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 dividing 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 rise 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 by 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 @code{rms}. Can be @code{peak} or @code{rms}.
964 Choose if the average level between all channels or the louder channel affects
966 Default is @code{average}. Can be @code{average} or @code{maximum}.
971 The limiter prevents an input signal from rising 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 the 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
2557 If enabled, use fixed number of audio samples. This improves speed when
2558 filtering with large delay. Default is disabled.
2561 Enable multichannels evaluation on gain. Default is disabled.
2564 Enable zero phase mode by subtracting timestamp to compensate delay.
2565 Default is disabled.
2568 @subsection Examples
2573 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2576 lowpass at 1000 Hz with gain_entry:
2578 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2581 custom equalization:
2583 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2586 higher delay with zero phase to compensate delay:
2588 firequalizer=delay=0.1:fixed=on:zero_phase=on
2591 lowpass on left channel, highpass on right channel:
2593 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2594 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2599 Apply a flanging effect to the audio.
2601 The filter accepts the following options:
2605 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2608 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2611 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2615 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2616 Default value is 71.
2619 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2622 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2623 Default value is @var{sinusoidal}.
2626 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2627 Default value is 25.
2630 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2631 Default is @var{linear}.
2636 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2637 embedded HDCD codes is expanded into a 20-bit PCM stream.
2639 The filter supports the Peak Extend and Low-level Gain Adjustment features
2640 of HDCD, and detects the Transient Filter flag.
2643 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2646 When using the filter with wav, note the default encoding for wav is 16-bit,
2647 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2648 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2650 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2651 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2654 The filter accepts the following options:
2657 @item disable_autoconvert
2658 Disable any automatic format conversion or resampling in the filter graph.
2660 @item process_stereo
2661 Process the stereo channels together. If target_gain does not match between
2662 channels, consider it invalid and use the last valid target_gain.
2665 Set the code detect timer period in ms.
2668 Always extend peaks above -3dBFS even if PE isn't signaled.
2671 Replace audio with a solid tone and adjust the amplitude to signal some
2672 specific aspect of the decoding process. The output file can be loaded in
2673 an audio editor alongside the original to aid analysis.
2675 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2682 Gain adjustment level at each sample
2684 Samples where peak extend occurs
2686 Samples where the code detect timer is active
2688 Samples where the target gain does not match between channels
2694 Apply a high-pass filter with 3dB point frequency.
2695 The filter can be either single-pole, or double-pole (the default).
2696 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2698 The filter accepts the following options:
2702 Set frequency in Hz. Default is 3000.
2705 Set number of poles. Default is 2.
2708 Set method to specify band-width of filter.
2721 Specify the band-width of a filter in width_type units.
2722 Applies only to double-pole filter.
2723 The default is 0.707q and gives a Butterworth response.
2728 Join multiple input streams into one multi-channel stream.
2730 It accepts the following parameters:
2734 The number of input streams. It defaults to 2.
2736 @item channel_layout
2737 The desired output channel layout. It defaults to stereo.
2740 Map channels from inputs to output. The argument is a '|'-separated list of
2741 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2742 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2743 can be either the name of the input channel (e.g. FL for front left) or its
2744 index in the specified input stream. @var{out_channel} is the name of the output
2748 The filter will attempt to guess the mappings when they are not specified
2749 explicitly. It does so by first trying to find an unused matching input channel
2750 and if that fails it picks the first unused input channel.
2752 Join 3 inputs (with properly set channel layouts):
2754 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2757 Build a 5.1 output from 6 single-channel streams:
2759 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2760 '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'
2766 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2768 To enable compilation of this filter you need to configure FFmpeg with
2769 @code{--enable-ladspa}.
2773 Specifies the name of LADSPA plugin library to load. If the environment
2774 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2775 each one of the directories specified by the colon separated list in
2776 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2777 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2778 @file{/usr/lib/ladspa/}.
2781 Specifies the plugin within the library. Some libraries contain only
2782 one plugin, but others contain many of them. If this is not set filter
2783 will list all available plugins within the specified library.
2786 Set the '|' separated list of controls which are zero or more floating point
2787 values that determine the behavior of the loaded plugin (for example delay,
2789 Controls need to be defined using the following syntax:
2790 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2791 @var{valuei} is the value set on the @var{i}-th control.
2792 Alternatively they can be also defined using the following syntax:
2793 @var{value0}|@var{value1}|@var{value2}|..., where
2794 @var{valuei} is the value set on the @var{i}-th control.
2795 If @option{controls} is set to @code{help}, all available controls and
2796 their valid ranges are printed.
2798 @item sample_rate, s
2799 Specify the sample rate, default to 44100. Only used if plugin have
2803 Set the number of samples per channel per each output frame, default
2804 is 1024. Only used if plugin have zero inputs.
2807 Set the minimum duration of the sourced audio. See
2808 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2809 for the accepted syntax.
2810 Note that the resulting duration may be greater than the specified duration,
2811 as the generated audio is always cut at the end of a complete frame.
2812 If not specified, or the expressed duration is negative, the audio is
2813 supposed to be generated forever.
2814 Only used if plugin have zero inputs.
2818 @subsection Examples
2822 List all available plugins within amp (LADSPA example plugin) library:
2828 List all available controls and their valid ranges for @code{vcf_notch}
2829 plugin from @code{VCF} library:
2831 ladspa=f=vcf:p=vcf_notch:c=help
2835 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2838 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2842 Add reverberation to the audio using TAP-plugins
2843 (Tom's Audio Processing plugins):
2845 ladspa=file=tap_reverb:tap_reverb
2849 Generate white noise, with 0.2 amplitude:
2851 ladspa=file=cmt:noise_source_white:c=c0=.2
2855 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2856 @code{C* Audio Plugin Suite} (CAPS) library:
2858 ladspa=file=caps:Click:c=c1=20'
2862 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2864 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2868 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2869 @code{SWH Plugins} collection:
2871 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2875 Attenuate low frequencies using Multiband EQ from Steve Harris
2876 @code{SWH Plugins} collection:
2878 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2882 @subsection Commands
2884 This filter supports the following commands:
2887 Modify the @var{N}-th control value.
2889 If the specified value is not valid, it is ignored and prior one is kept.
2894 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2895 Support for both single pass (livestreams, files) and double pass (files) modes.
2896 This algorithm can target IL, LRA, and maximum true peak.
2898 To enable compilation of this filter you need to configure FFmpeg with
2899 @code{--enable-libebur128}.
2901 The filter accepts the following options:
2905 Set integrated loudness target.
2906 Range is -70.0 - -5.0. Default value is -24.0.
2909 Set loudness range target.
2910 Range is 1.0 - 20.0. Default value is 7.0.
2913 Set maximum true peak.
2914 Range is -9.0 - +0.0. Default value is -2.0.
2916 @item measured_I, measured_i
2917 Measured IL of input file.
2918 Range is -99.0 - +0.0.
2920 @item measured_LRA, measured_lra
2921 Measured LRA of input file.
2922 Range is 0.0 - 99.0.
2924 @item measured_TP, measured_tp
2925 Measured true peak of input file.
2926 Range is -99.0 - +99.0.
2928 @item measured_thresh
2929 Measured threshold of input file.
2930 Range is -99.0 - +0.0.
2933 Set offset gain. Gain is applied before the true-peak limiter.
2934 Range is -99.0 - +99.0. Default is +0.0.
2937 Normalize linearly if possible.
2938 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2939 to be specified in order to use this mode.
2940 Options are true or false. Default is true.
2943 Treat mono input files as "dual-mono". If a mono file is intended for playback
2944 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2945 If set to @code{true}, this option will compensate for this effect.
2946 Multi-channel input files are not affected by this option.
2947 Options are true or false. Default is false.
2950 Set print format for stats. Options are summary, json, or none.
2951 Default value is none.
2956 Apply a low-pass filter with 3dB point frequency.
2957 The filter can be either single-pole or double-pole (the default).
2958 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2960 The filter accepts the following options:
2964 Set frequency in Hz. Default is 500.
2967 Set number of poles. Default is 2.
2970 Set method to specify band-width of filter.
2983 Specify the band-width of a filter in width_type units.
2984 Applies only to double-pole filter.
2985 The default is 0.707q and gives a Butterworth response.
2991 Mix channels with specific gain levels. The filter accepts the output
2992 channel layout followed by a set of channels definitions.
2994 This filter is also designed to efficiently remap the channels of an audio
2997 The filter accepts parameters of the form:
2998 "@var{l}|@var{outdef}|@var{outdef}|..."
3002 output channel layout or number of channels
3005 output channel specification, of the form:
3006 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
3009 output channel to define, either a channel name (FL, FR, etc.) or a channel
3010 number (c0, c1, etc.)
3013 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3016 input channel to use, see out_name for details; it is not possible to mix
3017 named and numbered input channels
3020 If the `=' in a channel specification is replaced by `<', then the gains for
3021 that specification will be renormalized so that the total is 1, thus
3022 avoiding clipping noise.
3024 @subsection Mixing examples
3026 For example, if you want to down-mix from stereo to mono, but with a bigger
3027 factor for the left channel:
3029 pan=1c|c0=0.9*c0+0.1*c1
3032 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3033 7-channels surround:
3035 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3038 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3039 that should be preferred (see "-ac" option) unless you have very specific
3042 @subsection Remapping examples
3044 The channel remapping will be effective if, and only if:
3047 @item gain coefficients are zeroes or ones,
3048 @item only one input per channel output,
3051 If all these conditions are satisfied, the filter will notify the user ("Pure
3052 channel mapping detected"), and use an optimized and lossless method to do the
3055 For example, if you have a 5.1 source and want a stereo audio stream by
3056 dropping the extra channels:
3058 pan="stereo| c0=FL | c1=FR"
3061 Given the same source, you can also switch front left and front right channels
3062 and keep the input channel layout:
3064 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3067 If the input is a stereo audio stream, you can mute the front left channel (and
3068 still keep the stereo channel layout) with:
3073 Still with a stereo audio stream input, you can copy the right channel in both
3074 front left and right:
3076 pan="stereo| c0=FR | c1=FR"
3081 ReplayGain scanner filter. This filter takes an audio stream as an input and
3082 outputs it unchanged.
3083 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3087 Convert the audio sample format, sample rate and channel layout. It is
3088 not meant to be used directly.
3091 Apply time-stretching and pitch-shifting with librubberband.
3093 The filter accepts the following options:
3097 Set tempo scale factor.
3100 Set pitch scale factor.
3103 Set transients detector.
3104 Possible values are:
3113 Possible values are:
3122 Possible values are:
3129 Set processing window size.
3130 Possible values are:
3139 Possible values are:
3146 Enable formant preservation when shift pitching.
3147 Possible values are:
3155 Possible values are:
3164 Possible values are:
3171 @section sidechaincompress
3173 This filter acts like normal compressor but has the ability to compress
3174 detected signal using second input signal.
3175 It needs two input streams and returns one output stream.
3176 First input stream will be processed depending on second stream signal.
3177 The filtered signal then can be filtered with other filters in later stages of
3178 processing. See @ref{pan} and @ref{amerge} filter.
3180 The filter accepts the following options:
3184 Set input gain. Default is 1. Range is between 0.015625 and 64.
3187 If a signal of second stream raises above this level it will affect the gain
3188 reduction of first stream.
3189 By default is 0.125. Range is between 0.00097563 and 1.
3192 Set a ratio about which the signal is reduced. 1:2 means that if the level
3193 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3194 Default is 2. Range is between 1 and 20.
3197 Amount of milliseconds the signal has to rise above the threshold before gain
3198 reduction starts. Default is 20. Range is between 0.01 and 2000.
3201 Amount of milliseconds the signal has to fall below the threshold before
3202 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3205 Set the amount by how much signal will be amplified after processing.
3206 Default is 2. Range is from 1 and 64.
3209 Curve the sharp knee around the threshold to enter gain reduction more softly.
3210 Default is 2.82843. Range is between 1 and 8.
3213 Choose if the @code{average} level between all channels of side-chain stream
3214 or the louder(@code{maximum}) channel of side-chain stream affects the
3215 reduction. Default is @code{average}.
3218 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3219 of @code{rms}. Default is @code{rms} which is mainly smoother.
3222 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3225 How much to use compressed signal in output. Default is 1.
3226 Range is between 0 and 1.
3229 @subsection Examples
3233 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3234 depending on the signal of 2nd input and later compressed signal to be
3235 merged with 2nd input:
3237 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3241 @section sidechaingate
3243 A sidechain gate acts like a normal (wideband) gate but has the ability to
3244 filter the detected signal before sending it to the gain reduction stage.
3245 Normally a gate uses the full range signal to detect a level above the
3247 For example: If you cut all lower frequencies from your sidechain signal
3248 the gate will decrease the volume of your track only if not enough highs
3249 appear. With this technique you are able to reduce the resonation of a
3250 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3252 It needs two input streams and returns one output stream.
3253 First input stream will be processed depending on second stream signal.
3255 The filter accepts the following options:
3259 Set input level before filtering.
3260 Default is 1. Allowed range is from 0.015625 to 64.
3263 Set the level of gain reduction when the signal is below the threshold.
3264 Default is 0.06125. Allowed range is from 0 to 1.
3267 If a signal rises above this level the gain reduction is released.
3268 Default is 0.125. Allowed range is from 0 to 1.
3271 Set a ratio about which the signal is reduced.
3272 Default is 2. Allowed range is from 1 to 9000.
3275 Amount of milliseconds the signal has to rise above the threshold before gain
3277 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3280 Amount of milliseconds the signal has to fall below the threshold before the
3281 reduction is increased again. Default is 250 milliseconds.
3282 Allowed range is from 0.01 to 9000.
3285 Set amount of amplification of signal after processing.
3286 Default is 1. Allowed range is from 1 to 64.
3289 Curve the sharp knee around the threshold to enter gain reduction more softly.
3290 Default is 2.828427125. Allowed range is from 1 to 8.
3293 Choose if exact signal should be taken for detection or an RMS like one.
3294 Default is rms. Can be peak or rms.
3297 Choose if the average level between all channels or the louder channel affects
3299 Default is average. Can be average or maximum.
3302 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3305 @section silencedetect
3307 Detect silence in an audio stream.
3309 This filter logs a message when it detects that the input audio volume is less
3310 or equal to a noise tolerance value for a duration greater or equal to the
3311 minimum detected noise duration.
3313 The printed times and duration are expressed in seconds.
3315 The filter accepts the following options:
3319 Set silence duration until notification (default is 2 seconds).
3322 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3323 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3326 @subsection Examples
3330 Detect 5 seconds of silence with -50dB noise tolerance:
3332 silencedetect=n=-50dB:d=5
3336 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3337 tolerance in @file{silence.mp3}:
3339 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3343 @section silenceremove
3345 Remove silence from the beginning, middle or end of the audio.
3347 The filter accepts the following options:
3351 This value is used to indicate if audio should be trimmed at beginning of
3352 the audio. A value of zero indicates no silence should be trimmed from the
3353 beginning. When specifying a non-zero value, it trims audio up until it
3354 finds non-silence. Normally, when trimming silence from beginning of audio
3355 the @var{start_periods} will be @code{1} but it can be increased to higher
3356 values to trim all audio up to specific count of non-silence periods.
3357 Default value is @code{0}.
3359 @item start_duration
3360 Specify the amount of time that non-silence must be detected before it stops
3361 trimming audio. By increasing the duration, bursts of noises can be treated
3362 as silence and trimmed off. Default value is @code{0}.
3364 @item start_threshold
3365 This indicates what sample value should be treated as silence. For digital
3366 audio, a value of @code{0} may be fine but for audio recorded from analog,
3367 you may wish to increase the value to account for background noise.
3368 Can be specified in dB (in case "dB" is appended to the specified value)
3369 or amplitude ratio. Default value is @code{0}.
3372 Set the count for trimming silence from the end of audio.
3373 To remove silence from the middle of a file, specify a @var{stop_periods}
3374 that is negative. This value is then treated as a positive value and is
3375 used to indicate the effect should restart processing as specified by
3376 @var{start_periods}, making it suitable for removing periods of silence
3377 in the middle of the audio.
3378 Default value is @code{0}.
3381 Specify a duration of silence that must exist before audio is not copied any
3382 more. By specifying a higher duration, silence that is wanted can be left in
3384 Default value is @code{0}.
3386 @item stop_threshold
3387 This is the same as @option{start_threshold} but for trimming silence from
3389 Can be specified in dB (in case "dB" is appended to the specified value)
3390 or amplitude ratio. Default value is @code{0}.
3393 This indicates that @var{stop_duration} length of audio should be left intact
3394 at the beginning of each period of silence.
3395 For example, if you want to remove long pauses between words but do not want
3396 to remove the pauses completely. Default value is @code{0}.
3399 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3400 and works better with digital silence which is exactly 0.
3401 Default value is @code{rms}.
3404 Set ratio used to calculate size of window for detecting silence.
3405 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3408 @subsection Examples
3412 The following example shows how this filter can be used to start a recording
3413 that does not contain the delay at the start which usually occurs between
3414 pressing the record button and the start of the performance:
3416 silenceremove=1:5:0.02
3420 Trim all silence encountered from beginning to end where there is more than 1
3421 second of silence in audio:
3423 silenceremove=0:0:0:-1:1:-90dB
3429 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3430 loudspeakers around the user for binaural listening via headphones (audio
3431 formats up to 9 channels supported).
3432 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3433 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3434 Austrian Academy of Sciences.
3436 To enable compilation of this filter you need to configure FFmpeg with
3437 @code{--enable-netcdf}.
3439 The filter accepts the following options:
3443 Set the SOFA file used for rendering.
3446 Set gain applied to audio. Value is in dB. Default is 0.
3449 Set rotation of virtual loudspeakers in deg. Default is 0.
3452 Set elevation of virtual speakers in deg. Default is 0.
3455 Set distance in meters between loudspeakers and the listener with near-field
3456 HRTFs. Default is 1.
3459 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3460 processing audio in time domain which is slow.
3461 @var{freq} is processing audio in frequency domain which is fast.
3462 Default is @var{freq}.
3465 Set custom positions of virtual loudspeakers. Syntax for this option is:
3466 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3467 Each virtual loudspeaker is described with short channel name following with
3468 azimuth and elevation in degreees.
3469 Each virtual loudspeaker description is separated by '|'.
3470 For example to override front left and front right channel positions use:
3471 'speakers=FL 45 15|FR 345 15'.
3472 Descriptions with unrecognised channel names are ignored.
3475 @subsection Examples
3479 Using ClubFritz6 sofa file:
3481 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3485 Using ClubFritz12 sofa file and bigger radius with small rotation:
3487 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3491 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3492 and also with custom gain:
3494 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3498 @section stereotools
3500 This filter has some handy utilities to manage stereo signals, for converting
3501 M/S stereo recordings to L/R signal while having control over the parameters
3502 or spreading the stereo image of master track.
3504 The filter accepts the following options:
3508 Set input level before filtering for both channels. Defaults is 1.
3509 Allowed range is from 0.015625 to 64.
3512 Set output level after filtering for both channels. Defaults is 1.
3513 Allowed range is from 0.015625 to 64.
3516 Set input balance between both channels. Default is 0.
3517 Allowed range is from -1 to 1.
3520 Set output balance between both channels. Default is 0.
3521 Allowed range is from -1 to 1.
3524 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3525 clipping. Disabled by default.
3528 Mute the left channel. Disabled by default.
3531 Mute the right channel. Disabled by default.
3534 Change the phase of the left channel. Disabled by default.
3537 Change the phase of the right channel. Disabled by default.
3540 Set stereo mode. Available values are:
3544 Left/Right to Left/Right, this is default.
3547 Left/Right to Mid/Side.
3550 Mid/Side to Left/Right.
3553 Left/Right to Left/Left.
3556 Left/Right to Right/Right.
3559 Left/Right to Left + Right.
3562 Left/Right to Right/Left.
3566 Set level of side signal. Default is 1.
3567 Allowed range is from 0.015625 to 64.
3570 Set balance of side signal. Default is 0.
3571 Allowed range is from -1 to 1.
3574 Set level of the middle signal. Default is 1.
3575 Allowed range is from 0.015625 to 64.
3578 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3581 Set stereo base between mono and inversed channels. Default is 0.
3582 Allowed range is from -1 to 1.
3585 Set delay in milliseconds how much to delay left from right channel and
3586 vice versa. Default is 0. Allowed range is from -20 to 20.
3589 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3592 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3595 @subsection Examples
3599 Apply karaoke like effect:
3601 stereotools=mlev=0.015625
3605 Convert M/S signal to L/R:
3607 "stereotools=mode=ms>lr"
3611 @section stereowiden
3613 This filter enhance the stereo effect by suppressing signal common to both
3614 channels and by delaying the signal of left into right and vice versa,
3615 thereby widening the stereo effect.
3617 The filter accepts the following options:
3621 Time in milliseconds of the delay of left signal into right and vice versa.
3622 Default is 20 milliseconds.
3625 Amount of gain in delayed signal into right and vice versa. Gives a delay
3626 effect of left signal in right output and vice versa which gives widening
3627 effect. Default is 0.3.
3630 Cross feed of left into right with inverted phase. This helps in suppressing
3631 the mono. If the value is 1 it will cancel all the signal common to both
3632 channels. Default is 0.3.
3635 Set level of input signal of original channel. Default is 0.8.
3640 Boost or cut treble (upper) frequencies of the audio using a two-pole
3641 shelving filter with a response similar to that of a standard
3642 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3644 The filter accepts the following options:
3648 Give the gain at whichever is the lower of ~22 kHz and the
3649 Nyquist frequency. Its useful range is about -20 (for a large cut)
3650 to +20 (for a large boost). Beware of clipping when using a positive gain.
3653 Set the filter's central frequency and so can be used
3654 to extend or reduce the frequency range to be boosted or cut.
3655 The default value is @code{3000} Hz.
3658 Set method to specify band-width of filter.
3671 Determine how steep is the filter's shelf transition.
3676 Sinusoidal amplitude modulation.
3678 The filter accepts the following options:
3682 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3683 (20 Hz or lower) will result in a tremolo effect.
3684 This filter may also be used as a ring modulator by specifying
3685 a modulation frequency higher than 20 Hz.
3686 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3689 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3690 Default value is 0.5.
3695 Sinusoidal phase modulation.
3697 The filter accepts the following options:
3701 Modulation frequency in Hertz.
3702 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3705 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3706 Default value is 0.5.
3711 Adjust the input audio volume.
3713 It accepts the following parameters:
3717 Set audio volume expression.
3719 Output values are clipped to the maximum value.
3721 The output audio volume is given by the relation:
3723 @var{output_volume} = @var{volume} * @var{input_volume}
3726 The default value for @var{volume} is "1.0".
3729 This parameter represents the mathematical precision.
3731 It determines which input sample formats will be allowed, which affects the
3732 precision of the volume scaling.
3736 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3738 32-bit floating-point; this limits input sample format to FLT. (default)
3740 64-bit floating-point; this limits input sample format to DBL.
3744 Choose the behaviour on encountering ReplayGain side data in input frames.
3748 Remove ReplayGain side data, ignoring its contents (the default).
3751 Ignore ReplayGain side data, but leave it in the frame.
3754 Prefer the track gain, if present.
3757 Prefer the album gain, if present.
3760 @item replaygain_preamp
3761 Pre-amplification gain in dB to apply to the selected replaygain gain.
3763 Default value for @var{replaygain_preamp} is 0.0.
3766 Set when the volume expression is evaluated.
3768 It accepts the following values:
3771 only evaluate expression once during the filter initialization, or
3772 when the @samp{volume} command is sent
3775 evaluate expression for each incoming frame
3778 Default value is @samp{once}.
3781 The volume expression can contain the following parameters.
3785 frame number (starting at zero)
3788 @item nb_consumed_samples
3789 number of samples consumed by the filter
3791 number of samples in the current frame
3793 original frame position in the file
3799 PTS at start of stream
3801 time at start of stream
3807 last set volume value
3810 Note that when @option{eval} is set to @samp{once} only the
3811 @var{sample_rate} and @var{tb} variables are available, all other
3812 variables will evaluate to NAN.
3814 @subsection Commands
3816 This filter supports the following commands:
3819 Modify the volume expression.
3820 The command accepts the same syntax of the corresponding option.
3822 If the specified expression is not valid, it is kept at its current
3824 @item replaygain_noclip
3825 Prevent clipping by limiting the gain applied.
3827 Default value for @var{replaygain_noclip} is 1.
3831 @subsection Examples
3835 Halve the input audio volume:
3839 volume=volume=-6.0206dB
3842 In all the above example the named key for @option{volume} can be
3843 omitted, for example like in:
3849 Increase input audio power by 6 decibels using fixed-point precision:
3851 volume=volume=6dB:precision=fixed
3855 Fade volume after time 10 with an annihilation period of 5 seconds:
3857 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3861 @section volumedetect
3863 Detect the volume of the input video.
3865 The filter has no parameters. The input is not modified. Statistics about
3866 the volume will be printed in the log when the input stream end is reached.
3868 In particular it will show the mean volume (root mean square), maximum
3869 volume (on a per-sample basis), and the beginning of a histogram of the
3870 registered volume values (from the maximum value to a cumulated 1/1000 of
3873 All volumes are in decibels relative to the maximum PCM value.
3875 @subsection Examples
3877 Here is an excerpt of the output:
3879 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3880 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3881 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3882 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3883 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3884 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3885 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3886 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3887 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3893 The mean square energy is approximately -27 dB, or 10^-2.7.
3895 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3897 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3900 In other words, raising the volume by +4 dB does not cause any clipping,
3901 raising it by +5 dB causes clipping for 6 samples, etc.
3903 @c man end AUDIO FILTERS
3905 @chapter Audio Sources
3906 @c man begin AUDIO SOURCES
3908 Below is a description of the currently available audio sources.
3912 Buffer audio frames, and make them available to the filter chain.
3914 This source is mainly intended for a programmatic use, in particular
3915 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3917 It accepts the following parameters:
3921 The timebase which will be used for timestamps of submitted frames. It must be
3922 either a floating-point number or in @var{numerator}/@var{denominator} form.
3925 The sample rate of the incoming audio buffers.
3928 The sample format of the incoming audio buffers.
3929 Either a sample format name or its corresponding integer representation from
3930 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3932 @item channel_layout
3933 The channel layout of the incoming audio buffers.
3934 Either a channel layout name from channel_layout_map in
3935 @file{libavutil/channel_layout.c} or its corresponding integer representation
3936 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3939 The number of channels of the incoming audio buffers.
3940 If both @var{channels} and @var{channel_layout} are specified, then they
3945 @subsection Examples
3948 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3951 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3952 Since the sample format with name "s16p" corresponds to the number
3953 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3956 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3961 Generate an audio signal specified by an expression.
3963 This source accepts in input one or more expressions (one for each
3964 channel), which are evaluated and used to generate a corresponding
3967 This source accepts the following options:
3971 Set the '|'-separated expressions list for each separate channel. In case the
3972 @option{channel_layout} option is not specified, the selected channel layout
3973 depends on the number of provided expressions. Otherwise the last
3974 specified expression is applied to the remaining output channels.
3976 @item channel_layout, c
3977 Set the channel layout. The number of channels in the specified layout
3978 must be equal to the number of specified expressions.
3981 Set the minimum duration of the sourced audio. See
3982 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3983 for the accepted syntax.
3984 Note that the resulting duration may be greater than the specified
3985 duration, as the generated audio is always cut at the end of a
3988 If not specified, or the expressed duration is negative, the audio is
3989 supposed to be generated forever.
3992 Set the number of samples per channel per each output frame,
3995 @item sample_rate, s
3996 Specify the sample rate, default to 44100.
3999 Each expression in @var{exprs} can contain the following constants:
4003 number of the evaluated sample, starting from 0
4006 time of the evaluated sample expressed in seconds, starting from 0
4013 @subsection Examples
4023 Generate a sin signal with frequency of 440 Hz, set sample rate to
4026 aevalsrc="sin(440*2*PI*t):s=8000"
4030 Generate a two channels signal, specify the channel layout (Front
4031 Center + Back Center) explicitly:
4033 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4037 Generate white noise:
4039 aevalsrc="-2+random(0)"
4043 Generate an amplitude modulated signal:
4045 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4049 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4051 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4058 The null audio source, return unprocessed audio frames. It is mainly useful
4059 as a template and to be employed in analysis / debugging tools, or as
4060 the source for filters which ignore the input data (for example the sox
4063 This source accepts the following options:
4067 @item channel_layout, cl
4069 Specifies the channel layout, and can be either an integer or a string
4070 representing a channel layout. The default value of @var{channel_layout}
4073 Check the channel_layout_map definition in
4074 @file{libavutil/channel_layout.c} for the mapping between strings and
4075 channel layout values.
4077 @item sample_rate, r
4078 Specifies the sample rate, and defaults to 44100.
4081 Set the number of samples per requested frames.
4085 @subsection Examples
4089 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4091 anullsrc=r=48000:cl=4
4095 Do the same operation with a more obvious syntax:
4097 anullsrc=r=48000:cl=mono
4101 All the parameters need to be explicitly defined.
4105 Synthesize a voice utterance using the libflite library.
4107 To enable compilation of this filter you need to configure FFmpeg with
4108 @code{--enable-libflite}.
4110 Note that the flite library is not thread-safe.
4112 The filter accepts the following options:
4117 If set to 1, list the names of the available voices and exit
4118 immediately. Default value is 0.
4121 Set the maximum number of samples per frame. Default value is 512.
4124 Set the filename containing the text to speak.
4127 Set the text to speak.
4130 Set the voice to use for the speech synthesis. Default value is
4131 @code{kal}. See also the @var{list_voices} option.
4134 @subsection Examples
4138 Read from file @file{speech.txt}, and synthesize the text using the
4139 standard flite voice:
4141 flite=textfile=speech.txt
4145 Read the specified text selecting the @code{slt} voice:
4147 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4151 Input text to ffmpeg:
4153 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4157 Make @file{ffplay} speak the specified text, using @code{flite} and
4158 the @code{lavfi} device:
4160 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4164 For more information about libflite, check:
4165 @url{http://www.speech.cs.cmu.edu/flite/}
4169 Generate a noise audio signal.
4171 The filter accepts the following options:
4174 @item sample_rate, r
4175 Specify the sample rate. Default value is 48000 Hz.
4178 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4182 Specify the duration of the generated audio stream. Not specifying this option
4183 results in noise with an infinite length.
4185 @item color, colour, c
4186 Specify the color of noise. Available noise colors are white, pink, and brown.
4187 Default color is white.
4190 Specify a value used to seed the PRNG.
4193 Set the number of samples per each output frame, default is 1024.
4196 @subsection Examples
4201 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4203 anoisesrc=d=60:c=pink:r=44100:a=0.5
4209 Generate an audio signal made of a sine wave with amplitude 1/8.
4211 The audio signal is bit-exact.
4213 The filter accepts the following options:
4218 Set the carrier frequency. Default is 440 Hz.
4220 @item beep_factor, b
4221 Enable a periodic beep every second with frequency @var{beep_factor} times
4222 the carrier frequency. Default is 0, meaning the beep is disabled.
4224 @item sample_rate, r
4225 Specify the sample rate, default is 44100.
4228 Specify the duration of the generated audio stream.
4230 @item samples_per_frame
4231 Set the number of samples per output frame.
4233 The expression can contain the following constants:
4237 The (sequential) number of the output audio frame, starting from 0.
4240 The PTS (Presentation TimeStamp) of the output audio frame,
4241 expressed in @var{TB} units.
4244 The PTS of the output audio frame, expressed in seconds.
4247 The timebase of the output audio frames.
4250 Default is @code{1024}.
4253 @subsection Examples
4258 Generate a simple 440 Hz sine wave:
4264 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4268 sine=frequency=220:beep_factor=4:duration=5
4272 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4275 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4279 @c man end AUDIO SOURCES
4281 @chapter Audio Sinks
4282 @c man begin AUDIO SINKS
4284 Below is a description of the currently available audio sinks.
4286 @section abuffersink
4288 Buffer audio frames, and make them available to the end of filter chain.
4290 This sink is mainly intended for programmatic use, in particular
4291 through the interface defined in @file{libavfilter/buffersink.h}
4292 or the options system.
4294 It accepts a pointer to an AVABufferSinkContext structure, which
4295 defines the incoming buffers' formats, to be passed as the opaque
4296 parameter to @code{avfilter_init_filter} for initialization.
4299 Null audio sink; do absolutely nothing with the input audio. It is
4300 mainly useful as a template and for use in analysis / debugging
4303 @c man end AUDIO SINKS
4305 @chapter Video Filters
4306 @c man begin VIDEO FILTERS
4308 When you configure your FFmpeg build, you can disable any of the
4309 existing filters using @code{--disable-filters}.
4310 The configure output will show the video filters included in your
4313 Below is a description of the currently available video filters.
4315 @section alphaextract
4317 Extract the alpha component from the input as a grayscale video. This
4318 is especially useful with the @var{alphamerge} filter.
4322 Add or replace the alpha component of the primary input with the
4323 grayscale value of a second input. This is intended for use with
4324 @var{alphaextract} to allow the transmission or storage of frame
4325 sequences that have alpha in a format that doesn't support an alpha
4328 For example, to reconstruct full frames from a normal YUV-encoded video
4329 and a separate video created with @var{alphaextract}, you might use:
4331 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4334 Since this filter is designed for reconstruction, it operates on frame
4335 sequences without considering timestamps, and terminates when either
4336 input reaches end of stream. This will cause problems if your encoding
4337 pipeline drops frames. If you're trying to apply an image as an
4338 overlay to a video stream, consider the @var{overlay} filter instead.
4342 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4343 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4344 Substation Alpha) subtitles files.
4346 This filter accepts the following option in addition to the common options from
4347 the @ref{subtitles} filter:
4351 Set the shaping engine
4353 Available values are:
4356 The default libass shaping engine, which is the best available.
4358 Fast, font-agnostic shaper that can do only substitutions
4360 Slower shaper using OpenType for substitutions and positioning
4363 The default is @code{auto}.
4367 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4369 The filter accepts the following options:
4373 Set threshold A for 1st plane. Default is 0.02.
4374 Valid range is 0 to 0.3.
4377 Set threshold B for 1st plane. Default is 0.04.
4378 Valid range is 0 to 5.
4381 Set threshold A for 2nd plane. Default is 0.02.
4382 Valid range is 0 to 0.3.
4385 Set threshold B for 2nd plane. Default is 0.04.
4386 Valid range is 0 to 5.
4389 Set threshold A for 3rd plane. Default is 0.02.
4390 Valid range is 0 to 0.3.
4393 Set threshold B for 3rd plane. Default is 0.04.
4394 Valid range is 0 to 5.
4396 Threshold A is designed to react on abrupt changes in the input signal and
4397 threshold B is designed to react on continuous changes in the input signal.
4400 Set number of frames filter will use for averaging. Default is 33. Must be odd
4401 number in range [5, 129].
4404 Set what planes of frame filter will use for averaging. Default is all.
4409 Apply average blur filter.
4411 The filter accepts the following options:
4415 Set horizontal kernel size.
4418 Set which planes to filter. By default all planes are filtered.
4421 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4422 Default is @code{0}.
4427 Compute the bounding box for the non-black pixels in the input frame
4430 This filter computes the bounding box containing all the pixels with a
4431 luminance value greater than the minimum allowed value.
4432 The parameters describing the bounding box are printed on the filter
4435 The filter accepts the following option:
4439 Set the minimal luminance value. Default is @code{16}.
4442 @section bitplanenoise
4444 Show and measure bit plane noise.
4446 The filter accepts the following options:
4450 Set which plane to analyze. Default is @code{1}.
4453 Filter out noisy pixels from @code{bitplane} set above.
4454 Default is disabled.
4457 @section blackdetect
4459 Detect video intervals that are (almost) completely black. Can be
4460 useful to detect chapter transitions, commercials, or invalid
4461 recordings. Output lines contains the time for the start, end and
4462 duration of the detected black interval expressed in seconds.
4464 In order to display the output lines, you need to set the loglevel at
4465 least to the AV_LOG_INFO value.
4467 The filter accepts the following options:
4470 @item black_min_duration, d
4471 Set the minimum detected black duration expressed in seconds. It must
4472 be a non-negative floating point number.
4474 Default value is 2.0.
4476 @item picture_black_ratio_th, pic_th
4477 Set the threshold for considering a picture "black".
4478 Express the minimum value for the ratio:
4480 @var{nb_black_pixels} / @var{nb_pixels}
4483 for which a picture is considered black.
4484 Default value is 0.98.
4486 @item pixel_black_th, pix_th
4487 Set the threshold for considering a pixel "black".
4489 The threshold expresses the maximum pixel luminance value for which a
4490 pixel is considered "black". The provided value is scaled according to
4491 the following equation:
4493 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4496 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4497 the input video format, the range is [0-255] for YUV full-range
4498 formats and [16-235] for YUV non full-range formats.
4500 Default value is 0.10.
4503 The following example sets the maximum pixel threshold to the minimum
4504 value, and detects only black intervals of 2 or more seconds:
4506 blackdetect=d=2:pix_th=0.00
4511 Detect frames that are (almost) completely black. Can be useful to
4512 detect chapter transitions or commercials. Output lines consist of
4513 the frame number of the detected frame, the percentage of blackness,
4514 the position in the file if known or -1 and the timestamp in seconds.
4516 In order to display the output lines, you need to set the loglevel at
4517 least to the AV_LOG_INFO value.
4519 It accepts the following parameters:
4524 The percentage of the pixels that have to be below the threshold; it defaults to
4527 @item threshold, thresh
4528 The threshold below which a pixel value is considered black; it defaults to
4533 @section blend, tblend
4535 Blend two video frames into each other.
4537 The @code{blend} filter takes two input streams and outputs one
4538 stream, the first input is the "top" layer and second input is
4539 "bottom" layer. By default, the output terminates when the longest input terminates.
4541 The @code{tblend} (time blend) filter takes two consecutive frames
4542 from one single stream, and outputs the result obtained by blending
4543 the new frame on top of the old frame.
4545 A description of the accepted options follows.
4553 Set blend mode for specific pixel component or all pixel components in case
4554 of @var{all_mode}. Default value is @code{normal}.
4556 Available values for component modes are:
4597 Set blend opacity for specific pixel component or all pixel components in case
4598 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4605 Set blend expression for specific pixel component or all pixel components in case
4606 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4608 The expressions can use the following variables:
4612 The sequential number of the filtered frame, starting from @code{0}.
4616 the coordinates of the current sample
4620 the width and height of currently filtered plane
4624 Width and height scale depending on the currently filtered plane. It is the
4625 ratio between the corresponding luma plane number of pixels and the current
4626 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4627 @code{0.5,0.5} for chroma planes.
4630 Time of the current frame, expressed in seconds.
4633 Value of pixel component at current location for first video frame (top layer).
4636 Value of pixel component at current location for second video frame (bottom layer).
4640 Force termination when the shortest input terminates. Default is
4641 @code{0}. This option is only defined for the @code{blend} filter.
4644 Continue applying the last bottom frame after the end of the stream. A value of
4645 @code{0} disable the filter after the last frame of the bottom layer is reached.
4646 Default is @code{1}. This option is only defined for the @code{blend} filter.
4649 @subsection Examples
4653 Apply transition from bottom layer to top layer in first 10 seconds:
4655 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4659 Apply 1x1 checkerboard effect:
4661 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4665 Apply uncover left effect:
4667 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4671 Apply uncover down effect:
4673 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4677 Apply uncover up-left effect:
4679 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4683 Split diagonally video and shows top and bottom layer on each side:
4685 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4689 Display differences between the current and the previous frame:
4691 tblend=all_mode=difference128
4697 Apply a boxblur algorithm to the input video.
4699 It accepts the following parameters:
4703 @item luma_radius, lr
4704 @item luma_power, lp
4705 @item chroma_radius, cr
4706 @item chroma_power, cp
4707 @item alpha_radius, ar
4708 @item alpha_power, ap
4712 A description of the accepted options follows.
4715 @item luma_radius, lr
4716 @item chroma_radius, cr
4717 @item alpha_radius, ar
4718 Set an expression for the box radius in pixels used for blurring the
4719 corresponding input plane.
4721 The radius value must be a non-negative number, and must not be
4722 greater than the value of the expression @code{min(w,h)/2} for the
4723 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4726 Default value for @option{luma_radius} is "2". If not specified,
4727 @option{chroma_radius} and @option{alpha_radius} default to the
4728 corresponding value set for @option{luma_radius}.
4730 The expressions can contain the following constants:
4734 The input width and height in pixels.
4738 The input chroma image width and height in pixels.
4742 The horizontal and vertical chroma subsample values. For example, for the
4743 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4746 @item luma_power, lp
4747 @item chroma_power, cp
4748 @item alpha_power, ap
4749 Specify how many times the boxblur filter is applied to the
4750 corresponding plane.
4752 Default value for @option{luma_power} is 2. If not specified,
4753 @option{chroma_power} and @option{alpha_power} default to the
4754 corresponding value set for @option{luma_power}.
4756 A value of 0 will disable the effect.
4759 @subsection Examples
4763 Apply a boxblur filter with the luma, chroma, and alpha radii
4766 boxblur=luma_radius=2:luma_power=1
4771 Set the luma radius to 2, and alpha and chroma radius to 0:
4773 boxblur=2:1:cr=0:ar=0
4777 Set the luma and chroma radii to a fraction of the video dimension:
4779 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4785 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4786 Deinterlacing Filter").
4788 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4789 interpolation algorithms.
4790 It accepts the following parameters:
4794 The interlacing mode to adopt. It accepts one of the following values:
4798 Output one frame for each frame.
4800 Output one frame for each field.
4803 The default value is @code{send_field}.
4806 The picture field parity assumed for the input interlaced video. It accepts one
4807 of the following values:
4811 Assume the top field is first.
4813 Assume the bottom field is first.
4815 Enable automatic detection of field parity.
4818 The default value is @code{auto}.
4819 If the interlacing is unknown or the decoder does not export this information,
4820 top field first will be assumed.
4823 Specify which frames to deinterlace. Accept one of the following
4828 Deinterlace all frames.
4830 Only deinterlace frames marked as interlaced.
4833 The default value is @code{all}.
4837 YUV colorspace color/chroma keying.
4839 The filter accepts the following options:
4843 The color which will be replaced with transparency.
4846 Similarity percentage with the key color.
4848 0.01 matches only the exact key color, while 1.0 matches everything.
4853 0.0 makes pixels either fully transparent, or not transparent at all.
4855 Higher values result in semi-transparent pixels, with a higher transparency
4856 the more similar the pixels color is to the key color.
4859 Signals that the color passed is already in YUV instead of RGB.
4861 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4862 This can be used to pass exact YUV values as hexadecimal numbers.
4865 @subsection Examples
4869 Make every green pixel in the input image transparent:
4871 ffmpeg -i input.png -vf chromakey=green out.png
4875 Overlay a greenscreen-video on top of a static black background.
4877 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
4883 Display CIE color diagram with pixels overlaid onto it.
4885 The filter accepts the following options:
4900 @item uhdtv, rec2020
4913 Set what gamuts to draw.
4915 See @code{system} option for available values.
4918 Set ciescope size, by default set to 512.
4921 Set intensity used to map input pixel values to CIE diagram.
4924 Set contrast used to draw tongue colors that are out of active color system gamut.
4927 Correct gamma displayed on scope, by default enabled.
4930 Show white point on CIE diagram, by default disabled.
4933 Set input gamma. Used only with XYZ input color space.
4938 Visualize information exported by some codecs.
4940 Some codecs can export information through frames using side-data or other
4941 means. For example, some MPEG based codecs export motion vectors through the
4942 @var{export_mvs} flag in the codec @option{flags2} option.
4944 The filter accepts the following option:
4948 Set motion vectors to visualize.
4950 Available flags for @var{mv} are:
4954 forward predicted MVs of P-frames
4956 forward predicted MVs of B-frames
4958 backward predicted MVs of B-frames
4962 Display quantization parameters using the chroma planes.
4965 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4967 Available flags for @var{mv_type} are:
4971 forward predicted MVs
4973 backward predicted MVs
4976 @item frame_type, ft
4977 Set frame type to visualize motion vectors of.
4979 Available flags for @var{frame_type} are:
4983 intra-coded frames (I-frames)
4985 predicted frames (P-frames)
4987 bi-directionally predicted frames (B-frames)
4991 @subsection Examples
4995 Visualize forward predicted MVs of all frames using @command{ffplay}:
4997 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5001 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5003 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5007 @section colorbalance
5008 Modify intensity of primary colors (red, green and blue) of input frames.
5010 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5011 regions for the red-cyan, green-magenta or blue-yellow balance.
5013 A positive adjustment value shifts the balance towards the primary color, a negative
5014 value towards the complementary color.
5016 The filter accepts the following options:
5022 Adjust red, green and blue shadows (darkest pixels).
5027 Adjust red, green and blue midtones (medium pixels).
5032 Adjust red, green and blue highlights (brightest pixels).
5034 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5037 @subsection Examples
5041 Add red color cast to shadows:
5048 RGB colorspace color keying.
5050 The filter accepts the following options:
5054 The color which will be replaced with transparency.
5057 Similarity percentage with the key color.
5059 0.01 matches only the exact key color, while 1.0 matches everything.
5064 0.0 makes pixels either fully transparent, or not transparent at all.
5066 Higher values result in semi-transparent pixels, with a higher transparency
5067 the more similar the pixels color is to the key color.
5070 @subsection Examples
5074 Make every green pixel in the input image transparent:
5076 ffmpeg -i input.png -vf colorkey=green out.png
5080 Overlay a greenscreen-video on top of a static background image.
5082 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
5086 @section colorlevels
5088 Adjust video input frames using levels.
5090 The filter accepts the following options:
5097 Adjust red, green, blue and alpha input black point.
5098 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5104 Adjust red, green, blue and alpha input white point.
5105 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5107 Input levels are used to lighten highlights (bright tones), darken shadows
5108 (dark tones), change the balance of bright and dark tones.
5114 Adjust red, green, blue and alpha output black point.
5115 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5121 Adjust red, green, blue and alpha output white point.
5122 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5124 Output levels allows manual selection of a constrained output level range.
5127 @subsection Examples
5131 Make video output darker:
5133 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5139 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5143 Make video output lighter:
5145 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5149 Increase brightness:
5151 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5155 @section colorchannelmixer
5157 Adjust video input frames by re-mixing color channels.
5159 This filter modifies a color channel by adding the values associated to
5160 the other channels of the same pixels. For example if the value to
5161 modify is red, the output value will be:
5163 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5166 The filter accepts the following options:
5173 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5174 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5180 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5181 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5187 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5188 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5194 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5195 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5197 Allowed ranges for options are @code{[-2.0, 2.0]}.
5200 @subsection Examples
5204 Convert source to grayscale:
5206 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5209 Simulate sepia tones:
5211 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5215 @section colormatrix
5217 Convert color matrix.
5219 The filter accepts the following options:
5224 Specify the source and destination color matrix. Both values must be
5227 The accepted values are:
5246 For example to convert from BT.601 to SMPTE-240M, use the command:
5248 colormatrix=bt601:smpte240m
5253 Convert colorspace, transfer characteristics or color primaries.
5255 The filter accepts the following options:
5260 Specify all color properties at once.
5262 The accepted values are:
5292 Specify output colorspace.
5294 The accepted values are:
5303 BT.470BG or BT.601-6 625
5306 SMPTE-170M or BT.601-6 525
5312 BT.2020 with non-constant luminance
5318 Specify output transfer characteristics.
5320 The accepted values are:
5326 Constant gamma of 2.2
5329 Constant gamma of 2.8
5332 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5338 BT.2020 for 10-bits content
5341 BT.2020 for 12-bits content
5347 Specify output color primaries.
5349 The accepted values are:
5358 BT.470BG or BT.601-6 625
5361 SMPTE-170M or BT.601-6 525
5373 Specify output color range.
5375 The accepted values are:
5378 MPEG (restricted) range
5386 Specify output color format.
5388 The accepted values are:
5391 YUV 4:2:0 planar 8-bits
5394 YUV 4:2:0 planar 10-bits
5397 YUV 4:2:0 planar 12-bits
5400 YUV 4:2:2 planar 8-bits
5403 YUV 4:2:2 planar 10-bits
5406 YUV 4:2:2 planar 12-bits
5409 YUV 4:4:4 planar 8-bits
5412 YUV 4:4:4 planar 10-bits
5415 YUV 4:4:4 planar 12-bits
5420 Do a fast conversion, which skips gamma/primary correction. This will take
5421 significantly less CPU, but will be mathematically incorrect. To get output
5422 compatible with that produced by the colormatrix filter, use fast=1.
5425 Specify dithering mode.
5427 The accepted values are:
5433 Floyd-Steinberg dithering
5437 Whitepoint adaptation mode.
5439 The accepted values are:
5442 Bradford whitepoint adaptation
5445 von Kries whitepoint adaptation
5448 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5452 Override all input properties at once. Same accepted values as @ref{all}.
5455 Override input colorspace. Same accepted values as @ref{space}.
5458 Override input color primaries. Same accepted values as @ref{primaries}.
5461 Override input transfer characteristics. Same accepted values as @ref{trc}.
5464 Override input color range. Same accepted values as @ref{range}.
5468 The filter converts the transfer characteristics, color space and color
5469 primaries to the specified user values. The output value, if not specified,
5470 is set to a default value based on the "all" property. If that property is
5471 also not specified, the filter will log an error. The output color range and
5472 format default to the same value as the input color range and format. The
5473 input transfer characteristics, color space, color primaries and color range
5474 should be set on the input data. If any of these are missing, the filter will
5475 log an error and no conversion will take place.
5477 For example to convert the input to SMPTE-240M, use the command:
5479 colorspace=smpte240m
5482 @section convolution
5484 Apply convolution 3x3 or 5x5 filter.
5486 The filter accepts the following options:
5493 Set matrix for each plane.
5494 Matrix is sequence of 9 or 25 signed integers.
5500 Set multiplier for calculated value for each plane.
5506 Set bias for each plane. This value is added to the result of the multiplication.
5507 Useful for making the overall image brighter or darker. Default is 0.0.
5510 @subsection Examples
5516 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"
5522 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"
5528 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"
5534 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"
5540 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"
5546 Copy the input source unchanged to the output. This is mainly useful for
5551 Video filtering on GPU using Apple's CoreImage API on OSX.
5553 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5554 processed by video hardware. However, software-based OpenGL implementations
5555 exist which means there is no guarantee for hardware processing. It depends on
5558 There are many filters and image generators provided by Apple that come with a
5559 large variety of options. The filter has to be referenced by its name along
5562 The coreimage filter accepts the following options:
5565 List all available filters and generators along with all their respective
5566 options as well as possible minimum and maximum values along with the default
5573 Specify all filters by their respective name and options.
5574 Use @var{list_filters} to determine all valid filter names and options.
5575 Numerical options are specified by a float value and are automatically clamped
5576 to their respective value range. Vector and color options have to be specified
5577 by a list of space separated float values. Character escaping has to be done.
5578 A special option name @code{default} is available to use default options for a
5581 It is required to specify either @code{default} or at least one of the filter options.
5582 All omitted options are used with their default values.
5583 The syntax of the filter string is as follows:
5585 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5589 Specify a rectangle where the output of the filter chain is copied into the
5590 input image. It is given by a list of space separated float values:
5592 output_rect=x\ y\ width\ height
5594 If not given, the output rectangle equals the dimensions of the input image.
5595 The output rectangle is automatically cropped at the borders of the input
5596 image. Negative values are valid for each component.
5598 output_rect=25\ 25\ 100\ 100
5602 Several filters can be chained for successive processing without GPU-HOST
5603 transfers allowing for fast processing of complex filter chains.
5604 Currently, only filters with zero (generators) or exactly one (filters) input
5605 image and one output image are supported. Also, transition filters are not yet
5608 Some filters generate output images with additional padding depending on the
5609 respective filter kernel. The padding is automatically removed to ensure the
5610 filter output has the same size as the input image.
5612 For image generators, the size of the output image is determined by the
5613 previous output image of the filter chain or the input image of the whole
5614 filterchain, respectively. The generators do not use the pixel information of
5615 this image to generate their output. However, the generated output is
5616 blended onto this image, resulting in partial or complete coverage of the
5619 The @ref{coreimagesrc} video source can be used for generating input images
5620 which are directly fed into the filter chain. By using it, providing input
5621 images by another video source or an input video is not required.
5623 @subsection Examples
5628 List all filters available:
5630 coreimage=list_filters=true
5634 Use the CIBoxBlur filter with default options to blur an image:
5636 coreimage=filter=CIBoxBlur@@default
5640 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5641 its center at 100x100 and a radius of 50 pixels:
5643 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5647 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5648 given as complete and escaped command-line for Apple's standard bash shell:
5650 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5656 Crop the input video to given dimensions.
5658 It accepts the following parameters:
5662 The width of the output video. It defaults to @code{iw}.
5663 This expression is evaluated only once during the filter
5664 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5667 The height of the output video. It defaults to @code{ih}.
5668 This expression is evaluated only once during the filter
5669 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5672 The horizontal position, in the input video, of the left edge of the output
5673 video. It defaults to @code{(in_w-out_w)/2}.
5674 This expression is evaluated per-frame.
5677 The vertical position, in the input video, of the top edge of the output video.
5678 It defaults to @code{(in_h-out_h)/2}.
5679 This expression is evaluated per-frame.
5682 If set to 1 will force the output display aspect ratio
5683 to be the same of the input, by changing the output sample aspect
5684 ratio. It defaults to 0.
5687 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5688 width/height/x/y as specified and will not be rounded to nearest smaller value.
5692 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5693 expressions containing the following constants:
5698 The computed values for @var{x} and @var{y}. They are evaluated for
5703 The input width and height.
5707 These are the same as @var{in_w} and @var{in_h}.
5711 The output (cropped) width and height.
5715 These are the same as @var{out_w} and @var{out_h}.
5718 same as @var{iw} / @var{ih}
5721 input sample aspect ratio
5724 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5728 horizontal and vertical chroma subsample values. For example for the
5729 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5732 The number of the input frame, starting from 0.
5735 the position in the file of the input frame, NAN if unknown
5738 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5742 The expression for @var{out_w} may depend on the value of @var{out_h},
5743 and the expression for @var{out_h} may depend on @var{out_w}, but they
5744 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5745 evaluated after @var{out_w} and @var{out_h}.
5747 The @var{x} and @var{y} parameters specify the expressions for the
5748 position of the top-left corner of the output (non-cropped) area. They
5749 are evaluated for each frame. If the evaluated value is not valid, it
5750 is approximated to the nearest valid value.
5752 The expression for @var{x} may depend on @var{y}, and the expression
5753 for @var{y} may depend on @var{x}.
5755 @subsection Examples
5759 Crop area with size 100x100 at position (12,34).
5764 Using named options, the example above becomes:
5766 crop=w=100:h=100:x=12:y=34
5770 Crop the central input area with size 100x100:
5776 Crop the central input area with size 2/3 of the input video:
5778 crop=2/3*in_w:2/3*in_h
5782 Crop the input video central square:
5789 Delimit the rectangle with the top-left corner placed at position
5790 100:100 and the right-bottom corner corresponding to the right-bottom
5791 corner of the input image.
5793 crop=in_w-100:in_h-100:100:100
5797 Crop 10 pixels from the left and right borders, and 20 pixels from
5798 the top and bottom borders
5800 crop=in_w-2*10:in_h-2*20
5804 Keep only the bottom right quarter of the input image:
5806 crop=in_w/2:in_h/2:in_w/2:in_h/2
5810 Crop height for getting Greek harmony:
5812 crop=in_w:1/PHI*in_w
5816 Apply trembling effect:
5818 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)
5822 Apply erratic camera effect depending on timestamp:
5824 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)"
5828 Set x depending on the value of y:
5830 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5834 @subsection Commands
5836 This filter supports the following commands:
5842 Set width/height of the output video and the horizontal/vertical position
5844 The command accepts the same syntax of the corresponding option.
5846 If the specified expression is not valid, it is kept at its current
5852 Auto-detect the crop size.
5854 It calculates the necessary cropping parameters and prints the
5855 recommended parameters via the logging system. The detected dimensions
5856 correspond to the non-black area of the input video.
5858 It accepts the following parameters:
5863 Set higher black value threshold, which can be optionally specified
5864 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5865 value greater to the set value is considered non-black. It defaults to 24.
5866 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5867 on the bitdepth of the pixel format.
5870 The value which the width/height should be divisible by. It defaults to
5871 16. The offset is automatically adjusted to center the video. Use 2 to
5872 get only even dimensions (needed for 4:2:2 video). 16 is best when
5873 encoding to most video codecs.
5875 @item reset_count, reset
5876 Set the counter that determines after how many frames cropdetect will
5877 reset the previously detected largest video area and start over to
5878 detect the current optimal crop area. Default value is 0.
5880 This can be useful when channel logos distort the video area. 0
5881 indicates 'never reset', and returns the largest area encountered during
5888 Apply color adjustments using curves.
5890 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5891 component (red, green and blue) has its values defined by @var{N} key points
5892 tied from each other using a smooth curve. The x-axis represents the pixel
5893 values from the input frame, and the y-axis the new pixel values to be set for
5896 By default, a component curve is defined by the two points @var{(0;0)} and
5897 @var{(1;1)}. This creates a straight line where each original pixel value is
5898 "adjusted" to its own value, which means no change to the image.
5900 The filter allows you to redefine these two points and add some more. A new
5901 curve (using a natural cubic spline interpolation) will be define to pass
5902 smoothly through all these new coordinates. The new defined points needs to be
5903 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5904 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5905 the vector spaces, the values will be clipped accordingly.
5907 The filter accepts the following options:
5911 Select one of the available color presets. This option can be used in addition
5912 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5913 options takes priority on the preset values.
5914 Available presets are:
5917 @item color_negative
5920 @item increase_contrast
5922 @item linear_contrast
5923 @item medium_contrast
5925 @item strong_contrast
5928 Default is @code{none}.
5930 Set the master key points. These points will define a second pass mapping. It
5931 is sometimes called a "luminance" or "value" mapping. It can be used with
5932 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5933 post-processing LUT.
5935 Set the key points for the red component.
5937 Set the key points for the green component.
5939 Set the key points for the blue component.
5941 Set the key points for all components (not including master).
5942 Can be used in addition to the other key points component
5943 options. In this case, the unset component(s) will fallback on this
5944 @option{all} setting.
5946 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5948 Save Gnuplot script of the curves in specified file.
5951 To avoid some filtergraph syntax conflicts, each key points list need to be
5952 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5954 @subsection Examples
5958 Increase slightly the middle level of blue:
5960 curves=blue='0/0 0.5/0.58 1/1'
5966 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'
5968 Here we obtain the following coordinates for each components:
5971 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5973 @code{(0;0) (0.50;0.48) (1;1)}
5975 @code{(0;0.22) (0.49;0.44) (1;0.80)}
5979 The previous example can also be achieved with the associated built-in preset:
5981 curves=preset=vintage
5991 Use a Photoshop preset and redefine the points of the green component:
5993 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
5997 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
5998 and @command{gnuplot}:
6000 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6001 gnuplot -p /tmp/curves.plt
6007 Video data analysis filter.
6009 This filter shows hexadecimal pixel values of part of video.
6011 The filter accepts the following options:
6015 Set output video size.
6018 Set x offset from where to pick pixels.
6021 Set y offset from where to pick pixels.
6024 Set scope mode, can be one of the following:
6027 Draw hexadecimal pixel values with white color on black background.
6030 Draw hexadecimal pixel values with input video pixel color on black
6034 Draw hexadecimal pixel values on color background picked from input video,
6035 the text color is picked in such way so its always visible.
6039 Draw rows and columns numbers on left and top of video.
6042 Set background opacity.
6047 Denoise frames using 2D DCT (frequency domain filtering).
6049 This filter is not designed for real time.
6051 The filter accepts the following options:
6055 Set the noise sigma constant.
6057 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6058 coefficient (absolute value) below this threshold with be dropped.
6060 If you need a more advanced filtering, see @option{expr}.
6062 Default is @code{0}.
6065 Set number overlapping pixels for each block. Since the filter can be slow, you
6066 may want to reduce this value, at the cost of a less effective filter and the
6067 risk of various artefacts.
6069 If the overlapping value doesn't permit processing the whole input width or
6070 height, a warning will be displayed and according borders won't be denoised.
6072 Default value is @var{blocksize}-1, which is the best possible setting.
6075 Set the coefficient factor expression.
6077 For each coefficient of a DCT block, this expression will be evaluated as a
6078 multiplier value for the coefficient.
6080 If this is option is set, the @option{sigma} option will be ignored.
6082 The absolute value of the coefficient can be accessed through the @var{c}
6086 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6087 @var{blocksize}, which is the width and height of the processed blocks.
6089 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6090 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6091 on the speed processing. Also, a larger block size does not necessarily means a
6095 @subsection Examples
6097 Apply a denoise with a @option{sigma} of @code{4.5}:
6102 The same operation can be achieved using the expression system:
6104 dctdnoiz=e='gte(c, 4.5*3)'
6107 Violent denoise using a block size of @code{16x16}:
6114 Remove banding artifacts from input video.
6115 It works by replacing banded pixels with average value of referenced pixels.
6117 The filter accepts the following options:
6124 Set banding detection threshold for each plane. Default is 0.02.
6125 Valid range is 0.00003 to 0.5.
6126 If difference between current pixel and reference pixel is less than threshold,
6127 it will be considered as banded.
6130 Banding detection range in pixels. Default is 16. If positive, random number
6131 in range 0 to set value will be used. If negative, exact absolute value
6133 The range defines square of four pixels around current pixel.
6136 Set direction in radians from which four pixel will be compared. If positive,
6137 random direction from 0 to set direction will be picked. If negative, exact of
6138 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6139 will pick only pixels on same row and -PI/2 will pick only pixels on same
6143 If enabled, current pixel is compared with average value of all four
6144 surrounding pixels. The default is enabled. If disabled current pixel is
6145 compared with all four surrounding pixels. The pixel is considered banded
6146 if only all four differences with surrounding pixels are less than threshold.
6152 Drop duplicated frames at regular intervals.
6154 The filter accepts the following options:
6158 Set the number of frames from which one will be dropped. Setting this to
6159 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6160 Default is @code{5}.
6163 Set the threshold for duplicate detection. If the difference metric for a frame
6164 is less than or equal to this value, then it is declared as duplicate. Default
6168 Set scene change threshold. Default is @code{15}.
6172 Set the size of the x and y-axis blocks used during metric calculations.
6173 Larger blocks give better noise suppression, but also give worse detection of
6174 small movements. Must be a power of two. Default is @code{32}.
6177 Mark main input as a pre-processed input and activate clean source input
6178 stream. This allows the input to be pre-processed with various filters to help
6179 the metrics calculation while keeping the frame selection lossless. When set to
6180 @code{1}, the first stream is for the pre-processed input, and the second
6181 stream is the clean source from where the kept frames are chosen. Default is
6185 Set whether or not chroma is considered in the metric calculations. Default is
6191 Apply deflate effect to the video.
6193 This filter replaces the pixel by the local(3x3) average by taking into account
6194 only values lower than the pixel.
6196 It accepts the following options:
6203 Limit the maximum change for each plane, default is 65535.
6204 If 0, plane will remain unchanged.
6209 Remove judder produced by partially interlaced telecined content.
6211 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6212 source was partially telecined content then the output of @code{pullup,dejudder}
6213 will have a variable frame rate. May change the recorded frame rate of the
6214 container. Aside from that change, this filter will not affect constant frame
6217 The option available in this filter is:
6221 Specify the length of the window over which the judder repeats.
6223 Accepts any integer greater than 1. Useful values are:
6227 If the original was telecined from 24 to 30 fps (Film to NTSC).
6230 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6233 If a mixture of the two.
6236 The default is @samp{4}.
6241 Suppress a TV station logo by a simple interpolation of the surrounding
6242 pixels. Just set a rectangle covering the logo and watch it disappear
6243 (and sometimes something even uglier appear - your mileage may vary).
6245 It accepts the following parameters:
6250 Specify the top left corner coordinates of the logo. They must be
6255 Specify the width and height of the logo to clear. They must be
6259 Specify the thickness of the fuzzy edge of the rectangle (added to
6260 @var{w} and @var{h}). The default value is 1. This option is
6261 deprecated, setting higher values should no longer be necessary and
6265 When set to 1, a green rectangle is drawn on the screen to simplify
6266 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6267 The default value is 0.
6269 The rectangle is drawn on the outermost pixels which will be (partly)
6270 replaced with interpolated values. The values of the next pixels
6271 immediately outside this rectangle in each direction will be used to
6272 compute the interpolated pixel values inside the rectangle.
6276 @subsection Examples
6280 Set a rectangle covering the area with top left corner coordinates 0,0
6281 and size 100x77, and a band of size 10:
6283 delogo=x=0:y=0:w=100:h=77:band=10
6290 Attempt to fix small changes in horizontal and/or vertical shift. This
6291 filter helps remove camera shake from hand-holding a camera, bumping a
6292 tripod, moving on a vehicle, etc.
6294 The filter accepts the following options:
6302 Specify a rectangular area where to limit the search for motion
6304 If desired the search for motion vectors can be limited to a
6305 rectangular area of the frame defined by its top left corner, width
6306 and height. These parameters have the same meaning as the drawbox
6307 filter which can be used to visualise the position of the bounding
6310 This is useful when simultaneous movement of subjects within the frame
6311 might be confused for camera motion by the motion vector search.
6313 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6314 then the full frame is used. This allows later options to be set
6315 without specifying the bounding box for the motion vector search.
6317 Default - search the whole frame.
6321 Specify the maximum extent of movement in x and y directions in the
6322 range 0-64 pixels. Default 16.
6325 Specify how to generate pixels to fill blanks at the edge of the
6326 frame. Available values are:
6329 Fill zeroes at blank locations
6331 Original image at blank locations
6333 Extruded edge value at blank locations
6335 Mirrored edge at blank locations
6337 Default value is @samp{mirror}.
6340 Specify the blocksize to use for motion search. Range 4-128 pixels,
6344 Specify the contrast threshold for blocks. Only blocks with more than
6345 the specified contrast (difference between darkest and lightest
6346 pixels) will be considered. Range 1-255, default 125.
6349 Specify the search strategy. Available values are:
6352 Set exhaustive search
6354 Set less exhaustive search.
6356 Default value is @samp{exhaustive}.
6359 If set then a detailed log of the motion search is written to the
6363 If set to 1, specify using OpenCL capabilities, only available if
6364 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6370 Apply an exact inverse of the telecine operation. It requires a predefined
6371 pattern specified using the pattern option which must be the same as that passed
6372 to the telecine filter.
6374 This filter accepts the following options:
6383 The default value is @code{top}.
6387 A string of numbers representing the pulldown pattern you wish to apply.
6388 The default value is @code{23}.
6391 A number representing position of the first frame with respect to the telecine
6392 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6397 Apply dilation effect to the video.
6399 This filter replaces the pixel by the local(3x3) maximum.
6401 It accepts the following options:
6408 Limit the maximum change for each plane, default is 65535.
6409 If 0, plane will remain unchanged.
6412 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6415 Flags to local 3x3 coordinates maps like this:
6424 Displace pixels as indicated by second and third input stream.
6426 It takes three input streams and outputs one stream, the first input is the
6427 source, and second and third input are displacement maps.
6429 The second input specifies how much to displace pixels along the
6430 x-axis, while the third input specifies how much to displace pixels
6432 If one of displacement map streams terminates, last frame from that
6433 displacement map will be used.
6435 Note that once generated, displacements maps can be reused over and over again.
6437 A description of the accepted options follows.
6441 Set displace behavior for pixels that are out of range.
6443 Available values are:
6446 Missing pixels are replaced by black pixels.
6449 Adjacent pixels will spread out to replace missing pixels.
6452 Out of range pixels are wrapped so they point to pixels of other side.
6454 Default is @samp{smear}.
6458 @subsection Examples
6462 Add ripple effect to rgb input of video size hd720:
6464 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
6468 Add wave effect to rgb input of video size hd720:
6470 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
6476 Draw a colored box on the input image.
6478 It accepts the following parameters:
6483 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6487 The expressions which specify the width and height of the box; if 0 they are interpreted as
6488 the input width and height. It defaults to 0.
6491 Specify the color of the box to write. For the general syntax of this option,
6492 check the "Color" section in the ffmpeg-utils manual. If the special
6493 value @code{invert} is used, the box edge color is the same as the
6494 video with inverted luma.
6497 The expression which sets the thickness of the box edge. Default value is @code{3}.
6499 See below for the list of accepted constants.
6502 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6503 following constants:
6507 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6511 horizontal and vertical chroma subsample values. For example for the
6512 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6516 The input width and height.
6519 The input sample aspect ratio.
6523 The x and y offset coordinates where the box is drawn.
6527 The width and height of the drawn box.
6530 The thickness of the drawn box.
6532 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6533 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6537 @subsection Examples
6541 Draw a black box around the edge of the input image:
6547 Draw a box with color red and an opacity of 50%:
6549 drawbox=10:20:200:60:red@@0.5
6552 The previous example can be specified as:
6554 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6558 Fill the box with pink color:
6560 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6564 Draw a 2-pixel red 2.40:1 mask:
6566 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
6572 Draw a grid on the input image.
6574 It accepts the following parameters:
6579 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6583 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6584 input width and height, respectively, minus @code{thickness}, so image gets
6585 framed. Default to 0.
6588 Specify the color of the grid. For the general syntax of this option,
6589 check the "Color" section in the ffmpeg-utils manual. If the special
6590 value @code{invert} is used, the grid color is the same as the
6591 video with inverted luma.
6594 The expression which sets the thickness of the grid line. Default value is @code{1}.
6596 See below for the list of accepted constants.
6599 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6600 following constants:
6604 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6608 horizontal and vertical chroma subsample values. For example for the
6609 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6613 The input grid cell width and height.
6616 The input sample aspect ratio.
6620 The x and y coordinates of some point of grid intersection (meant to configure offset).
6624 The width and height of the drawn cell.
6627 The thickness of the drawn cell.
6629 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6630 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6634 @subsection Examples
6638 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6640 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6644 Draw a white 3x3 grid with an opacity of 50%:
6646 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6653 Draw a text string or text from a specified file on top of a video, using the
6654 libfreetype library.
6656 To enable compilation of this filter, you need to configure FFmpeg with
6657 @code{--enable-libfreetype}.
6658 To enable default font fallback and the @var{font} option you need to
6659 configure FFmpeg with @code{--enable-libfontconfig}.
6660 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6661 @code{--enable-libfribidi}.
6665 It accepts the following parameters:
6670 Used to draw a box around text using the background color.
6671 The value must be either 1 (enable) or 0 (disable).
6672 The default value of @var{box} is 0.
6675 Set the width of the border to be drawn around the box using @var{boxcolor}.
6676 The default value of @var{boxborderw} is 0.
6679 The color to be used for drawing box around text. For the syntax of this
6680 option, check the "Color" section in the ffmpeg-utils manual.
6682 The default value of @var{boxcolor} is "white".
6685 Set the width of the border to be drawn around the text using @var{bordercolor}.
6686 The default value of @var{borderw} is 0.
6689 Set the color to be used for drawing border around text. For the syntax of this
6690 option, check the "Color" section in the ffmpeg-utils manual.
6692 The default value of @var{bordercolor} is "black".
6695 Select how the @var{text} is expanded. Can be either @code{none},
6696 @code{strftime} (deprecated) or
6697 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6701 If true, check and fix text coords to avoid clipping.
6704 The color to be used for drawing fonts. For the syntax of this option, check
6705 the "Color" section in the ffmpeg-utils manual.
6707 The default value of @var{fontcolor} is "black".
6709 @item fontcolor_expr
6710 String which is expanded the same way as @var{text} to obtain dynamic
6711 @var{fontcolor} value. By default this option has empty value and is not
6712 processed. When this option is set, it overrides @var{fontcolor} option.
6715 The font family to be used for drawing text. By default Sans.
6718 The font file to be used for drawing text. The path must be included.
6719 This parameter is mandatory if the fontconfig support is disabled.
6722 This option does not exist, please see the timeline system
6725 Draw the text applying alpha blending. The value can
6726 be a number between 0.0 and 1.0.
6727 The expression accepts the same variables @var{x, y} as well.
6728 The default value is 1.
6729 Please see @var{fontcolor_expr}.
6732 The font size to be used for drawing text.
6733 The default value of @var{fontsize} is 16.
6736 If set to 1, attempt to shape the text (for example, reverse the order of
6737 right-to-left text and join Arabic characters) before drawing it.
6738 Otherwise, just draw the text exactly as given.
6739 By default 1 (if supported).
6742 The flags to be used for loading the fonts.
6744 The flags map the corresponding flags supported by libfreetype, and are
6745 a combination of the following values:
6752 @item vertical_layout
6753 @item force_autohint
6756 @item ignore_global_advance_width
6758 @item ignore_transform
6764 Default value is "default".
6766 For more information consult the documentation for the FT_LOAD_*
6770 The color to be used for drawing a shadow behind the drawn text. For the
6771 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6773 The default value of @var{shadowcolor} is "black".
6777 The x and y offsets for the text shadow position with respect to the
6778 position of the text. They can be either positive or negative
6779 values. The default value for both is "0".
6782 The starting frame number for the n/frame_num variable. The default value
6786 The size in number of spaces to use for rendering the tab.
6790 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6791 format. It can be used with or without text parameter. @var{timecode_rate}
6792 option must be specified.
6794 @item timecode_rate, rate, r
6795 Set the timecode frame rate (timecode only).
6798 The text string to be drawn. The text must be a sequence of UTF-8
6800 This parameter is mandatory if no file is specified with the parameter
6804 A text file containing text to be drawn. The text must be a sequence
6805 of UTF-8 encoded characters.
6807 This parameter is mandatory if no text string is specified with the
6808 parameter @var{text}.
6810 If both @var{text} and @var{textfile} are specified, an error is thrown.
6813 If set to 1, the @var{textfile} will be reloaded before each frame.
6814 Be sure to update it atomically, or it may be read partially, or even fail.
6818 The expressions which specify the offsets where text will be drawn
6819 within the video frame. They are relative to the top/left border of the
6822 The default value of @var{x} and @var{y} is "0".
6824 See below for the list of accepted constants and functions.
6827 The parameters for @var{x} and @var{y} are expressions containing the
6828 following constants and functions:
6832 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6836 horizontal and vertical chroma subsample values. For example for the
6837 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6840 the height of each text line
6848 @item max_glyph_a, ascent
6849 the maximum distance from the baseline to the highest/upper grid
6850 coordinate used to place a glyph outline point, for all the rendered
6852 It is a positive value, due to the grid's orientation with the Y axis
6855 @item max_glyph_d, descent
6856 the maximum distance from the baseline to the lowest grid coordinate
6857 used to place a glyph outline point, for all the rendered glyphs.
6858 This is a negative value, due to the grid's orientation, with the Y axis
6862 maximum glyph height, that is the maximum height for all the glyphs
6863 contained in the rendered text, it is equivalent to @var{ascent} -
6867 maximum glyph width, that is the maximum width for all the glyphs
6868 contained in the rendered text
6871 the number of input frame, starting from 0
6873 @item rand(min, max)
6874 return a random number included between @var{min} and @var{max}
6877 The input sample aspect ratio.
6880 timestamp expressed in seconds, NAN if the input timestamp is unknown
6883 the height of the rendered text
6886 the width of the rendered text
6890 the x and y offset coordinates where the text is drawn.
6892 These parameters allow the @var{x} and @var{y} expressions to refer
6893 each other, so you can for example specify @code{y=x/dar}.
6896 @anchor{drawtext_expansion}
6897 @subsection Text expansion
6899 If @option{expansion} is set to @code{strftime},
6900 the filter recognizes strftime() sequences in the provided text and
6901 expands them accordingly. Check the documentation of strftime(). This
6902 feature is deprecated.
6904 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6906 If @option{expansion} is set to @code{normal} (which is the default),
6907 the following expansion mechanism is used.
6909 The backslash character @samp{\}, followed by any character, always expands to
6910 the second character.
6912 Sequences of the form @code{%@{...@}} are expanded. The text between the
6913 braces is a function name, possibly followed by arguments separated by ':'.
6914 If the arguments contain special characters or delimiters (':' or '@}'),
6915 they should be escaped.
6917 Note that they probably must also be escaped as the value for the
6918 @option{text} option in the filter argument string and as the filter
6919 argument in the filtergraph description, and possibly also for the shell,
6920 that makes up to four levels of escaping; using a text file avoids these
6923 The following functions are available:
6928 The expression evaluation result.
6930 It must take one argument specifying the expression to be evaluated,
6931 which accepts the same constants and functions as the @var{x} and
6932 @var{y} values. Note that not all constants should be used, for
6933 example the text size is not known when evaluating the expression, so
6934 the constants @var{text_w} and @var{text_h} will have an undefined
6937 @item expr_int_format, eif
6938 Evaluate the expression's value and output as formatted integer.
6940 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6941 The second argument specifies the output format. Allowed values are @samp{x},
6942 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6943 @code{printf} function.
6944 The third parameter is optional and sets the number of positions taken by the output.
6945 It can be used to add padding with zeros from the left.
6948 The time at which the filter is running, expressed in UTC.
6949 It can accept an argument: a strftime() format string.
6952 The time at which the filter is running, expressed in the local time zone.
6953 It can accept an argument: a strftime() format string.
6956 Frame metadata. Takes one or two arguments.
6958 The first argument is mandatory and specifies the metadata key.
6960 The second argument is optional and specifies a default value, used when the
6961 metadata key is not found or empty.
6964 The frame number, starting from 0.
6967 A 1 character description of the current picture type.
6970 The timestamp of the current frame.
6971 It can take up to three arguments.
6973 The first argument is the format of the timestamp; it defaults to @code{flt}
6974 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6975 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6976 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
6977 @code{localtime} stands for the timestamp of the frame formatted as
6978 local time zone time.
6980 The second argument is an offset added to the timestamp.
6982 If the format is set to @code{localtime} or @code{gmtime},
6983 a third argument may be supplied: a strftime() format string.
6984 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
6987 @subsection Examples
6991 Draw "Test Text" with font FreeSerif, using the default values for the
6992 optional parameters.
6995 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
6999 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7000 and y=50 (counting from the top-left corner of the screen), text is
7001 yellow with a red box around it. Both the text and the box have an
7005 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7006 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7009 Note that the double quotes are not necessary if spaces are not used
7010 within the parameter list.
7013 Show the text at the center of the video frame:
7015 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7019 Show the text at a random position, switching to a new position every 30 seconds:
7021 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)"
7025 Show a text line sliding from right to left in the last row of the video
7026 frame. The file @file{LONG_LINE} is assumed to contain a single line
7029 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7033 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7035 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7039 Draw a single green letter "g", at the center of the input video.
7040 The glyph baseline is placed at half screen height.
7042 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7046 Show text for 1 second every 3 seconds:
7048 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7052 Use fontconfig to set the font. Note that the colons need to be escaped.
7054 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7058 Print the date of a real-time encoding (see strftime(3)):
7060 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7064 Show text fading in and out (appearing/disappearing):
7067 DS=1.0 # display start
7068 DE=10.0 # display end
7069 FID=1.5 # fade in duration
7070 FOD=5 # fade out duration
7071 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 @}"
7076 For more information about libfreetype, check:
7077 @url{http://www.freetype.org/}.
7079 For more information about fontconfig, check:
7080 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7082 For more information about libfribidi, check:
7083 @url{http://fribidi.org/}.
7087 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7089 The filter accepts the following options:
7094 Set low and high threshold values used by the Canny thresholding
7097 The high threshold selects the "strong" edge pixels, which are then
7098 connected through 8-connectivity with the "weak" edge pixels selected
7099 by the low threshold.
7101 @var{low} and @var{high} threshold values must be chosen in the range
7102 [0,1], and @var{low} should be lesser or equal to @var{high}.
7104 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7108 Define the drawing mode.
7112 Draw white/gray wires on black background.
7115 Mix the colors to create a paint/cartoon effect.
7118 Default value is @var{wires}.
7121 @subsection Examples
7125 Standard edge detection with custom values for the hysteresis thresholding:
7127 edgedetect=low=0.1:high=0.4
7131 Painting effect without thresholding:
7133 edgedetect=mode=colormix:high=0
7138 Set brightness, contrast, saturation and approximate gamma adjustment.
7140 The filter accepts the following options:
7144 Set the contrast expression. The value must be a float value in range
7145 @code{-2.0} to @code{2.0}. The default value is "1".
7148 Set the brightness expression. The value must be a float value in
7149 range @code{-1.0} to @code{1.0}. The default value is "0".
7152 Set the saturation expression. The value must be a float in
7153 range @code{0.0} to @code{3.0}. The default value is "1".
7156 Set the gamma expression. The value must be a float in range
7157 @code{0.1} to @code{10.0}. The default value is "1".
7160 Set the gamma expression for red. The value must be a float in
7161 range @code{0.1} to @code{10.0}. The default value is "1".
7164 Set the gamma expression for green. The value must be a float in range
7165 @code{0.1} to @code{10.0}. The default value is "1".
7168 Set the gamma expression for blue. The value must be a float in range
7169 @code{0.1} to @code{10.0}. The default value is "1".
7172 Set the gamma weight expression. It can be used to reduce the effect
7173 of a high gamma value on bright image areas, e.g. keep them from
7174 getting overamplified and just plain white. The value must be a float
7175 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7176 gamma correction all the way down while @code{1.0} leaves it at its
7177 full strength. Default is "1".
7180 Set when the expressions for brightness, contrast, saturation and
7181 gamma expressions are evaluated.
7183 It accepts the following values:
7186 only evaluate expressions once during the filter initialization or
7187 when a command is processed
7190 evaluate expressions for each incoming frame
7193 Default value is @samp{init}.
7196 The expressions accept the following parameters:
7199 frame count of the input frame starting from 0
7202 byte position of the corresponding packet in the input file, NAN if
7206 frame rate of the input video, NAN if the input frame rate is unknown
7209 timestamp expressed in seconds, NAN if the input timestamp is unknown
7212 @subsection Commands
7213 The filter supports the following commands:
7217 Set the contrast expression.
7220 Set the brightness expression.
7223 Set the saturation expression.
7226 Set the gamma expression.
7229 Set the gamma_r expression.
7232 Set gamma_g expression.
7235 Set gamma_b expression.
7238 Set gamma_weight expression.
7240 The command accepts the same syntax of the corresponding option.
7242 If the specified expression is not valid, it is kept at its current
7249 Apply erosion effect to the video.
7251 This filter replaces the pixel by the local(3x3) minimum.
7253 It accepts the following options:
7260 Limit the maximum change for each plane, default is 65535.
7261 If 0, plane will remain unchanged.
7264 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7267 Flags to local 3x3 coordinates maps like this:
7274 @section extractplanes
7276 Extract color channel components from input video stream into
7277 separate grayscale video streams.
7279 The filter accepts the following option:
7283 Set plane(s) to extract.
7285 Available values for planes are:
7296 Choosing planes not available in the input will result in an error.
7297 That means you cannot select @code{r}, @code{g}, @code{b} planes
7298 with @code{y}, @code{u}, @code{v} planes at same time.
7301 @subsection Examples
7305 Extract luma, u and v color channel component from input video frame
7306 into 3 grayscale outputs:
7308 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
7314 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7316 For each input image, the filter will compute the optimal mapping from
7317 the input to the output given the codebook length, that is the number
7318 of distinct output colors.
7320 This filter accepts the following options.
7323 @item codebook_length, l
7324 Set codebook length. The value must be a positive integer, and
7325 represents the number of distinct output colors. Default value is 256.
7328 Set the maximum number of iterations to apply for computing the optimal
7329 mapping. The higher the value the better the result and the higher the
7330 computation time. Default value is 1.
7333 Set a random seed, must be an integer included between 0 and
7334 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7335 will try to use a good random seed on a best effort basis.
7338 Set pal8 output pixel format. This option does not work with codebook
7339 length greater than 256.
7344 Apply a fade-in/out effect to the input video.
7346 It accepts the following parameters:
7350 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7352 Default is @code{in}.
7354 @item start_frame, s
7355 Specify the number of the frame to start applying the fade
7356 effect at. Default is 0.
7359 The number of frames that the fade effect lasts. At the end of the
7360 fade-in effect, the output video will have the same intensity as the input video.
7361 At the end of the fade-out transition, the output video will be filled with the
7362 selected @option{color}.
7366 If set to 1, fade only alpha channel, if one exists on the input.
7369 @item start_time, st
7370 Specify the timestamp (in seconds) of the frame to start to apply the fade
7371 effect. If both start_frame and start_time are specified, the fade will start at
7372 whichever comes last. Default is 0.
7375 The number of seconds for which the fade effect has to last. At the end of the
7376 fade-in effect the output video will have the same intensity as the input video,
7377 at the end of the fade-out transition the output video will be filled with the
7378 selected @option{color}.
7379 If both duration and nb_frames are specified, duration is used. Default is 0
7380 (nb_frames is used by default).
7383 Specify the color of the fade. Default is "black".
7386 @subsection Examples
7390 Fade in the first 30 frames of video:
7395 The command above is equivalent to:
7401 Fade out the last 45 frames of a 200-frame video:
7404 fade=type=out:start_frame=155:nb_frames=45
7408 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7410 fade=in:0:25, fade=out:975:25
7414 Make the first 5 frames yellow, then fade in from frame 5-24:
7416 fade=in:5:20:color=yellow
7420 Fade in alpha over first 25 frames of video:
7422 fade=in:0:25:alpha=1
7426 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7428 fade=t=in:st=5.5:d=0.5
7434 Apply arbitrary expressions to samples in frequency domain
7438 Adjust the dc value (gain) of the luma plane of the image. The filter
7439 accepts an integer value in range @code{0} to @code{1000}. The default
7440 value is set to @code{0}.
7443 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7444 filter accepts an integer value in range @code{0} to @code{1000}. The
7445 default value is set to @code{0}.
7448 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7449 filter accepts an integer value in range @code{0} to @code{1000}. The
7450 default value is set to @code{0}.
7453 Set the frequency domain weight expression for the luma plane.
7456 Set the frequency domain weight expression for the 1st chroma plane.
7459 Set the frequency domain weight expression for the 2nd chroma plane.
7461 The filter accepts the following variables:
7464 The coordinates of the current sample.
7468 The width and height of the image.
7471 @subsection Examples
7477 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7483 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7489 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7495 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7502 Extract a single field from an interlaced image using stride
7503 arithmetic to avoid wasting CPU time. The output frames are marked as
7506 The filter accepts the following options:
7510 Specify whether to extract the top (if the value is @code{0} or
7511 @code{top}) or the bottom field (if the value is @code{1} or
7517 Create new frames by copying the top and bottom fields from surrounding frames
7518 supplied as numbers by the hint file.
7522 Set file containing hints: absolute/relative frame numbers.
7524 There must be one line for each frame in a clip. Each line must contain two
7525 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7526 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7527 is current frame number for @code{absolute} mode or out of [-1, 1] range
7528 for @code{relative} mode. First number tells from which frame to pick up top
7529 field and second number tells from which frame to pick up bottom field.
7531 If optionally followed by @code{+} output frame will be marked as interlaced,
7532 else if followed by @code{-} output frame will be marked as progressive, else
7533 it will be marked same as input frame.
7534 If line starts with @code{#} or @code{;} that line is skipped.
7537 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7540 Example of first several lines of @code{hint} file for @code{relative} mode:
7543 1,0 - # second frame, use third's frame top field and second's frame bottom field
7544 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7561 Field matching filter for inverse telecine. It is meant to reconstruct the
7562 progressive frames from a telecined stream. The filter does not drop duplicated
7563 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7564 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7566 The separation of the field matching and the decimation is notably motivated by
7567 the possibility of inserting a de-interlacing filter fallback between the two.
7568 If the source has mixed telecined and real interlaced content,
7569 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7570 But these remaining combed frames will be marked as interlaced, and thus can be
7571 de-interlaced by a later filter such as @ref{yadif} before decimation.
7573 In addition to the various configuration options, @code{fieldmatch} can take an
7574 optional second stream, activated through the @option{ppsrc} option. If
7575 enabled, the frames reconstruction will be based on the fields and frames from
7576 this second stream. This allows the first input to be pre-processed in order to
7577 help the various algorithms of the filter, while keeping the output lossless
7578 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7579 or brightness/contrast adjustments can help.
7581 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7582 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7583 which @code{fieldmatch} is based on. While the semantic and usage are very
7584 close, some behaviour and options names can differ.
7586 The @ref{decimate} filter currently only works for constant frame rate input.
7587 If your input has mixed telecined (30fps) and progressive content with a lower
7588 framerate like 24fps use the following filterchain to produce the necessary cfr
7589 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7591 The filter accepts the following options:
7595 Specify the assumed field order of the input stream. Available values are:
7599 Auto detect parity (use FFmpeg's internal parity value).
7601 Assume bottom field first.
7603 Assume top field first.
7606 Note that it is sometimes recommended not to trust the parity announced by the
7609 Default value is @var{auto}.
7612 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7613 sense that it won't risk creating jerkiness due to duplicate frames when
7614 possible, but if there are bad edits or blended fields it will end up
7615 outputting combed frames when a good match might actually exist. On the other
7616 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7617 but will almost always find a good frame if there is one. The other values are
7618 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7619 jerkiness and creating duplicate frames versus finding good matches in sections
7620 with bad edits, orphaned fields, blended fields, etc.
7622 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7624 Available values are:
7628 2-way matching (p/c)
7630 2-way matching, and trying 3rd match if still combed (p/c + n)
7632 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7634 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7635 still combed (p/c + n + u/b)
7637 3-way matching (p/c/n)
7639 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7640 detected as combed (p/c/n + u/b)
7643 The parenthesis at the end indicate the matches that would be used for that
7644 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7647 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7650 Default value is @var{pc_n}.
7653 Mark the main input stream as a pre-processed input, and enable the secondary
7654 input stream as the clean source to pick the fields from. See the filter
7655 introduction for more details. It is similar to the @option{clip2} feature from
7658 Default value is @code{0} (disabled).
7661 Set the field to match from. It is recommended to set this to the same value as
7662 @option{order} unless you experience matching failures with that setting. In
7663 certain circumstances changing the field that is used to match from can have a
7664 large impact on matching performance. Available values are:
7668 Automatic (same value as @option{order}).
7670 Match from the bottom field.
7672 Match from the top field.
7675 Default value is @var{auto}.
7678 Set whether or not chroma is included during the match comparisons. In most
7679 cases it is recommended to leave this enabled. You should set this to @code{0}
7680 only if your clip has bad chroma problems such as heavy rainbowing or other
7681 artifacts. Setting this to @code{0} could also be used to speed things up at
7682 the cost of some accuracy.
7684 Default value is @code{1}.
7688 These define an exclusion band which excludes the lines between @option{y0} and
7689 @option{y1} from being included in the field matching decision. An exclusion
7690 band can be used to ignore subtitles, a logo, or other things that may
7691 interfere with the matching. @option{y0} sets the starting scan line and
7692 @option{y1} sets the ending line; all lines in between @option{y0} and
7693 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7694 @option{y0} and @option{y1} to the same value will disable the feature.
7695 @option{y0} and @option{y1} defaults to @code{0}.
7698 Set the scene change detection threshold as a percentage of maximum change on
7699 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7700 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7701 @option{scthresh} is @code{[0.0, 100.0]}.
7703 Default value is @code{12.0}.
7706 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7707 account the combed scores of matches when deciding what match to use as the
7708 final match. Available values are:
7712 No final matching based on combed scores.
7714 Combed scores are only used when a scene change is detected.
7716 Use combed scores all the time.
7719 Default is @var{sc}.
7722 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7723 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7724 Available values are:
7728 No forced calculation.
7730 Force p/c/n calculations.
7732 Force p/c/n/u/b calculations.
7735 Default value is @var{none}.
7738 This is the area combing threshold used for combed frame detection. This
7739 essentially controls how "strong" or "visible" combing must be to be detected.
7740 Larger values mean combing must be more visible and smaller values mean combing
7741 can be less visible or strong and still be detected. Valid settings are from
7742 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7743 be detected as combed). This is basically a pixel difference value. A good
7744 range is @code{[8, 12]}.
7746 Default value is @code{9}.
7749 Sets whether or not chroma is considered in the combed frame decision. Only
7750 disable this if your source has chroma problems (rainbowing, etc.) that are
7751 causing problems for the combed frame detection with chroma enabled. Actually,
7752 using @option{chroma}=@var{0} is usually more reliable, except for the case
7753 where there is chroma only combing in the source.
7755 Default value is @code{0}.
7759 Respectively set the x-axis and y-axis size of the window used during combed
7760 frame detection. This has to do with the size of the area in which
7761 @option{combpel} pixels are required to be detected as combed for a frame to be
7762 declared combed. See the @option{combpel} parameter description for more info.
7763 Possible values are any number that is a power of 2 starting at 4 and going up
7766 Default value is @code{16}.
7769 The number of combed pixels inside any of the @option{blocky} by
7770 @option{blockx} size blocks on the frame for the frame to be detected as
7771 combed. While @option{cthresh} controls how "visible" the combing must be, this
7772 setting controls "how much" combing there must be in any localized area (a
7773 window defined by the @option{blockx} and @option{blocky} settings) on the
7774 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7775 which point no frames will ever be detected as combed). This setting is known
7776 as @option{MI} in TFM/VFM vocabulary.
7778 Default value is @code{80}.
7781 @anchor{p/c/n/u/b meaning}
7782 @subsection p/c/n/u/b meaning
7784 @subsubsection p/c/n
7786 We assume the following telecined stream:
7789 Top fields: 1 2 2 3 4
7790 Bottom fields: 1 2 3 4 4
7793 The numbers correspond to the progressive frame the fields relate to. Here, the
7794 first two frames are progressive, the 3rd and 4th are combed, and so on.
7796 When @code{fieldmatch} is configured to run a matching from bottom
7797 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7802 B 1 2 3 4 4 <-- matching reference
7811 As a result of the field matching, we can see that some frames get duplicated.
7812 To perform a complete inverse telecine, you need to rely on a decimation filter
7813 after this operation. See for instance the @ref{decimate} filter.
7815 The same operation now matching from top fields (@option{field}=@var{top})
7820 T 1 2 2 3 4 <-- matching reference
7830 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7831 basically, they refer to the frame and field of the opposite parity:
7834 @item @var{p} matches the field of the opposite parity in the previous frame
7835 @item @var{c} matches the field of the opposite parity in the current frame
7836 @item @var{n} matches the field of the opposite parity in the next frame
7841 The @var{u} and @var{b} matching are a bit special in the sense that they match
7842 from the opposite parity flag. In the following examples, we assume that we are
7843 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7844 'x' is placed above and below each matched fields.
7846 With bottom matching (@option{field}=@var{bottom}):
7851 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7852 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7860 With top matching (@option{field}=@var{top}):
7865 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7866 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7874 @subsection Examples
7876 Simple IVTC of a top field first telecined stream:
7878 fieldmatch=order=tff:combmatch=none, decimate
7881 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7883 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7888 Transform the field order of the input video.
7890 It accepts the following parameters:
7895 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7896 for bottom field first.
7899 The default value is @samp{tff}.
7901 The transformation is done by shifting the picture content up or down
7902 by one line, and filling the remaining line with appropriate picture content.
7903 This method is consistent with most broadcast field order converters.
7905 If the input video is not flagged as being interlaced, or it is already
7906 flagged as being of the required output field order, then this filter does
7907 not alter the incoming video.
7909 It is very useful when converting to or from PAL DV material,
7910 which is bottom field first.
7914 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7917 @section fifo, afifo
7919 Buffer input images and send them when they are requested.
7921 It is mainly useful when auto-inserted by the libavfilter
7924 It does not take parameters.
7928 Find a rectangular object
7930 It accepts the following options:
7934 Filepath of the object image, needs to be in gray8.
7937 Detection threshold, default is 0.5.
7940 Number of mipmaps, default is 3.
7942 @item xmin, ymin, xmax, ymax
7943 Specifies the rectangle in which to search.
7946 @subsection Examples
7950 Generate a representative palette of a given video using @command{ffmpeg}:
7952 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7958 Cover a rectangular object
7960 It accepts the following options:
7964 Filepath of the optional cover image, needs to be in yuv420.
7969 It accepts the following values:
7972 cover it by the supplied image
7974 cover it by interpolating the surrounding pixels
7977 Default value is @var{blur}.
7980 @subsection Examples
7984 Generate a representative palette of a given video using @command{ffmpeg}:
7986 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7993 Convert the input video to one of the specified pixel formats.
7994 Libavfilter will try to pick one that is suitable as input to
7997 It accepts the following parameters:
8001 A '|'-separated list of pixel format names, such as
8002 "pix_fmts=yuv420p|monow|rgb24".
8006 @subsection Examples
8010 Convert the input video to the @var{yuv420p} format
8012 format=pix_fmts=yuv420p
8015 Convert the input video to any of the formats in the list
8017 format=pix_fmts=yuv420p|yuv444p|yuv410p
8024 Convert the video to specified constant frame rate by duplicating or dropping
8025 frames as necessary.
8027 It accepts the following parameters:
8031 The desired output frame rate. The default is @code{25}.
8036 Possible values are:
8039 zero round towards 0
8043 round towards -infinity
8045 round towards +infinity
8049 The default is @code{near}.
8052 Assume the first PTS should be the given value, in seconds. This allows for
8053 padding/trimming at the start of stream. By default, no assumption is made
8054 about the first frame's expected PTS, so no padding or trimming is done.
8055 For example, this could be set to 0 to pad the beginning with duplicates of
8056 the first frame if a video stream starts after the audio stream or to trim any
8057 frames with a negative PTS.
8061 Alternatively, the options can be specified as a flat string:
8062 @var{fps}[:@var{round}].
8064 See also the @ref{setpts} filter.
8066 @subsection Examples
8070 A typical usage in order to set the fps to 25:
8076 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8078 fps=fps=film:round=near
8084 Pack two different video streams into a stereoscopic video, setting proper
8085 metadata on supported codecs. The two views should have the same size and
8086 framerate and processing will stop when the shorter video ends. Please note
8087 that you may conveniently adjust view properties with the @ref{scale} and
8090 It accepts the following parameters:
8094 The desired packing format. Supported values are:
8099 The views are next to each other (default).
8102 The views are on top of each other.
8105 The views are packed by line.
8108 The views are packed by column.
8111 The views are temporally interleaved.
8120 # Convert left and right views into a frame-sequential video
8121 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8123 # Convert views into a side-by-side video with the same output resolution as the input
8124 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
8129 Change the frame rate by interpolating new video output frames from the source
8132 This filter is not designed to function correctly with interlaced media. If
8133 you wish to change the frame rate of interlaced media then you are required
8134 to deinterlace before this filter and re-interlace after this filter.
8136 A description of the accepted options follows.
8140 Specify the output frames per second. This option can also be specified
8141 as a value alone. The default is @code{50}.
8144 Specify the start of a range where the output frame will be created as a
8145 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8146 the default is @code{15}.
8149 Specify the end of a range where the output frame will be created as a
8150 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8151 the default is @code{240}.
8154 Specify the level at which a scene change is detected as a value between
8155 0 and 100 to indicate a new scene; a low value reflects a low
8156 probability for the current frame to introduce a new scene, while a higher
8157 value means the current frame is more likely to be one.
8158 The default is @code{7}.
8161 Specify flags influencing the filter process.
8163 Available value for @var{flags} is:
8166 @item scene_change_detect, scd
8167 Enable scene change detection using the value of the option @var{scene}.
8168 This flag is enabled by default.
8174 Select one frame every N-th frame.
8176 This filter accepts the following option:
8179 Select frame after every @code{step} frames.
8180 Allowed values are positive integers higher than 0. Default value is @code{1}.
8186 Apply a frei0r effect to the input video.
8188 To enable the compilation of this filter, you need to install the frei0r
8189 header and configure FFmpeg with @code{--enable-frei0r}.
8191 It accepts the following parameters:
8196 The name of the frei0r effect to load. If the environment variable
8197 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8198 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8199 Otherwise, the standard frei0r paths are searched, in this order:
8200 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8201 @file{/usr/lib/frei0r-1/}.
8204 A '|'-separated list of parameters to pass to the frei0r effect.
8208 A frei0r effect parameter can be a boolean (its value is either
8209 "y" or "n"), a double, a color (specified as
8210 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8211 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8212 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8213 @var{X} and @var{Y} are floating point numbers) and/or a string.
8215 The number and types of parameters depend on the loaded effect. If an
8216 effect parameter is not specified, the default value is set.
8218 @subsection Examples
8222 Apply the distort0r effect, setting the first two double parameters:
8224 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8228 Apply the colordistance effect, taking a color as the first parameter:
8230 frei0r=colordistance:0.2/0.3/0.4
8231 frei0r=colordistance:violet
8232 frei0r=colordistance:0x112233
8236 Apply the perspective effect, specifying the top left and top right image
8239 frei0r=perspective:0.2/0.2|0.8/0.2
8243 For more information, see
8244 @url{http://frei0r.dyne.org}
8248 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8250 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8251 processing filter, one of them is performed once per block, not per pixel.
8252 This allows for much higher speed.
8254 The filter accepts the following options:
8258 Set quality. This option defines the number of levels for averaging. It accepts
8259 an integer in the range 4-5. Default value is @code{4}.
8262 Force a constant quantization parameter. It accepts an integer in range 0-63.
8263 If not set, the filter will use the QP from the video stream (if available).
8266 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8267 more details but also more artifacts, while higher values make the image smoother
8268 but also blurrier. Default value is @code{0} − PSNR optimal.
8271 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8272 option may cause flicker since the B-Frames have often larger QP. Default is
8273 @code{0} (not enabled).
8279 Apply Gaussian blur filter.
8281 The filter accepts the following options:
8285 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8288 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8291 Set which planes to filter. By default all planes are filtered.
8294 Set vertical sigma, if negative it will be same as @code{sigma}.
8295 Default is @code{-1}.
8300 The filter accepts the following options:
8304 Set the luminance expression.
8306 Set the chrominance blue expression.
8308 Set the chrominance red expression.
8310 Set the alpha expression.
8312 Set the red expression.
8314 Set the green expression.
8316 Set the blue expression.
8319 The colorspace is selected according to the specified options. If one
8320 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8321 options is specified, the filter will automatically select a YCbCr
8322 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8323 @option{blue_expr} options is specified, it will select an RGB
8326 If one of the chrominance expression is not defined, it falls back on the other
8327 one. If no alpha expression is specified it will evaluate to opaque value.
8328 If none of chrominance expressions are specified, they will evaluate
8329 to the luminance expression.
8331 The expressions can use the following variables and functions:
8335 The sequential number of the filtered frame, starting from @code{0}.
8339 The coordinates of the current sample.
8343 The width and height of the image.
8347 Width and height scale depending on the currently filtered plane. It is the
8348 ratio between the corresponding luma plane number of pixels and the current
8349 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8350 @code{0.5,0.5} for chroma planes.
8353 Time of the current frame, expressed in seconds.
8356 Return the value of the pixel at location (@var{x},@var{y}) of the current
8360 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8364 Return the value of the pixel at location (@var{x},@var{y}) of the
8365 blue-difference chroma plane. Return 0 if there is no such plane.
8368 Return the value of the pixel at location (@var{x},@var{y}) of the
8369 red-difference chroma plane. Return 0 if there is no such plane.
8374 Return the value of the pixel at location (@var{x},@var{y}) of the
8375 red/green/blue component. Return 0 if there is no such component.
8378 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8379 plane. Return 0 if there is no such plane.
8382 For functions, if @var{x} and @var{y} are outside the area, the value will be
8383 automatically clipped to the closer edge.
8385 @subsection Examples
8389 Flip the image horizontally:
8395 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8396 wavelength of 100 pixels:
8398 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8402 Generate a fancy enigmatic moving light:
8404 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
8408 Generate a quick emboss effect:
8410 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8414 Modify RGB components depending on pixel position:
8416 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8420 Create a radial gradient that is the same size as the input (also see
8421 the @ref{vignette} filter):
8423 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8429 Fix the banding artifacts that are sometimes introduced into nearly flat
8430 regions by truncation to 8-bit color depth.
8431 Interpolate the gradients that should go where the bands are, and
8434 It is designed for playback only. Do not use it prior to
8435 lossy compression, because compression tends to lose the dither and
8436 bring back the bands.
8438 It accepts the following parameters:
8443 The maximum amount by which the filter will change any one pixel. This is also
8444 the threshold for detecting nearly flat regions. Acceptable values range from
8445 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8449 The neighborhood to fit the gradient to. A larger radius makes for smoother
8450 gradients, but also prevents the filter from modifying the pixels near detailed
8451 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8452 values will be clipped to the valid range.
8456 Alternatively, the options can be specified as a flat string:
8457 @var{strength}[:@var{radius}]
8459 @subsection Examples
8463 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8469 Specify radius, omitting the strength (which will fall-back to the default
8480 Apply a Hald CLUT to a video stream.
8482 First input is the video stream to process, and second one is the Hald CLUT.
8483 The Hald CLUT input can be a simple picture or a complete video stream.
8485 The filter accepts the following options:
8489 Force termination when the shortest input terminates. Default is @code{0}.
8491 Continue applying the last CLUT after the end of the stream. A value of
8492 @code{0} disable the filter after the last frame of the CLUT is reached.
8493 Default is @code{1}.
8496 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8497 filters share the same internals).
8499 More information about the Hald CLUT can be found on Eskil Steenberg's website
8500 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8502 @subsection Workflow examples
8504 @subsubsection Hald CLUT video stream
8506 Generate an identity Hald CLUT stream altered with various effects:
8508 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
8511 Note: make sure you use a lossless codec.
8513 Then use it with @code{haldclut} to apply it on some random stream:
8515 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8518 The Hald CLUT will be applied to the 10 first seconds (duration of
8519 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8520 to the remaining frames of the @code{mandelbrot} stream.
8522 @subsubsection Hald CLUT with preview
8524 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8525 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8526 biggest possible square starting at the top left of the picture. The remaining
8527 padding pixels (bottom or right) will be ignored. This area can be used to add
8528 a preview of the Hald CLUT.
8530 Typically, the following generated Hald CLUT will be supported by the
8531 @code{haldclut} filter:
8534 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8535 pad=iw+320 [padded_clut];
8536 smptebars=s=320x256, split [a][b];
8537 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8538 [main][b] overlay=W-320" -frames:v 1 clut.png
8541 It contains the original and a preview of the effect of the CLUT: SMPTE color
8542 bars are displayed on the right-top, and below the same color bars processed by
8545 Then, the effect of this Hald CLUT can be visualized with:
8547 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8552 Flip the input video horizontally.
8554 For example, to horizontally flip the input video with @command{ffmpeg}:
8556 ffmpeg -i in.avi -vf "hflip" out.avi
8560 This filter applies a global color histogram equalization on a
8563 It can be used to correct video that has a compressed range of pixel
8564 intensities. The filter redistributes the pixel intensities to
8565 equalize their distribution across the intensity range. It may be
8566 viewed as an "automatically adjusting contrast filter". This filter is
8567 useful only for correcting degraded or poorly captured source
8570 The filter accepts the following options:
8574 Determine the amount of equalization to be applied. As the strength
8575 is reduced, the distribution of pixel intensities more-and-more
8576 approaches that of the input frame. The value must be a float number
8577 in the range [0,1] and defaults to 0.200.
8580 Set the maximum intensity that can generated and scale the output
8581 values appropriately. The strength should be set as desired and then
8582 the intensity can be limited if needed to avoid washing-out. The value
8583 must be a float number in the range [0,1] and defaults to 0.210.
8586 Set the antibanding level. If enabled the filter will randomly vary
8587 the luminance of output pixels by a small amount to avoid banding of
8588 the histogram. Possible values are @code{none}, @code{weak} or
8589 @code{strong}. It defaults to @code{none}.
8594 Compute and draw a color distribution histogram for the input video.
8596 The computed histogram is a representation of the color component
8597 distribution in an image.
8599 Standard histogram displays the color components distribution in an image.
8600 Displays color graph for each color component. Shows distribution of
8601 the Y, U, V, A or R, G, B components, depending on input format, in the
8602 current frame. Below each graph a color component scale meter is shown.
8604 The filter accepts the following options:
8608 Set height of level. Default value is @code{200}.
8609 Allowed range is [50, 2048].
8612 Set height of color scale. Default value is @code{12}.
8613 Allowed range is [0, 40].
8617 It accepts the following values:
8620 Per color component graphs are placed below each other.
8623 Presents information identical to that in the @code{parade}, except
8624 that the graphs representing color components are superimposed directly
8627 Default is @code{parade}.
8630 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8631 Default is @code{linear}.
8634 Set what color components to display.
8635 Default is @code{7}.
8638 Set foreground opacity. Default is @code{0.7}.
8641 Set background opacity. Default is @code{0.5}.
8644 @subsection Examples
8649 Calculate and draw histogram:
8651 ffplay -i input -vf histogram
8659 This is a high precision/quality 3d denoise filter. It aims to reduce
8660 image noise, producing smooth images and making still images really
8661 still. It should enhance compressibility.
8663 It accepts the following optional parameters:
8667 A non-negative floating point number which specifies spatial luma strength.
8670 @item chroma_spatial
8671 A non-negative floating point number which specifies spatial chroma strength.
8672 It defaults to 3.0*@var{luma_spatial}/4.0.
8675 A floating point number which specifies luma temporal strength. It defaults to
8676 6.0*@var{luma_spatial}/4.0.
8679 A floating point number which specifies chroma temporal strength. It defaults to
8680 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8683 @anchor{hwupload_cuda}
8684 @section hwupload_cuda
8686 Upload system memory frames to a CUDA device.
8688 It accepts the following optional parameters:
8692 The number of the CUDA device to use
8697 Apply a high-quality magnification filter designed for pixel art. This filter
8698 was originally created by Maxim Stepin.
8700 It accepts the following option:
8704 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8705 @code{hq3x} and @code{4} for @code{hq4x}.
8706 Default is @code{3}.
8710 Stack input videos horizontally.
8712 All streams must be of same pixel format and of same height.
8714 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8715 to create same output.
8717 The filter accept the following option:
8721 Set number of input streams. Default is 2.
8724 If set to 1, force the output to terminate when the shortest input
8725 terminates. Default value is 0.
8730 Modify the hue and/or the saturation of the input.
8732 It accepts the following parameters:
8736 Specify the hue angle as a number of degrees. It accepts an expression,
8737 and defaults to "0".
8740 Specify the saturation in the [-10,10] range. It accepts an expression and
8744 Specify the hue angle as a number of radians. It accepts an
8745 expression, and defaults to "0".
8748 Specify the brightness in the [-10,10] range. It accepts an expression and
8752 @option{h} and @option{H} are mutually exclusive, and can't be
8753 specified at the same time.
8755 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8756 expressions containing the following constants:
8760 frame count of the input frame starting from 0
8763 presentation timestamp of the input frame expressed in time base units
8766 frame rate of the input video, NAN if the input frame rate is unknown
8769 timestamp expressed in seconds, NAN if the input timestamp is unknown
8772 time base of the input video
8775 @subsection Examples
8779 Set the hue to 90 degrees and the saturation to 1.0:
8785 Same command but expressing the hue in radians:
8791 Rotate hue and make the saturation swing between 0
8792 and 2 over a period of 1 second:
8794 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8798 Apply a 3 seconds saturation fade-in effect starting at 0:
8803 The general fade-in expression can be written as:
8805 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8809 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8811 hue="s=max(0\, min(1\, (8-t)/3))"
8814 The general fade-out expression can be written as:
8816 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8821 @subsection Commands
8823 This filter supports the following commands:
8829 Modify the hue and/or the saturation and/or brightness of the input video.
8830 The command accepts the same syntax of the corresponding option.
8832 If the specified expression is not valid, it is kept at its current
8838 Grow first stream into second stream by connecting components.
8839 This allows to build more robust edge masks.
8841 This filter accepts the following options:
8845 Set which planes will be processed as bitmap, unprocessed planes will be
8846 copied from first stream.
8847 By default value 0xf, all planes will be processed.
8850 Set threshold which is used in filtering. If pixel component value is higher than
8851 this value filter algorithm for connecting components is activated.
8852 By default value is 0.
8857 Detect video interlacing type.
8859 This filter tries to detect if the input frames are interlaced, progressive,
8860 top or bottom field first. It will also try to detect fields that are
8861 repeated between adjacent frames (a sign of telecine).
8863 Single frame detection considers only immediately adjacent frames when classifying each frame.
8864 Multiple frame detection incorporates the classification history of previous frames.
8866 The filter will log these metadata values:
8869 @item single.current_frame
8870 Detected type of current frame using single-frame detection. One of:
8871 ``tff'' (top field first), ``bff'' (bottom field first),
8872 ``progressive'', or ``undetermined''
8875 Cumulative number of frames detected as top field first using single-frame detection.
8878 Cumulative number of frames detected as top field first using multiple-frame detection.
8881 Cumulative number of frames detected as bottom field first using single-frame detection.
8883 @item multiple.current_frame
8884 Detected type of current frame using multiple-frame detection. One of:
8885 ``tff'' (top field first), ``bff'' (bottom field first),
8886 ``progressive'', or ``undetermined''
8889 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8891 @item single.progressive
8892 Cumulative number of frames detected as progressive using single-frame detection.
8894 @item multiple.progressive
8895 Cumulative number of frames detected as progressive using multiple-frame detection.
8897 @item single.undetermined
8898 Cumulative number of frames that could not be classified using single-frame detection.
8900 @item multiple.undetermined
8901 Cumulative number of frames that could not be classified using multiple-frame detection.
8903 @item repeated.current_frame
8904 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8906 @item repeated.neither
8907 Cumulative number of frames with no repeated field.
8910 Cumulative number of frames with the top field repeated from the previous frame's top field.
8912 @item repeated.bottom
8913 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8916 The filter accepts the following options:
8920 Set interlacing threshold.
8922 Set progressive threshold.
8924 Threshold for repeated field detection.
8926 Number of frames after which a given frame's contribution to the
8927 statistics is halved (i.e., it contributes only 0.5 to its
8928 classification). The default of 0 means that all frames seen are given
8929 full weight of 1.0 forever.
8930 @item analyze_interlaced_flag
8931 When this is not 0 then idet will use the specified number of frames to determine
8932 if the interlaced flag is accurate, it will not count undetermined frames.
8933 If the flag is found to be accurate it will be used without any further
8934 computations, if it is found to be inaccurate it will be cleared without any
8935 further computations. This allows inserting the idet filter as a low computational
8936 method to clean up the interlaced flag
8941 Deinterleave or interleave fields.
8943 This filter allows one to process interlaced images fields without
8944 deinterlacing them. Deinterleaving splits the input frame into 2
8945 fields (so called half pictures). Odd lines are moved to the top
8946 half of the output image, even lines to the bottom half.
8947 You can process (filter) them independently and then re-interleave them.
8949 The filter accepts the following options:
8953 @item chroma_mode, c
8955 Available values for @var{luma_mode}, @var{chroma_mode} and
8956 @var{alpha_mode} are:
8962 @item deinterleave, d
8963 Deinterleave fields, placing one above the other.
8966 Interleave fields. Reverse the effect of deinterleaving.
8968 Default value is @code{none}.
8971 @item chroma_swap, cs
8972 @item alpha_swap, as
8973 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
8978 Apply inflate effect to the video.
8980 This filter replaces the pixel by the local(3x3) average by taking into account
8981 only values higher than the pixel.
8983 It accepts the following options:
8990 Limit the maximum change for each plane, default is 65535.
8991 If 0, plane will remain unchanged.
8996 Simple interlacing filter from progressive contents. This interleaves upper (or
8997 lower) lines from odd frames with lower (or upper) lines from even frames,
8998 halving the frame rate and preserving image height.
9001 Original Original New Frame
9002 Frame 'j' Frame 'j+1' (tff)
9003 ========== =========== ==================
9004 Line 0 --------------------> Frame 'j' Line 0
9005 Line 1 Line 1 ----> Frame 'j+1' Line 1
9006 Line 2 ---------------------> Frame 'j' Line 2
9007 Line 3 Line 3 ----> Frame 'j+1' Line 3
9009 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9012 It accepts the following optional parameters:
9016 This determines whether the interlaced frame is taken from the even
9017 (tff - default) or odd (bff) lines of the progressive frame.
9020 Enable (default) or disable the vertical lowpass filter to avoid twitter
9021 interlacing and reduce moire patterns.
9026 Deinterlace input video by applying Donald Graft's adaptive kernel
9027 deinterling. Work on interlaced parts of a video to produce
9030 The description of the accepted parameters follows.
9034 Set the threshold which affects the filter's tolerance when
9035 determining if a pixel line must be processed. It must be an integer
9036 in the range [0,255] and defaults to 10. A value of 0 will result in
9037 applying the process on every pixels.
9040 Paint pixels exceeding the threshold value to white if set to 1.
9044 Set the fields order. Swap fields if set to 1, leave fields alone if
9048 Enable additional sharpening if set to 1. Default is 0.
9051 Enable twoway sharpening if set to 1. Default is 0.
9054 @subsection Examples
9058 Apply default values:
9060 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9064 Enable additional sharpening:
9070 Paint processed pixels in white:
9076 @section lenscorrection
9078 Correct radial lens distortion
9080 This filter can be used to correct for radial distortion as can result from the use
9081 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9082 one can use tools available for example as part of opencv or simply trial-and-error.
9083 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9084 and extract the k1 and k2 coefficients from the resulting matrix.
9086 Note that effectively the same filter is available in the open-source tools Krita and
9087 Digikam from the KDE project.
9089 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9090 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9091 brightness distribution, so you may want to use both filters together in certain
9092 cases, though you will have to take care of ordering, i.e. whether vignetting should
9093 be applied before or after lens correction.
9097 The filter accepts the following options:
9101 Relative x-coordinate of the focal point of the image, and thereby the center of the
9102 distortion. This value has a range [0,1] and is expressed as fractions of the image
9105 Relative y-coordinate of the focal point of the image, and thereby the center of the
9106 distortion. This value has a range [0,1] and is expressed as fractions of the image
9109 Coefficient of the quadratic correction term. 0.5 means no correction.
9111 Coefficient of the double quadratic correction term. 0.5 means no correction.
9114 The formula that generates the correction is:
9116 @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)
9118 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9119 distances from the focal point in the source and target images, respectively.
9125 The filter accepts the following options:
9129 Set the number of loops.
9132 Set maximal size in number of frames.
9135 Set first frame of loop.
9141 Apply a 3D LUT to an input video.
9143 The filter accepts the following options:
9147 Set the 3D LUT file name.
9149 Currently supported formats:
9161 Select interpolation mode.
9163 Available values are:
9167 Use values from the nearest defined point.
9169 Interpolate values using the 8 points defining a cube.
9171 Interpolate values using a tetrahedron.
9175 @section lut, lutrgb, lutyuv
9177 Compute a look-up table for binding each pixel component input value
9178 to an output value, and apply it to the input video.
9180 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9181 to an RGB input video.
9183 These filters accept the following parameters:
9186 set first pixel component expression
9188 set second pixel component expression
9190 set third pixel component expression
9192 set fourth pixel component expression, corresponds to the alpha component
9195 set red component expression
9197 set green component expression
9199 set blue component expression
9201 alpha component expression
9204 set Y/luminance component expression
9206 set U/Cb component expression
9208 set V/Cr component expression
9211 Each of them specifies the expression to use for computing the lookup table for
9212 the corresponding pixel component values.
9214 The exact component associated to each of the @var{c*} options depends on the
9217 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9218 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9220 The expressions can contain the following constants and functions:
9225 The input width and height.
9228 The input value for the pixel component.
9231 The input value, clipped to the @var{minval}-@var{maxval} range.
9234 The maximum value for the pixel component.
9237 The minimum value for the pixel component.
9240 The negated value for the pixel component value, clipped to the
9241 @var{minval}-@var{maxval} range; it corresponds to the expression
9242 "maxval-clipval+minval".
9245 The computed value in @var{val}, clipped to the
9246 @var{minval}-@var{maxval} range.
9248 @item gammaval(gamma)
9249 The computed gamma correction value of the pixel component value,
9250 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9252 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9256 All expressions default to "val".
9258 @subsection Examples
9264 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9265 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9268 The above is the same as:
9270 lutrgb="r=negval:g=negval:b=negval"
9271 lutyuv="y=negval:u=negval:v=negval"
9281 Remove chroma components, turning the video into a graytone image:
9283 lutyuv="u=128:v=128"
9287 Apply a luma burning effect:
9293 Remove green and blue components:
9299 Set a constant alpha channel value on input:
9301 format=rgba,lutrgb=a="maxval-minval/2"
9305 Correct luminance gamma by a factor of 0.5:
9307 lutyuv=y=gammaval(0.5)
9311 Discard least significant bits of luma:
9313 lutyuv=y='bitand(val, 128+64+32)'
9317 Technicolor like effect:
9319 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9325 Compute and apply a lookup table from two video inputs.
9327 This filter accepts the following parameters:
9330 set first pixel component expression
9332 set second pixel component expression
9334 set third pixel component expression
9336 set fourth pixel component expression, corresponds to the alpha component
9339 Each of them specifies the expression to use for computing the lookup table for
9340 the corresponding pixel component values.
9342 The exact component associated to each of the @var{c*} options depends on the
9345 The expressions can contain the following constants:
9350 The input width and height.
9353 The first input value for the pixel component.
9356 The second input value for the pixel component.
9359 The first input video bit depth.
9362 The second input video bit depth.
9365 All expressions default to "x".
9367 @subsection Examples
9371 Highlight differences between two RGB video streams:
9373 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1)'
9377 Highlight differences between two YUV video streams:
9379 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1)'
9383 @section maskedclamp
9385 Clamp the first input stream with the second input and third input stream.
9387 Returns the value of first stream to be between second input
9388 stream - @code{undershoot} and third input stream + @code{overshoot}.
9390 This filter accepts the following options:
9393 Default value is @code{0}.
9396 Default value is @code{0}.
9399 Set which planes will be processed as bitmap, unprocessed planes will be
9400 copied from first stream.
9401 By default value 0xf, all planes will be processed.
9404 @section maskedmerge
9406 Merge the first input stream with the second input stream using per pixel
9407 weights in the third input stream.
9409 A value of 0 in the third stream pixel component means that pixel component
9410 from first stream is returned unchanged, while maximum value (eg. 255 for
9411 8-bit videos) means that pixel component from second stream is returned
9412 unchanged. Intermediate values define the amount of merging between both
9413 input stream's pixel components.
9415 This filter accepts the following options:
9418 Set which planes will be processed as bitmap, unprocessed planes will be
9419 copied from first stream.
9420 By default value 0xf, all planes will be processed.
9425 Apply motion-compensation deinterlacing.
9427 It needs one field per frame as input and must thus be used together
9428 with yadif=1/3 or equivalent.
9430 This filter accepts the following options:
9433 Set the deinterlacing mode.
9435 It accepts one of the following values:
9440 use iterative motion estimation
9442 like @samp{slow}, but use multiple reference frames.
9444 Default value is @samp{fast}.
9447 Set the picture field parity assumed for the input video. It must be
9448 one of the following values:
9452 assume top field first
9454 assume bottom field first
9457 Default value is @samp{bff}.
9460 Set per-block quantization parameter (QP) used by the internal
9463 Higher values should result in a smoother motion vector field but less
9464 optimal individual vectors. Default value is 1.
9467 @section mergeplanes
9469 Merge color channel components from several video streams.
9471 The filter accepts up to 4 input streams, and merge selected input
9472 planes to the output video.
9474 This filter accepts the following options:
9477 Set input to output plane mapping. Default is @code{0}.
9479 The mappings is specified as a bitmap. It should be specified as a
9480 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9481 mapping for the first plane of the output stream. 'A' sets the number of
9482 the input stream to use (from 0 to 3), and 'a' the plane number of the
9483 corresponding input to use (from 0 to 3). The rest of the mappings is
9484 similar, 'Bb' describes the mapping for the output stream second
9485 plane, 'Cc' describes the mapping for the output stream third plane and
9486 'Dd' describes the mapping for the output stream fourth plane.
9489 Set output pixel format. Default is @code{yuva444p}.
9492 @subsection Examples
9496 Merge three gray video streams of same width and height into single video stream:
9498 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9502 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9504 [a0][a1]mergeplanes=0x00010210:yuva444p
9508 Swap Y and A plane in yuva444p stream:
9510 format=yuva444p,mergeplanes=0x03010200:yuva444p
9514 Swap U and V plane in yuv420p stream:
9516 format=yuv420p,mergeplanes=0x000201:yuv420p
9520 Cast a rgb24 clip to yuv444p:
9522 format=rgb24,mergeplanes=0x000102:yuv444p
9528 Estimate and export motion vectors using block matching algorithms.
9529 Motion vectors are stored in frame side data to be used by other filters.
9531 This filter accepts the following options:
9534 Specify the motion estimation method. Accepts one of the following values:
9538 Exhaustive search algorithm.
9540 Three step search algorithm.
9542 Two dimensional logarithmic search algorithm.
9544 New three step search algorithm.
9546 Four step search algorithm.
9548 Diamond search algorithm.
9550 Hexagon-based search algorithm.
9552 Enhanced predictive zonal search algorithm.
9554 Uneven multi-hexagon search algorithm.
9556 Default value is @samp{esa}.
9559 Macroblock size. Default @code{16}.
9562 Search parameter. Default @code{7}.
9565 @section minterpolate
9567 Convert the video to specified frame rate using motion interpolation.
9569 This filter accepts the following options:
9572 Specify the output frame rate. This can be rational e.g. @code{60000/1001}. Frames are dropped if @var{fps} is lower than source fps. Default @code{60}.
9575 Motion interpolation mode. Following values are accepted:
9578 Duplicate previous or next frame for interpolating new ones.
9580 Blend source frames. Interpolated frame is mean of previous and next frames.
9582 Motion compensated interpolation. Following options are effective when this mode is selected:
9586 Motion compensation mode. Following values are accepted:
9589 Overlapped block motion compensation.
9591 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
9593 Default mode is @samp{obmc}.
9596 Motion estimation mode. Following values are accepted:
9599 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
9601 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
9603 Default mode is @samp{bilat}.
9606 The algorithm to be used for motion estimation. Following values are accepted:
9609 Exhaustive search algorithm.
9611 Three step search algorithm.
9613 Two dimensional logarithmic search algorithm.
9615 New three step search algorithm.
9617 Four step search algorithm.
9619 Diamond search algorithm.
9621 Hexagon-based search algorithm.
9623 Enhanced predictive zonal search algorithm.
9625 Uneven multi-hexagon search algorithm.
9627 Default algorithm is @samp{epzs}.
9630 Macroblock size. Default @code{16}.
9633 Motion estimation search parameter. Default @code{32}.
9636 Enable variable-size block motion compensation. Motion estimation is applied with smaller block sizes at object boundaries in order to make the them less blur. Default is @code{0} (disabled).
9641 Scene change detection method. Scene change leads motion vectors to be in random direction. Scene change detection replace interpolated frames by duplicate ones. May not be needed for other modes. Following values are accepted:
9644 Disable scene change detection.
9646 Frame difference. Corresponding pixel values are compared and if it statisfies @var{scd_threshold} scene change is detected.
9648 Default method is @samp{fdiff}.
9651 Scene change detection threshold. Default is @code{5.0}.
9656 Drop frames that do not differ greatly from the previous frame in
9657 order to reduce frame rate.
9659 The main use of this filter is for very-low-bitrate encoding
9660 (e.g. streaming over dialup modem), but it could in theory be used for
9661 fixing movies that were inverse-telecined incorrectly.
9663 A description of the accepted options follows.
9667 Set the maximum number of consecutive frames which can be dropped (if
9668 positive), or the minimum interval between dropped frames (if
9669 negative). If the value is 0, the frame is dropped unregarding the
9670 number of previous sequentially dropped frames.
9677 Set the dropping threshold values.
9679 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9680 represent actual pixel value differences, so a threshold of 64
9681 corresponds to 1 unit of difference for each pixel, or the same spread
9682 out differently over the block.
9684 A frame is a candidate for dropping if no 8x8 blocks differ by more
9685 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9686 meaning the whole image) differ by more than a threshold of @option{lo}.
9688 Default value for @option{hi} is 64*12, default value for @option{lo} is
9689 64*5, and default value for @option{frac} is 0.33.
9697 It accepts an integer in input; if non-zero it negates the
9698 alpha component (if available). The default value in input is 0.
9702 Denoise frames using Non-Local Means algorithm.
9704 Each pixel is adjusted by looking for other pixels with similar contexts. This
9705 context similarity is defined by comparing their surrounding patches of size
9706 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
9709 Note that the research area defines centers for patches, which means some
9710 patches will be made of pixels outside that research area.
9712 The filter accepts the following options.
9716 Set denoising strength.
9722 Same as @option{p} but for chroma planes.
9724 The default value is @var{0} and means automatic.
9730 Same as @option{r} but for chroma planes.
9732 The default value is @var{0} and means automatic.
9737 Deinterlace video using neural network edge directed interpolation.
9739 This filter accepts the following options:
9743 Mandatory option, without binary file filter can not work.
9744 Currently file can be found here:
9745 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9748 Set which frames to deinterlace, by default it is @code{all}.
9749 Can be @code{all} or @code{interlaced}.
9752 Set mode of operation.
9754 Can be one of the following:
9758 Use frame flags, both fields.
9760 Use frame flags, single field.
9764 Use bottom field only.
9766 Use both fields, top first.
9768 Use both fields, bottom first.
9772 Set which planes to process, by default filter process all frames.
9775 Set size of local neighborhood around each pixel, used by the predictor neural
9778 Can be one of the following:
9791 Set the number of neurons in predicctor neural network.
9792 Can be one of the following:
9803 Controls the number of different neural network predictions that are blended
9804 together to compute the final output value. Can be @code{fast}, default or
9808 Set which set of weights to use in the predictor.
9809 Can be one of the following:
9813 weights trained to minimize absolute error
9815 weights trained to minimize squared error
9819 Controls whether or not the prescreener neural network is used to decide
9820 which pixels should be processed by the predictor neural network and which
9821 can be handled by simple cubic interpolation.
9822 The prescreener is trained to know whether cubic interpolation will be
9823 sufficient for a pixel or whether it should be predicted by the predictor nn.
9824 The computational complexity of the prescreener nn is much less than that of
9825 the predictor nn. Since most pixels can be handled by cubic interpolation,
9826 using the prescreener generally results in much faster processing.
9827 The prescreener is pretty accurate, so the difference between using it and not
9828 using it is almost always unnoticeable.
9830 Can be one of the following:
9838 Default is @code{new}.
9841 Set various debugging flags.
9846 Force libavfilter not to use any of the specified pixel formats for the
9847 input to the next filter.
9849 It accepts the following parameters:
9853 A '|'-separated list of pixel format names, such as
9854 apix_fmts=yuv420p|monow|rgb24".
9858 @subsection Examples
9862 Force libavfilter to use a format different from @var{yuv420p} for the
9863 input to the vflip filter:
9865 noformat=pix_fmts=yuv420p,vflip
9869 Convert the input video to any of the formats not contained in the list:
9871 noformat=yuv420p|yuv444p|yuv410p
9877 Add noise on video input frame.
9879 The filter accepts the following options:
9887 Set noise seed for specific pixel component or all pixel components in case
9888 of @var{all_seed}. Default value is @code{123457}.
9890 @item all_strength, alls
9891 @item c0_strength, c0s
9892 @item c1_strength, c1s
9893 @item c2_strength, c2s
9894 @item c3_strength, c3s
9895 Set noise strength for specific pixel component or all pixel components in case
9896 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9898 @item all_flags, allf
9903 Set pixel component flags or set flags for all components if @var{all_flags}.
9904 Available values for component flags are:
9907 averaged temporal noise (smoother)
9909 mix random noise with a (semi)regular pattern
9911 temporal noise (noise pattern changes between frames)
9913 uniform noise (gaussian otherwise)
9917 @subsection Examples
9919 Add temporal and uniform noise to input video:
9921 noise=alls=20:allf=t+u
9926 Pass the video source unchanged to the output.
9929 Optical Character Recognition
9931 This filter uses Tesseract for optical character recognition.
9933 It accepts the following options:
9937 Set datapath to tesseract data. Default is to use whatever was
9938 set at installation.
9941 Set language, default is "eng".
9944 Set character whitelist.
9947 Set character blacklist.
9950 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9954 Apply a video transform using libopencv.
9956 To enable this filter, install the libopencv library and headers and
9957 configure FFmpeg with @code{--enable-libopencv}.
9959 It accepts the following parameters:
9964 The name of the libopencv filter to apply.
9967 The parameters to pass to the libopencv filter. If not specified, the default
9972 Refer to the official libopencv documentation for more precise
9974 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9976 Several libopencv filters are supported; see the following subsections.
9981 Dilate an image by using a specific structuring element.
9982 It corresponds to the libopencv function @code{cvDilate}.
9984 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
9986 @var{struct_el} represents a structuring element, and has the syntax:
9987 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
9989 @var{cols} and @var{rows} represent the number of columns and rows of
9990 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
9991 point, and @var{shape} the shape for the structuring element. @var{shape}
9992 must be "rect", "cross", "ellipse", or "custom".
9994 If the value for @var{shape} is "custom", it must be followed by a
9995 string of the form "=@var{filename}". The file with name
9996 @var{filename} is assumed to represent a binary image, with each
9997 printable character corresponding to a bright pixel. When a custom
9998 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
9999 or columns and rows of the read file are assumed instead.
10001 The default value for @var{struct_el} is "3x3+0x0/rect".
10003 @var{nb_iterations} specifies the number of times the transform is
10004 applied to the image, and defaults to 1.
10008 # Use the default values
10011 # Dilate using a structuring element with a 5x5 cross, iterating two times
10012 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10014 # Read the shape from the file diamond.shape, iterating two times.
10015 # The file diamond.shape may contain a pattern of characters like this
10021 # The specified columns and rows are ignored
10022 # but the anchor point coordinates are not
10023 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10028 Erode an image by using a specific structuring element.
10029 It corresponds to the libopencv function @code{cvErode}.
10031 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10032 with the same syntax and semantics as the @ref{dilate} filter.
10036 Smooth the input video.
10038 The filter takes the following parameters:
10039 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10041 @var{type} is the type of smooth filter to apply, and must be one of
10042 the following values: "blur", "blur_no_scale", "median", "gaussian",
10043 or "bilateral". The default value is "gaussian".
10045 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10046 depend on the smooth type. @var{param1} and
10047 @var{param2} accept integer positive values or 0. @var{param3} and
10048 @var{param4} accept floating point values.
10050 The default value for @var{param1} is 3. The default value for the
10051 other parameters is 0.
10053 These parameters correspond to the parameters assigned to the
10054 libopencv function @code{cvSmooth}.
10059 Overlay one video on top of another.
10061 It takes two inputs and has one output. The first input is the "main"
10062 video on which the second input is overlaid.
10064 It accepts the following parameters:
10066 A description of the accepted options follows.
10071 Set the expression for the x and y coordinates of the overlaid video
10072 on the main video. Default value is "0" for both expressions. In case
10073 the expression is invalid, it is set to a huge value (meaning that the
10074 overlay will not be displayed within the output visible area).
10077 The action to take when EOF is encountered on the secondary input; it accepts
10078 one of the following values:
10082 Repeat the last frame (the default).
10086 Pass the main input through.
10090 Set when the expressions for @option{x}, and @option{y} are evaluated.
10092 It accepts the following values:
10095 only evaluate expressions once during the filter initialization or
10096 when a command is processed
10099 evaluate expressions for each incoming frame
10102 Default value is @samp{frame}.
10105 If set to 1, force the output to terminate when the shortest input
10106 terminates. Default value is 0.
10109 Set the format for the output video.
10111 It accepts the following values:
10114 force YUV420 output
10117 force YUV422 output
10120 force YUV444 output
10126 Default value is @samp{yuv420}.
10128 @item rgb @emph{(deprecated)}
10129 If set to 1, force the filter to accept inputs in the RGB
10130 color space. Default value is 0. This option is deprecated, use
10131 @option{format} instead.
10134 If set to 1, force the filter to draw the last overlay frame over the
10135 main input until the end of the stream. A value of 0 disables this
10136 behavior. Default value is 1.
10139 The @option{x}, and @option{y} expressions can contain the following
10145 The main input width and height.
10149 The overlay input width and height.
10153 The computed values for @var{x} and @var{y}. They are evaluated for
10158 horizontal and vertical chroma subsample values of the output
10159 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
10163 the number of input frame, starting from 0
10166 the position in the file of the input frame, NAN if unknown
10169 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
10173 Note that the @var{n}, @var{pos}, @var{t} variables are available only
10174 when evaluation is done @emph{per frame}, and will evaluate to NAN
10175 when @option{eval} is set to @samp{init}.
10177 Be aware that frames are taken from each input video in timestamp
10178 order, hence, if their initial timestamps differ, it is a good idea
10179 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
10180 have them begin in the same zero timestamp, as the example for
10181 the @var{movie} filter does.
10183 You can chain together more overlays but you should test the
10184 efficiency of such approach.
10186 @subsection Commands
10188 This filter supports the following commands:
10192 Modify the x and y of the overlay input.
10193 The command accepts the same syntax of the corresponding option.
10195 If the specified expression is not valid, it is kept at its current
10199 @subsection Examples
10203 Draw the overlay at 10 pixels from the bottom right corner of the main
10206 overlay=main_w-overlay_w-10:main_h-overlay_h-10
10209 Using named options the example above becomes:
10211 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
10215 Insert a transparent PNG logo in the bottom left corner of the input,
10216 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
10218 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
10222 Insert 2 different transparent PNG logos (second logo on bottom
10223 right corner) using the @command{ffmpeg} tool:
10225 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
10229 Add a transparent color layer on top of the main video; @code{WxH}
10230 must specify the size of the main input to the overlay filter:
10232 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
10236 Play an original video and a filtered version (here with the deshake
10237 filter) side by side using the @command{ffplay} tool:
10239 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
10242 The above command is the same as:
10244 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
10248 Make a sliding overlay appearing from the left to the right top part of the
10249 screen starting since time 2:
10251 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
10255 Compose output by putting two input videos side to side:
10257 ffmpeg -i left.avi -i right.avi -filter_complex "
10258 nullsrc=size=200x100 [background];
10259 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
10260 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
10261 [background][left] overlay=shortest=1 [background+left];
10262 [background+left][right] overlay=shortest=1:x=100 [left+right]
10267 Mask 10-20 seconds of a video by applying the delogo filter to a section
10269 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
10270 -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]'
10275 Chain several overlays in cascade:
10277 nullsrc=s=200x200 [bg];
10278 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
10279 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
10280 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
10281 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
10282 [in3] null, [mid2] overlay=100:100 [out0]
10289 Apply Overcomplete Wavelet denoiser.
10291 The filter accepts the following options:
10297 Larger depth values will denoise lower frequency components more, but
10298 slow down filtering.
10300 Must be an int in the range 8-16, default is @code{8}.
10302 @item luma_strength, ls
10305 Must be a double value in the range 0-1000, default is @code{1.0}.
10307 @item chroma_strength, cs
10308 Set chroma strength.
10310 Must be a double value in the range 0-1000, default is @code{1.0}.
10316 Add paddings to the input image, and place the original input at the
10317 provided @var{x}, @var{y} coordinates.
10319 It accepts the following parameters:
10324 Specify an expression for the size of the output image with the
10325 paddings added. If the value for @var{width} or @var{height} is 0, the
10326 corresponding input size is used for the output.
10328 The @var{width} expression can reference the value set by the
10329 @var{height} expression, and vice versa.
10331 The default value of @var{width} and @var{height} is 0.
10335 Specify the offsets to place the input image at within the padded area,
10336 with respect to the top/left border of the output image.
10338 The @var{x} expression can reference the value set by the @var{y}
10339 expression, and vice versa.
10341 The default value of @var{x} and @var{y} is 0.
10344 Specify the color of the padded area. For the syntax of this option,
10345 check the "Color" section in the ffmpeg-utils manual.
10347 The default value of @var{color} is "black".
10350 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10351 options are expressions containing the following constants:
10356 The input video width and height.
10360 These are the same as @var{in_w} and @var{in_h}.
10364 The output width and height (the size of the padded area), as
10365 specified by the @var{width} and @var{height} expressions.
10369 These are the same as @var{out_w} and @var{out_h}.
10373 The x and y offsets as specified by the @var{x} and @var{y}
10374 expressions, or NAN if not yet specified.
10377 same as @var{iw} / @var{ih}
10380 input sample aspect ratio
10383 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10387 The horizontal and vertical chroma subsample values. For example for the
10388 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10391 @subsection Examples
10395 Add paddings with the color "violet" to the input video. The output video
10396 size is 640x480, and the top-left corner of the input video is placed at
10399 pad=640:480:0:40:violet
10402 The example above is equivalent to the following command:
10404 pad=width=640:height=480:x=0:y=40:color=violet
10408 Pad the input to get an output with dimensions increased by 3/2,
10409 and put the input video at the center of the padded area:
10411 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10415 Pad the input to get a squared output with size equal to the maximum
10416 value between the input width and height, and put the input video at
10417 the center of the padded area:
10419 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10423 Pad the input to get a final w/h ratio of 16:9:
10425 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10429 In case of anamorphic video, in order to set the output display aspect
10430 correctly, it is necessary to use @var{sar} in the expression,
10431 according to the relation:
10433 (ih * X / ih) * sar = output_dar
10434 X = output_dar / sar
10437 Thus the previous example needs to be modified to:
10439 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10443 Double the output size and put the input video in the bottom-right
10444 corner of the output padded area:
10446 pad="2*iw:2*ih:ow-iw:oh-ih"
10450 @anchor{palettegen}
10451 @section palettegen
10453 Generate one palette for a whole video stream.
10455 It accepts the following options:
10459 Set the maximum number of colors to quantize in the palette.
10460 Note: the palette will still contain 256 colors; the unused palette entries
10463 @item reserve_transparent
10464 Create a palette of 255 colors maximum and reserve the last one for
10465 transparency. Reserving the transparency color is useful for GIF optimization.
10466 If not set, the maximum of colors in the palette will be 256. You probably want
10467 to disable this option for a standalone image.
10471 Set statistics mode.
10473 It accepts the following values:
10476 Compute full frame histograms.
10478 Compute histograms only for the part that differs from previous frame. This
10479 might be relevant to give more importance to the moving part of your input if
10480 the background is static.
10482 Compute new histogram for each frame.
10485 Default value is @var{full}.
10488 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10489 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10490 color quantization of the palette. This information is also visible at
10491 @var{info} logging level.
10493 @subsection Examples
10497 Generate a representative palette of a given video using @command{ffmpeg}:
10499 ffmpeg -i input.mkv -vf palettegen palette.png
10503 @section paletteuse
10505 Use a palette to downsample an input video stream.
10507 The filter takes two inputs: one video stream and a palette. The palette must
10508 be a 256 pixels image.
10510 It accepts the following options:
10514 Select dithering mode. Available algorithms are:
10517 Ordered 8x8 bayer dithering (deterministic)
10519 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10520 Note: this dithering is sometimes considered "wrong" and is included as a
10522 @item floyd_steinberg
10523 Floyd and Steingberg dithering (error diffusion)
10525 Frankie Sierra dithering v2 (error diffusion)
10527 Frankie Sierra dithering v2 "Lite" (error diffusion)
10530 Default is @var{sierra2_4a}.
10533 When @var{bayer} dithering is selected, this option defines the scale of the
10534 pattern (how much the crosshatch pattern is visible). A low value means more
10535 visible pattern for less banding, and higher value means less visible pattern
10536 at the cost of more banding.
10538 The option must be an integer value in the range [0,5]. Default is @var{2}.
10541 If set, define the zone to process
10545 Only the changing rectangle will be reprocessed. This is similar to GIF
10546 cropping/offsetting compression mechanism. This option can be useful for speed
10547 if only a part of the image is changing, and has use cases such as limiting the
10548 scope of the error diffusal @option{dither} to the rectangle that bounds the
10549 moving scene (it leads to more deterministic output if the scene doesn't change
10550 much, and as a result less moving noise and better GIF compression).
10553 Default is @var{none}.
10556 Take new palette for each output frame.
10559 @subsection Examples
10563 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10564 using @command{ffmpeg}:
10566 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10570 @section perspective
10572 Correct perspective of video not recorded perpendicular to the screen.
10574 A description of the accepted parameters follows.
10585 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10586 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10587 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10588 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10589 then the corners of the source will be sent to the specified coordinates.
10591 The expressions can use the following variables:
10596 the width and height of video frame.
10600 Output frame count.
10603 @item interpolation
10604 Set interpolation for perspective correction.
10606 It accepts the following values:
10612 Default value is @samp{linear}.
10615 Set interpretation of coordinate options.
10617 It accepts the following values:
10621 Send point in the source specified by the given coordinates to
10622 the corners of the destination.
10624 @item 1, destination
10626 Send the corners of the source to the point in the destination specified
10627 by the given coordinates.
10629 Default value is @samp{source}.
10633 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10635 It accepts the following values:
10638 only evaluate expressions once during the filter initialization or
10639 when a command is processed
10642 evaluate expressions for each incoming frame
10645 Default value is @samp{init}.
10650 Delay interlaced video by one field time so that the field order changes.
10652 The intended use is to fix PAL movies that have been captured with the
10653 opposite field order to the film-to-video transfer.
10655 A description of the accepted parameters follows.
10661 It accepts the following values:
10664 Capture field order top-first, transfer bottom-first.
10665 Filter will delay the bottom field.
10668 Capture field order bottom-first, transfer top-first.
10669 Filter will delay the top field.
10672 Capture and transfer with the same field order. This mode only exists
10673 for the documentation of the other options to refer to, but if you
10674 actually select it, the filter will faithfully do nothing.
10677 Capture field order determined automatically by field flags, transfer
10679 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10680 basis using field flags. If no field information is available,
10681 then this works just like @samp{u}.
10684 Capture unknown or varying, transfer opposite.
10685 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10686 analyzing the images and selecting the alternative that produces best
10687 match between the fields.
10690 Capture top-first, transfer unknown or varying.
10691 Filter selects among @samp{t} and @samp{p} using image analysis.
10694 Capture bottom-first, transfer unknown or varying.
10695 Filter selects among @samp{b} and @samp{p} using image analysis.
10698 Capture determined by field flags, transfer unknown or varying.
10699 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10700 image analysis. If no field information is available, then this works just
10701 like @samp{U}. This is the default mode.
10704 Both capture and transfer unknown or varying.
10705 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10709 @section pixdesctest
10711 Pixel format descriptor test filter, mainly useful for internal
10712 testing. The output video should be equal to the input video.
10716 format=monow, pixdesctest
10719 can be used to test the monowhite pixel format descriptor definition.
10723 Enable the specified chain of postprocessing subfilters using libpostproc. This
10724 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10725 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10726 Each subfilter and some options have a short and a long name that can be used
10727 interchangeably, i.e. dr/dering are the same.
10729 The filters accept the following options:
10733 Set postprocessing subfilters string.
10736 All subfilters share common options to determine their scope:
10740 Honor the quality commands for this subfilter.
10743 Do chrominance filtering, too (default).
10746 Do luminance filtering only (no chrominance).
10749 Do chrominance filtering only (no luminance).
10752 These options can be appended after the subfilter name, separated by a '|'.
10754 Available subfilters are:
10757 @item hb/hdeblock[|difference[|flatness]]
10758 Horizontal deblocking filter
10761 Difference factor where higher values mean more deblocking (default: @code{32}).
10763 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10766 @item vb/vdeblock[|difference[|flatness]]
10767 Vertical deblocking filter
10770 Difference factor where higher values mean more deblocking (default: @code{32}).
10772 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10775 @item ha/hadeblock[|difference[|flatness]]
10776 Accurate horizontal deblocking filter
10779 Difference factor where higher values mean more deblocking (default: @code{32}).
10781 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10784 @item va/vadeblock[|difference[|flatness]]
10785 Accurate vertical deblocking filter
10788 Difference factor where higher values mean more deblocking (default: @code{32}).
10790 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10794 The horizontal and vertical deblocking filters share the difference and
10795 flatness values so you cannot set different horizontal and vertical
10799 @item h1/x1hdeblock
10800 Experimental horizontal deblocking filter
10802 @item v1/x1vdeblock
10803 Experimental vertical deblocking filter
10808 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10811 larger -> stronger filtering
10813 larger -> stronger filtering
10815 larger -> stronger filtering
10818 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10821 Stretch luminance to @code{0-255}.
10824 @item lb/linblenddeint
10825 Linear blend deinterlacing filter that deinterlaces the given block by
10826 filtering all lines with a @code{(1 2 1)} filter.
10828 @item li/linipoldeint
10829 Linear interpolating deinterlacing filter that deinterlaces the given block by
10830 linearly interpolating every second line.
10832 @item ci/cubicipoldeint
10833 Cubic interpolating deinterlacing filter deinterlaces the given block by
10834 cubically interpolating every second line.
10836 @item md/mediandeint
10837 Median deinterlacing filter that deinterlaces the given block by applying a
10838 median filter to every second line.
10840 @item fd/ffmpegdeint
10841 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10842 second line with a @code{(-1 4 2 4 -1)} filter.
10845 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10846 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10848 @item fq/forceQuant[|quantizer]
10849 Overrides the quantizer table from the input with the constant quantizer you
10857 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10860 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10863 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10866 @subsection Examples
10870 Apply horizontal and vertical deblocking, deringing and automatic
10871 brightness/contrast:
10877 Apply default filters without brightness/contrast correction:
10883 Apply default filters and temporal denoiser:
10885 pp=default/tmpnoise|1|2|3
10889 Apply deblocking on luminance only, and switch vertical deblocking on or off
10890 automatically depending on available CPU time:
10897 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10898 similar to spp = 6 with 7 point DCT, where only the center sample is
10901 The filter accepts the following options:
10905 Force a constant quantization parameter. It accepts an integer in range
10906 0 to 63. If not set, the filter will use the QP from the video stream
10910 Set thresholding mode. Available modes are:
10914 Set hard thresholding.
10916 Set soft thresholding (better de-ringing effect, but likely blurrier).
10918 Set medium thresholding (good results, default).
10923 Apply prewitt operator to input video stream.
10925 The filter accepts the following option:
10929 Set which planes will be processed, unprocessed planes will be copied.
10930 By default value 0xf, all planes will be processed.
10933 Set value which will be multiplied with filtered result.
10936 Set value which will be added to filtered result.
10941 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10942 Ratio) between two input videos.
10944 This filter takes in input two input videos, the first input is
10945 considered the "main" source and is passed unchanged to the
10946 output. The second input is used as a "reference" video for computing
10949 Both video inputs must have the same resolution and pixel format for
10950 this filter to work correctly. Also it assumes that both inputs
10951 have the same number of frames, which are compared one by one.
10953 The obtained average PSNR is printed through the logging system.
10955 The filter stores the accumulated MSE (mean squared error) of each
10956 frame, and at the end of the processing it is averaged across all frames
10957 equally, and the following formula is applied to obtain the PSNR:
10960 PSNR = 10*log10(MAX^2/MSE)
10963 Where MAX is the average of the maximum values of each component of the
10966 The description of the accepted parameters follows.
10969 @item stats_file, f
10970 If specified the filter will use the named file to save the PSNR of
10971 each individual frame. When filename equals "-" the data is sent to
10974 @item stats_version
10975 Specifies which version of the stats file format to use. Details of
10976 each format are written below.
10977 Default value is 1.
10979 @item stats_add_max
10980 Determines whether the max value is output to the stats log.
10981 Default value is 0.
10982 Requires stats_version >= 2. If this is set and stats_version < 2,
10983 the filter will return an error.
10986 The file printed if @var{stats_file} is selected, contains a sequence of
10987 key/value pairs of the form @var{key}:@var{value} for each compared
10990 If a @var{stats_version} greater than 1 is specified, a header line precedes
10991 the list of per-frame-pair stats, with key value pairs following the frame
10992 format with the following parameters:
10995 @item psnr_log_version
10996 The version of the log file format. Will match @var{stats_version}.
10999 A comma separated list of the per-frame-pair parameters included in
11003 A description of each shown per-frame-pair parameter follows:
11007 sequential number of the input frame, starting from 1
11010 Mean Square Error pixel-by-pixel average difference of the compared
11011 frames, averaged over all the image components.
11013 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
11014 Mean Square Error pixel-by-pixel average difference of the compared
11015 frames for the component specified by the suffix.
11017 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
11018 Peak Signal to Noise ratio of the compared frames for the component
11019 specified by the suffix.
11021 @item max_avg, max_y, max_u, max_v
11022 Maximum allowed value for each channel, and average over all
11028 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
11029 [main][ref] psnr="stats_file=stats.log" [out]
11032 On this example the input file being processed is compared with the
11033 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
11034 is stored in @file{stats.log}.
11039 Pulldown reversal (inverse telecine) filter, capable of handling mixed
11040 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
11043 The pullup filter is designed to take advantage of future context in making
11044 its decisions. This filter is stateless in the sense that it does not lock
11045 onto a pattern to follow, but it instead looks forward to the following
11046 fields in order to identify matches and rebuild progressive frames.
11048 To produce content with an even framerate, insert the fps filter after
11049 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
11050 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
11052 The filter accepts the following options:
11059 These options set the amount of "junk" to ignore at the left, right, top, and
11060 bottom of the image, respectively. Left and right are in units of 8 pixels,
11061 while top and bottom are in units of 2 lines.
11062 The default is 8 pixels on each side.
11065 Set the strict breaks. Setting this option to 1 will reduce the chances of
11066 filter generating an occasional mismatched frame, but it may also cause an
11067 excessive number of frames to be dropped during high motion sequences.
11068 Conversely, setting it to -1 will make filter match fields more easily.
11069 This may help processing of video where there is slight blurring between
11070 the fields, but may also cause there to be interlaced frames in the output.
11071 Default value is @code{0}.
11074 Set the metric plane to use. It accepts the following values:
11080 Use chroma blue plane.
11083 Use chroma red plane.
11086 This option may be set to use chroma plane instead of the default luma plane
11087 for doing filter's computations. This may improve accuracy on very clean
11088 source material, but more likely will decrease accuracy, especially if there
11089 is chroma noise (rainbow effect) or any grayscale video.
11090 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
11091 load and make pullup usable in realtime on slow machines.
11094 For best results (without duplicated frames in the output file) it is
11095 necessary to change the output frame rate. For example, to inverse
11096 telecine NTSC input:
11098 ffmpeg -i input -vf pullup -r 24000/1001 ...
11103 Change video quantization parameters (QP).
11105 The filter accepts the following option:
11109 Set expression for quantization parameter.
11112 The expression is evaluated through the eval API and can contain, among others,
11113 the following constants:
11117 1 if index is not 129, 0 otherwise.
11120 Sequentional index starting from -129 to 128.
11123 @subsection Examples
11127 Some equation like:
11135 Flush video frames from internal cache of frames into a random order.
11136 No frame is discarded.
11137 Inspired by @ref{frei0r} nervous filter.
11141 Set size in number of frames of internal cache, in range from @code{2} to
11142 @code{512}. Default is @code{30}.
11145 Set seed for random number generator, must be an integer included between
11146 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
11147 less than @code{0}, the filter will try to use a good random seed on a
11153 Read vertical interval timecode (VITC) information from the top lines of a
11156 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
11157 timecode value, if a valid timecode has been detected. Further metadata key
11158 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
11159 timecode data has been found or not.
11161 This filter accepts the following options:
11165 Set the maximum number of lines to scan for VITC data. If the value is set to
11166 @code{-1} the full video frame is scanned. Default is @code{45}.
11169 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
11170 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
11173 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
11174 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
11177 @subsection Examples
11181 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
11182 draw @code{--:--:--:--} as a placeholder:
11184 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
11190 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
11192 Destination pixel at position (X, Y) will be picked from source (x, y) position
11193 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
11194 value for pixel will be used for destination pixel.
11196 Xmap and Ymap input video streams must be of same dimensions. Output video stream
11197 will have Xmap/Ymap video stream dimensions.
11198 Xmap and Ymap input video streams are 16bit depth, single channel.
11200 @section removegrain
11202 The removegrain filter is a spatial denoiser for progressive video.
11206 Set mode for the first plane.
11209 Set mode for the second plane.
11212 Set mode for the third plane.
11215 Set mode for the fourth plane.
11218 Range of mode is from 0 to 24. Description of each mode follows:
11222 Leave input plane unchanged. Default.
11225 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
11228 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
11231 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
11234 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
11235 This is equivalent to a median filter.
11238 Line-sensitive clipping giving the minimal change.
11241 Line-sensitive clipping, intermediate.
11244 Line-sensitive clipping, intermediate.
11247 Line-sensitive clipping, intermediate.
11250 Line-sensitive clipping on a line where the neighbours pixels are the closest.
11253 Replaces the target pixel with the closest neighbour.
11256 [1 2 1] horizontal and vertical kernel blur.
11262 Bob mode, interpolates top field from the line where the neighbours
11263 pixels are the closest.
11266 Bob mode, interpolates bottom field from the line where the neighbours
11267 pixels are the closest.
11270 Bob mode, interpolates top field. Same as 13 but with a more complicated
11271 interpolation formula.
11274 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
11275 interpolation formula.
11278 Clips the pixel with the minimum and maximum of respectively the maximum and
11279 minimum of each pair of opposite neighbour pixels.
11282 Line-sensitive clipping using opposite neighbours whose greatest distance from
11283 the current pixel is minimal.
11286 Replaces the pixel with the average of its 8 neighbours.
11289 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
11292 Clips pixels using the averages of opposite neighbour.
11295 Same as mode 21 but simpler and faster.
11298 Small edge and halo removal, but reputed useless.
11304 @section removelogo
11306 Suppress a TV station logo, using an image file to determine which
11307 pixels comprise the logo. It works by filling in the pixels that
11308 comprise the logo with neighboring pixels.
11310 The filter accepts the following options:
11314 Set the filter bitmap file, which can be any image format supported by
11315 libavformat. The width and height of the image file must match those of the
11316 video stream being processed.
11319 Pixels in the provided bitmap image with a value of zero are not
11320 considered part of the logo, non-zero pixels are considered part of
11321 the logo. If you use white (255) for the logo and black (0) for the
11322 rest, you will be safe. For making the filter bitmap, it is
11323 recommended to take a screen capture of a black frame with the logo
11324 visible, and then using a threshold filter followed by the erode
11325 filter once or twice.
11327 If needed, little splotches can be fixed manually. Remember that if
11328 logo pixels are not covered, the filter quality will be much
11329 reduced. Marking too many pixels as part of the logo does not hurt as
11330 much, but it will increase the amount of blurring needed to cover over
11331 the image and will destroy more information than necessary, and extra
11332 pixels will slow things down on a large logo.
11334 @section repeatfields
11336 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11337 fields based on its value.
11341 Reverse a video clip.
11343 Warning: This filter requires memory to buffer the entire clip, so trimming
11346 @subsection Examples
11350 Take the first 5 seconds of a clip, and reverse it.
11358 Rotate video by an arbitrary angle expressed in radians.
11360 The filter accepts the following options:
11362 A description of the optional parameters follows.
11365 Set an expression for the angle by which to rotate the input video
11366 clockwise, expressed as a number of radians. A negative value will
11367 result in a counter-clockwise rotation. By default it is set to "0".
11369 This expression is evaluated for each frame.
11372 Set the output width expression, default value is "iw".
11373 This expression is evaluated just once during configuration.
11376 Set the output height expression, default value is "ih".
11377 This expression is evaluated just once during configuration.
11380 Enable bilinear interpolation if set to 1, a value of 0 disables
11381 it. Default value is 1.
11384 Set the color used to fill the output area not covered by the rotated
11385 image. For the general syntax of this option, check the "Color" section in the
11386 ffmpeg-utils manual. If the special value "none" is selected then no
11387 background is printed (useful for example if the background is never shown).
11389 Default value is "black".
11392 The expressions for the angle and the output size can contain the
11393 following constants and functions:
11397 sequential number of the input frame, starting from 0. It is always NAN
11398 before the first frame is filtered.
11401 time in seconds of the input frame, it is set to 0 when the filter is
11402 configured. It is always NAN before the first frame is filtered.
11406 horizontal and vertical chroma subsample values. For example for the
11407 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11411 the input video width and height
11415 the output width and height, that is the size of the padded area as
11416 specified by the @var{width} and @var{height} expressions
11420 the minimal width/height required for completely containing the input
11421 video rotated by @var{a} radians.
11423 These are only available when computing the @option{out_w} and
11424 @option{out_h} expressions.
11427 @subsection Examples
11431 Rotate the input by PI/6 radians clockwise:
11437 Rotate the input by PI/6 radians counter-clockwise:
11443 Rotate the input by 45 degrees clockwise:
11449 Apply a constant rotation with period T, starting from an angle of PI/3:
11451 rotate=PI/3+2*PI*t/T
11455 Make the input video rotation oscillating with a period of T
11456 seconds and an amplitude of A radians:
11458 rotate=A*sin(2*PI/T*t)
11462 Rotate the video, output size is chosen so that the whole rotating
11463 input video is always completely contained in the output:
11465 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11469 Rotate the video, reduce the output size so that no background is ever
11472 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11476 @subsection Commands
11478 The filter supports the following commands:
11482 Set the angle expression.
11483 The command accepts the same syntax of the corresponding option.
11485 If the specified expression is not valid, it is kept at its current
11491 Apply Shape Adaptive Blur.
11493 The filter accepts the following options:
11496 @item luma_radius, lr
11497 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11498 value is 1.0. A greater value will result in a more blurred image, and
11499 in slower processing.
11501 @item luma_pre_filter_radius, lpfr
11502 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11505 @item luma_strength, ls
11506 Set luma maximum difference between pixels to still be considered, must
11507 be a value in the 0.1-100.0 range, default value is 1.0.
11509 @item chroma_radius, cr
11510 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11511 greater value will result in a more blurred image, and in slower
11514 @item chroma_pre_filter_radius, cpfr
11515 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11517 @item chroma_strength, cs
11518 Set chroma maximum difference between pixels to still be considered,
11519 must be a value in the -0.9-100.0 range.
11522 Each chroma option value, if not explicitly specified, is set to the
11523 corresponding luma option value.
11528 Scale (resize) the input video, using the libswscale library.
11530 The scale filter forces the output display aspect ratio to be the same
11531 of the input, by changing the output sample aspect ratio.
11533 If the input image format is different from the format requested by
11534 the next filter, the scale filter will convert the input to the
11537 @subsection Options
11538 The filter accepts the following options, or any of the options
11539 supported by the libswscale scaler.
11541 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11542 the complete list of scaler options.
11547 Set the output video dimension expression. Default value is the input
11550 If the value is 0, the input width is used for the output.
11552 If one of the values is -1, the scale filter will use a value that
11553 maintains the aspect ratio of the input image, calculated from the
11554 other specified dimension. If both of them are -1, the input size is
11557 If one of the values is -n with n > 1, the scale filter will also use a value
11558 that maintains the aspect ratio of the input image, calculated from the other
11559 specified dimension. After that it will, however, make sure that the calculated
11560 dimension is divisible by n and adjust the value if necessary.
11562 See below for the list of accepted constants for use in the dimension
11566 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11570 Only evaluate expressions once during the filter initialization or when a command is processed.
11573 Evaluate expressions for each incoming frame.
11577 Default value is @samp{init}.
11581 Set the interlacing mode. It accepts the following values:
11585 Force interlaced aware scaling.
11588 Do not apply interlaced scaling.
11591 Select interlaced aware scaling depending on whether the source frames
11592 are flagged as interlaced or not.
11595 Default value is @samp{0}.
11598 Set libswscale scaling flags. See
11599 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11600 complete list of values. If not explicitly specified the filter applies
11604 @item param0, param1
11605 Set libswscale input parameters for scaling algorithms that need them. See
11606 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11607 complete documentation. If not explicitly specified the filter applies
11613 Set the video size. For the syntax of this option, check the
11614 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11616 @item in_color_matrix
11617 @item out_color_matrix
11618 Set in/output YCbCr color space type.
11620 This allows the autodetected value to be overridden as well as allows forcing
11621 a specific value used for the output and encoder.
11623 If not specified, the color space type depends on the pixel format.
11629 Choose automatically.
11632 Format conforming to International Telecommunication Union (ITU)
11633 Recommendation BT.709.
11636 Set color space conforming to the United States Federal Communications
11637 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11640 Set color space conforming to:
11644 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11647 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11650 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11655 Set color space conforming to SMPTE ST 240:1999.
11660 Set in/output YCbCr sample range.
11662 This allows the autodetected value to be overridden as well as allows forcing
11663 a specific value used for the output and encoder. If not specified, the
11664 range depends on the pixel format. Possible values:
11668 Choose automatically.
11671 Set full range (0-255 in case of 8-bit luma).
11674 Set "MPEG" range (16-235 in case of 8-bit luma).
11677 @item force_original_aspect_ratio
11678 Enable decreasing or increasing output video width or height if necessary to
11679 keep the original aspect ratio. Possible values:
11683 Scale the video as specified and disable this feature.
11686 The output video dimensions will automatically be decreased if needed.
11689 The output video dimensions will automatically be increased if needed.
11693 One useful instance of this option is that when you know a specific device's
11694 maximum allowed resolution, you can use this to limit the output video to
11695 that, while retaining the aspect ratio. For example, device A allows
11696 1280x720 playback, and your video is 1920x800. Using this option (set it to
11697 decrease) and specifying 1280x720 to the command line makes the output
11700 Please note that this is a different thing than specifying -1 for @option{w}
11701 or @option{h}, you still need to specify the output resolution for this option
11706 The values of the @option{w} and @option{h} options are expressions
11707 containing the following constants:
11712 The input width and height
11716 These are the same as @var{in_w} and @var{in_h}.
11720 The output (scaled) width and height
11724 These are the same as @var{out_w} and @var{out_h}
11727 The same as @var{iw} / @var{ih}
11730 input sample aspect ratio
11733 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11737 horizontal and vertical input chroma subsample values. For example for the
11738 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11742 horizontal and vertical output chroma subsample values. For example for the
11743 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11746 @subsection Examples
11750 Scale the input video to a size of 200x100
11755 This is equivalent to:
11766 Specify a size abbreviation for the output size:
11771 which can also be written as:
11777 Scale the input to 2x:
11779 scale=w=2*iw:h=2*ih
11783 The above is the same as:
11785 scale=2*in_w:2*in_h
11789 Scale the input to 2x with forced interlaced scaling:
11791 scale=2*iw:2*ih:interl=1
11795 Scale the input to half size:
11797 scale=w=iw/2:h=ih/2
11801 Increase the width, and set the height to the same size:
11807 Seek Greek harmony:
11814 Increase the height, and set the width to 3/2 of the height:
11816 scale=w=3/2*oh:h=3/5*ih
11820 Increase the size, making the size a multiple of the chroma
11823 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11827 Increase the width to a maximum of 500 pixels,
11828 keeping the same aspect ratio as the input:
11830 scale=w='min(500\, iw*3/2):h=-1'
11834 @subsection Commands
11836 This filter supports the following commands:
11840 Set the output video dimension expression.
11841 The command accepts the same syntax of the corresponding option.
11843 If the specified expression is not valid, it is kept at its current
11849 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11850 format conversion on CUDA video frames. Setting the output width and height
11851 works in the same way as for the @var{scale} filter.
11853 The following additional options are accepted:
11856 The pixel format of the output CUDA frames. If set to the string "same" (the
11857 default), the input format will be kept. Note that automatic format negotiation
11858 and conversion is not yet supported for hardware frames
11861 The interpolation algorithm used for resizing. One of the following:
11868 @item cubic2p_bspline
11869 2-parameter cubic (B=1, C=0)
11871 @item cubic2p_catmullrom
11872 2-parameter cubic (B=0, C=1/2)
11874 @item cubic2p_b05c03
11875 2-parameter cubic (B=1/2, C=3/10)
11887 Scale (resize) the input video, based on a reference video.
11889 See the scale filter for available options, scale2ref supports the same but
11890 uses the reference video instead of the main input as basis.
11892 @subsection Examples
11896 Scale a subtitle stream to match the main video in size before overlaying
11898 'scale2ref[b][a];[a][b]overlay'
11902 @anchor{selectivecolor}
11903 @section selectivecolor
11905 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11906 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11907 by the "purity" of the color (that is, how saturated it already is).
11909 This filter is similar to the Adobe Photoshop Selective Color tool.
11911 The filter accepts the following options:
11914 @item correction_method
11915 Select color correction method.
11917 Available values are:
11920 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11923 Specified adjustments are relative to the original component value.
11925 Default is @code{absolute}.
11927 Adjustments for red pixels (pixels where the red component is the maximum)
11929 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11931 Adjustments for green pixels (pixels where the green component is the maximum)
11933 Adjustments for cyan pixels (pixels where the red component is the minimum)
11935 Adjustments for blue pixels (pixels where the blue component is the maximum)
11937 Adjustments for magenta pixels (pixels where the green component is the minimum)
11939 Adjustments for white pixels (pixels where all components are greater than 128)
11941 Adjustments for all pixels except pure black and pure white
11943 Adjustments for black pixels (pixels where all components are lesser than 128)
11945 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11948 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11949 4 space separated floating point adjustment values in the [-1,1] range,
11950 respectively to adjust the amount of cyan, magenta, yellow and black for the
11951 pixels of its range.
11953 @subsection Examples
11957 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11958 increase magenta by 27% in blue areas:
11960 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11964 Use a Photoshop selective color preset:
11966 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11970 @anchor{separatefields}
11971 @section separatefields
11973 The @code{separatefields} takes a frame-based video input and splits
11974 each frame into its components fields, producing a new half height clip
11975 with twice the frame rate and twice the frame count.
11977 This filter use field-dominance information in frame to decide which
11978 of each pair of fields to place first in the output.
11979 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
11981 @section setdar, setsar
11983 The @code{setdar} filter sets the Display Aspect Ratio for the filter
11986 This is done by changing the specified Sample (aka Pixel) Aspect
11987 Ratio, according to the following equation:
11989 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
11992 Keep in mind that the @code{setdar} filter does not modify the pixel
11993 dimensions of the video frame. Also, the display aspect ratio set by
11994 this filter may be changed by later filters in the filterchain,
11995 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
11998 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
11999 the filter output video.
12001 Note that as a consequence of the application of this filter, the
12002 output display aspect ratio will change according to the equation
12005 Keep in mind that the sample aspect ratio set by the @code{setsar}
12006 filter may be changed by later filters in the filterchain, e.g. if
12007 another "setsar" or a "setdar" filter is applied.
12009 It accepts the following parameters:
12012 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
12013 Set the aspect ratio used by the filter.
12015 The parameter can be a floating point number string, an expression, or
12016 a string of the form @var{num}:@var{den}, where @var{num} and
12017 @var{den} are the numerator and denominator of the aspect ratio. If
12018 the parameter is not specified, it is assumed the value "0".
12019 In case the form "@var{num}:@var{den}" is used, the @code{:} character
12023 Set the maximum integer value to use for expressing numerator and
12024 denominator when reducing the expressed aspect ratio to a rational.
12025 Default value is @code{100}.
12029 The parameter @var{sar} is an expression containing
12030 the following constants:
12034 These are approximated values for the mathematical constants e
12035 (Euler's number), pi (Greek pi), and phi (the golden ratio).
12038 The input width and height.
12041 These are the same as @var{w} / @var{h}.
12044 The input sample aspect ratio.
12047 The input display aspect ratio. It is the same as
12048 (@var{w} / @var{h}) * @var{sar}.
12051 Horizontal and vertical chroma subsample values. For example, for the
12052 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12055 @subsection Examples
12060 To change the display aspect ratio to 16:9, specify one of the following:
12067 To change the sample aspect ratio to 10:11, specify:
12073 To set a display aspect ratio of 16:9, and specify a maximum integer value of
12074 1000 in the aspect ratio reduction, use the command:
12076 setdar=ratio=16/9:max=1000
12084 Force field for the output video frame.
12086 The @code{setfield} filter marks the interlace type field for the
12087 output frames. It does not change the input frame, but only sets the
12088 corresponding property, which affects how the frame is treated by
12089 following filters (e.g. @code{fieldorder} or @code{yadif}).
12091 The filter accepts the following options:
12096 Available values are:
12100 Keep the same field property.
12103 Mark the frame as bottom-field-first.
12106 Mark the frame as top-field-first.
12109 Mark the frame as progressive.
12115 Show a line containing various information for each input video frame.
12116 The input video is not modified.
12118 The shown line contains a sequence of key/value pairs of the form
12119 @var{key}:@var{value}.
12121 The following values are shown in the output:
12125 The (sequential) number of the input frame, starting from 0.
12128 The Presentation TimeStamp of the input frame, expressed as a number of
12129 time base units. The time base unit depends on the filter input pad.
12132 The Presentation TimeStamp of the input frame, expressed as a number of
12136 The position of the frame in the input stream, or -1 if this information is
12137 unavailable and/or meaningless (for example in case of synthetic video).
12140 The pixel format name.
12143 The sample aspect ratio of the input frame, expressed in the form
12144 @var{num}/@var{den}.
12147 The size of the input frame. For the syntax of this option, check the
12148 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12151 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
12152 for bottom field first).
12155 This is 1 if the frame is a key frame, 0 otherwise.
12158 The picture type of the input frame ("I" for an I-frame, "P" for a
12159 P-frame, "B" for a B-frame, or "?" for an unknown type).
12160 Also refer to the documentation of the @code{AVPictureType} enum and of
12161 the @code{av_get_picture_type_char} function defined in
12162 @file{libavutil/avutil.h}.
12165 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
12167 @item plane_checksum
12168 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
12169 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
12172 @section showpalette
12174 Displays the 256 colors palette of each frame. This filter is only relevant for
12175 @var{pal8} pixel format frames.
12177 It accepts the following option:
12181 Set the size of the box used to represent one palette color entry. Default is
12182 @code{30} (for a @code{30x30} pixel box).
12185 @section shuffleframes
12187 Reorder and/or duplicate video frames.
12189 It accepts the following parameters:
12193 Set the destination indexes of input frames.
12194 This is space or '|' separated list of indexes that maps input frames to output
12195 frames. Number of indexes also sets maximal value that each index may have.
12198 The first frame has the index 0. The default is to keep the input unchanged.
12200 @subsection Examples
12204 Swap second and third frame of every three frames of the input:
12206 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
12210 Swap 10th and 1st frame of every ten frames of the input:
12212 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
12216 @section shuffleplanes
12218 Reorder and/or duplicate video planes.
12220 It accepts the following parameters:
12225 The index of the input plane to be used as the first output plane.
12228 The index of the input plane to be used as the second output plane.
12231 The index of the input plane to be used as the third output plane.
12234 The index of the input plane to be used as the fourth output plane.
12238 The first plane has the index 0. The default is to keep the input unchanged.
12240 @subsection Examples
12244 Swap the second and third planes of the input:
12246 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
12250 @anchor{signalstats}
12251 @section signalstats
12252 Evaluate various visual metrics that assist in determining issues associated
12253 with the digitization of analog video media.
12255 By default the filter will log these metadata values:
12259 Display the minimal Y value contained within the input frame. Expressed in
12263 Display the Y value at the 10% percentile within the input frame. Expressed in
12267 Display the average Y value within the input frame. Expressed in range of
12271 Display the Y value at the 90% percentile within the input frame. Expressed in
12275 Display the maximum Y value contained within the input frame. Expressed in
12279 Display the minimal U value contained within the input frame. Expressed in
12283 Display the U value at the 10% percentile within the input frame. Expressed in
12287 Display the average U value within the input frame. Expressed in range of
12291 Display the U value at the 90% percentile within the input frame. Expressed in
12295 Display the maximum U value contained within the input frame. Expressed in
12299 Display the minimal V value contained within the input frame. Expressed in
12303 Display the V value at the 10% percentile within the input frame. Expressed in
12307 Display the average V value within the input frame. Expressed in range of
12311 Display the V value at the 90% percentile within the input frame. Expressed in
12315 Display the maximum V value contained within the input frame. Expressed in
12319 Display the minimal saturation value contained within the input frame.
12320 Expressed in range of [0-~181.02].
12323 Display the saturation value at the 10% percentile within the input frame.
12324 Expressed in range of [0-~181.02].
12327 Display the average saturation value within the input frame. Expressed in range
12331 Display the saturation value at the 90% percentile within the input frame.
12332 Expressed in range of [0-~181.02].
12335 Display the maximum saturation value contained within the input frame.
12336 Expressed in range of [0-~181.02].
12339 Display the median value for hue within the input frame. Expressed in range of
12343 Display the average value for hue within the input frame. Expressed in range of
12347 Display the average of sample value difference between all values of the Y
12348 plane in the current frame and corresponding values of the previous input frame.
12349 Expressed in range of [0-255].
12352 Display the average of sample value difference between all values of the U
12353 plane in the current frame and corresponding values of the previous input frame.
12354 Expressed in range of [0-255].
12357 Display the average of sample value difference between all values of the V
12358 plane in the current frame and corresponding values of the previous input frame.
12359 Expressed in range of [0-255].
12362 Display bit depth of Y plane in current frame.
12363 Expressed in range of [0-16].
12366 Display bit depth of U plane in current frame.
12367 Expressed in range of [0-16].
12370 Display bit depth of V plane in current frame.
12371 Expressed in range of [0-16].
12374 The filter accepts the following options:
12380 @option{stat} specify an additional form of image analysis.
12381 @option{out} output video with the specified type of pixel highlighted.
12383 Both options accept the following values:
12387 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12388 unlike the neighboring pixels of the same field. Examples of temporal outliers
12389 include the results of video dropouts, head clogs, or tape tracking issues.
12392 Identify @var{vertical line repetition}. Vertical line repetition includes
12393 similar rows of pixels within a frame. In born-digital video vertical line
12394 repetition is common, but this pattern is uncommon in video digitized from an
12395 analog source. When it occurs in video that results from the digitization of an
12396 analog source it can indicate concealment from a dropout compensator.
12399 Identify pixels that fall outside of legal broadcast range.
12403 Set the highlight color for the @option{out} option. The default color is
12407 @subsection Examples
12411 Output data of various video metrics:
12413 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12417 Output specific data about the minimum and maximum values of the Y plane per frame:
12419 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12423 Playback video while highlighting pixels that are outside of broadcast range in red.
12425 ffplay example.mov -vf signalstats="out=brng:color=red"
12429 Playback video with signalstats metadata drawn over the frame.
12431 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12434 The contents of signalstat_drawtext.txt used in the command are:
12437 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12438 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12439 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12440 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12448 Blur the input video without impacting the outlines.
12450 It accepts the following options:
12453 @item luma_radius, lr
12454 Set the luma radius. The option value must be a float number in
12455 the range [0.1,5.0] that specifies the variance of the gaussian filter
12456 used to blur the image (slower if larger). Default value is 1.0.
12458 @item luma_strength, ls
12459 Set the luma strength. The option value must be a float number
12460 in the range [-1.0,1.0] that configures the blurring. A value included
12461 in [0.0,1.0] will blur the image whereas a value included in
12462 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12464 @item luma_threshold, lt
12465 Set the luma threshold used as a coefficient to determine
12466 whether a pixel should be blurred or not. The option value must be an
12467 integer in the range [-30,30]. A value of 0 will filter all the image,
12468 a value included in [0,30] will filter flat areas and a value included
12469 in [-30,0] will filter edges. Default value is 0.
12471 @item chroma_radius, cr
12472 Set the chroma radius. The option value must be a float number in
12473 the range [0.1,5.0] that specifies the variance of the gaussian filter
12474 used to blur the image (slower if larger). Default value is 1.0.
12476 @item chroma_strength, cs
12477 Set the chroma strength. The option value must be a float number
12478 in the range [-1.0,1.0] that configures the blurring. A value included
12479 in [0.0,1.0] will blur the image whereas a value included in
12480 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12482 @item chroma_threshold, ct
12483 Set the chroma threshold used as a coefficient to determine
12484 whether a pixel should be blurred or not. The option value must be an
12485 integer in the range [-30,30]. A value of 0 will filter all the image,
12486 a value included in [0,30] will filter flat areas and a value included
12487 in [-30,0] will filter edges. Default value is 0.
12490 If a chroma option is not explicitly set, the corresponding luma value
12495 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12497 This filter takes in input two input videos, the first input is
12498 considered the "main" source and is passed unchanged to the
12499 output. The second input is used as a "reference" video for computing
12502 Both video inputs must have the same resolution and pixel format for
12503 this filter to work correctly. Also it assumes that both inputs
12504 have the same number of frames, which are compared one by one.
12506 The filter stores the calculated SSIM of each frame.
12508 The description of the accepted parameters follows.
12511 @item stats_file, f
12512 If specified the filter will use the named file to save the SSIM of
12513 each individual frame. When filename equals "-" the data is sent to
12517 The file printed if @var{stats_file} is selected, contains a sequence of
12518 key/value pairs of the form @var{key}:@var{value} for each compared
12521 A description of each shown parameter follows:
12525 sequential number of the input frame, starting from 1
12527 @item Y, U, V, R, G, B
12528 SSIM of the compared frames for the component specified by the suffix.
12531 SSIM of the compared frames for the whole frame.
12534 Same as above but in dB representation.
12539 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12540 [main][ref] ssim="stats_file=stats.log" [out]
12543 On this example the input file being processed is compared with the
12544 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12545 is stored in @file{stats.log}.
12547 Another example with both psnr and ssim at same time:
12549 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12554 Convert between different stereoscopic image formats.
12556 The filters accept the following options:
12560 Set stereoscopic image format of input.
12562 Available values for input image formats are:
12565 side by side parallel (left eye left, right eye right)
12568 side by side crosseye (right eye left, left eye right)
12571 side by side parallel with half width resolution
12572 (left eye left, right eye right)
12575 side by side crosseye with half width resolution
12576 (right eye left, left eye right)
12579 above-below (left eye above, right eye below)
12582 above-below (right eye above, left eye below)
12585 above-below with half height resolution
12586 (left eye above, right eye below)
12589 above-below with half height resolution
12590 (right eye above, left eye below)
12593 alternating frames (left eye first, right eye second)
12596 alternating frames (right eye first, left eye second)
12599 interleaved rows (left eye has top row, right eye starts on next row)
12602 interleaved rows (right eye has top row, left eye starts on next row)
12605 interleaved columns, left eye first
12608 interleaved columns, right eye first
12610 Default value is @samp{sbsl}.
12614 Set stereoscopic image format of output.
12618 side by side parallel (left eye left, right eye right)
12621 side by side crosseye (right eye left, left eye right)
12624 side by side parallel with half width resolution
12625 (left eye left, right eye right)
12628 side by side crosseye with half width resolution
12629 (right eye left, left eye right)
12632 above-below (left eye above, right eye below)
12635 above-below (right eye above, left eye below)
12638 above-below with half height resolution
12639 (left eye above, right eye below)
12642 above-below with half height resolution
12643 (right eye above, left eye below)
12646 alternating frames (left eye first, right eye second)
12649 alternating frames (right eye first, left eye second)
12652 interleaved rows (left eye has top row, right eye starts on next row)
12655 interleaved rows (right eye has top row, left eye starts on next row)
12658 anaglyph red/blue gray
12659 (red filter on left eye, blue filter on right eye)
12662 anaglyph red/green gray
12663 (red filter on left eye, green filter on right eye)
12666 anaglyph red/cyan gray
12667 (red filter on left eye, cyan filter on right eye)
12670 anaglyph red/cyan half colored
12671 (red filter on left eye, cyan filter on right eye)
12674 anaglyph red/cyan color
12675 (red filter on left eye, cyan filter on right eye)
12678 anaglyph red/cyan color optimized with the least squares projection of dubois
12679 (red filter on left eye, cyan filter on right eye)
12682 anaglyph green/magenta gray
12683 (green filter on left eye, magenta filter on right eye)
12686 anaglyph green/magenta half colored
12687 (green filter on left eye, magenta filter on right eye)
12690 anaglyph green/magenta colored
12691 (green filter on left eye, magenta filter on right eye)
12694 anaglyph green/magenta color optimized with the least squares projection of dubois
12695 (green filter on left eye, magenta filter on right eye)
12698 anaglyph yellow/blue gray
12699 (yellow filter on left eye, blue filter on right eye)
12702 anaglyph yellow/blue half colored
12703 (yellow filter on left eye, blue filter on right eye)
12706 anaglyph yellow/blue colored
12707 (yellow filter on left eye, blue filter on right eye)
12710 anaglyph yellow/blue color optimized with the least squares projection of dubois
12711 (yellow filter on left eye, blue filter on right eye)
12714 mono output (left eye only)
12717 mono output (right eye only)
12720 checkerboard, left eye first
12723 checkerboard, right eye first
12726 interleaved columns, left eye first
12729 interleaved columns, right eye first
12735 Default value is @samp{arcd}.
12738 @subsection Examples
12742 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12748 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12754 @section streamselect, astreamselect
12755 Select video or audio streams.
12757 The filter accepts the following options:
12761 Set number of inputs. Default is 2.
12764 Set input indexes to remap to outputs.
12767 @subsection Commands
12769 The @code{streamselect} and @code{astreamselect} filter supports the following
12774 Set input indexes to remap to outputs.
12777 @subsection Examples
12781 Select first 5 seconds 1st stream and rest of time 2nd stream:
12783 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12787 Same as above, but for audio:
12789 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12794 Apply sobel operator to input video stream.
12796 The filter accepts the following option:
12800 Set which planes will be processed, unprocessed planes will be copied.
12801 By default value 0xf, all planes will be processed.
12804 Set value which will be multiplied with filtered result.
12807 Set value which will be added to filtered result.
12813 Apply a simple postprocessing filter that compresses and decompresses the image
12814 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12815 and average the results.
12817 The filter accepts the following options:
12821 Set quality. This option defines the number of levels for averaging. It accepts
12822 an integer in the range 0-6. If set to @code{0}, the filter will have no
12823 effect. A value of @code{6} means the higher quality. For each increment of
12824 that value the speed drops by a factor of approximately 2. Default value is
12828 Force a constant quantization parameter. If not set, the filter will use the QP
12829 from the video stream (if available).
12832 Set thresholding mode. Available modes are:
12836 Set hard thresholding (default).
12838 Set soft thresholding (better de-ringing effect, but likely blurrier).
12841 @item use_bframe_qp
12842 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12843 option may cause flicker since the B-Frames have often larger QP. Default is
12844 @code{0} (not enabled).
12850 Draw subtitles on top of input video using the libass library.
12852 To enable compilation of this filter you need to configure FFmpeg with
12853 @code{--enable-libass}. This filter also requires a build with libavcodec and
12854 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12855 Alpha) subtitles format.
12857 The filter accepts the following options:
12861 Set the filename of the subtitle file to read. It must be specified.
12863 @item original_size
12864 Specify the size of the original video, the video for which the ASS file
12865 was composed. For the syntax of this option, check the
12866 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12867 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12868 correctly scale the fonts if the aspect ratio has been changed.
12871 Set a directory path containing fonts that can be used by the filter.
12872 These fonts will be used in addition to whatever the font provider uses.
12875 Set subtitles input character encoding. @code{subtitles} filter only. Only
12876 useful if not UTF-8.
12878 @item stream_index, si
12879 Set subtitles stream index. @code{subtitles} filter only.
12882 Override default style or script info parameters of the subtitles. It accepts a
12883 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12886 If the first key is not specified, it is assumed that the first value
12887 specifies the @option{filename}.
12889 For example, to render the file @file{sub.srt} on top of the input
12890 video, use the command:
12895 which is equivalent to:
12897 subtitles=filename=sub.srt
12900 To render the default subtitles stream from file @file{video.mkv}, use:
12902 subtitles=video.mkv
12905 To render the second subtitles stream from that file, use:
12907 subtitles=video.mkv:si=1
12910 To make the subtitles stream from @file{sub.srt} appear in transparent green
12911 @code{DejaVu Serif}, use:
12913 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12916 @section super2xsai
12918 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12919 Interpolate) pixel art scaling algorithm.
12921 Useful for enlarging pixel art images without reducing sharpness.
12925 Swap two rectangular objects in video.
12927 This filter accepts the following options:
12937 Set 1st rect x coordinate.
12940 Set 1st rect y coordinate.
12943 Set 2nd rect x coordinate.
12946 Set 2nd rect y coordinate.
12948 All expressions are evaluated once for each frame.
12951 The all options are expressions containing the following constants:
12956 The input width and height.
12959 same as @var{w} / @var{h}
12962 input sample aspect ratio
12965 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12968 The number of the input frame, starting from 0.
12971 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12974 the position in the file of the input frame, NAN if unknown
12982 Apply telecine process to the video.
12984 This filter accepts the following options:
12993 The default value is @code{top}.
12997 A string of numbers representing the pulldown pattern you wish to apply.
12998 The default value is @code{23}.
13002 Some typical patterns:
13007 24p: 2332 (preferred)
13014 24p: 222222222223 ("Euro pulldown")
13020 Select the most representative frame in a given sequence of consecutive frames.
13022 The filter accepts the following options:
13026 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
13027 will pick one of them, and then handle the next batch of @var{n} frames until
13028 the end. Default is @code{100}.
13031 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
13032 value will result in a higher memory usage, so a high value is not recommended.
13034 @subsection Examples
13038 Extract one picture each 50 frames:
13044 Complete example of a thumbnail creation with @command{ffmpeg}:
13046 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
13052 Tile several successive frames together.
13054 The filter accepts the following options:
13059 Set the grid size (i.e. the number of lines and columns). For the syntax of
13060 this option, check the
13061 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13064 Set the maximum number of frames to render in the given area. It must be less
13065 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
13066 the area will be used.
13069 Set the outer border margin in pixels.
13072 Set the inner border thickness (i.e. the number of pixels between frames). For
13073 more advanced padding options (such as having different values for the edges),
13074 refer to the pad video filter.
13077 Specify the color of the unused area. For the syntax of this option, check the
13078 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
13082 @subsection Examples
13086 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
13088 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
13090 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
13091 duplicating each output frame to accommodate the originally detected frame
13095 Display @code{5} pictures in an area of @code{3x2} frames,
13096 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
13097 mixed flat and named options:
13099 tile=3x2:nb_frames=5:padding=7:margin=2
13103 @section tinterlace
13105 Perform various types of temporal field interlacing.
13107 Frames are counted starting from 1, so the first input frame is
13110 The filter accepts the following options:
13115 Specify the mode of the interlacing. This option can also be specified
13116 as a value alone. See below for a list of values for this option.
13118 Available values are:
13122 Move odd frames into the upper field, even into the lower field,
13123 generating a double height frame at half frame rate.
13127 Frame 1 Frame 2 Frame 3 Frame 4
13129 11111 22222 33333 44444
13130 11111 22222 33333 44444
13131 11111 22222 33333 44444
13132 11111 22222 33333 44444
13146 Only output odd frames, even frames are dropped, generating a frame with
13147 unchanged height at half frame rate.
13152 Frame 1 Frame 2 Frame 3 Frame 4
13154 11111 22222 33333 44444
13155 11111 22222 33333 44444
13156 11111 22222 33333 44444
13157 11111 22222 33333 44444
13167 Only output even frames, odd frames are dropped, generating a frame with
13168 unchanged height at half frame rate.
13173 Frame 1 Frame 2 Frame 3 Frame 4
13175 11111 22222 33333 44444
13176 11111 22222 33333 44444
13177 11111 22222 33333 44444
13178 11111 22222 33333 44444
13188 Expand each frame to full height, but pad alternate lines with black,
13189 generating a frame with double height at the same input frame rate.
13194 Frame 1 Frame 2 Frame 3 Frame 4
13196 11111 22222 33333 44444
13197 11111 22222 33333 44444
13198 11111 22222 33333 44444
13199 11111 22222 33333 44444
13202 11111 ..... 33333 .....
13203 ..... 22222 ..... 44444
13204 11111 ..... 33333 .....
13205 ..... 22222 ..... 44444
13206 11111 ..... 33333 .....
13207 ..... 22222 ..... 44444
13208 11111 ..... 33333 .....
13209 ..... 22222 ..... 44444
13213 @item interleave_top, 4
13214 Interleave the upper field from odd frames with the lower field from
13215 even frames, generating a frame with unchanged height at half frame rate.
13220 Frame 1 Frame 2 Frame 3 Frame 4
13222 11111<- 22222 33333<- 44444
13223 11111 22222<- 33333 44444<-
13224 11111<- 22222 33333<- 44444
13225 11111 22222<- 33333 44444<-
13235 @item interleave_bottom, 5
13236 Interleave the lower field from odd frames with the upper field from
13237 even frames, generating a frame with unchanged height at half frame rate.
13242 Frame 1 Frame 2 Frame 3 Frame 4
13244 11111 22222<- 33333 44444<-
13245 11111<- 22222 33333<- 44444
13246 11111 22222<- 33333 44444<-
13247 11111<- 22222 33333<- 44444
13257 @item interlacex2, 6
13258 Double frame rate with unchanged height. Frames are inserted each
13259 containing the second temporal field from the previous input frame and
13260 the first temporal field from the next input frame. This mode relies on
13261 the top_field_first flag. Useful for interlaced video displays with no
13262 field synchronisation.
13267 Frame 1 Frame 2 Frame 3 Frame 4
13269 11111 22222 33333 44444
13270 11111 22222 33333 44444
13271 11111 22222 33333 44444
13272 11111 22222 33333 44444
13275 11111 22222 22222 33333 33333 44444 44444
13276 11111 11111 22222 22222 33333 33333 44444
13277 11111 22222 22222 33333 33333 44444 44444
13278 11111 11111 22222 22222 33333 33333 44444
13283 Move odd frames into the upper field, even into the lower field,
13284 generating a double height frame at same frame rate.
13289 Frame 1 Frame 2 Frame 3 Frame 4
13291 11111 22222 33333 44444
13292 11111 22222 33333 44444
13293 11111 22222 33333 44444
13294 11111 22222 33333 44444
13297 11111 33333 33333 55555
13298 22222 22222 44444 44444
13299 11111 33333 33333 55555
13300 22222 22222 44444 44444
13301 11111 33333 33333 55555
13302 22222 22222 44444 44444
13303 11111 33333 33333 55555
13304 22222 22222 44444 44444
13309 Numeric values are deprecated but are accepted for backward
13310 compatibility reasons.
13312 Default mode is @code{merge}.
13315 Specify flags influencing the filter process.
13317 Available value for @var{flags} is:
13320 @item low_pass_filter, vlfp
13321 Enable vertical low-pass filtering in the filter.
13322 Vertical low-pass filtering is required when creating an interlaced
13323 destination from a progressive source which contains high-frequency
13324 vertical detail. Filtering will reduce interlace 'twitter' and Moire
13327 Vertical low-pass filtering can only be enabled for @option{mode}
13328 @var{interleave_top} and @var{interleave_bottom}.
13335 Transpose rows with columns in the input video and optionally flip it.
13337 It accepts the following parameters:
13342 Specify the transposition direction.
13344 Can assume the following values:
13346 @item 0, 4, cclock_flip
13347 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
13355 Rotate by 90 degrees clockwise, that is:
13363 Rotate by 90 degrees counterclockwise, that is:
13370 @item 3, 7, clock_flip
13371 Rotate by 90 degrees clockwise and vertically flip, that is:
13379 For values between 4-7, the transposition is only done if the input
13380 video geometry is portrait and not landscape. These values are
13381 deprecated, the @code{passthrough} option should be used instead.
13383 Numerical values are deprecated, and should be dropped in favor of
13384 symbolic constants.
13387 Do not apply the transposition if the input geometry matches the one
13388 specified by the specified value. It accepts the following values:
13391 Always apply transposition.
13393 Preserve portrait geometry (when @var{height} >= @var{width}).
13395 Preserve landscape geometry (when @var{width} >= @var{height}).
13398 Default value is @code{none}.
13401 For example to rotate by 90 degrees clockwise and preserve portrait
13404 transpose=dir=1:passthrough=portrait
13407 The command above can also be specified as:
13409 transpose=1:portrait
13413 Trim the input so that the output contains one continuous subpart of the input.
13415 It accepts the following parameters:
13418 Specify the time of the start of the kept section, i.e. the frame with the
13419 timestamp @var{start} will be the first frame in the output.
13422 Specify the time of the first frame that will be dropped, i.e. the frame
13423 immediately preceding the one with the timestamp @var{end} will be the last
13424 frame in the output.
13427 This is the same as @var{start}, except this option sets the start timestamp
13428 in timebase units instead of seconds.
13431 This is the same as @var{end}, except this option sets the end timestamp
13432 in timebase units instead of seconds.
13435 The maximum duration of the output in seconds.
13438 The number of the first frame that should be passed to the output.
13441 The number of the first frame that should be dropped.
13444 @option{start}, @option{end}, and @option{duration} are expressed as time
13445 duration specifications; see
13446 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13447 for the accepted syntax.
13449 Note that the first two sets of the start/end options and the @option{duration}
13450 option look at the frame timestamp, while the _frame variants simply count the
13451 frames that pass through the filter. Also note that this filter does not modify
13452 the timestamps. If you wish for the output timestamps to start at zero, insert a
13453 setpts filter after the trim filter.
13455 If multiple start or end options are set, this filter tries to be greedy and
13456 keep all the frames that match at least one of the specified constraints. To keep
13457 only the part that matches all the constraints at once, chain multiple trim
13460 The defaults are such that all the input is kept. So it is possible to set e.g.
13461 just the end values to keep everything before the specified time.
13466 Drop everything except the second minute of input:
13468 ffmpeg -i INPUT -vf trim=60:120
13472 Keep only the first second:
13474 ffmpeg -i INPUT -vf trim=duration=1
13483 Sharpen or blur the input video.
13485 It accepts the following parameters:
13488 @item luma_msize_x, lx
13489 Set the luma matrix horizontal size. It must be an odd integer between
13490 3 and 23. The default value is 5.
13492 @item luma_msize_y, ly
13493 Set the luma matrix vertical size. It must be an odd integer between 3
13494 and 23. The default value is 5.
13496 @item luma_amount, la
13497 Set the luma effect strength. It must be a floating point number, reasonable
13498 values lay between -1.5 and 1.5.
13500 Negative values will blur the input video, while positive values will
13501 sharpen it, a value of zero will disable the effect.
13503 Default value is 1.0.
13505 @item chroma_msize_x, cx
13506 Set the chroma matrix horizontal size. It must be an odd integer
13507 between 3 and 23. The default value is 5.
13509 @item chroma_msize_y, cy
13510 Set the chroma matrix vertical size. It must be an odd integer
13511 between 3 and 23. The default value is 5.
13513 @item chroma_amount, ca
13514 Set the chroma effect strength. It must be a floating point number, reasonable
13515 values lay between -1.5 and 1.5.
13517 Negative values will blur the input video, while positive values will
13518 sharpen it, a value of zero will disable the effect.
13520 Default value is 0.0.
13523 If set to 1, specify using OpenCL capabilities, only available if
13524 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13528 All parameters are optional and default to the equivalent of the
13529 string '5:5:1.0:5:5:0.0'.
13531 @subsection Examples
13535 Apply strong luma sharpen effect:
13537 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13541 Apply a strong blur of both luma and chroma parameters:
13543 unsharp=7:7:-2:7:7:-2
13549 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13550 the image at several (or - in the case of @option{quality} level @code{8} - all)
13551 shifts and average the results.
13553 The way this differs from the behavior of spp is that uspp actually encodes &
13554 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13555 DCT similar to MJPEG.
13557 The filter accepts the following options:
13561 Set quality. This option defines the number of levels for averaging. It accepts
13562 an integer in the range 0-8. If set to @code{0}, the filter will have no
13563 effect. A value of @code{8} means the higher quality. For each increment of
13564 that value the speed drops by a factor of approximately 2. Default value is
13568 Force a constant quantization parameter. If not set, the filter will use the QP
13569 from the video stream (if available).
13572 @section vaguedenoiser
13574 Apply a wavelet based denoiser.
13576 It transforms each frame from the video input into the wavelet domain,
13577 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
13578 the obtained coefficients. It does an inverse wavelet transform after.
13579 Due to wavelet properties, it should give a nice smoothed result, and
13580 reduced noise, without blurring picture features.
13582 This filter accepts the following options:
13586 The filtering strength. The higher, the more filtered the video will be.
13587 Hard thresholding can use a higher threshold than soft thresholding
13588 before the video looks overfiltered.
13591 The filtering method the filter will use.
13593 It accepts the following values:
13596 All values under the threshold will be zeroed.
13599 All values under the threshold will be zeroed. All values above will be
13600 reduced by the threshold.
13603 Scales or nullifies coefficients - intermediary between (more) soft and
13604 (less) hard thresholding.
13608 Number of times, the wavelet will decompose the picture. Picture can't
13609 be decomposed beyond a particular point (typically, 8 for a 640x480
13610 frame - as 2^9 = 512 > 480)
13613 Partial of full denoising (limited coefficients shrinking), from 0 to 100.
13616 A list of the planes to process. By default all planes are processed.
13619 @section vectorscope
13621 Display 2 color component values in the two dimensional graph (which is called
13624 This filter accepts the following options:
13628 Set vectorscope mode.
13630 It accepts the following values:
13633 Gray values are displayed on graph, higher brightness means more pixels have
13634 same component color value on location in graph. This is the default mode.
13637 Gray values are displayed on graph. Surrounding pixels values which are not
13638 present in video frame are drawn in gradient of 2 color components which are
13639 set by option @code{x} and @code{y}. The 3rd color component is static.
13642 Actual color components values present in video frame are displayed on graph.
13645 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13646 on graph increases value of another color component, which is luminance by
13647 default values of @code{x} and @code{y}.
13650 Actual colors present in video frame are displayed on graph. If two different
13651 colors map to same position on graph then color with higher value of component
13652 not present in graph is picked.
13655 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13656 component picked from radial gradient.
13660 Set which color component will be represented on X-axis. Default is @code{1}.
13663 Set which color component will be represented on Y-axis. Default is @code{2}.
13666 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13667 of color component which represents frequency of (X, Y) location in graph.
13672 No envelope, this is default.
13675 Instant envelope, even darkest single pixel will be clearly highlighted.
13678 Hold maximum and minimum values presented in graph over time. This way you
13679 can still spot out of range values without constantly looking at vectorscope.
13682 Peak and instant envelope combined together.
13686 Set what kind of graticule to draw.
13694 Set graticule opacity.
13697 Set graticule flags.
13701 Draw graticule for white point.
13704 Draw graticule for black point.
13707 Draw color points short names.
13711 Set background opacity.
13713 @item lthreshold, l
13714 Set low threshold for color component not represented on X or Y axis.
13715 Values lower than this value will be ignored. Default is 0.
13716 Note this value is multiplied with actual max possible value one pixel component
13717 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13720 @item hthreshold, h
13721 Set high threshold for color component not represented on X or Y axis.
13722 Values higher than this value will be ignored. Default is 1.
13723 Note this value is multiplied with actual max possible value one pixel component
13724 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13725 is 0.9 * 255 = 230.
13727 @item colorspace, c
13728 Set what kind of colorspace to use when drawing graticule.
13737 @anchor{vidstabdetect}
13738 @section vidstabdetect
13740 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13741 @ref{vidstabtransform} for pass 2.
13743 This filter generates a file with relative translation and rotation
13744 transform information about subsequent frames, which is then used by
13745 the @ref{vidstabtransform} filter.
13747 To enable compilation of this filter you need to configure FFmpeg with
13748 @code{--enable-libvidstab}.
13750 This filter accepts the following options:
13754 Set the path to the file used to write the transforms information.
13755 Default value is @file{transforms.trf}.
13758 Set how shaky the video is and how quick the camera is. It accepts an
13759 integer in the range 1-10, a value of 1 means little shakiness, a
13760 value of 10 means strong shakiness. Default value is 5.
13763 Set the accuracy of the detection process. It must be a value in the
13764 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13765 accuracy. Default value is 15.
13768 Set stepsize of the search process. The region around minimum is
13769 scanned with 1 pixel resolution. Default value is 6.
13772 Set minimum contrast. Below this value a local measurement field is
13773 discarded. Must be a floating point value in the range 0-1. Default
13777 Set reference frame number for tripod mode.
13779 If enabled, the motion of the frames is compared to a reference frame
13780 in the filtered stream, identified by the specified number. The idea
13781 is to compensate all movements in a more-or-less static scene and keep
13782 the camera view absolutely still.
13784 If set to 0, it is disabled. The frames are counted starting from 1.
13787 Show fields and transforms in the resulting frames. It accepts an
13788 integer in the range 0-2. Default value is 0, which disables any
13792 @subsection Examples
13796 Use default values:
13802 Analyze strongly shaky movie and put the results in file
13803 @file{mytransforms.trf}:
13805 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13809 Visualize the result of internal transformations in the resulting
13812 vidstabdetect=show=1
13816 Analyze a video with medium shakiness using @command{ffmpeg}:
13818 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13822 @anchor{vidstabtransform}
13823 @section vidstabtransform
13825 Video stabilization/deshaking: pass 2 of 2,
13826 see @ref{vidstabdetect} for pass 1.
13828 Read a file with transform information for each frame and
13829 apply/compensate them. Together with the @ref{vidstabdetect}
13830 filter this can be used to deshake videos. See also
13831 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13832 the @ref{unsharp} filter, see below.
13834 To enable compilation of this filter you need to configure FFmpeg with
13835 @code{--enable-libvidstab}.
13837 @subsection Options
13841 Set path to the file used to read the transforms. Default value is
13842 @file{transforms.trf}.
13845 Set the number of frames (value*2 + 1) used for lowpass filtering the
13846 camera movements. Default value is 10.
13848 For example a number of 10 means that 21 frames are used (10 in the
13849 past and 10 in the future) to smoothen the motion in the video. A
13850 larger value leads to a smoother video, but limits the acceleration of
13851 the camera (pan/tilt movements). 0 is a special case where a static
13852 camera is simulated.
13855 Set the camera path optimization algorithm.
13857 Accepted values are:
13860 gaussian kernel low-pass filter on camera motion (default)
13862 averaging on transformations
13866 Set maximal number of pixels to translate frames. Default value is -1,
13870 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13871 value is -1, meaning no limit.
13874 Specify how to deal with borders that may be visible due to movement
13877 Available values are:
13880 keep image information from previous frame (default)
13882 fill the border black
13886 Invert transforms if set to 1. Default value is 0.
13889 Consider transforms as relative to previous frame if set to 1,
13890 absolute if set to 0. Default value is 0.
13893 Set percentage to zoom. A positive value will result in a zoom-in
13894 effect, a negative value in a zoom-out effect. Default value is 0 (no
13898 Set optimal zooming to avoid borders.
13900 Accepted values are:
13905 optimal static zoom value is determined (only very strong movements
13906 will lead to visible borders) (default)
13908 optimal adaptive zoom value is determined (no borders will be
13909 visible), see @option{zoomspeed}
13912 Note that the value given at zoom is added to the one calculated here.
13915 Set percent to zoom maximally each frame (enabled when
13916 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13920 Specify type of interpolation.
13922 Available values are:
13927 linear only horizontal
13929 linear in both directions (default)
13931 cubic in both directions (slow)
13935 Enable virtual tripod mode if set to 1, which is equivalent to
13936 @code{relative=0:smoothing=0}. Default value is 0.
13938 Use also @code{tripod} option of @ref{vidstabdetect}.
13941 Increase log verbosity if set to 1. Also the detected global motions
13942 are written to the temporary file @file{global_motions.trf}. Default
13946 @subsection Examples
13950 Use @command{ffmpeg} for a typical stabilization with default values:
13952 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13955 Note the use of the @ref{unsharp} filter which is always recommended.
13958 Zoom in a bit more and load transform data from a given file:
13960 vidstabtransform=zoom=5:input="mytransforms.trf"
13964 Smoothen the video even more:
13966 vidstabtransform=smoothing=30
13972 Flip the input video vertically.
13974 For example, to vertically flip a video with @command{ffmpeg}:
13976 ffmpeg -i in.avi -vf "vflip" out.avi
13982 Make or reverse a natural vignetting effect.
13984 The filter accepts the following options:
13988 Set lens angle expression as a number of radians.
13990 The value is clipped in the @code{[0,PI/2]} range.
13992 Default value: @code{"PI/5"}
13996 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
14000 Set forward/backward mode.
14002 Available modes are:
14005 The larger the distance from the central point, the darker the image becomes.
14008 The larger the distance from the central point, the brighter the image becomes.
14009 This can be used to reverse a vignette effect, though there is no automatic
14010 detection to extract the lens @option{angle} and other settings (yet). It can
14011 also be used to create a burning effect.
14014 Default value is @samp{forward}.
14017 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
14019 It accepts the following values:
14022 Evaluate expressions only once during the filter initialization.
14025 Evaluate expressions for each incoming frame. This is way slower than the
14026 @samp{init} mode since it requires all the scalers to be re-computed, but it
14027 allows advanced dynamic expressions.
14030 Default value is @samp{init}.
14033 Set dithering to reduce the circular banding effects. Default is @code{1}
14037 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
14038 Setting this value to the SAR of the input will make a rectangular vignetting
14039 following the dimensions of the video.
14041 Default is @code{1/1}.
14044 @subsection Expressions
14046 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
14047 following parameters.
14052 input width and height
14055 the number of input frame, starting from 0
14058 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
14059 @var{TB} units, NAN if undefined
14062 frame rate of the input video, NAN if the input frame rate is unknown
14065 the PTS (Presentation TimeStamp) of the filtered video frame,
14066 expressed in seconds, NAN if undefined
14069 time base of the input video
14073 @subsection Examples
14077 Apply simple strong vignetting effect:
14083 Make a flickering vignetting:
14085 vignette='PI/4+random(1)*PI/50':eval=frame
14091 Stack input videos vertically.
14093 All streams must be of same pixel format and of same width.
14095 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
14096 to create same output.
14098 The filter accept the following option:
14102 Set number of input streams. Default is 2.
14105 If set to 1, force the output to terminate when the shortest input
14106 terminates. Default value is 0.
14111 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
14112 Deinterlacing Filter").
14114 Based on the process described by Martin Weston for BBC R&D, and
14115 implemented based on the de-interlace algorithm written by Jim
14116 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
14117 uses filter coefficients calculated by BBC R&D.
14119 There are two sets of filter coefficients, so called "simple":
14120 and "complex". Which set of filter coefficients is used can
14121 be set by passing an optional parameter:
14125 Set the interlacing filter coefficients. Accepts one of the following values:
14129 Simple filter coefficient set.
14131 More-complex filter coefficient set.
14133 Default value is @samp{complex}.
14136 Specify which frames to deinterlace. Accept one of the following values:
14140 Deinterlace all frames,
14142 Only deinterlace frames marked as interlaced.
14145 Default value is @samp{all}.
14149 Video waveform monitor.
14151 The waveform monitor plots color component intensity. By default luminance
14152 only. Each column of the waveform corresponds to a column of pixels in the
14155 It accepts the following options:
14159 Can be either @code{row}, or @code{column}. Default is @code{column}.
14160 In row mode, the graph on the left side represents color component value 0 and
14161 the right side represents value = 255. In column mode, the top side represents
14162 color component value = 0 and bottom side represents value = 255.
14165 Set intensity. Smaller values are useful to find out how many values of the same
14166 luminance are distributed across input rows/columns.
14167 Default value is @code{0.04}. Allowed range is [0, 1].
14170 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
14171 In mirrored mode, higher values will be represented on the left
14172 side for @code{row} mode and at the top for @code{column} mode. Default is
14173 @code{1} (mirrored).
14177 It accepts the following values:
14180 Presents information identical to that in the @code{parade}, except
14181 that the graphs representing color components are superimposed directly
14184 This display mode makes it easier to spot relative differences or similarities
14185 in overlapping areas of the color components that are supposed to be identical,
14186 such as neutral whites, grays, or blacks.
14189 Display separate graph for the color components side by side in
14190 @code{row} mode or one below the other in @code{column} mode.
14193 Display separate graph for the color components side by side in
14194 @code{column} mode or one below the other in @code{row} mode.
14196 Using this display mode makes it easy to spot color casts in the highlights
14197 and shadows of an image, by comparing the contours of the top and the bottom
14198 graphs of each waveform. Since whites, grays, and blacks are characterized
14199 by exactly equal amounts of red, green, and blue, neutral areas of the picture
14200 should display three waveforms of roughly equal width/height. If not, the
14201 correction is easy to perform by making level adjustments the three waveforms.
14203 Default is @code{stack}.
14205 @item components, c
14206 Set which color components to display. Default is 1, which means only luminance
14207 or red color component if input is in RGB colorspace. If is set for example to
14208 7 it will display all 3 (if) available color components.
14213 No envelope, this is default.
14216 Instant envelope, minimum and maximum values presented in graph will be easily
14217 visible even with small @code{step} value.
14220 Hold minimum and maximum values presented in graph across time. This way you
14221 can still spot out of range values without constantly looking at waveforms.
14224 Peak and instant envelope combined together.
14230 No filtering, this is default.
14233 Luma and chroma combined together.
14236 Similar as above, but shows difference between blue and red chroma.
14239 Displays only chroma.
14242 Displays actual color value on waveform.
14245 Similar as above, but with luma showing frequency of chroma values.
14249 Set which graticule to display.
14253 Do not display graticule.
14256 Display green graticule showing legal broadcast ranges.
14260 Set graticule opacity.
14263 Set graticule flags.
14267 Draw numbers above lines. By default enabled.
14270 Draw dots instead of lines.
14274 Set scale used for displaying graticule.
14281 Default is digital.
14284 Set background opacity.
14289 The @code{weave} takes a field-based video input and join
14290 each two sequential fields into single frame, producing a new double
14291 height clip with half the frame rate and half the frame count.
14293 It accepts the following option:
14297 Set first field. Available values are:
14301 Set the frame as top-field-first.
14304 Set the frame as bottom-field-first.
14308 @subsection Examples
14312 Interlace video using @ref{select} and @ref{separatefields} filter:
14314 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
14319 Apply the xBR high-quality magnification filter which is designed for pixel
14320 art. It follows a set of edge-detection rules, see
14321 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
14323 It accepts the following option:
14327 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
14328 @code{3xBR} and @code{4} for @code{4xBR}.
14329 Default is @code{3}.
14335 Deinterlace the input video ("yadif" means "yet another deinterlacing
14338 It accepts the following parameters:
14344 The interlacing mode to adopt. It accepts one of the following values:
14347 @item 0, send_frame
14348 Output one frame for each frame.
14349 @item 1, send_field
14350 Output one frame for each field.
14351 @item 2, send_frame_nospatial
14352 Like @code{send_frame}, but it skips the spatial interlacing check.
14353 @item 3, send_field_nospatial
14354 Like @code{send_field}, but it skips the spatial interlacing check.
14357 The default value is @code{send_frame}.
14360 The picture field parity assumed for the input interlaced video. It accepts one
14361 of the following values:
14365 Assume the top field is first.
14367 Assume the bottom field is first.
14369 Enable automatic detection of field parity.
14372 The default value is @code{auto}.
14373 If the interlacing is unknown or the decoder does not export this information,
14374 top field first will be assumed.
14377 Specify which frames to deinterlace. Accept one of the following
14382 Deinterlace all frames.
14383 @item 1, interlaced
14384 Only deinterlace frames marked as interlaced.
14387 The default value is @code{all}.
14392 Apply Zoom & Pan effect.
14394 This filter accepts the following options:
14398 Set the zoom expression. Default is 1.
14402 Set the x and y expression. Default is 0.
14405 Set the duration expression in number of frames.
14406 This sets for how many number of frames effect will last for
14407 single input image.
14410 Set the output image size, default is 'hd720'.
14413 Set the output frame rate, default is '25'.
14416 Each expression can contain the following constants:
14435 Output frame count.
14439 Last calculated 'x' and 'y' position from 'x' and 'y' expression
14440 for current input frame.
14444 'x' and 'y' of last output frame of previous input frame or 0 when there was
14445 not yet such frame (first input frame).
14448 Last calculated zoom from 'z' expression for current input frame.
14451 Last calculated zoom of last output frame of previous input frame.
14454 Number of output frames for current input frame. Calculated from 'd' expression
14455 for each input frame.
14458 number of output frames created for previous input frame
14461 Rational number: input width / input height
14464 sample aspect ratio
14467 display aspect ratio
14471 @subsection Examples
14475 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
14477 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
14481 Zoom-in up to 1.5 and pan always at center of picture:
14483 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14487 Same as above but without pausing:
14489 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14494 Scale (resize) the input video, using the z.lib library:
14495 https://github.com/sekrit-twc/zimg.
14497 The zscale filter forces the output display aspect ratio to be the same
14498 as the input, by changing the output sample aspect ratio.
14500 If the input image format is different from the format requested by
14501 the next filter, the zscale filter will convert the input to the
14504 @subsection Options
14505 The filter accepts the following options.
14510 Set the output video dimension expression. Default value is the input
14513 If the @var{width} or @var{w} is 0, the input width is used for the output.
14514 If the @var{height} or @var{h} is 0, the input height is used for the output.
14516 If one of the values is -1, the zscale filter will use a value that
14517 maintains the aspect ratio of the input image, calculated from the
14518 other specified dimension. If both of them are -1, the input size is
14521 If one of the values is -n with n > 1, the zscale filter will also use a value
14522 that maintains the aspect ratio of the input image, calculated from the other
14523 specified dimension. After that it will, however, make sure that the calculated
14524 dimension is divisible by n and adjust the value if necessary.
14526 See below for the list of accepted constants for use in the dimension
14530 Set the video size. For the syntax of this option, check the
14531 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14534 Set the dither type.
14536 Possible values are:
14541 @item error_diffusion
14547 Set the resize filter type.
14549 Possible values are:
14559 Default is bilinear.
14562 Set the color range.
14564 Possible values are:
14571 Default is same as input.
14574 Set the color primaries.
14576 Possible values are:
14586 Default is same as input.
14589 Set the transfer characteristics.
14591 Possible values are:
14602 Default is same as input.
14605 Set the colorspace matrix.
14607 Possible value are:
14618 Default is same as input.
14621 Set the input color range.
14623 Possible values are:
14630 Default is same as input.
14632 @item primariesin, pin
14633 Set the input color primaries.
14635 Possible values are:
14645 Default is same as input.
14647 @item transferin, tin
14648 Set the input transfer characteristics.
14650 Possible values are:
14661 Default is same as input.
14663 @item matrixin, min
14664 Set the input colorspace matrix.
14666 Possible value are:
14678 Set the output chroma location.
14680 Possible values are:
14691 @item chromalin, cin
14692 Set the input chroma location.
14694 Possible values are:
14706 The values of the @option{w} and @option{h} options are expressions
14707 containing the following constants:
14712 The input width and height
14716 These are the same as @var{in_w} and @var{in_h}.
14720 The output (scaled) width and height
14724 These are the same as @var{out_w} and @var{out_h}
14727 The same as @var{iw} / @var{ih}
14730 input sample aspect ratio
14733 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14737 horizontal and vertical input chroma subsample values. For example for the
14738 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14742 horizontal and vertical output chroma subsample values. For example for the
14743 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14749 @c man end VIDEO FILTERS
14751 @chapter Video Sources
14752 @c man begin VIDEO SOURCES
14754 Below is a description of the currently available video sources.
14758 Buffer video frames, and make them available to the filter chain.
14760 This source is mainly intended for a programmatic use, in particular
14761 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14763 It accepts the following parameters:
14768 Specify the size (width and height) of the buffered video frames. For the
14769 syntax of this option, check the
14770 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14773 The input video width.
14776 The input video height.
14779 A string representing the pixel format of the buffered video frames.
14780 It may be a number corresponding to a pixel format, or a pixel format
14784 Specify the timebase assumed by the timestamps of the buffered frames.
14787 Specify the frame rate expected for the video stream.
14789 @item pixel_aspect, sar
14790 The sample (pixel) aspect ratio of the input video.
14793 Specify the optional parameters to be used for the scale filter which
14794 is automatically inserted when an input change is detected in the
14795 input size or format.
14797 @item hw_frames_ctx
14798 When using a hardware pixel format, this should be a reference to an
14799 AVHWFramesContext describing input frames.
14804 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14807 will instruct the source to accept video frames with size 320x240 and
14808 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14809 square pixels (1:1 sample aspect ratio).
14810 Since the pixel format with name "yuv410p" corresponds to the number 6
14811 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14812 this example corresponds to:
14814 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14817 Alternatively, the options can be specified as a flat string, but this
14818 syntax is deprecated:
14820 @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}]
14824 Create a pattern generated by an elementary cellular automaton.
14826 The initial state of the cellular automaton can be defined through the
14827 @option{filename} and @option{pattern} options. If such options are
14828 not specified an initial state is created randomly.
14830 At each new frame a new row in the video is filled with the result of
14831 the cellular automaton next generation. The behavior when the whole
14832 frame is filled is defined by the @option{scroll} option.
14834 This source accepts the following options:
14838 Read the initial cellular automaton state, i.e. the starting row, from
14839 the specified file.
14840 In the file, each non-whitespace character is considered an alive
14841 cell, a newline will terminate the row, and further characters in the
14842 file will be ignored.
14845 Read the initial cellular automaton state, i.e. the starting row, from
14846 the specified string.
14848 Each non-whitespace character in the string is considered an alive
14849 cell, a newline will terminate the row, and further characters in the
14850 string will be ignored.
14853 Set the video rate, that is the number of frames generated per second.
14856 @item random_fill_ratio, ratio
14857 Set the random fill ratio for the initial cellular automaton row. It
14858 is a floating point number value ranging from 0 to 1, defaults to
14861 This option is ignored when a file or a pattern is specified.
14863 @item random_seed, seed
14864 Set the seed for filling randomly the initial row, must be an integer
14865 included between 0 and UINT32_MAX. If not specified, or if explicitly
14866 set to -1, the filter will try to use a good random seed on a best
14870 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14871 Default value is 110.
14874 Set the size of the output video. For the syntax of this option, check the
14875 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14877 If @option{filename} or @option{pattern} is specified, the size is set
14878 by default to the width of the specified initial state row, and the
14879 height is set to @var{width} * PHI.
14881 If @option{size} is set, it must contain the width of the specified
14882 pattern string, and the specified pattern will be centered in the
14885 If a filename or a pattern string is not specified, the size value
14886 defaults to "320x518" (used for a randomly generated initial state).
14889 If set to 1, scroll the output upward when all the rows in the output
14890 have been already filled. If set to 0, the new generated row will be
14891 written over the top row just after the bottom row is filled.
14894 @item start_full, full
14895 If set to 1, completely fill the output with generated rows before
14896 outputting the first frame.
14897 This is the default behavior, for disabling set the value to 0.
14900 If set to 1, stitch the left and right row edges together.
14901 This is the default behavior, for disabling set the value to 0.
14904 @subsection Examples
14908 Read the initial state from @file{pattern}, and specify an output of
14911 cellauto=f=pattern:s=200x400
14915 Generate a random initial row with a width of 200 cells, with a fill
14918 cellauto=ratio=2/3:s=200x200
14922 Create a pattern generated by rule 18 starting by a single alive cell
14923 centered on an initial row with width 100:
14925 cellauto=p=@@:s=100x400:full=0:rule=18
14929 Specify a more elaborated initial pattern:
14931 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14936 @anchor{coreimagesrc}
14937 @section coreimagesrc
14938 Video source generated on GPU using Apple's CoreImage API on OSX.
14940 This video source is a specialized version of the @ref{coreimage} video filter.
14941 Use a core image generator at the beginning of the applied filterchain to
14942 generate the content.
14944 The coreimagesrc video source accepts the following options:
14946 @item list_generators
14947 List all available generators along with all their respective options as well as
14948 possible minimum and maximum values along with the default values.
14950 list_generators=true
14954 Specify the size of the sourced video. For the syntax of this option, check the
14955 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14956 The default value is @code{320x240}.
14959 Specify the frame rate of the sourced video, as the number of frames
14960 generated per second. It has to be a string in the format
14961 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14962 number or a valid video frame rate abbreviation. The default value is
14966 Set the sample aspect ratio of the sourced video.
14969 Set the duration of the sourced video. See
14970 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14971 for the accepted syntax.
14973 If not specified, or the expressed duration is negative, the video is
14974 supposed to be generated forever.
14977 Additionally, all options of the @ref{coreimage} video filter are accepted.
14978 A complete filterchain can be used for further processing of the
14979 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
14980 and examples for details.
14982 @subsection Examples
14987 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
14988 given as complete and escaped command-line for Apple's standard bash shell:
14990 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
14992 This example is equivalent to the QRCode example of @ref{coreimage} without the
14993 need for a nullsrc video source.
14997 @section mandelbrot
14999 Generate a Mandelbrot set fractal, and progressively zoom towards the
15000 point specified with @var{start_x} and @var{start_y}.
15002 This source accepts the following options:
15007 Set the terminal pts value. Default value is 400.
15010 Set the terminal scale value.
15011 Must be a floating point value. Default value is 0.3.
15014 Set the inner coloring mode, that is the algorithm used to draw the
15015 Mandelbrot fractal internal region.
15017 It shall assume one of the following values:
15022 Show time until convergence.
15024 Set color based on point closest to the origin of the iterations.
15029 Default value is @var{mincol}.
15032 Set the bailout value. Default value is 10.0.
15035 Set the maximum of iterations performed by the rendering
15036 algorithm. Default value is 7189.
15039 Set outer coloring mode.
15040 It shall assume one of following values:
15042 @item iteration_count
15043 Set iteration cound mode.
15044 @item normalized_iteration_count
15045 set normalized iteration count mode.
15047 Default value is @var{normalized_iteration_count}.
15050 Set frame rate, expressed as number of frames per second. Default
15054 Set frame size. For the syntax of this option, check the "Video
15055 size" section in the ffmpeg-utils manual. Default value is "640x480".
15058 Set the initial scale value. Default value is 3.0.
15061 Set the initial x position. Must be a floating point value between
15062 -100 and 100. Default value is -0.743643887037158704752191506114774.
15065 Set the initial y position. Must be a floating point value between
15066 -100 and 100. Default value is -0.131825904205311970493132056385139.
15071 Generate various test patterns, as generated by the MPlayer test filter.
15073 The size of the generated video is fixed, and is 256x256.
15074 This source is useful in particular for testing encoding features.
15076 This source accepts the following options:
15081 Specify the frame rate of the sourced video, as the number of frames
15082 generated per second. It has to be a string in the format
15083 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15084 number or a valid video frame rate abbreviation. The default value is
15088 Set the duration of the sourced video. See
15089 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15090 for the accepted syntax.
15092 If not specified, or the expressed duration is negative, the video is
15093 supposed to be generated forever.
15097 Set the number or the name of the test to perform. Supported tests are:
15113 Default value is "all", which will cycle through the list of all tests.
15118 mptestsrc=t=dc_luma
15121 will generate a "dc_luma" test pattern.
15123 @section frei0r_src
15125 Provide a frei0r source.
15127 To enable compilation of this filter you need to install the frei0r
15128 header and configure FFmpeg with @code{--enable-frei0r}.
15130 This source accepts the following parameters:
15135 The size of the video to generate. For the syntax of this option, check the
15136 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15139 The framerate of the generated video. It may be a string of the form
15140 @var{num}/@var{den} or a frame rate abbreviation.
15143 The name to the frei0r source to load. For more information regarding frei0r and
15144 how to set the parameters, read the @ref{frei0r} section in the video filters
15147 @item filter_params
15148 A '|'-separated list of parameters to pass to the frei0r source.
15152 For example, to generate a frei0r partik0l source with size 200x200
15153 and frame rate 10 which is overlaid on the overlay filter main input:
15155 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
15160 Generate a life pattern.
15162 This source is based on a generalization of John Conway's life game.
15164 The sourced input represents a life grid, each pixel represents a cell
15165 which can be in one of two possible states, alive or dead. Every cell
15166 interacts with its eight neighbours, which are the cells that are
15167 horizontally, vertically, or diagonally adjacent.
15169 At each interaction the grid evolves according to the adopted rule,
15170 which specifies the number of neighbor alive cells which will make a
15171 cell stay alive or born. The @option{rule} option allows one to specify
15174 This source accepts the following options:
15178 Set the file from which to read the initial grid state. In the file,
15179 each non-whitespace character is considered an alive cell, and newline
15180 is used to delimit the end of each row.
15182 If this option is not specified, the initial grid is generated
15186 Set the video rate, that is the number of frames generated per second.
15189 @item random_fill_ratio, ratio
15190 Set the random fill ratio for the initial random grid. It is a
15191 floating point number value ranging from 0 to 1, defaults to 1/PHI.
15192 It is ignored when a file is specified.
15194 @item random_seed, seed
15195 Set the seed for filling the initial random grid, must be an integer
15196 included between 0 and UINT32_MAX. If not specified, or if explicitly
15197 set to -1, the filter will try to use a good random seed on a best
15203 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
15204 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
15205 @var{NS} specifies the number of alive neighbor cells which make a
15206 live cell stay alive, and @var{NB} the number of alive neighbor cells
15207 which make a dead cell to become alive (i.e. to "born").
15208 "s" and "b" can be used in place of "S" and "B", respectively.
15210 Alternatively a rule can be specified by an 18-bits integer. The 9
15211 high order bits are used to encode the next cell state if it is alive
15212 for each number of neighbor alive cells, the low order bits specify
15213 the rule for "borning" new cells. Higher order bits encode for an
15214 higher number of neighbor cells.
15215 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
15216 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
15218 Default value is "S23/B3", which is the original Conway's game of life
15219 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
15220 cells, and will born a new cell if there are three alive cells around
15224 Set the size of the output video. For the syntax of this option, check the
15225 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15227 If @option{filename} is specified, the size is set by default to the
15228 same size of the input file. If @option{size} is set, it must contain
15229 the size specified in the input file, and the initial grid defined in
15230 that file is centered in the larger resulting area.
15232 If a filename is not specified, the size value defaults to "320x240"
15233 (used for a randomly generated initial grid).
15236 If set to 1, stitch the left and right grid edges together, and the
15237 top and bottom edges also. Defaults to 1.
15240 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
15241 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
15242 value from 0 to 255.
15245 Set the color of living (or new born) cells.
15248 Set the color of dead cells. If @option{mold} is set, this is the first color
15249 used to represent a dead cell.
15252 Set mold color, for definitely dead and moldy cells.
15254 For the syntax of these 3 color options, check the "Color" section in the
15255 ffmpeg-utils manual.
15258 @subsection Examples
15262 Read a grid from @file{pattern}, and center it on a grid of size
15265 life=f=pattern:s=300x300
15269 Generate a random grid of size 200x200, with a fill ratio of 2/3:
15271 life=ratio=2/3:s=200x200
15275 Specify a custom rule for evolving a randomly generated grid:
15281 Full example with slow death effect (mold) using @command{ffplay}:
15283 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
15290 @anchor{haldclutsrc}
15292 @anchor{rgbtestsrc}
15294 @anchor{smptehdbars}
15297 @anchor{yuvtestsrc}
15298 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
15300 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
15302 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
15304 The @code{color} source provides an uniformly colored input.
15306 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
15307 @ref{haldclut} filter.
15309 The @code{nullsrc} source returns unprocessed video frames. It is
15310 mainly useful to be employed in analysis / debugging tools, or as the
15311 source for filters which ignore the input data.
15313 The @code{rgbtestsrc} source generates an RGB test pattern useful for
15314 detecting RGB vs BGR issues. You should see a red, green and blue
15315 stripe from top to bottom.
15317 The @code{smptebars} source generates a color bars pattern, based on
15318 the SMPTE Engineering Guideline EG 1-1990.
15320 The @code{smptehdbars} source generates a color bars pattern, based on
15321 the SMPTE RP 219-2002.
15323 The @code{testsrc} source generates a test video pattern, showing a
15324 color pattern, a scrolling gradient and a timestamp. This is mainly
15325 intended for testing purposes.
15327 The @code{testsrc2} source is similar to testsrc, but supports more
15328 pixel formats instead of just @code{rgb24}. This allows using it as an
15329 input for other tests without requiring a format conversion.
15331 The @code{yuvtestsrc} source generates an YUV test pattern. You should
15332 see a y, cb and cr stripe from top to bottom.
15334 The sources accept the following parameters:
15339 Specify the color of the source, only available in the @code{color}
15340 source. For the syntax of this option, check the "Color" section in the
15341 ffmpeg-utils manual.
15344 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
15345 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
15346 pixels to be used as identity matrix for 3D lookup tables. Each component is
15347 coded on a @code{1/(N*N)} scale.
15350 Specify the size of the sourced video. For the syntax of this option, check the
15351 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15352 The default value is @code{320x240}.
15354 This option is not available with the @code{haldclutsrc} filter.
15357 Specify the frame rate of the sourced video, as the number of frames
15358 generated per second. It has to be a string in the format
15359 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15360 number or a valid video frame rate abbreviation. The default value is
15364 Set the sample aspect ratio of the sourced video.
15367 Set the duration of the sourced video. See
15368 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15369 for the accepted syntax.
15371 If not specified, or the expressed duration is negative, the video is
15372 supposed to be generated forever.
15375 Set the number of decimals to show in the timestamp, only available in the
15376 @code{testsrc} source.
15378 The displayed timestamp value will correspond to the original
15379 timestamp value multiplied by the power of 10 of the specified
15380 value. Default value is 0.
15383 For example the following:
15385 testsrc=duration=5.3:size=qcif:rate=10
15388 will generate a video with a duration of 5.3 seconds, with size
15389 176x144 and a frame rate of 10 frames per second.
15391 The following graph description will generate a red source
15392 with an opacity of 0.2, with size "qcif" and a frame rate of 10
15395 color=c=red@@0.2:s=qcif:r=10
15398 If the input content is to be ignored, @code{nullsrc} can be used. The
15399 following command generates noise in the luminance plane by employing
15400 the @code{geq} filter:
15402 nullsrc=s=256x256, geq=random(1)*255:128:128
15405 @subsection Commands
15407 The @code{color} source supports the following commands:
15411 Set the color of the created image. Accepts the same syntax of the
15412 corresponding @option{color} option.
15415 @c man end VIDEO SOURCES
15417 @chapter Video Sinks
15418 @c man begin VIDEO SINKS
15420 Below is a description of the currently available video sinks.
15422 @section buffersink
15424 Buffer video frames, and make them available to the end of the filter
15427 This sink is mainly intended for programmatic use, in particular
15428 through the interface defined in @file{libavfilter/buffersink.h}
15429 or the options system.
15431 It accepts a pointer to an AVBufferSinkContext structure, which
15432 defines the incoming buffers' formats, to be passed as the opaque
15433 parameter to @code{avfilter_init_filter} for initialization.
15437 Null video sink: do absolutely nothing with the input video. It is
15438 mainly useful as a template and for use in analysis / debugging
15441 @c man end VIDEO SINKS
15443 @chapter Multimedia Filters
15444 @c man begin MULTIMEDIA FILTERS
15446 Below is a description of the currently available multimedia filters.
15448 @section ahistogram
15450 Convert input audio to a video output, displaying the volume histogram.
15452 The filter accepts the following options:
15456 Specify how histogram is calculated.
15458 It accepts the following values:
15461 Use single histogram for all channels.
15463 Use separate histogram for each channel.
15465 Default is @code{single}.
15468 Set frame rate, expressed as number of frames per second. Default
15472 Specify the video size for the output. For the syntax of this option, check the
15473 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15474 Default value is @code{hd720}.
15479 It accepts the following values:
15490 reverse logarithmic
15492 Default is @code{log}.
15495 Set amplitude scale.
15497 It accepts the following values:
15504 Default is @code{log}.
15507 Set how much frames to accumulate in histogram.
15508 Defauls is 1. Setting this to -1 accumulates all frames.
15511 Set histogram ratio of window height.
15514 Set sonogram sliding.
15516 It accepts the following values:
15519 replace old rows with new ones.
15521 scroll from top to bottom.
15523 Default is @code{replace}.
15526 @section aphasemeter
15528 Convert input audio to a video output, displaying the audio phase.
15530 The filter accepts the following options:
15534 Set the output frame rate. Default value is @code{25}.
15537 Set the video size for the output. For the syntax of this option, check the
15538 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15539 Default value is @code{800x400}.
15544 Specify the red, green, blue contrast. Default values are @code{2},
15545 @code{7} and @code{1}.
15546 Allowed range is @code{[0, 255]}.
15549 Set color which will be used for drawing median phase. If color is
15550 @code{none} which is default, no median phase value will be drawn.
15553 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15554 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15555 The @code{-1} means left and right channels are completely out of phase and
15556 @code{1} means channels are in phase.
15558 @section avectorscope
15560 Convert input audio to a video output, representing the audio vector
15563 The filter is used to measure the difference between channels of stereo
15564 audio stream. A monoaural signal, consisting of identical left and right
15565 signal, results in straight vertical line. Any stereo separation is visible
15566 as a deviation from this line, creating a Lissajous figure.
15567 If the straight (or deviation from it) but horizontal line appears this
15568 indicates that the left and right channels are out of phase.
15570 The filter accepts the following options:
15574 Set the vectorscope mode.
15576 Available values are:
15579 Lissajous rotated by 45 degrees.
15582 Same as above but not rotated.
15585 Shape resembling half of circle.
15588 Default value is @samp{lissajous}.
15591 Set the video size for the output. For the syntax of this option, check the
15592 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15593 Default value is @code{400x400}.
15596 Set the output frame rate. Default value is @code{25}.
15602 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15603 @code{160}, @code{80} and @code{255}.
15604 Allowed range is @code{[0, 255]}.
15610 Specify the red, green, blue and alpha fade. Default values are @code{15},
15611 @code{10}, @code{5} and @code{5}.
15612 Allowed range is @code{[0, 255]}.
15615 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15618 Set the vectorscope drawing mode.
15620 Available values are:
15623 Draw dot for each sample.
15626 Draw line between previous and current sample.
15629 Default value is @samp{dot}.
15632 Specify amplitude scale of audio samples.
15634 Available values are:
15651 @subsection Examples
15655 Complete example using @command{ffplay}:
15657 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15658 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15662 @section bench, abench
15664 Benchmark part of a filtergraph.
15666 The filter accepts the following options:
15670 Start or stop a timer.
15672 Available values are:
15675 Get the current time, set it as frame metadata (using the key
15676 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15679 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15680 the input frame metadata to get the time difference. Time difference, average,
15681 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15682 @code{min}) are then printed. The timestamps are expressed in seconds.
15686 @subsection Examples
15690 Benchmark @ref{selectivecolor} filter:
15692 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15698 Concatenate audio and video streams, joining them together one after the
15701 The filter works on segments of synchronized video and audio streams. All
15702 segments must have the same number of streams of each type, and that will
15703 also be the number of streams at output.
15705 The filter accepts the following options:
15710 Set the number of segments. Default is 2.
15713 Set the number of output video streams, that is also the number of video
15714 streams in each segment. Default is 1.
15717 Set the number of output audio streams, that is also the number of audio
15718 streams in each segment. Default is 0.
15721 Activate unsafe mode: do not fail if segments have a different format.
15725 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15726 @var{a} audio outputs.
15728 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15729 segment, in the same order as the outputs, then the inputs for the second
15732 Related streams do not always have exactly the same duration, for various
15733 reasons including codec frame size or sloppy authoring. For that reason,
15734 related synchronized streams (e.g. a video and its audio track) should be
15735 concatenated at once. The concat filter will use the duration of the longest
15736 stream in each segment (except the last one), and if necessary pad shorter
15737 audio streams with silence.
15739 For this filter to work correctly, all segments must start at timestamp 0.
15741 All corresponding streams must have the same parameters in all segments; the
15742 filtering system will automatically select a common pixel format for video
15743 streams, and a common sample format, sample rate and channel layout for
15744 audio streams, but other settings, such as resolution, must be converted
15745 explicitly by the user.
15747 Different frame rates are acceptable but will result in variable frame rate
15748 at output; be sure to configure the output file to handle it.
15750 @subsection Examples
15754 Concatenate an opening, an episode and an ending, all in bilingual version
15755 (video in stream 0, audio in streams 1 and 2):
15757 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15758 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15759 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15760 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15764 Concatenate two parts, handling audio and video separately, using the
15765 (a)movie sources, and adjusting the resolution:
15767 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15768 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15769 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15771 Note that a desync will happen at the stitch if the audio and video streams
15772 do not have exactly the same duration in the first file.
15776 @section drawgraph, adrawgraph
15778 Draw a graph using input video or audio metadata.
15780 It accepts the following parameters:
15784 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15787 Set 1st foreground color expression.
15790 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15793 Set 2nd foreground color expression.
15796 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15799 Set 3rd foreground color expression.
15802 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15805 Set 4th foreground color expression.
15808 Set minimal value of metadata value.
15811 Set maximal value of metadata value.
15814 Set graph background color. Default is white.
15819 Available values for mode is:
15826 Default is @code{line}.
15831 Available values for slide is:
15834 Draw new frame when right border is reached.
15837 Replace old columns with new ones.
15840 Scroll from right to left.
15843 Scroll from left to right.
15846 Draw single picture.
15849 Default is @code{frame}.
15852 Set size of graph video. For the syntax of this option, check the
15853 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15854 The default value is @code{900x256}.
15856 The foreground color expressions can use the following variables:
15859 Minimal value of metadata value.
15862 Maximal value of metadata value.
15865 Current metadata key value.
15868 The color is defined as 0xAABBGGRR.
15871 Example using metadata from @ref{signalstats} filter:
15873 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15876 Example using metadata from @ref{ebur128} filter:
15878 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15884 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15885 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15886 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15887 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15889 The filter also has a video output (see the @var{video} option) with a real
15890 time graph to observe the loudness evolution. The graphic contains the logged
15891 message mentioned above, so it is not printed anymore when this option is set,
15892 unless the verbose logging is set. The main graphing area contains the
15893 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15894 the momentary loudness (400 milliseconds).
15896 More information about the Loudness Recommendation EBU R128 on
15897 @url{http://tech.ebu.ch/loudness}.
15899 The filter accepts the following options:
15904 Activate the video output. The audio stream is passed unchanged whether this
15905 option is set or no. The video stream will be the first output stream if
15906 activated. Default is @code{0}.
15909 Set the video size. This option is for video only. For the syntax of this
15911 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15912 Default and minimum resolution is @code{640x480}.
15915 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15916 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15917 other integer value between this range is allowed.
15920 Set metadata injection. If set to @code{1}, the audio input will be segmented
15921 into 100ms output frames, each of them containing various loudness information
15922 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15924 Default is @code{0}.
15927 Force the frame logging level.
15929 Available values are:
15932 information logging level
15934 verbose logging level
15937 By default, the logging level is set to @var{info}. If the @option{video} or
15938 the @option{metadata} options are set, it switches to @var{verbose}.
15943 Available modes can be cumulated (the option is a @code{flag} type). Possible
15947 Disable any peak mode (default).
15949 Enable sample-peak mode.
15951 Simple peak mode looking for the higher sample value. It logs a message
15952 for sample-peak (identified by @code{SPK}).
15954 Enable true-peak mode.
15956 If enabled, the peak lookup is done on an over-sampled version of the input
15957 stream for better peak accuracy. It logs a message for true-peak.
15958 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15959 This mode requires a build with @code{libswresample}.
15963 Treat mono input files as "dual mono". If a mono file is intended for playback
15964 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15965 If set to @code{true}, this option will compensate for this effect.
15966 Multi-channel input files are not affected by this option.
15969 Set a specific pan law to be used for the measurement of dual mono files.
15970 This parameter is optional, and has a default value of -3.01dB.
15973 @subsection Examples
15977 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
15979 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
15983 Run an analysis with @command{ffmpeg}:
15985 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
15989 @section interleave, ainterleave
15991 Temporally interleave frames from several inputs.
15993 @code{interleave} works with video inputs, @code{ainterleave} with audio.
15995 These filters read frames from several inputs and send the oldest
15996 queued frame to the output.
15998 Input streams must have well defined, monotonically increasing frame
16001 In order to submit one frame to output, these filters need to enqueue
16002 at least one frame for each input, so they cannot work in case one
16003 input is not yet terminated and will not receive incoming frames.
16005 For example consider the case when one input is a @code{select} filter
16006 which always drops input frames. The @code{interleave} filter will keep
16007 reading from that input, but it will never be able to send new frames
16008 to output until the input sends an end-of-stream signal.
16010 Also, depending on inputs synchronization, the filters will drop
16011 frames in case one input receives more frames than the other ones, and
16012 the queue is already filled.
16014 These filters accept the following options:
16018 Set the number of different inputs, it is 2 by default.
16021 @subsection Examples
16025 Interleave frames belonging to different streams using @command{ffmpeg}:
16027 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
16031 Add flickering blur effect:
16033 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
16037 @section metadata, ametadata
16039 Manipulate frame metadata.
16041 This filter accepts the following options:
16045 Set mode of operation of the filter.
16047 Can be one of the following:
16051 If both @code{value} and @code{key} is set, select frames
16052 which have such metadata. If only @code{key} is set, select
16053 every frame that has such key in metadata.
16056 Add new metadata @code{key} and @code{value}. If key is already available
16060 Modify value of already present key.
16063 If @code{value} is set, delete only keys that have such value.
16064 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
16068 Print key and its value if metadata was found. If @code{key} is not set print all
16069 metadata values available in frame.
16073 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
16076 Set metadata value which will be used. This option is mandatory for
16077 @code{modify} and @code{add} mode.
16080 Which function to use when comparing metadata value and @code{value}.
16082 Can be one of following:
16086 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
16089 Values are interpreted as strings, returns true if metadata value starts with
16090 the @code{value} option string.
16093 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
16096 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
16099 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
16102 Values are interpreted as floats, returns true if expression from option @code{expr}
16107 Set expression which is used when @code{function} is set to @code{expr}.
16108 The expression is evaluated through the eval API and can contain the following
16113 Float representation of @code{value} from metadata key.
16116 Float representation of @code{value} as supplied by user in @code{value} option.
16119 If specified in @code{print} mode, output is written to the named file. Instead of
16120 plain filename any writable url can be specified. Filename ``-'' is a shorthand
16121 for standard output. If @code{file} option is not set, output is written to the log
16122 with AV_LOG_INFO loglevel.
16127 @subsection Examples
16131 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
16134 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
16137 Print silencedetect output to file @file{metadata.txt}.
16139 silencedetect,ametadata=mode=print:file=metadata.txt
16142 Direct all metadata to a pipe with file descriptor 4.
16144 metadata=mode=print:file='pipe\:4'
16148 @section perms, aperms
16150 Set read/write permissions for the output frames.
16152 These filters are mainly aimed at developers to test direct path in the
16153 following filter in the filtergraph.
16155 The filters accept the following options:
16159 Select the permissions mode.
16161 It accepts the following values:
16164 Do nothing. This is the default.
16166 Set all the output frames read-only.
16168 Set all the output frames directly writable.
16170 Make the frame read-only if writable, and writable if read-only.
16172 Set each output frame read-only or writable randomly.
16176 Set the seed for the @var{random} mode, must be an integer included between
16177 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16178 @code{-1}, the filter will try to use a good random seed on a best effort
16182 Note: in case of auto-inserted filter between the permission filter and the
16183 following one, the permission might not be received as expected in that
16184 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
16185 perms/aperms filter can avoid this problem.
16187 @section realtime, arealtime
16189 Slow down filtering to match real time approximatively.
16191 These filters will pause the filtering for a variable amount of time to
16192 match the output rate with the input timestamps.
16193 They are similar to the @option{re} option to @code{ffmpeg}.
16195 They accept the following options:
16199 Time limit for the pauses. Any pause longer than that will be considered
16200 a timestamp discontinuity and reset the timer. Default is 2 seconds.
16204 @section select, aselect
16206 Select frames to pass in output.
16208 This filter accepts the following options:
16213 Set expression, which is evaluated for each input frame.
16215 If the expression is evaluated to zero, the frame is discarded.
16217 If the evaluation result is negative or NaN, the frame is sent to the
16218 first output; otherwise it is sent to the output with index
16219 @code{ceil(val)-1}, assuming that the input index starts from 0.
16221 For example a value of @code{1.2} corresponds to the output with index
16222 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
16225 Set the number of outputs. The output to which to send the selected
16226 frame is based on the result of the evaluation. Default value is 1.
16229 The expression can contain the following constants:
16233 The (sequential) number of the filtered frame, starting from 0.
16236 The (sequential) number of the selected frame, starting from 0.
16238 @item prev_selected_n
16239 The sequential number of the last selected frame. It's NAN if undefined.
16242 The timebase of the input timestamps.
16245 The PTS (Presentation TimeStamp) of the filtered video frame,
16246 expressed in @var{TB} units. It's NAN if undefined.
16249 The PTS of the filtered video frame,
16250 expressed in seconds. It's NAN if undefined.
16253 The PTS of the previously filtered video frame. It's NAN if undefined.
16255 @item prev_selected_pts
16256 The PTS of the last previously filtered video frame. It's NAN if undefined.
16258 @item prev_selected_t
16259 The PTS of the last previously selected video frame. It's NAN if undefined.
16262 The PTS of the first video frame in the video. It's NAN if undefined.
16265 The time of the first video frame in the video. It's NAN if undefined.
16267 @item pict_type @emph{(video only)}
16268 The type of the filtered frame. It can assume one of the following
16280 @item interlace_type @emph{(video only)}
16281 The frame interlace type. It can assume one of the following values:
16284 The frame is progressive (not interlaced).
16286 The frame is top-field-first.
16288 The frame is bottom-field-first.
16291 @item consumed_sample_n @emph{(audio only)}
16292 the number of selected samples before the current frame
16294 @item samples_n @emph{(audio only)}
16295 the number of samples in the current frame
16297 @item sample_rate @emph{(audio only)}
16298 the input sample rate
16301 This is 1 if the filtered frame is a key-frame, 0 otherwise.
16304 the position in the file of the filtered frame, -1 if the information
16305 is not available (e.g. for synthetic video)
16307 @item scene @emph{(video only)}
16308 value between 0 and 1 to indicate a new scene; a low value reflects a low
16309 probability for the current frame to introduce a new scene, while a higher
16310 value means the current frame is more likely to be one (see the example below)
16312 @item concatdec_select
16313 The concat demuxer can select only part of a concat input file by setting an
16314 inpoint and an outpoint, but the output packets may not be entirely contained
16315 in the selected interval. By using this variable, it is possible to skip frames
16316 generated by the concat demuxer which are not exactly contained in the selected
16319 This works by comparing the frame pts against the @var{lavf.concat.start_time}
16320 and the @var{lavf.concat.duration} packet metadata values which are also
16321 present in the decoded frames.
16323 The @var{concatdec_select} variable is -1 if the frame pts is at least
16324 start_time and either the duration metadata is missing or the frame pts is less
16325 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
16328 That basically means that an input frame is selected if its pts is within the
16329 interval set by the concat demuxer.
16333 The default value of the select expression is "1".
16335 @subsection Examples
16339 Select all frames in input:
16344 The example above is the same as:
16356 Select only I-frames:
16358 select='eq(pict_type\,I)'
16362 Select one frame every 100:
16364 select='not(mod(n\,100))'
16368 Select only frames contained in the 10-20 time interval:
16370 select=between(t\,10\,20)
16374 Select only I-frames contained in the 10-20 time interval:
16376 select=between(t\,10\,20)*eq(pict_type\,I)
16380 Select frames with a minimum distance of 10 seconds:
16382 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
16386 Use aselect to select only audio frames with samples number > 100:
16388 aselect='gt(samples_n\,100)'
16392 Create a mosaic of the first scenes:
16394 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
16397 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
16401 Send even and odd frames to separate outputs, and compose them:
16403 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
16407 Select useful frames from an ffconcat file which is using inpoints and
16408 outpoints but where the source files are not intra frame only.
16410 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
16414 @section sendcmd, asendcmd
16416 Send commands to filters in the filtergraph.
16418 These filters read commands to be sent to other filters in the
16421 @code{sendcmd} must be inserted between two video filters,
16422 @code{asendcmd} must be inserted between two audio filters, but apart
16423 from that they act the same way.
16425 The specification of commands can be provided in the filter arguments
16426 with the @var{commands} option, or in a file specified by the
16427 @var{filename} option.
16429 These filters accept the following options:
16432 Set the commands to be read and sent to the other filters.
16434 Set the filename of the commands to be read and sent to the other
16438 @subsection Commands syntax
16440 A commands description consists of a sequence of interval
16441 specifications, comprising a list of commands to be executed when a
16442 particular event related to that interval occurs. The occurring event
16443 is typically the current frame time entering or leaving a given time
16446 An interval is specified by the following syntax:
16448 @var{START}[-@var{END}] @var{COMMANDS};
16451 The time interval is specified by the @var{START} and @var{END} times.
16452 @var{END} is optional and defaults to the maximum time.
16454 The current frame time is considered within the specified interval if
16455 it is included in the interval [@var{START}, @var{END}), that is when
16456 the time is greater or equal to @var{START} and is lesser than
16459 @var{COMMANDS} consists of a sequence of one or more command
16460 specifications, separated by ",", relating to that interval. The
16461 syntax of a command specification is given by:
16463 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
16466 @var{FLAGS} is optional and specifies the type of events relating to
16467 the time interval which enable sending the specified command, and must
16468 be a non-null sequence of identifier flags separated by "+" or "|" and
16469 enclosed between "[" and "]".
16471 The following flags are recognized:
16474 The command is sent when the current frame timestamp enters the
16475 specified interval. In other words, the command is sent when the
16476 previous frame timestamp was not in the given interval, and the
16480 The command is sent when the current frame timestamp leaves the
16481 specified interval. In other words, the command is sent when the
16482 previous frame timestamp was in the given interval, and the
16486 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
16489 @var{TARGET} specifies the target of the command, usually the name of
16490 the filter class or a specific filter instance name.
16492 @var{COMMAND} specifies the name of the command for the target filter.
16494 @var{ARG} is optional and specifies the optional list of argument for
16495 the given @var{COMMAND}.
16497 Between one interval specification and another, whitespaces, or
16498 sequences of characters starting with @code{#} until the end of line,
16499 are ignored and can be used to annotate comments.
16501 A simplified BNF description of the commands specification syntax
16504 @var{COMMAND_FLAG} ::= "enter" | "leave"
16505 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
16506 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
16507 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
16508 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
16509 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
16512 @subsection Examples
16516 Specify audio tempo change at second 4:
16518 asendcmd=c='4.0 atempo tempo 1.5',atempo
16522 Specify a list of drawtext and hue commands in a file.
16524 # show text in the interval 5-10
16525 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
16526 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16528 # desaturate the image in the interval 15-20
16529 15.0-20.0 [enter] hue s 0,
16530 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16532 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16534 # apply an exponential saturation fade-out effect, starting from time 25
16535 25 [enter] hue s exp(25-t)
16538 A filtergraph allowing to read and process the above command list
16539 stored in a file @file{test.cmd}, can be specified with:
16541 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16546 @section setpts, asetpts
16548 Change the PTS (presentation timestamp) of the input frames.
16550 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16552 This filter accepts the following options:
16557 The expression which is evaluated for each frame to construct its timestamp.
16561 The expression is evaluated through the eval API and can contain the following
16566 frame rate, only defined for constant frame-rate video
16569 The presentation timestamp in input
16572 The count of the input frame for video or the number of consumed samples,
16573 not including the current frame for audio, starting from 0.
16575 @item NB_CONSUMED_SAMPLES
16576 The number of consumed samples, not including the current frame (only
16579 @item NB_SAMPLES, S
16580 The number of samples in the current frame (only audio)
16582 @item SAMPLE_RATE, SR
16583 The audio sample rate.
16586 The PTS of the first frame.
16589 the time in seconds of the first frame
16592 State whether the current frame is interlaced.
16595 the time in seconds of the current frame
16598 original position in the file of the frame, or undefined if undefined
16599 for the current frame
16602 The previous input PTS.
16605 previous input time in seconds
16608 The previous output PTS.
16611 previous output time in seconds
16614 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16618 The wallclock (RTC) time at the start of the movie in microseconds.
16621 The timebase of the input timestamps.
16625 @subsection Examples
16629 Start counting PTS from zero
16631 setpts=PTS-STARTPTS
16635 Apply fast motion effect:
16641 Apply slow motion effect:
16647 Set fixed rate of 25 frames per second:
16653 Set fixed rate 25 fps with some jitter:
16655 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16659 Apply an offset of 10 seconds to the input PTS:
16665 Generate timestamps from a "live source" and rebase onto the current timebase:
16667 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16671 Generate timestamps by counting samples:
16678 @section settb, asettb
16680 Set the timebase to use for the output frames timestamps.
16681 It is mainly useful for testing timebase configuration.
16683 It accepts the following parameters:
16688 The expression which is evaluated into the output timebase.
16692 The value for @option{tb} is an arithmetic expression representing a
16693 rational. The expression can contain the constants "AVTB" (the default
16694 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16695 audio only). Default value is "intb".
16697 @subsection Examples
16701 Set the timebase to 1/25:
16707 Set the timebase to 1/10:
16713 Set the timebase to 1001/1000:
16719 Set the timebase to 2*intb:
16725 Set the default timebase value:
16732 Convert input audio to a video output representing frequency spectrum
16733 logarithmically using Brown-Puckette constant Q transform algorithm with
16734 direct frequency domain coefficient calculation (but the transform itself
16735 is not really constant Q, instead the Q factor is actually variable/clamped),
16736 with musical tone scale, from E0 to D#10.
16738 The filter accepts the following options:
16742 Specify the video size for the output. It must be even. For the syntax of this option,
16743 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16744 Default value is @code{1920x1080}.
16747 Set the output frame rate. Default value is @code{25}.
16750 Set the bargraph height. It must be even. Default value is @code{-1} which
16751 computes the bargraph height automatically.
16754 Set the axis height. It must be even. Default value is @code{-1} which computes
16755 the axis height automatically.
16758 Set the sonogram height. It must be even. Default value is @code{-1} which
16759 computes the sonogram height automatically.
16762 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16763 instead. Default value is @code{1}.
16765 @item sono_v, volume
16766 Specify the sonogram volume expression. It can contain variables:
16769 the @var{bar_v} evaluated expression
16770 @item frequency, freq, f
16771 the frequency where it is evaluated
16772 @item timeclamp, tc
16773 the value of @var{timeclamp} option
16777 @item a_weighting(f)
16778 A-weighting of equal loudness
16779 @item b_weighting(f)
16780 B-weighting of equal loudness
16781 @item c_weighting(f)
16782 C-weighting of equal loudness.
16784 Default value is @code{16}.
16786 @item bar_v, volume2
16787 Specify the bargraph volume expression. It can contain variables:
16790 the @var{sono_v} evaluated expression
16791 @item frequency, freq, f
16792 the frequency where it is evaluated
16793 @item timeclamp, tc
16794 the value of @var{timeclamp} option
16798 @item a_weighting(f)
16799 A-weighting of equal loudness
16800 @item b_weighting(f)
16801 B-weighting of equal loudness
16802 @item c_weighting(f)
16803 C-weighting of equal loudness.
16805 Default value is @code{sono_v}.
16807 @item sono_g, gamma
16808 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16809 higher gamma makes the spectrum having more range. Default value is @code{3}.
16810 Acceptable range is @code{[1, 7]}.
16812 @item bar_g, gamma2
16813 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16816 @item timeclamp, tc
16817 Specify the transform timeclamp. At low frequency, there is trade-off between
16818 accuracy in time domain and frequency domain. If timeclamp is lower,
16819 event in time domain is represented more accurately (such as fast bass drum),
16820 otherwise event in frequency domain is represented more accurately
16821 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16824 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16825 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16828 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16829 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16832 This option is deprecated and ignored.
16835 Specify the transform length in time domain. Use this option to control accuracy
16836 trade-off between time domain and frequency domain at every frequency sample.
16837 It can contain variables:
16839 @item frequency, freq, f
16840 the frequency where it is evaluated
16841 @item timeclamp, tc
16842 the value of @var{timeclamp} option.
16844 Default value is @code{384*tc/(384+tc*f)}.
16847 Specify the transform count for every video frame. Default value is @code{6}.
16848 Acceptable range is @code{[1, 30]}.
16851 Specify the transform count for every single pixel. Default value is @code{0},
16852 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16855 Specify font file for use with freetype to draw the axis. If not specified,
16856 use embedded font. Note that drawing with font file or embedded font is not
16857 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16861 Specify font color expression. This is arithmetic expression that should return
16862 integer value 0xRRGGBB. It can contain variables:
16864 @item frequency, freq, f
16865 the frequency where it is evaluated
16866 @item timeclamp, tc
16867 the value of @var{timeclamp} option
16872 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16873 @item r(x), g(x), b(x)
16874 red, green, and blue value of intensity x.
16876 Default value is @code{st(0, (midi(f)-59.5)/12);
16877 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16878 r(1-ld(1)) + b(ld(1))}.
16881 Specify image file to draw the axis. This option override @var{fontfile} and
16882 @var{fontcolor} option.
16885 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16886 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16887 Default value is @code{1}.
16890 Set colorspace. The accepted values are:
16893 Unspecified (default)
16902 BT.470BG or BT.601-6 625
16905 SMPTE-170M or BT.601-6 525
16911 BT.2020 with non-constant luminance
16917 @subsection Examples
16921 Playing audio while showing the spectrum:
16923 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16927 Same as above, but with frame rate 30 fps:
16929 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16933 Playing at 1280x720:
16935 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16939 Disable sonogram display:
16945 A1 and its harmonics: A1, A2, (near)E3, A3:
16947 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),
16948 asplit[a][out1]; [a] showcqt [out0]'
16952 Same as above, but with more accuracy in frequency domain:
16954 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),
16955 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16961 bar_v=10:sono_v=bar_v*a_weighting(f)
16965 Custom gamma, now spectrum is linear to the amplitude.
16971 Custom tlength equation:
16973 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)))'
16977 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
16979 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
16983 Custom frequency range with custom axis using image file:
16985 axisfile=myaxis.png:basefreq=40:endfreq=10000
16991 Convert input audio to video output representing the audio power spectrum.
16992 Audio amplitude is on Y-axis while frequency is on X-axis.
16994 The filter accepts the following options:
16998 Specify size of video. For the syntax of this option, check the
16999 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17000 Default is @code{1024x512}.
17004 This set how each frequency bin will be represented.
17006 It accepts the following values:
17012 Default is @code{bar}.
17015 Set amplitude scale.
17017 It accepts the following values:
17031 Default is @code{log}.
17034 Set frequency scale.
17036 It accepts the following values:
17045 Reverse logarithmic scale.
17047 Default is @code{lin}.
17052 It accepts the following values:
17068 Default is @code{w2048}
17071 Set windowing function.
17073 It accepts the following values:
17095 Default is @code{hanning}.
17098 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17099 which means optimal overlap for selected window function will be picked.
17102 Set time averaging. Setting this to 0 will display current maximal peaks.
17103 Default is @code{1}, which means time averaging is disabled.
17106 Specify list of colors separated by space or by '|' which will be used to
17107 draw channel frequencies. Unrecognized or missing colors will be replaced
17111 Set channel display mode.
17113 It accepts the following values:
17118 Default is @code{combined}.
17121 Set minimum amplitude used in @code{log} amplitude scaler.
17125 @anchor{showspectrum}
17126 @section showspectrum
17128 Convert input audio to a video output, representing the audio frequency
17131 The filter accepts the following options:
17135 Specify the video size for the output. For the syntax of this option, check the
17136 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17137 Default value is @code{640x512}.
17140 Specify how the spectrum should slide along the window.
17142 It accepts the following values:
17145 the samples start again on the left when they reach the right
17147 the samples scroll from right to left
17149 frames are only produced when the samples reach the right
17151 the samples scroll from left to right
17154 Default value is @code{replace}.
17157 Specify display mode.
17159 It accepts the following values:
17162 all channels are displayed in the same row
17164 all channels are displayed in separate rows
17167 Default value is @samp{combined}.
17170 Specify display color mode.
17172 It accepts the following values:
17175 each channel is displayed in a separate color
17177 each channel is displayed using the same color scheme
17179 each channel is displayed using the rainbow color scheme
17181 each channel is displayed using the moreland color scheme
17183 each channel is displayed using the nebulae color scheme
17185 each channel is displayed using the fire color scheme
17187 each channel is displayed using the fiery color scheme
17189 each channel is displayed using the fruit color scheme
17191 each channel is displayed using the cool color scheme
17194 Default value is @samp{channel}.
17197 Specify scale used for calculating intensity color values.
17199 It accepts the following values:
17204 square root, default
17215 Default value is @samp{sqrt}.
17218 Set saturation modifier for displayed colors. Negative values provide
17219 alternative color scheme. @code{0} is no saturation at all.
17220 Saturation must be in [-10.0, 10.0] range.
17221 Default value is @code{1}.
17224 Set window function.
17226 It accepts the following values:
17250 Default value is @code{hann}.
17253 Set orientation of time vs frequency axis. Can be @code{vertical} or
17254 @code{horizontal}. Default is @code{vertical}.
17257 Set ratio of overlap window. Default value is @code{0}.
17258 When value is @code{1} overlap is set to recommended size for specific
17259 window function currently used.
17262 Set scale gain for calculating intensity color values.
17263 Default value is @code{1}.
17266 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
17269 Set color rotation, must be in [-1.0, 1.0] range.
17270 Default value is @code{0}.
17273 The usage is very similar to the showwaves filter; see the examples in that
17276 @subsection Examples
17280 Large window with logarithmic color scaling:
17282 showspectrum=s=1280x480:scale=log
17286 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
17288 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17289 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
17293 @section showspectrumpic
17295 Convert input audio to a single video frame, representing the audio frequency
17298 The filter accepts the following options:
17302 Specify the video size for the output. For the syntax of this option, check the
17303 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17304 Default value is @code{4096x2048}.
17307 Specify display mode.
17309 It accepts the following values:
17312 all channels are displayed in the same row
17314 all channels are displayed in separate rows
17316 Default value is @samp{combined}.
17319 Specify display color mode.
17321 It accepts the following values:
17324 each channel is displayed in a separate color
17326 each channel is displayed using the same color scheme
17328 each channel is displayed using the rainbow color scheme
17330 each channel is displayed using the moreland color scheme
17332 each channel is displayed using the nebulae color scheme
17334 each channel is displayed using the fire color scheme
17336 each channel is displayed using the fiery color scheme
17338 each channel is displayed using the fruit color scheme
17340 each channel is displayed using the cool color scheme
17342 Default value is @samp{intensity}.
17345 Specify scale used for calculating intensity color values.
17347 It accepts the following values:
17352 square root, default
17362 Default value is @samp{log}.
17365 Set saturation modifier for displayed colors. Negative values provide
17366 alternative color scheme. @code{0} is no saturation at all.
17367 Saturation must be in [-10.0, 10.0] range.
17368 Default value is @code{1}.
17371 Set window function.
17373 It accepts the following values:
17396 Default value is @code{hann}.
17399 Set orientation of time vs frequency axis. Can be @code{vertical} or
17400 @code{horizontal}. Default is @code{vertical}.
17403 Set scale gain for calculating intensity color values.
17404 Default value is @code{1}.
17407 Draw time and frequency axes and legends. Default is enabled.
17410 Set color rotation, must be in [-1.0, 1.0] range.
17411 Default value is @code{0}.
17414 @subsection Examples
17418 Extract an audio spectrogram of a whole audio track
17419 in a 1024x1024 picture using @command{ffmpeg}:
17421 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
17425 @section showvolume
17427 Convert input audio volume to a video output.
17429 The filter accepts the following options:
17436 Set border width, allowed range is [0, 5]. Default is 1.
17439 Set channel width, allowed range is [80, 8192]. Default is 400.
17442 Set channel height, allowed range is [1, 900]. Default is 20.
17445 Set fade, allowed range is [0.001, 1]. Default is 0.95.
17448 Set volume color expression.
17450 The expression can use the following variables:
17454 Current max volume of channel in dB.
17460 Current channel number, starting from 0.
17464 If set, displays channel names. Default is enabled.
17467 If set, displays volume values. Default is enabled.
17470 Set orientation, can be @code{horizontal} or @code{vertical},
17471 default is @code{horizontal}.
17474 Set step size, allowed range s [0, 5]. Default is 0, which means
17480 Convert input audio to a video output, representing the samples waves.
17482 The filter accepts the following options:
17486 Specify the video size for the output. For the syntax of this option, check the
17487 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17488 Default value is @code{600x240}.
17493 Available values are:
17496 Draw a point for each sample.
17499 Draw a vertical line for each sample.
17502 Draw a point for each sample and a line between them.
17505 Draw a centered vertical line for each sample.
17508 Default value is @code{point}.
17511 Set the number of samples which are printed on the same column. A
17512 larger value will decrease the frame rate. Must be a positive
17513 integer. This option can be set only if the value for @var{rate}
17514 is not explicitly specified.
17517 Set the (approximate) output frame rate. This is done by setting the
17518 option @var{n}. Default value is "25".
17520 @item split_channels
17521 Set if channels should be drawn separately or overlap. Default value is 0.
17524 Set colors separated by '|' which are going to be used for drawing of each channel.
17527 Set amplitude scale.
17529 Available values are:
17547 @subsection Examples
17551 Output the input file audio and the corresponding video representation
17554 amovie=a.mp3,asplit[out0],showwaves[out1]
17558 Create a synthetic signal and show it with showwaves, forcing a
17559 frame rate of 30 frames per second:
17561 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17565 @section showwavespic
17567 Convert input audio to a single video frame, representing the samples waves.
17569 The filter accepts the following options:
17573 Specify the video size for the output. For the syntax of this option, check the
17574 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17575 Default value is @code{600x240}.
17577 @item split_channels
17578 Set if channels should be drawn separately or overlap. Default value is 0.
17581 Set colors separated by '|' which are going to be used for drawing of each channel.
17584 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17588 @subsection Examples
17592 Extract a channel split representation of the wave form of a whole audio track
17593 in a 1024x800 picture using @command{ffmpeg}:
17595 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17599 @section sidedata, asidedata
17601 Delete frame side data, or select frames based on it.
17603 This filter accepts the following options:
17607 Set mode of operation of the filter.
17609 Can be one of the following:
17613 Select every frame with side data of @code{type}.
17616 Delete side data of @code{type}. If @code{type} is not set, delete all side
17622 Set side data type used with all modes. Must be set for @code{select} mode. For
17623 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
17624 in @file{libavutil/frame.h}. For example, to choose
17625 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
17629 @section spectrumsynth
17631 Sythesize audio from 2 input video spectrums, first input stream represents
17632 magnitude across time and second represents phase across time.
17633 The filter will transform from frequency domain as displayed in videos back
17634 to time domain as presented in audio output.
17636 This filter is primarly created for reversing processed @ref{showspectrum}
17637 filter outputs, but can synthesize sound from other spectrograms too.
17638 But in such case results are going to be poor if the phase data is not
17639 available, because in such cases phase data need to be recreated, usually
17640 its just recreated from random noise.
17641 For best results use gray only output (@code{channel} color mode in
17642 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17643 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17644 @code{data} option. Inputs videos should generally use @code{fullframe}
17645 slide mode as that saves resources needed for decoding video.
17647 The filter accepts the following options:
17651 Specify sample rate of output audio, the sample rate of audio from which
17652 spectrum was generated may differ.
17655 Set number of channels represented in input video spectrums.
17658 Set scale which was used when generating magnitude input spectrum.
17659 Can be @code{lin} or @code{log}. Default is @code{log}.
17662 Set slide which was used when generating inputs spectrums.
17663 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17664 Default is @code{fullframe}.
17667 Set window function used for resynthesis.
17670 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17671 which means optimal overlap for selected window function will be picked.
17674 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17675 Default is @code{vertical}.
17678 @subsection Examples
17682 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17683 then resynthesize videos back to audio with spectrumsynth:
17685 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
17686 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
17687 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17691 @section split, asplit
17693 Split input into several identical outputs.
17695 @code{asplit} works with audio input, @code{split} with video.
17697 The filter accepts a single parameter which specifies the number of outputs. If
17698 unspecified, it defaults to 2.
17700 @subsection Examples
17704 Create two separate outputs from the same input:
17706 [in] split [out0][out1]
17710 To create 3 or more outputs, you need to specify the number of
17713 [in] asplit=3 [out0][out1][out2]
17717 Create two separate outputs from the same input, one cropped and
17720 [in] split [splitout1][splitout2];
17721 [splitout1] crop=100:100:0:0 [cropout];
17722 [splitout2] pad=200:200:100:100 [padout];
17726 Create 5 copies of the input audio with @command{ffmpeg}:
17728 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17734 Receive commands sent through a libzmq client, and forward them to
17735 filters in the filtergraph.
17737 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17738 must be inserted between two video filters, @code{azmq} between two
17741 To enable these filters you need to install the libzmq library and
17742 headers and configure FFmpeg with @code{--enable-libzmq}.
17744 For more information about libzmq see:
17745 @url{http://www.zeromq.org/}
17747 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17748 receives messages sent through a network interface defined by the
17749 @option{bind_address} option.
17751 The received message must be in the form:
17753 @var{TARGET} @var{COMMAND} [@var{ARG}]
17756 @var{TARGET} specifies the target of the command, usually the name of
17757 the filter class or a specific filter instance name.
17759 @var{COMMAND} specifies the name of the command for the target filter.
17761 @var{ARG} is optional and specifies the optional argument list for the
17762 given @var{COMMAND}.
17764 Upon reception, the message is processed and the corresponding command
17765 is injected into the filtergraph. Depending on the result, the filter
17766 will send a reply to the client, adopting the format:
17768 @var{ERROR_CODE} @var{ERROR_REASON}
17772 @var{MESSAGE} is optional.
17774 @subsection Examples
17776 Look at @file{tools/zmqsend} for an example of a zmq client which can
17777 be used to send commands processed by these filters.
17779 Consider the following filtergraph generated by @command{ffplay}
17781 ffplay -dumpgraph 1 -f lavfi "
17782 color=s=100x100:c=red [l];
17783 color=s=100x100:c=blue [r];
17784 nullsrc=s=200x100, zmq [bg];
17785 [bg][l] overlay [bg+l];
17786 [bg+l][r] overlay=x=100 "
17789 To change the color of the left side of the video, the following
17790 command can be used:
17792 echo Parsed_color_0 c yellow | tools/zmqsend
17795 To change the right side:
17797 echo Parsed_color_1 c pink | tools/zmqsend
17800 @c man end MULTIMEDIA FILTERS
17802 @chapter Multimedia Sources
17803 @c man begin MULTIMEDIA SOURCES
17805 Below is a description of the currently available multimedia sources.
17809 This is the same as @ref{movie} source, except it selects an audio
17815 Read audio and/or video stream(s) from a movie container.
17817 It accepts the following parameters:
17821 The name of the resource to read (not necessarily a file; it can also be a
17822 device or a stream accessed through some protocol).
17824 @item format_name, f
17825 Specifies the format assumed for the movie to read, and can be either
17826 the name of a container or an input device. If not specified, the
17827 format is guessed from @var{movie_name} or by probing.
17829 @item seek_point, sp
17830 Specifies the seek point in seconds. The frames will be output
17831 starting from this seek point. The parameter is evaluated with
17832 @code{av_strtod}, so the numerical value may be suffixed by an IS
17833 postfix. The default value is "0".
17836 Specifies the streams to read. Several streams can be specified,
17837 separated by "+". The source will then have as many outputs, in the
17838 same order. The syntax is explained in the ``Stream specifiers''
17839 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17840 respectively the default (best suited) video and audio stream. Default
17841 is "dv", or "da" if the filter is called as "amovie".
17843 @item stream_index, si
17844 Specifies the index of the video stream to read. If the value is -1,
17845 the most suitable video stream will be automatically selected. The default
17846 value is "-1". Deprecated. If the filter is called "amovie", it will select
17847 audio instead of video.
17850 Specifies how many times to read the stream in sequence.
17851 If the value is less than 1, the stream will be read again and again.
17852 Default value is "1".
17854 Note that when the movie is looped the source timestamps are not
17855 changed, so it will generate non monotonically increasing timestamps.
17857 @item discontinuity
17858 Specifies the time difference between frames above which the point is
17859 considered a timestamp discontinuity which is removed by adjusting the later
17863 It allows overlaying a second video on top of the main input of
17864 a filtergraph, as shown in this graph:
17866 input -----------> deltapts0 --> overlay --> output
17869 movie --> scale--> deltapts1 -------+
17871 @subsection Examples
17875 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17876 on top of the input labelled "in":
17878 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17879 [in] setpts=PTS-STARTPTS [main];
17880 [main][over] overlay=16:16 [out]
17884 Read from a video4linux2 device, and overlay it on top of the input
17887 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17888 [in] setpts=PTS-STARTPTS [main];
17889 [main][over] overlay=16:16 [out]
17893 Read the first video stream and the audio stream with id 0x81 from
17894 dvd.vob; the video is connected to the pad named "video" and the audio is
17895 connected to the pad named "audio":
17897 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17901 @subsection Commands
17903 Both movie and amovie support the following commands:
17906 Perform seek using "av_seek_frame".
17907 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17910 @var{stream_index}: If stream_index is -1, a default
17911 stream is selected, and @var{timestamp} is automatically converted
17912 from AV_TIME_BASE units to the stream specific time_base.
17914 @var{timestamp}: Timestamp in AVStream.time_base units
17915 or, if no stream is specified, in AV_TIME_BASE units.
17917 @var{flags}: Flags which select direction and seeking mode.
17921 Get movie duration in AV_TIME_BASE units.
17925 @c man end MULTIMEDIA SOURCES