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 process_stereo
2656 Process the stereo channels together. If target_gain does not match between
2657 channels, consider it invalid and use the last valid target_gain.
2660 Always extend peaks above -3dBFS even if PE isn't signaled.
2663 Replace audio with a solid tone and adjust the amplitude to signal some
2664 specific aspect of the decoding process. The output file can be loaded in
2665 an audio editor alongside the original to aid analysis.
2667 @code{analyze_mode=pe:force_pe=1} can be used to see all samples above the PE level.
2674 Gain adjustment level at each sample
2676 Samples where peak extend occurs
2678 Samples where the code detect timer is active
2680 Samples where the target gain does not match between channels
2686 Apply a high-pass filter with 3dB point frequency.
2687 The filter can be either single-pole, or double-pole (the default).
2688 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2690 The filter accepts the following options:
2694 Set frequency in Hz. Default is 3000.
2697 Set number of poles. Default is 2.
2700 Set method to specify band-width of filter.
2713 Specify the band-width of a filter in width_type units.
2714 Applies only to double-pole filter.
2715 The default is 0.707q and gives a Butterworth response.
2720 Join multiple input streams into one multi-channel stream.
2722 It accepts the following parameters:
2726 The number of input streams. It defaults to 2.
2728 @item channel_layout
2729 The desired output channel layout. It defaults to stereo.
2732 Map channels from inputs to output. The argument is a '|'-separated list of
2733 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2734 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2735 can be either the name of the input channel (e.g. FL for front left) or its
2736 index in the specified input stream. @var{out_channel} is the name of the output
2740 The filter will attempt to guess the mappings when they are not specified
2741 explicitly. It does so by first trying to find an unused matching input channel
2742 and if that fails it picks the first unused input channel.
2744 Join 3 inputs (with properly set channel layouts):
2746 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2749 Build a 5.1 output from 6 single-channel streams:
2751 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2752 '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'
2758 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2760 To enable compilation of this filter you need to configure FFmpeg with
2761 @code{--enable-ladspa}.
2765 Specifies the name of LADSPA plugin library to load. If the environment
2766 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2767 each one of the directories specified by the colon separated list in
2768 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2769 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2770 @file{/usr/lib/ladspa/}.
2773 Specifies the plugin within the library. Some libraries contain only
2774 one plugin, but others contain many of them. If this is not set filter
2775 will list all available plugins within the specified library.
2778 Set the '|' separated list of controls which are zero or more floating point
2779 values that determine the behavior of the loaded plugin (for example delay,
2781 Controls need to be defined using the following syntax:
2782 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2783 @var{valuei} is the value set on the @var{i}-th control.
2784 Alternatively they can be also defined using the following syntax:
2785 @var{value0}|@var{value1}|@var{value2}|..., where
2786 @var{valuei} is the value set on the @var{i}-th control.
2787 If @option{controls} is set to @code{help}, all available controls and
2788 their valid ranges are printed.
2790 @item sample_rate, s
2791 Specify the sample rate, default to 44100. Only used if plugin have
2795 Set the number of samples per channel per each output frame, default
2796 is 1024. Only used if plugin have zero inputs.
2799 Set the minimum duration of the sourced audio. See
2800 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2801 for the accepted syntax.
2802 Note that the resulting duration may be greater than the specified duration,
2803 as the generated audio is always cut at the end of a complete frame.
2804 If not specified, or the expressed duration is negative, the audio is
2805 supposed to be generated forever.
2806 Only used if plugin have zero inputs.
2810 @subsection Examples
2814 List all available plugins within amp (LADSPA example plugin) library:
2820 List all available controls and their valid ranges for @code{vcf_notch}
2821 plugin from @code{VCF} library:
2823 ladspa=f=vcf:p=vcf_notch:c=help
2827 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2830 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2834 Add reverberation to the audio using TAP-plugins
2835 (Tom's Audio Processing plugins):
2837 ladspa=file=tap_reverb:tap_reverb
2841 Generate white noise, with 0.2 amplitude:
2843 ladspa=file=cmt:noise_source_white:c=c0=.2
2847 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2848 @code{C* Audio Plugin Suite} (CAPS) library:
2850 ladspa=file=caps:Click:c=c1=20'
2854 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2856 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2860 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2861 @code{SWH Plugins} collection:
2863 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2867 Attenuate low frequencies using Multiband EQ from Steve Harris
2868 @code{SWH Plugins} collection:
2870 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2874 @subsection Commands
2876 This filter supports the following commands:
2879 Modify the @var{N}-th control value.
2881 If the specified value is not valid, it is ignored and prior one is kept.
2886 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2887 Support for both single pass (livestreams, files) and double pass (files) modes.
2888 This algorithm can target IL, LRA, and maximum true peak.
2890 To enable compilation of this filter you need to configure FFmpeg with
2891 @code{--enable-libebur128}.
2893 The filter accepts the following options:
2897 Set integrated loudness target.
2898 Range is -70.0 - -5.0. Default value is -24.0.
2901 Set loudness range target.
2902 Range is 1.0 - 20.0. Default value is 7.0.
2905 Set maximum true peak.
2906 Range is -9.0 - +0.0. Default value is -2.0.
2908 @item measured_I, measured_i
2909 Measured IL of input file.
2910 Range is -99.0 - +0.0.
2912 @item measured_LRA, measured_lra
2913 Measured LRA of input file.
2914 Range is 0.0 - 99.0.
2916 @item measured_TP, measured_tp
2917 Measured true peak of input file.
2918 Range is -99.0 - +99.0.
2920 @item measured_thresh
2921 Measured threshold of input file.
2922 Range is -99.0 - +0.0.
2925 Set offset gain. Gain is applied before the true-peak limiter.
2926 Range is -99.0 - +99.0. Default is +0.0.
2929 Normalize linearly if possible.
2930 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2931 to be specified in order to use this mode.
2932 Options are true or false. Default is true.
2935 Treat mono input files as "dual-mono". If a mono file is intended for playback
2936 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2937 If set to @code{true}, this option will compensate for this effect.
2938 Multi-channel input files are not affected by this option.
2939 Options are true or false. Default is false.
2942 Set print format for stats. Options are summary, json, or none.
2943 Default value is none.
2948 Apply a low-pass filter with 3dB point frequency.
2949 The filter can be either single-pole or double-pole (the default).
2950 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2952 The filter accepts the following options:
2956 Set frequency in Hz. Default is 500.
2959 Set number of poles. Default is 2.
2962 Set method to specify band-width of filter.
2975 Specify the band-width of a filter in width_type units.
2976 Applies only to double-pole filter.
2977 The default is 0.707q and gives a Butterworth response.
2983 Mix channels with specific gain levels. The filter accepts the output
2984 channel layout followed by a set of channels definitions.
2986 This filter is also designed to efficiently remap the channels of an audio
2989 The filter accepts parameters of the form:
2990 "@var{l}|@var{outdef}|@var{outdef}|..."
2994 output channel layout or number of channels
2997 output channel specification, of the form:
2998 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
3001 output channel to define, either a channel name (FL, FR, etc.) or a channel
3002 number (c0, c1, etc.)
3005 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3008 input channel to use, see out_name for details; it is not possible to mix
3009 named and numbered input channels
3012 If the `=' in a channel specification is replaced by `<', then the gains for
3013 that specification will be renormalized so that the total is 1, thus
3014 avoiding clipping noise.
3016 @subsection Mixing examples
3018 For example, if you want to down-mix from stereo to mono, but with a bigger
3019 factor for the left channel:
3021 pan=1c|c0=0.9*c0+0.1*c1
3024 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3025 7-channels surround:
3027 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3030 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3031 that should be preferred (see "-ac" option) unless you have very specific
3034 @subsection Remapping examples
3036 The channel remapping will be effective if, and only if:
3039 @item gain coefficients are zeroes or ones,
3040 @item only one input per channel output,
3043 If all these conditions are satisfied, the filter will notify the user ("Pure
3044 channel mapping detected"), and use an optimized and lossless method to do the
3047 For example, if you have a 5.1 source and want a stereo audio stream by
3048 dropping the extra channels:
3050 pan="stereo| c0=FL | c1=FR"
3053 Given the same source, you can also switch front left and front right channels
3054 and keep the input channel layout:
3056 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3059 If the input is a stereo audio stream, you can mute the front left channel (and
3060 still keep the stereo channel layout) with:
3065 Still with a stereo audio stream input, you can copy the right channel in both
3066 front left and right:
3068 pan="stereo| c0=FR | c1=FR"
3073 ReplayGain scanner filter. This filter takes an audio stream as an input and
3074 outputs it unchanged.
3075 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3079 Convert the audio sample format, sample rate and channel layout. It is
3080 not meant to be used directly.
3083 Apply time-stretching and pitch-shifting with librubberband.
3085 The filter accepts the following options:
3089 Set tempo scale factor.
3092 Set pitch scale factor.
3095 Set transients detector.
3096 Possible values are:
3105 Possible values are:
3114 Possible values are:
3121 Set processing window size.
3122 Possible values are:
3131 Possible values are:
3138 Enable formant preservation when shift pitching.
3139 Possible values are:
3147 Possible values are:
3156 Possible values are:
3163 @section sidechaincompress
3165 This filter acts like normal compressor but has the ability to compress
3166 detected signal using second input signal.
3167 It needs two input streams and returns one output stream.
3168 First input stream will be processed depending on second stream signal.
3169 The filtered signal then can be filtered with other filters in later stages of
3170 processing. See @ref{pan} and @ref{amerge} filter.
3172 The filter accepts the following options:
3176 Set input gain. Default is 1. Range is between 0.015625 and 64.
3179 If a signal of second stream raises above this level it will affect the gain
3180 reduction of first stream.
3181 By default is 0.125. Range is between 0.00097563 and 1.
3184 Set a ratio about which the signal is reduced. 1:2 means that if the level
3185 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3186 Default is 2. Range is between 1 and 20.
3189 Amount of milliseconds the signal has to rise above the threshold before gain
3190 reduction starts. Default is 20. Range is between 0.01 and 2000.
3193 Amount of milliseconds the signal has to fall below the threshold before
3194 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3197 Set the amount by how much signal will be amplified after processing.
3198 Default is 2. Range is from 1 and 64.
3201 Curve the sharp knee around the threshold to enter gain reduction more softly.
3202 Default is 2.82843. Range is between 1 and 8.
3205 Choose if the @code{average} level between all channels of side-chain stream
3206 or the louder(@code{maximum}) channel of side-chain stream affects the
3207 reduction. Default is @code{average}.
3210 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3211 of @code{rms}. Default is @code{rms} which is mainly smoother.
3214 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3217 How much to use compressed signal in output. Default is 1.
3218 Range is between 0 and 1.
3221 @subsection Examples
3225 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3226 depending on the signal of 2nd input and later compressed signal to be
3227 merged with 2nd input:
3229 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3233 @section sidechaingate
3235 A sidechain gate acts like a normal (wideband) gate but has the ability to
3236 filter the detected signal before sending it to the gain reduction stage.
3237 Normally a gate uses the full range signal to detect a level above the
3239 For example: If you cut all lower frequencies from your sidechain signal
3240 the gate will decrease the volume of your track only if not enough highs
3241 appear. With this technique you are able to reduce the resonation of a
3242 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3244 It needs two input streams and returns one output stream.
3245 First input stream will be processed depending on second stream signal.
3247 The filter accepts the following options:
3251 Set input level before filtering.
3252 Default is 1. Allowed range is from 0.015625 to 64.
3255 Set the level of gain reduction when the signal is below the threshold.
3256 Default is 0.06125. Allowed range is from 0 to 1.
3259 If a signal rises above this level the gain reduction is released.
3260 Default is 0.125. Allowed range is from 0 to 1.
3263 Set a ratio about which the signal is reduced.
3264 Default is 2. Allowed range is from 1 to 9000.
3267 Amount of milliseconds the signal has to rise above the threshold before gain
3269 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3272 Amount of milliseconds the signal has to fall below the threshold before the
3273 reduction is increased again. Default is 250 milliseconds.
3274 Allowed range is from 0.01 to 9000.
3277 Set amount of amplification of signal after processing.
3278 Default is 1. Allowed range is from 1 to 64.
3281 Curve the sharp knee around the threshold to enter gain reduction more softly.
3282 Default is 2.828427125. Allowed range is from 1 to 8.
3285 Choose if exact signal should be taken for detection or an RMS like one.
3286 Default is rms. Can be peak or rms.
3289 Choose if the average level between all channels or the louder channel affects
3291 Default is average. Can be average or maximum.
3294 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3297 @section silencedetect
3299 Detect silence in an audio stream.
3301 This filter logs a message when it detects that the input audio volume is less
3302 or equal to a noise tolerance value for a duration greater or equal to the
3303 minimum detected noise duration.
3305 The printed times and duration are expressed in seconds.
3307 The filter accepts the following options:
3311 Set silence duration until notification (default is 2 seconds).
3314 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3315 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3318 @subsection Examples
3322 Detect 5 seconds of silence with -50dB noise tolerance:
3324 silencedetect=n=-50dB:d=5
3328 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3329 tolerance in @file{silence.mp3}:
3331 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3335 @section silenceremove
3337 Remove silence from the beginning, middle or end of the audio.
3339 The filter accepts the following options:
3343 This value is used to indicate if audio should be trimmed at beginning of
3344 the audio. A value of zero indicates no silence should be trimmed from the
3345 beginning. When specifying a non-zero value, it trims audio up until it
3346 finds non-silence. Normally, when trimming silence from beginning of audio
3347 the @var{start_periods} will be @code{1} but it can be increased to higher
3348 values to trim all audio up to specific count of non-silence periods.
3349 Default value is @code{0}.
3351 @item start_duration
3352 Specify the amount of time that non-silence must be detected before it stops
3353 trimming audio. By increasing the duration, bursts of noises can be treated
3354 as silence and trimmed off. Default value is @code{0}.
3356 @item start_threshold
3357 This indicates what sample value should be treated as silence. For digital
3358 audio, a value of @code{0} may be fine but for audio recorded from analog,
3359 you may wish to increase the value to account for background noise.
3360 Can be specified in dB (in case "dB" is appended to the specified value)
3361 or amplitude ratio. Default value is @code{0}.
3364 Set the count for trimming silence from the end of audio.
3365 To remove silence from the middle of a file, specify a @var{stop_periods}
3366 that is negative. This value is then treated as a positive value and is
3367 used to indicate the effect should restart processing as specified by
3368 @var{start_periods}, making it suitable for removing periods of silence
3369 in the middle of the audio.
3370 Default value is @code{0}.
3373 Specify a duration of silence that must exist before audio is not copied any
3374 more. By specifying a higher duration, silence that is wanted can be left in
3376 Default value is @code{0}.
3378 @item stop_threshold
3379 This is the same as @option{start_threshold} but for trimming silence from
3381 Can be specified in dB (in case "dB" is appended to the specified value)
3382 or amplitude ratio. Default value is @code{0}.
3385 This indicate that @var{stop_duration} length of audio should be left intact
3386 at the beginning of each period of silence.
3387 For example, if you want to remove long pauses between words but do not want
3388 to remove the pauses completely. Default value is @code{0}.
3391 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3392 and works better with digital silence which is exactly 0.
3393 Default value is @code{rms}.
3396 Set ratio used to calculate size of window for detecting silence.
3397 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3400 @subsection Examples
3404 The following example shows how this filter can be used to start a recording
3405 that does not contain the delay at the start which usually occurs between
3406 pressing the record button and the start of the performance:
3408 silenceremove=1:5:0.02
3412 Trim all silence encountered from beginning to end where there is more than 1
3413 second of silence in audio:
3415 silenceremove=0:0:0:-1:1:-90dB
3421 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3422 loudspeakers around the user for binaural listening via headphones (audio
3423 formats up to 9 channels supported).
3424 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3425 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3426 Austrian Academy of Sciences.
3428 To enable compilation of this filter you need to configure FFmpeg with
3429 @code{--enable-netcdf}.
3431 The filter accepts the following options:
3435 Set the SOFA file used for rendering.
3438 Set gain applied to audio. Value is in dB. Default is 0.
3441 Set rotation of virtual loudspeakers in deg. Default is 0.
3444 Set elevation of virtual speakers in deg. Default is 0.
3447 Set distance in meters between loudspeakers and the listener with near-field
3448 HRTFs. Default is 1.
3451 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3452 processing audio in time domain which is slow.
3453 @var{freq} is processing audio in frequency domain which is fast.
3454 Default is @var{freq}.
3457 Set custom positions of virtual loudspeakers. Syntax for this option is:
3458 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3459 Each virtual loudspeaker is described with short channel name following with
3460 azimuth and elevation in degreees.
3461 Each virtual loudspeaker description is separated by '|'.
3462 For example to override front left and front right channel positions use:
3463 'speakers=FL 45 15|FR 345 15'.
3464 Descriptions with unrecognised channel names are ignored.
3467 @subsection Examples
3471 Using ClubFritz6 sofa file:
3473 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3477 Using ClubFritz12 sofa file and bigger radius with small rotation:
3479 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3483 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3484 and also with custom gain:
3486 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3490 @section stereotools
3492 This filter has some handy utilities to manage stereo signals, for converting
3493 M/S stereo recordings to L/R signal while having control over the parameters
3494 or spreading the stereo image of master track.
3496 The filter accepts the following options:
3500 Set input level before filtering for both channels. Defaults is 1.
3501 Allowed range is from 0.015625 to 64.
3504 Set output level after filtering for both channels. Defaults is 1.
3505 Allowed range is from 0.015625 to 64.
3508 Set input balance between both channels. Default is 0.
3509 Allowed range is from -1 to 1.
3512 Set output balance between both channels. Default is 0.
3513 Allowed range is from -1 to 1.
3516 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3517 clipping. Disabled by default.
3520 Mute the left channel. Disabled by default.
3523 Mute the right channel. Disabled by default.
3526 Change the phase of the left channel. Disabled by default.
3529 Change the phase of the right channel. Disabled by default.
3532 Set stereo mode. Available values are:
3536 Left/Right to Left/Right, this is default.
3539 Left/Right to Mid/Side.
3542 Mid/Side to Left/Right.
3545 Left/Right to Left/Left.
3548 Left/Right to Right/Right.
3551 Left/Right to Left + Right.
3554 Left/Right to Right/Left.
3558 Set level of side signal. Default is 1.
3559 Allowed range is from 0.015625 to 64.
3562 Set balance of side signal. Default is 0.
3563 Allowed range is from -1 to 1.
3566 Set level of the middle signal. Default is 1.
3567 Allowed range is from 0.015625 to 64.
3570 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3573 Set stereo base between mono and inversed channels. Default is 0.
3574 Allowed range is from -1 to 1.
3577 Set delay in milliseconds how much to delay left from right channel and
3578 vice versa. Default is 0. Allowed range is from -20 to 20.
3581 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3584 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3587 @subsection Examples
3591 Apply karaoke like effect:
3593 stereotools=mlev=0.015625
3597 Convert M/S signal to L/R:
3599 "stereotools=mode=ms>lr"
3603 @section stereowiden
3605 This filter enhance the stereo effect by suppressing signal common to both
3606 channels and by delaying the signal of left into right and vice versa,
3607 thereby widening the stereo effect.
3609 The filter accepts the following options:
3613 Time in milliseconds of the delay of left signal into right and vice versa.
3614 Default is 20 milliseconds.
3617 Amount of gain in delayed signal into right and vice versa. Gives a delay
3618 effect of left signal in right output and vice versa which gives widening
3619 effect. Default is 0.3.
3622 Cross feed of left into right with inverted phase. This helps in suppressing
3623 the mono. If the value is 1 it will cancel all the signal common to both
3624 channels. Default is 0.3.
3627 Set level of input signal of original channel. Default is 0.8.
3632 Boost or cut treble (upper) frequencies of the audio using a two-pole
3633 shelving filter with a response similar to that of a standard
3634 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3636 The filter accepts the following options:
3640 Give the gain at whichever is the lower of ~22 kHz and the
3641 Nyquist frequency. Its useful range is about -20 (for a large cut)
3642 to +20 (for a large boost). Beware of clipping when using a positive gain.
3645 Set the filter's central frequency and so can be used
3646 to extend or reduce the frequency range to be boosted or cut.
3647 The default value is @code{3000} Hz.
3650 Set method to specify band-width of filter.
3663 Determine how steep is the filter's shelf transition.
3668 Sinusoidal amplitude modulation.
3670 The filter accepts the following options:
3674 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3675 (20 Hz or lower) will result in a tremolo effect.
3676 This filter may also be used as a ring modulator by specifying
3677 a modulation frequency higher than 20 Hz.
3678 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3681 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3682 Default value is 0.5.
3687 Sinusoidal phase modulation.
3689 The filter accepts the following options:
3693 Modulation frequency in Hertz.
3694 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3697 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3698 Default value is 0.5.
3703 Adjust the input audio volume.
3705 It accepts the following parameters:
3709 Set audio volume expression.
3711 Output values are clipped to the maximum value.
3713 The output audio volume is given by the relation:
3715 @var{output_volume} = @var{volume} * @var{input_volume}
3718 The default value for @var{volume} is "1.0".
3721 This parameter represents the mathematical precision.
3723 It determines which input sample formats will be allowed, which affects the
3724 precision of the volume scaling.
3728 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3730 32-bit floating-point; this limits input sample format to FLT. (default)
3732 64-bit floating-point; this limits input sample format to DBL.
3736 Choose the behaviour on encountering ReplayGain side data in input frames.
3740 Remove ReplayGain side data, ignoring its contents (the default).
3743 Ignore ReplayGain side data, but leave it in the frame.
3746 Prefer the track gain, if present.
3749 Prefer the album gain, if present.
3752 @item replaygain_preamp
3753 Pre-amplification gain in dB to apply to the selected replaygain gain.
3755 Default value for @var{replaygain_preamp} is 0.0.
3758 Set when the volume expression is evaluated.
3760 It accepts the following values:
3763 only evaluate expression once during the filter initialization, or
3764 when the @samp{volume} command is sent
3767 evaluate expression for each incoming frame
3770 Default value is @samp{once}.
3773 The volume expression can contain the following parameters.
3777 frame number (starting at zero)
3780 @item nb_consumed_samples
3781 number of samples consumed by the filter
3783 number of samples in the current frame
3785 original frame position in the file
3791 PTS at start of stream
3793 time at start of stream
3799 last set volume value
3802 Note that when @option{eval} is set to @samp{once} only the
3803 @var{sample_rate} and @var{tb} variables are available, all other
3804 variables will evaluate to NAN.
3806 @subsection Commands
3808 This filter supports the following commands:
3811 Modify the volume expression.
3812 The command accepts the same syntax of the corresponding option.
3814 If the specified expression is not valid, it is kept at its current
3816 @item replaygain_noclip
3817 Prevent clipping by limiting the gain applied.
3819 Default value for @var{replaygain_noclip} is 1.
3823 @subsection Examples
3827 Halve the input audio volume:
3831 volume=volume=-6.0206dB
3834 In all the above example the named key for @option{volume} can be
3835 omitted, for example like in:
3841 Increase input audio power by 6 decibels using fixed-point precision:
3843 volume=volume=6dB:precision=fixed
3847 Fade volume after time 10 with an annihilation period of 5 seconds:
3849 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3853 @section volumedetect
3855 Detect the volume of the input video.
3857 The filter has no parameters. The input is not modified. Statistics about
3858 the volume will be printed in the log when the input stream end is reached.
3860 In particular it will show the mean volume (root mean square), maximum
3861 volume (on a per-sample basis), and the beginning of a histogram of the
3862 registered volume values (from the maximum value to a cumulated 1/1000 of
3865 All volumes are in decibels relative to the maximum PCM value.
3867 @subsection Examples
3869 Here is an excerpt of the output:
3871 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3872 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3873 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3874 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3875 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3876 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3877 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3878 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3879 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3885 The mean square energy is approximately -27 dB, or 10^-2.7.
3887 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3889 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3892 In other words, raising the volume by +4 dB does not cause any clipping,
3893 raising it by +5 dB causes clipping for 6 samples, etc.
3895 @c man end AUDIO FILTERS
3897 @chapter Audio Sources
3898 @c man begin AUDIO SOURCES
3900 Below is a description of the currently available audio sources.
3904 Buffer audio frames, and make them available to the filter chain.
3906 This source is mainly intended for a programmatic use, in particular
3907 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3909 It accepts the following parameters:
3913 The timebase which will be used for timestamps of submitted frames. It must be
3914 either a floating-point number or in @var{numerator}/@var{denominator} form.
3917 The sample rate of the incoming audio buffers.
3920 The sample format of the incoming audio buffers.
3921 Either a sample format name or its corresponding integer representation from
3922 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3924 @item channel_layout
3925 The channel layout of the incoming audio buffers.
3926 Either a channel layout name from channel_layout_map in
3927 @file{libavutil/channel_layout.c} or its corresponding integer representation
3928 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3931 The number of channels of the incoming audio buffers.
3932 If both @var{channels} and @var{channel_layout} are specified, then they
3937 @subsection Examples
3940 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3943 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3944 Since the sample format with name "s16p" corresponds to the number
3945 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3948 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3953 Generate an audio signal specified by an expression.
3955 This source accepts in input one or more expressions (one for each
3956 channel), which are evaluated and used to generate a corresponding
3959 This source accepts the following options:
3963 Set the '|'-separated expressions list for each separate channel. In case the
3964 @option{channel_layout} option is not specified, the selected channel layout
3965 depends on the number of provided expressions. Otherwise the last
3966 specified expression is applied to the remaining output channels.
3968 @item channel_layout, c
3969 Set the channel layout. The number of channels in the specified layout
3970 must be equal to the number of specified expressions.
3973 Set the minimum duration of the sourced audio. See
3974 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3975 for the accepted syntax.
3976 Note that the resulting duration may be greater than the specified
3977 duration, as the generated audio is always cut at the end of a
3980 If not specified, or the expressed duration is negative, the audio is
3981 supposed to be generated forever.
3984 Set the number of samples per channel per each output frame,
3987 @item sample_rate, s
3988 Specify the sample rate, default to 44100.
3991 Each expression in @var{exprs} can contain the following constants:
3995 number of the evaluated sample, starting from 0
3998 time of the evaluated sample expressed in seconds, starting from 0
4005 @subsection Examples
4015 Generate a sin signal with frequency of 440 Hz, set sample rate to
4018 aevalsrc="sin(440*2*PI*t):s=8000"
4022 Generate a two channels signal, specify the channel layout (Front
4023 Center + Back Center) explicitly:
4025 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4029 Generate white noise:
4031 aevalsrc="-2+random(0)"
4035 Generate an amplitude modulated signal:
4037 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4041 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4043 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4050 The null audio source, return unprocessed audio frames. It is mainly useful
4051 as a template and to be employed in analysis / debugging tools, or as
4052 the source for filters which ignore the input data (for example the sox
4055 This source accepts the following options:
4059 @item channel_layout, cl
4061 Specifies the channel layout, and can be either an integer or a string
4062 representing a channel layout. The default value of @var{channel_layout}
4065 Check the channel_layout_map definition in
4066 @file{libavutil/channel_layout.c} for the mapping between strings and
4067 channel layout values.
4069 @item sample_rate, r
4070 Specifies the sample rate, and defaults to 44100.
4073 Set the number of samples per requested frames.
4077 @subsection Examples
4081 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4083 anullsrc=r=48000:cl=4
4087 Do the same operation with a more obvious syntax:
4089 anullsrc=r=48000:cl=mono
4093 All the parameters need to be explicitly defined.
4097 Synthesize a voice utterance using the libflite library.
4099 To enable compilation of this filter you need to configure FFmpeg with
4100 @code{--enable-libflite}.
4102 Note that the flite library is not thread-safe.
4104 The filter accepts the following options:
4109 If set to 1, list the names of the available voices and exit
4110 immediately. Default value is 0.
4113 Set the maximum number of samples per frame. Default value is 512.
4116 Set the filename containing the text to speak.
4119 Set the text to speak.
4122 Set the voice to use for the speech synthesis. Default value is
4123 @code{kal}. See also the @var{list_voices} option.
4126 @subsection Examples
4130 Read from file @file{speech.txt}, and synthesize the text using the
4131 standard flite voice:
4133 flite=textfile=speech.txt
4137 Read the specified text selecting the @code{slt} voice:
4139 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4143 Input text to ffmpeg:
4145 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4149 Make @file{ffplay} speak the specified text, using @code{flite} and
4150 the @code{lavfi} device:
4152 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4156 For more information about libflite, check:
4157 @url{http://www.speech.cs.cmu.edu/flite/}
4161 Generate a noise audio signal.
4163 The filter accepts the following options:
4166 @item sample_rate, r
4167 Specify the sample rate. Default value is 48000 Hz.
4170 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4174 Specify the duration of the generated audio stream. Not specifying this option
4175 results in noise with an infinite length.
4177 @item color, colour, c
4178 Specify the color of noise. Available noise colors are white, pink, and brown.
4179 Default color is white.
4182 Specify a value used to seed the PRNG.
4185 Set the number of samples per each output frame, default is 1024.
4188 @subsection Examples
4193 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4195 anoisesrc=d=60:c=pink:r=44100:a=0.5
4201 Generate an audio signal made of a sine wave with amplitude 1/8.
4203 The audio signal is bit-exact.
4205 The filter accepts the following options:
4210 Set the carrier frequency. Default is 440 Hz.
4212 @item beep_factor, b
4213 Enable a periodic beep every second with frequency @var{beep_factor} times
4214 the carrier frequency. Default is 0, meaning the beep is disabled.
4216 @item sample_rate, r
4217 Specify the sample rate, default is 44100.
4220 Specify the duration of the generated audio stream.
4222 @item samples_per_frame
4223 Set the number of samples per output frame.
4225 The expression can contain the following constants:
4229 The (sequential) number of the output audio frame, starting from 0.
4232 The PTS (Presentation TimeStamp) of the output audio frame,
4233 expressed in @var{TB} units.
4236 The PTS of the output audio frame, expressed in seconds.
4239 The timebase of the output audio frames.
4242 Default is @code{1024}.
4245 @subsection Examples
4250 Generate a simple 440 Hz sine wave:
4256 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4260 sine=frequency=220:beep_factor=4:duration=5
4264 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4267 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4271 @c man end AUDIO SOURCES
4273 @chapter Audio Sinks
4274 @c man begin AUDIO SINKS
4276 Below is a description of the currently available audio sinks.
4278 @section abuffersink
4280 Buffer audio frames, and make them available to the end of filter chain.
4282 This sink is mainly intended for programmatic use, in particular
4283 through the interface defined in @file{libavfilter/buffersink.h}
4284 or the options system.
4286 It accepts a pointer to an AVABufferSinkContext structure, which
4287 defines the incoming buffers' formats, to be passed as the opaque
4288 parameter to @code{avfilter_init_filter} for initialization.
4291 Null audio sink; do absolutely nothing with the input audio. It is
4292 mainly useful as a template and for use in analysis / debugging
4295 @c man end AUDIO SINKS
4297 @chapter Video Filters
4298 @c man begin VIDEO FILTERS
4300 When you configure your FFmpeg build, you can disable any of the
4301 existing filters using @code{--disable-filters}.
4302 The configure output will show the video filters included in your
4305 Below is a description of the currently available video filters.
4307 @section alphaextract
4309 Extract the alpha component from the input as a grayscale video. This
4310 is especially useful with the @var{alphamerge} filter.
4314 Add or replace the alpha component of the primary input with the
4315 grayscale value of a second input. This is intended for use with
4316 @var{alphaextract} to allow the transmission or storage of frame
4317 sequences that have alpha in a format that doesn't support an alpha
4320 For example, to reconstruct full frames from a normal YUV-encoded video
4321 and a separate video created with @var{alphaextract}, you might use:
4323 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4326 Since this filter is designed for reconstruction, it operates on frame
4327 sequences without considering timestamps, and terminates when either
4328 input reaches end of stream. This will cause problems if your encoding
4329 pipeline drops frames. If you're trying to apply an image as an
4330 overlay to a video stream, consider the @var{overlay} filter instead.
4334 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4335 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4336 Substation Alpha) subtitles files.
4338 This filter accepts the following option in addition to the common options from
4339 the @ref{subtitles} filter:
4343 Set the shaping engine
4345 Available values are:
4348 The default libass shaping engine, which is the best available.
4350 Fast, font-agnostic shaper that can do only substitutions
4352 Slower shaper using OpenType for substitutions and positioning
4355 The default is @code{auto}.
4359 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4361 The filter accepts the following options:
4365 Set threshold A for 1st plane. Default is 0.02.
4366 Valid range is 0 to 0.3.
4369 Set threshold B for 1st plane. Default is 0.04.
4370 Valid range is 0 to 5.
4373 Set threshold A for 2nd plane. Default is 0.02.
4374 Valid range is 0 to 0.3.
4377 Set threshold B for 2nd plane. Default is 0.04.
4378 Valid range is 0 to 5.
4381 Set threshold A for 3rd plane. Default is 0.02.
4382 Valid range is 0 to 0.3.
4385 Set threshold B for 3rd plane. Default is 0.04.
4386 Valid range is 0 to 5.
4388 Threshold A is designed to react on abrupt changes in the input signal and
4389 threshold B is designed to react on continuous changes in the input signal.
4392 Set number of frames filter will use for averaging. Default is 33. Must be odd
4393 number in range [5, 129].
4398 Compute the bounding box for the non-black pixels in the input frame
4401 This filter computes the bounding box containing all the pixels with a
4402 luminance value greater than the minimum allowed value.
4403 The parameters describing the bounding box are printed on the filter
4406 The filter accepts the following option:
4410 Set the minimal luminance value. Default is @code{16}.
4413 @section bitplanenoise
4415 Show and measure bit plane noise.
4417 The filter accepts the following options:
4421 Set which plane to analyze. Default is @code{1}.
4424 Filter out noisy pixels from @code{bitplane} set above.
4425 Default is disabled.
4428 @section blackdetect
4430 Detect video intervals that are (almost) completely black. Can be
4431 useful to detect chapter transitions, commercials, or invalid
4432 recordings. Output lines contains the time for the start, end and
4433 duration of the detected black interval expressed in seconds.
4435 In order to display the output lines, you need to set the loglevel at
4436 least to the AV_LOG_INFO value.
4438 The filter accepts the following options:
4441 @item black_min_duration, d
4442 Set the minimum detected black duration expressed in seconds. It must
4443 be a non-negative floating point number.
4445 Default value is 2.0.
4447 @item picture_black_ratio_th, pic_th
4448 Set the threshold for considering a picture "black".
4449 Express the minimum value for the ratio:
4451 @var{nb_black_pixels} / @var{nb_pixels}
4454 for which a picture is considered black.
4455 Default value is 0.98.
4457 @item pixel_black_th, pix_th
4458 Set the threshold for considering a pixel "black".
4460 The threshold expresses the maximum pixel luminance value for which a
4461 pixel is considered "black". The provided value is scaled according to
4462 the following equation:
4464 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4467 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4468 the input video format, the range is [0-255] for YUV full-range
4469 formats and [16-235] for YUV non full-range formats.
4471 Default value is 0.10.
4474 The following example sets the maximum pixel threshold to the minimum
4475 value, and detects only black intervals of 2 or more seconds:
4477 blackdetect=d=2:pix_th=0.00
4482 Detect frames that are (almost) completely black. Can be useful to
4483 detect chapter transitions or commercials. Output lines consist of
4484 the frame number of the detected frame, the percentage of blackness,
4485 the position in the file if known or -1 and the timestamp in seconds.
4487 In order to display the output lines, you need to set the loglevel at
4488 least to the AV_LOG_INFO value.
4490 It accepts the following parameters:
4495 The percentage of the pixels that have to be below the threshold; it defaults to
4498 @item threshold, thresh
4499 The threshold below which a pixel value is considered black; it defaults to
4504 @section blend, tblend
4506 Blend two video frames into each other.
4508 The @code{blend} filter takes two input streams and outputs one
4509 stream, the first input is the "top" layer and second input is
4510 "bottom" layer. Output terminates when shortest input terminates.
4512 The @code{tblend} (time blend) filter takes two consecutive frames
4513 from one single stream, and outputs the result obtained by blending
4514 the new frame on top of the old frame.
4516 A description of the accepted options follows.
4524 Set blend mode for specific pixel component or all pixel components in case
4525 of @var{all_mode}. Default value is @code{normal}.
4527 Available values for component modes are:
4568 Set blend opacity for specific pixel component or all pixel components in case
4569 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4576 Set blend expression for specific pixel component or all pixel components in case
4577 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4579 The expressions can use the following variables:
4583 The sequential number of the filtered frame, starting from @code{0}.
4587 the coordinates of the current sample
4591 the width and height of currently filtered plane
4595 Width and height scale depending on the currently filtered plane. It is the
4596 ratio between the corresponding luma plane number of pixels and the current
4597 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4598 @code{0.5,0.5} for chroma planes.
4601 Time of the current frame, expressed in seconds.
4604 Value of pixel component at current location for first video frame (top layer).
4607 Value of pixel component at current location for second video frame (bottom layer).
4611 Force termination when the shortest input terminates. Default is
4612 @code{0}. This option is only defined for the @code{blend} filter.
4615 Continue applying the last bottom frame after the end of the stream. A value of
4616 @code{0} disable the filter after the last frame of the bottom layer is reached.
4617 Default is @code{1}. This option is only defined for the @code{blend} filter.
4620 @subsection Examples
4624 Apply transition from bottom layer to top layer in first 10 seconds:
4626 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4630 Apply 1x1 checkerboard effect:
4632 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4636 Apply uncover left effect:
4638 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4642 Apply uncover down effect:
4644 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4648 Apply uncover up-left effect:
4650 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4654 Split diagonally video and shows top and bottom layer on each side:
4656 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4660 Display differences between the current and the previous frame:
4662 tblend=all_mode=difference128
4668 Apply a boxblur algorithm to the input video.
4670 It accepts the following parameters:
4674 @item luma_radius, lr
4675 @item luma_power, lp
4676 @item chroma_radius, cr
4677 @item chroma_power, cp
4678 @item alpha_radius, ar
4679 @item alpha_power, ap
4683 A description of the accepted options follows.
4686 @item luma_radius, lr
4687 @item chroma_radius, cr
4688 @item alpha_radius, ar
4689 Set an expression for the box radius in pixels used for blurring the
4690 corresponding input plane.
4692 The radius value must be a non-negative number, and must not be
4693 greater than the value of the expression @code{min(w,h)/2} for the
4694 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4697 Default value for @option{luma_radius} is "2". If not specified,
4698 @option{chroma_radius} and @option{alpha_radius} default to the
4699 corresponding value set for @option{luma_radius}.
4701 The expressions can contain the following constants:
4705 The input width and height in pixels.
4709 The input chroma image width and height in pixels.
4713 The horizontal and vertical chroma subsample values. For example, for the
4714 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4717 @item luma_power, lp
4718 @item chroma_power, cp
4719 @item alpha_power, ap
4720 Specify how many times the boxblur filter is applied to the
4721 corresponding plane.
4723 Default value for @option{luma_power} is 2. If not specified,
4724 @option{chroma_power} and @option{alpha_power} default to the
4725 corresponding value set for @option{luma_power}.
4727 A value of 0 will disable the effect.
4730 @subsection Examples
4734 Apply a boxblur filter with the luma, chroma, and alpha radii
4737 boxblur=luma_radius=2:luma_power=1
4742 Set the luma radius to 2, and alpha and chroma radius to 0:
4744 boxblur=2:1:cr=0:ar=0
4748 Set the luma and chroma radii to a fraction of the video dimension:
4750 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4756 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4757 Deinterlacing Filter").
4759 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4760 interpolation algorithms.
4761 It accepts the following parameters:
4765 The interlacing mode to adopt. It accepts one of the following values:
4769 Output one frame for each frame.
4771 Output one frame for each field.
4774 The default value is @code{send_field}.
4777 The picture field parity assumed for the input interlaced video. It accepts one
4778 of the following values:
4782 Assume the top field is first.
4784 Assume the bottom field is first.
4786 Enable automatic detection of field parity.
4789 The default value is @code{auto}.
4790 If the interlacing is unknown or the decoder does not export this information,
4791 top field first will be assumed.
4794 Specify which frames to deinterlace. Accept one of the following
4799 Deinterlace all frames.
4801 Only deinterlace frames marked as interlaced.
4804 The default value is @code{all}.
4808 YUV colorspace color/chroma keying.
4810 The filter accepts the following options:
4814 The color which will be replaced with transparency.
4817 Similarity percentage with the key color.
4819 0.01 matches only the exact key color, while 1.0 matches everything.
4824 0.0 makes pixels either fully transparent, or not transparent at all.
4826 Higher values result in semi-transparent pixels, with a higher transparency
4827 the more similar the pixels color is to the key color.
4830 Signals that the color passed is already in YUV instead of RGB.
4832 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4833 This can be used to pass exact YUV values as hexadecimal numbers.
4836 @subsection Examples
4840 Make every green pixel in the input image transparent:
4842 ffmpeg -i input.png -vf chromakey=green out.png
4846 Overlay a greenscreen-video on top of a static black background.
4848 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
4854 Display CIE color diagram with pixels overlaid onto it.
4856 The filter accepts the following options:
4871 @item uhdtv, rec2020
4884 Set what gamuts to draw.
4886 See @code{system} option for available values.
4889 Set ciescope size, by default set to 512.
4892 Set intensity used to map input pixel values to CIE diagram.
4895 Set contrast used to draw tongue colors that are out of active color system gamut.
4898 Correct gamma displayed on scope, by default enabled.
4901 Show white point on CIE diagram, by default disabled.
4904 Set input gamma. Used only with XYZ input color space.
4909 Visualize information exported by some codecs.
4911 Some codecs can export information through frames using side-data or other
4912 means. For example, some MPEG based codecs export motion vectors through the
4913 @var{export_mvs} flag in the codec @option{flags2} option.
4915 The filter accepts the following option:
4919 Set motion vectors to visualize.
4921 Available flags for @var{mv} are:
4925 forward predicted MVs of P-frames
4927 forward predicted MVs of B-frames
4929 backward predicted MVs of B-frames
4933 Display quantization parameters using the chroma planes.
4936 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4938 Available flags for @var{mv_type} are:
4942 forward predicted MVs
4944 backward predicted MVs
4947 @item frame_type, ft
4948 Set frame type to visualize motion vectors of.
4950 Available flags for @var{frame_type} are:
4954 intra-coded frames (I-frames)
4956 predicted frames (P-frames)
4958 bi-directionally predicted frames (B-frames)
4962 @subsection Examples
4966 Visualize forward predicted MVs of all frames using @command{ffplay}:
4968 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
4972 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
4974 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
4978 @section colorbalance
4979 Modify intensity of primary colors (red, green and blue) of input frames.
4981 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
4982 regions for the red-cyan, green-magenta or blue-yellow balance.
4984 A positive adjustment value shifts the balance towards the primary color, a negative
4985 value towards the complementary color.
4987 The filter accepts the following options:
4993 Adjust red, green and blue shadows (darkest pixels).
4998 Adjust red, green and blue midtones (medium pixels).
5003 Adjust red, green and blue highlights (brightest pixels).
5005 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5008 @subsection Examples
5012 Add red color cast to shadows:
5019 RGB colorspace color keying.
5021 The filter accepts the following options:
5025 The color which will be replaced with transparency.
5028 Similarity percentage with the key color.
5030 0.01 matches only the exact key color, while 1.0 matches everything.
5035 0.0 makes pixels either fully transparent, or not transparent at all.
5037 Higher values result in semi-transparent pixels, with a higher transparency
5038 the more similar the pixels color is to the key color.
5041 @subsection Examples
5045 Make every green pixel in the input image transparent:
5047 ffmpeg -i input.png -vf colorkey=green out.png
5051 Overlay a greenscreen-video on top of a static background image.
5053 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
5057 @section colorlevels
5059 Adjust video input frames using levels.
5061 The filter accepts the following options:
5068 Adjust red, green, blue and alpha input black point.
5069 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5075 Adjust red, green, blue and alpha input white point.
5076 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5078 Input levels are used to lighten highlights (bright tones), darken shadows
5079 (dark tones), change the balance of bright and dark tones.
5085 Adjust red, green, blue and alpha output black point.
5086 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5092 Adjust red, green, blue and alpha output white point.
5093 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5095 Output levels allows manual selection of a constrained output level range.
5098 @subsection Examples
5102 Make video output darker:
5104 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5110 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5114 Make video output lighter:
5116 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5120 Increase brightness:
5122 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5126 @section colorchannelmixer
5128 Adjust video input frames by re-mixing color channels.
5130 This filter modifies a color channel by adding the values associated to
5131 the other channels of the same pixels. For example if the value to
5132 modify is red, the output value will be:
5134 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5137 The filter accepts the following options:
5144 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5145 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5151 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5152 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5158 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5159 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5165 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5166 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5168 Allowed ranges for options are @code{[-2.0, 2.0]}.
5171 @subsection Examples
5175 Convert source to grayscale:
5177 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5180 Simulate sepia tones:
5182 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5186 @section colormatrix
5188 Convert color matrix.
5190 The filter accepts the following options:
5195 Specify the source and destination color matrix. Both values must be
5198 The accepted values are:
5217 For example to convert from BT.601 to SMPTE-240M, use the command:
5219 colormatrix=bt601:smpte240m
5224 Convert colorspace, transfer characteristics or color primaries.
5226 The filter accepts the following options:
5230 Specify all color properties at once.
5232 The accepted values are:
5261 Specify output colorspace.
5263 The accepted values are:
5272 BT.470BG or BT.601-6 625
5275 SMPTE-170M or BT.601-6 525
5281 BT.2020 with non-constant luminance
5286 Specify output transfer characteristics.
5288 The accepted values are:
5294 Constant gamma of 2.2
5297 Constant gamma of 2.8
5300 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5306 BT.2020 for 10-bits content
5309 BT.2020 for 12-bits content
5314 Specify output color primaries.
5316 The accepted values are:
5325 BT.470BG or BT.601-6 625
5328 SMPTE-170M or BT.601-6 525
5339 Specify output color range.
5341 The accepted values are:
5344 MPEG (restricted) range
5352 Specify output color format.
5354 The accepted values are:
5357 YUV 4:2:0 planar 8-bits
5360 YUV 4:2:0 planar 10-bits
5363 YUV 4:2:0 planar 12-bits
5366 YUV 4:2:2 planar 8-bits
5369 YUV 4:2:2 planar 10-bits
5372 YUV 4:2:2 planar 12-bits
5375 YUV 4:4:4 planar 8-bits
5378 YUV 4:4:4 planar 10-bits
5381 YUV 4:4:4 planar 12-bits
5386 Do a fast conversion, which skips gamma/primary correction. This will take
5387 significantly less CPU, but will be mathematically incorrect. To get output
5388 compatible with that produced by the colormatrix filter, use fast=1.
5391 Specify dithering mode.
5393 The accepted values are:
5399 Floyd-Steinberg dithering
5403 Whitepoint adaptation mode.
5405 The accepted values are:
5408 Bradford whitepoint adaptation
5411 von Kries whitepoint adaptation
5414 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5419 The filter converts the transfer characteristics, color space and color
5420 primaries to the specified user values. The output value, if not specified,
5421 is set to a default value based on the "all" property. If that property is
5422 also not specified, the filter will log an error. The output color range and
5423 format default to the same value as the input color range and format. The
5424 input transfer characteristics, color space, color primaries and color range
5425 should be set on the input data. If any of these are missing, the filter will
5426 log an error and no conversion will take place.
5428 For example to convert the input to SMPTE-240M, use the command:
5430 colorspace=smpte240m
5433 @section convolution
5435 Apply convolution 3x3 or 5x5 filter.
5437 The filter accepts the following options:
5444 Set matrix for each plane.
5445 Matrix is sequence of 9 or 25 signed integers.
5451 Set multiplier for calculated value for each plane.
5457 Set bias for each plane. This value is added to the result of the multiplication.
5458 Useful for making the overall image brighter or darker. Default is 0.0.
5461 @subsection Examples
5467 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"
5473 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"
5479 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"
5485 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"
5491 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"
5497 Copy the input source unchanged to the output. This is mainly useful for
5502 Video filtering on GPU using Apple's CoreImage API on OSX.
5504 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5505 processed by video hardware. However, software-based OpenGL implementations
5506 exist which means there is no guarantee for hardware processing. It depends on
5509 There are many filters and image generators provided by Apple that come with a
5510 large variety of options. The filter has to be referenced by its name along
5513 The coreimage filter accepts the following options:
5516 List all available filters and generators along with all their respective
5517 options as well as possible minimum and maximum values along with the default
5524 Specify all filters by their respective name and options.
5525 Use @var{list_filters} to determine all valid filter names and options.
5526 Numerical options are specified by a float value and are automatically clamped
5527 to their respective value range. Vector and color options have to be specified
5528 by a list of space separated float values. Character escaping has to be done.
5529 A special option name @code{default} is available to use default options for a
5532 It is required to specify either @code{default} or at least one of the filter options.
5533 All omitted options are used with their default values.
5534 The syntax of the filter string is as follows:
5536 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5540 Specify a rectangle where the output of the filter chain is copied into the
5541 input image. It is given by a list of space separated float values:
5543 output_rect=x\ y\ width\ height
5545 If not given, the output rectangle equals the dimensions of the input image.
5546 The output rectangle is automatically cropped at the borders of the input
5547 image. Negative values are valid for each component.
5549 output_rect=25\ 25\ 100\ 100
5553 Several filters can be chained for successive processing without GPU-HOST
5554 transfers allowing for fast processing of complex filter chains.
5555 Currently, only filters with zero (generators) or exactly one (filters) input
5556 image and one output image are supported. Also, transition filters are not yet
5559 Some filters generate output images with additional padding depending on the
5560 respective filter kernel. The padding is automatically removed to ensure the
5561 filter output has the same size as the input image.
5563 For image generators, the size of the output image is determined by the
5564 previous output image of the filter chain or the input image of the whole
5565 filterchain, respectively. The generators do not use the pixel information of
5566 this image to generate their output. However, the generated output is
5567 blended onto this image, resulting in partial or complete coverage of the
5570 The @ref{coreimagesrc} video source can be used for generating input images
5571 which are directly fed into the filter chain. By using it, providing input
5572 images by another video source or an input video is not required.
5574 @subsection Examples
5579 List all filters available:
5581 coreimage=list_filters=true
5585 Use the CIBoxBlur filter with default options to blur an image:
5587 coreimage=filter=CIBoxBlur@@default
5591 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5592 its center at 100x100 and a radius of 50 pixels:
5594 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5598 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5599 given as complete and escaped command-line for Apple's standard bash shell:
5601 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5607 Crop the input video to given dimensions.
5609 It accepts the following parameters:
5613 The width of the output video. It defaults to @code{iw}.
5614 This expression is evaluated only once during the filter
5615 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5618 The height of the output video. It defaults to @code{ih}.
5619 This expression is evaluated only once during the filter
5620 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5623 The horizontal position, in the input video, of the left edge of the output
5624 video. It defaults to @code{(in_w-out_w)/2}.
5625 This expression is evaluated per-frame.
5628 The vertical position, in the input video, of the top edge of the output video.
5629 It defaults to @code{(in_h-out_h)/2}.
5630 This expression is evaluated per-frame.
5633 If set to 1 will force the output display aspect ratio
5634 to be the same of the input, by changing the output sample aspect
5635 ratio. It defaults to 0.
5638 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5639 expressions containing the following constants:
5644 The computed values for @var{x} and @var{y}. They are evaluated for
5649 The input width and height.
5653 These are the same as @var{in_w} and @var{in_h}.
5657 The output (cropped) width and height.
5661 These are the same as @var{out_w} and @var{out_h}.
5664 same as @var{iw} / @var{ih}
5667 input sample aspect ratio
5670 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5674 horizontal and vertical chroma subsample values. For example for the
5675 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5678 The number of the input frame, starting from 0.
5681 the position in the file of the input frame, NAN if unknown
5684 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5688 The expression for @var{out_w} may depend on the value of @var{out_h},
5689 and the expression for @var{out_h} may depend on @var{out_w}, but they
5690 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5691 evaluated after @var{out_w} and @var{out_h}.
5693 The @var{x} and @var{y} parameters specify the expressions for the
5694 position of the top-left corner of the output (non-cropped) area. They
5695 are evaluated for each frame. If the evaluated value is not valid, it
5696 is approximated to the nearest valid value.
5698 The expression for @var{x} may depend on @var{y}, and the expression
5699 for @var{y} may depend on @var{x}.
5701 @subsection Examples
5705 Crop area with size 100x100 at position (12,34).
5710 Using named options, the example above becomes:
5712 crop=w=100:h=100:x=12:y=34
5716 Crop the central input area with size 100x100:
5722 Crop the central input area with size 2/3 of the input video:
5724 crop=2/3*in_w:2/3*in_h
5728 Crop the input video central square:
5735 Delimit the rectangle with the top-left corner placed at position
5736 100:100 and the right-bottom corner corresponding to the right-bottom
5737 corner of the input image.
5739 crop=in_w-100:in_h-100:100:100
5743 Crop 10 pixels from the left and right borders, and 20 pixels from
5744 the top and bottom borders
5746 crop=in_w-2*10:in_h-2*20
5750 Keep only the bottom right quarter of the input image:
5752 crop=in_w/2:in_h/2:in_w/2:in_h/2
5756 Crop height for getting Greek harmony:
5758 crop=in_w:1/PHI*in_w
5762 Apply trembling effect:
5764 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)
5768 Apply erratic camera effect depending on timestamp:
5770 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)"
5774 Set x depending on the value of y:
5776 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5780 @subsection Commands
5782 This filter supports the following commands:
5788 Set width/height of the output video and the horizontal/vertical position
5790 The command accepts the same syntax of the corresponding option.
5792 If the specified expression is not valid, it is kept at its current
5798 Auto-detect the crop size.
5800 It calculates the necessary cropping parameters and prints the
5801 recommended parameters via the logging system. The detected dimensions
5802 correspond to the non-black area of the input video.
5804 It accepts the following parameters:
5809 Set higher black value threshold, which can be optionally specified
5810 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5811 value greater to the set value is considered non-black. It defaults to 24.
5812 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5813 on the bitdepth of the pixel format.
5816 The value which the width/height should be divisible by. It defaults to
5817 16. The offset is automatically adjusted to center the video. Use 2 to
5818 get only even dimensions (needed for 4:2:2 video). 16 is best when
5819 encoding to most video codecs.
5821 @item reset_count, reset
5822 Set the counter that determines after how many frames cropdetect will
5823 reset the previously detected largest video area and start over to
5824 detect the current optimal crop area. Default value is 0.
5826 This can be useful when channel logos distort the video area. 0
5827 indicates 'never reset', and returns the largest area encountered during
5834 Apply color adjustments using curves.
5836 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5837 component (red, green and blue) has its values defined by @var{N} key points
5838 tied from each other using a smooth curve. The x-axis represents the pixel
5839 values from the input frame, and the y-axis the new pixel values to be set for
5842 By default, a component curve is defined by the two points @var{(0;0)} and
5843 @var{(1;1)}. This creates a straight line where each original pixel value is
5844 "adjusted" to its own value, which means no change to the image.
5846 The filter allows you to redefine these two points and add some more. A new
5847 curve (using a natural cubic spline interpolation) will be define to pass
5848 smoothly through all these new coordinates. The new defined points needs to be
5849 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5850 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5851 the vector spaces, the values will be clipped accordingly.
5853 The filter accepts the following options:
5857 Select one of the available color presets. This option can be used in addition
5858 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5859 options takes priority on the preset values.
5860 Available presets are:
5863 @item color_negative
5866 @item increase_contrast
5868 @item linear_contrast
5869 @item medium_contrast
5871 @item strong_contrast
5874 Default is @code{none}.
5876 Set the master key points. These points will define a second pass mapping. It
5877 is sometimes called a "luminance" or "value" mapping. It can be used with
5878 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5879 post-processing LUT.
5881 Set the key points for the red component.
5883 Set the key points for the green component.
5885 Set the key points for the blue component.
5887 Set the key points for all components (not including master).
5888 Can be used in addition to the other key points component
5889 options. In this case, the unset component(s) will fallback on this
5890 @option{all} setting.
5892 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5894 Save Gnuplot script of the curves in specified file.
5897 To avoid some filtergraph syntax conflicts, each key points list need to be
5898 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5900 @subsection Examples
5904 Increase slightly the middle level of blue:
5906 curves=blue='0/0 0.5/0.58 1/1'
5912 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'
5914 Here we obtain the following coordinates for each components:
5917 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5919 @code{(0;0) (0.50;0.48) (1;1)}
5921 @code{(0;0.22) (0.49;0.44) (1;0.80)}
5925 The previous example can also be achieved with the associated built-in preset:
5927 curves=preset=vintage
5937 Use a Photoshop preset and redefine the points of the green component:
5939 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
5943 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
5944 and @command{gnuplot}:
5946 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
5947 gnuplot -p /tmp/curves.plt
5953 Video data analysis filter.
5955 This filter shows hexadecimal pixel values of part of video.
5957 The filter accepts the following options:
5961 Set output video size.
5964 Set x offset from where to pick pixels.
5967 Set y offset from where to pick pixels.
5970 Set scope mode, can be one of the following:
5973 Draw hexadecimal pixel values with white color on black background.
5976 Draw hexadecimal pixel values with input video pixel color on black
5980 Draw hexadecimal pixel values on color background picked from input video,
5981 the text color is picked in such way so its always visible.
5985 Draw rows and columns numbers on left and top of video.
5990 Denoise frames using 2D DCT (frequency domain filtering).
5992 This filter is not designed for real time.
5994 The filter accepts the following options:
5998 Set the noise sigma constant.
6000 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6001 coefficient (absolute value) below this threshold with be dropped.
6003 If you need a more advanced filtering, see @option{expr}.
6005 Default is @code{0}.
6008 Set number overlapping pixels for each block. Since the filter can be slow, you
6009 may want to reduce this value, at the cost of a less effective filter and the
6010 risk of various artefacts.
6012 If the overlapping value doesn't permit processing the whole input width or
6013 height, a warning will be displayed and according borders won't be denoised.
6015 Default value is @var{blocksize}-1, which is the best possible setting.
6018 Set the coefficient factor expression.
6020 For each coefficient of a DCT block, this expression will be evaluated as a
6021 multiplier value for the coefficient.
6023 If this is option is set, the @option{sigma} option will be ignored.
6025 The absolute value of the coefficient can be accessed through the @var{c}
6029 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6030 @var{blocksize}, which is the width and height of the processed blocks.
6032 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6033 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6034 on the speed processing. Also, a larger block size does not necessarily means a
6038 @subsection Examples
6040 Apply a denoise with a @option{sigma} of @code{4.5}:
6045 The same operation can be achieved using the expression system:
6047 dctdnoiz=e='gte(c, 4.5*3)'
6050 Violent denoise using a block size of @code{16x16}:
6057 Remove banding artifacts from input video.
6058 It works by replacing banded pixels with average value of referenced pixels.
6060 The filter accepts the following options:
6067 Set banding detection threshold for each plane. Default is 0.02.
6068 Valid range is 0.00003 to 0.5.
6069 If difference between current pixel and reference pixel is less than threshold,
6070 it will be considered as banded.
6073 Banding detection range in pixels. Default is 16. If positive, random number
6074 in range 0 to set value will be used. If negative, exact absolute value
6076 The range defines square of four pixels around current pixel.
6079 Set direction in radians from which four pixel will be compared. If positive,
6080 random direction from 0 to set direction will be picked. If negative, exact of
6081 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6082 will pick only pixels on same row and -PI/2 will pick only pixels on same
6086 If enabled, current pixel is compared with average value of all four
6087 surrounding pixels. The default is enabled. If disabled current pixel is
6088 compared with all four surrounding pixels. The pixel is considered banded
6089 if only all four differences with surrounding pixels are less than threshold.
6095 Drop duplicated frames at regular intervals.
6097 The filter accepts the following options:
6101 Set the number of frames from which one will be dropped. Setting this to
6102 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6103 Default is @code{5}.
6106 Set the threshold for duplicate detection. If the difference metric for a frame
6107 is less than or equal to this value, then it is declared as duplicate. Default
6111 Set scene change threshold. Default is @code{15}.
6115 Set the size of the x and y-axis blocks used during metric calculations.
6116 Larger blocks give better noise suppression, but also give worse detection of
6117 small movements. Must be a power of two. Default is @code{32}.
6120 Mark main input as a pre-processed input and activate clean source input
6121 stream. This allows the input to be pre-processed with various filters to help
6122 the metrics calculation while keeping the frame selection lossless. When set to
6123 @code{1}, the first stream is for the pre-processed input, and the second
6124 stream is the clean source from where the kept frames are chosen. Default is
6128 Set whether or not chroma is considered in the metric calculations. Default is
6134 Apply deflate effect to the video.
6136 This filter replaces the pixel by the local(3x3) average by taking into account
6137 only values lower than the pixel.
6139 It accepts the following options:
6146 Limit the maximum change for each plane, default is 65535.
6147 If 0, plane will remain unchanged.
6152 Remove judder produced by partially interlaced telecined content.
6154 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6155 source was partially telecined content then the output of @code{pullup,dejudder}
6156 will have a variable frame rate. May change the recorded frame rate of the
6157 container. Aside from that change, this filter will not affect constant frame
6160 The option available in this filter is:
6164 Specify the length of the window over which the judder repeats.
6166 Accepts any integer greater than 1. Useful values are:
6170 If the original was telecined from 24 to 30 fps (Film to NTSC).
6173 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6176 If a mixture of the two.
6179 The default is @samp{4}.
6184 Suppress a TV station logo by a simple interpolation of the surrounding
6185 pixels. Just set a rectangle covering the logo and watch it disappear
6186 (and sometimes something even uglier appear - your mileage may vary).
6188 It accepts the following parameters:
6193 Specify the top left corner coordinates of the logo. They must be
6198 Specify the width and height of the logo to clear. They must be
6202 Specify the thickness of the fuzzy edge of the rectangle (added to
6203 @var{w} and @var{h}). The default value is 1. This option is
6204 deprecated, setting higher values should no longer be necessary and
6208 When set to 1, a green rectangle is drawn on the screen to simplify
6209 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6210 The default value is 0.
6212 The rectangle is drawn on the outermost pixels which will be (partly)
6213 replaced with interpolated values. The values of the next pixels
6214 immediately outside this rectangle in each direction will be used to
6215 compute the interpolated pixel values inside the rectangle.
6219 @subsection Examples
6223 Set a rectangle covering the area with top left corner coordinates 0,0
6224 and size 100x77, and a band of size 10:
6226 delogo=x=0:y=0:w=100:h=77:band=10
6233 Attempt to fix small changes in horizontal and/or vertical shift. This
6234 filter helps remove camera shake from hand-holding a camera, bumping a
6235 tripod, moving on a vehicle, etc.
6237 The filter accepts the following options:
6245 Specify a rectangular area where to limit the search for motion
6247 If desired the search for motion vectors can be limited to a
6248 rectangular area of the frame defined by its top left corner, width
6249 and height. These parameters have the same meaning as the drawbox
6250 filter which can be used to visualise the position of the bounding
6253 This is useful when simultaneous movement of subjects within the frame
6254 might be confused for camera motion by the motion vector search.
6256 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6257 then the full frame is used. This allows later options to be set
6258 without specifying the bounding box for the motion vector search.
6260 Default - search the whole frame.
6264 Specify the maximum extent of movement in x and y directions in the
6265 range 0-64 pixels. Default 16.
6268 Specify how to generate pixels to fill blanks at the edge of the
6269 frame. Available values are:
6272 Fill zeroes at blank locations
6274 Original image at blank locations
6276 Extruded edge value at blank locations
6278 Mirrored edge at blank locations
6280 Default value is @samp{mirror}.
6283 Specify the blocksize to use for motion search. Range 4-128 pixels,
6287 Specify the contrast threshold for blocks. Only blocks with more than
6288 the specified contrast (difference between darkest and lightest
6289 pixels) will be considered. Range 1-255, default 125.
6292 Specify the search strategy. Available values are:
6295 Set exhaustive search
6297 Set less exhaustive search.
6299 Default value is @samp{exhaustive}.
6302 If set then a detailed log of the motion search is written to the
6306 If set to 1, specify using OpenCL capabilities, only available if
6307 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6313 Apply an exact inverse of the telecine operation. It requires a predefined
6314 pattern specified using the pattern option which must be the same as that passed
6315 to the telecine filter.
6317 This filter accepts the following options:
6326 The default value is @code{top}.
6330 A string of numbers representing the pulldown pattern you wish to apply.
6331 The default value is @code{23}.
6334 A number representing position of the first frame with respect to the telecine
6335 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6340 Apply dilation effect to the video.
6342 This filter replaces the pixel by the local(3x3) maximum.
6344 It accepts the following options:
6351 Limit the maximum change for each plane, default is 65535.
6352 If 0, plane will remain unchanged.
6355 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6358 Flags to local 3x3 coordinates maps like this:
6367 Displace pixels as indicated by second and third input stream.
6369 It takes three input streams and outputs one stream, the first input is the
6370 source, and second and third input are displacement maps.
6372 The second input specifies how much to displace pixels along the
6373 x-axis, while the third input specifies how much to displace pixels
6375 If one of displacement map streams terminates, last frame from that
6376 displacement map will be used.
6378 Note that once generated, displacements maps can be reused over and over again.
6380 A description of the accepted options follows.
6384 Set displace behavior for pixels that are out of range.
6386 Available values are:
6389 Missing pixels are replaced by black pixels.
6392 Adjacent pixels will spread out to replace missing pixels.
6395 Out of range pixels are wrapped so they point to pixels of other side.
6397 Default is @samp{smear}.
6401 @subsection Examples
6405 Add ripple effect to rgb input of video size hd720:
6407 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
6411 Add wave effect to rgb input of video size hd720:
6413 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
6419 Draw a colored box on the input image.
6421 It accepts the following parameters:
6426 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6430 The expressions which specify the width and height of the box; if 0 they are interpreted as
6431 the input width and height. It defaults to 0.
6434 Specify the color of the box to write. For the general syntax of this option,
6435 check the "Color" section in the ffmpeg-utils manual. If the special
6436 value @code{invert} is used, the box edge color is the same as the
6437 video with inverted luma.
6440 The expression which sets the thickness of the box edge. Default value is @code{3}.
6442 See below for the list of accepted constants.
6445 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6446 following constants:
6450 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6454 horizontal and vertical chroma subsample values. For example for the
6455 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6459 The input width and height.
6462 The input sample aspect ratio.
6466 The x and y offset coordinates where the box is drawn.
6470 The width and height of the drawn box.
6473 The thickness of the drawn box.
6475 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6476 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6480 @subsection Examples
6484 Draw a black box around the edge of the input image:
6490 Draw a box with color red and an opacity of 50%:
6492 drawbox=10:20:200:60:red@@0.5
6495 The previous example can be specified as:
6497 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6501 Fill the box with pink color:
6503 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6507 Draw a 2-pixel red 2.40:1 mask:
6509 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
6515 Draw a grid on the input image.
6517 It accepts the following parameters:
6522 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6526 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6527 input width and height, respectively, minus @code{thickness}, so image gets
6528 framed. Default to 0.
6531 Specify the color of the grid. For the general syntax of this option,
6532 check the "Color" section in the ffmpeg-utils manual. If the special
6533 value @code{invert} is used, the grid color is the same as the
6534 video with inverted luma.
6537 The expression which sets the thickness of the grid line. Default value is @code{1}.
6539 See below for the list of accepted constants.
6542 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6543 following constants:
6547 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6551 horizontal and vertical chroma subsample values. For example for the
6552 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6556 The input grid cell width and height.
6559 The input sample aspect ratio.
6563 The x and y coordinates of some point of grid intersection (meant to configure offset).
6567 The width and height of the drawn cell.
6570 The thickness of the drawn cell.
6572 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6573 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6577 @subsection Examples
6581 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6583 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6587 Draw a white 3x3 grid with an opacity of 50%:
6589 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6596 Draw a text string or text from a specified file on top of a video, using the
6597 libfreetype library.
6599 To enable compilation of this filter, you need to configure FFmpeg with
6600 @code{--enable-libfreetype}.
6601 To enable default font fallback and the @var{font} option you need to
6602 configure FFmpeg with @code{--enable-libfontconfig}.
6603 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6604 @code{--enable-libfribidi}.
6608 It accepts the following parameters:
6613 Used to draw a box around text using the background color.
6614 The value must be either 1 (enable) or 0 (disable).
6615 The default value of @var{box} is 0.
6618 Set the width of the border to be drawn around the box using @var{boxcolor}.
6619 The default value of @var{boxborderw} is 0.
6622 The color to be used for drawing box around text. For the syntax of this
6623 option, check the "Color" section in the ffmpeg-utils manual.
6625 The default value of @var{boxcolor} is "white".
6628 Set the width of the border to be drawn around the text using @var{bordercolor}.
6629 The default value of @var{borderw} is 0.
6632 Set the color to be used for drawing border around text. For the syntax of this
6633 option, check the "Color" section in the ffmpeg-utils manual.
6635 The default value of @var{bordercolor} is "black".
6638 Select how the @var{text} is expanded. Can be either @code{none},
6639 @code{strftime} (deprecated) or
6640 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6644 If true, check and fix text coords to avoid clipping.
6647 The color to be used for drawing fonts. For the syntax of this option, check
6648 the "Color" section in the ffmpeg-utils manual.
6650 The default value of @var{fontcolor} is "black".
6652 @item fontcolor_expr
6653 String which is expanded the same way as @var{text} to obtain dynamic
6654 @var{fontcolor} value. By default this option has empty value and is not
6655 processed. When this option is set, it overrides @var{fontcolor} option.
6658 The font family to be used for drawing text. By default Sans.
6661 The font file to be used for drawing text. The path must be included.
6662 This parameter is mandatory if the fontconfig support is disabled.
6665 This option does not exist, please see the timeline system
6668 Draw the text applying alpha blending. The value can
6669 be either a number between 0.0 and 1.0
6670 The expression accepts the same variables @var{x, y} do.
6671 The default value is 1.
6672 Please see fontcolor_expr
6675 The font size to be used for drawing text.
6676 The default value of @var{fontsize} is 16.
6679 If set to 1, attempt to shape the text (for example, reverse the order of
6680 right-to-left text and join Arabic characters) before drawing it.
6681 Otherwise, just draw the text exactly as given.
6682 By default 1 (if supported).
6685 The flags to be used for loading the fonts.
6687 The flags map the corresponding flags supported by libfreetype, and are
6688 a combination of the following values:
6695 @item vertical_layout
6696 @item force_autohint
6699 @item ignore_global_advance_width
6701 @item ignore_transform
6707 Default value is "default".
6709 For more information consult the documentation for the FT_LOAD_*
6713 The color to be used for drawing a shadow behind the drawn text. For the
6714 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6716 The default value of @var{shadowcolor} is "black".
6720 The x and y offsets for the text shadow position with respect to the
6721 position of the text. They can be either positive or negative
6722 values. The default value for both is "0".
6725 The starting frame number for the n/frame_num variable. The default value
6729 The size in number of spaces to use for rendering the tab.
6733 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6734 format. It can be used with or without text parameter. @var{timecode_rate}
6735 option must be specified.
6737 @item timecode_rate, rate, r
6738 Set the timecode frame rate (timecode only).
6741 The text string to be drawn. The text must be a sequence of UTF-8
6743 This parameter is mandatory if no file is specified with the parameter
6747 A text file containing text to be drawn. The text must be a sequence
6748 of UTF-8 encoded characters.
6750 This parameter is mandatory if no text string is specified with the
6751 parameter @var{text}.
6753 If both @var{text} and @var{textfile} are specified, an error is thrown.
6756 If set to 1, the @var{textfile} will be reloaded before each frame.
6757 Be sure to update it atomically, or it may be read partially, or even fail.
6761 The expressions which specify the offsets where text will be drawn
6762 within the video frame. They are relative to the top/left border of the
6765 The default value of @var{x} and @var{y} is "0".
6767 See below for the list of accepted constants and functions.
6770 The parameters for @var{x} and @var{y} are expressions containing the
6771 following constants and functions:
6775 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6779 horizontal and vertical chroma subsample values. For example for the
6780 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6783 the height of each text line
6791 @item max_glyph_a, ascent
6792 the maximum distance from the baseline to the highest/upper grid
6793 coordinate used to place a glyph outline point, for all the rendered
6795 It is a positive value, due to the grid's orientation with the Y axis
6798 @item max_glyph_d, descent
6799 the maximum distance from the baseline to the lowest grid coordinate
6800 used to place a glyph outline point, for all the rendered glyphs.
6801 This is a negative value, due to the grid's orientation, with the Y axis
6805 maximum glyph height, that is the maximum height for all the glyphs
6806 contained in the rendered text, it is equivalent to @var{ascent} -
6810 maximum glyph width, that is the maximum width for all the glyphs
6811 contained in the rendered text
6814 the number of input frame, starting from 0
6816 @item rand(min, max)
6817 return a random number included between @var{min} and @var{max}
6820 The input sample aspect ratio.
6823 timestamp expressed in seconds, NAN if the input timestamp is unknown
6826 the height of the rendered text
6829 the width of the rendered text
6833 the x and y offset coordinates where the text is drawn.
6835 These parameters allow the @var{x} and @var{y} expressions to refer
6836 each other, so you can for example specify @code{y=x/dar}.
6839 @anchor{drawtext_expansion}
6840 @subsection Text expansion
6842 If @option{expansion} is set to @code{strftime},
6843 the filter recognizes strftime() sequences in the provided text and
6844 expands them accordingly. Check the documentation of strftime(). This
6845 feature is deprecated.
6847 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6849 If @option{expansion} is set to @code{normal} (which is the default),
6850 the following expansion mechanism is used.
6852 The backslash character @samp{\}, followed by any character, always expands to
6853 the second character.
6855 Sequence of the form @code{%@{...@}} are expanded. The text between the
6856 braces is a function name, possibly followed by arguments separated by ':'.
6857 If the arguments contain special characters or delimiters (':' or '@}'),
6858 they should be escaped.
6860 Note that they probably must also be escaped as the value for the
6861 @option{text} option in the filter argument string and as the filter
6862 argument in the filtergraph description, and possibly also for the shell,
6863 that makes up to four levels of escaping; using a text file avoids these
6866 The following functions are available:
6871 The expression evaluation result.
6873 It must take one argument specifying the expression to be evaluated,
6874 which accepts the same constants and functions as the @var{x} and
6875 @var{y} values. Note that not all constants should be used, for
6876 example the text size is not known when evaluating the expression, so
6877 the constants @var{text_w} and @var{text_h} will have an undefined
6880 @item expr_int_format, eif
6881 Evaluate the expression's value and output as formatted integer.
6883 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6884 The second argument specifies the output format. Allowed values are @samp{x},
6885 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6886 @code{printf} function.
6887 The third parameter is optional and sets the number of positions taken by the output.
6888 It can be used to add padding with zeros from the left.
6891 The time at which the filter is running, expressed in UTC.
6892 It can accept an argument: a strftime() format string.
6895 The time at which the filter is running, expressed in the local time zone.
6896 It can accept an argument: a strftime() format string.
6899 Frame metadata. Takes one or two arguments.
6901 The first argument is mandatory and specifies the metadata key.
6903 The second argument is optional and specifies a default value, used when the
6904 metadata key is not found or empty.
6907 The frame number, starting from 0.
6910 A 1 character description of the current picture type.
6913 The timestamp of the current frame.
6914 It can take up to three arguments.
6916 The first argument is the format of the timestamp; it defaults to @code{flt}
6917 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6918 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6919 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
6920 @code{localtime} stands for the timestamp of the frame formatted as
6921 local time zone time.
6923 The second argument is an offset added to the timestamp.
6925 If the format is set to @code{localtime} or @code{gmtime},
6926 a third argument may be supplied: a strftime() format string.
6927 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
6930 @subsection Examples
6934 Draw "Test Text" with font FreeSerif, using the default values for the
6935 optional parameters.
6938 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
6942 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
6943 and y=50 (counting from the top-left corner of the screen), text is
6944 yellow with a red box around it. Both the text and the box have an
6948 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
6949 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
6952 Note that the double quotes are not necessary if spaces are not used
6953 within the parameter list.
6956 Show the text at the center of the video frame:
6958 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
6962 Show the text at a random position, switching to a new position every 30 seconds:
6964 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)"
6968 Show a text line sliding from right to left in the last row of the video
6969 frame. The file @file{LONG_LINE} is assumed to contain a single line
6972 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
6976 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
6978 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
6982 Draw a single green letter "g", at the center of the input video.
6983 The glyph baseline is placed at half screen height.
6985 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
6989 Show text for 1 second every 3 seconds:
6991 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
6995 Use fontconfig to set the font. Note that the colons need to be escaped.
6997 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7001 Print the date of a real-time encoding (see strftime(3)):
7003 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7007 Show text fading in and out (appearing/disappearing):
7010 DS=1.0 # display start
7011 DE=10.0 # display end
7012 FID=1.5 # fade in duration
7013 FOD=5 # fade out duration
7014 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 @}"
7019 For more information about libfreetype, check:
7020 @url{http://www.freetype.org/}.
7022 For more information about fontconfig, check:
7023 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7025 For more information about libfribidi, check:
7026 @url{http://fribidi.org/}.
7030 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7032 The filter accepts the following options:
7037 Set low and high threshold values used by the Canny thresholding
7040 The high threshold selects the "strong" edge pixels, which are then
7041 connected through 8-connectivity with the "weak" edge pixels selected
7042 by the low threshold.
7044 @var{low} and @var{high} threshold values must be chosen in the range
7045 [0,1], and @var{low} should be lesser or equal to @var{high}.
7047 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7051 Define the drawing mode.
7055 Draw white/gray wires on black background.
7058 Mix the colors to create a paint/cartoon effect.
7061 Default value is @var{wires}.
7064 @subsection Examples
7068 Standard edge detection with custom values for the hysteresis thresholding:
7070 edgedetect=low=0.1:high=0.4
7074 Painting effect without thresholding:
7076 edgedetect=mode=colormix:high=0
7081 Set brightness, contrast, saturation and approximate gamma adjustment.
7083 The filter accepts the following options:
7087 Set the contrast expression. The value must be a float value in range
7088 @code{-2.0} to @code{2.0}. The default value is "1".
7091 Set the brightness expression. The value must be a float value in
7092 range @code{-1.0} to @code{1.0}. The default value is "0".
7095 Set the saturation expression. The value must be a float in
7096 range @code{0.0} to @code{3.0}. The default value is "1".
7099 Set the gamma expression. The value must be a float in range
7100 @code{0.1} to @code{10.0}. The default value is "1".
7103 Set the gamma expression for red. The value must be a float in
7104 range @code{0.1} to @code{10.0}. The default value is "1".
7107 Set the gamma expression for green. The value must be a float in range
7108 @code{0.1} to @code{10.0}. The default value is "1".
7111 Set the gamma expression for blue. The value must be a float in range
7112 @code{0.1} to @code{10.0}. The default value is "1".
7115 Set the gamma weight expression. It can be used to reduce the effect
7116 of a high gamma value on bright image areas, e.g. keep them from
7117 getting overamplified and just plain white. The value must be a float
7118 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7119 gamma correction all the way down while @code{1.0} leaves it at its
7120 full strength. Default is "1".
7123 Set when the expressions for brightness, contrast, saturation and
7124 gamma expressions are evaluated.
7126 It accepts the following values:
7129 only evaluate expressions once during the filter initialization or
7130 when a command is processed
7133 evaluate expressions for each incoming frame
7136 Default value is @samp{init}.
7139 The expressions accept the following parameters:
7142 frame count of the input frame starting from 0
7145 byte position of the corresponding packet in the input file, NAN if
7149 frame rate of the input video, NAN if the input frame rate is unknown
7152 timestamp expressed in seconds, NAN if the input timestamp is unknown
7155 @subsection Commands
7156 The filter supports the following commands:
7160 Set the contrast expression.
7163 Set the brightness expression.
7166 Set the saturation expression.
7169 Set the gamma expression.
7172 Set the gamma_r expression.
7175 Set gamma_g expression.
7178 Set gamma_b expression.
7181 Set gamma_weight expression.
7183 The command accepts the same syntax of the corresponding option.
7185 If the specified expression is not valid, it is kept at its current
7192 Apply erosion effect to the video.
7194 This filter replaces the pixel by the local(3x3) minimum.
7196 It accepts the following options:
7203 Limit the maximum change for each plane, default is 65535.
7204 If 0, plane will remain unchanged.
7207 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7210 Flags to local 3x3 coordinates maps like this:
7217 @section extractplanes
7219 Extract color channel components from input video stream into
7220 separate grayscale video streams.
7222 The filter accepts the following option:
7226 Set plane(s) to extract.
7228 Available values for planes are:
7239 Choosing planes not available in the input will result in an error.
7240 That means you cannot select @code{r}, @code{g}, @code{b} planes
7241 with @code{y}, @code{u}, @code{v} planes at same time.
7244 @subsection Examples
7248 Extract luma, u and v color channel component from input video frame
7249 into 3 grayscale outputs:
7251 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
7257 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7259 For each input image, the filter will compute the optimal mapping from
7260 the input to the output given the codebook length, that is the number
7261 of distinct output colors.
7263 This filter accepts the following options.
7266 @item codebook_length, l
7267 Set codebook length. The value must be a positive integer, and
7268 represents the number of distinct output colors. Default value is 256.
7271 Set the maximum number of iterations to apply for computing the optimal
7272 mapping. The higher the value the better the result and the higher the
7273 computation time. Default value is 1.
7276 Set a random seed, must be an integer included between 0 and
7277 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7278 will try to use a good random seed on a best effort basis.
7281 Set pal8 output pixel format. This option does not work with codebook
7282 length greater than 256.
7287 Apply a fade-in/out effect to the input video.
7289 It accepts the following parameters:
7293 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7295 Default is @code{in}.
7297 @item start_frame, s
7298 Specify the number of the frame to start applying the fade
7299 effect at. Default is 0.
7302 The number of frames that the fade effect lasts. At the end of the
7303 fade-in effect, the output video will have the same intensity as the input video.
7304 At the end of the fade-out transition, the output video will be filled with the
7305 selected @option{color}.
7309 If set to 1, fade only alpha channel, if one exists on the input.
7312 @item start_time, st
7313 Specify the timestamp (in seconds) of the frame to start to apply the fade
7314 effect. If both start_frame and start_time are specified, the fade will start at
7315 whichever comes last. Default is 0.
7318 The number of seconds for which the fade effect has to last. At the end of the
7319 fade-in effect the output video will have the same intensity as the input video,
7320 at the end of the fade-out transition the output video will be filled with the
7321 selected @option{color}.
7322 If both duration and nb_frames are specified, duration is used. Default is 0
7323 (nb_frames is used by default).
7326 Specify the color of the fade. Default is "black".
7329 @subsection Examples
7333 Fade in the first 30 frames of video:
7338 The command above is equivalent to:
7344 Fade out the last 45 frames of a 200-frame video:
7347 fade=type=out:start_frame=155:nb_frames=45
7351 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7353 fade=in:0:25, fade=out:975:25
7357 Make the first 5 frames yellow, then fade in from frame 5-24:
7359 fade=in:5:20:color=yellow
7363 Fade in alpha over first 25 frames of video:
7365 fade=in:0:25:alpha=1
7369 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7371 fade=t=in:st=5.5:d=0.5
7377 Apply arbitrary expressions to samples in frequency domain
7381 Adjust the dc value (gain) of the luma plane of the image. The filter
7382 accepts an integer value in range @code{0} to @code{1000}. The default
7383 value is set to @code{0}.
7386 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7387 filter accepts an integer value in range @code{0} to @code{1000}. The
7388 default value is set to @code{0}.
7391 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7392 filter accepts an integer value in range @code{0} to @code{1000}. The
7393 default value is set to @code{0}.
7396 Set the frequency domain weight expression for the luma plane.
7399 Set the frequency domain weight expression for the 1st chroma plane.
7402 Set the frequency domain weight expression for the 2nd chroma plane.
7404 The filter accepts the following variables:
7407 The coordinates of the current sample.
7411 The width and height of the image.
7414 @subsection Examples
7420 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7426 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7432 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7438 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7445 Extract a single field from an interlaced image using stride
7446 arithmetic to avoid wasting CPU time. The output frames are marked as
7449 The filter accepts the following options:
7453 Specify whether to extract the top (if the value is @code{0} or
7454 @code{top}) or the bottom field (if the value is @code{1} or
7460 Create new frames by copying the top and bottom fields from surrounding frames
7461 supplied as numbers by the hint file.
7465 Set file containing hints: absolute/relative frame numbers.
7467 There must be one line for each frame in a clip. Each line must contain two
7468 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7469 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7470 is current frame number for @code{absolute} mode or out of [-1, 1] range
7471 for @code{relative} mode. First number tells from which frame to pick up top
7472 field and second number tells from which frame to pick up bottom field.
7474 If optionally followed by @code{+} output frame will be marked as interlaced,
7475 else if followed by @code{-} output frame will be marked as progressive, else
7476 it will be marked same as input frame.
7477 If line starts with @code{#} or @code{;} that line is skipped.
7480 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7483 Example of first several lines of @code{hint} file for @code{relative} mode:
7486 1,0 - # second frame, use third's frame top field and second's frame bottom field
7487 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7504 Field matching filter for inverse telecine. It is meant to reconstruct the
7505 progressive frames from a telecined stream. The filter does not drop duplicated
7506 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7507 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7509 The separation of the field matching and the decimation is notably motivated by
7510 the possibility of inserting a de-interlacing filter fallback between the two.
7511 If the source has mixed telecined and real interlaced content,
7512 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7513 But these remaining combed frames will be marked as interlaced, and thus can be
7514 de-interlaced by a later filter such as @ref{yadif} before decimation.
7516 In addition to the various configuration options, @code{fieldmatch} can take an
7517 optional second stream, activated through the @option{ppsrc} option. If
7518 enabled, the frames reconstruction will be based on the fields and frames from
7519 this second stream. This allows the first input to be pre-processed in order to
7520 help the various algorithms of the filter, while keeping the output lossless
7521 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7522 or brightness/contrast adjustments can help.
7524 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7525 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7526 which @code{fieldmatch} is based on. While the semantic and usage are very
7527 close, some behaviour and options names can differ.
7529 The @ref{decimate} filter currently only works for constant frame rate input.
7530 If your input has mixed telecined (30fps) and progressive content with a lower
7531 framerate like 24fps use the following filterchain to produce the necessary cfr
7532 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7534 The filter accepts the following options:
7538 Specify the assumed field order of the input stream. Available values are:
7542 Auto detect parity (use FFmpeg's internal parity value).
7544 Assume bottom field first.
7546 Assume top field first.
7549 Note that it is sometimes recommended not to trust the parity announced by the
7552 Default value is @var{auto}.
7555 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7556 sense that it won't risk creating jerkiness due to duplicate frames when
7557 possible, but if there are bad edits or blended fields it will end up
7558 outputting combed frames when a good match might actually exist. On the other
7559 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7560 but will almost always find a good frame if there is one. The other values are
7561 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7562 jerkiness and creating duplicate frames versus finding good matches in sections
7563 with bad edits, orphaned fields, blended fields, etc.
7565 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7567 Available values are:
7571 2-way matching (p/c)
7573 2-way matching, and trying 3rd match if still combed (p/c + n)
7575 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7577 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7578 still combed (p/c + n + u/b)
7580 3-way matching (p/c/n)
7582 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7583 detected as combed (p/c/n + u/b)
7586 The parenthesis at the end indicate the matches that would be used for that
7587 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7590 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7593 Default value is @var{pc_n}.
7596 Mark the main input stream as a pre-processed input, and enable the secondary
7597 input stream as the clean source to pick the fields from. See the filter
7598 introduction for more details. It is similar to the @option{clip2} feature from
7601 Default value is @code{0} (disabled).
7604 Set the field to match from. It is recommended to set this to the same value as
7605 @option{order} unless you experience matching failures with that setting. In
7606 certain circumstances changing the field that is used to match from can have a
7607 large impact on matching performance. Available values are:
7611 Automatic (same value as @option{order}).
7613 Match from the bottom field.
7615 Match from the top field.
7618 Default value is @var{auto}.
7621 Set whether or not chroma is included during the match comparisons. In most
7622 cases it is recommended to leave this enabled. You should set this to @code{0}
7623 only if your clip has bad chroma problems such as heavy rainbowing or other
7624 artifacts. Setting this to @code{0} could also be used to speed things up at
7625 the cost of some accuracy.
7627 Default value is @code{1}.
7631 These define an exclusion band which excludes the lines between @option{y0} and
7632 @option{y1} from being included in the field matching decision. An exclusion
7633 band can be used to ignore subtitles, a logo, or other things that may
7634 interfere with the matching. @option{y0} sets the starting scan line and
7635 @option{y1} sets the ending line; all lines in between @option{y0} and
7636 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7637 @option{y0} and @option{y1} to the same value will disable the feature.
7638 @option{y0} and @option{y1} defaults to @code{0}.
7641 Set the scene change detection threshold as a percentage of maximum change on
7642 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7643 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7644 @option{scthresh} is @code{[0.0, 100.0]}.
7646 Default value is @code{12.0}.
7649 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7650 account the combed scores of matches when deciding what match to use as the
7651 final match. Available values are:
7655 No final matching based on combed scores.
7657 Combed scores are only used when a scene change is detected.
7659 Use combed scores all the time.
7662 Default is @var{sc}.
7665 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7666 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7667 Available values are:
7671 No forced calculation.
7673 Force p/c/n calculations.
7675 Force p/c/n/u/b calculations.
7678 Default value is @var{none}.
7681 This is the area combing threshold used for combed frame detection. This
7682 essentially controls how "strong" or "visible" combing must be to be detected.
7683 Larger values mean combing must be more visible and smaller values mean combing
7684 can be less visible or strong and still be detected. Valid settings are from
7685 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7686 be detected as combed). This is basically a pixel difference value. A good
7687 range is @code{[8, 12]}.
7689 Default value is @code{9}.
7692 Sets whether or not chroma is considered in the combed frame decision. Only
7693 disable this if your source has chroma problems (rainbowing, etc.) that are
7694 causing problems for the combed frame detection with chroma enabled. Actually,
7695 using @option{chroma}=@var{0} is usually more reliable, except for the case
7696 where there is chroma only combing in the source.
7698 Default value is @code{0}.
7702 Respectively set the x-axis and y-axis size of the window used during combed
7703 frame detection. This has to do with the size of the area in which
7704 @option{combpel} pixels are required to be detected as combed for a frame to be
7705 declared combed. See the @option{combpel} parameter description for more info.
7706 Possible values are any number that is a power of 2 starting at 4 and going up
7709 Default value is @code{16}.
7712 The number of combed pixels inside any of the @option{blocky} by
7713 @option{blockx} size blocks on the frame for the frame to be detected as
7714 combed. While @option{cthresh} controls how "visible" the combing must be, this
7715 setting controls "how much" combing there must be in any localized area (a
7716 window defined by the @option{blockx} and @option{blocky} settings) on the
7717 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7718 which point no frames will ever be detected as combed). This setting is known
7719 as @option{MI} in TFM/VFM vocabulary.
7721 Default value is @code{80}.
7724 @anchor{p/c/n/u/b meaning}
7725 @subsection p/c/n/u/b meaning
7727 @subsubsection p/c/n
7729 We assume the following telecined stream:
7732 Top fields: 1 2 2 3 4
7733 Bottom fields: 1 2 3 4 4
7736 The numbers correspond to the progressive frame the fields relate to. Here, the
7737 first two frames are progressive, the 3rd and 4th are combed, and so on.
7739 When @code{fieldmatch} is configured to run a matching from bottom
7740 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7745 B 1 2 3 4 4 <-- matching reference
7754 As a result of the field matching, we can see that some frames get duplicated.
7755 To perform a complete inverse telecine, you need to rely on a decimation filter
7756 after this operation. See for instance the @ref{decimate} filter.
7758 The same operation now matching from top fields (@option{field}=@var{top})
7763 T 1 2 2 3 4 <-- matching reference
7773 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7774 basically, they refer to the frame and field of the opposite parity:
7777 @item @var{p} matches the field of the opposite parity in the previous frame
7778 @item @var{c} matches the field of the opposite parity in the current frame
7779 @item @var{n} matches the field of the opposite parity in the next frame
7784 The @var{u} and @var{b} matching are a bit special in the sense that they match
7785 from the opposite parity flag. In the following examples, we assume that we are
7786 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7787 'x' is placed above and below each matched fields.
7789 With bottom matching (@option{field}=@var{bottom}):
7794 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7795 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7803 With top matching (@option{field}=@var{top}):
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 @subsection Examples
7819 Simple IVTC of a top field first telecined stream:
7821 fieldmatch=order=tff:combmatch=none, decimate
7824 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7826 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7831 Transform the field order of the input video.
7833 It accepts the following parameters:
7838 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7839 for bottom field first.
7842 The default value is @samp{tff}.
7844 The transformation is done by shifting the picture content up or down
7845 by one line, and filling the remaining line with appropriate picture content.
7846 This method is consistent with most broadcast field order converters.
7848 If the input video is not flagged as being interlaced, or it is already
7849 flagged as being of the required output field order, then this filter does
7850 not alter the incoming video.
7852 It is very useful when converting to or from PAL DV material,
7853 which is bottom field first.
7857 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7860 @section fifo, afifo
7862 Buffer input images and send them when they are requested.
7864 It is mainly useful when auto-inserted by the libavfilter
7867 It does not take parameters.
7871 Find a rectangular object
7873 It accepts the following options:
7877 Filepath of the object image, needs to be in gray8.
7880 Detection threshold, default is 0.5.
7883 Number of mipmaps, default is 3.
7885 @item xmin, ymin, xmax, ymax
7886 Specifies the rectangle in which to search.
7889 @subsection Examples
7893 Generate a representative palette of a given video using @command{ffmpeg}:
7895 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7901 Cover a rectangular object
7903 It accepts the following options:
7907 Filepath of the optional cover image, needs to be in yuv420.
7912 It accepts the following values:
7915 cover it by the supplied image
7917 cover it by interpolating the surrounding pixels
7920 Default value is @var{blur}.
7923 @subsection Examples
7927 Generate a representative palette of a given video using @command{ffmpeg}:
7929 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7936 Convert the input video to one of the specified pixel formats.
7937 Libavfilter will try to pick one that is suitable as input to
7940 It accepts the following parameters:
7944 A '|'-separated list of pixel format names, such as
7945 "pix_fmts=yuv420p|monow|rgb24".
7949 @subsection Examples
7953 Convert the input video to the @var{yuv420p} format
7955 format=pix_fmts=yuv420p
7958 Convert the input video to any of the formats in the list
7960 format=pix_fmts=yuv420p|yuv444p|yuv410p
7967 Convert the video to specified constant frame rate by duplicating or dropping
7968 frames as necessary.
7970 It accepts the following parameters:
7974 The desired output frame rate. The default is @code{25}.
7979 Possible values are:
7982 zero round towards 0
7986 round towards -infinity
7988 round towards +infinity
7992 The default is @code{near}.
7995 Assume the first PTS should be the given value, in seconds. This allows for
7996 padding/trimming at the start of stream. By default, no assumption is made
7997 about the first frame's expected PTS, so no padding or trimming is done.
7998 For example, this could be set to 0 to pad the beginning with duplicates of
7999 the first frame if a video stream starts after the audio stream or to trim any
8000 frames with a negative PTS.
8004 Alternatively, the options can be specified as a flat string:
8005 @var{fps}[:@var{round}].
8007 See also the @ref{setpts} filter.
8009 @subsection Examples
8013 A typical usage in order to set the fps to 25:
8019 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8021 fps=fps=film:round=near
8027 Pack two different video streams into a stereoscopic video, setting proper
8028 metadata on supported codecs. The two views should have the same size and
8029 framerate and processing will stop when the shorter video ends. Please note
8030 that you may conveniently adjust view properties with the @ref{scale} and
8033 It accepts the following parameters:
8037 The desired packing format. Supported values are:
8042 The views are next to each other (default).
8045 The views are on top of each other.
8048 The views are packed by line.
8051 The views are packed by column.
8054 The views are temporally interleaved.
8063 # Convert left and right views into a frame-sequential video
8064 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8066 # Convert views into a side-by-side video with the same output resolution as the input
8067 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
8072 Change the frame rate by interpolating new video output frames from the source
8075 This filter is not designed to function correctly with interlaced media. If
8076 you wish to change the frame rate of interlaced media then you are required
8077 to deinterlace before this filter and re-interlace after this filter.
8079 A description of the accepted options follows.
8083 Specify the output frames per second. This option can also be specified
8084 as a value alone. The default is @code{50}.
8087 Specify the start of a range where the output frame will be created as a
8088 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8089 the default is @code{15}.
8092 Specify the end of a range where the output frame will be created as a
8093 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8094 the default is @code{240}.
8097 Specify the level at which a scene change is detected as a value between
8098 0 and 100 to indicate a new scene; a low value reflects a low
8099 probability for the current frame to introduce a new scene, while a higher
8100 value means the current frame is more likely to be one.
8101 The default is @code{7}.
8104 Specify flags influencing the filter process.
8106 Available value for @var{flags} is:
8109 @item scene_change_detect, scd
8110 Enable scene change detection using the value of the option @var{scene}.
8111 This flag is enabled by default.
8117 Select one frame every N-th frame.
8119 This filter accepts the following option:
8122 Select frame after every @code{step} frames.
8123 Allowed values are positive integers higher than 0. Default value is @code{1}.
8129 Apply a frei0r effect to the input video.
8131 To enable the compilation of this filter, you need to install the frei0r
8132 header and configure FFmpeg with @code{--enable-frei0r}.
8134 It accepts the following parameters:
8139 The name of the frei0r effect to load. If the environment variable
8140 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8141 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8142 Otherwise, the standard frei0r paths are searched, in this order:
8143 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8144 @file{/usr/lib/frei0r-1/}.
8147 A '|'-separated list of parameters to pass to the frei0r effect.
8151 A frei0r effect parameter can be a boolean (its value is either
8152 "y" or "n"), a double, a color (specified as
8153 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8154 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8155 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8156 @var{X} and @var{Y} are floating point numbers) and/or a string.
8158 The number and types of parameters depend on the loaded effect. If an
8159 effect parameter is not specified, the default value is set.
8161 @subsection Examples
8165 Apply the distort0r effect, setting the first two double parameters:
8167 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8171 Apply the colordistance effect, taking a color as the first parameter:
8173 frei0r=colordistance:0.2/0.3/0.4
8174 frei0r=colordistance:violet
8175 frei0r=colordistance:0x112233
8179 Apply the perspective effect, specifying the top left and top right image
8182 frei0r=perspective:0.2/0.2|0.8/0.2
8186 For more information, see
8187 @url{http://frei0r.dyne.org}
8191 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8193 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8194 processing filter, one of them is performed once per block, not per pixel.
8195 This allows for much higher speed.
8197 The filter accepts the following options:
8201 Set quality. This option defines the number of levels for averaging. It accepts
8202 an integer in the range 4-5. Default value is @code{4}.
8205 Force a constant quantization parameter. It accepts an integer in range 0-63.
8206 If not set, the filter will use the QP from the video stream (if available).
8209 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8210 more details but also more artifacts, while higher values make the image smoother
8211 but also blurrier. Default value is @code{0} − PSNR optimal.
8214 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8215 option may cause flicker since the B-Frames have often larger QP. Default is
8216 @code{0} (not enabled).
8222 The filter accepts the following options:
8226 Set the luminance expression.
8228 Set the chrominance blue expression.
8230 Set the chrominance red expression.
8232 Set the alpha expression.
8234 Set the red expression.
8236 Set the green expression.
8238 Set the blue expression.
8241 The colorspace is selected according to the specified options. If one
8242 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8243 options is specified, the filter will automatically select a YCbCr
8244 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8245 @option{blue_expr} options is specified, it will select an RGB
8248 If one of the chrominance expression is not defined, it falls back on the other
8249 one. If no alpha expression is specified it will evaluate to opaque value.
8250 If none of chrominance expressions are specified, they will evaluate
8251 to the luminance expression.
8253 The expressions can use the following variables and functions:
8257 The sequential number of the filtered frame, starting from @code{0}.
8261 The coordinates of the current sample.
8265 The width and height of the image.
8269 Width and height scale depending on the currently filtered plane. It is the
8270 ratio between the corresponding luma plane number of pixels and the current
8271 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8272 @code{0.5,0.5} for chroma planes.
8275 Time of the current frame, expressed in seconds.
8278 Return the value of the pixel at location (@var{x},@var{y}) of the current
8282 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8286 Return the value of the pixel at location (@var{x},@var{y}) of the
8287 blue-difference chroma plane. Return 0 if there is no such plane.
8290 Return the value of the pixel at location (@var{x},@var{y}) of the
8291 red-difference chroma plane. Return 0 if there is no such plane.
8296 Return the value of the pixel at location (@var{x},@var{y}) of the
8297 red/green/blue component. Return 0 if there is no such component.
8300 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8301 plane. Return 0 if there is no such plane.
8304 For functions, if @var{x} and @var{y} are outside the area, the value will be
8305 automatically clipped to the closer edge.
8307 @subsection Examples
8311 Flip the image horizontally:
8317 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8318 wavelength of 100 pixels:
8320 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8324 Generate a fancy enigmatic moving light:
8326 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
8330 Generate a quick emboss effect:
8332 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8336 Modify RGB components depending on pixel position:
8338 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8342 Create a radial gradient that is the same size as the input (also see
8343 the @ref{vignette} filter):
8345 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8351 Fix the banding artifacts that are sometimes introduced into nearly flat
8352 regions by truncation to 8-bit color depth.
8353 Interpolate the gradients that should go where the bands are, and
8356 It is designed for playback only. Do not use it prior to
8357 lossy compression, because compression tends to lose the dither and
8358 bring back the bands.
8360 It accepts the following parameters:
8365 The maximum amount by which the filter will change any one pixel. This is also
8366 the threshold for detecting nearly flat regions. Acceptable values range from
8367 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8371 The neighborhood to fit the gradient to. A larger radius makes for smoother
8372 gradients, but also prevents the filter from modifying the pixels near detailed
8373 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8374 values will be clipped to the valid range.
8378 Alternatively, the options can be specified as a flat string:
8379 @var{strength}[:@var{radius}]
8381 @subsection Examples
8385 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8391 Specify radius, omitting the strength (which will fall-back to the default
8402 Apply a Hald CLUT to a video stream.
8404 First input is the video stream to process, and second one is the Hald CLUT.
8405 The Hald CLUT input can be a simple picture or a complete video stream.
8407 The filter accepts the following options:
8411 Force termination when the shortest input terminates. Default is @code{0}.
8413 Continue applying the last CLUT after the end of the stream. A value of
8414 @code{0} disable the filter after the last frame of the CLUT is reached.
8415 Default is @code{1}.
8418 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8419 filters share the same internals).
8421 More information about the Hald CLUT can be found on Eskil Steenberg's website
8422 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8424 @subsection Workflow examples
8426 @subsubsection Hald CLUT video stream
8428 Generate an identity Hald CLUT stream altered with various effects:
8430 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
8433 Note: make sure you use a lossless codec.
8435 Then use it with @code{haldclut} to apply it on some random stream:
8437 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8440 The Hald CLUT will be applied to the 10 first seconds (duration of
8441 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8442 to the remaining frames of the @code{mandelbrot} stream.
8444 @subsubsection Hald CLUT with preview
8446 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8447 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8448 biggest possible square starting at the top left of the picture. The remaining
8449 padding pixels (bottom or right) will be ignored. This area can be used to add
8450 a preview of the Hald CLUT.
8452 Typically, the following generated Hald CLUT will be supported by the
8453 @code{haldclut} filter:
8456 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8457 pad=iw+320 [padded_clut];
8458 smptebars=s=320x256, split [a][b];
8459 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8460 [main][b] overlay=W-320" -frames:v 1 clut.png
8463 It contains the original and a preview of the effect of the CLUT: SMPTE color
8464 bars are displayed on the right-top, and below the same color bars processed by
8467 Then, the effect of this Hald CLUT can be visualized with:
8469 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8474 Flip the input video horizontally.
8476 For example, to horizontally flip the input video with @command{ffmpeg}:
8478 ffmpeg -i in.avi -vf "hflip" out.avi
8482 This filter applies a global color histogram equalization on a
8485 It can be used to correct video that has a compressed range of pixel
8486 intensities. The filter redistributes the pixel intensities to
8487 equalize their distribution across the intensity range. It may be
8488 viewed as an "automatically adjusting contrast filter". This filter is
8489 useful only for correcting degraded or poorly captured source
8492 The filter accepts the following options:
8496 Determine the amount of equalization to be applied. As the strength
8497 is reduced, the distribution of pixel intensities more-and-more
8498 approaches that of the input frame. The value must be a float number
8499 in the range [0,1] and defaults to 0.200.
8502 Set the maximum intensity that can generated and scale the output
8503 values appropriately. The strength should be set as desired and then
8504 the intensity can be limited if needed to avoid washing-out. The value
8505 must be a float number in the range [0,1] and defaults to 0.210.
8508 Set the antibanding level. If enabled the filter will randomly vary
8509 the luminance of output pixels by a small amount to avoid banding of
8510 the histogram. Possible values are @code{none}, @code{weak} or
8511 @code{strong}. It defaults to @code{none}.
8516 Compute and draw a color distribution histogram for the input video.
8518 The computed histogram is a representation of the color component
8519 distribution in an image.
8521 Standard histogram displays the color components distribution in an image.
8522 Displays color graph for each color component. Shows distribution of
8523 the Y, U, V, A or R, G, B components, depending on input format, in the
8524 current frame. Below each graph a color component scale meter is shown.
8526 The filter accepts the following options:
8530 Set height of level. Default value is @code{200}.
8531 Allowed range is [50, 2048].
8534 Set height of color scale. Default value is @code{12}.
8535 Allowed range is [0, 40].
8539 It accepts the following values:
8542 Per color component graphs are placed below each other.
8545 Presents information identical to that in the @code{parade}, except
8546 that the graphs representing color components are superimposed directly
8549 Default is @code{parade}.
8552 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8553 Default is @code{linear}.
8556 Set what color components to display.
8557 Default is @code{7}.
8560 @subsection Examples
8565 Calculate and draw histogram:
8567 ffplay -i input -vf histogram
8575 This is a high precision/quality 3d denoise filter. It aims to reduce
8576 image noise, producing smooth images and making still images really
8577 still. It should enhance compressibility.
8579 It accepts the following optional parameters:
8583 A non-negative floating point number which specifies spatial luma strength.
8586 @item chroma_spatial
8587 A non-negative floating point number which specifies spatial chroma strength.
8588 It defaults to 3.0*@var{luma_spatial}/4.0.
8591 A floating point number which specifies luma temporal strength. It defaults to
8592 6.0*@var{luma_spatial}/4.0.
8595 A floating point number which specifies chroma temporal strength. It defaults to
8596 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8599 @anchor{hwupload_cuda}
8600 @section hwupload_cuda
8602 Upload system memory frames to a CUDA device.
8604 It accepts the following optional parameters:
8608 The number of the CUDA device to use
8613 Apply a high-quality magnification filter designed for pixel art. This filter
8614 was originally created by Maxim Stepin.
8616 It accepts the following option:
8620 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8621 @code{hq3x} and @code{4} for @code{hq4x}.
8622 Default is @code{3}.
8626 Stack input videos horizontally.
8628 All streams must be of same pixel format and of same height.
8630 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8631 to create same output.
8633 The filter accept the following option:
8637 Set number of input streams. Default is 2.
8640 If set to 1, force the output to terminate when the shortest input
8641 terminates. Default value is 0.
8646 Modify the hue and/or the saturation of the input.
8648 It accepts the following parameters:
8652 Specify the hue angle as a number of degrees. It accepts an expression,
8653 and defaults to "0".
8656 Specify the saturation in the [-10,10] range. It accepts an expression and
8660 Specify the hue angle as a number of radians. It accepts an
8661 expression, and defaults to "0".
8664 Specify the brightness in the [-10,10] range. It accepts an expression and
8668 @option{h} and @option{H} are mutually exclusive, and can't be
8669 specified at the same time.
8671 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8672 expressions containing the following constants:
8676 frame count of the input frame starting from 0
8679 presentation timestamp of the input frame expressed in time base units
8682 frame rate of the input video, NAN if the input frame rate is unknown
8685 timestamp expressed in seconds, NAN if the input timestamp is unknown
8688 time base of the input video
8691 @subsection Examples
8695 Set the hue to 90 degrees and the saturation to 1.0:
8701 Same command but expressing the hue in radians:
8707 Rotate hue and make the saturation swing between 0
8708 and 2 over a period of 1 second:
8710 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8714 Apply a 3 seconds saturation fade-in effect starting at 0:
8719 The general fade-in expression can be written as:
8721 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8725 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8727 hue="s=max(0\, min(1\, (8-t)/3))"
8730 The general fade-out expression can be written as:
8732 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8737 @subsection Commands
8739 This filter supports the following commands:
8745 Modify the hue and/or the saturation and/or brightness of the input video.
8746 The command accepts the same syntax of the corresponding option.
8748 If the specified expression is not valid, it is kept at its current
8754 Detect video interlacing type.
8756 This filter tries to detect if the input frames as interlaced, progressive,
8757 top or bottom field first. It will also try and detect fields that are
8758 repeated between adjacent frames (a sign of telecine).
8760 Single frame detection considers only immediately adjacent frames when classifying each frame.
8761 Multiple frame detection incorporates the classification history of previous frames.
8763 The filter will log these metadata values:
8766 @item single.current_frame
8767 Detected type of current frame using single-frame detection. One of:
8768 ``tff'' (top field first), ``bff'' (bottom field first),
8769 ``progressive'', or ``undetermined''
8772 Cumulative number of frames detected as top field first using single-frame detection.
8775 Cumulative number of frames detected as top field first using multiple-frame detection.
8778 Cumulative number of frames detected as bottom field first using single-frame detection.
8780 @item multiple.current_frame
8781 Detected type of current frame using multiple-frame detection. One of:
8782 ``tff'' (top field first), ``bff'' (bottom field first),
8783 ``progressive'', or ``undetermined''
8786 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8788 @item single.progressive
8789 Cumulative number of frames detected as progressive using single-frame detection.
8791 @item multiple.progressive
8792 Cumulative number of frames detected as progressive using multiple-frame detection.
8794 @item single.undetermined
8795 Cumulative number of frames that could not be classified using single-frame detection.
8797 @item multiple.undetermined
8798 Cumulative number of frames that could not be classified using multiple-frame detection.
8800 @item repeated.current_frame
8801 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8803 @item repeated.neither
8804 Cumulative number of frames with no repeated field.
8807 Cumulative number of frames with the top field repeated from the previous frame's top field.
8809 @item repeated.bottom
8810 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8813 The filter accepts the following options:
8817 Set interlacing threshold.
8819 Set progressive threshold.
8821 Threshold for repeated field detection.
8823 Number of frames after which a given frame's contribution to the
8824 statistics is halved (i.e., it contributes only 0.5 to it's
8825 classification). The default of 0 means that all frames seen are given
8826 full weight of 1.0 forever.
8827 @item analyze_interlaced_flag
8828 When this is not 0 then idet will use the specified number of frames to determine
8829 if the interlaced flag is accurate, it will not count undetermined frames.
8830 If the flag is found to be accurate it will be used without any further
8831 computations, if it is found to be inaccurate it will be cleared without any
8832 further computations. This allows inserting the idet filter as a low computational
8833 method to clean up the interlaced flag
8838 Deinterleave or interleave fields.
8840 This filter allows one to process interlaced images fields without
8841 deinterlacing them. Deinterleaving splits the input frame into 2
8842 fields (so called half pictures). Odd lines are moved to the top
8843 half of the output image, even lines to the bottom half.
8844 You can process (filter) them independently and then re-interleave them.
8846 The filter accepts the following options:
8850 @item chroma_mode, c
8852 Available values for @var{luma_mode}, @var{chroma_mode} and
8853 @var{alpha_mode} are:
8859 @item deinterleave, d
8860 Deinterleave fields, placing one above the other.
8863 Interleave fields. Reverse the effect of deinterleaving.
8865 Default value is @code{none}.
8868 @item chroma_swap, cs
8869 @item alpha_swap, as
8870 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
8875 Apply inflate effect to the video.
8877 This filter replaces the pixel by the local(3x3) average by taking into account
8878 only values higher than the pixel.
8880 It accepts the following options:
8887 Limit the maximum change for each plane, default is 65535.
8888 If 0, plane will remain unchanged.
8893 Simple interlacing filter from progressive contents. This interleaves upper (or
8894 lower) lines from odd frames with lower (or upper) lines from even frames,
8895 halving the frame rate and preserving image height.
8898 Original Original New Frame
8899 Frame 'j' Frame 'j+1' (tff)
8900 ========== =========== ==================
8901 Line 0 --------------------> Frame 'j' Line 0
8902 Line 1 Line 1 ----> Frame 'j+1' Line 1
8903 Line 2 ---------------------> Frame 'j' Line 2
8904 Line 3 Line 3 ----> Frame 'j+1' Line 3
8906 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
8909 It accepts the following optional parameters:
8913 This determines whether the interlaced frame is taken from the even
8914 (tff - default) or odd (bff) lines of the progressive frame.
8917 Enable (default) or disable the vertical lowpass filter to avoid twitter
8918 interlacing and reduce moire patterns.
8923 Deinterlace input video by applying Donald Graft's adaptive kernel
8924 deinterling. Work on interlaced parts of a video to produce
8927 The description of the accepted parameters follows.
8931 Set the threshold which affects the filter's tolerance when
8932 determining if a pixel line must be processed. It must be an integer
8933 in the range [0,255] and defaults to 10. A value of 0 will result in
8934 applying the process on every pixels.
8937 Paint pixels exceeding the threshold value to white if set to 1.
8941 Set the fields order. Swap fields if set to 1, leave fields alone if
8945 Enable additional sharpening if set to 1. Default is 0.
8948 Enable twoway sharpening if set to 1. Default is 0.
8951 @subsection Examples
8955 Apply default values:
8957 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
8961 Enable additional sharpening:
8967 Paint processed pixels in white:
8973 @section lenscorrection
8975 Correct radial lens distortion
8977 This filter can be used to correct for radial distortion as can result from the use
8978 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
8979 one can use tools available for example as part of opencv or simply trial-and-error.
8980 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
8981 and extract the k1 and k2 coefficients from the resulting matrix.
8983 Note that effectively the same filter is available in the open-source tools Krita and
8984 Digikam from the KDE project.
8986 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
8987 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
8988 brightness distribution, so you may want to use both filters together in certain
8989 cases, though you will have to take care of ordering, i.e. whether vignetting should
8990 be applied before or after lens correction.
8994 The filter accepts the following options:
8998 Relative x-coordinate of the focal point of the image, and thereby the center of the
8999 distortion. This value has a range [0,1] and is expressed as fractions of the image
9002 Relative y-coordinate of the focal point of the image, and thereby the center of the
9003 distortion. This value has a range [0,1] and is expressed as fractions of the image
9006 Coefficient of the quadratic correction term. 0.5 means no correction.
9008 Coefficient of the double quadratic correction term. 0.5 means no correction.
9011 The formula that generates the correction is:
9013 @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)
9015 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9016 distances from the focal point in the source and target images, respectively.
9022 The filter accepts the following options:
9026 Set the number of loops.
9029 Set maximal size in number of frames.
9032 Set first frame of loop.
9038 Apply a 3D LUT to an input video.
9040 The filter accepts the following options:
9044 Set the 3D LUT file name.
9046 Currently supported formats:
9058 Select interpolation mode.
9060 Available values are:
9064 Use values from the nearest defined point.
9066 Interpolate values using the 8 points defining a cube.
9068 Interpolate values using a tetrahedron.
9072 @section lut, lutrgb, lutyuv
9074 Compute a look-up table for binding each pixel component input value
9075 to an output value, and apply it to the input video.
9077 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9078 to an RGB input video.
9080 These filters accept the following parameters:
9083 set first pixel component expression
9085 set second pixel component expression
9087 set third pixel component expression
9089 set fourth pixel component expression, corresponds to the alpha component
9092 set red component expression
9094 set green component expression
9096 set blue component expression
9098 alpha component expression
9101 set Y/luminance component expression
9103 set U/Cb component expression
9105 set V/Cr component expression
9108 Each of them specifies the expression to use for computing the lookup table for
9109 the corresponding pixel component values.
9111 The exact component associated to each of the @var{c*} options depends on the
9114 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9115 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9117 The expressions can contain the following constants and functions:
9122 The input width and height.
9125 The input value for the pixel component.
9128 The input value, clipped to the @var{minval}-@var{maxval} range.
9131 The maximum value for the pixel component.
9134 The minimum value for the pixel component.
9137 The negated value for the pixel component value, clipped to the
9138 @var{minval}-@var{maxval} range; it corresponds to the expression
9139 "maxval-clipval+minval".
9142 The computed value in @var{val}, clipped to the
9143 @var{minval}-@var{maxval} range.
9145 @item gammaval(gamma)
9146 The computed gamma correction value of the pixel component value,
9147 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9149 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9153 All expressions default to "val".
9155 @subsection Examples
9161 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9162 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9165 The above is the same as:
9167 lutrgb="r=negval:g=negval:b=negval"
9168 lutyuv="y=negval:u=negval:v=negval"
9178 Remove chroma components, turning the video into a graytone image:
9180 lutyuv="u=128:v=128"
9184 Apply a luma burning effect:
9190 Remove green and blue components:
9196 Set a constant alpha channel value on input:
9198 format=rgba,lutrgb=a="maxval-minval/2"
9202 Correct luminance gamma by a factor of 0.5:
9204 lutyuv=y=gammaval(0.5)
9208 Discard least significant bits of luma:
9210 lutyuv=y='bitand(val, 128+64+32)'
9214 Technicolor like effect:
9216 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9220 @section maskedmerge
9222 Merge the first input stream with the second input stream using per pixel
9223 weights in the third input stream.
9225 A value of 0 in the third stream pixel component means that pixel component
9226 from first stream is returned unchanged, while maximum value (eg. 255 for
9227 8-bit videos) means that pixel component from second stream is returned
9228 unchanged. Intermediate values define the amount of merging between both
9229 input stream's pixel components.
9231 This filter accepts the following options:
9234 Set which planes will be processed as bitmap, unprocessed planes will be
9235 copied from first stream.
9236 By default value 0xf, all planes will be processed.
9241 Apply motion-compensation deinterlacing.
9243 It needs one field per frame as input and must thus be used together
9244 with yadif=1/3 or equivalent.
9246 This filter accepts the following options:
9249 Set the deinterlacing mode.
9251 It accepts one of the following values:
9256 use iterative motion estimation
9258 like @samp{slow}, but use multiple reference frames.
9260 Default value is @samp{fast}.
9263 Set the picture field parity assumed for the input video. It must be
9264 one of the following values:
9268 assume top field first
9270 assume bottom field first
9273 Default value is @samp{bff}.
9276 Set per-block quantization parameter (QP) used by the internal
9279 Higher values should result in a smoother motion vector field but less
9280 optimal individual vectors. Default value is 1.
9283 @section mergeplanes
9285 Merge color channel components from several video streams.
9287 The filter accepts up to 4 input streams, and merge selected input
9288 planes to the output video.
9290 This filter accepts the following options:
9293 Set input to output plane mapping. Default is @code{0}.
9295 The mappings is specified as a bitmap. It should be specified as a
9296 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9297 mapping for the first plane of the output stream. 'A' sets the number of
9298 the input stream to use (from 0 to 3), and 'a' the plane number of the
9299 corresponding input to use (from 0 to 3). The rest of the mappings is
9300 similar, 'Bb' describes the mapping for the output stream second
9301 plane, 'Cc' describes the mapping for the output stream third plane and
9302 'Dd' describes the mapping for the output stream fourth plane.
9305 Set output pixel format. Default is @code{yuva444p}.
9308 @subsection Examples
9312 Merge three gray video streams of same width and height into single video stream:
9314 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9318 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9320 [a0][a1]mergeplanes=0x00010210:yuva444p
9324 Swap Y and A plane in yuva444p stream:
9326 format=yuva444p,mergeplanes=0x03010200:yuva444p
9330 Swap U and V plane in yuv420p stream:
9332 format=yuv420p,mergeplanes=0x000201:yuv420p
9336 Cast a rgb24 clip to yuv444p:
9338 format=rgb24,mergeplanes=0x000102:yuv444p
9344 Drop frames that do not differ greatly from the previous frame in
9345 order to reduce frame rate.
9347 The main use of this filter is for very-low-bitrate encoding
9348 (e.g. streaming over dialup modem), but it could in theory be used for
9349 fixing movies that were inverse-telecined incorrectly.
9351 A description of the accepted options follows.
9355 Set the maximum number of consecutive frames which can be dropped (if
9356 positive), or the minimum interval between dropped frames (if
9357 negative). If the value is 0, the frame is dropped unregarding the
9358 number of previous sequentially dropped frames.
9365 Set the dropping threshold values.
9367 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9368 represent actual pixel value differences, so a threshold of 64
9369 corresponds to 1 unit of difference for each pixel, or the same spread
9370 out differently over the block.
9372 A frame is a candidate for dropping if no 8x8 blocks differ by more
9373 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9374 meaning the whole image) differ by more than a threshold of @option{lo}.
9376 Default value for @option{hi} is 64*12, default value for @option{lo} is
9377 64*5, and default value for @option{frac} is 0.33.
9385 It accepts an integer in input; if non-zero it negates the
9386 alpha component (if available). The default value in input is 0.
9390 Deinterlace video using neural network edge directed interpolation.
9392 This filter accepts the following options:
9396 Mandatory option, without binary file filter can not work.
9397 Currently file can be found here:
9398 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9401 Set which frames to deinterlace, by default it is @code{all}.
9402 Can be @code{all} or @code{interlaced}.
9405 Set mode of operation.
9407 Can be one of the following:
9411 Use frame flags, both fields.
9413 Use frame flags, single field.
9417 Use bottom field only.
9419 Use both fields, top first.
9421 Use both fields, bottom first.
9425 Set which planes to process, by default filter process all frames.
9428 Set size of local neighborhood around each pixel, used by the predictor neural
9431 Can be one of the following:
9444 Set the number of neurons in predicctor neural network.
9445 Can be one of the following:
9456 Controls the number of different neural network predictions that are blended
9457 together to compute the final output value. Can be @code{fast}, default or
9461 Set which set of weights to use in the predictor.
9462 Can be one of the following:
9466 weights trained to minimize absolute error
9468 weights trained to minimize squared error
9472 Controls whether or not the prescreener neural network is used to decide
9473 which pixels should be processed by the predictor neural network and which
9474 can be handled by simple cubic interpolation.
9475 The prescreener is trained to know whether cubic interpolation will be
9476 sufficient for a pixel or whether it should be predicted by the predictor nn.
9477 The computational complexity of the prescreener nn is much less than that of
9478 the predictor nn. Since most pixels can be handled by cubic interpolation,
9479 using the prescreener generally results in much faster processing.
9480 The prescreener is pretty accurate, so the difference between using it and not
9481 using it is almost always unnoticeable.
9483 Can be one of the following:
9491 Default is @code{new}.
9494 Set various debugging flags.
9499 Force libavfilter not to use any of the specified pixel formats for the
9500 input to the next filter.
9502 It accepts the following parameters:
9506 A '|'-separated list of pixel format names, such as
9507 apix_fmts=yuv420p|monow|rgb24".
9511 @subsection Examples
9515 Force libavfilter to use a format different from @var{yuv420p} for the
9516 input to the vflip filter:
9518 noformat=pix_fmts=yuv420p,vflip
9522 Convert the input video to any of the formats not contained in the list:
9524 noformat=yuv420p|yuv444p|yuv410p
9530 Add noise on video input frame.
9532 The filter accepts the following options:
9540 Set noise seed for specific pixel component or all pixel components in case
9541 of @var{all_seed}. Default value is @code{123457}.
9543 @item all_strength, alls
9544 @item c0_strength, c0s
9545 @item c1_strength, c1s
9546 @item c2_strength, c2s
9547 @item c3_strength, c3s
9548 Set noise strength for specific pixel component or all pixel components in case
9549 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9551 @item all_flags, allf
9556 Set pixel component flags or set flags for all components if @var{all_flags}.
9557 Available values for component flags are:
9560 averaged temporal noise (smoother)
9562 mix random noise with a (semi)regular pattern
9564 temporal noise (noise pattern changes between frames)
9566 uniform noise (gaussian otherwise)
9570 @subsection Examples
9572 Add temporal and uniform noise to input video:
9574 noise=alls=20:allf=t+u
9579 Pass the video source unchanged to the output.
9582 Optical Character Recognition
9584 This filter uses Tesseract for optical character recognition.
9586 It accepts the following options:
9590 Set datapath to tesseract data. Default is to use whatever was
9591 set at installation.
9594 Set language, default is "eng".
9597 Set character whitelist.
9600 Set character blacklist.
9603 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9607 Apply a video transform using libopencv.
9609 To enable this filter, install the libopencv library and headers and
9610 configure FFmpeg with @code{--enable-libopencv}.
9612 It accepts the following parameters:
9617 The name of the libopencv filter to apply.
9620 The parameters to pass to the libopencv filter. If not specified, the default
9625 Refer to the official libopencv documentation for more precise
9627 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9629 Several libopencv filters are supported; see the following subsections.
9634 Dilate an image by using a specific structuring element.
9635 It corresponds to the libopencv function @code{cvDilate}.
9637 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
9639 @var{struct_el} represents a structuring element, and has the syntax:
9640 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
9642 @var{cols} and @var{rows} represent the number of columns and rows of
9643 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
9644 point, and @var{shape} the shape for the structuring element. @var{shape}
9645 must be "rect", "cross", "ellipse", or "custom".
9647 If the value for @var{shape} is "custom", it must be followed by a
9648 string of the form "=@var{filename}". The file with name
9649 @var{filename} is assumed to represent a binary image, with each
9650 printable character corresponding to a bright pixel. When a custom
9651 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
9652 or columns and rows of the read file are assumed instead.
9654 The default value for @var{struct_el} is "3x3+0x0/rect".
9656 @var{nb_iterations} specifies the number of times the transform is
9657 applied to the image, and defaults to 1.
9661 # Use the default values
9664 # Dilate using a structuring element with a 5x5 cross, iterating two times
9665 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
9667 # Read the shape from the file diamond.shape, iterating two times.
9668 # The file diamond.shape may contain a pattern of characters like this
9674 # The specified columns and rows are ignored
9675 # but the anchor point coordinates are not
9676 ocv=dilate:0x0+2x2/custom=diamond.shape|2
9681 Erode an image by using a specific structuring element.
9682 It corresponds to the libopencv function @code{cvErode}.
9684 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
9685 with the same syntax and semantics as the @ref{dilate} filter.
9689 Smooth the input video.
9691 The filter takes the following parameters:
9692 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
9694 @var{type} is the type of smooth filter to apply, and must be one of
9695 the following values: "blur", "blur_no_scale", "median", "gaussian",
9696 or "bilateral". The default value is "gaussian".
9698 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
9699 depend on the smooth type. @var{param1} and
9700 @var{param2} accept integer positive values or 0. @var{param3} and
9701 @var{param4} accept floating point values.
9703 The default value for @var{param1} is 3. The default value for the
9704 other parameters is 0.
9706 These parameters correspond to the parameters assigned to the
9707 libopencv function @code{cvSmooth}.
9712 Overlay one video on top of another.
9714 It takes two inputs and has one output. The first input is the "main"
9715 video on which the second input is overlaid.
9717 It accepts the following parameters:
9719 A description of the accepted options follows.
9724 Set the expression for the x and y coordinates of the overlaid video
9725 on the main video. Default value is "0" for both expressions. In case
9726 the expression is invalid, it is set to a huge value (meaning that the
9727 overlay will not be displayed within the output visible area).
9730 The action to take when EOF is encountered on the secondary input; it accepts
9731 one of the following values:
9735 Repeat the last frame (the default).
9739 Pass the main input through.
9743 Set when the expressions for @option{x}, and @option{y} are evaluated.
9745 It accepts the following values:
9748 only evaluate expressions once during the filter initialization or
9749 when a command is processed
9752 evaluate expressions for each incoming frame
9755 Default value is @samp{frame}.
9758 If set to 1, force the output to terminate when the shortest input
9759 terminates. Default value is 0.
9762 Set the format for the output video.
9764 It accepts the following values:
9779 Default value is @samp{yuv420}.
9781 @item rgb @emph{(deprecated)}
9782 If set to 1, force the filter to accept inputs in the RGB
9783 color space. Default value is 0. This option is deprecated, use
9784 @option{format} instead.
9787 If set to 1, force the filter to draw the last overlay frame over the
9788 main input until the end of the stream. A value of 0 disables this
9789 behavior. Default value is 1.
9792 The @option{x}, and @option{y} expressions can contain the following
9798 The main input width and height.
9802 The overlay input width and height.
9806 The computed values for @var{x} and @var{y}. They are evaluated for
9811 horizontal and vertical chroma subsample values of the output
9812 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
9816 the number of input frame, starting from 0
9819 the position in the file of the input frame, NAN if unknown
9822 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
9826 Note that the @var{n}, @var{pos}, @var{t} variables are available only
9827 when evaluation is done @emph{per frame}, and will evaluate to NAN
9828 when @option{eval} is set to @samp{init}.
9830 Be aware that frames are taken from each input video in timestamp
9831 order, hence, if their initial timestamps differ, it is a good idea
9832 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
9833 have them begin in the same zero timestamp, as the example for
9834 the @var{movie} filter does.
9836 You can chain together more overlays but you should test the
9837 efficiency of such approach.
9839 @subsection Commands
9841 This filter supports the following commands:
9845 Modify the x and y of the overlay input.
9846 The command accepts the same syntax of the corresponding option.
9848 If the specified expression is not valid, it is kept at its current
9852 @subsection Examples
9856 Draw the overlay at 10 pixels from the bottom right corner of the main
9859 overlay=main_w-overlay_w-10:main_h-overlay_h-10
9862 Using named options the example above becomes:
9864 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
9868 Insert a transparent PNG logo in the bottom left corner of the input,
9869 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
9871 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
9875 Insert 2 different transparent PNG logos (second logo on bottom
9876 right corner) using the @command{ffmpeg} tool:
9878 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
9882 Add a transparent color layer on top of the main video; @code{WxH}
9883 must specify the size of the main input to the overlay filter:
9885 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
9889 Play an original video and a filtered version (here with the deshake
9890 filter) side by side using the @command{ffplay} tool:
9892 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
9895 The above command is the same as:
9897 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
9901 Make a sliding overlay appearing from the left to the right top part of the
9902 screen starting since time 2:
9904 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
9908 Compose output by putting two input videos side to side:
9910 ffmpeg -i left.avi -i right.avi -filter_complex "
9911 nullsrc=size=200x100 [background];
9912 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
9913 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
9914 [background][left] overlay=shortest=1 [background+left];
9915 [background+left][right] overlay=shortest=1:x=100 [left+right]
9920 Mask 10-20 seconds of a video by applying the delogo filter to a section
9922 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
9923 -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]'
9928 Chain several overlays in cascade:
9930 nullsrc=s=200x200 [bg];
9931 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
9932 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
9933 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
9934 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
9935 [in3] null, [mid2] overlay=100:100 [out0]
9942 Apply Overcomplete Wavelet denoiser.
9944 The filter accepts the following options:
9950 Larger depth values will denoise lower frequency components more, but
9951 slow down filtering.
9953 Must be an int in the range 8-16, default is @code{8}.
9955 @item luma_strength, ls
9958 Must be a double value in the range 0-1000, default is @code{1.0}.
9960 @item chroma_strength, cs
9961 Set chroma strength.
9963 Must be a double value in the range 0-1000, default is @code{1.0}.
9969 Add paddings to the input image, and place the original input at the
9970 provided @var{x}, @var{y} coordinates.
9972 It accepts the following parameters:
9977 Specify an expression for the size of the output image with the
9978 paddings added. If the value for @var{width} or @var{height} is 0, the
9979 corresponding input size is used for the output.
9981 The @var{width} expression can reference the value set by the
9982 @var{height} expression, and vice versa.
9984 The default value of @var{width} and @var{height} is 0.
9988 Specify the offsets to place the input image at within the padded area,
9989 with respect to the top/left border of the output image.
9991 The @var{x} expression can reference the value set by the @var{y}
9992 expression, and vice versa.
9994 The default value of @var{x} and @var{y} is 0.
9997 Specify the color of the padded area. For the syntax of this option,
9998 check the "Color" section in the ffmpeg-utils manual.
10000 The default value of @var{color} is "black".
10003 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10004 options are expressions containing the following constants:
10009 The input video width and height.
10013 These are the same as @var{in_w} and @var{in_h}.
10017 The output width and height (the size of the padded area), as
10018 specified by the @var{width} and @var{height} expressions.
10022 These are the same as @var{out_w} and @var{out_h}.
10026 The x and y offsets as specified by the @var{x} and @var{y}
10027 expressions, or NAN if not yet specified.
10030 same as @var{iw} / @var{ih}
10033 input sample aspect ratio
10036 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10040 The horizontal and vertical chroma subsample values. For example for the
10041 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10044 @subsection Examples
10048 Add paddings with the color "violet" to the input video. The output video
10049 size is 640x480, and the top-left corner of the input video is placed at
10052 pad=640:480:0:40:violet
10055 The example above is equivalent to the following command:
10057 pad=width=640:height=480:x=0:y=40:color=violet
10061 Pad the input to get an output with dimensions increased by 3/2,
10062 and put the input video at the center of the padded area:
10064 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10068 Pad the input to get a squared output with size equal to the maximum
10069 value between the input width and height, and put the input video at
10070 the center of the padded area:
10072 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10076 Pad the input to get a final w/h ratio of 16:9:
10078 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10082 In case of anamorphic video, in order to set the output display aspect
10083 correctly, it is necessary to use @var{sar} in the expression,
10084 according to the relation:
10086 (ih * X / ih) * sar = output_dar
10087 X = output_dar / sar
10090 Thus the previous example needs to be modified to:
10092 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10096 Double the output size and put the input video in the bottom-right
10097 corner of the output padded area:
10099 pad="2*iw:2*ih:ow-iw:oh-ih"
10103 @anchor{palettegen}
10104 @section palettegen
10106 Generate one palette for a whole video stream.
10108 It accepts the following options:
10112 Set the maximum number of colors to quantize in the palette.
10113 Note: the palette will still contain 256 colors; the unused palette entries
10116 @item reserve_transparent
10117 Create a palette of 255 colors maximum and reserve the last one for
10118 transparency. Reserving the transparency color is useful for GIF optimization.
10119 If not set, the maximum of colors in the palette will be 256. You probably want
10120 to disable this option for a standalone image.
10124 Set statistics mode.
10126 It accepts the following values:
10129 Compute full frame histograms.
10131 Compute histograms only for the part that differs from previous frame. This
10132 might be relevant to give more importance to the moving part of your input if
10133 the background is static.
10136 Default value is @var{full}.
10139 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10140 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10141 color quantization of the palette. This information is also visible at
10142 @var{info} logging level.
10144 @subsection Examples
10148 Generate a representative palette of a given video using @command{ffmpeg}:
10150 ffmpeg -i input.mkv -vf palettegen palette.png
10154 @section paletteuse
10156 Use a palette to downsample an input video stream.
10158 The filter takes two inputs: one video stream and a palette. The palette must
10159 be a 256 pixels image.
10161 It accepts the following options:
10165 Select dithering mode. Available algorithms are:
10168 Ordered 8x8 bayer dithering (deterministic)
10170 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10171 Note: this dithering is sometimes considered "wrong" and is included as a
10173 @item floyd_steinberg
10174 Floyd and Steingberg dithering (error diffusion)
10176 Frankie Sierra dithering v2 (error diffusion)
10178 Frankie Sierra dithering v2 "Lite" (error diffusion)
10181 Default is @var{sierra2_4a}.
10184 When @var{bayer} dithering is selected, this option defines the scale of the
10185 pattern (how much the crosshatch pattern is visible). A low value means more
10186 visible pattern for less banding, and higher value means less visible pattern
10187 at the cost of more banding.
10189 The option must be an integer value in the range [0,5]. Default is @var{2}.
10192 If set, define the zone to process
10196 Only the changing rectangle will be reprocessed. This is similar to GIF
10197 cropping/offsetting compression mechanism. This option can be useful for speed
10198 if only a part of the image is changing, and has use cases such as limiting the
10199 scope of the error diffusal @option{dither} to the rectangle that bounds the
10200 moving scene (it leads to more deterministic output if the scene doesn't change
10201 much, and as a result less moving noise and better GIF compression).
10204 Default is @var{none}.
10207 @subsection Examples
10211 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10212 using @command{ffmpeg}:
10214 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10218 @section perspective
10220 Correct perspective of video not recorded perpendicular to the screen.
10222 A description of the accepted parameters follows.
10233 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10234 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10235 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10236 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10237 then the corners of the source will be sent to the specified coordinates.
10239 The expressions can use the following variables:
10244 the width and height of video frame.
10248 Output frame count.
10251 @item interpolation
10252 Set interpolation for perspective correction.
10254 It accepts the following values:
10260 Default value is @samp{linear}.
10263 Set interpretation of coordinate options.
10265 It accepts the following values:
10269 Send point in the source specified by the given coordinates to
10270 the corners of the destination.
10272 @item 1, destination
10274 Send the corners of the source to the point in the destination specified
10275 by the given coordinates.
10277 Default value is @samp{source}.
10281 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10283 It accepts the following values:
10286 only evaluate expressions once during the filter initialization or
10287 when a command is processed
10290 evaluate expressions for each incoming frame
10293 Default value is @samp{init}.
10298 Delay interlaced video by one field time so that the field order changes.
10300 The intended use is to fix PAL movies that have been captured with the
10301 opposite field order to the film-to-video transfer.
10303 A description of the accepted parameters follows.
10309 It accepts the following values:
10312 Capture field order top-first, transfer bottom-first.
10313 Filter will delay the bottom field.
10316 Capture field order bottom-first, transfer top-first.
10317 Filter will delay the top field.
10320 Capture and transfer with the same field order. This mode only exists
10321 for the documentation of the other options to refer to, but if you
10322 actually select it, the filter will faithfully do nothing.
10325 Capture field order determined automatically by field flags, transfer
10327 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10328 basis using field flags. If no field information is available,
10329 then this works just like @samp{u}.
10332 Capture unknown or varying, transfer opposite.
10333 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10334 analyzing the images and selecting the alternative that produces best
10335 match between the fields.
10338 Capture top-first, transfer unknown or varying.
10339 Filter selects among @samp{t} and @samp{p} using image analysis.
10342 Capture bottom-first, transfer unknown or varying.
10343 Filter selects among @samp{b} and @samp{p} using image analysis.
10346 Capture determined by field flags, transfer unknown or varying.
10347 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10348 image analysis. If no field information is available, then this works just
10349 like @samp{U}. This is the default mode.
10352 Both capture and transfer unknown or varying.
10353 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10357 @section pixdesctest
10359 Pixel format descriptor test filter, mainly useful for internal
10360 testing. The output video should be equal to the input video.
10364 format=monow, pixdesctest
10367 can be used to test the monowhite pixel format descriptor definition.
10371 Enable the specified chain of postprocessing subfilters using libpostproc. This
10372 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10373 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10374 Each subfilter and some options have a short and a long name that can be used
10375 interchangeably, i.e. dr/dering are the same.
10377 The filters accept the following options:
10381 Set postprocessing subfilters string.
10384 All subfilters share common options to determine their scope:
10388 Honor the quality commands for this subfilter.
10391 Do chrominance filtering, too (default).
10394 Do luminance filtering only (no chrominance).
10397 Do chrominance filtering only (no luminance).
10400 These options can be appended after the subfilter name, separated by a '|'.
10402 Available subfilters are:
10405 @item hb/hdeblock[|difference[|flatness]]
10406 Horizontal deblocking filter
10409 Difference factor where higher values mean more deblocking (default: @code{32}).
10411 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10414 @item vb/vdeblock[|difference[|flatness]]
10415 Vertical deblocking filter
10418 Difference factor where higher values mean more deblocking (default: @code{32}).
10420 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10423 @item ha/hadeblock[|difference[|flatness]]
10424 Accurate horizontal deblocking filter
10427 Difference factor where higher values mean more deblocking (default: @code{32}).
10429 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10432 @item va/vadeblock[|difference[|flatness]]
10433 Accurate vertical deblocking filter
10436 Difference factor where higher values mean more deblocking (default: @code{32}).
10438 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10442 The horizontal and vertical deblocking filters share the difference and
10443 flatness values so you cannot set different horizontal and vertical
10447 @item h1/x1hdeblock
10448 Experimental horizontal deblocking filter
10450 @item v1/x1vdeblock
10451 Experimental vertical deblocking filter
10456 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10459 larger -> stronger filtering
10461 larger -> stronger filtering
10463 larger -> stronger filtering
10466 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10469 Stretch luminance to @code{0-255}.
10472 @item lb/linblenddeint
10473 Linear blend deinterlacing filter that deinterlaces the given block by
10474 filtering all lines with a @code{(1 2 1)} filter.
10476 @item li/linipoldeint
10477 Linear interpolating deinterlacing filter that deinterlaces the given block by
10478 linearly interpolating every second line.
10480 @item ci/cubicipoldeint
10481 Cubic interpolating deinterlacing filter deinterlaces the given block by
10482 cubically interpolating every second line.
10484 @item md/mediandeint
10485 Median deinterlacing filter that deinterlaces the given block by applying a
10486 median filter to every second line.
10488 @item fd/ffmpegdeint
10489 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10490 second line with a @code{(-1 4 2 4 -1)} filter.
10493 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10494 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10496 @item fq/forceQuant[|quantizer]
10497 Overrides the quantizer table from the input with the constant quantizer you
10505 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10508 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10511 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10514 @subsection Examples
10518 Apply horizontal and vertical deblocking, deringing and automatic
10519 brightness/contrast:
10525 Apply default filters without brightness/contrast correction:
10531 Apply default filters and temporal denoiser:
10533 pp=default/tmpnoise|1|2|3
10537 Apply deblocking on luminance only, and switch vertical deblocking on or off
10538 automatically depending on available CPU time:
10545 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10546 similar to spp = 6 with 7 point DCT, where only the center sample is
10549 The filter accepts the following options:
10553 Force a constant quantization parameter. It accepts an integer in range
10554 0 to 63. If not set, the filter will use the QP from the video stream
10558 Set thresholding mode. Available modes are:
10562 Set hard thresholding.
10564 Set soft thresholding (better de-ringing effect, but likely blurrier).
10566 Set medium thresholding (good results, default).
10572 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10573 Ratio) between two input videos.
10575 This filter takes in input two input videos, the first input is
10576 considered the "main" source and is passed unchanged to the
10577 output. The second input is used as a "reference" video for computing
10580 Both video inputs must have the same resolution and pixel format for
10581 this filter to work correctly. Also it assumes that both inputs
10582 have the same number of frames, which are compared one by one.
10584 The obtained average PSNR is printed through the logging system.
10586 The filter stores the accumulated MSE (mean squared error) of each
10587 frame, and at the end of the processing it is averaged across all frames
10588 equally, and the following formula is applied to obtain the PSNR:
10591 PSNR = 10*log10(MAX^2/MSE)
10594 Where MAX is the average of the maximum values of each component of the
10597 The description of the accepted parameters follows.
10600 @item stats_file, f
10601 If specified the filter will use the named file to save the PSNR of
10602 each individual frame. When filename equals "-" the data is sent to
10605 @item stats_version
10606 Specifies which version of the stats file format to use. Details of
10607 each format are written below.
10608 Default value is 1.
10611 The file printed if @var{stats_file} is selected, contains a sequence of
10612 key/value pairs of the form @var{key}:@var{value} for each compared
10615 If a @var{stats_version} greater than 1 is specified, a header line precedes
10616 the list of per-frame-pair stats, with key value pairs following the frame
10617 format with the following parameters:
10620 @item psnr_log_version
10621 The version of the log file format. Will match @var{stats_version}.
10624 A comma separated list of the per-frame-pair parameters included in
10628 A description of each shown per-frame-pair parameter follows:
10632 sequential number of the input frame, starting from 1
10635 Mean Square Error pixel-by-pixel average difference of the compared
10636 frames, averaged over all the image components.
10638 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
10639 Mean Square Error pixel-by-pixel average difference of the compared
10640 frames for the component specified by the suffix.
10642 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
10643 Peak Signal to Noise ratio of the compared frames for the component
10644 specified by the suffix.
10649 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
10650 [main][ref] psnr="stats_file=stats.log" [out]
10653 On this example the input file being processed is compared with the
10654 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
10655 is stored in @file{stats.log}.
10660 Pulldown reversal (inverse telecine) filter, capable of handling mixed
10661 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
10664 The pullup filter is designed to take advantage of future context in making
10665 its decisions. This filter is stateless in the sense that it does not lock
10666 onto a pattern to follow, but it instead looks forward to the following
10667 fields in order to identify matches and rebuild progressive frames.
10669 To produce content with an even framerate, insert the fps filter after
10670 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
10671 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
10673 The filter accepts the following options:
10680 These options set the amount of "junk" to ignore at the left, right, top, and
10681 bottom of the image, respectively. Left and right are in units of 8 pixels,
10682 while top and bottom are in units of 2 lines.
10683 The default is 8 pixels on each side.
10686 Set the strict breaks. Setting this option to 1 will reduce the chances of
10687 filter generating an occasional mismatched frame, but it may also cause an
10688 excessive number of frames to be dropped during high motion sequences.
10689 Conversely, setting it to -1 will make filter match fields more easily.
10690 This may help processing of video where there is slight blurring between
10691 the fields, but may also cause there to be interlaced frames in the output.
10692 Default value is @code{0}.
10695 Set the metric plane to use. It accepts the following values:
10701 Use chroma blue plane.
10704 Use chroma red plane.
10707 This option may be set to use chroma plane instead of the default luma plane
10708 for doing filter's computations. This may improve accuracy on very clean
10709 source material, but more likely will decrease accuracy, especially if there
10710 is chroma noise (rainbow effect) or any grayscale video.
10711 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
10712 load and make pullup usable in realtime on slow machines.
10715 For best results (without duplicated frames in the output file) it is
10716 necessary to change the output frame rate. For example, to inverse
10717 telecine NTSC input:
10719 ffmpeg -i input -vf pullup -r 24000/1001 ...
10724 Change video quantization parameters (QP).
10726 The filter accepts the following option:
10730 Set expression for quantization parameter.
10733 The expression is evaluated through the eval API and can contain, among others,
10734 the following constants:
10738 1 if index is not 129, 0 otherwise.
10741 Sequentional index starting from -129 to 128.
10744 @subsection Examples
10748 Some equation like:
10756 Flush video frames from internal cache of frames into a random order.
10757 No frame is discarded.
10758 Inspired by @ref{frei0r} nervous filter.
10762 Set size in number of frames of internal cache, in range from @code{2} to
10763 @code{512}. Default is @code{30}.
10766 Set seed for random number generator, must be an integer included between
10767 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
10768 less than @code{0}, the filter will try to use a good random seed on a
10774 Read vertical interval timecode (VITC) information from the top lines of a
10777 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
10778 timecode value, if a valid timecode has been detected. Further metadata key
10779 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
10780 timecode data has been found or not.
10782 This filter accepts the following options:
10786 Set the maximum number of lines to scan for VITC data. If the value is set to
10787 @code{-1} the full video frame is scanned. Default is @code{45}.
10790 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
10791 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
10794 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
10795 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
10798 @subsection Examples
10802 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
10803 draw @code{--:--:--:--} as a placeholder:
10805 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
10811 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
10813 Destination pixel at position (X, Y) will be picked from source (x, y) position
10814 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
10815 value for pixel will be used for destination pixel.
10817 Xmap and Ymap input video streams must be of same dimensions. Output video stream
10818 will have Xmap/Ymap video stream dimensions.
10819 Xmap and Ymap input video streams are 16bit depth, single channel.
10821 @section removegrain
10823 The removegrain filter is a spatial denoiser for progressive video.
10827 Set mode for the first plane.
10830 Set mode for the second plane.
10833 Set mode for the third plane.
10836 Set mode for the fourth plane.
10839 Range of mode is from 0 to 24. Description of each mode follows:
10843 Leave input plane unchanged. Default.
10846 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
10849 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
10852 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
10855 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
10856 This is equivalent to a median filter.
10859 Line-sensitive clipping giving the minimal change.
10862 Line-sensitive clipping, intermediate.
10865 Line-sensitive clipping, intermediate.
10868 Line-sensitive clipping, intermediate.
10871 Line-sensitive clipping on a line where the neighbours pixels are the closest.
10874 Replaces the target pixel with the closest neighbour.
10877 [1 2 1] horizontal and vertical kernel blur.
10883 Bob mode, interpolates top field from the line where the neighbours
10884 pixels are the closest.
10887 Bob mode, interpolates bottom field from the line where the neighbours
10888 pixels are the closest.
10891 Bob mode, interpolates top field. Same as 13 but with a more complicated
10892 interpolation formula.
10895 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
10896 interpolation formula.
10899 Clips the pixel with the minimum and maximum of respectively the maximum and
10900 minimum of each pair of opposite neighbour pixels.
10903 Line-sensitive clipping using opposite neighbours whose greatest distance from
10904 the current pixel is minimal.
10907 Replaces the pixel with the average of its 8 neighbours.
10910 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
10913 Clips pixels using the averages of opposite neighbour.
10916 Same as mode 21 but simpler and faster.
10919 Small edge and halo removal, but reputed useless.
10925 @section removelogo
10927 Suppress a TV station logo, using an image file to determine which
10928 pixels comprise the logo. It works by filling in the pixels that
10929 comprise the logo with neighboring pixels.
10931 The filter accepts the following options:
10935 Set the filter bitmap file, which can be any image format supported by
10936 libavformat. The width and height of the image file must match those of the
10937 video stream being processed.
10940 Pixels in the provided bitmap image with a value of zero are not
10941 considered part of the logo, non-zero pixels are considered part of
10942 the logo. If you use white (255) for the logo and black (0) for the
10943 rest, you will be safe. For making the filter bitmap, it is
10944 recommended to take a screen capture of a black frame with the logo
10945 visible, and then using a threshold filter followed by the erode
10946 filter once or twice.
10948 If needed, little splotches can be fixed manually. Remember that if
10949 logo pixels are not covered, the filter quality will be much
10950 reduced. Marking too many pixels as part of the logo does not hurt as
10951 much, but it will increase the amount of blurring needed to cover over
10952 the image and will destroy more information than necessary, and extra
10953 pixels will slow things down on a large logo.
10955 @section repeatfields
10957 This filter uses the repeat_field flag from the Video ES headers and hard repeats
10958 fields based on its value.
10962 Reverse a video clip.
10964 Warning: This filter requires memory to buffer the entire clip, so trimming
10967 @subsection Examples
10971 Take the first 5 seconds of a clip, and reverse it.
10979 Rotate video by an arbitrary angle expressed in radians.
10981 The filter accepts the following options:
10983 A description of the optional parameters follows.
10986 Set an expression for the angle by which to rotate the input video
10987 clockwise, expressed as a number of radians. A negative value will
10988 result in a counter-clockwise rotation. By default it is set to "0".
10990 This expression is evaluated for each frame.
10993 Set the output width expression, default value is "iw".
10994 This expression is evaluated just once during configuration.
10997 Set the output height expression, default value is "ih".
10998 This expression is evaluated just once during configuration.
11001 Enable bilinear interpolation if set to 1, a value of 0 disables
11002 it. Default value is 1.
11005 Set the color used to fill the output area not covered by the rotated
11006 image. For the general syntax of this option, check the "Color" section in the
11007 ffmpeg-utils manual. If the special value "none" is selected then no
11008 background is printed (useful for example if the background is never shown).
11010 Default value is "black".
11013 The expressions for the angle and the output size can contain the
11014 following constants and functions:
11018 sequential number of the input frame, starting from 0. It is always NAN
11019 before the first frame is filtered.
11022 time in seconds of the input frame, it is set to 0 when the filter is
11023 configured. It is always NAN before the first frame is filtered.
11027 horizontal and vertical chroma subsample values. For example for the
11028 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11032 the input video width and height
11036 the output width and height, that is the size of the padded area as
11037 specified by the @var{width} and @var{height} expressions
11041 the minimal width/height required for completely containing the input
11042 video rotated by @var{a} radians.
11044 These are only available when computing the @option{out_w} and
11045 @option{out_h} expressions.
11048 @subsection Examples
11052 Rotate the input by PI/6 radians clockwise:
11058 Rotate the input by PI/6 radians counter-clockwise:
11064 Rotate the input by 45 degrees clockwise:
11070 Apply a constant rotation with period T, starting from an angle of PI/3:
11072 rotate=PI/3+2*PI*t/T
11076 Make the input video rotation oscillating with a period of T
11077 seconds and an amplitude of A radians:
11079 rotate=A*sin(2*PI/T*t)
11083 Rotate the video, output size is chosen so that the whole rotating
11084 input video is always completely contained in the output:
11086 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11090 Rotate the video, reduce the output size so that no background is ever
11093 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11097 @subsection Commands
11099 The filter supports the following commands:
11103 Set the angle expression.
11104 The command accepts the same syntax of the corresponding option.
11106 If the specified expression is not valid, it is kept at its current
11112 Apply Shape Adaptive Blur.
11114 The filter accepts the following options:
11117 @item luma_radius, lr
11118 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11119 value is 1.0. A greater value will result in a more blurred image, and
11120 in slower processing.
11122 @item luma_pre_filter_radius, lpfr
11123 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11126 @item luma_strength, ls
11127 Set luma maximum difference between pixels to still be considered, must
11128 be a value in the 0.1-100.0 range, default value is 1.0.
11130 @item chroma_radius, cr
11131 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11132 greater value will result in a more blurred image, and in slower
11135 @item chroma_pre_filter_radius, cpfr
11136 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11138 @item chroma_strength, cs
11139 Set chroma maximum difference between pixels to still be considered,
11140 must be a value in the -0.9-100.0 range.
11143 Each chroma option value, if not explicitly specified, is set to the
11144 corresponding luma option value.
11149 Scale (resize) the input video, using the libswscale library.
11151 The scale filter forces the output display aspect ratio to be the same
11152 of the input, by changing the output sample aspect ratio.
11154 If the input image format is different from the format requested by
11155 the next filter, the scale filter will convert the input to the
11158 @subsection Options
11159 The filter accepts the following options, or any of the options
11160 supported by the libswscale scaler.
11162 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11163 the complete list of scaler options.
11168 Set the output video dimension expression. Default value is the input
11171 If the value is 0, the input width is used for the output.
11173 If one of the values is -1, the scale filter will use a value that
11174 maintains the aspect ratio of the input image, calculated from the
11175 other specified dimension. If both of them are -1, the input size is
11178 If one of the values is -n with n > 1, the scale filter will also use a value
11179 that maintains the aspect ratio of the input image, calculated from the other
11180 specified dimension. After that it will, however, make sure that the calculated
11181 dimension is divisible by n and adjust the value if necessary.
11183 See below for the list of accepted constants for use in the dimension
11187 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11191 Only evaluate expressions once during the filter initialization or when a command is processed.
11194 Evaluate expressions for each incoming frame.
11198 Default value is @samp{init}.
11202 Set the interlacing mode. It accepts the following values:
11206 Force interlaced aware scaling.
11209 Do not apply interlaced scaling.
11212 Select interlaced aware scaling depending on whether the source frames
11213 are flagged as interlaced or not.
11216 Default value is @samp{0}.
11219 Set libswscale scaling flags. See
11220 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11221 complete list of values. If not explicitly specified the filter applies
11225 @item param0, param1
11226 Set libswscale input parameters for scaling algorithms that need them. See
11227 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11228 complete documentation. If not explicitly specified the filter applies
11234 Set the video size. For the syntax of this option, check the
11235 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11237 @item in_color_matrix
11238 @item out_color_matrix
11239 Set in/output YCbCr color space type.
11241 This allows the autodetected value to be overridden as well as allows forcing
11242 a specific value used for the output and encoder.
11244 If not specified, the color space type depends on the pixel format.
11250 Choose automatically.
11253 Format conforming to International Telecommunication Union (ITU)
11254 Recommendation BT.709.
11257 Set color space conforming to the United States Federal Communications
11258 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11261 Set color space conforming to:
11265 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11268 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11271 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11276 Set color space conforming to SMPTE ST 240:1999.
11281 Set in/output YCbCr sample range.
11283 This allows the autodetected value to be overridden as well as allows forcing
11284 a specific value used for the output and encoder. If not specified, the
11285 range depends on the pixel format. Possible values:
11289 Choose automatically.
11292 Set full range (0-255 in case of 8-bit luma).
11295 Set "MPEG" range (16-235 in case of 8-bit luma).
11298 @item force_original_aspect_ratio
11299 Enable decreasing or increasing output video width or height if necessary to
11300 keep the original aspect ratio. Possible values:
11304 Scale the video as specified and disable this feature.
11307 The output video dimensions will automatically be decreased if needed.
11310 The output video dimensions will automatically be increased if needed.
11314 One useful instance of this option is that when you know a specific device's
11315 maximum allowed resolution, you can use this to limit the output video to
11316 that, while retaining the aspect ratio. For example, device A allows
11317 1280x720 playback, and your video is 1920x800. Using this option (set it to
11318 decrease) and specifying 1280x720 to the command line makes the output
11321 Please note that this is a different thing than specifying -1 for @option{w}
11322 or @option{h}, you still need to specify the output resolution for this option
11327 The values of the @option{w} and @option{h} options are expressions
11328 containing the following constants:
11333 The input width and height
11337 These are the same as @var{in_w} and @var{in_h}.
11341 The output (scaled) width and height
11345 These are the same as @var{out_w} and @var{out_h}
11348 The same as @var{iw} / @var{ih}
11351 input sample aspect ratio
11354 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11358 horizontal and vertical input chroma subsample values. For example for the
11359 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11363 horizontal and vertical output chroma subsample values. For example for the
11364 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11367 @subsection Examples
11371 Scale the input video to a size of 200x100
11376 This is equivalent to:
11387 Specify a size abbreviation for the output size:
11392 which can also be written as:
11398 Scale the input to 2x:
11400 scale=w=2*iw:h=2*ih
11404 The above is the same as:
11406 scale=2*in_w:2*in_h
11410 Scale the input to 2x with forced interlaced scaling:
11412 scale=2*iw:2*ih:interl=1
11416 Scale the input to half size:
11418 scale=w=iw/2:h=ih/2
11422 Increase the width, and set the height to the same size:
11428 Seek Greek harmony:
11435 Increase the height, and set the width to 3/2 of the height:
11437 scale=w=3/2*oh:h=3/5*ih
11441 Increase the size, making the size a multiple of the chroma
11444 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11448 Increase the width to a maximum of 500 pixels,
11449 keeping the same aspect ratio as the input:
11451 scale=w='min(500\, iw*3/2):h=-1'
11455 @subsection Commands
11457 This filter supports the following commands:
11461 Set the output video dimension expression.
11462 The command accepts the same syntax of the corresponding option.
11464 If the specified expression is not valid, it is kept at its current
11470 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11471 format conversion on CUDA video frames. Setting the output width and height
11472 works in the same way as for the @var{scale} filter.
11474 The following additional options are accepted:
11477 The pixel format of the output CUDA frames. If set to the string "same" (the
11478 default), the input format will be kept. Note that automatic format negotiation
11479 and conversion is not yet supported for hardware frames
11482 The interpolation algorithm used for resizing. One of the following:
11489 @item cubic2p_bspline
11490 2-parameter cubic (B=1, C=0)
11492 @item cubic2p_catmullrom
11493 2-parameter cubic (B=0, C=1/2)
11495 @item cubic2p_b05c03
11496 2-parameter cubic (B=1/2, C=3/10)
11508 Scale (resize) the input video, based on a reference video.
11510 See the scale filter for available options, scale2ref supports the same but
11511 uses the reference video instead of the main input as basis.
11513 @subsection Examples
11517 Scale a subtitle stream to match the main video in size before overlaying
11519 'scale2ref[b][a];[a][b]overlay'
11523 @anchor{selectivecolor}
11524 @section selectivecolor
11526 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11527 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11528 by the "purity" of the color (that is, how saturated it already is).
11530 This filter is similar to the Adobe Photoshop Selective Color tool.
11532 The filter accepts the following options:
11535 @item correction_method
11536 Select color correction method.
11538 Available values are:
11541 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11544 Specified adjustments are relative to the original component value.
11546 Default is @code{absolute}.
11548 Adjustments for red pixels (pixels where the red component is the maximum)
11550 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11552 Adjustments for green pixels (pixels where the green component is the maximum)
11554 Adjustments for cyan pixels (pixels where the red component is the minimum)
11556 Adjustments for blue pixels (pixels where the blue component is the maximum)
11558 Adjustments for magenta pixels (pixels where the green component is the minimum)
11560 Adjustments for white pixels (pixels where all components are greater than 128)
11562 Adjustments for all pixels except pure black and pure white
11564 Adjustments for black pixels (pixels where all components are lesser than 128)
11566 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11569 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11570 4 space separated floating point adjustment values in the [-1,1] range,
11571 respectively to adjust the amount of cyan, magenta, yellow and black for the
11572 pixels of its range.
11574 @subsection Examples
11578 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11579 increase magenta by 27% in blue areas:
11581 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11585 Use a Photoshop selective color preset:
11587 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11591 @section separatefields
11593 The @code{separatefields} takes a frame-based video input and splits
11594 each frame into its components fields, producing a new half height clip
11595 with twice the frame rate and twice the frame count.
11597 This filter use field-dominance information in frame to decide which
11598 of each pair of fields to place first in the output.
11599 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
11601 @section setdar, setsar
11603 The @code{setdar} filter sets the Display Aspect Ratio for the filter
11606 This is done by changing the specified Sample (aka Pixel) Aspect
11607 Ratio, according to the following equation:
11609 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
11612 Keep in mind that the @code{setdar} filter does not modify the pixel
11613 dimensions of the video frame. Also, the display aspect ratio set by
11614 this filter may be changed by later filters in the filterchain,
11615 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
11618 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
11619 the filter output video.
11621 Note that as a consequence of the application of this filter, the
11622 output display aspect ratio will change according to the equation
11625 Keep in mind that the sample aspect ratio set by the @code{setsar}
11626 filter may be changed by later filters in the filterchain, e.g. if
11627 another "setsar" or a "setdar" filter is applied.
11629 It accepts the following parameters:
11632 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
11633 Set the aspect ratio used by the filter.
11635 The parameter can be a floating point number string, an expression, or
11636 a string of the form @var{num}:@var{den}, where @var{num} and
11637 @var{den} are the numerator and denominator of the aspect ratio. If
11638 the parameter is not specified, it is assumed the value "0".
11639 In case the form "@var{num}:@var{den}" is used, the @code{:} character
11643 Set the maximum integer value to use for expressing numerator and
11644 denominator when reducing the expressed aspect ratio to a rational.
11645 Default value is @code{100}.
11649 The parameter @var{sar} is an expression containing
11650 the following constants:
11654 These are approximated values for the mathematical constants e
11655 (Euler's number), pi (Greek pi), and phi (the golden ratio).
11658 The input width and height.
11661 These are the same as @var{w} / @var{h}.
11664 The input sample aspect ratio.
11667 The input display aspect ratio. It is the same as
11668 (@var{w} / @var{h}) * @var{sar}.
11671 Horizontal and vertical chroma subsample values. For example, for the
11672 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11675 @subsection Examples
11680 To change the display aspect ratio to 16:9, specify one of the following:
11687 To change the sample aspect ratio to 10:11, specify:
11693 To set a display aspect ratio of 16:9, and specify a maximum integer value of
11694 1000 in the aspect ratio reduction, use the command:
11696 setdar=ratio=16/9:max=1000
11704 Force field for the output video frame.
11706 The @code{setfield} filter marks the interlace type field for the
11707 output frames. It does not change the input frame, but only sets the
11708 corresponding property, which affects how the frame is treated by
11709 following filters (e.g. @code{fieldorder} or @code{yadif}).
11711 The filter accepts the following options:
11716 Available values are:
11720 Keep the same field property.
11723 Mark the frame as bottom-field-first.
11726 Mark the frame as top-field-first.
11729 Mark the frame as progressive.
11735 Show a line containing various information for each input video frame.
11736 The input video is not modified.
11738 The shown line contains a sequence of key/value pairs of the form
11739 @var{key}:@var{value}.
11741 The following values are shown in the output:
11745 The (sequential) number of the input frame, starting from 0.
11748 The Presentation TimeStamp of the input frame, expressed as a number of
11749 time base units. The time base unit depends on the filter input pad.
11752 The Presentation TimeStamp of the input frame, expressed as a number of
11756 The position of the frame in the input stream, or -1 if this information is
11757 unavailable and/or meaningless (for example in case of synthetic video).
11760 The pixel format name.
11763 The sample aspect ratio of the input frame, expressed in the form
11764 @var{num}/@var{den}.
11767 The size of the input frame. For the syntax of this option, check the
11768 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11771 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
11772 for bottom field first).
11775 This is 1 if the frame is a key frame, 0 otherwise.
11778 The picture type of the input frame ("I" for an I-frame, "P" for a
11779 P-frame, "B" for a B-frame, or "?" for an unknown type).
11780 Also refer to the documentation of the @code{AVPictureType} enum and of
11781 the @code{av_get_picture_type_char} function defined in
11782 @file{libavutil/avutil.h}.
11785 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
11787 @item plane_checksum
11788 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
11789 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
11792 @section showpalette
11794 Displays the 256 colors palette of each frame. This filter is only relevant for
11795 @var{pal8} pixel format frames.
11797 It accepts the following option:
11801 Set the size of the box used to represent one palette color entry. Default is
11802 @code{30} (for a @code{30x30} pixel box).
11805 @section shuffleframes
11807 Reorder and/or duplicate video frames.
11809 It accepts the following parameters:
11813 Set the destination indexes of input frames.
11814 This is space or '|' separated list of indexes that maps input frames to output
11815 frames. Number of indexes also sets maximal value that each index may have.
11818 The first frame has the index 0. The default is to keep the input unchanged.
11820 Swap second and third frame of every three frames of the input:
11822 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
11825 @section shuffleplanes
11827 Reorder and/or duplicate video planes.
11829 It accepts the following parameters:
11834 The index of the input plane to be used as the first output plane.
11837 The index of the input plane to be used as the second output plane.
11840 The index of the input plane to be used as the third output plane.
11843 The index of the input plane to be used as the fourth output plane.
11847 The first plane has the index 0. The default is to keep the input unchanged.
11849 Swap the second and third planes of the input:
11851 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
11854 @anchor{signalstats}
11855 @section signalstats
11856 Evaluate various visual metrics that assist in determining issues associated
11857 with the digitization of analog video media.
11859 By default the filter will log these metadata values:
11863 Display the minimal Y value contained within the input frame. Expressed in
11867 Display the Y value at the 10% percentile within the input frame. Expressed in
11871 Display the average Y value within the input frame. Expressed in range of
11875 Display the Y value at the 90% percentile within the input frame. Expressed in
11879 Display the maximum Y value contained within the input frame. Expressed in
11883 Display the minimal U value contained within the input frame. Expressed in
11887 Display the U value at the 10% percentile within the input frame. Expressed in
11891 Display the average U value within the input frame. Expressed in range of
11895 Display the U value at the 90% percentile within the input frame. Expressed in
11899 Display the maximum U value contained within the input frame. Expressed in
11903 Display the minimal V value contained within the input frame. Expressed in
11907 Display the V value at the 10% percentile within the input frame. Expressed in
11911 Display the average V value within the input frame. Expressed in range of
11915 Display the V value at the 90% percentile within the input frame. Expressed in
11919 Display the maximum V value contained within the input frame. Expressed in
11923 Display the minimal saturation value contained within the input frame.
11924 Expressed in range of [0-~181.02].
11927 Display the saturation value at the 10% percentile within the input frame.
11928 Expressed in range of [0-~181.02].
11931 Display the average saturation value within the input frame. Expressed in range
11935 Display the saturation value at the 90% percentile within the input frame.
11936 Expressed in range of [0-~181.02].
11939 Display the maximum saturation value contained within the input frame.
11940 Expressed in range of [0-~181.02].
11943 Display the median value for hue within the input frame. Expressed in range of
11947 Display the average value for hue within the input frame. Expressed in range of
11951 Display the average of sample value difference between all values of the Y
11952 plane in the current frame and corresponding values of the previous input frame.
11953 Expressed in range of [0-255].
11956 Display the average of sample value difference between all values of the U
11957 plane in the current frame and corresponding values of the previous input frame.
11958 Expressed in range of [0-255].
11961 Display the average of sample value difference between all values of the V
11962 plane in the current frame and corresponding values of the previous input frame.
11963 Expressed in range of [0-255].
11966 Display bit depth of Y plane in current frame.
11967 Expressed in range of [0-16].
11970 Display bit depth of U plane in current frame.
11971 Expressed in range of [0-16].
11974 Display bit depth of V plane in current frame.
11975 Expressed in range of [0-16].
11978 The filter accepts the following options:
11984 @option{stat} specify an additional form of image analysis.
11985 @option{out} output video with the specified type of pixel highlighted.
11987 Both options accept the following values:
11991 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
11992 unlike the neighboring pixels of the same field. Examples of temporal outliers
11993 include the results of video dropouts, head clogs, or tape tracking issues.
11996 Identify @var{vertical line repetition}. Vertical line repetition includes
11997 similar rows of pixels within a frame. In born-digital video vertical line
11998 repetition is common, but this pattern is uncommon in video digitized from an
11999 analog source. When it occurs in video that results from the digitization of an
12000 analog source it can indicate concealment from a dropout compensator.
12003 Identify pixels that fall outside of legal broadcast range.
12007 Set the highlight color for the @option{out} option. The default color is
12011 @subsection Examples
12015 Output data of various video metrics:
12017 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12021 Output specific data about the minimum and maximum values of the Y plane per frame:
12023 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12027 Playback video while highlighting pixels that are outside of broadcast range in red.
12029 ffplay example.mov -vf signalstats="out=brng:color=red"
12033 Playback video with signalstats metadata drawn over the frame.
12035 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12038 The contents of signalstat_drawtext.txt used in the command are:
12041 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12042 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12043 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12044 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12052 Blur the input video without impacting the outlines.
12054 It accepts the following options:
12057 @item luma_radius, lr
12058 Set the luma radius. The option value must be a float number in
12059 the range [0.1,5.0] that specifies the variance of the gaussian filter
12060 used to blur the image (slower if larger). Default value is 1.0.
12062 @item luma_strength, ls
12063 Set the luma strength. The option value must be a float number
12064 in the range [-1.0,1.0] that configures the blurring. A value included
12065 in [0.0,1.0] will blur the image whereas a value included in
12066 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12068 @item luma_threshold, lt
12069 Set the luma threshold used as a coefficient to determine
12070 whether a pixel should be blurred or not. The option value must be an
12071 integer in the range [-30,30]. A value of 0 will filter all the image,
12072 a value included in [0,30] will filter flat areas and a value included
12073 in [-30,0] will filter edges. Default value is 0.
12075 @item chroma_radius, cr
12076 Set the chroma radius. The option value must be a float number in
12077 the range [0.1,5.0] that specifies the variance of the gaussian filter
12078 used to blur the image (slower if larger). Default value is 1.0.
12080 @item chroma_strength, cs
12081 Set the chroma strength. The option value must be a float number
12082 in the range [-1.0,1.0] that configures the blurring. A value included
12083 in [0.0,1.0] will blur the image whereas a value included in
12084 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12086 @item chroma_threshold, ct
12087 Set the chroma threshold used as a coefficient to determine
12088 whether a pixel should be blurred or not. The option value must be an
12089 integer in the range [-30,30]. A value of 0 will filter all the image,
12090 a value included in [0,30] will filter flat areas and a value included
12091 in [-30,0] will filter edges. Default value is 0.
12094 If a chroma option is not explicitly set, the corresponding luma value
12099 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12101 This filter takes in input two input videos, the first input is
12102 considered the "main" source and is passed unchanged to the
12103 output. The second input is used as a "reference" video for computing
12106 Both video inputs must have the same resolution and pixel format for
12107 this filter to work correctly. Also it assumes that both inputs
12108 have the same number of frames, which are compared one by one.
12110 The filter stores the calculated SSIM of each frame.
12112 The description of the accepted parameters follows.
12115 @item stats_file, f
12116 If specified the filter will use the named file to save the SSIM of
12117 each individual frame. When filename equals "-" the data is sent to
12121 The file printed if @var{stats_file} is selected, contains a sequence of
12122 key/value pairs of the form @var{key}:@var{value} for each compared
12125 A description of each shown parameter follows:
12129 sequential number of the input frame, starting from 1
12131 @item Y, U, V, R, G, B
12132 SSIM of the compared frames for the component specified by the suffix.
12135 SSIM of the compared frames for the whole frame.
12138 Same as above but in dB representation.
12143 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12144 [main][ref] ssim="stats_file=stats.log" [out]
12147 On this example the input file being processed is compared with the
12148 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12149 is stored in @file{stats.log}.
12151 Another example with both psnr and ssim at same time:
12153 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12158 Convert between different stereoscopic image formats.
12160 The filters accept the following options:
12164 Set stereoscopic image format of input.
12166 Available values for input image formats are:
12169 side by side parallel (left eye left, right eye right)
12172 side by side crosseye (right eye left, left eye right)
12175 side by side parallel with half width resolution
12176 (left eye left, right eye right)
12179 side by side crosseye with half width resolution
12180 (right eye left, left eye right)
12183 above-below (left eye above, right eye below)
12186 above-below (right eye above, left eye below)
12189 above-below with half height resolution
12190 (left eye above, right eye below)
12193 above-below with half height resolution
12194 (right eye above, left eye below)
12197 alternating frames (left eye first, right eye second)
12200 alternating frames (right eye first, left eye second)
12203 interleaved rows (left eye has top row, right eye starts on next row)
12206 interleaved rows (right eye has top row, left eye starts on next row)
12209 interleaved columns, left eye first
12212 interleaved columns, right eye first
12214 Default value is @samp{sbsl}.
12218 Set stereoscopic image format of output.
12222 side by side parallel (left eye left, right eye right)
12225 side by side crosseye (right eye left, left eye right)
12228 side by side parallel with half width resolution
12229 (left eye left, right eye right)
12232 side by side crosseye with half width resolution
12233 (right eye left, left eye right)
12236 above-below (left eye above, right eye below)
12239 above-below (right eye above, left eye below)
12242 above-below with half height resolution
12243 (left eye above, right eye below)
12246 above-below with half height resolution
12247 (right eye above, left eye below)
12250 alternating frames (left eye first, right eye second)
12253 alternating frames (right eye first, left eye second)
12256 interleaved rows (left eye has top row, right eye starts on next row)
12259 interleaved rows (right eye has top row, left eye starts on next row)
12262 anaglyph red/blue gray
12263 (red filter on left eye, blue filter on right eye)
12266 anaglyph red/green gray
12267 (red filter on left eye, green filter on right eye)
12270 anaglyph red/cyan gray
12271 (red filter on left eye, cyan filter on right eye)
12274 anaglyph red/cyan half colored
12275 (red filter on left eye, cyan filter on right eye)
12278 anaglyph red/cyan color
12279 (red filter on left eye, cyan filter on right eye)
12282 anaglyph red/cyan color optimized with the least squares projection of dubois
12283 (red filter on left eye, cyan filter on right eye)
12286 anaglyph green/magenta gray
12287 (green filter on left eye, magenta filter on right eye)
12290 anaglyph green/magenta half colored
12291 (green filter on left eye, magenta filter on right eye)
12294 anaglyph green/magenta colored
12295 (green filter on left eye, magenta filter on right eye)
12298 anaglyph green/magenta color optimized with the least squares projection of dubois
12299 (green filter on left eye, magenta filter on right eye)
12302 anaglyph yellow/blue gray
12303 (yellow filter on left eye, blue filter on right eye)
12306 anaglyph yellow/blue half colored
12307 (yellow filter on left eye, blue filter on right eye)
12310 anaglyph yellow/blue colored
12311 (yellow filter on left eye, blue filter on right eye)
12314 anaglyph yellow/blue color optimized with the least squares projection of dubois
12315 (yellow filter on left eye, blue filter on right eye)
12318 mono output (left eye only)
12321 mono output (right eye only)
12324 checkerboard, left eye first
12327 checkerboard, right eye first
12330 interleaved columns, left eye first
12333 interleaved columns, right eye first
12339 Default value is @samp{arcd}.
12342 @subsection Examples
12346 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12352 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12358 @section streamselect, astreamselect
12359 Select video or audio streams.
12361 The filter accepts the following options:
12365 Set number of inputs. Default is 2.
12368 Set input indexes to remap to outputs.
12371 @subsection Commands
12373 The @code{streamselect} and @code{astreamselect} filter supports the following
12378 Set input indexes to remap to outputs.
12381 @subsection Examples
12385 Select first 5 seconds 1st stream and rest of time 2nd stream:
12387 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12391 Same as above, but for audio:
12393 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12400 Apply a simple postprocessing filter that compresses and decompresses the image
12401 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12402 and average the results.
12404 The filter accepts the following options:
12408 Set quality. This option defines the number of levels for averaging. It accepts
12409 an integer in the range 0-6. If set to @code{0}, the filter will have no
12410 effect. A value of @code{6} means the higher quality. For each increment of
12411 that value the speed drops by a factor of approximately 2. Default value is
12415 Force a constant quantization parameter. If not set, the filter will use the QP
12416 from the video stream (if available).
12419 Set thresholding mode. Available modes are:
12423 Set hard thresholding (default).
12425 Set soft thresholding (better de-ringing effect, but likely blurrier).
12428 @item use_bframe_qp
12429 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12430 option may cause flicker since the B-Frames have often larger QP. Default is
12431 @code{0} (not enabled).
12437 Draw subtitles on top of input video using the libass library.
12439 To enable compilation of this filter you need to configure FFmpeg with
12440 @code{--enable-libass}. This filter also requires a build with libavcodec and
12441 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12442 Alpha) subtitles format.
12444 The filter accepts the following options:
12448 Set the filename of the subtitle file to read. It must be specified.
12450 @item original_size
12451 Specify the size of the original video, the video for which the ASS file
12452 was composed. For the syntax of this option, check the
12453 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12454 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12455 correctly scale the fonts if the aspect ratio has been changed.
12458 Set a directory path containing fonts that can be used by the filter.
12459 These fonts will be used in addition to whatever the font provider uses.
12462 Set subtitles input character encoding. @code{subtitles} filter only. Only
12463 useful if not UTF-8.
12465 @item stream_index, si
12466 Set subtitles stream index. @code{subtitles} filter only.
12469 Override default style or script info parameters of the subtitles. It accepts a
12470 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12473 If the first key is not specified, it is assumed that the first value
12474 specifies the @option{filename}.
12476 For example, to render the file @file{sub.srt} on top of the input
12477 video, use the command:
12482 which is equivalent to:
12484 subtitles=filename=sub.srt
12487 To render the default subtitles stream from file @file{video.mkv}, use:
12489 subtitles=video.mkv
12492 To render the second subtitles stream from that file, use:
12494 subtitles=video.mkv:si=1
12497 To make the subtitles stream from @file{sub.srt} appear in transparent green
12498 @code{DejaVu Serif}, use:
12500 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12503 @section super2xsai
12505 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12506 Interpolate) pixel art scaling algorithm.
12508 Useful for enlarging pixel art images without reducing sharpness.
12512 Swap two rectangular objects in video.
12514 This filter accepts the following options:
12524 Set 1st rect x coordinate.
12527 Set 1st rect y coordinate.
12530 Set 2nd rect x coordinate.
12533 Set 2nd rect y coordinate.
12535 All expressions are evaluated once for each frame.
12538 The all options are expressions containing the following constants:
12543 The input width and height.
12546 same as @var{w} / @var{h}
12549 input sample aspect ratio
12552 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12555 The number of the input frame, starting from 0.
12558 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12561 the position in the file of the input frame, NAN if unknown
12569 Apply telecine process to the video.
12571 This filter accepts the following options:
12580 The default value is @code{top}.
12584 A string of numbers representing the pulldown pattern you wish to apply.
12585 The default value is @code{23}.
12589 Some typical patterns:
12594 24p: 2332 (preferred)
12601 24p: 222222222223 ("Euro pulldown")
12607 Select the most representative frame in a given sequence of consecutive frames.
12609 The filter accepts the following options:
12613 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
12614 will pick one of them, and then handle the next batch of @var{n} frames until
12615 the end. Default is @code{100}.
12618 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
12619 value will result in a higher memory usage, so a high value is not recommended.
12621 @subsection Examples
12625 Extract one picture each 50 frames:
12631 Complete example of a thumbnail creation with @command{ffmpeg}:
12633 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
12639 Tile several successive frames together.
12641 The filter accepts the following options:
12646 Set the grid size (i.e. the number of lines and columns). For the syntax of
12647 this option, check the
12648 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12651 Set the maximum number of frames to render in the given area. It must be less
12652 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
12653 the area will be used.
12656 Set the outer border margin in pixels.
12659 Set the inner border thickness (i.e. the number of pixels between frames). For
12660 more advanced padding options (such as having different values for the edges),
12661 refer to the pad video filter.
12664 Specify the color of the unused area. For the syntax of this option, check the
12665 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
12669 @subsection Examples
12673 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
12675 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
12677 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
12678 duplicating each output frame to accommodate the originally detected frame
12682 Display @code{5} pictures in an area of @code{3x2} frames,
12683 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
12684 mixed flat and named options:
12686 tile=3x2:nb_frames=5:padding=7:margin=2
12690 @section tinterlace
12692 Perform various types of temporal field interlacing.
12694 Frames are counted starting from 1, so the first input frame is
12697 The filter accepts the following options:
12702 Specify the mode of the interlacing. This option can also be specified
12703 as a value alone. See below for a list of values for this option.
12705 Available values are:
12709 Move odd frames into the upper field, even into the lower field,
12710 generating a double height frame at half frame rate.
12714 Frame 1 Frame 2 Frame 3 Frame 4
12716 11111 22222 33333 44444
12717 11111 22222 33333 44444
12718 11111 22222 33333 44444
12719 11111 22222 33333 44444
12733 Only output odd frames, even frames are dropped, generating a frame with
12734 unchanged height at half frame rate.
12739 Frame 1 Frame 2 Frame 3 Frame 4
12741 11111 22222 33333 44444
12742 11111 22222 33333 44444
12743 11111 22222 33333 44444
12744 11111 22222 33333 44444
12754 Only output even frames, odd frames are dropped, generating a frame with
12755 unchanged height at half frame rate.
12760 Frame 1 Frame 2 Frame 3 Frame 4
12762 11111 22222 33333 44444
12763 11111 22222 33333 44444
12764 11111 22222 33333 44444
12765 11111 22222 33333 44444
12775 Expand each frame to full height, but pad alternate lines with black,
12776 generating a frame with double height at the same input frame rate.
12781 Frame 1 Frame 2 Frame 3 Frame 4
12783 11111 22222 33333 44444
12784 11111 22222 33333 44444
12785 11111 22222 33333 44444
12786 11111 22222 33333 44444
12789 11111 ..... 33333 .....
12790 ..... 22222 ..... 44444
12791 11111 ..... 33333 .....
12792 ..... 22222 ..... 44444
12793 11111 ..... 33333 .....
12794 ..... 22222 ..... 44444
12795 11111 ..... 33333 .....
12796 ..... 22222 ..... 44444
12800 @item interleave_top, 4
12801 Interleave the upper field from odd frames with the lower field from
12802 even frames, generating a frame with unchanged height at half frame rate.
12807 Frame 1 Frame 2 Frame 3 Frame 4
12809 11111<- 22222 33333<- 44444
12810 11111 22222<- 33333 44444<-
12811 11111<- 22222 33333<- 44444
12812 11111 22222<- 33333 44444<-
12822 @item interleave_bottom, 5
12823 Interleave the lower field from odd frames with the upper field from
12824 even frames, generating a frame with unchanged height at half frame rate.
12829 Frame 1 Frame 2 Frame 3 Frame 4
12831 11111 22222<- 33333 44444<-
12832 11111<- 22222 33333<- 44444
12833 11111 22222<- 33333 44444<-
12834 11111<- 22222 33333<- 44444
12844 @item interlacex2, 6
12845 Double frame rate with unchanged height. Frames are inserted each
12846 containing the second temporal field from the previous input frame and
12847 the first temporal field from the next input frame. This mode relies on
12848 the top_field_first flag. Useful for interlaced video displays with no
12849 field synchronisation.
12854 Frame 1 Frame 2 Frame 3 Frame 4
12856 11111 22222 33333 44444
12857 11111 22222 33333 44444
12858 11111 22222 33333 44444
12859 11111 22222 33333 44444
12862 11111 22222 22222 33333 33333 44444 44444
12863 11111 11111 22222 22222 33333 33333 44444
12864 11111 22222 22222 33333 33333 44444 44444
12865 11111 11111 22222 22222 33333 33333 44444
12870 Move odd frames into the upper field, even into the lower field,
12871 generating a double height frame at same frame rate.
12876 Frame 1 Frame 2 Frame 3 Frame 4
12878 11111 22222 33333 44444
12879 11111 22222 33333 44444
12880 11111 22222 33333 44444
12881 11111 22222 33333 44444
12884 11111 33333 33333 55555
12885 22222 22222 44444 44444
12886 11111 33333 33333 55555
12887 22222 22222 44444 44444
12888 11111 33333 33333 55555
12889 22222 22222 44444 44444
12890 11111 33333 33333 55555
12891 22222 22222 44444 44444
12896 Numeric values are deprecated but are accepted for backward
12897 compatibility reasons.
12899 Default mode is @code{merge}.
12902 Specify flags influencing the filter process.
12904 Available value for @var{flags} is:
12907 @item low_pass_filter, vlfp
12908 Enable vertical low-pass filtering in the filter.
12909 Vertical low-pass filtering is required when creating an interlaced
12910 destination from a progressive source which contains high-frequency
12911 vertical detail. Filtering will reduce interlace 'twitter' and Moire
12914 Vertical low-pass filtering can only be enabled for @option{mode}
12915 @var{interleave_top} and @var{interleave_bottom}.
12922 Transpose rows with columns in the input video and optionally flip it.
12924 It accepts the following parameters:
12929 Specify the transposition direction.
12931 Can assume the following values:
12933 @item 0, 4, cclock_flip
12934 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
12942 Rotate by 90 degrees clockwise, that is:
12950 Rotate by 90 degrees counterclockwise, that is:
12957 @item 3, 7, clock_flip
12958 Rotate by 90 degrees clockwise and vertically flip, that is:
12966 For values between 4-7, the transposition is only done if the input
12967 video geometry is portrait and not landscape. These values are
12968 deprecated, the @code{passthrough} option should be used instead.
12970 Numerical values are deprecated, and should be dropped in favor of
12971 symbolic constants.
12974 Do not apply the transposition if the input geometry matches the one
12975 specified by the specified value. It accepts the following values:
12978 Always apply transposition.
12980 Preserve portrait geometry (when @var{height} >= @var{width}).
12982 Preserve landscape geometry (when @var{width} >= @var{height}).
12985 Default value is @code{none}.
12988 For example to rotate by 90 degrees clockwise and preserve portrait
12991 transpose=dir=1:passthrough=portrait
12994 The command above can also be specified as:
12996 transpose=1:portrait
13000 Trim the input so that the output contains one continuous subpart of the input.
13002 It accepts the following parameters:
13005 Specify the time of the start of the kept section, i.e. the frame with the
13006 timestamp @var{start} will be the first frame in the output.
13009 Specify the time of the first frame that will be dropped, i.e. the frame
13010 immediately preceding the one with the timestamp @var{end} will be the last
13011 frame in the output.
13014 This is the same as @var{start}, except this option sets the start timestamp
13015 in timebase units instead of seconds.
13018 This is the same as @var{end}, except this option sets the end timestamp
13019 in timebase units instead of seconds.
13022 The maximum duration of the output in seconds.
13025 The number of the first frame that should be passed to the output.
13028 The number of the first frame that should be dropped.
13031 @option{start}, @option{end}, and @option{duration} are expressed as time
13032 duration specifications; see
13033 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13034 for the accepted syntax.
13036 Note that the first two sets of the start/end options and the @option{duration}
13037 option look at the frame timestamp, while the _frame variants simply count the
13038 frames that pass through the filter. Also note that this filter does not modify
13039 the timestamps. If you wish for the output timestamps to start at zero, insert a
13040 setpts filter after the trim filter.
13042 If multiple start or end options are set, this filter tries to be greedy and
13043 keep all the frames that match at least one of the specified constraints. To keep
13044 only the part that matches all the constraints at once, chain multiple trim
13047 The defaults are such that all the input is kept. So it is possible to set e.g.
13048 just the end values to keep everything before the specified time.
13053 Drop everything except the second minute of input:
13055 ffmpeg -i INPUT -vf trim=60:120
13059 Keep only the first second:
13061 ffmpeg -i INPUT -vf trim=duration=1
13070 Sharpen or blur the input video.
13072 It accepts the following parameters:
13075 @item luma_msize_x, lx
13076 Set the luma matrix horizontal size. It must be an odd integer between
13077 3 and 63. The default value is 5.
13079 @item luma_msize_y, ly
13080 Set the luma matrix vertical size. It must be an odd integer between 3
13081 and 63. The default value is 5.
13083 @item luma_amount, la
13084 Set the luma effect strength. It must be a floating point number, reasonable
13085 values lay between -1.5 and 1.5.
13087 Negative values will blur the input video, while positive values will
13088 sharpen it, a value of zero will disable the effect.
13090 Default value is 1.0.
13092 @item chroma_msize_x, cx
13093 Set the chroma matrix horizontal size. It must be an odd integer
13094 between 3 and 63. The default value is 5.
13096 @item chroma_msize_y, cy
13097 Set the chroma matrix vertical size. It must be an odd integer
13098 between 3 and 63. The default value is 5.
13100 @item chroma_amount, ca
13101 Set the chroma effect strength. It must be a floating point number, reasonable
13102 values lay between -1.5 and 1.5.
13104 Negative values will blur the input video, while positive values will
13105 sharpen it, a value of zero will disable the effect.
13107 Default value is 0.0.
13110 If set to 1, specify using OpenCL capabilities, only available if
13111 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13115 All parameters are optional and default to the equivalent of the
13116 string '5:5:1.0:5:5:0.0'.
13118 @subsection Examples
13122 Apply strong luma sharpen effect:
13124 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13128 Apply a strong blur of both luma and chroma parameters:
13130 unsharp=7:7:-2:7:7:-2
13136 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13137 the image at several (or - in the case of @option{quality} level @code{8} - all)
13138 shifts and average the results.
13140 The way this differs from the behavior of spp is that uspp actually encodes &
13141 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13142 DCT similar to MJPEG.
13144 The filter accepts the following options:
13148 Set quality. This option defines the number of levels for averaging. It accepts
13149 an integer in the range 0-8. If set to @code{0}, the filter will have no
13150 effect. A value of @code{8} means the higher quality. For each increment of
13151 that value the speed drops by a factor of approximately 2. Default value is
13155 Force a constant quantization parameter. If not set, the filter will use the QP
13156 from the video stream (if available).
13159 @section vectorscope
13161 Display 2 color component values in the two dimensional graph (which is called
13164 This filter accepts the following options:
13168 Set vectorscope mode.
13170 It accepts the following values:
13173 Gray values are displayed on graph, higher brightness means more pixels have
13174 same component color value on location in graph. This is the default mode.
13177 Gray values are displayed on graph. Surrounding pixels values which are not
13178 present in video frame are drawn in gradient of 2 color components which are
13179 set by option @code{x} and @code{y}. The 3rd color component is static.
13182 Actual color components values present in video frame are displayed on graph.
13185 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13186 on graph increases value of another color component, which is luminance by
13187 default values of @code{x} and @code{y}.
13190 Actual colors present in video frame are displayed on graph. If two different
13191 colors map to same position on graph then color with higher value of component
13192 not present in graph is picked.
13195 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13196 component picked from radial gradient.
13200 Set which color component will be represented on X-axis. Default is @code{1}.
13203 Set which color component will be represented on Y-axis. Default is @code{2}.
13206 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13207 of color component which represents frequency of (X, Y) location in graph.
13212 No envelope, this is default.
13215 Instant envelope, even darkest single pixel will be clearly highlighted.
13218 Hold maximum and minimum values presented in graph over time. This way you
13219 can still spot out of range values without constantly looking at vectorscope.
13222 Peak and instant envelope combined together.
13226 Set what kind of graticule to draw.
13234 Set graticule opacity.
13237 Set graticule flags.
13241 Draw graticule for white point.
13244 Draw graticule for black point.
13247 Draw color points short names.
13251 Set background opacity.
13253 @item lthreshold, l
13254 Set low threshold for color component not represented on X or Y axis.
13255 Values lower than this value will be ignored. Default is 0.
13256 Note this value is multiplied with actual max possible value one pixel component
13257 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13260 @item hthreshold, h
13261 Set high threshold for color component not represented on X or Y axis.
13262 Values higher than this value will be ignored. Default is 1.
13263 Note this value is multiplied with actual max possible value one pixel component
13264 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13265 is 0.9 * 255 = 230.
13267 @item colorspace, c
13268 Set what kind of colorspace to use when drawing graticule.
13277 @anchor{vidstabdetect}
13278 @section vidstabdetect
13280 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13281 @ref{vidstabtransform} for pass 2.
13283 This filter generates a file with relative translation and rotation
13284 transform information about subsequent frames, which is then used by
13285 the @ref{vidstabtransform} filter.
13287 To enable compilation of this filter you need to configure FFmpeg with
13288 @code{--enable-libvidstab}.
13290 This filter accepts the following options:
13294 Set the path to the file used to write the transforms information.
13295 Default value is @file{transforms.trf}.
13298 Set how shaky the video is and how quick the camera is. It accepts an
13299 integer in the range 1-10, a value of 1 means little shakiness, a
13300 value of 10 means strong shakiness. Default value is 5.
13303 Set the accuracy of the detection process. It must be a value in the
13304 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13305 accuracy. Default value is 15.
13308 Set stepsize of the search process. The region around minimum is
13309 scanned with 1 pixel resolution. Default value is 6.
13312 Set minimum contrast. Below this value a local measurement field is
13313 discarded. Must be a floating point value in the range 0-1. Default
13317 Set reference frame number for tripod mode.
13319 If enabled, the motion of the frames is compared to a reference frame
13320 in the filtered stream, identified by the specified number. The idea
13321 is to compensate all movements in a more-or-less static scene and keep
13322 the camera view absolutely still.
13324 If set to 0, it is disabled. The frames are counted starting from 1.
13327 Show fields and transforms in the resulting frames. It accepts an
13328 integer in the range 0-2. Default value is 0, which disables any
13332 @subsection Examples
13336 Use default values:
13342 Analyze strongly shaky movie and put the results in file
13343 @file{mytransforms.trf}:
13345 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13349 Visualize the result of internal transformations in the resulting
13352 vidstabdetect=show=1
13356 Analyze a video with medium shakiness using @command{ffmpeg}:
13358 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13362 @anchor{vidstabtransform}
13363 @section vidstabtransform
13365 Video stabilization/deshaking: pass 2 of 2,
13366 see @ref{vidstabdetect} for pass 1.
13368 Read a file with transform information for each frame and
13369 apply/compensate them. Together with the @ref{vidstabdetect}
13370 filter this can be used to deshake videos. See also
13371 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13372 the @ref{unsharp} filter, see below.
13374 To enable compilation of this filter you need to configure FFmpeg with
13375 @code{--enable-libvidstab}.
13377 @subsection Options
13381 Set path to the file used to read the transforms. Default value is
13382 @file{transforms.trf}.
13385 Set the number of frames (value*2 + 1) used for lowpass filtering the
13386 camera movements. Default value is 10.
13388 For example a number of 10 means that 21 frames are used (10 in the
13389 past and 10 in the future) to smoothen the motion in the video. A
13390 larger value leads to a smoother video, but limits the acceleration of
13391 the camera (pan/tilt movements). 0 is a special case where a static
13392 camera is simulated.
13395 Set the camera path optimization algorithm.
13397 Accepted values are:
13400 gaussian kernel low-pass filter on camera motion (default)
13402 averaging on transformations
13406 Set maximal number of pixels to translate frames. Default value is -1,
13410 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13411 value is -1, meaning no limit.
13414 Specify how to deal with borders that may be visible due to movement
13417 Available values are:
13420 keep image information from previous frame (default)
13422 fill the border black
13426 Invert transforms if set to 1. Default value is 0.
13429 Consider transforms as relative to previous frame if set to 1,
13430 absolute if set to 0. Default value is 0.
13433 Set percentage to zoom. A positive value will result in a zoom-in
13434 effect, a negative value in a zoom-out effect. Default value is 0 (no
13438 Set optimal zooming to avoid borders.
13440 Accepted values are:
13445 optimal static zoom value is determined (only very strong movements
13446 will lead to visible borders) (default)
13448 optimal adaptive zoom value is determined (no borders will be
13449 visible), see @option{zoomspeed}
13452 Note that the value given at zoom is added to the one calculated here.
13455 Set percent to zoom maximally each frame (enabled when
13456 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13460 Specify type of interpolation.
13462 Available values are:
13467 linear only horizontal
13469 linear in both directions (default)
13471 cubic in both directions (slow)
13475 Enable virtual tripod mode if set to 1, which is equivalent to
13476 @code{relative=0:smoothing=0}. Default value is 0.
13478 Use also @code{tripod} option of @ref{vidstabdetect}.
13481 Increase log verbosity if set to 1. Also the detected global motions
13482 are written to the temporary file @file{global_motions.trf}. Default
13486 @subsection Examples
13490 Use @command{ffmpeg} for a typical stabilization with default values:
13492 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13495 Note the use of the @ref{unsharp} filter which is always recommended.
13498 Zoom in a bit more and load transform data from a given file:
13500 vidstabtransform=zoom=5:input="mytransforms.trf"
13504 Smoothen the video even more:
13506 vidstabtransform=smoothing=30
13512 Flip the input video vertically.
13514 For example, to vertically flip a video with @command{ffmpeg}:
13516 ffmpeg -i in.avi -vf "vflip" out.avi
13522 Make or reverse a natural vignetting effect.
13524 The filter accepts the following options:
13528 Set lens angle expression as a number of radians.
13530 The value is clipped in the @code{[0,PI/2]} range.
13532 Default value: @code{"PI/5"}
13536 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
13540 Set forward/backward mode.
13542 Available modes are:
13545 The larger the distance from the central point, the darker the image becomes.
13548 The larger the distance from the central point, the brighter the image becomes.
13549 This can be used to reverse a vignette effect, though there is no automatic
13550 detection to extract the lens @option{angle} and other settings (yet). It can
13551 also be used to create a burning effect.
13554 Default value is @samp{forward}.
13557 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
13559 It accepts the following values:
13562 Evaluate expressions only once during the filter initialization.
13565 Evaluate expressions for each incoming frame. This is way slower than the
13566 @samp{init} mode since it requires all the scalers to be re-computed, but it
13567 allows advanced dynamic expressions.
13570 Default value is @samp{init}.
13573 Set dithering to reduce the circular banding effects. Default is @code{1}
13577 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
13578 Setting this value to the SAR of the input will make a rectangular vignetting
13579 following the dimensions of the video.
13581 Default is @code{1/1}.
13584 @subsection Expressions
13586 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
13587 following parameters.
13592 input width and height
13595 the number of input frame, starting from 0
13598 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
13599 @var{TB} units, NAN if undefined
13602 frame rate of the input video, NAN if the input frame rate is unknown
13605 the PTS (Presentation TimeStamp) of the filtered video frame,
13606 expressed in seconds, NAN if undefined
13609 time base of the input video
13613 @subsection Examples
13617 Apply simple strong vignetting effect:
13623 Make a flickering vignetting:
13625 vignette='PI/4+random(1)*PI/50':eval=frame
13631 Stack input videos vertically.
13633 All streams must be of same pixel format and of same width.
13635 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13636 to create same output.
13638 The filter accept the following option:
13642 Set number of input streams. Default is 2.
13645 If set to 1, force the output to terminate when the shortest input
13646 terminates. Default value is 0.
13651 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
13652 Deinterlacing Filter").
13654 Based on the process described by Martin Weston for BBC R&D, and
13655 implemented based on the de-interlace algorithm written by Jim
13656 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
13657 uses filter coefficients calculated by BBC R&D.
13659 There are two sets of filter coefficients, so called "simple":
13660 and "complex". Which set of filter coefficients is used can
13661 be set by passing an optional parameter:
13665 Set the interlacing filter coefficients. Accepts one of the following values:
13669 Simple filter coefficient set.
13671 More-complex filter coefficient set.
13673 Default value is @samp{complex}.
13676 Specify which frames to deinterlace. Accept one of the following values:
13680 Deinterlace all frames,
13682 Only deinterlace frames marked as interlaced.
13685 Default value is @samp{all}.
13689 Video waveform monitor.
13691 The waveform monitor plots color component intensity. By default luminance
13692 only. Each column of the waveform corresponds to a column of pixels in the
13695 It accepts the following options:
13699 Can be either @code{row}, or @code{column}. Default is @code{column}.
13700 In row mode, the graph on the left side represents color component value 0 and
13701 the right side represents value = 255. In column mode, the top side represents
13702 color component value = 0 and bottom side represents value = 255.
13705 Set intensity. Smaller values are useful to find out how many values of the same
13706 luminance are distributed across input rows/columns.
13707 Default value is @code{0.04}. Allowed range is [0, 1].
13710 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
13711 In mirrored mode, higher values will be represented on the left
13712 side for @code{row} mode and at the top for @code{column} mode. Default is
13713 @code{1} (mirrored).
13717 It accepts the following values:
13720 Presents information identical to that in the @code{parade}, except
13721 that the graphs representing color components are superimposed directly
13724 This display mode makes it easier to spot relative differences or similarities
13725 in overlapping areas of the color components that are supposed to be identical,
13726 such as neutral whites, grays, or blacks.
13729 Display separate graph for the color components side by side in
13730 @code{row} mode or one below the other in @code{column} mode.
13733 Display separate graph for the color components side by side in
13734 @code{column} mode or one below the other in @code{row} mode.
13736 Using this display mode makes it easy to spot color casts in the highlights
13737 and shadows of an image, by comparing the contours of the top and the bottom
13738 graphs of each waveform. Since whites, grays, and blacks are characterized
13739 by exactly equal amounts of red, green, and blue, neutral areas of the picture
13740 should display three waveforms of roughly equal width/height. If not, the
13741 correction is easy to perform by making level adjustments the three waveforms.
13743 Default is @code{stack}.
13745 @item components, c
13746 Set which color components to display. Default is 1, which means only luminance
13747 or red color component if input is in RGB colorspace. If is set for example to
13748 7 it will display all 3 (if) available color components.
13753 No envelope, this is default.
13756 Instant envelope, minimum and maximum values presented in graph will be easily
13757 visible even with small @code{step} value.
13760 Hold minimum and maximum values presented in graph across time. This way you
13761 can still spot out of range values without constantly looking at waveforms.
13764 Peak and instant envelope combined together.
13770 No filtering, this is default.
13773 Luma and chroma combined together.
13776 Similar as above, but shows difference between blue and red chroma.
13779 Displays only chroma.
13782 Displays actual color value on waveform.
13785 Similar as above, but with luma showing frequency of chroma values.
13789 Set which graticule to display.
13793 Do not display graticule.
13796 Display green graticule showing legal broadcast ranges.
13800 Set graticule opacity.
13803 Set graticule flags.
13807 Draw numbers above lines. By default enabled.
13810 Draw dots instead of lines.
13814 Set scale used for displaying graticule.
13821 Default is digital.
13825 Apply the xBR high-quality magnification filter which is designed for pixel
13826 art. It follows a set of edge-detection rules, see
13827 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
13829 It accepts the following option:
13833 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
13834 @code{3xBR} and @code{4} for @code{4xBR}.
13835 Default is @code{3}.
13841 Deinterlace the input video ("yadif" means "yet another deinterlacing
13844 It accepts the following parameters:
13850 The interlacing mode to adopt. It accepts one of the following values:
13853 @item 0, send_frame
13854 Output one frame for each frame.
13855 @item 1, send_field
13856 Output one frame for each field.
13857 @item 2, send_frame_nospatial
13858 Like @code{send_frame}, but it skips the spatial interlacing check.
13859 @item 3, send_field_nospatial
13860 Like @code{send_field}, but it skips the spatial interlacing check.
13863 The default value is @code{send_frame}.
13866 The picture field parity assumed for the input interlaced video. It accepts one
13867 of the following values:
13871 Assume the top field is first.
13873 Assume the bottom field is first.
13875 Enable automatic detection of field parity.
13878 The default value is @code{auto}.
13879 If the interlacing is unknown or the decoder does not export this information,
13880 top field first will be assumed.
13883 Specify which frames to deinterlace. Accept one of the following
13888 Deinterlace all frames.
13889 @item 1, interlaced
13890 Only deinterlace frames marked as interlaced.
13893 The default value is @code{all}.
13898 Apply Zoom & Pan effect.
13900 This filter accepts the following options:
13904 Set the zoom expression. Default is 1.
13908 Set the x and y expression. Default is 0.
13911 Set the duration expression in number of frames.
13912 This sets for how many number of frames effect will last for
13913 single input image.
13916 Set the output image size, default is 'hd720'.
13919 Set the output frame rate, default is '25'.
13922 Each expression can contain the following constants:
13941 Output frame count.
13945 Last calculated 'x' and 'y' position from 'x' and 'y' expression
13946 for current input frame.
13950 'x' and 'y' of last output frame of previous input frame or 0 when there was
13951 not yet such frame (first input frame).
13954 Last calculated zoom from 'z' expression for current input frame.
13957 Last calculated zoom of last output frame of previous input frame.
13960 Number of output frames for current input frame. Calculated from 'd' expression
13961 for each input frame.
13964 number of output frames created for previous input frame
13967 Rational number: input width / input height
13970 sample aspect ratio
13973 display aspect ratio
13977 @subsection Examples
13981 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
13983 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
13987 Zoom-in up to 1.5 and pan always at center of picture:
13989 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
13993 Same as above but without pausing:
13995 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14000 Scale (resize) the input video, using the z.lib library:
14001 https://github.com/sekrit-twc/zimg.
14003 The zscale filter forces the output display aspect ratio to be the same
14004 as the input, by changing the output sample aspect ratio.
14006 If the input image format is different from the format requested by
14007 the next filter, the zscale filter will convert the input to the
14010 @subsection Options
14011 The filter accepts the following options.
14016 Set the output video dimension expression. Default value is the input
14019 If the @var{width} or @var{w} is 0, the input width is used for the output.
14020 If the @var{height} or @var{h} is 0, the input height is used for the output.
14022 If one of the values is -1, the zscale filter will use a value that
14023 maintains the aspect ratio of the input image, calculated from the
14024 other specified dimension. If both of them are -1, the input size is
14027 If one of the values is -n with n > 1, the zscale filter will also use a value
14028 that maintains the aspect ratio of the input image, calculated from the other
14029 specified dimension. After that it will, however, make sure that the calculated
14030 dimension is divisible by n and adjust the value if necessary.
14032 See below for the list of accepted constants for use in the dimension
14036 Set the video size. For the syntax of this option, check the
14037 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14040 Set the dither type.
14042 Possible values are:
14047 @item error_diffusion
14053 Set the resize filter type.
14055 Possible values are:
14065 Default is bilinear.
14068 Set the color range.
14070 Possible values are:
14077 Default is same as input.
14080 Set the color primaries.
14082 Possible values are:
14092 Default is same as input.
14095 Set the transfer characteristics.
14097 Possible values are:
14108 Default is same as input.
14111 Set the colorspace matrix.
14113 Possible value are:
14124 Default is same as input.
14127 Set the input color range.
14129 Possible values are:
14136 Default is same as input.
14138 @item primariesin, pin
14139 Set the input color primaries.
14141 Possible values are:
14151 Default is same as input.
14153 @item transferin, tin
14154 Set the input transfer characteristics.
14156 Possible values are:
14167 Default is same as input.
14169 @item matrixin, min
14170 Set the input colorspace matrix.
14172 Possible value are:
14184 The values of the @option{w} and @option{h} options are expressions
14185 containing the following constants:
14190 The input width and height
14194 These are the same as @var{in_w} and @var{in_h}.
14198 The output (scaled) width and height
14202 These are the same as @var{out_w} and @var{out_h}
14205 The same as @var{iw} / @var{ih}
14208 input sample aspect ratio
14211 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14215 horizontal and vertical input chroma subsample values. For example for the
14216 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14220 horizontal and vertical output chroma subsample values. For example for the
14221 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14227 @c man end VIDEO FILTERS
14229 @chapter Video Sources
14230 @c man begin VIDEO SOURCES
14232 Below is a description of the currently available video sources.
14236 Buffer video frames, and make them available to the filter chain.
14238 This source is mainly intended for a programmatic use, in particular
14239 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14241 It accepts the following parameters:
14246 Specify the size (width and height) of the buffered video frames. For the
14247 syntax of this option, check the
14248 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14251 The input video width.
14254 The input video height.
14257 A string representing the pixel format of the buffered video frames.
14258 It may be a number corresponding to a pixel format, or a pixel format
14262 Specify the timebase assumed by the timestamps of the buffered frames.
14265 Specify the frame rate expected for the video stream.
14267 @item pixel_aspect, sar
14268 The sample (pixel) aspect ratio of the input video.
14271 Specify the optional parameters to be used for the scale filter which
14272 is automatically inserted when an input change is detected in the
14273 input size or format.
14275 @item hw_frames_ctx
14276 When using a hardware pixel format, this should be a reference to an
14277 AVHWFramesContext describing input frames.
14282 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14285 will instruct the source to accept video frames with size 320x240 and
14286 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14287 square pixels (1:1 sample aspect ratio).
14288 Since the pixel format with name "yuv410p" corresponds to the number 6
14289 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14290 this example corresponds to:
14292 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14295 Alternatively, the options can be specified as a flat string, but this
14296 syntax is deprecated:
14298 @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}]
14302 Create a pattern generated by an elementary cellular automaton.
14304 The initial state of the cellular automaton can be defined through the
14305 @option{filename}, and @option{pattern} options. If such options are
14306 not specified an initial state is created randomly.
14308 At each new frame a new row in the video is filled with the result of
14309 the cellular automaton next generation. The behavior when the whole
14310 frame is filled is defined by the @option{scroll} option.
14312 This source accepts the following options:
14316 Read the initial cellular automaton state, i.e. the starting row, from
14317 the specified file.
14318 In the file, each non-whitespace character is considered an alive
14319 cell, a newline will terminate the row, and further characters in the
14320 file will be ignored.
14323 Read the initial cellular automaton state, i.e. the starting row, from
14324 the specified string.
14326 Each non-whitespace character in the string is considered an alive
14327 cell, a newline will terminate the row, and further characters in the
14328 string will be ignored.
14331 Set the video rate, that is the number of frames generated per second.
14334 @item random_fill_ratio, ratio
14335 Set the random fill ratio for the initial cellular automaton row. It
14336 is a floating point number value ranging from 0 to 1, defaults to
14339 This option is ignored when a file or a pattern is specified.
14341 @item random_seed, seed
14342 Set the seed for filling randomly the initial row, must be an integer
14343 included between 0 and UINT32_MAX. If not specified, or if explicitly
14344 set to -1, the filter will try to use a good random seed on a best
14348 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14349 Default value is 110.
14352 Set the size of the output video. For the syntax of this option, check the
14353 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14355 If @option{filename} or @option{pattern} is specified, the size is set
14356 by default to the width of the specified initial state row, and the
14357 height is set to @var{width} * PHI.
14359 If @option{size} is set, it must contain the width of the specified
14360 pattern string, and the specified pattern will be centered in the
14363 If a filename or a pattern string is not specified, the size value
14364 defaults to "320x518" (used for a randomly generated initial state).
14367 If set to 1, scroll the output upward when all the rows in the output
14368 have been already filled. If set to 0, the new generated row will be
14369 written over the top row just after the bottom row is filled.
14372 @item start_full, full
14373 If set to 1, completely fill the output with generated rows before
14374 outputting the first frame.
14375 This is the default behavior, for disabling set the value to 0.
14378 If set to 1, stitch the left and right row edges together.
14379 This is the default behavior, for disabling set the value to 0.
14382 @subsection Examples
14386 Read the initial state from @file{pattern}, and specify an output of
14389 cellauto=f=pattern:s=200x400
14393 Generate a random initial row with a width of 200 cells, with a fill
14396 cellauto=ratio=2/3:s=200x200
14400 Create a pattern generated by rule 18 starting by a single alive cell
14401 centered on an initial row with width 100:
14403 cellauto=p=@@:s=100x400:full=0:rule=18
14407 Specify a more elaborated initial pattern:
14409 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14414 @anchor{coreimagesrc}
14415 @section coreimagesrc
14416 Video source generated on GPU using Apple's CoreImage API on OSX.
14418 This video source is a specialized version of the @ref{coreimage} video filter.
14419 Use a core image generator at the beginning of the applied filterchain to
14420 generate the content.
14422 The coreimagesrc video source accepts the following options:
14424 @item list_generators
14425 List all available generators along with all their respective options as well as
14426 possible minimum and maximum values along with the default values.
14428 list_generators=true
14432 Specify the size of the sourced video. For the syntax of this option, check the
14433 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14434 The default value is @code{320x240}.
14437 Specify the frame rate of the sourced video, as the number of frames
14438 generated per second. It has to be a string in the format
14439 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14440 number or a valid video frame rate abbreviation. The default value is
14444 Set the sample aspect ratio of the sourced video.
14447 Set the duration of the sourced video. See
14448 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14449 for the accepted syntax.
14451 If not specified, or the expressed duration is negative, the video is
14452 supposed to be generated forever.
14455 Additionally, all options of the @ref{coreimage} video filter are accepted.
14456 A complete filterchain can be used for further processing of the
14457 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
14458 and examples for details.
14460 @subsection Examples
14465 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
14466 given as complete and escaped command-line for Apple's standard bash shell:
14468 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
14470 This example is equivalent to the QRCode example of @ref{coreimage} without the
14471 need for a nullsrc video source.
14475 @section mandelbrot
14477 Generate a Mandelbrot set fractal, and progressively zoom towards the
14478 point specified with @var{start_x} and @var{start_y}.
14480 This source accepts the following options:
14485 Set the terminal pts value. Default value is 400.
14488 Set the terminal scale value.
14489 Must be a floating point value. Default value is 0.3.
14492 Set the inner coloring mode, that is the algorithm used to draw the
14493 Mandelbrot fractal internal region.
14495 It shall assume one of the following values:
14500 Show time until convergence.
14502 Set color based on point closest to the origin of the iterations.
14507 Default value is @var{mincol}.
14510 Set the bailout value. Default value is 10.0.
14513 Set the maximum of iterations performed by the rendering
14514 algorithm. Default value is 7189.
14517 Set outer coloring mode.
14518 It shall assume one of following values:
14520 @item iteration_count
14521 Set iteration cound mode.
14522 @item normalized_iteration_count
14523 set normalized iteration count mode.
14525 Default value is @var{normalized_iteration_count}.
14528 Set frame rate, expressed as number of frames per second. Default
14532 Set frame size. For the syntax of this option, check the "Video
14533 size" section in the ffmpeg-utils manual. Default value is "640x480".
14536 Set the initial scale value. Default value is 3.0.
14539 Set the initial x position. Must be a floating point value between
14540 -100 and 100. Default value is -0.743643887037158704752191506114774.
14543 Set the initial y position. Must be a floating point value between
14544 -100 and 100. Default value is -0.131825904205311970493132056385139.
14549 Generate various test patterns, as generated by the MPlayer test filter.
14551 The size of the generated video is fixed, and is 256x256.
14552 This source is useful in particular for testing encoding features.
14554 This source accepts the following options:
14559 Specify the frame rate of the sourced video, as the number of frames
14560 generated per second. It has to be a string in the format
14561 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14562 number or a valid video frame rate abbreviation. The default value is
14566 Set the duration of the sourced video. See
14567 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14568 for the accepted syntax.
14570 If not specified, or the expressed duration is negative, the video is
14571 supposed to be generated forever.
14575 Set the number or the name of the test to perform. Supported tests are:
14591 Default value is "all", which will cycle through the list of all tests.
14596 mptestsrc=t=dc_luma
14599 will generate a "dc_luma" test pattern.
14601 @section frei0r_src
14603 Provide a frei0r source.
14605 To enable compilation of this filter you need to install the frei0r
14606 header and configure FFmpeg with @code{--enable-frei0r}.
14608 This source accepts the following parameters:
14613 The size of the video to generate. For the syntax of this option, check the
14614 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14617 The framerate of the generated video. It may be a string of the form
14618 @var{num}/@var{den} or a frame rate abbreviation.
14621 The name to the frei0r source to load. For more information regarding frei0r and
14622 how to set the parameters, read the @ref{frei0r} section in the video filters
14625 @item filter_params
14626 A '|'-separated list of parameters to pass to the frei0r source.
14630 For example, to generate a frei0r partik0l source with size 200x200
14631 and frame rate 10 which is overlaid on the overlay filter main input:
14633 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
14638 Generate a life pattern.
14640 This source is based on a generalization of John Conway's life game.
14642 The sourced input represents a life grid, each pixel represents a cell
14643 which can be in one of two possible states, alive or dead. Every cell
14644 interacts with its eight neighbours, which are the cells that are
14645 horizontally, vertically, or diagonally adjacent.
14647 At each interaction the grid evolves according to the adopted rule,
14648 which specifies the number of neighbor alive cells which will make a
14649 cell stay alive or born. The @option{rule} option allows one to specify
14652 This source accepts the following options:
14656 Set the file from which to read the initial grid state. In the file,
14657 each non-whitespace character is considered an alive cell, and newline
14658 is used to delimit the end of each row.
14660 If this option is not specified, the initial grid is generated
14664 Set the video rate, that is the number of frames generated per second.
14667 @item random_fill_ratio, ratio
14668 Set the random fill ratio for the initial random grid. It is a
14669 floating point number value ranging from 0 to 1, defaults to 1/PHI.
14670 It is ignored when a file is specified.
14672 @item random_seed, seed
14673 Set the seed for filling the initial random grid, must be an integer
14674 included between 0 and UINT32_MAX. If not specified, or if explicitly
14675 set to -1, the filter will try to use a good random seed on a best
14681 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
14682 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
14683 @var{NS} specifies the number of alive neighbor cells which make a
14684 live cell stay alive, and @var{NB} the number of alive neighbor cells
14685 which make a dead cell to become alive (i.e. to "born").
14686 "s" and "b" can be used in place of "S" and "B", respectively.
14688 Alternatively a rule can be specified by an 18-bits integer. The 9
14689 high order bits are used to encode the next cell state if it is alive
14690 for each number of neighbor alive cells, the low order bits specify
14691 the rule for "borning" new cells. Higher order bits encode for an
14692 higher number of neighbor cells.
14693 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
14694 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
14696 Default value is "S23/B3", which is the original Conway's game of life
14697 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
14698 cells, and will born a new cell if there are three alive cells around
14702 Set the size of the output video. For the syntax of this option, check the
14703 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14705 If @option{filename} is specified, the size is set by default to the
14706 same size of the input file. If @option{size} is set, it must contain
14707 the size specified in the input file, and the initial grid defined in
14708 that file is centered in the larger resulting area.
14710 If a filename is not specified, the size value defaults to "320x240"
14711 (used for a randomly generated initial grid).
14714 If set to 1, stitch the left and right grid edges together, and the
14715 top and bottom edges also. Defaults to 1.
14718 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
14719 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
14720 value from 0 to 255.
14723 Set the color of living (or new born) cells.
14726 Set the color of dead cells. If @option{mold} is set, this is the first color
14727 used to represent a dead cell.
14730 Set mold color, for definitely dead and moldy cells.
14732 For the syntax of these 3 color options, check the "Color" section in the
14733 ffmpeg-utils manual.
14736 @subsection Examples
14740 Read a grid from @file{pattern}, and center it on a grid of size
14743 life=f=pattern:s=300x300
14747 Generate a random grid of size 200x200, with a fill ratio of 2/3:
14749 life=ratio=2/3:s=200x200
14753 Specify a custom rule for evolving a randomly generated grid:
14759 Full example with slow death effect (mold) using @command{ffplay}:
14761 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
14768 @anchor{haldclutsrc}
14770 @anchor{rgbtestsrc}
14772 @anchor{smptehdbars}
14775 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2
14777 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
14779 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
14781 The @code{color} source provides an uniformly colored input.
14783 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
14784 @ref{haldclut} filter.
14786 The @code{nullsrc} source returns unprocessed video frames. It is
14787 mainly useful to be employed in analysis / debugging tools, or as the
14788 source for filters which ignore the input data.
14790 The @code{rgbtestsrc} source generates an RGB test pattern useful for
14791 detecting RGB vs BGR issues. You should see a red, green and blue
14792 stripe from top to bottom.
14794 The @code{smptebars} source generates a color bars pattern, based on
14795 the SMPTE Engineering Guideline EG 1-1990.
14797 The @code{smptehdbars} source generates a color bars pattern, based on
14798 the SMPTE RP 219-2002.
14800 The @code{testsrc} source generates a test video pattern, showing a
14801 color pattern, a scrolling gradient and a timestamp. This is mainly
14802 intended for testing purposes.
14804 The @code{testsrc2} source is similar to testsrc, but supports more
14805 pixel formats instead of just @code{rgb24}. This allows using it as an
14806 input for other tests without requiring a format conversion.
14808 The sources accept the following parameters:
14813 Specify the color of the source, only available in the @code{color}
14814 source. For the syntax of this option, check the "Color" section in the
14815 ffmpeg-utils manual.
14818 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
14819 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
14820 pixels to be used as identity matrix for 3D lookup tables. Each component is
14821 coded on a @code{1/(N*N)} scale.
14824 Specify the size of the sourced video. For the syntax of this option, check the
14825 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14826 The default value is @code{320x240}.
14828 This option is not available with the @code{haldclutsrc} filter.
14831 Specify the frame rate of the sourced video, as the number of frames
14832 generated per second. It has to be a string in the format
14833 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14834 number or a valid video frame rate abbreviation. The default value is
14838 Set the sample aspect ratio of the sourced video.
14841 Set the duration of the sourced video. See
14842 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14843 for the accepted syntax.
14845 If not specified, or the expressed duration is negative, the video is
14846 supposed to be generated forever.
14849 Set the number of decimals to show in the timestamp, only available in the
14850 @code{testsrc} source.
14852 The displayed timestamp value will correspond to the original
14853 timestamp value multiplied by the power of 10 of the specified
14854 value. Default value is 0.
14857 For example the following:
14859 testsrc=duration=5.3:size=qcif:rate=10
14862 will generate a video with a duration of 5.3 seconds, with size
14863 176x144 and a frame rate of 10 frames per second.
14865 The following graph description will generate a red source
14866 with an opacity of 0.2, with size "qcif" and a frame rate of 10
14869 color=c=red@@0.2:s=qcif:r=10
14872 If the input content is to be ignored, @code{nullsrc} can be used. The
14873 following command generates noise in the luminance plane by employing
14874 the @code{geq} filter:
14876 nullsrc=s=256x256, geq=random(1)*255:128:128
14879 @subsection Commands
14881 The @code{color} source supports the following commands:
14885 Set the color of the created image. Accepts the same syntax of the
14886 corresponding @option{color} option.
14889 @c man end VIDEO SOURCES
14891 @chapter Video Sinks
14892 @c man begin VIDEO SINKS
14894 Below is a description of the currently available video sinks.
14896 @section buffersink
14898 Buffer video frames, and make them available to the end of the filter
14901 This sink is mainly intended for programmatic use, in particular
14902 through the interface defined in @file{libavfilter/buffersink.h}
14903 or the options system.
14905 It accepts a pointer to an AVBufferSinkContext structure, which
14906 defines the incoming buffers' formats, to be passed as the opaque
14907 parameter to @code{avfilter_init_filter} for initialization.
14911 Null video sink: do absolutely nothing with the input video. It is
14912 mainly useful as a template and for use in analysis / debugging
14915 @c man end VIDEO SINKS
14917 @chapter Multimedia Filters
14918 @c man begin MULTIMEDIA FILTERS
14920 Below is a description of the currently available multimedia filters.
14922 @section ahistogram
14924 Convert input audio to a video output, displaying the volume histogram.
14926 The filter accepts the following options:
14930 Specify how histogram is calculated.
14932 It accepts the following values:
14935 Use single histogram for all channels.
14937 Use separate histogram for each channel.
14939 Default is @code{single}.
14942 Set frame rate, expressed as number of frames per second. Default
14946 Specify the video size for the output. For the syntax of this option, check the
14947 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14948 Default value is @code{hd720}.
14953 It accepts the following values:
14964 reverse logarithmic
14966 Default is @code{log}.
14969 Set amplitude scale.
14971 It accepts the following values:
14978 Default is @code{log}.
14981 Set how much frames to accumulate in histogram.
14982 Defauls is 1. Setting this to -1 accumulates all frames.
14985 Set histogram ratio of window height.
14988 Set sonogram sliding.
14990 It accepts the following values:
14993 replace old rows with new ones.
14995 scroll from top to bottom.
14997 Default is @code{replace}.
15000 @section aphasemeter
15002 Convert input audio to a video output, displaying the audio phase.
15004 The filter accepts the following options:
15008 Set the output frame rate. Default value is @code{25}.
15011 Set the video size for the output. For the syntax of this option, check the
15012 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15013 Default value is @code{800x400}.
15018 Specify the red, green, blue contrast. Default values are @code{2},
15019 @code{7} and @code{1}.
15020 Allowed range is @code{[0, 255]}.
15023 Set color which will be used for drawing median phase. If color is
15024 @code{none} which is default, no median phase value will be drawn.
15027 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15028 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15029 The @code{-1} means left and right channels are completely out of phase and
15030 @code{1} means channels are in phase.
15032 @section avectorscope
15034 Convert input audio to a video output, representing the audio vector
15037 The filter is used to measure the difference between channels of stereo
15038 audio stream. A monoaural signal, consisting of identical left and right
15039 signal, results in straight vertical line. Any stereo separation is visible
15040 as a deviation from this line, creating a Lissajous figure.
15041 If the straight (or deviation from it) but horizontal line appears this
15042 indicates that the left and right channels are out of phase.
15044 The filter accepts the following options:
15048 Set the vectorscope mode.
15050 Available values are:
15053 Lissajous rotated by 45 degrees.
15056 Same as above but not rotated.
15059 Shape resembling half of circle.
15062 Default value is @samp{lissajous}.
15065 Set the video size for the output. For the syntax of this option, check the
15066 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15067 Default value is @code{400x400}.
15070 Set the output frame rate. Default value is @code{25}.
15076 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15077 @code{160}, @code{80} and @code{255}.
15078 Allowed range is @code{[0, 255]}.
15084 Specify the red, green, blue and alpha fade. Default values are @code{15},
15085 @code{10}, @code{5} and @code{5}.
15086 Allowed range is @code{[0, 255]}.
15089 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15092 Set the vectorscope drawing mode.
15094 Available values are:
15097 Draw dot for each sample.
15100 Draw line between previous and current sample.
15103 Default value is @samp{dot}.
15106 Specify amplitude scale of audio samples.
15108 Available values are:
15125 @subsection Examples
15129 Complete example using @command{ffplay}:
15131 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15132 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15136 @section bench, abench
15138 Benchmark part of a filtergraph.
15140 The filter accepts the following options:
15144 Start or stop a timer.
15146 Available values are:
15149 Get the current time, set it as frame metadata (using the key
15150 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15153 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15154 the input frame metadata to get the time difference. Time difference, average,
15155 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15156 @code{min}) are then printed. The timestamps are expressed in seconds.
15160 @subsection Examples
15164 Benchmark @ref{selectivecolor} filter:
15166 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15172 Concatenate audio and video streams, joining them together one after the
15175 The filter works on segments of synchronized video and audio streams. All
15176 segments must have the same number of streams of each type, and that will
15177 also be the number of streams at output.
15179 The filter accepts the following options:
15184 Set the number of segments. Default is 2.
15187 Set the number of output video streams, that is also the number of video
15188 streams in each segment. Default is 1.
15191 Set the number of output audio streams, that is also the number of audio
15192 streams in each segment. Default is 0.
15195 Activate unsafe mode: do not fail if segments have a different format.
15199 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15200 @var{a} audio outputs.
15202 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15203 segment, in the same order as the outputs, then the inputs for the second
15206 Related streams do not always have exactly the same duration, for various
15207 reasons including codec frame size or sloppy authoring. For that reason,
15208 related synchronized streams (e.g. a video and its audio track) should be
15209 concatenated at once. The concat filter will use the duration of the longest
15210 stream in each segment (except the last one), and if necessary pad shorter
15211 audio streams with silence.
15213 For this filter to work correctly, all segments must start at timestamp 0.
15215 All corresponding streams must have the same parameters in all segments; the
15216 filtering system will automatically select a common pixel format for video
15217 streams, and a common sample format, sample rate and channel layout for
15218 audio streams, but other settings, such as resolution, must be converted
15219 explicitly by the user.
15221 Different frame rates are acceptable but will result in variable frame rate
15222 at output; be sure to configure the output file to handle it.
15224 @subsection Examples
15228 Concatenate an opening, an episode and an ending, all in bilingual version
15229 (video in stream 0, audio in streams 1 and 2):
15231 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15232 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15233 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15234 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15238 Concatenate two parts, handling audio and video separately, using the
15239 (a)movie sources, and adjusting the resolution:
15241 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15242 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15243 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15245 Note that a desync will happen at the stitch if the audio and video streams
15246 do not have exactly the same duration in the first file.
15250 @section drawgraph, adrawgraph
15252 Draw a graph using input video or audio metadata.
15254 It accepts the following parameters:
15258 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15261 Set 1st foreground color expression.
15264 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15267 Set 2nd foreground color expression.
15270 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15273 Set 3rd foreground color expression.
15276 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15279 Set 4th foreground color expression.
15282 Set minimal value of metadata value.
15285 Set maximal value of metadata value.
15288 Set graph background color. Default is white.
15293 Available values for mode is:
15300 Default is @code{line}.
15305 Available values for slide is:
15308 Draw new frame when right border is reached.
15311 Replace old columns with new ones.
15314 Scroll from right to left.
15317 Scroll from left to right.
15320 Draw single picture.
15323 Default is @code{frame}.
15326 Set size of graph video. For the syntax of this option, check the
15327 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15328 The default value is @code{900x256}.
15330 The foreground color expressions can use the following variables:
15333 Minimal value of metadata value.
15336 Maximal value of metadata value.
15339 Current metadata key value.
15342 The color is defined as 0xAABBGGRR.
15345 Example using metadata from @ref{signalstats} filter:
15347 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15350 Example using metadata from @ref{ebur128} filter:
15352 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15358 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15359 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15360 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15361 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15363 The filter also has a video output (see the @var{video} option) with a real
15364 time graph to observe the loudness evolution. The graphic contains the logged
15365 message mentioned above, so it is not printed anymore when this option is set,
15366 unless the verbose logging is set. The main graphing area contains the
15367 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15368 the momentary loudness (400 milliseconds).
15370 More information about the Loudness Recommendation EBU R128 on
15371 @url{http://tech.ebu.ch/loudness}.
15373 The filter accepts the following options:
15378 Activate the video output. The audio stream is passed unchanged whether this
15379 option is set or no. The video stream will be the first output stream if
15380 activated. Default is @code{0}.
15383 Set the video size. This option is for video only. For the syntax of this
15385 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15386 Default and minimum resolution is @code{640x480}.
15389 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15390 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15391 other integer value between this range is allowed.
15394 Set metadata injection. If set to @code{1}, the audio input will be segmented
15395 into 100ms output frames, each of them containing various loudness information
15396 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15398 Default is @code{0}.
15401 Force the frame logging level.
15403 Available values are:
15406 information logging level
15408 verbose logging level
15411 By default, the logging level is set to @var{info}. If the @option{video} or
15412 the @option{metadata} options are set, it switches to @var{verbose}.
15417 Available modes can be cumulated (the option is a @code{flag} type). Possible
15421 Disable any peak mode (default).
15423 Enable sample-peak mode.
15425 Simple peak mode looking for the higher sample value. It logs a message
15426 for sample-peak (identified by @code{SPK}).
15428 Enable true-peak mode.
15430 If enabled, the peak lookup is done on an over-sampled version of the input
15431 stream for better peak accuracy. It logs a message for true-peak.
15432 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15433 This mode requires a build with @code{libswresample}.
15437 Treat mono input files as "dual mono". If a mono file is intended for playback
15438 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15439 If set to @code{true}, this option will compensate for this effect.
15440 Multi-channel input files are not affected by this option.
15443 Set a specific pan law to be used for the measurement of dual mono files.
15444 This parameter is optional, and has a default value of -3.01dB.
15447 @subsection Examples
15451 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
15453 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
15457 Run an analysis with @command{ffmpeg}:
15459 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
15463 @section interleave, ainterleave
15465 Temporally interleave frames from several inputs.
15467 @code{interleave} works with video inputs, @code{ainterleave} with audio.
15469 These filters read frames from several inputs and send the oldest
15470 queued frame to the output.
15472 Input streams must have a well defined, monotonically increasing frame
15475 In order to submit one frame to output, these filters need to enqueue
15476 at least one frame for each input, so they cannot work in case one
15477 input is not yet terminated and will not receive incoming frames.
15479 For example consider the case when one input is a @code{select} filter
15480 which always drop input frames. The @code{interleave} filter will keep
15481 reading from that input, but it will never be able to send new frames
15482 to output until the input will send an end-of-stream signal.
15484 Also, depending on inputs synchronization, the filters will drop
15485 frames in case one input receives more frames than the other ones, and
15486 the queue is already filled.
15488 These filters accept the following options:
15492 Set the number of different inputs, it is 2 by default.
15495 @subsection Examples
15499 Interleave frames belonging to different streams using @command{ffmpeg}:
15501 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
15505 Add flickering blur effect:
15507 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
15511 @section metadata, ametadata
15513 Manipulate frame metadata.
15515 This filter accepts the following options:
15519 Set mode of operation of the filter.
15521 Can be one of the following:
15525 If both @code{value} and @code{key} is set, select frames
15526 which have such metadata. If only @code{key} is set, select
15527 every frame that has such key in metadata.
15530 Add new metadata @code{key} and @code{value}. If key is already available
15534 Modify value of already present key.
15537 If @code{value} is set, delete only keys that have such value.
15538 Otherwise, delete key.
15541 Print key and its value if metadata was found. If @code{key} is not set print all
15542 metadata values available in frame.
15546 Set key used with all modes. Must be set for all modes except @code{print}.
15549 Set metadata value which will be used. This option is mandatory for
15550 @code{modify} and @code{add} mode.
15553 Which function to use when comparing metadata value and @code{value}.
15555 Can be one of following:
15559 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
15562 Values are interpreted as strings, returns true if metadata value starts with
15563 the @code{value} option string.
15566 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
15569 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
15572 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
15575 Values are interpreted as floats, returns true if expression from option @code{expr}
15580 Set expression which is used when @code{function} is set to @code{expr}.
15581 The expression is evaluated through the eval API and can contain the following
15586 Float representation of @code{value} from metadata key.
15589 Float representation of @code{value} as supplied by user in @code{value} option.
15592 If specified in @code{print} mode, output is written to the named file. Instead of
15593 plain filename any writable url can be specified. Filename ``-'' is a shorthand
15594 for standard output. If @code{file} option is not set, output is written to the log
15595 with AV_LOG_INFO loglevel.
15600 @subsection Examples
15604 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
15607 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
15610 Print silencedetect output to file @file{metadata.txt}.
15612 silencedetect,ametadata=mode=print:file=metadata.txt
15615 Direct all metadata to a pipe with file descriptor 4.
15617 metadata=mode=print:file='pipe\:4'
15621 @section perms, aperms
15623 Set read/write permissions for the output frames.
15625 These filters are mainly aimed at developers to test direct path in the
15626 following filter in the filtergraph.
15628 The filters accept the following options:
15632 Select the permissions mode.
15634 It accepts the following values:
15637 Do nothing. This is the default.
15639 Set all the output frames read-only.
15641 Set all the output frames directly writable.
15643 Make the frame read-only if writable, and writable if read-only.
15645 Set each output frame read-only or writable randomly.
15649 Set the seed for the @var{random} mode, must be an integer included between
15650 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
15651 @code{-1}, the filter will try to use a good random seed on a best effort
15655 Note: in case of auto-inserted filter between the permission filter and the
15656 following one, the permission might not be received as expected in that
15657 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
15658 perms/aperms filter can avoid this problem.
15660 @section realtime, arealtime
15662 Slow down filtering to match real time approximatively.
15664 These filters will pause the filtering for a variable amount of time to
15665 match the output rate with the input timestamps.
15666 They are similar to the @option{re} option to @code{ffmpeg}.
15668 They accept the following options:
15672 Time limit for the pauses. Any pause longer than that will be considered
15673 a timestamp discontinuity and reset the timer. Default is 2 seconds.
15676 @section select, aselect
15678 Select frames to pass in output.
15680 This filter accepts the following options:
15685 Set expression, which is evaluated for each input frame.
15687 If the expression is evaluated to zero, the frame is discarded.
15689 If the evaluation result is negative or NaN, the frame is sent to the
15690 first output; otherwise it is sent to the output with index
15691 @code{ceil(val)-1}, assuming that the input index starts from 0.
15693 For example a value of @code{1.2} corresponds to the output with index
15694 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
15697 Set the number of outputs. The output to which to send the selected
15698 frame is based on the result of the evaluation. Default value is 1.
15701 The expression can contain the following constants:
15705 The (sequential) number of the filtered frame, starting from 0.
15708 The (sequential) number of the selected frame, starting from 0.
15710 @item prev_selected_n
15711 The sequential number of the last selected frame. It's NAN if undefined.
15714 The timebase of the input timestamps.
15717 The PTS (Presentation TimeStamp) of the filtered video frame,
15718 expressed in @var{TB} units. It's NAN if undefined.
15721 The PTS of the filtered video frame,
15722 expressed in seconds. It's NAN if undefined.
15725 The PTS of the previously filtered video frame. It's NAN if undefined.
15727 @item prev_selected_pts
15728 The PTS of the last previously filtered video frame. It's NAN if undefined.
15730 @item prev_selected_t
15731 The PTS of the last previously selected video frame. It's NAN if undefined.
15734 The PTS of the first video frame in the video. It's NAN if undefined.
15737 The time of the first video frame in the video. It's NAN if undefined.
15739 @item pict_type @emph{(video only)}
15740 The type of the filtered frame. It can assume one of the following
15752 @item interlace_type @emph{(video only)}
15753 The frame interlace type. It can assume one of the following values:
15756 The frame is progressive (not interlaced).
15758 The frame is top-field-first.
15760 The frame is bottom-field-first.
15763 @item consumed_sample_n @emph{(audio only)}
15764 the number of selected samples before the current frame
15766 @item samples_n @emph{(audio only)}
15767 the number of samples in the current frame
15769 @item sample_rate @emph{(audio only)}
15770 the input sample rate
15773 This is 1 if the filtered frame is a key-frame, 0 otherwise.
15776 the position in the file of the filtered frame, -1 if the information
15777 is not available (e.g. for synthetic video)
15779 @item scene @emph{(video only)}
15780 value between 0 and 1 to indicate a new scene; a low value reflects a low
15781 probability for the current frame to introduce a new scene, while a higher
15782 value means the current frame is more likely to be one (see the example below)
15784 @item concatdec_select
15785 The concat demuxer can select only part of a concat input file by setting an
15786 inpoint and an outpoint, but the output packets may not be entirely contained
15787 in the selected interval. By using this variable, it is possible to skip frames
15788 generated by the concat demuxer which are not exactly contained in the selected
15791 This works by comparing the frame pts against the @var{lavf.concat.start_time}
15792 and the @var{lavf.concat.duration} packet metadata values which are also
15793 present in the decoded frames.
15795 The @var{concatdec_select} variable is -1 if the frame pts is at least
15796 start_time and either the duration metadata is missing or the frame pts is less
15797 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
15800 That basically means that an input frame is selected if its pts is within the
15801 interval set by the concat demuxer.
15805 The default value of the select expression is "1".
15807 @subsection Examples
15811 Select all frames in input:
15816 The example above is the same as:
15828 Select only I-frames:
15830 select='eq(pict_type\,I)'
15834 Select one frame every 100:
15836 select='not(mod(n\,100))'
15840 Select only frames contained in the 10-20 time interval:
15842 select=between(t\,10\,20)
15846 Select only I-frames contained in the 10-20 time interval:
15848 select=between(t\,10\,20)*eq(pict_type\,I)
15852 Select frames with a minimum distance of 10 seconds:
15854 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
15858 Use aselect to select only audio frames with samples number > 100:
15860 aselect='gt(samples_n\,100)'
15864 Create a mosaic of the first scenes:
15866 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
15869 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
15873 Send even and odd frames to separate outputs, and compose them:
15875 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
15879 Select useful frames from an ffconcat file which is using inpoints and
15880 outpoints but where the source files are not intra frame only.
15882 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
15886 @section sendcmd, asendcmd
15888 Send commands to filters in the filtergraph.
15890 These filters read commands to be sent to other filters in the
15893 @code{sendcmd} must be inserted between two video filters,
15894 @code{asendcmd} must be inserted between two audio filters, but apart
15895 from that they act the same way.
15897 The specification of commands can be provided in the filter arguments
15898 with the @var{commands} option, or in a file specified by the
15899 @var{filename} option.
15901 These filters accept the following options:
15904 Set the commands to be read and sent to the other filters.
15906 Set the filename of the commands to be read and sent to the other
15910 @subsection Commands syntax
15912 A commands description consists of a sequence of interval
15913 specifications, comprising a list of commands to be executed when a
15914 particular event related to that interval occurs. The occurring event
15915 is typically the current frame time entering or leaving a given time
15918 An interval is specified by the following syntax:
15920 @var{START}[-@var{END}] @var{COMMANDS};
15923 The time interval is specified by the @var{START} and @var{END} times.
15924 @var{END} is optional and defaults to the maximum time.
15926 The current frame time is considered within the specified interval if
15927 it is included in the interval [@var{START}, @var{END}), that is when
15928 the time is greater or equal to @var{START} and is lesser than
15931 @var{COMMANDS} consists of a sequence of one or more command
15932 specifications, separated by ",", relating to that interval. The
15933 syntax of a command specification is given by:
15935 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
15938 @var{FLAGS} is optional and specifies the type of events relating to
15939 the time interval which enable sending the specified command, and must
15940 be a non-null sequence of identifier flags separated by "+" or "|" and
15941 enclosed between "[" and "]".
15943 The following flags are recognized:
15946 The command is sent when the current frame timestamp enters the
15947 specified interval. In other words, the command is sent when the
15948 previous frame timestamp was not in the given interval, and the
15952 The command is sent when the current frame timestamp leaves the
15953 specified interval. In other words, the command is sent when the
15954 previous frame timestamp was in the given interval, and the
15958 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
15961 @var{TARGET} specifies the target of the command, usually the name of
15962 the filter class or a specific filter instance name.
15964 @var{COMMAND} specifies the name of the command for the target filter.
15966 @var{ARG} is optional and specifies the optional list of argument for
15967 the given @var{COMMAND}.
15969 Between one interval specification and another, whitespaces, or
15970 sequences of characters starting with @code{#} until the end of line,
15971 are ignored and can be used to annotate comments.
15973 A simplified BNF description of the commands specification syntax
15976 @var{COMMAND_FLAG} ::= "enter" | "leave"
15977 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
15978 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
15979 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
15980 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
15981 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
15984 @subsection Examples
15988 Specify audio tempo change at second 4:
15990 asendcmd=c='4.0 atempo tempo 1.5',atempo
15994 Specify a list of drawtext and hue commands in a file.
15996 # show text in the interval 5-10
15997 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
15998 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16000 # desaturate the image in the interval 15-20
16001 15.0-20.0 [enter] hue s 0,
16002 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16004 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16006 # apply an exponential saturation fade-out effect, starting from time 25
16007 25 [enter] hue s exp(25-t)
16010 A filtergraph allowing to read and process the above command list
16011 stored in a file @file{test.cmd}, can be specified with:
16013 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16018 @section setpts, asetpts
16020 Change the PTS (presentation timestamp) of the input frames.
16022 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16024 This filter accepts the following options:
16029 The expression which is evaluated for each frame to construct its timestamp.
16033 The expression is evaluated through the eval API and can contain the following
16038 frame rate, only defined for constant frame-rate video
16041 The presentation timestamp in input
16044 The count of the input frame for video or the number of consumed samples,
16045 not including the current frame for audio, starting from 0.
16047 @item NB_CONSUMED_SAMPLES
16048 The number of consumed samples, not including the current frame (only
16051 @item NB_SAMPLES, S
16052 The number of samples in the current frame (only audio)
16054 @item SAMPLE_RATE, SR
16055 The audio sample rate.
16058 The PTS of the first frame.
16061 the time in seconds of the first frame
16064 State whether the current frame is interlaced.
16067 the time in seconds of the current frame
16070 original position in the file of the frame, or undefined if undefined
16071 for the current frame
16074 The previous input PTS.
16077 previous input time in seconds
16080 The previous output PTS.
16083 previous output time in seconds
16086 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16090 The wallclock (RTC) time at the start of the movie in microseconds.
16093 The timebase of the input timestamps.
16097 @subsection Examples
16101 Start counting PTS from zero
16103 setpts=PTS-STARTPTS
16107 Apply fast motion effect:
16113 Apply slow motion effect:
16119 Set fixed rate of 25 frames per second:
16125 Set fixed rate 25 fps with some jitter:
16127 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16131 Apply an offset of 10 seconds to the input PTS:
16137 Generate timestamps from a "live source" and rebase onto the current timebase:
16139 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16143 Generate timestamps by counting samples:
16150 @section settb, asettb
16152 Set the timebase to use for the output frames timestamps.
16153 It is mainly useful for testing timebase configuration.
16155 It accepts the following parameters:
16160 The expression which is evaluated into the output timebase.
16164 The value for @option{tb} is an arithmetic expression representing a
16165 rational. The expression can contain the constants "AVTB" (the default
16166 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16167 audio only). Default value is "intb".
16169 @subsection Examples
16173 Set the timebase to 1/25:
16179 Set the timebase to 1/10:
16185 Set the timebase to 1001/1000:
16191 Set the timebase to 2*intb:
16197 Set the default timebase value:
16204 Convert input audio to a video output representing frequency spectrum
16205 logarithmically using Brown-Puckette constant Q transform algorithm with
16206 direct frequency domain coefficient calculation (but the transform itself
16207 is not really constant Q, instead the Q factor is actually variable/clamped),
16208 with musical tone scale, from E0 to D#10.
16210 The filter accepts the following options:
16214 Specify the video size for the output. It must be even. For the syntax of this option,
16215 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16216 Default value is @code{1920x1080}.
16219 Set the output frame rate. Default value is @code{25}.
16222 Set the bargraph height. It must be even. Default value is @code{-1} which
16223 computes the bargraph height automatically.
16226 Set the axis height. It must be even. Default value is @code{-1} which computes
16227 the axis height automatically.
16230 Set the sonogram height. It must be even. Default value is @code{-1} which
16231 computes the sonogram height automatically.
16234 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16235 instead. Default value is @code{1}.
16237 @item sono_v, volume
16238 Specify the sonogram volume expression. It can contain variables:
16241 the @var{bar_v} evaluated expression
16242 @item frequency, freq, f
16243 the frequency where it is evaluated
16244 @item timeclamp, tc
16245 the value of @var{timeclamp} option
16249 @item a_weighting(f)
16250 A-weighting of equal loudness
16251 @item b_weighting(f)
16252 B-weighting of equal loudness
16253 @item c_weighting(f)
16254 C-weighting of equal loudness.
16256 Default value is @code{16}.
16258 @item bar_v, volume2
16259 Specify the bargraph volume expression. It can contain variables:
16262 the @var{sono_v} evaluated expression
16263 @item frequency, freq, f
16264 the frequency where it is evaluated
16265 @item timeclamp, tc
16266 the value of @var{timeclamp} option
16270 @item a_weighting(f)
16271 A-weighting of equal loudness
16272 @item b_weighting(f)
16273 B-weighting of equal loudness
16274 @item c_weighting(f)
16275 C-weighting of equal loudness.
16277 Default value is @code{sono_v}.
16279 @item sono_g, gamma
16280 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16281 higher gamma makes the spectrum having more range. Default value is @code{3}.
16282 Acceptable range is @code{[1, 7]}.
16284 @item bar_g, gamma2
16285 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16288 @item timeclamp, tc
16289 Specify the transform timeclamp. At low frequency, there is trade-off between
16290 accuracy in time domain and frequency domain. If timeclamp is lower,
16291 event in time domain is represented more accurately (such as fast bass drum),
16292 otherwise event in frequency domain is represented more accurately
16293 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16296 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16297 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16300 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16301 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16304 This option is deprecated and ignored.
16307 Specify the transform length in time domain. Use this option to control accuracy
16308 trade-off between time domain and frequency domain at every frequency sample.
16309 It can contain variables:
16311 @item frequency, freq, f
16312 the frequency where it is evaluated
16313 @item timeclamp, tc
16314 the value of @var{timeclamp} option.
16316 Default value is @code{384*tc/(384+tc*f)}.
16319 Specify the transform count for every video frame. Default value is @code{6}.
16320 Acceptable range is @code{[1, 30]}.
16323 Specify the transform count for every single pixel. Default value is @code{0},
16324 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16327 Specify font file for use with freetype to draw the axis. If not specified,
16328 use embedded font. Note that drawing with font file or embedded font is not
16329 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16333 Specify font color expression. This is arithmetic expression that should return
16334 integer value 0xRRGGBB. It can contain variables:
16336 @item frequency, freq, f
16337 the frequency where it is evaluated
16338 @item timeclamp, tc
16339 the value of @var{timeclamp} option
16344 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16345 @item r(x), g(x), b(x)
16346 red, green, and blue value of intensity x.
16348 Default value is @code{st(0, (midi(f)-59.5)/12);
16349 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16350 r(1-ld(1)) + b(ld(1))}.
16353 Specify image file to draw the axis. This option override @var{fontfile} and
16354 @var{fontcolor} option.
16357 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16358 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16359 Default value is @code{1}.
16363 @subsection Examples
16367 Playing audio while showing the spectrum:
16369 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16373 Same as above, but with frame rate 30 fps:
16375 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16379 Playing at 1280x720:
16381 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16385 Disable sonogram display:
16391 A1 and its harmonics: A1, A2, (near)E3, A3:
16393 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),
16394 asplit[a][out1]; [a] showcqt [out0]'
16398 Same as above, but with more accuracy in frequency domain:
16400 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),
16401 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16407 bar_v=10:sono_v=bar_v*a_weighting(f)
16411 Custom gamma, now spectrum is linear to the amplitude.
16417 Custom tlength equation:
16419 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)))'
16423 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
16425 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
16429 Custom frequency range with custom axis using image file:
16431 axisfile=myaxis.png:basefreq=40:endfreq=10000
16437 Convert input audio to video output representing the audio power spectrum.
16438 Audio amplitude is on Y-axis while frequency is on X-axis.
16440 The filter accepts the following options:
16444 Specify size of video. For the syntax of this option, check the
16445 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16446 Default is @code{1024x512}.
16450 This set how each frequency bin will be represented.
16452 It accepts the following values:
16458 Default is @code{bar}.
16461 Set amplitude scale.
16463 It accepts the following values:
16477 Default is @code{log}.
16480 Set frequency scale.
16482 It accepts the following values:
16491 Reverse logarithmic scale.
16493 Default is @code{lin}.
16498 It accepts the following values:
16514 Default is @code{w2048}
16517 Set windowing function.
16519 It accepts the following values:
16541 Default is @code{hanning}.
16544 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
16545 which means optimal overlap for selected window function will be picked.
16548 Set time averaging. Setting this to 0 will display current maximal peaks.
16549 Default is @code{1}, which means time averaging is disabled.
16552 Specify list of colors separated by space or by '|' which will be used to
16553 draw channel frequencies. Unrecognized or missing colors will be replaced
16557 Set channel display mode.
16559 It accepts the following values:
16564 Default is @code{combined}.
16567 Set minimum amplitude used in @code{log} amplitude scaler.
16571 @anchor{showspectrum}
16572 @section showspectrum
16574 Convert input audio to a video output, representing the audio frequency
16577 The filter accepts the following options:
16581 Specify the video size for the output. For the syntax of this option, check the
16582 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16583 Default value is @code{640x512}.
16586 Specify how the spectrum should slide along the window.
16588 It accepts the following values:
16591 the samples start again on the left when they reach the right
16593 the samples scroll from right to left
16595 the samples scroll from left to right
16597 frames are only produced when the samples reach the right
16600 Default value is @code{replace}.
16603 Specify display mode.
16605 It accepts the following values:
16608 all channels are displayed in the same row
16610 all channels are displayed in separate rows
16613 Default value is @samp{combined}.
16616 Specify display color mode.
16618 It accepts the following values:
16621 each channel is displayed in a separate color
16623 each channel is displayed using the same color scheme
16625 each channel is displayed using the rainbow color scheme
16627 each channel is displayed using the moreland color scheme
16629 each channel is displayed using the nebulae color scheme
16631 each channel is displayed using the fire color scheme
16633 each channel is displayed using the fiery color scheme
16635 each channel is displayed using the fruit color scheme
16637 each channel is displayed using the cool color scheme
16640 Default value is @samp{channel}.
16643 Specify scale used for calculating intensity color values.
16645 It accepts the following values:
16650 square root, default
16661 Default value is @samp{sqrt}.
16664 Set saturation modifier for displayed colors. Negative values provide
16665 alternative color scheme. @code{0} is no saturation at all.
16666 Saturation must be in [-10.0, 10.0] range.
16667 Default value is @code{1}.
16670 Set window function.
16672 It accepts the following values:
16696 Default value is @code{hann}.
16699 Set orientation of time vs frequency axis. Can be @code{vertical} or
16700 @code{horizontal}. Default is @code{vertical}.
16703 Set ratio of overlap window. Default value is @code{0}.
16704 When value is @code{1} overlap is set to recommended size for specific
16705 window function currently used.
16708 Set scale gain for calculating intensity color values.
16709 Default value is @code{1}.
16712 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
16715 Set color rotation, must be in [-1.0, 1.0] range.
16716 Default value is @code{0}.
16719 The usage is very similar to the showwaves filter; see the examples in that
16722 @subsection Examples
16726 Large window with logarithmic color scaling:
16728 showspectrum=s=1280x480:scale=log
16732 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
16734 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
16735 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
16739 @section showspectrumpic
16741 Convert input audio to a single video frame, representing the audio frequency
16744 The filter accepts the following options:
16748 Specify the video size for the output. For the syntax of this option, check the
16749 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16750 Default value is @code{4096x2048}.
16753 Specify display mode.
16755 It accepts the following values:
16758 all channels are displayed in the same row
16760 all channels are displayed in separate rows
16762 Default value is @samp{combined}.
16765 Specify display color mode.
16767 It accepts the following values:
16770 each channel is displayed in a separate color
16772 each channel is displayed using the same color scheme
16774 each channel is displayed using the rainbow color scheme
16776 each channel is displayed using the moreland color scheme
16778 each channel is displayed using the nebulae color scheme
16780 each channel is displayed using the fire color scheme
16782 each channel is displayed using the fiery color scheme
16784 each channel is displayed using the fruit color scheme
16786 each channel is displayed using the cool color scheme
16788 Default value is @samp{intensity}.
16791 Specify scale used for calculating intensity color values.
16793 It accepts the following values:
16798 square root, default
16808 Default value is @samp{log}.
16811 Set saturation modifier for displayed colors. Negative values provide
16812 alternative color scheme. @code{0} is no saturation at all.
16813 Saturation must be in [-10.0, 10.0] range.
16814 Default value is @code{1}.
16817 Set window function.
16819 It accepts the following values:
16842 Default value is @code{hann}.
16845 Set orientation of time vs frequency axis. Can be @code{vertical} or
16846 @code{horizontal}. Default is @code{vertical}.
16849 Set scale gain for calculating intensity color values.
16850 Default value is @code{1}.
16853 Draw time and frequency axes and legends. Default is enabled.
16856 Set color rotation, must be in [-1.0, 1.0] range.
16857 Default value is @code{0}.
16860 @subsection Examples
16864 Extract an audio spectrogram of a whole audio track
16865 in a 1024x1024 picture using @command{ffmpeg}:
16867 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
16871 @section showvolume
16873 Convert input audio volume to a video output.
16875 The filter accepts the following options:
16882 Set border width, allowed range is [0, 5]. Default is 1.
16885 Set channel width, allowed range is [80, 8192]. Default is 400.
16888 Set channel height, allowed range is [1, 900]. Default is 20.
16891 Set fade, allowed range is [0.001, 1]. Default is 0.95.
16894 Set volume color expression.
16896 The expression can use the following variables:
16900 Current max volume of channel in dB.
16906 Current channel number, starting from 0.
16910 If set, displays channel names. Default is enabled.
16913 If set, displays volume values. Default is enabled.
16916 Set orientation, can be @code{horizontal} or @code{vertical},
16917 default is @code{horizontal}.
16920 Set step size, allowed range s [0, 5]. Default is 0, which means
16926 Convert input audio to a video output, representing the samples waves.
16928 The filter accepts the following options:
16932 Specify the video size for the output. For the syntax of this option, check the
16933 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16934 Default value is @code{600x240}.
16939 Available values are:
16942 Draw a point for each sample.
16945 Draw a vertical line for each sample.
16948 Draw a point for each sample and a line between them.
16951 Draw a centered vertical line for each sample.
16954 Default value is @code{point}.
16957 Set the number of samples which are printed on the same column. A
16958 larger value will decrease the frame rate. Must be a positive
16959 integer. This option can be set only if the value for @var{rate}
16960 is not explicitly specified.
16963 Set the (approximate) output frame rate. This is done by setting the
16964 option @var{n}. Default value is "25".
16966 @item split_channels
16967 Set if channels should be drawn separately or overlap. Default value is 0.
16970 Set colors separated by '|' which are going to be used for drawing of each channel.
16973 Set amplitude scale.
16975 Available values are:
16993 @subsection Examples
16997 Output the input file audio and the corresponding video representation
17000 amovie=a.mp3,asplit[out0],showwaves[out1]
17004 Create a synthetic signal and show it with showwaves, forcing a
17005 frame rate of 30 frames per second:
17007 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17011 @section showwavespic
17013 Convert input audio to a single video frame, representing the samples waves.
17015 The filter accepts the following options:
17019 Specify the video size for the output. For the syntax of this option, check the
17020 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17021 Default value is @code{600x240}.
17023 @item split_channels
17024 Set if channels should be drawn separately or overlap. Default value is 0.
17027 Set colors separated by '|' which are going to be used for drawing of each channel.
17030 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17034 @subsection Examples
17038 Extract a channel split representation of the wave form of a whole audio track
17039 in a 1024x800 picture using @command{ffmpeg}:
17041 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17045 @section spectrumsynth
17047 Sythesize audio from 2 input video spectrums, first input stream represents
17048 magnitude across time and second represents phase across time.
17049 The filter will transform from frequency domain as displayed in videos back
17050 to time domain as presented in audio output.
17052 This filter is primarly created for reversing processed @ref{showspectrum}
17053 filter outputs, but can synthesize sound from other spectrograms too.
17054 But in such case results are going to be poor if the phase data is not
17055 available, because in such cases phase data need to be recreated, usually
17056 its just recreated from random noise.
17057 For best results use gray only output (@code{channel} color mode in
17058 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17059 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17060 @code{data} option. Inputs videos should generally use @code{fullframe}
17061 slide mode as that saves resources needed for decoding video.
17063 The filter accepts the following options:
17067 Specify sample rate of output audio, the sample rate of audio from which
17068 spectrum was generated may differ.
17071 Set number of channels represented in input video spectrums.
17074 Set scale which was used when generating magnitude input spectrum.
17075 Can be @code{lin} or @code{log}. Default is @code{log}.
17078 Set slide which was used when generating inputs spectrums.
17079 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17080 Default is @code{fullframe}.
17083 Set window function used for resynthesis.
17086 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17087 which means optimal overlap for selected window function will be picked.
17090 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17091 Default is @code{vertical}.
17094 @subsection Examples
17098 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17099 then resynthesize videos back to audio with spectrumsynth:
17101 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
17102 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
17103 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17107 @section split, asplit
17109 Split input into several identical outputs.
17111 @code{asplit} works with audio input, @code{split} with video.
17113 The filter accepts a single parameter which specifies the number of outputs. If
17114 unspecified, it defaults to 2.
17116 @subsection Examples
17120 Create two separate outputs from the same input:
17122 [in] split [out0][out1]
17126 To create 3 or more outputs, you need to specify the number of
17129 [in] asplit=3 [out0][out1][out2]
17133 Create two separate outputs from the same input, one cropped and
17136 [in] split [splitout1][splitout2];
17137 [splitout1] crop=100:100:0:0 [cropout];
17138 [splitout2] pad=200:200:100:100 [padout];
17142 Create 5 copies of the input audio with @command{ffmpeg}:
17144 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17150 Receive commands sent through a libzmq client, and forward them to
17151 filters in the filtergraph.
17153 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17154 must be inserted between two video filters, @code{azmq} between two
17157 To enable these filters you need to install the libzmq library and
17158 headers and configure FFmpeg with @code{--enable-libzmq}.
17160 For more information about libzmq see:
17161 @url{http://www.zeromq.org/}
17163 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17164 receives messages sent through a network interface defined by the
17165 @option{bind_address} option.
17167 The received message must be in the form:
17169 @var{TARGET} @var{COMMAND} [@var{ARG}]
17172 @var{TARGET} specifies the target of the command, usually the name of
17173 the filter class or a specific filter instance name.
17175 @var{COMMAND} specifies the name of the command for the target filter.
17177 @var{ARG} is optional and specifies the optional argument list for the
17178 given @var{COMMAND}.
17180 Upon reception, the message is processed and the corresponding command
17181 is injected into the filtergraph. Depending on the result, the filter
17182 will send a reply to the client, adopting the format:
17184 @var{ERROR_CODE} @var{ERROR_REASON}
17188 @var{MESSAGE} is optional.
17190 @subsection Examples
17192 Look at @file{tools/zmqsend} for an example of a zmq client which can
17193 be used to send commands processed by these filters.
17195 Consider the following filtergraph generated by @command{ffplay}
17197 ffplay -dumpgraph 1 -f lavfi "
17198 color=s=100x100:c=red [l];
17199 color=s=100x100:c=blue [r];
17200 nullsrc=s=200x100, zmq [bg];
17201 [bg][l] overlay [bg+l];
17202 [bg+l][r] overlay=x=100 "
17205 To change the color of the left side of the video, the following
17206 command can be used:
17208 echo Parsed_color_0 c yellow | tools/zmqsend
17211 To change the right side:
17213 echo Parsed_color_1 c pink | tools/zmqsend
17216 @c man end MULTIMEDIA FILTERS
17218 @chapter Multimedia Sources
17219 @c man begin MULTIMEDIA SOURCES
17221 Below is a description of the currently available multimedia sources.
17225 This is the same as @ref{movie} source, except it selects an audio
17231 Read audio and/or video stream(s) from a movie container.
17233 It accepts the following parameters:
17237 The name of the resource to read (not necessarily a file; it can also be a
17238 device or a stream accessed through some protocol).
17240 @item format_name, f
17241 Specifies the format assumed for the movie to read, and can be either
17242 the name of a container or an input device. If not specified, the
17243 format is guessed from @var{movie_name} or by probing.
17245 @item seek_point, sp
17246 Specifies the seek point in seconds. The frames will be output
17247 starting from this seek point. The parameter is evaluated with
17248 @code{av_strtod}, so the numerical value may be suffixed by an IS
17249 postfix. The default value is "0".
17252 Specifies the streams to read. Several streams can be specified,
17253 separated by "+". The source will then have as many outputs, in the
17254 same order. The syntax is explained in the ``Stream specifiers''
17255 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17256 respectively the default (best suited) video and audio stream. Default
17257 is "dv", or "da" if the filter is called as "amovie".
17259 @item stream_index, si
17260 Specifies the index of the video stream to read. If the value is -1,
17261 the most suitable video stream will be automatically selected. The default
17262 value is "-1". Deprecated. If the filter is called "amovie", it will select
17263 audio instead of video.
17266 Specifies how many times to read the stream in sequence.
17267 If the value is less than 1, the stream will be read again and again.
17268 Default value is "1".
17270 Note that when the movie is looped the source timestamps are not
17271 changed, so it will generate non monotonically increasing timestamps.
17273 @item discontinuity
17274 Specifies the time difference between frames above which the point is
17275 considered a timestamp discontinuity which is removed by adjusting the later
17279 It allows overlaying a second video on top of the main input of
17280 a filtergraph, as shown in this graph:
17282 input -----------> deltapts0 --> overlay --> output
17285 movie --> scale--> deltapts1 -------+
17287 @subsection Examples
17291 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17292 on top of the input labelled "in":
17294 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17295 [in] setpts=PTS-STARTPTS [main];
17296 [main][over] overlay=16:16 [out]
17300 Read from a video4linux2 device, and overlay it on top of the input
17303 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17304 [in] setpts=PTS-STARTPTS [main];
17305 [main][over] overlay=16:16 [out]
17309 Read the first video stream and the audio stream with id 0x81 from
17310 dvd.vob; the video is connected to the pad named "video" and the audio is
17311 connected to the pad named "audio":
17313 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17317 @subsection Commands
17319 Both movie and amovie support the following commands:
17322 Perform seek using "av_seek_frame".
17323 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17326 @var{stream_index}: If stream_index is -1, a default
17327 stream is selected, and @var{timestamp} is automatically converted
17328 from AV_TIME_BASE units to the stream specific time_base.
17330 @var{timestamp}: Timestamp in AVStream.time_base units
17331 or, if no stream is specified, in AV_TIME_BASE units.
17333 @var{flags}: Flags which select direction and seeking mode.
17337 Get movie duration in AV_TIME_BASE units.
17341 @c man end MULTIMEDIA SOURCES