatness]]
10632 Horizontal deblocking filter
10635 Difference factor where higher values mean more deblocking (default: @code{32}).
10637 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10640 @item vb/vdeblock[|difference[|flatness]]
10641 Vertical deblocking filter
10644 Difference factor where higher values mean more deblocking (default: @code{32}).
10646 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10649 @item ha/hadeblock[|difference[|flatness]]
10650 Accurate horizontal deblocking filter
10653 Difference factor where higher values mean more deblocking (default: @code{32}).
10655 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10658 @item va/vadeblock[|difference[|flatness]]
10659 Accurate vertical deblocking filter
10662 Difference factor where higher values mean more deblocking (default: @code{32}).
10664 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10668 The horizontal and vertical deblocking filters share the difference and
10669 flatness values so you cannot set different horizontal and vertical
10673 @item h1/x1hdeblock
10674 Experimental horizontal deblocking filter
10676 @item v1/x1vdeblock
10677 Experimental vertical deblocking filter
10682 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10685 larger -> stronger filtering
10687 larger -> stronger filtering
10689 larger -> stronger filtering
10692 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10695 Stretch luminance to @code{0-255}.
10698 @item lb/linblenddeint
10699 Linear blend deinterlacing filter that deinterlaces the given block by
10700 filtering all lines with a @code{(1 2 1)} filter.
10702 @item li/linipoldeint
10703 Linear interpolating deinterlacing filter that deinterlaces the given block by
10704 linearly interpolating every second line.
10706 @item ci/cubicipoldeint
10707 Cubic interpolating deinterlacing filter deinterlaces the given block by
10708 cubically interpolating every second line.
10710 @item md/mediandeint
10711 Median deinterlacing filter that deinterlaces the given block by applying a
10712 median filter to every second line.
10714 @item fd/ffmpegdeint
10715 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10716 second line with a @code{(-1 4 2 4 -1)} filter.
10719 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10720 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10722 @item fq/forceQuant[|quantizer]
10723 Overrides the quantizer table from the input with the constant quantizer you
10731 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10734 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10737 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10740 @subsection Examples
10744 Apply horizontal and vertical deblocking, deringing and automatic
10745 brightness/contrast:
10751 Apply default filters without brightness/contrast correction:
10757 Apply default filters and temporal denoiser:
10759 pp=default/tmpnoise|1|2|3
10763 Apply deblocking on luminance only, and switch vertical deblocking on or off
10764 automatically depending on available CPU time:
10771 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10772 similar to spp = 6 with 7 point DCT, where only the center sample is
10775 The filter accepts the following options:
10779 Force a constant quantization parameter. It accepts an integer in range
10780 0 to 63. If not set, the filter will use the QP from the video stream
10784 Set thresholding mode. Available modes are:
10788 Set hard thresholding.
10790 Set soft thresholding (better de-ringing effect, but likely blurrier).
10792 Set medium thresholding (good results, default).
10798 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10799 Ratio) between two input videos.
10801 This filter takes in input two input videos, the first input is
10802 considered the "main" source and is passed unchanged to the
10803 output. The second input is used as a "reference" video for computing
10806 Both video inputs must have the same resolution and pixel format for
10807 this filter to work correctly. Also it assumes that both inputs
10808 have the same number of frames, which are compared one by one.
10810 The obtained average PSNR is printed through the logging system.
10812 The filter stores the accumulated MSE (mean squared error) of each
10813 frame, and at the end of the processing it is averaged across all frames
10814 equally, and the following formula is applied to obtain the PSNR:
10817 PSNR = 10*log10(MAX^2/MSE)
10820 Where MAX is the average of the maximum values of each component of the
10823 The description of the accepted parameters follows.
10826 @item stats_file, f
10827 If specified the filter will use the named file to save the PSNR of
10828 each individual frame. When filename equals "-" the data is sent to
10831 @item stats_version
10832 Specifies which version of the stats file format to use. Details of
10833 each format are written below.
10834 Default value is 1.
10836 @item stats_add_max
10837 Determines whether the max value is output to the stats log.
10838 Default value is 0.
10839 Requires stats_version >= 2. If this is set and stats_version < 2,
10840 the filter will return an error.
10843 The file printed if @var{stats_file} is selected, contains a sequence of
10844 key/value pairs of the form @var{key}:@var{value} for each compared
10847 If a @var{stats_version} greater than 1 is specified, a header line precedes
10848 the list of per-frame-pair stats, with key value pairs following the frame
10849 format with the following parameters:
10852 @item psnr_log_version
10853 The version of the log file format. Will match @var{stats_version}.
10856 A comma separated list of the per-frame-pair parameters included in
10860 A description of each shown per-frame-pair parameter follows:
10864 sequential number of the input frame, starting from 1
10867 Mean Square Error pixel-by-pixel average difference of the compared
10868 frames, averaged over all the image components.
10870 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
10871 Mean Square Error pixel-by-pixel average difference of the compared
10872 frames for the component specified by the suffix.
10874 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
10875 Peak Signal to Noise ratio of the compared frames for the component
10876 specified by the suffix.
10878 @item max_avg, max_y, max_u, max_v
10879 Maximum allowed value for each channel, and average over all
10885 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
10886 [main][ref] psnr="stats_file=stats.log" [out]
10889 On this example the input file being processed is compared with the
10890 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
10891 is stored in @file{stats.log}.
10896 Pulldown reversal (inverse telecine) filter, capable of handling mixed
10897 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
10900 The pullup filter is designed to take advantage of future context in making
10901 its decisions. This filter is stateless in the sense that it does not lock
10902 onto a pattern to follow, but it instead looks forward to the following
10903 fields in order to identify matches and rebuild progressive frames.
10905 To produce content with an even framerate, insert the fps filter after
10906 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
10907 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
10909 The filter accepts the following options:
10916 These options set the amount of "junk" to ignore at the left, right, top, and
10917 bottom of the image, respectively. Left and right are in units of 8 pixels,
10918 while top and bottom are in units of 2 lines.
10919 The default is 8 pixels on each side.
10922 Set the strict breaks. Setting this option to 1 will reduce the chances of
10923 filter generating an occasional mismatched frame, but it may also cause an
10924 excessive number of frames to be dropped during high motion sequences.
10925 Conversely, setting it to -1 will make filter match fields more easily.
10926 This may help processing of video where there is slight blurring between
10927 the fields, but may also cause there to be interlaced frames in the output.
10928 Default value is @code{0}.
10931 Set the metric plane to use. It accepts the following values:
10937 Use chroma blue plane.
10940 Use chroma red plane.
10943 This option may be set to use chroma plane instead of the default luma plane
10944 for doing filter's computations. This may improve accuracy on very clean
10945 source material, but more likely will decrease accuracy, especially if there
10946 is chroma noise (rainbow effect) or any grayscale video.
10947 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
10948 load and make pullup usable in realtime on slow machines.
10951 For best results (without duplicated frames in the output file) it is
10952 necessary to change the output frame rate. For example, to inverse
10953 telecine NTSC input:
10955 ffmpeg -i input -vf pullup -r 24000/1001 ...
10960 Change video quantization parameters (QP).
10962 The filter accepts the following option:
10966 Set expression for quantization parameter.
10969 The expression is evaluated through the eval API and can contain, among others,
10970 the following constants:
10974 1 if index is not 129, 0 otherwise.
10977 Sequentional index starting from -129 to 128.
10980 @subsection Examples
10984 Some equation like:
10992 Flush video frames from internal cache of frames into a random order.
10993 No frame is discarded.
10994 Inspired by @ref{frei0r} nervous filter.
10998 Set size in number of frames of internal cache, in range from @code{2} to
10999 @code{512}. Default is @code{30}.
11002 Set seed for random number generator, must be an integer included between
11003 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
11004 less than @code{0}, the filter will try to use a good random seed on a
11010 Read vertical interval timecode (VITC) information from the top lines of a
11013 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
11014 timecode value, if a valid timecode has been detected. Further metadata key
11015 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
11016 timecode data has been found or not.
11018 This filter accepts the following options:
11022 Set the maximum number of lines to scan for VITC data. If the value is set to
11023 @code{-1} the full video frame is scanned. Default is @code{45}.
11026 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
11027 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
11030 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
11031 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
11034 @subsection Examples
11038 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
11039 draw @code{--:--:--:--} as a placeholder:
11041 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
11047 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
11049 Destination pixel at position (X, Y) will be picked from source (x, y) position
11050 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
11051 value for pixel will be used for destination pixel.
11053 Xmap and Ymap input video streams must be of same dimensions. Output video stream
11054 will have Xmap/Ymap video stream dimensions.
11055 Xmap and Ymap input video streams are 16bit depth, single channel.
11057 @section removegrain
11059 The removegrain filter is a spatial denoiser for progressive video.
11063 Set mode for the first plane.
11066 Set mode for the second plane.
11069 Set mode for the third plane.
11072 Set mode for the fourth plane.
11075 Range of mode is from 0 to 24. Description of each mode follows:
11079 Leave input plane unchanged. Default.
11082 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
11085 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
11088 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
11091 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
11092 This is equivalent to a median filter.
11095 Line-sensitive clipping giving the minimal change.
11098 Line-sensitive clipping, intermediate.
11101 Line-sensitive clipping, intermediate.
11104 Line-sensitive clipping, intermediate.
11107 Line-sensitive clipping on a line where the neighbours pixels are the closest.
11110 Replaces the target pixel with the closest neighbour.
11113 [1 2 1] horizontal and vertical kernel blur.
11119 Bob mode, interpolates top field from the line where the neighbours
11120 pixels are the closest.
11123 Bob mode, interpolates bottom field from the line where the neighbours
11124 pixels are the closest.
11127 Bob mode, interpolates top field. Same as 13 but with a more complicated
11128 interpolation formula.
11131 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
11132 interpolation formula.
11135 Clips the pixel with the minimum and maximum of respectively the maximum and
11136 minimum of each pair of opposite neighbour pixels.
11139 Line-sensitive clipping using opposite neighbours whose greatest distance from
11140 the current pixel is minimal.
11143 Replaces the pixel with the average of its 8 neighbours.
11146 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
11149 Clips pixels using the averages of opposite neighbour.
11152 Same as mode 21 but simpler and faster.
11155 Small edge and halo removal, but reputed useless.
11161 @section removelogo
11163 Suppress a TV station logo, using an image file to determine which
11164 pixels comprise the logo. It works by filling in the pixels that
11165 comprise the logo with neighboring pixels.
11167 The filter accepts the following options:
11171 Set the filter bitmap file, which can be any image format supported by
11172 libavformat. The width and height of the image file must match those of the
11173 video stream being processed.
11176 Pixels in the provided bitmap image with a value of zero are not
11177 considered part of the logo, non-zero pixels are considered part of
11178 the logo. If you use white (255) for the logo and black (0) for the
11179 rest, you will be safe. For making the filter bitmap, it is
11180 recommended to take a screen capture of a black frame with the logo
11181 visible, and then using a threshold filter followed by the erode
11182 filter once or twice.
11184 If needed, little splotches can be fixed manually. Remember that if
11185 logo pixels are not covered, the filter quality will be much
11186 reduced. Marking too many pixels as part of the logo does not hurt as
11187 much, but it will increase the amount of blurring needed to cover over
11188 the image and will destroy more information than necessary, and extra
11189 pixels will slow things down on a large logo.
11191 @section repeatfields
11193 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11194 fields based on its value.
11198 Reverse a video clip.
11200 Warning: This filter requires memory to buffer the entire clip, so trimming
11203 @subsection Examples
11207 Take the first 5 seconds of a clip, and reverse it.
11215 Rotate video by an arbitrary angle expressed in radians.
11217 The filter accepts the following options:
11219 A description of the optional parameters follows.
11222 Set an expression for the angle by which to rotate the input video
11223 clockwise, expressed as a number of radians. A negative value will
11224 result in a counter-clockwise rotation. By default it is set to "0".
11226 This expression is evaluated for each frame.
11229 Set the output width expression, default value is "iw".
11230 This expression is evaluated just once during configuration.
11233 Set the output height expression, default value is "ih".
11234 This expression is evaluated just once during configuration.
11237 Enable bilinear interpolation if set to 1, a value of 0 disables
11238 it. Default value is 1.
11241 Set the color used to fill the output area not covered by the rotated
11242 image. For the general syntax of this option, check the "Color" section in the
11243 ffmpeg-utils manual. If the special value "none" is selected then no
11244 background is printed (useful for example if the background is never shown).
11246 Default value is "black".
11249 The expressions for the angle and the output size can contain the
11250 following constants and functions:
11254 sequential number of the input frame, starting from 0. It is always NAN
11255 before the first frame is filtered.
11258 time in seconds of the input frame, it is set to 0 when the filter is
11259 configured. It is always NAN before the first frame is filtered.
11263 horizontal and vertical chroma subsample values. For example for the
11264 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11268 the input video width and height
11272 the output width and height, that is the size of the padded area as
11273 specified by the @var{width} and @var{height} expressions
11277 the minimal width/height required for completely containing the input
11278 video rotated by @var{a} radians.
11280 These are only available when computing the @option{out_w} and
11281 @option{out_h} expressions.
11284 @subsection Examples
11288 Rotate the input by PI/6 radians clockwise:
11294 Rotate the input by PI/6 radians counter-clockwise:
11300 Rotate the input by 45 degrees clockwise:
11306 Apply a constant rotation with period T, starting from an angle of PI/3:
11308 rotate=PI/3+2*PI*t/T
11312 Make the input video rotation oscillating with a period of T
11313 seconds and an amplitude of A radians:
11315 rotate=A*sin(2*PI/T*t)
11319 Rotate the video, output size is chosen so that the whole rotating
11320 input video is always completely contained in the output:
11322 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11326 Rotate the video, reduce the output size so that no background is ever
11329 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11333 @subsection Commands
11335 The filter supports the following commands:
11339 Set the angle expression.
11340 The command accepts the same syntax of the corresponding option.
11342 If the specified expression is not valid, it is kept at its current
11348 Apply Shape Adaptive Blur.
11350 The filter accepts the following options:
11353 @item luma_radius, lr
11354 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11355 value is 1.0. A greater value will result in a more blurred image, and
11356 in slower processing.
11358 @item luma_pre_filter_radius, lpfr
11359 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11362 @item luma_strength, ls
11363 Set luma maximum difference between pixels to still be considered, must
11364 be a value in the 0.1-100.0 range, default value is 1.0.
11366 @item chroma_radius, cr
11367 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11368 greater value will result in a more blurred image, and in slower
11371 @item chroma_pre_filter_radius, cpfr
11372 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11374 @item chroma_strength, cs
11375 Set chroma maximum difference between pixels to still be considered,
11376 must be a value in the -0.9-100.0 range.
11379 Each chroma option value, if not explicitly specified, is set to the
11380 corresponding luma option value.
11385 Scale (resize) the input video, using the libswscale library.
11387 The scale filter forces the output display aspect ratio to be the same
11388 of the input, by changing the output sample aspect ratio.
11390 If the input image format is different from the format requested by
11391 the next filter, the scale filter will convert the input to the
11394 @subsection Options
11395 The filter accepts the following options, or any of the options
11396 supported by the libswscale scaler.
11398 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11399 the complete list of scaler options.
11404 Set the output video dimension expression. Default value is the input
11407 If the value is 0, the input width is used for the output.
11409 If one of the values is -1, the scale filter will use a value that
11410 maintains the aspect ratio of the input image, calculated from the
11411 other specified dimension. If both of them are -1, the input size is
11414 If one of the values is -n with n > 1, the scale filter will also use a value
11415 that maintains the aspect ratio of the input image, calculated from the other
11416 specified dimension. After that it will, however, make sure that the calculated
11417 dimension is divisible by n and adjust the value if necessary.
11419 See below for the list of accepted constants for use in the dimension
11423 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11427 Only evaluate expressions once during the filter initialization or when a command is processed.
11430 Evaluate expressions for each incoming frame.
11434 Default value is @samp{init}.
11438 Set the interlacing mode. It accepts the following values:
11442 Force interlaced aware scaling.
11445 Do not apply interlaced scaling.
11448 Select interlaced aware scaling depending on whether the source frames
11449 are flagged as interlaced or not.
11452 Default value is @samp{0}.
11455 Set libswscale scaling flags. See
11456 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11457 complete list of values. If not explicitly specified the filter applies
11461 @item param0, param1
11462 Set libswscale input parameters for scaling algorithms that need them. See
11463 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11464 complete documentation. If not explicitly specified the filter applies
11470 Set the video size. For the syntax of this option, check the
11471 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11473 @item in_color_matrix
11474 @item out_color_matrix
11475 Set in/output YCbCr color space type.
11477 This allows the autodetected value to be overridden as well as allows forcing
11478 a specific value used for the output and encoder.
11480 If not specified, the color space type depends on the pixel format.
11486 Choose automatically.
11489 Format conforming to International Telecommunication Union (ITU)
11490 Recommendation BT.709.
11493 Set color space conforming to the United States Federal Communications
11494 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11497 Set color space conforming to:
11501 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11504 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11507 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11512 Set color space conforming to SMPTE ST 240:1999.
11517 Set in/output YCbCr sample range.
11519 This allows the autodetected value to be overridden as well as allows forcing
11520 a specific value used for the output and encoder. If not specified, the
11521 range depends on the pixel format. Possible values:
11525 Choose automatically.
11528 Set full range (0-255 in case of 8-bit luma).
11531 Set "MPEG" range (16-235 in case of 8-bit luma).
11534 @item force_original_aspect_ratio
11535 Enable decreasing or increasing output video width or height if necessary to
11536 keep the original aspect ratio. Possible values:
11540 Scale the video as specified and disable this feature.
11543 The output video dimensions will automatically be decreased if needed.
11546 The output video dimensions will automatically be increased if needed.
11550 One useful instance of this option is that when you know a specific device's
11551 maximum allowed resolution, you can use this to limit the output video to
11552 that, while retaining the aspect ratio. For example, device A allows
11553 1280x720 playback, and your video is 1920x800. Using this option (set it to
11554 decrease) and specifying 1280x720 to the command line makes the output
11557 Please note that this is a different thing than specifying -1 for @option{w}
11558 or @option{h}, you still need to specify the output resolution for this option
11563 The values of the @option{w} and @option{h} options are expressions
11564 containing the following constants:
11569 The input width and height
11573 These are the same as @var{in_w} and @var{in_h}.
11577 The output (scaled) width and height
11581 These are the same as @var{out_w} and @var{out_h}
11584 The same as @var{iw} / @var{ih}
11587 input sample aspect ratio
11590 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11594 horizontal and vertical input chroma subsample values. For example for the
11595 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11599 horizontal and vertical output chroma subsample values. For example for the
11600 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11603 @subsection Examples
11607 Scale the input video to a size of 200x100
11612 This is equivalent to:
11623 Specify a size abbreviation for the output size:
11628 which can also be written as:
11634 Scale the input to 2x:
11636 scale=w=2*iw:h=2*ih
11640 The above is the same as:
11642 scale=2*in_w:2*in_h
11646 Scale the input to 2x with forced interlaced scaling:
11648 scale=2*iw:2*ih:interl=1
11652 Scale the input to half size:
11654 scale=w=iw/2:h=ih/2
11658 Increase the width, and set the height to the same size:
11664 Seek Greek harmony:
11671 Increase the height, and set the width to 3/2 of the height:
11673 scale=w=3/2*oh:h=3/5*ih
11677 Increase the size, making the size a multiple of the chroma
11680 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11684 Increase the width to a maximum of 500 pixels,
11685 keeping the same aspect ratio as the input:
11687 scale=w='min(500\, iw*3/2):h=-1'
11691 @subsection Commands
11693 This filter supports the following commands:
11697 Set the output video dimension expression.
11698 The command accepts the same syntax of the corresponding option.
11700 If the specified expression is not valid, it is kept at its current
11706 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11707 format conversion on CUDA video frames. Setting the output width and height
11708 works in the same way as for the @var{scale} filter.
11710 The following additional options are accepted:
11713 The pixel format of the output CUDA frames. If set to the string "same" (the
11714 default), the input format will be kept. Note that automatic format negotiation
11715 and conversion is not yet supported for hardware frames
11718 The interpolation algorithm used for resizing. One of the following:
11725 @item cubic2p_bspline
11726 2-parameter cubic (B=1, C=0)
11728 @item cubic2p_catmullrom
11729 2-parameter cubic (B=0, C=1/2)
11731 @item cubic2p_b05c03
11732 2-parameter cubic (B=1/2, C=3/10)
11744 Scale (resize) the input video, based on a reference video.
11746 See the scale filter for available options, scale2ref supports the same but
11747 uses the reference video instead of the main input as basis.
11749 @subsection Examples
11753 Scale a subtitle stream to match the main video in size before overlaying
11755 'scale2ref[b][a];[a][b]overlay'
11759 @anchor{selectivecolor}
11760 @section selectivecolor
11762 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11763 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11764 by the "purity" of the color (that is, how saturated it already is).
11766 This filter is similar to the Adobe Photoshop Selective Color tool.
11768 The filter accepts the following options:
11771 @item correction_method
11772 Select color correction method.
11774 Available values are:
11777 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11780 Specified adjustments are relative to the original component value.
11782 Default is @code{absolute}.
11784 Adjustments for red pixels (pixels where the red component is the maximum)
11786 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11788 Adjustments for green pixels (pixels where the green component is the maximum)
11790 Adjustments for cyan pixels (pixels where the red component is the minimum)
11792 Adjustments for blue pixels (pixels where the blue component is the maximum)
11794 Adjustments for magenta pixels (pixels where the green component is the minimum)
11796 Adjustments for white pixels (pixels where all components are greater than 128)
11798 Adjustments for all pixels except pure black and pure white
11800 Adjustments for black pixels (pixels where all components are lesser than 128)
11802 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11805 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11806 4 space separated floating point adjustment values in the [-1,1] range,
11807 respectively to adjust the amount of cyan, magenta, yellow and black for the
11808 pixels of its range.
11810 @subsection Examples
11814 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11815 increase magenta by 27% in blue areas:
11817 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11821 Use a Photoshop selective color preset:
11823 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11827 @section separatefields
11829 The @code{separatefields} takes a frame-based video input and splits
11830 each frame into its components fields, producing a new half height clip
11831 with twice the frame rate and twice the frame count.
11833 This filter use field-dominance information in frame to decide which
11834 of each pair of fields to place first in the output.
11835 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
11837 @section setdar, setsar
11839 The @code{setdar} filter sets the Display Aspect Ratio for the filter
11842 This is done by changing the specified Sample (aka Pixel) Aspect
11843 Ratio, according to the following equation:
11845 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
11848 Keep in mind that the @code{setdar} filter does not modify the pixel
11849 dimensions of the video frame. Also, the display aspect ratio set by
11850 this filter may be changed by later filters in the filterchain,
11851 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
11854 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
11855 the filter output video.
11857 Note that as a consequence of the application of this filter, the
11858 output display aspect ratio will change according to the equation
11861 Keep in mind that the sample aspect ratio set by the @code{setsar}
11862 filter may be changed by later filters in the filterchain, e.g. if
11863 another "setsar" or a "setdar" filter is applied.
11865 It accepts the following parameters:
11868 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
11869 Set the aspect ratio used by the filter.
11871 The parameter can be a floating point number string, an expression, or
11872 a string of the form @var{num}:@var{den}, where @var{num} and
11873 @var{den} are the numerator and denominator of the aspect ratio. If
11874 the parameter is not specified, it is assumed the value "0".
11875 In case the form "@var{num}:@var{den}" is used, the @code{:} character
11879 Set the maximum integer value to use for expressing numerator and
11880 denominator when reducing the expressed aspect ratio to a rational.
11881 Default value is @code{100}.
11885 The parameter @var{sar} is an expression containing
11886 the following constants:
11890 These are approximated values for the mathematical constants e
11891 (Euler's number), pi (Greek pi), and phi (the golden ratio).
11894 The input width and height.
11897 These are the same as @var{w} / @var{h}.
11900 The input sample aspect ratio.
11903 The input display aspect ratio. It is the same as
11904 (@var{w} / @var{h}) * @var{sar}.
11907 Horizontal and vertical chroma subsample values. For example, for the
11908 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11911 @subsection Examples
11916 To change the display aspect ratio to 16:9, specify one of the following:
11923 To change the sample aspect ratio to 10:11, specify:
11929 To set a display aspect ratio of 16:9, and specify a maximum integer value of
11930 1000 in the aspect ratio reduction, use the command:
11932 setdar=ratio=16/9:max=1000
11940 Force field for the output video frame.
11942 The @code{setfield} filter marks the interlace type field for the
11943 output frames. It does not change the input frame, but only sets the
11944 corresponding property, which affects how the frame is treated by
11945 following filters (e.g. @code{fieldorder} or @code{yadif}).
11947 The filter accepts the following options:
11952 Available values are:
11956 Keep the same field property.
11959 Mark the frame as bottom-field-first.
11962 Mark the frame as top-field-first.
11965 Mark the frame as progressive.
11971 Show a line containing various information for each input video frame.
11972 The input video is not modified.
11974 The shown line contains a sequence of key/value pairs of the form
11975 @var{key}:@var{value}.
11977 The following values are shown in the output:
11981 The (sequential) number of the input frame, starting from 0.
11984 The Presentation TimeStamp of the input frame, expressed as a number of
11985 time base units. The time base unit depends on the filter input pad.
11988 The Presentation TimeStamp of the input frame, expressed as a number of
11992 The position of the frame in the input stream, or -1 if this information is
11993 unavailable and/or meaningless (for example in case of synthetic video).
11996 The pixel format name.
11999 The sample aspect ratio of the input frame, expressed in the form
12000 @var{num}/@var{den}.
12003 The size of the input frame. For the syntax of this option, check the
12004 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12007 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
12008 for bottom field first).
12011 This is 1 if the frame is a key frame, 0 otherwise.
12014 The picture type of the input frame ("I" for an I-frame, "P" for a
12015 P-frame, "B" for a B-frame, or "?" for an unknown type).
12016 Also refer to the documentation of the @code{AVPictureType} enum and of
12017 the @code{av_get_picture_type_char} function defined in
12018 @file{libavutil/avutil.h}.
12021 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
12023 @item plane_checksum
12024 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
12025 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
12028 @section showpalette
12030 Displays the 256 colors palette of each frame. This filter is only relevant for
12031 @var{pal8} pixel format frames.
12033 It accepts the following option:
12037 Set the size of the box used to represent one palette color entry. Default is
12038 @code{30} (for a @code{30x30} pixel box).
12041 @section shuffleframes
12043 Reorder and/or duplicate video frames.
12045 It accepts the following parameters:
12049 Set the destination indexes of input frames.
12050 This is space or '|' separated list of indexes that maps input frames to output
12051 frames. Number of indexes also sets maximal value that each index may have.
12054 The first frame has the index 0. The default is to keep the input unchanged.
12056 Swap second and third frame of every three frames of the input:
12058 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
12061 @section shuffleplanes
12063 Reorder and/or duplicate video planes.
12065 It accepts the following parameters:
12070 The index of the input plane to be used as the first output plane.
12073 The index of the input plane to be used as the second output plane.
12076 The index of the input plane to be used as the third output plane.
12079 The index of the input plane to be used as the fourth output plane.
12083 The first plane has the index 0. The default is to keep the input unchanged.
12085 Swap the second and third planes of the input:
12087 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
12090 @anchor{signalstats}
12091 @section signalstats
12092 Evaluate various visual metrics that assist in determining issues associated
12093 with the digitization of analog video media.
12095 By default the filter will log these metadata values:
12099 Display the minimal Y value contained within the input frame. Expressed in
12103 Display the Y value at the 10% percentile within the input frame. Expressed in
12107 Display the average Y value within the input frame. Expressed in range of
12111 Display the Y value at the 90% percentile within the input frame. Expressed in
12115 Display the maximum Y value contained within the input frame. Expressed in
12119 Display the minimal U value contained within the input frame. Expressed in
12123 Display the U value at the 10% percentile within the input frame. Expressed in
12127 Display the average U value within the input frame. Expressed in range of
12131 Display the U value at the 90% percentile within the input frame. Expressed in
12135 Display the maximum U value contained within the input frame. Expressed in
12139 Display the minimal V value contained within the input frame. Expressed in
12143 Display the V value at the 10% percentile within the input frame. Expressed in
12147 Display the average V value within the input frame. Expressed in range of
12151 Display the V value at the 90% percentile within the input frame. Expressed in
12155 Display the maximum V value contained within the input frame. Expressed in
12159 Display the minimal saturation value contained within the input frame.
12160 Expressed in range of [0-~181.02].
12163 Display the saturation value at the 10% percentile within the input frame.
12164 Expressed in range of [0-~181.02].
12167 Display the average saturation value within the input frame. Expressed in range
12171 Display the saturation value at the 90% percentile within the input frame.
12172 Expressed in range of [0-~181.02].
12175 Display the maximum saturation value contained within the input frame.
12176 Expressed in range of [0-~181.02].
12179 Display the median value for hue within the input frame. Expressed in range of
12183 Display the average value for hue within the input frame. Expressed in range of
12187 Display the average of sample value difference between all values of the Y
12188 plane in the current frame and corresponding values of the previous input frame.
12189 Expressed in range of [0-255].
12192 Display the average of sample value difference between all values of the U
12193 plane in the current frame and corresponding values of the previous input frame.
12194 Expressed in range of [0-255].
12197 Display the average of sample value difference between all values of the V
12198 plane in the current frame and corresponding values of the previous input frame.
12199 Expressed in range of [0-255].
12202 Display bit depth of Y plane in current frame.
12203 Expressed in range of [0-16].
12206 Display bit depth of U plane in current frame.
12207 Expressed in range of [0-16].
12210 Display bit depth of V plane in current frame.
12211 Expressed in range of [0-16].
12214 The filter accepts the following options:
12220 @option{stat} specify an additional form of image analysis.
12221 @option{out} output video with the specified type of pixel highlighted.
12223 Both options accept the following values:
12227 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12228 unlike the neighboring pixels of the same field. Examples of temporal outliers
12229 include the results of video dropouts, head clogs, or tape tracking issues.
12232 Identify @var{vertical line repetition}. Vertical line repetition includes
12233 similar rows of pixels within a frame. In born-digital video vertical line
12234 repetition is common, but this pattern is uncommon in video digitized from an
12235 analog source. When it occurs in video that results from the digitization of an
12236 analog source it can indicate concealment from a dropout compensator.
12239 Identify pixels that fall outside of legal broadcast range.
12243 Set the highlight color for the @option{out} option. The default color is
12247 @subsection Examples
12251 Output data of various video metrics:
12253 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12257 Output specific data about the minimum and maximum values of the Y plane per frame:
12259 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12263 Playback video while highlighting pixels that are outside of broadcast range in red.
12265 ffplay example.mov -vf signalstats="out=brng:color=red"
12269 Playback video with signalstats metadata drawn over the frame.
12271 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12274 The contents of signalstat_drawtext.txt used in the command are:
12277 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12278 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12279 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12280 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12288 Blur the input video without impacting the outlines.
12290 It accepts the following options:
12293 @item luma_radius, lr
12294 Set the luma radius. The option value must be a float number in
12295 the range [0.1,5.0] that specifies the variance of the gaussian filter
12296 used to blur the image (slower if larger). Default value is 1.0.
12298 @item luma_strength, ls
12299 Set the luma strength. The option value must be a float number
12300 in the range [-1.0,1.0] that configures the blurring. A value included
12301 in [0.0,1.0] will blur the image whereas a value included in
12302 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12304 @item luma_threshold, lt
12305 Set the luma threshold used as a coefficient to determine
12306 whether a pixel should be blurred or not. The option value must be an
12307 integer in the range [-30,30]. A value of 0 will filter all the image,
12308 a value included in [0,30] will filter flat areas and a value included
12309 in [-30,0] will filter edges. Default value is 0.
12311 @item chroma_radius, cr
12312 Set the chroma radius. The option value must be a float number in
12313 the range [0.1,5.0] that specifies the variance of the gaussian filter
12314 used to blur the image (slower if larger). Default value is 1.0.
12316 @item chroma_strength, cs
12317 Set the chroma strength. The option value must be a float number
12318 in the range [-1.0,1.0] that configures the blurring. A value included
12319 in [0.0,1.0] will blur the image whereas a value included in
12320 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12322 @item chroma_threshold, ct
12323 Set the chroma threshold used as a coefficient to determine
12324 whether a pixel should be blurred or not. The option value must be an
12325 integer in the range [-30,30]. A value of 0 will filter all the image,
12326 a value included in [0,30] will filter flat areas and a value included
12327 in [-30,0] will filter edges. Default value is 0.
12330 If a chroma option is not explicitly set, the corresponding luma value
12335 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12337 This filter takes in input two input videos, the first input is
12338 considered the "main" source and is passed unchanged to the
12339 output. The second input is used as a "reference" video for computing
12342 Both video inputs must have the same resolution and pixel format for
12343 this filter to work correctly. Also it assumes that both inputs
12344 have the same number of frames, which are compared one by one.
12346 The filter stores the calculated SSIM of each frame.
12348 The description of the accepted parameters follows.
12351 @item stats_file, f
12352 If specified the filter will use the named file to save the SSIM of
12353 each individual frame. When filename equals "-" the data is sent to
12357 The file printed if @var{stats_file} is selected, contains a sequence of
12358 key/value pairs of the form @var{key}:@var{value} for each compared
12361 A description of each shown parameter follows:
12365 sequential number of the input frame, starting from 1
12367 @item Y, U, V, R, G, B
12368 SSIM of the compared frames for the component specified by the suffix.
12371 SSIM of the compared frames for the whole frame.
12374 Same as above but in dB representation.
12379 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12380 [main][ref] ssim="stats_file=stats.log" [out]
12383 On this example the input file being processed is compared with the
12384 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12385 is stored in @file{stats.log}.
12387 Another example with both psnr and ssim at same time:
12389 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12394 Convert between different stereoscopic image formats.
12396 The filters accept the following options:
12400 Set stereoscopic image format of input.
12402 Available values for input image formats are:
12405 side by side parallel (left eye left, right eye right)
12408 side by side crosseye (right eye left, left eye right)
12411 side by side parallel with half width resolution
12412 (left eye left, right eye right)
12415 side by side crosseye with half width resolution
12416 (right eye left, left eye right)
12419 above-below (left eye above, right eye below)
12422 above-below (right eye above, left eye below)
12425 above-below with half height resolution
12426 (left eye above, right eye below)
12429 above-below with half height resolution
12430 (right eye above, left eye below)
12433 alternating frames (left eye first, right eye second)
12436 alternating frames (right eye first, left eye second)
12439 interleaved rows (left eye has top row, right eye starts on next row)
12442 interleaved rows (right eye has top row, left eye starts on next row)
12445 interleaved columns, left eye first
12448 interleaved columns, right eye first
12450 Default value is @samp{sbsl}.
12454 Set stereoscopic image format of output.
12458 side by side parallel (left eye left, right eye right)
12461 side by side crosseye (right eye left, left eye right)
12464 side by side parallel with half width resolution
12465 (left eye left, right eye right)
12468 side by side crosseye with half width resolution
12469 (right eye left, left eye right)
12472 above-below (left eye above, right eye below)
12475 above-below (right eye above, left eye below)
12478 above-below with half height resolution
12479 (left eye above, right eye below)
12482 above-below with half height resolution
12483 (right eye above, left eye below)
12486 alternating frames (left eye first, right eye second)
12489 alternating frames (right eye first, left eye second)
12492 interleaved rows (left eye has top row, right eye starts on next row)
12495 interleaved rows (right eye has top row, left eye starts on next row)
12498 anaglyph red/blue gray
12499 (red filter on left eye, blue filter on right eye)
12502 anaglyph red/green gray
12503 (red filter on left eye, green filter on right eye)
12506 anaglyph red/cyan gray
12507 (red filter on left eye, cyan filter on right eye)
12510 anaglyph red/cyan half colored
12511 (red filter on left eye, cyan filter on right eye)
12514 anaglyph red/cyan color
12515 (red filter on left eye, cyan filter on right eye)
12518 anaglyph red/cyan color optimized with the least squares projection of dubois
12519 (red filter on left eye, cyan filter on right eye)
12522 anaglyph green/magenta gray
12523 (green filter on left eye, magenta filter on right eye)
12526 anaglyph green/magenta half colored
12527 (green filter on left eye, magenta filter on right eye)
12530 anaglyph green/magenta colored
12531 (green filter on left eye, magenta filter on right eye)
12534 anaglyph green/magenta color optimized with the least squares projection of dubois
12535 (green filter on left eye, magenta filter on right eye)
12538 anaglyph yellow/blue gray
12539 (yellow filter on left eye, blue filter on right eye)
12542 anaglyph yellow/blue half colored
12543 (yellow filter on left eye, blue filter on right eye)
12546 anaglyph yellow/blue colored
12547 (yellow filter on left eye, blue filter on right eye)
12550 anaglyph yellow/blue color optimized with the least squares projection of dubois
12551 (yellow filter on left eye, blue filter on right eye)
12554 mono output (left eye only)
12557 mono output (right eye only)
12560 checkerboard, left eye first
12563 checkerboard, right eye first
12566 interleaved columns, left eye first
12569 interleaved columns, right eye first
12575 Default value is @samp{arcd}.
12578 @subsection Examples
12582 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12588 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12594 @section streamselect, astreamselect
12595 Select video or audio streams.
12597 The filter accepts the following options:
12601 Set number of inputs. Default is 2.
12604 Set input indexes to remap to outputs.
12607 @subsection Commands
12609 The @code{streamselect} and @code{astreamselect} filter supports the following
12614 Set input indexes to remap to outputs.
12617 @subsection Examples
12621 Select first 5 seconds 1st stream and rest of time 2nd stream:
12623 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12627 Same as above, but for audio:
12629 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12636 Apply a simple postprocessing filter that compresses and decompresses the image
12637 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12638 and average the results.
12640 The filter accepts the following options:
12644 Set quality. This option defines the number of levels for averaging. It accepts
12645 an integer in the range 0-6. If set to @code{0}, the filter will have no
12646 effect. A value of @code{6} means the higher quality. For each increment of
12647 that value the speed drops by a factor of approximately 2. Default value is
12651 Force a constant quantization parameter. If not set, the filter will use the QP
12652 from the video stream (if available).
12655 Set thresholding mode. Available modes are:
12659 Set hard thresholding (default).
12661 Set soft thresholding (better de-ringing effect, but likely blurrier).
12664 @item use_bframe_qp
12665 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12666 option may cause flicker since the B-Frames have often larger QP. Default is
12667 @code{0} (not enabled).
12673 Draw subtitles on top of input video using the libass library.
12675 To enable compilation of this filter you need to configure FFmpeg with
12676 @code{--enable-libass}. This filter also requires a build with libavcodec and
12677 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12678 Alpha) subtitles format.
12680 The filter accepts the following options:
12684 Set the filename of the subtitle file to read. It must be specified.
12686 @item original_size
12687 Specify the size of the original video, the video for which the ASS file
12688 was composed. For the syntax of this option, check the
12689 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12690 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12691 correctly scale the fonts if the aspect ratio has been changed.
12694 Set a directory path containing fonts that can be used by the filter.
12695 These fonts will be used in addition to whatever the font provider uses.
12698 Set subtitles input character encoding. @code{subtitles} filter only. Only
12699 useful if not UTF-8.
12701 @item stream_index, si
12702 Set subtitles stream index. @code{subtitles} filter only.
12705 Override default style or script info parameters of the subtitles. It accepts a
12706 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12709 If the first key is not specified, it is assumed that the first value
12710 specifies the @option{filename}.
12712 For example, to render the file @file{sub.srt} on top of the input
12713 video, use the command:
12718 which is equivalent to:
12720 subtitles=filename=sub.srt
12723 To render the default subtitles stream from file @file{video.mkv}, use:
12725 subtitles=video.mkv
12728 To render the second subtitles stream from that file, use:
12730 subtitles=video.mkv:si=1
12733 To make the subtitles stream from @file{sub.srt} appear in transparent green
12734 @code{DejaVu Serif}, use:
12736 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12739 @section super2xsai
12741 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12742 Interpolate) pixel art scaling algorithm.
12744 Useful for enlarging pixel art images without reducing sharpness.
12748 Swap two rectangular objects in video.
12750 This filter accepts the following options:
12760 Set 1st rect x coordinate.
12763 Set 1st rect y coordinate.
12766 Set 2nd rect x coordinate.
12769 Set 2nd rect y coordinate.
12771 All expressions are evaluated once for each frame.
12774 The all options are expressions containing the following constants:
12779 The input width and height.
12782 same as @var{w} / @var{h}
12785 input sample aspect ratio
12788 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12791 The number of the input frame, starting from 0.
12794 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12797 the position in the file of the input frame, NAN if unknown
12805 Apply telecine process to the video.
12807 This filter accepts the following options:
12816 The default value is @code{top}.
12820 A string of numbers representing the pulldown pattern you wish to apply.
12821 The default value is @code{23}.
12825 Some typical patterns:
12830 24p: 2332 (preferred)
12837 24p: 222222222223 ("Euro pulldown")
12843 Select the most representative frame in a given sequence of consecutive frames.
12845 The filter accepts the following options:
12849 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
12850 will pick one of them, and then handle the next batch of @var{n} frames until
12851 the end. Default is @code{100}.
12854 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
12855 value will result in a higher memory usage, so a high value is not recommended.
12857 @subsection Examples
12861 Extract one picture each 50 frames:
12867 Complete example of a thumbnail creation with @command{ffmpeg}:
12869 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
12875 Tile several successive frames together.
12877 The filter accepts the following options:
12882 Set the grid size (i.e. the number of lines and columns). For the syntax of
12883 this option, check the
12884 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12887 Set the maximum number of frames to render in the given area. It must be less
12888 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
12889 the area will be used.
12892 Set the outer border margin in pixels.
12895 Set the inner border thickness (i.e. the number of pixels between frames). For
12896 more advanced padding options (such as having different values for the edges),
12897 refer to the pad video filter.
12900 Specify the color of the unused area. For the syntax of this option, check the
12901 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
12905 @subsection Examples
12909 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
12911 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
12913 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
12914 duplicating each output frame to accommodate the originally detected frame
12918 Display @code{5} pictures in an area of @code{3x2} frames,
12919 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
12920 mixed flat and named options:
12922 tile=3x2:nb_frames=5:padding=7:margin=2
12926 @section tinterlace
12928 Perform various types of temporal field interlacing.
12930 Frames are counted starting from 1, so the first input frame is
12933 The filter accepts the following options:
12938 Specify the mode of the interlacing. This option can also be specified
12939 as a value alone. See below for a list of values for this option.
12941 Available values are:
12945 Move odd frames into the upper field, even into the lower field,
12946 generating a double height frame at half frame rate.
12950 Frame 1 Frame 2 Frame 3 Frame 4
12952 11111 22222 33333 44444
12953 11111 22222 33333 44444
12954 11111 22222 33333 44444
12955 11111 22222 33333 44444
12969 Only output odd frames, even frames are dropped, generating a frame with
12970 unchanged height at half frame rate.
12975 Frame 1 Frame 2 Frame 3 Frame 4
12977 11111 22222 33333 44444
12978 11111 22222 33333 44444
12979 11111 22222 33333 44444
12980 11111 22222 33333 44444
12990 Only output even frames, odd frames are dropped, generating a frame with
12991 unchanged height at half frame rate.
12996 Frame 1 Frame 2 Frame 3 Frame 4
12998 11111 22222 33333 44444
12999 11111 22222 33333 44444
13000 11111 22222 33333 44444
13001 11111 22222 33333 44444
13011 Expand each frame to full height, but pad alternate lines with black,
13012 generating a frame with double height at the same input frame rate.
13017 Frame 1 Frame 2 Frame 3 Frame 4
13019 11111 22222 33333 44444
13020 11111 22222 33333 44444
13021 11111 22222 33333 44444
13022 11111 22222 33333 44444
13025 11111 ..... 33333 .....
13026 ..... 22222 ..... 44444
13027 11111 ..... 33333 .....
13028 ..... 22222 ..... 44444
13029 11111 ..... 33333 .....
13030 ..... 22222 ..... 44444
13031 11111 ..... 33333 .....
13032 ..... 22222 ..... 44444
13036 @item interleave_top, 4
13037 Interleave the upper field from odd frames with the lower field from
13038 even frames, generating a frame with unchanged height at half frame rate.
13043 Frame 1 Frame 2 Frame 3 Frame 4
13045 11111<- 22222 33333<- 44444
13046 11111 22222<- 33333 44444<-
13047 11111<- 22222 33333<- 44444
13048 11111 22222<- 33333 44444<-
13058 @item interleave_bottom, 5
13059 Interleave the lower field from odd frames with the upper field from
13060 even frames, generating a frame with unchanged height at half frame rate.
13065 Frame 1 Frame 2 Frame 3 Frame 4
13067 11111 22222<- 33333 44444<-
13068 11111<- 22222 33333<- 44444
13069 11111 22222<- 33333 44444<-
13070 11111<- 22222 33333<- 44444
13080 @item interlacex2, 6
13081 Double frame rate with unchanged height. Frames are inserted each
13082 containing the second temporal field from the previous input frame and
13083 the first temporal field from the next input frame. This mode relies on
13084 the top_field_first flag. Useful for interlaced video displays with no
13085 field synchronisation.
13090 Frame 1 Frame 2 Frame 3 Frame 4
13092 11111 22222 33333 44444
13093 11111 22222 33333 44444
13094 11111 22222 33333 44444
13095 11111 22222 33333 44444
13098 11111 22222 22222 33333 33333 44444 44444
13099 11111 11111 22222 22222 33333 33333 44444
13100 11111 22222 22222 33333 33333 44444 44444
13101 11111 11111 22222 22222 33333 33333 44444
13106 Move odd frames into the upper field, even into the lower field,
13107 generating a double height frame at same frame rate.
13112 Frame 1 Frame 2 Frame 3 Frame 4
13114 11111 22222 33333 44444
13115 11111 22222 33333 44444
13116 11111 22222 33333 44444
13117 11111 22222 33333 44444
13120 11111 33333 33333 55555
13121 22222 22222 44444 44444
13122 11111 33333 33333 55555
13123 22222 22222 44444 44444
13124 11111 33333 33333 55555
13125 22222 22222 44444 44444
13126 11111 33333 33333 55555
13127 22222 22222 44444 44444
13132 Numeric values are deprecated but are accepted for backward
13133 compatibility reasons.
13135 Default mode is @code{merge}.
13138 Specify flags influencing the filter process.
13140 Available value for @var{flags} is:
13143 @item low_pass_filter, vlfp
13144 Enable vertical low-pass filtering in the filter.
13145 Vertical low-pass filtering is required when creating an interlaced
13146 destination from a progressive source which contains high-frequency
13147 vertical detail. Filtering will reduce interlace 'twitter' and Moire
13150 Vertical low-pass filtering can only be enabled for @option{mode}
13151 @var{interleave_top} and @var{interleave_bottom}.
13158 Transpose rows with columns in the input video and optionally flip it.
13160 It accepts the following parameters:
13165 Specify the transposition direction.
13167 Can assume the following values:
13169 @item 0, 4, cclock_flip
13170 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
13178 Rotate by 90 degrees clockwise, that is:
13186 Rotate by 90 degrees counterclockwise, that is:
13193 @item 3, 7, clock_flip
13194 Rotate by 90 degrees clockwise and vertically flip, that is:
13202 For values between 4-7, the transposition is only done if the input
13203 video geometry is portrait and not landscape. These values are
13204 deprecated, the @code{passthrough} option should be used instead.
13206 Numerical values are deprecated, and should be dropped in favor of
13207 symbolic constants.
13210 Do not apply the transposition if the input geometry matches the one
13211 specified by the specified value. It accepts the following values:
13214 Always apply transposition.
13216 Preserve portrait geometry (when @var{height} >= @var{width}).
13218 Preserve landscape geometry (when @var{width} >= @var{height}).
13221 Default value is @code{none}.
13224 For example to rotate by 90 degrees clockwise and preserve portrait
13227 transpose=dir=1:passthrough=portrait
13230 The command above can also be specified as:
13232 transpose=1:portrait
13236 Trim the input so that the output contains one continuous subpart of the input.
13238 It accepts the following parameters:
13241 Specify the time of the start of the kept section, i.e. the frame with the
13242 timestamp @var{start} will be the first frame in the output.
13245 Specify the time of the first frame that will be dropped, i.e. the frame
13246 immediately preceding the one with the timestamp @var{end} will be the last
13247 frame in the output.
13250 This is the same as @var{start}, except this option sets the start timestamp
13251 in timebase units instead of seconds.
13254 This is the same as @var{end}, except this option sets the end timestamp
13255 in timebase units instead of seconds.
13258 The maximum duration of the output in seconds.
13261 The number of the first frame that should be passed to the output.
13264 The number of the first frame that should be dropped.
13267 @option{start}, @option{end}, and @option{duration} are expressed as time
13268 duration specifications; see
13269 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13270 for the accepted syntax.
13272 Note that the first two sets of the start/end options and the @option{duration}
13273 option look at the frame timestamp, while the _frame variants simply count the
13274 frames that pass through the filter. Also note that this filter does not modify
13275 the timestamps. If you wish for the output timestamps to start at zero, insert a
13276 setpts filter after the trim filter.
13278 If multiple start or end options are set, this filter tries to be greedy and
13279 keep all the frames that match at least one of the specified constraints. To keep
13280 only the part that matches all the constraints at once, chain multiple trim
13283 The defaults are such that all the input is kept. So it is possible to set e.g.
13284 just the end values to keep everything before the specified time.
13289 Drop everything except the second minute of input:
13291 ffmpeg -i INPUT -vf trim=60:120
13295 Keep only the first second:
13297 ffmpeg -i INPUT -vf trim=duration=1
13306 Sharpen or blur the input video.
13308 It accepts the following parameters:
13311 @item luma_msize_x, lx
13312 Set the luma matrix horizontal size. It must be an odd integer between
13313 3 and 63. The default value is 5.
13315 @item luma_msize_y, ly
13316 Set the luma matrix vertical size. It must be an odd integer between 3
13317 and 63. The default value is 5.
13319 @item luma_amount, la
13320 Set the luma effect strength. It must be a floating point number, reasonable
13321 values lay between -1.5 and 1.5.
13323 Negative values will blur the input video, while positive values will
13324 sharpen it, a value of zero will disable the effect.
13326 Default value is 1.0.
13328 @item chroma_msize_x, cx
13329 Set the chroma matrix horizontal size. It must be an odd integer
13330 between 3 and 63. The default value is 5.
13332 @item chroma_msize_y, cy
13333 Set the chroma matrix vertical size. It must be an odd integer
13334 between 3 and 63. The default value is 5.
13336 @item chroma_amount, ca
13337 Set the chroma effect strength. It must be a floating point number, reasonable
13338 values lay between -1.5 and 1.5.
13340 Negative values will blur the input video, while positive values will
13341 sharpen it, a value of zero will disable the effect.
13343 Default value is 0.0.
13346 If set to 1, specify using OpenCL capabilities, only available if
13347 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13351 All parameters are optional and default to the equivalent of the
13352 string '5:5:1.0:5:5:0.0'.
13354 @subsection Examples
13358 Apply strong luma sharpen effect:
13360 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13364 Apply a strong blur of both luma and chroma parameters:
13366 unsharp=7:7:-2:7:7:-2
13372 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13373 the image at several (or - in the case of @option{quality} level @code{8} - all)
13374 shifts and average the results.
13376 The way this differs from the behavior of spp is that uspp actually encodes &
13377 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13378 DCT similar to MJPEG.
13380 The filter accepts the following options:
13384 Set quality. This option defines the number of levels for averaging. It accepts
13385 an integer in the range 0-8. If set to @code{0}, the filter will have no
13386 effect. A value of @code{8} means the higher quality. For each increment of
13387 that value the speed drops by a factor of approximately 2. Default value is
13391 Force a constant quantization parameter. If not set, the filter will use the QP
13392 from the video stream (if available).
13395 @section vaguedenoiser
13397 Apply a wavelet based denoiser.
13399 It transforms each frame from the video input into the wavelet domain,
13400 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
13401 the obtained coefficients. It does an inverse wavelet transform after.
13402 Due to wavelet properties, it should give a nice smoothed result, and
13403 reduced noise, without blurring picture features.
13405 This filter accepts the following options:
13409 The filtering strength. The higher, the more filtered the video will be.
13410 Hard thresholding can use a higher threshold than soft thresholding
13411 before the video looks overfiltered.
13414 The filtering method the filter will use.
13416 It accepts the following values:
13419 All values under the threshold will be zeroed.
13422 All values under the threshold will be zeroed. All values above will be
13423 reduced by the threshold.
13426 Scales or nullifies coefficients - intermediary between (more) soft and
13427 (less) hard thresholding.
13431 Number of times, the wavelet will decompose the picture. Picture can't
13432 be decomposed beyond a particular point (typically, 8 for a 640x480
13433 frame - as 2^9 = 512 > 480)
13436 Partial of full denoising (limited coefficients shrinking), from 0 to 100.
13439 A list of the planes to process. By default all planes are processed.
13442 @section vectorscope
13444 Display 2 color component values in the two dimensional graph (which is called
13447 This filter accepts the following options:
13451 Set vectorscope mode.
13453 It accepts the following values:
13456 Gray values are displayed on graph, higher brightness means more pixels have
13457 same component color value on location in graph. This is the default mode.
13460 Gray values are displayed on graph. Surrounding pixels values which are not
13461 present in video frame are drawn in gradient of 2 color components which are
13462 set by option @code{x} and @code{y}. The 3rd color component is static.
13465 Actual color components values present in video frame are displayed on graph.
13468 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13469 on graph increases value of another color component, which is luminance by
13470 default values of @code{x} and @code{y}.
13473 Actual colors present in video frame are displayed on graph. If two different
13474 colors map to same position on graph then color with higher value of component
13475 not present in graph is picked.
13478 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13479 component picked from radial gradient.
13483 Set which color component will be represented on X-axis. Default is @code{1}.
13486 Set which color component will be represented on Y-axis. Default is @code{2}.
13489 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13490 of color component which represents frequency of (X, Y) location in graph.
13495 No envelope, this is default.
13498 Instant envelope, even darkest single pixel will be clearly highlighted.
13501 Hold maximum and minimum values presented in graph over time. This way you
13502 can still spot out of range values without constantly looking at vectorscope.
13505 Peak and instant envelope combined together.
13509 Set what kind of graticule to draw.
13517 Set graticule opacity.
13520 Set graticule flags.
13524 Draw graticule for white point.
13527 Draw graticule for black point.
13530 Draw color points short names.
13534 Set background opacity.
13536 @item lthreshold, l
13537 Set low threshold for color component not represented on X or Y axis.
13538 Values lower than this value will be ignored. Default is 0.
13539 Note this value is multiplied with actual max possible value one pixel component
13540 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13543 @item hthreshold, h
13544 Set high threshold for color component not represented on X or Y axis.
13545 Values higher than this value will be ignored. Default is 1.
13546 Note this value is multiplied with actual max possible value one pixel component
13547 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13548 is 0.9 * 255 = 230.
13550 @item colorspace, c
13551 Set what kind of colorspace to use when drawing graticule.
13560 @anchor{vidstabdetect}
13561 @section vidstabdetect
13563 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13564 @ref{vidstabtransform} for pass 2.
13566 This filter generates a file with relative translation and rotation
13567 transform information about subsequent frames, which is then used by
13568 the @ref{vidstabtransform} filter.
13570 To enable compilation of this filter you need to configure FFmpeg with
13571 @code{--enable-libvidstab}.
13573 This filter accepts the following options:
13577 Set the path to the file used to write the transforms information.
13578 Default value is @file{transforms.trf}.
13581 Set how shaky the video is and how quick the camera is. It accepts an
13582 integer in the range 1-10, a value of 1 means little shakiness, a
13583 value of 10 means strong shakiness. Default value is 5.
13586 Set the accuracy of the detection process. It must be a value in the
13587 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13588 accuracy. Default value is 15.
13591 Set stepsize of the search process. The region around minimum is
13592 scanned with 1 pixel resolution. Default value is 6.
13595 Set minimum contrast. Below this value a local measurement field is
13596 discarded. Must be a floating point value in the range 0-1. Default
13600 Set reference frame number for tripod mode.
13602 If enabled, the motion of the frames is compared to a reference frame
13603 in the filtered stream, identified by the specified number. The idea
13604 is to compensate all movements in a more-or-less static scene and keep
13605 the camera view absolutely still.
13607 If set to 0, it is disabled. The frames are counted starting from 1.
13610 Show fields and transforms in the resulting frames. It accepts an
13611 integer in the range 0-2. Default value is 0, which disables any
13615 @subsection Examples
13619 Use default values:
13625 Analyze strongly shaky movie and put the results in file
13626 @file{mytransforms.trf}:
13628 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13632 Visualize the result of internal transformations in the resulting
13635 vidstabdetect=show=1
13639 Analyze a video with medium shakiness using @command{ffmpeg}:
13641 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13645 @anchor{vidstabtransform}
13646 @section vidstabtransform
13648 Video stabilization/deshaking: pass 2 of 2,
13649 see @ref{vidstabdetect} for pass 1.
13651 Read a file with transform information for each frame and
13652 apply/compensate them. Together with the @ref{vidstabdetect}
13653 filter this can be used to deshake videos. See also
13654 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13655 the @ref{unsharp} filter, see below.
13657 To enable compilation of this filter you need to configure FFmpeg with
13658 @code{--enable-libvidstab}.
13660 @subsection Options
13664 Set path to the file used to read the transforms. Default value is
13665 @file{transforms.trf}.
13668 Set the number of frames (value*2 + 1) used for lowpass filtering the
13669 camera movements. Default value is 10.
13671 For example a number of 10 means that 21 frames are used (10 in the
13672 past and 10 in the future) to smoothen the motion in the video. A
13673 larger value leads to a smoother video, but limits the acceleration of
13674 the camera (pan/tilt movements). 0 is a special case where a static
13675 camera is simulated.
13678 Set the camera path optimization algorithm.
13680 Accepted values are:
13683 gaussian kernel low-pass filter on camera motion (default)
13685 averaging on transformations
13689 Set maximal number of pixels to translate frames. Default value is -1,
13693 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13694 value is -1, meaning no limit.
13697 Specify how to deal with borders that may be visible due to movement
13700 Available values are:
13703 keep image information from previous frame (default)
13705 fill the border black
13709 Invert transforms if set to 1. Default value is 0.
13712 Consider transforms as relative to previous frame if set to 1,
13713 absolute if set to 0. Default value is 0.
13716 Set percentage to zoom. A positive value will result in a zoom-in
13717 effect, a negative value in a zoom-out effect. Default value is 0 (no
13721 Set optimal zooming to avoid borders.
13723 Accepted values are:
13728 optimal static zoom value is determined (only very strong movements
13729 will lead to visible borders) (default)
13731 optimal adaptive zoom value is determined (no borders will be
13732 visible), see @option{zoomspeed}
13735 Note that the value given at zoom is added to the one calculated here.
13738 Set percent to zoom maximally each frame (enabled when
13739 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13743 Specify type of interpolation.
13745 Available values are:
13750 linear only horizontal
13752 linear in both directions (default)
13754 cubic in both directions (slow)
13758 Enable virtual tripod mode if set to 1, which is equivalent to
13759 @code{relative=0:smoothing=0}. Default value is 0.
13761 Use also @code{tripod} option of @ref{vidstabdetect}.
13764 Increase log verbosity if set to 1. Also the detected global motions
13765 are written to the temporary file @file{global_motions.trf}. Default
13769 @subsection Examples
13773 Use @command{ffmpeg} for a typical stabilization with default values:
13775 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13778 Note the use of the @ref{unsharp} filter which is always recommended.
13781 Zoom in a bit more and load transform data from a given file:
13783 vidstabtransform=zoom=5:input="mytransforms.trf"
13787 Smoothen the video even more:
13789 vidstabtransform=smoothing=30
13795 Flip the input video vertically.
13797 For example, to vertically flip a video with @command{ffmpeg}:
13799 ffmpeg -i in.avi -vf "vflip" out.avi
13805 Make or reverse a natural vignetting effect.
13807 The filter accepts the following options:
13811 Set lens angle expression as a number of radians.
13813 The value is clipped in the @code{[0,PI/2]} range.
13815 Default value: @code{"PI/5"}
13819 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
13823 Set forward/backward mode.
13825 Available modes are:
13828 The larger the distance from the central point, the darker the image becomes.
13831 The larger the distance from the central point, the brighter the image becomes.
13832 This can be used to reverse a vignette effect, though there is no automatic
13833 detection to extract the lens @option{angle} and other settings (yet). It can
13834 also be used to create a burning effect.
13837 Default value is @samp{forward}.
13840 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
13842 It accepts the following values:
13845 Evaluate expressions only once during the filter initialization.
13848 Evaluate expressions for each incoming frame. This is way slower than the
13849 @samp{init} mode since it requires all the scalers to be re-computed, but it
13850 allows advanced dynamic expressions.
13853 Default value is @samp{init}.
13856 Set dithering to reduce the circular banding effects. Default is @code{1}
13860 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
13861 Setting this value to the SAR of the input will make a rectangular vignetting
13862 following the dimensions of the video.
13864 Default is @code{1/1}.
13867 @subsection Expressions
13869 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
13870 following parameters.
13875 input width and height
13878 the number of input frame, starting from 0
13881 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
13882 @var{TB} units, NAN if undefined
13885 frame rate of the input video, NAN if the input frame rate is unknown
13888 the PTS (Presentation TimeStamp) of the filtered video frame,
13889 expressed in seconds, NAN if undefined
13892 time base of the input video
13896 @subsection Examples
13900 Apply simple strong vignetting effect:
13906 Make a flickering vignetting:
13908 vignette='PI/4+random(1)*PI/50':eval=frame
13914 Stack input videos vertically.
13916 All streams must be of same pixel format and of same width.
13918 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13919 to create same output.
13921 The filter accept the following option:
13925 Set number of input streams. Default is 2.
13928 If set to 1, force the output to terminate when the shortest input
13929 terminates. Default value is 0.
13934 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
13935 Deinterlacing Filter").
13937 Based on the process described by Martin Weston for BBC R&D, and
13938 implemented based on the de-interlace algorithm written by Jim
13939 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
13940 uses filter coefficients calculated by BBC R&D.
13942 There are two sets of filter coefficients, so called "simple":
13943 and "complex". Which set of filter coefficients is used can
13944 be set by passing an optional parameter:
13948 Set the interlacing filter coefficients. Accepts one of the following values:
13952 Simple filter coefficient set.
13954 More-complex filter coefficient set.
13956 Default value is @samp{complex}.
13959 Specify which frames to deinterlace. Accept one of the following values:
13963 Deinterlace all frames,
13965 Only deinterlace frames marked as interlaced.
13968 Default value is @samp{all}.
13972 Video waveform monitor.
13974 The waveform monitor plots color component intensity. By default luminance
13975 only. Each column of the waveform corresponds to a column of pixels in the
13978 It accepts the following options:
13982 Can be either @code{row}, or @code{column}. Default is @code{column}.
13983 In row mode, the graph on the left side represents color component value 0 and
13984 the right side represents value = 255. In column mode, the top side represents
13985 color component value = 0 and bottom side represents value = 255.
13988 Set intensity. Smaller values are useful to find out how many values of the same
13989 luminance are distributed across input rows/columns.
13990 Default value is @code{0.04}. Allowed range is [0, 1].
13993 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
13994 In mirrored mode, higher values will be represented on the left
13995 side for @code{row} mode and at the top for @code{column} mode. Default is
13996 @code{1} (mirrored).
14000 It accepts the following values:
14003 Presents information identical to that in the @code{parade}, except
14004 that the graphs representing color components are superimposed directly
14007 This display mode makes it easier to spot relative differences or similarities
14008 in overlapping areas of the color components that are supposed to be identical,
14009 such as neutral whites, grays, or blacks.
14012 Display separate graph for the color components side by side in
14013 @code{row} mode or one below the other in @code{column} mode.
14016 Display separate graph for the color components side by side in
14017 @code{column} mode or one below the other in @code{row} mode.
14019 Using this display mode makes it easy to spot color casts in the highlights
14020 and shadows of an image, by comparing the contours of the top and the bottom
14021 graphs of each waveform. Since whites, grays, and blacks are characterized
14022 by exactly equal amounts of red, green, and blue, neutral areas of the picture
14023 should display three waveforms of roughly equal width/height. If not, the
14024 correction is easy to perform by making level adjustments the three waveforms.
14026 Default is @code{stack}.
14028 @item components, c
14029 Set which color components to display. Default is 1, which means only luminance
14030 or red color component if input is in RGB colorspace. If is set for example to
14031 7 it will display all 3 (if) available color components.
14036 No envelope, this is default.
14039 Instant envelope, minimum and maximum values presented in graph will be easily
14040 visible even with small @code{step} value.
14043 Hold minimum and maximum values presented in graph across time. This way you
14044 can still spot out of range values without constantly looking at waveforms.
14047 Peak and instant envelope combined together.
14053 No filtering, this is default.
14056 Luma and chroma combined together.
14059 Similar as above, but shows difference between blue and red chroma.
14062 Displays only chroma.
14065 Displays actual color value on waveform.
14068 Similar as above, but with luma showing frequency of chroma values.
14072 Set which graticule to display.
14076 Do not display graticule.
14079 Display green graticule showing legal broadcast ranges.
14083 Set graticule opacity.
14086 Set graticule flags.
14090 Draw numbers above lines. By default enabled.
14093 Draw dots instead of lines.
14097 Set scale used for displaying graticule.
14104 Default is digital.
14109 The @code{weave} takes a field-based video input and join
14110 each two sequential fields into single frame, producing a new double
14111 height clip with half the frame rate and half the frame count.
14113 It accepts the following option:
14117 Set first field. Available values are:
14121 Set the frame as top-field-first.
14124 Set the frame as bottom-field-first.
14129 Apply the xBR high-quality magnification filter which is designed for pixel
14130 art. It follows a set of edge-detection rules, see
14131 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
14133 It accepts the following option:
14137 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
14138 @code{3xBR} and @code{4} for @code{4xBR}.
14139 Default is @code{3}.
14145 Deinterlace the input video ("yadif" means "yet another deinterlacing
14148 It accepts the following parameters:
14154 The interlacing mode to adopt. It accepts one of the following values:
14157 @item 0, send_frame
14158 Output one frame for each frame.
14159 @item 1, send_field
14160 Output one frame for each field.
14161 @item 2, send_frame_nospatial
14162 Like @code{send_frame}, but it skips the spatial interlacing check.
14163 @item 3, send_field_nospatial
14164 Like @code{send_field}, but it skips the spatial interlacing check.
14167 The default value is @code{send_frame}.
14170 The picture field parity assumed for the input interlaced video. It accepts one
14171 of the following values:
14175 Assume the top field is first.
14177 Assume the bottom field is first.
14179 Enable automatic detection of field parity.
14182 The default value is @code{auto}.
14183 If the interlacing is unknown or the decoder does not export this information,
14184 top field first will be assumed.
14187 Specify which frames to deinterlace. Accept one of the following
14192 Deinterlace all frames.
14193 @item 1, interlaced
14194 Only deinterlace frames marked as interlaced.
14197 The default value is @code{all}.
14202 Apply Zoom & Pan effect.
14204 This filter accepts the following options:
14208 Set the zoom expression. Default is 1.
14212 Set the x and y expression. Default is 0.
14215 Set the duration expression in number of frames.
14216 This sets for how many number of frames effect will last for
14217 single input image.
14220 Set the output image size, default is 'hd720'.
14223 Set the output frame rate, default is '25'.
14226 Each expression can contain the following constants:
14245 Output frame count.
14249 Last calculated 'x' and 'y' position from 'x' and 'y' expression
14250 for current input frame.
14254 'x' and 'y' of last output frame of previous input frame or 0 when there was
14255 not yet such frame (first input frame).
14258 Last calculated zoom from 'z' expression for current input frame.
14261 Last calculated zoom of last output frame of previous input frame.
14264 Number of output frames for current input frame. Calculated from 'd' expression
14265 for each input frame.
14268 number of output frames created for previous input frame
14271 Rational number: input width / input height
14274 sample aspect ratio
14277 display aspect ratio
14281 @subsection Examples
14285 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
14287 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
14291 Zoom-in up to 1.5 and pan always at center of picture:
14293 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14297 Same as above but without pausing:
14299 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14304 Scale (resize) the input video, using the z.lib library:
14305 https://github.com/sekrit-twc/zimg.
14307 The zscale filter forces the output display aspect ratio to be the same
14308 as the input, by changing the output sample aspect ratio.
14310 If the input image format is different from the format requested by
14311 the next filter, the zscale filter will convert the input to the
14314 @subsection Options
14315 The filter accepts the following options.
14320 Set the output video dimension expression. Default value is the input
14323 If the @var{width} or @var{w} is 0, the input width is used for the output.
14324 If the @var{height} or @var{h} is 0, the input height is used for the output.
14326 If one of the values is -1, the zscale filter will use a value that
14327 maintains the aspect ratio of the input image, calculated from the
14328 other specified dimension. If both of them are -1, the input size is
14331 If one of the values is -n with n > 1, the zscale filter will also use a value
14332 that maintains the aspect ratio of the input image, calculated from the other
14333 specified dimension. After that it will, however, make sure that the calculated
14334 dimension is divisible by n and adjust the value if necessary.
14336 See below for the list of accepted constants for use in the dimension
14340 Set the video size. For the syntax of this option, check the
14341 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14344 Set the dither type.
14346 Possible values are:
14351 @item error_diffusion
14357 Set the resize filter type.
14359 Possible values are:
14369 Default is bilinear.
14372 Set the color range.
14374 Possible values are:
14381 Default is same as input.
14384 Set the color primaries.
14386 Possible values are:
14396 Default is same as input.
14399 Set the transfer characteristics.
14401 Possible values are:
14412 Default is same as input.
14415 Set the colorspace matrix.
14417 Possible value are:
14428 Default is same as input.
14431 Set the input color range.
14433 Possible values are:
14440 Default is same as input.
14442 @item primariesin, pin
14443 Set the input color primaries.
14445 Possible values are:
14455 Default is same as input.
14457 @item transferin, tin
14458 Set the input transfer characteristics.
14460 Possible values are:
14471 Default is same as input.
14473 @item matrixin, min
14474 Set the input colorspace matrix.
14476 Possible value are:
14488 Set the output chroma location.
14490 Possible values are:
14501 @item chromalin, cin
14502 Set the input chroma location.
14504 Possible values are:
14516 The values of the @option{w} and @option{h} options are expressions
14517 containing the following constants:
14522 The input width and height
14526 These are the same as @var{in_w} and @var{in_h}.
14530 The output (scaled) width and height
14534 These are the same as @var{out_w} and @var{out_h}
14537 The same as @var{iw} / @var{ih}
14540 input sample aspect ratio
14543 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14547 horizontal and vertical input chroma subsample values. For example for the
14548 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14552 horizontal and vertical output chroma subsample values. For example for the
14553 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14559 @c man end VIDEO FILTERS
14561 @chapter Video Sources
14562 @c man begin VIDEO SOURCES
14564 Below is a description of the currently available video sources.
14568 Buffer video frames, and make them available to the filter chain.
14570 This source is mainly intended for a programmatic use, in particular
14571 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14573 It accepts the following parameters:
14578 Specify the size (width and height) of the buffered video frames. For the
14579 syntax of this option, check the
14580 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14583 The input video width.
14586 The input video height.
14589 A string representing the pixel format of the buffered video frames.
14590 It may be a number corresponding to a pixel format, or a pixel format
14594 Specify the timebase assumed by the timestamps of the buffered frames.
14597 Specify the frame rate expected for the video stream.
14599 @item pixel_aspect, sar
14600 The sample (pixel) aspect ratio of the input video.
14603 Specify the optional parameters to be used for the scale filter which
14604 is automatically inserted when an input change is detected in the
14605 input size or format.
14607 @item hw_frames_ctx
14608 When using a hardware pixel format, this should be a reference to an
14609 AVHWFramesContext describing input frames.
14614 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14617 will instruct the source to accept video frames with size 320x240 and
14618 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14619 square pixels (1:1 sample aspect ratio).
14620 Since the pixel format with name "yuv410p" corresponds to the number 6
14621 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14622 this example corresponds to:
14624 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14627 Alternatively, the options can be specified as a flat string, but this
14628 syntax is deprecated:
14630 @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}]
14634 Create a pattern generated by an elementary cellular automaton.
14636 The initial state of the cellular automaton can be defined through the
14637 @option{filename}, and @option{pattern} options. If such options are
14638 not specified an initial state is created randomly.
14640 At each new frame a new row in the video is filled with the result of
14641 the cellular automaton next generation. The behavior when the whole
14642 frame is filled is defined by the @option{scroll} option.
14644 This source accepts the following options:
14648 Read the initial cellular automaton state, i.e. the starting row, from
14649 the specified file.
14650 In the file, each non-whitespace character is considered an alive
14651 cell, a newline will terminate the row, and further characters in the
14652 file will be ignored.
14655 Read the initial cellular automaton state, i.e. the starting row, from
14656 the specified string.
14658 Each non-whitespace character in the string is considered an alive
14659 cell, a newline will terminate the row, and further characters in the
14660 string will be ignored.
14663 Set the video rate, that is the number of frames generated per second.
14666 @item random_fill_ratio, ratio
14667 Set the random fill ratio for the initial cellular automaton row. It
14668 is a floating point number value ranging from 0 to 1, defaults to
14671 This option is ignored when a file or a pattern is specified.
14673 @item random_seed, seed
14674 Set the seed for filling randomly the initial row, must be an integer
14675 included between 0 and UINT32_MAX. If not specified, or if explicitly
14676 set to -1, the filter will try to use a good random seed on a best
14680 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14681 Default value is 110.
14684 Set the size of the output video. For the syntax of this option, check the
14685 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14687 If @option{filename} or @option{pattern} is specified, the size is set
14688 by default to the width of the specified initial state row, and the
14689 height is set to @var{width} * PHI.
14691 If @option{size} is set, it must contain the width of the specified
14692 pattern string, and the specified pattern will be centered in the
14695 If a filename or a pattern string is not specified, the size value
14696 defaults to "320x518" (used for a randomly generated initial state).
14699 If set to 1, scroll the output upward when all the rows in the output
14700 have been already filled. If set to 0, the new generated row will be
14701 written over the top row just after the bottom row is filled.
14704 @item start_full, full
14705 If set to 1, completely fill the output with generated rows before
14706 outputting the first frame.
14707 This is the default behavior, for disabling set the value to 0.
14710 If set to 1, stitch the left and right row edges together.
14711 This is the default behavior, for disabling set the value to 0.
14714 @subsection Examples
14718 Read the initial state from @file{pattern}, and specify an output of
14721 cellauto=f=pattern:s=200x400
14725 Generate a random initial row with a width of 200 cells, with a fill
14728 cellauto=ratio=2/3:s=200x200
14732 Create a pattern generated by rule 18 starting by a single alive cell
14733 centered on an initial row with width 100:
14735 cellauto=p=@@:s=100x400:full=0:rule=18
14739 Specify a more elaborated initial pattern:
14741 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14746 @anchor{coreimagesrc}
14747 @section coreimagesrc
14748 Video source generated on GPU using Apple's CoreImage API on OSX.
14750 This video source is a specialized version of the @ref{coreimage} video filter.
14751 Use a core image generator at the beginning of the applied filterchain to
14752 generate the content.
14754 The coreimagesrc video source accepts the following options:
14756 @item list_generators
14757 List all available generators along with all their respective options as well as
14758 possible minimum and maximum values along with the default values.
14760 list_generators=true
14764 Specify the size of the sourced video. For the syntax of this option, check the
14765 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14766 The default value is @code{320x240}.
14769 Specify the frame rate of the sourced video, as the number of frames
14770 generated per second. It has to be a string in the format
14771 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14772 number or a valid video frame rate abbreviation. The default value is
14776 Set the sample aspect ratio of the sourced video.
14779 Set the duration of the sourced video. See
14780 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14781 for the accepted syntax.
14783 If not specified, or the expressed duration is negative, the video is
14784 supposed to be generated forever.
14787 Additionally, all options of the @ref{coreimage} video filter are accepted.
14788 A complete filterchain can be used for further processing of the
14789 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
14790 and examples for details.
14792 @subsection Examples
14797 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
14798 given as complete and escaped command-line for Apple's standard bash shell:
14800 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
14802 This example is equivalent to the QRCode example of @ref{coreimage} without the
14803 need for a nullsrc video source.
14807 @section mandelbrot
14809 Generate a Mandelbrot set fractal, and progressively zoom towards the
14810 point specified with @var{start_x} and @var{start_y}.
14812 This source accepts the following options:
14817 Set the terminal pts value. Default value is 400.
14820 Set the terminal scale value.
14821 Must be a floating point value. Default value is 0.3.
14824 Set the inner coloring mode, that is the algorithm used to draw the
14825 Mandelbrot fractal internal region.
14827 It shall assume one of the following values:
14832 Show time until convergence.
14834 Set color based on point closest to the origin of the iterations.
14839 Default value is @var{mincol}.
14842 Set the bailout value. Default value is 10.0.
14845 Set the maximum of iterations performed by the rendering
14846 algorithm. Default value is 7189.
14849 Set outer coloring mode.
14850 It shall assume one of following values:
14852 @item iteration_count
14853 Set iteration cound mode.
14854 @item normalized_iteration_count
14855 set normalized iteration count mode.
14857 Default value is @var{normalized_iteration_count}.
14860 Set frame rate, expressed as number of frames per second. Default
14864 Set frame size. For the syntax of this option, check the "Video
14865 size" section in the ffmpeg-utils manual. Default value is "640x480".
14868 Set the initial scale value. Default value is 3.0.
14871 Set the initial x position. Must be a floating point value between
14872 -100 and 100. Default value is -0.743643887037158704752191506114774.
14875 Set the initial y position. Must be a floating point value between
14876 -100 and 100. Default value is -0.131825904205311970493132056385139.
14881 Generate various test patterns, as generated by the MPlayer test filter.
14883 The size of the generated video is fixed, and is 256x256.
14884 This source is useful in particular for testing encoding features.
14886 This source accepts the following options:
14891 Specify the frame rate of the sourced video, as the number of frames
14892 generated per second. It has to be a string in the format
14893 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14894 number or a valid video frame rate abbreviation. The default value is
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.
14907 Set the number or the name of the test to perform. Supported tests are:
14923 Default value is "all", which will cycle through the list of all tests.
14928 mptestsrc=t=dc_luma
14931 will generate a "dc_luma" test pattern.
14933 @section frei0r_src
14935 Provide a frei0r source.
14937 To enable compilation of this filter you need to install the frei0r
14938 header and configure FFmpeg with @code{--enable-frei0r}.
14940 This source accepts the following parameters:
14945 The size of the video to generate. For the syntax of this option, check the
14946 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14949 The framerate of the generated video. It may be a string of the form
14950 @var{num}/@var{den} or a frame rate abbreviation.
14953 The name to the frei0r source to load. For more information regarding frei0r and
14954 how to set the parameters, read the @ref{frei0r} section in the video filters
14957 @item filter_params
14958 A '|'-separated list of parameters to pass to the frei0r source.
14962 For example, to generate a frei0r partik0l source with size 200x200
14963 and frame rate 10 which is overlaid on the overlay filter main input:
14965 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
14970 Generate a life pattern.
14972 This source is based on a generalization of John Conway's life game.
14974 The sourced input represents a life grid, each pixel represents a cell
14975 which can be in one of two possible states, alive or dead. Every cell
14976 interacts with its eight neighbours, which are the cells that are
14977 horizontally, vertically, or diagonally adjacent.
14979 At each interaction the grid evolves according to the adopted rule,
14980 which specifies the number of neighbor alive cells which will make a
14981 cell stay alive or born. The @option{rule} option allows one to specify
14984 This source accepts the following options:
14988 Set the file from which to read the initial grid state. In the file,
14989 each non-whitespace character is considered an alive cell, and newline
14990 is used to delimit the end of each row.
14992 If this option is not specified, the initial grid is generated
14996 Set the video rate, that is the number of frames generated per second.
14999 @item random_fill_ratio, ratio
15000 Set the random fill ratio for the initial random grid. It is a
15001 floating point number value ranging from 0 to 1, defaults to 1/PHI.
15002 It is ignored when a file is specified.
15004 @item random_seed, seed
15005 Set the seed for filling the initial random grid, must be an integer
15006 included between 0 and UINT32_MAX. If not specified, or if explicitly
15007 set to -1, the filter will try to use a good random seed on a best
15013 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
15014 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
15015 @var{NS} specifies the number of alive neighbor cells which make a
15016 live cell stay alive, and @var{NB} the number of alive neighbor cells
15017 which make a dead cell to become alive (i.e. to "born").
15018 "s" and "b" can be used in place of "S" and "B", respectively.
15020 Alternatively a rule can be specified by an 18-bits integer. The 9
15021 high order bits are used to encode the next cell state if it is alive
15022 for each number of neighbor alive cells, the low order bits specify
15023 the rule for "borning" new cells. Higher order bits encode for an
15024 higher number of neighbor cells.
15025 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
15026 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
15028 Default value is "S23/B3", which is the original Conway's game of life
15029 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
15030 cells, and will born a new cell if there are three alive cells around
15034 Set the size of the output video. For the syntax of this option, check the
15035 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15037 If @option{filename} is specified, the size is set by default to the
15038 same size of the input file. If @option{size} is set, it must contain
15039 the size specified in the input file, and the initial grid defined in
15040 that file is centered in the larger resulting area.
15042 If a filename is not specified, the size value defaults to "320x240"
15043 (used for a randomly generated initial grid).
15046 If set to 1, stitch the left and right grid edges together, and the
15047 top and bottom edges also. Defaults to 1.
15050 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
15051 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
15052 value from 0 to 255.
15055 Set the color of living (or new born) cells.
15058 Set the color of dead cells. If @option{mold} is set, this is the first color
15059 used to represent a dead cell.
15062 Set mold color, for definitely dead and moldy cells.
15064 For the syntax of these 3 color options, check the "Color" section in the
15065 ffmpeg-utils manual.
15068 @subsection Examples
15072 Read a grid from @file{pattern}, and center it on a grid of size
15075 life=f=pattern:s=300x300
15079 Generate a random grid of size 200x200, with a fill ratio of 2/3:
15081 life=ratio=2/3:s=200x200
15085 Specify a custom rule for evolving a randomly generated grid:
15091 Full example with slow death effect (mold) using @command{ffplay}:
15093 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
15100 @anchor{haldclutsrc}
15102 @anchor{rgbtestsrc}
15104 @anchor{smptehdbars}
15107 @anchor{yuvtestsrc}
15108 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
15110 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
15112 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
15114 The @code{color} source provides an uniformly colored input.
15116 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
15117 @ref{haldclut} filter.
15119 The @code{nullsrc} source returns unprocessed video frames. It is
15120 mainly useful to be employed in analysis / debugging tools, or as the
15121 source for filters which ignore the input data.
15123 The @code{rgbtestsrc} source generates an RGB test pattern useful for
15124 detecting RGB vs BGR issues. You should see a red, green and blue
15125 stripe from top to bottom.
15127 The @code{smptebars} source generates a color bars pattern, based on
15128 the SMPTE Engineering Guideline EG 1-1990.
15130 The @code{smptehdbars} source generates a color bars pattern, based on
15131 the SMPTE RP 219-2002.
15133 The @code{testsrc} source generates a test video pattern, showing a
15134 color pattern, a scrolling gradient and a timestamp. This is mainly
15135 intended for testing purposes.
15137 The @code{testsrc2} source is similar to testsrc, but supports more
15138 pixel formats instead of just @code{rgb24}. This allows using it as an
15139 input for other tests without requiring a format conversion.
15141 The @code{yuvtestsrc} source generates an YUV test pattern. You should
15142 see a y, cb and cr stripe from top to bottom.
15144 The sources accept the following parameters:
15149 Specify the color of the source, only available in the @code{color}
15150 source. For the syntax of this option, check the "Color" section in the
15151 ffmpeg-utils manual.
15154 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
15155 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
15156 pixels to be used as identity matrix for 3D lookup tables. Each component is
15157 coded on a @code{1/(N*N)} scale.
15160 Specify the size of the sourced video. For the syntax of this option, check the
15161 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15162 The default value is @code{320x240}.
15164 This option is not available with the @code{haldclutsrc} filter.
15167 Specify the frame rate of the sourced video, as the number of frames
15168 generated per second. It has to be a string in the format
15169 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15170 number or a valid video frame rate abbreviation. The default value is
15174 Set the sample aspect ratio of the sourced video.
15177 Set the duration of the sourced video. See
15178 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15179 for the accepted syntax.
15181 If not specified, or the expressed duration is negative, the video is
15182 supposed to be generated forever.
15185 Set the number of decimals to show in the timestamp, only available in the
15186 @code{testsrc} source.
15188 The displayed timestamp value will correspond to the original
15189 timestamp value multiplied by the power of 10 of the specified
15190 value. Default value is 0.
15193 For example the following:
15195 testsrc=duration=5.3:size=qcif:rate=10
15198 will generate a video with a duration of 5.3 seconds, with size
15199 176x144 and a frame rate of 10 frames per second.
15201 The following graph description will generate a red source
15202 with an opacity of 0.2, with size "qcif" and a frame rate of 10
15205 color=c=red@@0.2:s=qcif:r=10
15208 If the input content is to be ignored, @code{nullsrc} can be used. The
15209 following command generates noise in the luminance plane by employing
15210 the @code{geq} filter:
15212 nullsrc=s=256x256, geq=random(1)*255:128:128
15215 @subsection Commands
15217 The @code{color} source supports the following commands:
15221 Set the color of the created image. Accepts the same syntax of the
15222 corresponding @option{color} option.
15225 @c man end VIDEO SOURCES
15227 @chapter Video Sinks
15228 @c man begin VIDEO SINKS
15230 Below is a description of the currently available video sinks.
15232 @section buffersink
15234 Buffer video frames, and make them available to the end of the filter
15237 This sink is mainly intended for programmatic use, in particular
15238 through the interface defined in @file{libavfilter/buffersink.h}
15239 or the options system.
15241 It accepts a pointer to an AVBufferSinkContext structure, which
15242 defines the incoming buffers' formats, to be passed as the opaque
15243 parameter to @code{avfilter_init_filter} for initialization.
15247 Null video sink: do absolutely nothing with the input video. It is
15248 mainly useful as a template and for use in analysis / debugging
15251 @c man end VIDEO SINKS
15253 @chapter Multimedia Filters
15254 @c man begin MULTIMEDIA FILTERS
15256 Below is a description of the currently available multimedia filters.
15258 @section ahistogram
15260 Convert input audio to a video output, displaying the volume histogram.
15262 The filter accepts the following options:
15266 Specify how histogram is calculated.
15268 It accepts the following values:
15271 Use single histogram for all channels.
15273 Use separate histogram for each channel.
15275 Default is @code{single}.
15278 Set frame rate, expressed as number of frames per second. Default
15282 Specify the video size for the output. For the syntax of this option, check the
15283 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15284 Default value is @code{hd720}.
15289 It accepts the following values:
15300 reverse logarithmic
15302 Default is @code{log}.
15305 Set amplitude scale.
15307 It accepts the following values:
15314 Default is @code{log}.
15317 Set how much frames to accumulate in histogram.
15318 Defauls is 1. Setting this to -1 accumulates all frames.
15321 Set histogram ratio of window height.
15324 Set sonogram sliding.
15326 It accepts the following values:
15329 replace old rows with new ones.
15331 scroll from top to bottom.
15333 Default is @code{replace}.
15336 @section aphasemeter
15338 Convert input audio to a video output, displaying the audio phase.
15340 The filter accepts the following options:
15344 Set the output frame rate. Default value is @code{25}.
15347 Set the video size for the output. For the syntax of this option, check the
15348 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15349 Default value is @code{800x400}.
15354 Specify the red, green, blue contrast. Default values are @code{2},
15355 @code{7} and @code{1}.
15356 Allowed range is @code{[0, 255]}.
15359 Set color which will be used for drawing median phase. If color is
15360 @code{none} which is default, no median phase value will be drawn.
15363 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15364 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15365 The @code{-1} means left and right channels are completely out of phase and
15366 @code{1} means channels are in phase.
15368 @section avectorscope
15370 Convert input audio to a video output, representing the audio vector
15373 The filter is used to measure the difference between channels of stereo
15374 audio stream. A monoaural signal, consisting of identical left and right
15375 signal, results in straight vertical line. Any stereo separation is visible
15376 as a deviation from this line, creating a Lissajous figure.
15377 If the straight (or deviation from it) but horizontal line appears this
15378 indicates that the left and right channels are out of phase.
15380 The filter accepts the following options:
15384 Set the vectorscope mode.
15386 Available values are:
15389 Lissajous rotated by 45 degrees.
15392 Same as above but not rotated.
15395 Shape resembling half of circle.
15398 Default value is @samp{lissajous}.
15401 Set the video size for the output. For the syntax of this option, check the
15402 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15403 Default value is @code{400x400}.
15406 Set the output frame rate. Default value is @code{25}.
15412 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15413 @code{160}, @code{80} and @code{255}.
15414 Allowed range is @code{[0, 255]}.
15420 Specify the red, green, blue and alpha fade. Default values are @code{15},
15421 @code{10}, @code{5} and @code{5}.
15422 Allowed range is @code{[0, 255]}.
15425 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15428 Set the vectorscope drawing mode.
15430 Available values are:
15433 Draw dot for each sample.
15436 Draw line between previous and current sample.
15439 Default value is @samp{dot}.
15442 Specify amplitude scale of audio samples.
15444 Available values are:
15461 @subsection Examples
15465 Complete example using @command{ffplay}:
15467 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15468 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15472 @section bench, abench
15474 Benchmark part of a filtergraph.
15476 The filter accepts the following options:
15480 Start or stop a timer.
15482 Available values are:
15485 Get the current time, set it as frame metadata (using the key
15486 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15489 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15490 the input frame metadata to get the time difference. Time difference, average,
15491 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15492 @code{min}) are then printed. The timestamps are expressed in seconds.
15496 @subsection Examples
15500 Benchmark @ref{selectivecolor} filter:
15502 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15508 Concatenate audio and video streams, joining them together one after the
15511 The filter works on segments of synchronized video and audio streams. All
15512 segments must have the same number of streams of each type, and that will
15513 also be the number of streams at output.
15515 The filter accepts the following options:
15520 Set the number of segments. Default is 2.
15523 Set the number of output video streams, that is also the number of video
15524 streams in each segment. Default is 1.
15527 Set the number of output audio streams, that is also the number of audio
15528 streams in each segment. Default is 0.
15531 Activate unsafe mode: do not fail if segments have a different format.
15535 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15536 @var{a} audio outputs.
15538 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15539 segment, in the same order as the outputs, then the inputs for the second
15542 Related streams do not always have exactly the same duration, for various
15543 reasons including codec frame size or sloppy authoring. For that reason,
15544 related synchronized streams (e.g. a video and its audio track) should be
15545 concatenated at once. The concat filter will use the duration of the longest
15546 stream in each segment (except the last one), and if necessary pad shorter
15547 audio streams with silence.
15549 For this filter to work correctly, all segments must start at timestamp 0.
15551 All corresponding streams must have the same parameters in all segments; the
15552 filtering system will automatically select a common pixel format for video
15553 streams, and a common sample format, sample rate and channel layout for
15554 audio streams, but other settings, such as resolution, must be converted
15555 explicitly by the user.
15557 Different frame rates are acceptable but will result in variable frame rate
15558 at output; be sure to configure the output file to handle it.
15560 @subsection Examples
15564 Concatenate an opening, an episode and an ending, all in bilingual version
15565 (video in stream 0, audio in streams 1 and 2):
15567 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15568 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15569 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15570 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15574 Concatenate two parts, handling audio and video separately, using the
15575 (a)movie sources, and adjusting the resolution:
15577 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15578 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15579 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15581 Note that a desync will happen at the stitch if the audio and video streams
15582 do not have exactly the same duration in the first file.
15586 @section drawgraph, adrawgraph
15588 Draw a graph using input video or audio metadata.
15590 It accepts the following parameters:
15594 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15597 Set 1st foreground color expression.
15600 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15603 Set 2nd foreground color expression.
15606 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15609 Set 3rd foreground color expression.
15612 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15615 Set 4th foreground color expression.
15618 Set minimal value of metadata value.
15621 Set maximal value of metadata value.
15624 Set graph background color. Default is white.
15629 Available values for mode is:
15636 Default is @code{line}.
15641 Available values for slide is:
15644 Draw new frame when right border is reached.
15647 Replace old columns with new ones.
15650 Scroll from right to left.
15653 Scroll from left to right.
15656 Draw single picture.
15659 Default is @code{frame}.
15662 Set size of graph video. For the syntax of this option, check the
15663 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15664 The default value is @code{900x256}.
15666 The foreground color expressions can use the following variables:
15669 Minimal value of metadata value.
15672 Maximal value of metadata value.
15675 Current metadata key value.
15678 The color is defined as 0xAABBGGRR.
15681 Example using metadata from @ref{signalstats} filter:
15683 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15686 Example using metadata from @ref{ebur128} filter:
15688 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15694 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15695 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15696 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15697 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15699 The filter also has a video output (see the @var{video} option) with a real
15700 time graph to observe the loudness evolution. The graphic contains the logged
15701 message mentioned above, so it is not printed anymore when this option is set,
15702 unless the verbose logging is set. The main graphing area contains the
15703 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15704 the momentary loudness (400 milliseconds).
15706 More information about the Loudness Recommendation EBU R128 on
15707 @url{http://tech.ebu.ch/loudness}.
15709 The filter accepts the following options:
15714 Activate the video output. The audio stream is passed unchanged whether this
15715 option is set or no. The video stream will be the first output stream if
15716 activated. Default is @code{0}.
15719 Set the video size. This option is for video only. For the syntax of this
15721 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15722 Default and minimum resolution is @code{640x480}.
15725 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15726 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15727 other integer value between this range is allowed.
15730 Set metadata injection. If set to @code{1}, the audio input will be segmented
15731 into 100ms output frames, each of them containing various loudness information
15732 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15734 Default is @code{0}.
15737 Force the frame logging level.
15739 Available values are:
15742 information logging level
15744 verbose logging level
15747 By default, the logging level is set to @var{info}. If the @option{video} or
15748 the @option{metadata} options are set, it switches to @var{verbose}.
15753 Available modes can be cumulated (the option is a @code{flag} type). Possible
15757 Disable any peak mode (default).
15759 Enable sample-peak mode.
15761 Simple peak mode looking for the higher sample value. It logs a message
15762 for sample-peak (identified by @code{SPK}).
15764 Enable true-peak mode.
15766 If enabled, the peak lookup is done on an over-sampled version of the input
15767 stream for better peak accuracy. It logs a message for true-peak.
15768 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15769 This mode requires a build with @code{libswresample}.
15773 Treat mono input files as "dual mono". If a mono file is intended for playback
15774 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15775 If set to @code{true}, this option will compensate for this effect.
15776 Multi-channel input files are not affected by this option.
15779 Set a specific pan law to be used for the measurement of dual mono files.
15780 This parameter is optional, and has a default value of -3.01dB.
15783 @subsection Examples
15787 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
15789 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
15793 Run an analysis with @command{ffmpeg}:
15795 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
15799 @section interleave, ainterleave
15801 Temporally interleave frames from several inputs.
15803 @code{interleave} works with video inputs, @code{ainterleave} with audio.
15805 These filters read frames from several inputs and send the oldest
15806 queued frame to the output.
15808 Input streams must have a well defined, monotonically increasing frame
15811 In order to submit one frame to output, these filters need to enqueue
15812 at least one frame for each input, so they cannot work in case one
15813 input is not yet terminated and will not receive incoming frames.
15815 For example consider the case when one input is a @code{select} filter
15816 which always drop input frames. The @code{interleave} filter will keep
15817 reading from that input, but it will never be able to send new frames
15818 to output until the input will send an end-of-stream signal.
15820 Also, depending on inputs synchronization, the filters will drop
15821 frames in case one input receives more frames than the other ones, and
15822 the queue is already filled.
15824 These filters accept the following options:
15828 Set the number of different inputs, it is 2 by default.
15831 @subsection Examples
15835 Interleave frames belonging to different streams using @command{ffmpeg}:
15837 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
15841 Add flickering blur effect:
15843 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
15847 @section metadata, ametadata
15849 Manipulate frame metadata.
15851 This filter accepts the following options:
15855 Set mode of operation of the filter.
15857 Can be one of the following:
15861 If both @code{value} and @code{key} is set, select frames
15862 which have such metadata. If only @code{key} is set, select
15863 every frame that has such key in metadata.
15866 Add new metadata @code{key} and @code{value}. If key is already available
15870 Modify value of already present key.
15873 If @code{value} is set, delete only keys that have such value.
15874 Otherwise, delete key.
15877 Print key and its value if metadata was found. If @code{key} is not set print all
15878 metadata values available in frame.
15882 Set key used with all modes. Must be set for all modes except @code{print}.
15885 Set metadata value which will be used. This option is mandatory for
15886 @code{modify} and @code{add} mode.
15889 Which function to use when comparing metadata value and @code{value}.
15891 Can be one of following:
15895 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
15898 Values are interpreted as strings, returns true if metadata value starts with
15899 the @code{value} option string.
15902 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
15905 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
15908 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
15911 Values are interpreted as floats, returns true if expression from option @code{expr}
15916 Set expression which is used when @code{function} is set to @code{expr}.
15917 The expression is evaluated through the eval API and can contain the following
15922 Float representation of @code{value} from metadata key.
15925 Float representation of @code{value} as supplied by user in @code{value} option.
15928 If specified in @code{print} mode, output is written to the named file. Instead of
15929 plain filename any writable url can be specified. Filename ``-'' is a shorthand
15930 for standard output. If @code{file} option is not set, output is written to the log
15931 with AV_LOG_INFO loglevel.
15936 @subsection Examples
15940 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
15943 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
15946 Print silencedetect output to file @file{metadata.txt}.
15948 silencedetect,ametadata=mode=print:file=metadata.txt
15951 Direct all metadata to a pipe with file descriptor 4.
15953 metadata=mode=print:file='pipe\:4'
15957 @section perms, aperms
15959 Set read/write permissions for the output frames.
15961 These filters are mainly aimed at developers to test direct path in the
15962 following filter in the filtergraph.
15964 The filters accept the following options:
15968 Select the permissions mode.
15970 It accepts the following values:
15973 Do nothing. This is the default.
15975 Set all the output frames read-only.
15977 Set all the output frames directly writable.
15979 Make the frame read-only if writable, and writable if read-only.
15981 Set each output frame read-only or writable randomly.
15985 Set the seed for the @var{random} mode, must be an integer included between
15986 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
15987 @code{-1}, the filter will try to use a good random seed on a best effort
15991 Note: in case of auto-inserted filter between the permission filter and the
15992 following one, the permission might not be received as expected in that
15993 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
15994 perms/aperms filter can avoid this problem.
15996 @section realtime, arealtime
15998 Slow down filtering to match real time approximatively.
16000 These filters will pause the filtering for a variable amount of time to
16001 match the output rate with the input timestamps.
16002 They are similar to the @option{re} option to @code{ffmpeg}.
16004 They accept the following options:
16008 Time limit for the pauses. Any pause longer than that will be considered
16009 a timestamp discontinuity and reset the timer. Default is 2 seconds.
16012 @section select, aselect
16014 Select frames to pass in output.
16016 This filter accepts the following options:
16021 Set expression, which is evaluated for each input frame.
16023 If the expression is evaluated to zero, the frame is discarded.
16025 If the evaluation result is negative or NaN, the frame is sent to the
16026 first output; otherwise it is sent to the output with index
16027 @code{ceil(val)-1}, assuming that the input index starts from 0.
16029 For example a value of @code{1.2} corresponds to the output with index
16030 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
16033 Set the number of outputs. The output to which to send the selected
16034 frame is based on the result of the evaluation. Default value is 1.
16037 The expression can contain the following constants:
16041 The (sequential) number of the filtered frame, starting from 0.
16044 The (sequential) number of the selected frame, starting from 0.
16046 @item prev_selected_n
16047 The sequential number of the last selected frame. It's NAN if undefined.
16050 The timebase of the input timestamps.
16053 The PTS (Presentation TimeStamp) of the filtered video frame,
16054 expressed in @var{TB} units. It's NAN if undefined.
16057 The PTS of the filtered video frame,
16058 expressed in seconds. It's NAN if undefined.
16061 The PTS of the previously filtered video frame. It's NAN if undefined.
16063 @item prev_selected_pts
16064 The PTS of the last previously filtered video frame. It's NAN if undefined.
16066 @item prev_selected_t
16067 The PTS of the last previously selected video frame. It's NAN if undefined.
16070 The PTS of the first video frame in the video. It's NAN if undefined.
16073 The time of the first video frame in the video. It's NAN if undefined.
16075 @item pict_type @emph{(video only)}
16076 The type of the filtered frame. It can assume one of the following
16088 @item interlace_type @emph{(video only)}
16089 The frame interlace type. It can assume one of the following values:
16092 The frame is progressive (not interlaced).
16094 The frame is top-field-first.
16096 The frame is bottom-field-first.
16099 @item consumed_sample_n @emph{(audio only)}
16100 the number of selected samples before the current frame
16102 @item samples_n @emph{(audio only)}
16103 the number of samples in the current frame
16105 @item sample_rate @emph{(audio only)}
16106 the input sample rate
16109 This is 1 if the filtered frame is a key-frame, 0 otherwise.
16112 the position in the file of the filtered frame, -1 if the information
16113 is not available (e.g. for synthetic video)
16115 @item scene @emph{(video only)}
16116 value between 0 and 1 to indicate a new scene; a low value reflects a low
16117 probability for the current frame to introduce a new scene, while a higher
16118 value means the current frame is more likely to be one (see the example below)
16120 @item concatdec_select
16121 The concat demuxer can select only part of a concat input file by setting an
16122 inpoint and an outpoint, but the output packets may not be entirely contained
16123 in the selected interval. By using this variable, it is possible to skip frames
16124 generated by the concat demuxer which are not exactly contained in the selected
16127 This works by comparing the frame pts against the @var{lavf.concat.start_time}
16128 and the @var{lavf.concat.duration} packet metadata values which are also
16129 present in the decoded frames.
16131 The @var{concatdec_select} variable is -1 if the frame pts is at least
16132 start_time and either the duration metadata is missing or the frame pts is less
16133 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
16136 That basically means that an input frame is selected if its pts is within the
16137 interval set by the concat demuxer.
16141 The default value of the select expression is "1".
16143 @subsection Examples
16147 Select all frames in input:
16152 The example above is the same as:
16164 Select only I-frames:
16166 select='eq(pict_type\,I)'
16170 Select one frame every 100:
16172 select='not(mod(n\,100))'
16176 Select only frames contained in the 10-20 time interval:
16178 select=between(t\,10\,20)
16182 Select only I-frames contained in the 10-20 time interval:
16184 select=between(t\,10\,20)*eq(pict_type\,I)
16188 Select frames with a minimum distance of 10 seconds:
16190 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
16194 Use aselect to select only audio frames with samples number > 100:
16196 aselect='gt(samples_n\,100)'
16200 Create a mosaic of the first scenes:
16202 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
16205 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
16209 Send even and odd frames to separate outputs, and compose them:
16211 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
16215 Select useful frames from an ffconcat file which is using inpoints and
16216 outpoints but where the source files are not intra frame only.
16218 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
16222 @section sendcmd, asendcmd
16224 Send commands to filters in the filtergraph.
16226 These filters read commands to be sent to other filters in the
16229 @code{sendcmd} must be inserted between two video filters,
16230 @code{asendcmd} must be inserted between two audio filters, but apart
16231 from that they act the same way.
16233 The specification of commands can be provided in the filter arguments
16234 with the @var{commands} option, or in a file specified by the
16235 @var{filename} option.
16237 These filters accept the following options:
16240 Set the commands to be read and sent to the other filters.
16242 Set the filename of the commands to be read and sent to the other
16246 @subsection Commands syntax
16248 A commands description consists of a sequence of interval
16249 specifications, comprising a list of commands to be executed when a
16250 particular event related to that interval occurs. The occurring event
16251 is typically the current frame time entering or leaving a given time
16254 An interval is specified by the following syntax:
16256 @var{START}[-@var{END}] @var{COMMANDS};
16259 The time interval is specified by the @var{START} and @var{END} times.
16260 @var{END} is optional and defaults to the maximum time.
16262 The current frame time is considered within the specified interval if
16263 it is included in the interval [@var{START}, @var{END}), that is when
16264 the time is greater or equal to @var{START} and is lesser than
16267 @var{COMMANDS} consists of a sequence of one or more command
16268 specifications, separated by ",", relating to that interval. The
16269 syntax of a command specification is given by:
16271 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
16274 @var{FLAGS} is optional and specifies the type of events relating to
16275 the time interval which enable sending the specified command, and must
16276 be a non-null sequence of identifier flags separated by "+" or "|" and
16277 enclosed between "[" and "]".
16279 The following flags are recognized:
16282 The command is sent when the current frame timestamp enters the
16283 specified interval. In other words, the command is sent when the
16284 previous frame timestamp was not in the given interval, and the
16288 The command is sent when the current frame timestamp leaves the
16289 specified interval. In other words, the command is sent when the
16290 previous frame timestamp was in the given interval, and the
16294 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
16297 @var{TARGET} specifies the target of the command, usually the name of
16298 the filter class or a specific filter instance name.
16300 @var{COMMAND} specifies the name of the command for the target filter.
16302 @var{ARG} is optional and specifies the optional list of argument for
16303 the given @var{COMMAND}.
16305 Between one interval specification and another, whitespaces, or
16306 sequences of characters starting with @code{#} until the end of line,
16307 are ignored and can be used to annotate comments.
16309 A simplified BNF description of the commands specification syntax
16312 @var{COMMAND_FLAG} ::= "enter" | "leave"
16313 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
16314 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
16315 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
16316 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
16317 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
16320 @subsection Examples
16324 Specify audio tempo change at second 4:
16326 asendcmd=c='4.0 atempo tempo 1.5',atempo
16330 Specify a list of drawtext and hue commands in a file.
16332 # show text in the interval 5-10
16333 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
16334 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16336 # desaturate the image in the interval 15-20
16337 15.0-20.0 [enter] hue s 0,
16338 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16340 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16342 # apply an exponential saturation fade-out effect, starting from time 25
16343 25 [enter] hue s exp(25-t)
16346 A filtergraph allowing to read and process the above command list
16347 stored in a file @file{test.cmd}, can be specified with:
16349 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16354 @section setpts, asetpts
16356 Change the PTS (presentation timestamp) of the input frames.
16358 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16360 This filter accepts the following options:
16365 The expression which is evaluated for each frame to construct its timestamp.
16369 The expression is evaluated through the eval API and can contain the following
16374 frame rate, only defined for constant frame-rate video
16377 The presentation timestamp in input
16380 The count of the input frame for video or the number of consumed samples,
16381 not including the current frame for audio, starting from 0.
16383 @item NB_CONSUMED_SAMPLES
16384 The number of consumed samples, not including the current frame (only
16387 @item NB_SAMPLES, S
16388 The number of samples in the current frame (only audio)
16390 @item SAMPLE_RATE, SR
16391 The audio sample rate.
16394 The PTS of the first frame.
16397 the time in seconds of the first frame
16400 State whether the current frame is interlaced.
16403 the time in seconds of the current frame
16406 original position in the file of the frame, or undefined if undefined
16407 for the current frame
16410 The previous input PTS.
16413 previous input time in seconds
16416 The previous output PTS.
16419 previous output time in seconds
16422 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16426 The wallclock (RTC) time at the start of the movie in microseconds.
16429 The timebase of the input timestamps.
16433 @subsection Examples
16437 Start counting PTS from zero
16439 setpts=PTS-STARTPTS
16443 Apply fast motion effect:
16449 Apply slow motion effect:
16455 Set fixed rate of 25 frames per second:
16461 Set fixed rate 25 fps with some jitter:
16463 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16467 Apply an offset of 10 seconds to the input PTS:
16473 Generate timestamps from a "live source" and rebase onto the current timebase:
16475 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16479 Generate timestamps by counting samples:
16486 @section settb, asettb
16488 Set the timebase to use for the output frames timestamps.
16489 It is mainly useful for testing timebase configuration.
16491 It accepts the following parameters:
16496 The expression which is evaluated into the output timebase.
16500 The value for @option{tb} is an arithmetic expression representing a
16501 rational. The expression can contain the constants "AVTB" (the default
16502 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16503 audio only). Default value is "intb".
16505 @subsection Examples
16509 Set the timebase to 1/25:
16515 Set the timebase to 1/10:
16521 Set the timebase to 1001/1000:
16527 Set the timebase to 2*intb:
16533 Set the default timebase value:
16540 Convert input audio to a video output representing frequency spectrum
16541 logarithmically using Brown-Puckette constant Q transform algorithm with
16542 direct frequency domain coefficient calculation (but the transform itself
16543 is not really constant Q, instead the Q factor is actually variable/clamped),
16544 with musical tone scale, from E0 to D#10.
16546 The filter accepts the following options:
16550 Specify the video size for the output. It must be even. For the syntax of this option,
16551 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16552 Default value is @code{1920x1080}.
16555 Set the output frame rate. Default value is @code{25}.
16558 Set the bargraph height. It must be even. Default value is @code{-1} which
16559 computes the bargraph height automatically.
16562 Set the axis height. It must be even. Default value is @code{-1} which computes
16563 the axis height automatically.
16566 Set the sonogram height. It must be even. Default value is @code{-1} which
16567 computes the sonogram height automatically.
16570 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16571 instead. Default value is @code{1}.
16573 @item sono_v, volume
16574 Specify the sonogram volume expression. It can contain variables:
16577 the @var{bar_v} evaluated expression
16578 @item frequency, freq, f
16579 the frequency where it is evaluated
16580 @item timeclamp, tc
16581 the value of @var{timeclamp} option
16585 @item a_weighting(f)
16586 A-weighting of equal loudness
16587 @item b_weighting(f)
16588 B-weighting of equal loudness
16589 @item c_weighting(f)
16590 C-weighting of equal loudness.
16592 Default value is @code{16}.
16594 @item bar_v, volume2
16595 Specify the bargraph volume expression. It can contain variables:
16598 the @var{sono_v} evaluated expression
16599 @item frequency, freq, f
16600 the frequency where it is evaluated
16601 @item timeclamp, tc
16602 the value of @var{timeclamp} option
16606 @item a_weighting(f)
16607 A-weighting of equal loudness
16608 @item b_weighting(f)
16609 B-weighting of equal loudness
16610 @item c_weighting(f)
16611 C-weighting of equal loudness.
16613 Default value is @code{sono_v}.
16615 @item sono_g, gamma
16616 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16617 higher gamma makes the spectrum having more range. Default value is @code{3}.
16618 Acceptable range is @code{[1, 7]}.
16620 @item bar_g, gamma2
16621 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16624 @item timeclamp, tc
16625 Specify the transform timeclamp. At low frequency, there is trade-off between
16626 accuracy in time domain and frequency domain. If timeclamp is lower,
16627 event in time domain is represented more accurately (such as fast bass drum),
16628 otherwise event in frequency domain is represented more accurately
16629 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16632 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16633 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16636 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16637 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16640 This option is deprecated and ignored.
16643 Specify the transform length in time domain. Use this option to control accuracy
16644 trade-off between time domain and frequency domain at every frequency sample.
16645 It can contain variables:
16647 @item frequency, freq, f
16648 the frequency where it is evaluated
16649 @item timeclamp, tc
16650 the value of @var{timeclamp} option.
16652 Default value is @code{384*tc/(384+tc*f)}.
16655 Specify the transform count for every video frame. Default value is @code{6}.
16656 Acceptable range is @code{[1, 30]}.
16659 Specify the transform count for every single pixel. Default value is @code{0},
16660 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16663 Specify font file for use with freetype to draw the axis. If not specified,
16664 use embedded font. Note that drawing with font file or embedded font is not
16665 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16669 Specify font color expression. This is arithmetic expression that should return
16670 integer value 0xRRGGBB. It can contain variables:
16672 @item frequency, freq, f
16673 the frequency where it is evaluated
16674 @item timeclamp, tc
16675 the value of @var{timeclamp} option
16680 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16681 @item r(x), g(x), b(x)
16682 red, green, and blue value of intensity x.
16684 Default value is @code{st(0, (midi(f)-59.5)/12);
16685 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16686 r(1-ld(1)) + b(ld(1))}.
16689 Specify image file to draw the axis. This option override @var{fontfile} and
16690 @var{fontcolor} option.
16693 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16694 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16695 Default value is @code{1}.
16699 @subsection Examples
16703 Playing audio while showing the spectrum:
16705 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16709 Same as above, but with frame rate 30 fps:
16711 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16715 Playing at 1280x720:
16717 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16721 Disable sonogram display:
16727 A1 and its harmonics: A1, A2, (near)E3, A3:
16729 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),
16730 asplit[a][out1]; [a] showcqt [out0]'
16734 Same as above, but with more accuracy in frequency domain:
16736 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),
16737 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16743 bar_v=10:sono_v=bar_v*a_weighting(f)
16747 Custom gamma, now spectrum is linear to the amplitude.
16753 Custom tlength equation:
16755 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)))'
16759 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
16761 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
16765 Custom frequency range with custom axis using image file:
16767 axisfile=myaxis.png:basefreq=40:endfreq=10000
16773 Convert input audio to video output representing the audio power spectrum.
16774 Audio amplitude is on Y-axis while frequency is on X-axis.
16776 The filter accepts the following options:
16780 Specify size of video. For the syntax of this option, check the
16781 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16782 Default is @code{1024x512}.
16786 This set how each frequency bin will be represented.
16788 It accepts the following values:
16794 Default is @code{bar}.
16797 Set amplitude scale.
16799 It accepts the following values:
16813 Default is @code{log}.
16816 Set frequency scale.
16818 It accepts the following values:
16827 Reverse logarithmic scale.
16829 Default is @code{lin}.
16834 It accepts the following values:
16850 Default is @code{w2048}
16853 Set windowing function.
16855 It accepts the following values:
16877 Default is @code{hanning}.
16880 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
16881 which means optimal overlap for selected window function will be picked.
16884 Set time averaging. Setting this to 0 will display current maximal peaks.
16885 Default is @code{1}, which means time averaging is disabled.
16888 Specify list of colors separated by space or by '|' which will be used to
16889 draw channel frequencies. Unrecognized or missing colors will be replaced
16893 Set channel display mode.
16895 It accepts the following values:
16900 Default is @code{combined}.
16903 Set minimum amplitude used in @code{log} amplitude scaler.
16907 @anchor{showspectrum}
16908 @section showspectrum
16910 Convert input audio to a video output, representing the audio frequency
16913 The filter accepts the following options:
16917 Specify the video size for the output. For the syntax of this option, check the
16918 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16919 Default value is @code{640x512}.
16922 Specify how the spectrum should slide along the window.
16924 It accepts the following values:
16927 the samples start again on the left when they reach the right
16929 the samples scroll from right to left
16931 the samples scroll from left to right
16933 frames are only produced when the samples reach the right
16936 Default value is @code{replace}.
16939 Specify display mode.
16941 It accepts the following values:
16944 all channels are displayed in the same row
16946 all channels are displayed in separate rows
16949 Default value is @samp{combined}.
16952 Specify display color mode.
16954 It accepts the following values:
16957 each channel is displayed in a separate color
16959 each channel is displayed using the same color scheme
16961 each channel is displayed using the rainbow color scheme
16963 each channel is displayed using the moreland color scheme
16965 each channel is displayed using the nebulae color scheme
16967 each channel is displayed using the fire color scheme
16969 each channel is displayed using the fiery color scheme
16971 each channel is displayed using the fruit color scheme
16973 each channel is displayed using the cool color scheme
16976 Default value is @samp{channel}.
16979 Specify scale used for calculating intensity color values.
16981 It accepts the following values:
16986 square root, default
16997 Default value is @samp{sqrt}.
17000 Set saturation modifier for displayed colors. Negative values provide
17001 alternative color scheme. @code{0} is no saturation at all.
17002 Saturation must be in [-10.0, 10.0] range.
17003 Default value is @code{1}.
17006 Set window function.
17008 It accepts the following values:
17032 Default value is @code{hann}.
17035 Set orientation of time vs frequency axis. Can be @code{vertical} or
17036 @code{horizontal}. Default is @code{vertical}.
17039 Set ratio of overlap window. Default value is @code{0}.
17040 When value is @code{1} overlap is set to recommended size for specific
17041 window function currently used.
17044 Set scale gain for calculating intensity color values.
17045 Default value is @code{1}.
17048 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
17051 Set color rotation, must be in [-1.0, 1.0] range.
17052 Default value is @code{0}.
17055 The usage is very similar to the showwaves filter; see the examples in that
17058 @subsection Examples
17062 Large window with logarithmic color scaling:
17064 showspectrum=s=1280x480:scale=log
17068 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
17070 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17071 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
17075 @section showspectrumpic
17077 Convert input audio to a single video frame, representing the audio frequency
17080 The filter accepts the following options:
17084 Specify the video size for the output. For the syntax of this option, check the
17085 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17086 Default value is @code{4096x2048}.
17089 Specify display mode.
17091 It accepts the following values:
17094 all channels are displayed in the same row
17096 all channels are displayed in separate rows
17098 Default value is @samp{combined}.
17101 Specify display color mode.
17103 It accepts the following values:
17106 each channel is displayed in a separate color
17108 each channel is displayed using the same color scheme
17110 each channel is displayed using the rainbow color scheme
17112 each channel is displayed using the moreland color scheme
17114 each channel is displayed using the nebulae color scheme
17116 each channel is displayed using the fire color scheme
17118 each channel is displayed using the fiery color scheme
17120 each channel is displayed using the fruit color scheme
17122 each channel is displayed using the cool color scheme
17124 Default value is @samp{intensity}.
17127 Specify scale used for calculating intensity color values.
17129 It accepts the following values:
17134 square root, default
17144 Default value is @samp{log}.
17147 Set saturation modifier for displayed colors. Negative values provide
17148 alternative color scheme. @code{0} is no saturation at all.
17149 Saturation must be in [-10.0, 10.0] range.
17150 Default value is @code{1}.
17153 Set window function.
17155 It accepts the following values:
17178 Default value is @code{hann}.
17181 Set orientation of time vs frequency axis. Can be @code{vertical} or
17182 @code{horizontal}. Default is @code{vertical}.
17185 Set scale gain for calculating intensity color values.
17186 Default value is @code{1}.
17189 Draw time and frequency axes and legends. Default is enabled.
17192 Set color rotation, must be in [-1.0, 1.0] range.
17193 Default value is @code{0}.
17196 @subsection Examples
17200 Extract an audio spectrogram of a whole audio track
17201 in a 1024x1024 picture using @command{ffmpeg}:
17203 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
17207 @section showvolume
17209 Convert input audio volume to a video output.
17211 The filter accepts the following options:
17218 Set border width, allowed range is [0, 5]. Default is 1.
17221 Set channel width, allowed range is [80, 8192]. Default is 400.
17224 Set channel height, allowed range is [1, 900]. Default is 20.
17227 Set fade, allowed range is [0.001, 1]. Default is 0.95.
17230 Set volume color expression.
17232 The expression can use the following variables:
17236 Current max volume of channel in dB.
17242 Current channel number, starting from 0.
17246 If set, displays channel names. Default is enabled.
17249 If set, displays volume values. Default is enabled.
17252 Set orientation, can be @code{horizontal} or @code{vertical},
17253 default is @code{horizontal}.
17256 Set step size, allowed range s [0, 5]. Default is 0, which means
17262 Convert input audio to a video output, representing the samples waves.
17264 The filter accepts the following options:
17268 Specify the video size for the output. For the syntax of this option, check the
17269 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17270 Default value is @code{600x240}.
17275 Available values are:
17278 Draw a point for each sample.
17281 Draw a vertical line for each sample.
17284 Draw a point for each sample and a line between them.
17287 Draw a centered vertical line for each sample.
17290 Default value is @code{point}.
17293 Set the number of samples which are printed on the same column. A
17294 larger value will decrease the frame rate. Must be a positive
17295 integer. This option can be set only if the value for @var{rate}
17296 is not explicitly specified.
17299 Set the (approximate) output frame rate. This is done by setting the
17300 option @var{n}. Default value is "25".
17302 @item split_channels
17303 Set if channels should be drawn separately or overlap. Default value is 0.
17306 Set colors separated by '|' which are going to be used for drawing of each channel.
17309 Set amplitude scale.
17311 Available values are:
17329 @subsection Examples
17333 Output the input file audio and the corresponding video representation
17336 amovie=a.mp3,asplit[out0],showwaves[out1]
17340 Create a synthetic signal and show it with showwaves, forcing a
17341 frame rate of 30 frames per second:
17343 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17347 @section showwavespic
17349 Convert input audio to a single video frame, representing the samples waves.
17351 The filter accepts the following options:
17355 Specify the video size for the output. For the syntax of this option, check the
17356 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17357 Default value is @code{600x240}.
17359 @item split_channels
17360 Set if channels should be drawn separately or overlap. Default value is 0.
17363 Set colors separated by '|' which are going to be used for drawing of each channel.
17366 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17370 @subsection Examples
17374 Extract a channel split representation of the wave form of a whole audio track
17375 in a 1024x800 picture using @command{ffmpeg}:
17377 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17381 @section spectrumsynth
17383 Sythesize audio from 2 input video spectrums, first input stream represents
17384 magnitude across time and second represents phase across time.
17385 The filter will transform from frequency domain as displayed in videos back
17386 to time domain as presented in audio output.
17388 This filter is primarly created for reversing processed @ref{showspectrum}
17389 filter outputs, but can synthesize sound from other spectrograms too.
17390 But in such case results are going to be poor if the phase data is not
17391 available, because in such cases phase data need to be recreated, usually
17392 its just recreated from random noise.
17393 For best results use gray only output (@code{channel} color mode in
17394 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17395 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17396 @code{data} option. Inputs videos should generally use @code{fullframe}
17397 slide mode as that saves resources needed for decoding video.
17399 The filter accepts the following options:
17403 Specify sample rate of output audio, the sample rate of audio from which
17404 spectrum was generated may differ.
17407 Set number of channels represented in input video spectrums.
17410 Set scale which was used when generating magnitude input spectrum.
17411 Can be @code{lin} or @code{log}. Default is @code{log}.
17414 Set slide which was used when generating inputs spectrums.
17415 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17416 Default is @code{fullframe}.
17419 Set window function used for resynthesis.
17422 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17423 which means optimal overlap for selected window function will be picked.
17426 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17427 Default is @code{vertical}.
17430 @subsection Examples
17434 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17435 then resynthesize videos back to audio with spectrumsynth:
17437 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
17438 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
17439 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17443 @section split, asplit
17445 Split input into several identical outputs.
17447 @code{asplit} works with audio input, @code{split} with video.
17449 The filter accepts a single parameter which specifies the number of outputs. If
17450 unspecified, it defaults to 2.
17452 @subsection Examples
17456 Create two separate outputs from the same input:
17458 [in] split [out0][out1]
17462 To create 3 or more outputs, you need to specify the number of
17465 [in] asplit=3 [out0][out1][out2]
17469 Create two separate outputs from the same input, one cropped and
17472 [in] split [splitout1][splitout2];
17473 [splitout1] crop=100:100:0:0 [cropout];
17474 [splitout2] pad=200:200:100:100 [padout];
17478 Create 5 copies of the input audio with @command{ffmpeg}:
17480 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17486 Receive commands sent through a libzmq client, and forward them to
17487 filters in the filtergraph.
17489 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17490 must be inserted between two video filters, @code{azmq} between two
17493 To enable these filters you need to install the libzmq library and
17494 headers and configure FFmpeg with @code{--enable-libzmq}.
17496 For more information about libzmq see:
17497 @url{http://www.zeromq.org/}
17499 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17500 receives messages sent through a network interface defined by the
17501 @option{bind_address} option.
17503 The received message must be in the form:
17505 @var{TARGET} @var{COMMAND} [@var{ARG}]
17508 @var{TARGET} specifies the target of the command, usually the name of
17509 the filter class or a specific filter instance name.
17511 @var{COMMAND} specifies the name of the command for the target filter.
17513 @var{ARG} is optional and specifies the optional argument list for the
17514 given @var{COMMAND}.
17516 Upon reception, the message is processed and the corresponding command
17517 is injected into the filtergraph. Depending on the result, the filter
17518 will send a reply to the client, adopting the format:
17520 @var{ERROR_CODE} @var{ERROR_REASON}
17524 @var{MESSAGE} is optional.
17526 @subsection Examples
17528 Look at @file{tools/zmqsend} for an example of a zmq client which can
17529 be used to send commands processed by these filters.
17531 Consider the following filtergraph generated by @command{ffplay}
17533 ffplay -dumpgraph 1 -f lavfi "
17534 color=s=100x100:c=red [l];
17535 color=s=100x100:c=blue [r];
17536 nullsrc=s=200x100, zmq [bg];
17537 [bg][l] overlay [bg+l];
17538 [bg+l][r] overlay=x=100 "
17541 To change the color of the left side of the video, the following
17542 command can be used:
17544 echo Parsed_color_0 c yellow | tools/zmqsend
17547 To change the right side:
17549 echo Parsed_color_1 c pink | tools/zmqsend
17552 @c man end MULTIMEDIA FILTERS
17554 @chapter Multimedia Sources
17555 @c man begin MULTIMEDIA SOURCES
17557 Below is a description of the currently available multimedia sources.
17561 This is the same as @ref{movie} source, except it selects an audio
17567 Read audio and/or video stream(s) from a movie container.
17569 It accepts the following parameters:
17573 The name of the resource to read (not necessarily a file; it can also be a
17574 device or a stream accessed through some protocol).
17576 @item format_name, f
17577 Specifies the format assumed for the movie to read, and can be either
17578 the name of a container or an input device. If not specified, the
17579 format is guessed from @var{movie_name} or by probing.
17581 @item seek_point, sp
17582 Specifies the seek point in seconds. The frames will be output
17583 starting from this seek point. The parameter is evaluated with
17584 @code{av_strtod}, so the numerical value may be suffixed by an IS
17585 postfix. The default value is "0".
17588 Specifies the streams to read. Several streams can be specified,
17589 separated by "+". The source will then have as many outputs, in the
17590 same order. The syntax is explained in the ``Stream specifiers''
17591 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17592 respectively the default (best suited) video and audio stream. Default
17593 is "dv", or "da" if the filter is called as "amovie".
17595 @item stream_index, si
17596 Specifies the index of the video stream to read. If the value is -1,
17597 the most suitable video stream will be automatically selected. The default
17598 value is "-1". Deprecated. If the filter is called "amovie", it will select
17599 audio instead of video.
17602 Specifies how many times to read the stream in sequence.
17603 If the value is less than 1, the stream will be read again and again.
17604 Default value is "1".
17606 Note that when the movie is looped the source timestamps are not
17607 changed, so it will generate non monotonically increasing timestamps.
17609 @item discontinuity
17610 Specifies the time difference between frames above which the point is
17611 considered a timestamp discontinuity which is removed by adjusting the later
17615 It allows overlaying a second video on top of the main input of
17616 a filtergraph, as shown in this graph:
17618 input -----------> deltapts0 --> overlay --> output
17621 movie --> scale--> deltapts1 -------+
17623 @subsection Examples
17627 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17628 on top of the input labelled "in":
17630 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17631 [in] setpts=PTS-STARTPTS [main];
17632 [main][over] overlay=16:16 [out]
17636 Read from a video4linux2 device, and overlay it on top of the input
17639 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17640 [in] setpts=PTS-STARTPTS [main];
17641 [main][over] overlay=16:16 [out]
17645 Read the first video stream and the audio stream with id 0x81 from
17646 dvd.vob; the video is connected to the pad named "video" and the audio is
17647 connected to the pad named "audio":
17649 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17653 @subsection Commands
17655 Both movie and amovie support the following commands:
17658 Perform seek using "av_seek_frame".
17659 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17662 @var{stream_index}: If stream_index is -1, a default
17663 stream is selected, and @var{timestamp} is automatically converted
17664 from AV_TIME_BASE units to the stream specific time_base.
17666 @var{timestamp}: Timestamp in AVStream.time_base units
17667 or, if no stream is specified, in AV_TIME_BASE units.
17669 @var{flags}: Flags which select direction and seeking mode.
17673 Get movie duration in AV_TIME_BASE units.
17677 @c man end MULTIMEDIA SOURCES