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
2555 If enabled, use fixed number of audio samples. This improves speed when
2556 filtering with large delay. Default is disabled.
2559 Enable multichannels evaluation on gain. Default is disabled.
2562 Enable zero phase mode by subtracting timestamp to compensate delay.
2563 Default is disabled.
2566 @subsection Examples
2571 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2574 lowpass at 1000 Hz with gain_entry:
2576 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2579 custom equalization:
2581 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2584 higher delay with zero phase to compensate delay:
2586 firequalizer=delay=0.1:fixed=on:zero_phase=on
2589 lowpass on left channel, highpass on right channel:
2591 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2592 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2597 Apply a flanging effect to the audio.
2599 The filter accepts the following options:
2603 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2606 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2609 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2613 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2614 Default value is 71.
2617 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2620 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2621 Default value is @var{sinusoidal}.
2624 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2625 Default value is 25.
2628 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2629 Default is @var{linear}.
2634 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2635 embedded HDCD codes is expanded into a 20-bit PCM stream.
2637 The filter supports the Peak Extend and Low-level Gain Adjustment features
2638 of HDCD, and detects the Transient Filter flag.
2641 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2644 When using the filter with wav, note the default encoding for wav is 16-bit,
2645 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2646 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2648 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2649 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2652 The filter accepts the following options:
2655 @item disable_autoconvert
2656 Disable any automatic format conversion or resampling in the filter graph.
2658 @item process_stereo
2659 Process the stereo channels together. If target_gain does not match between
2660 channels, consider it invalid and use the last valid target_gain.
2663 Set the code detect timer period in ms.
2666 Always extend peaks above -3dBFS even if PE isn't signaled.
2669 Replace audio with a solid tone and adjust the amplitude to signal some
2670 specific aspect of the decoding process. The output file can be loaded in
2671 an audio editor alongside the original to aid analysis.
2673 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2680 Gain adjustment level at each sample
2682 Samples where peak extend occurs
2684 Samples where the code detect timer is active
2686 Samples where the target gain does not match between channels
2692 Apply a high-pass filter with 3dB point frequency.
2693 The filter can be either single-pole, or double-pole (the default).
2694 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2696 The filter accepts the following options:
2700 Set frequency in Hz. Default is 3000.
2703 Set number of poles. Default is 2.
2706 Set method to specify band-width of filter.
2719 Specify the band-width of a filter in width_type units.
2720 Applies only to double-pole filter.
2721 The default is 0.707q and gives a Butterworth response.
2726 Join multiple input streams into one multi-channel stream.
2728 It accepts the following parameters:
2732 The number of input streams. It defaults to 2.
2734 @item channel_layout
2735 The desired output channel layout. It defaults to stereo.
2738 Map channels from inputs to output. The argument is a '|'-separated list of
2739 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2740 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2741 can be either the name of the input channel (e.g. FL for front left) or its
2742 index in the specified input stream. @var{out_channel} is the name of the output
2746 The filter will attempt to guess the mappings when they are not specified
2747 explicitly. It does so by first trying to find an unused matching input channel
2748 and if that fails it picks the first unused input channel.
2750 Join 3 inputs (with properly set channel layouts):
2752 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2755 Build a 5.1 output from 6 single-channel streams:
2757 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2758 'join=inputs=6:channel_layout=5.1:map=0.0-FL|1.0-FR|2.0-FC|3.0-SL|4.0-SR|5.0-LFE'
2764 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2766 To enable compilation of this filter you need to configure FFmpeg with
2767 @code{--enable-ladspa}.
2771 Specifies the name of LADSPA plugin library to load. If the environment
2772 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2773 each one of the directories specified by the colon separated list in
2774 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2775 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2776 @file{/usr/lib/ladspa/}.
2779 Specifies the plugin within the library. Some libraries contain only
2780 one plugin, but others contain many of them. If this is not set filter
2781 will list all available plugins within the specified library.
2784 Set the '|' separated list of controls which are zero or more floating point
2785 values that determine the behavior of the loaded plugin (for example delay,
2787 Controls need to be defined using the following syntax:
2788 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2789 @var{valuei} is the value set on the @var{i}-th control.
2790 Alternatively they can be also defined using the following syntax:
2791 @var{value0}|@var{value1}|@var{value2}|..., where
2792 @var{valuei} is the value set on the @var{i}-th control.
2793 If @option{controls} is set to @code{help}, all available controls and
2794 their valid ranges are printed.
2796 @item sample_rate, s
2797 Specify the sample rate, default to 44100. Only used if plugin have
2801 Set the number of samples per channel per each output frame, default
2802 is 1024. Only used if plugin have zero inputs.
2805 Set the minimum duration of the sourced audio. See
2806 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2807 for the accepted syntax.
2808 Note that the resulting duration may be greater than the specified duration,
2809 as the generated audio is always cut at the end of a complete frame.
2810 If not specified, or the expressed duration is negative, the audio is
2811 supposed to be generated forever.
2812 Only used if plugin have zero inputs.
2816 @subsection Examples
2820 List all available plugins within amp (LADSPA example plugin) library:
2826 List all available controls and their valid ranges for @code{vcf_notch}
2827 plugin from @code{VCF} library:
2829 ladspa=f=vcf:p=vcf_notch:c=help
2833 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2836 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2840 Add reverberation to the audio using TAP-plugins
2841 (Tom's Audio Processing plugins):
2843 ladspa=file=tap_reverb:tap_reverb
2847 Generate white noise, with 0.2 amplitude:
2849 ladspa=file=cmt:noise_source_white:c=c0=.2
2853 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2854 @code{C* Audio Plugin Suite} (CAPS) library:
2856 ladspa=file=caps:Click:c=c1=20'
2860 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2862 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2866 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2867 @code{SWH Plugins} collection:
2869 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2873 Attenuate low frequencies using Multiband EQ from Steve Harris
2874 @code{SWH Plugins} collection:
2876 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2880 @subsection Commands
2882 This filter supports the following commands:
2885 Modify the @var{N}-th control value.
2887 If the specified value is not valid, it is ignored and prior one is kept.
2892 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2893 Support for both single pass (livestreams, files) and double pass (files) modes.
2894 This algorithm can target IL, LRA, and maximum true peak.
2896 To enable compilation of this filter you need to configure FFmpeg with
2897 @code{--enable-libebur128}.
2899 The filter accepts the following options:
2903 Set integrated loudness target.
2904 Range is -70.0 - -5.0. Default value is -24.0.
2907 Set loudness range target.
2908 Range is 1.0 - 20.0. Default value is 7.0.
2911 Set maximum true peak.
2912 Range is -9.0 - +0.0. Default value is -2.0.
2914 @item measured_I, measured_i
2915 Measured IL of input file.
2916 Range is -99.0 - +0.0.
2918 @item measured_LRA, measured_lra
2919 Measured LRA of input file.
2920 Range is 0.0 - 99.0.
2922 @item measured_TP, measured_tp
2923 Measured true peak of input file.
2924 Range is -99.0 - +99.0.
2926 @item measured_thresh
2927 Measured threshold of input file.
2928 Range is -99.0 - +0.0.
2931 Set offset gain. Gain is applied before the true-peak limiter.
2932 Range is -99.0 - +99.0. Default is +0.0.
2935 Normalize linearly if possible.
2936 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2937 to be specified in order to use this mode.
2938 Options are true or false. Default is true.
2941 Treat mono input files as "dual-mono". If a mono file is intended for playback
2942 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2943 If set to @code{true}, this option will compensate for this effect.
2944 Multi-channel input files are not affected by this option.
2945 Options are true or false. Default is false.
2948 Set print format for stats. Options are summary, json, or none.
2949 Default value is none.
2954 Apply a low-pass filter with 3dB point frequency.
2955 The filter can be either single-pole or double-pole (the default).
2956 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2958 The filter accepts the following options:
2962 Set frequency in Hz. Default is 500.
2965 Set number of poles. Default is 2.
2968 Set method to specify band-width of filter.
2981 Specify the band-width of a filter in width_type units.
2982 Applies only to double-pole filter.
2983 The default is 0.707q and gives a Butterworth response.
2989 Mix channels with specific gain levels. The filter accepts the output
2990 channel layout followed by a set of channels definitions.
2992 This filter is also designed to efficiently remap the channels of an audio
2995 The filter accepts parameters of the form:
2996 "@var{l}|@var{outdef}|@var{outdef}|..."
3000 output channel layout or number of channels
3003 output channel specification, of the form:
3004 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
3007 output channel to define, either a channel name (FL, FR, etc.) or a channel
3008 number (c0, c1, etc.)
3011 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3014 input channel to use, see out_name for details; it is not possible to mix
3015 named and numbered input channels
3018 If the `=' in a channel specification is replaced by `<', then the gains for
3019 that specification will be renormalized so that the total is 1, thus
3020 avoiding clipping noise.
3022 @subsection Mixing examples
3024 For example, if you want to down-mix from stereo to mono, but with a bigger
3025 factor for the left channel:
3027 pan=1c|c0=0.9*c0+0.1*c1
3030 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3031 7-channels surround:
3033 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3036 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3037 that should be preferred (see "-ac" option) unless you have very specific
3040 @subsection Remapping examples
3042 The channel remapping will be effective if, and only if:
3045 @item gain coefficients are zeroes or ones,
3046 @item only one input per channel output,
3049 If all these conditions are satisfied, the filter will notify the user ("Pure
3050 channel mapping detected"), and use an optimized and lossless method to do the
3053 For example, if you have a 5.1 source and want a stereo audio stream by
3054 dropping the extra channels:
3056 pan="stereo| c0=FL | c1=FR"
3059 Given the same source, you can also switch front left and front right channels
3060 and keep the input channel layout:
3062 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3065 If the input is a stereo audio stream, you can mute the front left channel (and
3066 still keep the stereo channel layout) with:
3071 Still with a stereo audio stream input, you can copy the right channel in both
3072 front left and right:
3074 pan="stereo| c0=FR | c1=FR"
3079 ReplayGain scanner filter. This filter takes an audio stream as an input and
3080 outputs it unchanged.
3081 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3085 Convert the audio sample format, sample rate and channel layout. It is
3086 not meant to be used directly.
3089 Apply time-stretching and pitch-shifting with librubberband.
3091 The filter accepts the following options:
3095 Set tempo scale factor.
3098 Set pitch scale factor.
3101 Set transients detector.
3102 Possible values are:
3111 Possible values are:
3120 Possible values are:
3127 Set processing window size.
3128 Possible values are:
3137 Possible values are:
3144 Enable formant preservation when shift pitching.
3145 Possible values are:
3153 Possible values are:
3162 Possible values are:
3169 @section sidechaincompress
3171 This filter acts like normal compressor but has the ability to compress
3172 detected signal using second input signal.
3173 It needs two input streams and returns one output stream.
3174 First input stream will be processed depending on second stream signal.
3175 The filtered signal then can be filtered with other filters in later stages of
3176 processing. See @ref{pan} and @ref{amerge} filter.
3178 The filter accepts the following options:
3182 Set input gain. Default is 1. Range is between 0.015625 and 64.
3185 If a signal of second stream raises above this level it will affect the gain
3186 reduction of first stream.
3187 By default is 0.125. Range is between 0.00097563 and 1.
3190 Set a ratio about which the signal is reduced. 1:2 means that if the level
3191 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3192 Default is 2. Range is between 1 and 20.
3195 Amount of milliseconds the signal has to rise above the threshold before gain
3196 reduction starts. Default is 20. Range is between 0.01 and 2000.
3199 Amount of milliseconds the signal has to fall below the threshold before
3200 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3203 Set the amount by how much signal will be amplified after processing.
3204 Default is 2. Range is from 1 and 64.
3207 Curve the sharp knee around the threshold to enter gain reduction more softly.
3208 Default is 2.82843. Range is between 1 and 8.
3211 Choose if the @code{average} level between all channels of side-chain stream
3212 or the louder(@code{maximum}) channel of side-chain stream affects the
3213 reduction. Default is @code{average}.
3216 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3217 of @code{rms}. Default is @code{rms} which is mainly smoother.
3220 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3223 How much to use compressed signal in output. Default is 1.
3224 Range is between 0 and 1.
3227 @subsection Examples
3231 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3232 depending on the signal of 2nd input and later compressed signal to be
3233 merged with 2nd input:
3235 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3239 @section sidechaingate
3241 A sidechain gate acts like a normal (wideband) gate but has the ability to
3242 filter the detected signal before sending it to the gain reduction stage.
3243 Normally a gate uses the full range signal to detect a level above the
3245 For example: If you cut all lower frequencies from your sidechain signal
3246 the gate will decrease the volume of your track only if not enough highs
3247 appear. With this technique you are able to reduce the resonation of a
3248 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3250 It needs two input streams and returns one output stream.
3251 First input stream will be processed depending on second stream signal.
3253 The filter accepts the following options:
3257 Set input level before filtering.
3258 Default is 1. Allowed range is from 0.015625 to 64.
3261 Set the level of gain reduction when the signal is below the threshold.
3262 Default is 0.06125. Allowed range is from 0 to 1.
3265 If a signal rises above this level the gain reduction is released.
3266 Default is 0.125. Allowed range is from 0 to 1.
3269 Set a ratio about which the signal is reduced.
3270 Default is 2. Allowed range is from 1 to 9000.
3273 Amount of milliseconds the signal has to rise above the threshold before gain
3275 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3278 Amount of milliseconds the signal has to fall below the threshold before the
3279 reduction is increased again. Default is 250 milliseconds.
3280 Allowed range is from 0.01 to 9000.
3283 Set amount of amplification of signal after processing.
3284 Default is 1. Allowed range is from 1 to 64.
3287 Curve the sharp knee around the threshold to enter gain reduction more softly.
3288 Default is 2.828427125. Allowed range is from 1 to 8.
3291 Choose if exact signal should be taken for detection or an RMS like one.
3292 Default is rms. Can be peak or rms.
3295 Choose if the average level between all channels or the louder channel affects
3297 Default is average. Can be average or maximum.
3300 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3303 @section silencedetect
3305 Detect silence in an audio stream.
3307 This filter logs a message when it detects that the input audio volume is less
3308 or equal to a noise tolerance value for a duration greater or equal to the
3309 minimum detected noise duration.
3311 The printed times and duration are expressed in seconds.
3313 The filter accepts the following options:
3317 Set silence duration until notification (default is 2 seconds).
3320 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3321 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3324 @subsection Examples
3328 Detect 5 seconds of silence with -50dB noise tolerance:
3330 silencedetect=n=-50dB:d=5
3334 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3335 tolerance in @file{silence.mp3}:
3337 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3341 @section silenceremove
3343 Remove silence from the beginning, middle or end of the audio.
3345 The filter accepts the following options:
3349 This value is used to indicate if audio should be trimmed at beginning of
3350 the audio. A value of zero indicates no silence should be trimmed from the
3351 beginning. When specifying a non-zero value, it trims audio up until it
3352 finds non-silence. Normally, when trimming silence from beginning of audio
3353 the @var{start_periods} will be @code{1} but it can be increased to higher
3354 values to trim all audio up to specific count of non-silence periods.
3355 Default value is @code{0}.
3357 @item start_duration
3358 Specify the amount of time that non-silence must be detected before it stops
3359 trimming audio. By increasing the duration, bursts of noises can be treated
3360 as silence and trimmed off. Default value is @code{0}.
3362 @item start_threshold
3363 This indicates what sample value should be treated as silence. For digital
3364 audio, a value of @code{0} may be fine but for audio recorded from analog,
3365 you may wish to increase the value to account for background noise.
3366 Can be specified in dB (in case "dB" is appended to the specified value)
3367 or amplitude ratio. Default value is @code{0}.
3370 Set the count for trimming silence from the end of audio.
3371 To remove silence from the middle of a file, specify a @var{stop_periods}
3372 that is negative. This value is then treated as a positive value and is
3373 used to indicate the effect should restart processing as specified by
3374 @var{start_periods}, making it suitable for removing periods of silence
3375 in the middle of the audio.
3376 Default value is @code{0}.
3379 Specify a duration of silence that must exist before audio is not copied any
3380 more. By specifying a higher duration, silence that is wanted can be left in
3382 Default value is @code{0}.
3384 @item stop_threshold
3385 This is the same as @option{start_threshold} but for trimming silence from
3387 Can be specified in dB (in case "dB" is appended to the specified value)
3388 or amplitude ratio. Default value is @code{0}.
3391 This indicates that @var{stop_duration} length of audio should be left intact
3392 at the beginning of each period of silence.
3393 For example, if you want to remove long pauses between words but do not want
3394 to remove the pauses completely. Default value is @code{0}.
3397 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3398 and works better with digital silence which is exactly 0.
3399 Default value is @code{rms}.
3402 Set ratio used to calculate size of window for detecting silence.
3403 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3406 @subsection Examples
3410 The following example shows how this filter can be used to start a recording
3411 that does not contain the delay at the start which usually occurs between
3412 pressing the record button and the start of the performance:
3414 silenceremove=1:5:0.02
3418 Trim all silence encountered from beginning to end where there is more than 1
3419 second of silence in audio:
3421 silenceremove=0:0:0:-1:1:-90dB
3427 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3428 loudspeakers around the user for binaural listening via headphones (audio
3429 formats up to 9 channels supported).
3430 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3431 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3432 Austrian Academy of Sciences.
3434 To enable compilation of this filter you need to configure FFmpeg with
3435 @code{--enable-netcdf}.
3437 The filter accepts the following options:
3441 Set the SOFA file used for rendering.
3444 Set gain applied to audio. Value is in dB. Default is 0.
3447 Set rotation of virtual loudspeakers in deg. Default is 0.
3450 Set elevation of virtual speakers in deg. Default is 0.
3453 Set distance in meters between loudspeakers and the listener with near-field
3454 HRTFs. Default is 1.
3457 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3458 processing audio in time domain which is slow.
3459 @var{freq} is processing audio in frequency domain which is fast.
3460 Default is @var{freq}.
3463 Set custom positions of virtual loudspeakers. Syntax for this option is:
3464 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3465 Each virtual loudspeaker is described with short channel name following with
3466 azimuth and elevation in degreees.
3467 Each virtual loudspeaker description is separated by '|'.
3468 For example to override front left and front right channel positions use:
3469 'speakers=FL 45 15|FR 345 15'.
3470 Descriptions with unrecognised channel names are ignored.
3473 @subsection Examples
3477 Using ClubFritz6 sofa file:
3479 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3483 Using ClubFritz12 sofa file and bigger radius with small rotation:
3485 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3489 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3490 and also with custom gain:
3492 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3496 @section stereotools
3498 This filter has some handy utilities to manage stereo signals, for converting
3499 M/S stereo recordings to L/R signal while having control over the parameters
3500 or spreading the stereo image of master track.
3502 The filter accepts the following options:
3506 Set input level before filtering for both channels. Defaults is 1.
3507 Allowed range is from 0.015625 to 64.
3510 Set output level after filtering for both channels. Defaults is 1.
3511 Allowed range is from 0.015625 to 64.
3514 Set input balance between both channels. Default is 0.
3515 Allowed range is from -1 to 1.
3518 Set output balance between both channels. Default is 0.
3519 Allowed range is from -1 to 1.
3522 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3523 clipping. Disabled by default.
3526 Mute the left channel. Disabled by default.
3529 Mute the right channel. Disabled by default.
3532 Change the phase of the left channel. Disabled by default.
3535 Change the phase of the right channel. Disabled by default.
3538 Set stereo mode. Available values are:
3542 Left/Right to Left/Right, this is default.
3545 Left/Right to Mid/Side.
3548 Mid/Side to Left/Right.
3551 Left/Right to Left/Left.
3554 Left/Right to Right/Right.
3557 Left/Right to Left + Right.
3560 Left/Right to Right/Left.
3564 Set level of side signal. Default is 1.
3565 Allowed range is from 0.015625 to 64.
3568 Set balance of side signal. Default is 0.
3569 Allowed range is from -1 to 1.
3572 Set level of the middle signal. Default is 1.
3573 Allowed range is from 0.015625 to 64.
3576 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3579 Set stereo base between mono and inversed channels. Default is 0.
3580 Allowed range is from -1 to 1.
3583 Set delay in milliseconds how much to delay left from right channel and
3584 vice versa. Default is 0. Allowed range is from -20 to 20.
3587 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3590 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3593 @subsection Examples
3597 Apply karaoke like effect:
3599 stereotools=mlev=0.015625
3603 Convert M/S signal to L/R:
3605 "stereotools=mode=ms>lr"
3609 @section stereowiden
3611 This filter enhance the stereo effect by suppressing signal common to both
3612 channels and by delaying the signal of left into right and vice versa,
3613 thereby widening the stereo effect.
3615 The filter accepts the following options:
3619 Time in milliseconds of the delay of left signal into right and vice versa.
3620 Default is 20 milliseconds.
3623 Amount of gain in delayed signal into right and vice versa. Gives a delay
3624 effect of left signal in right output and vice versa which gives widening
3625 effect. Default is 0.3.
3628 Cross feed of left into right with inverted phase. This helps in suppressing
3629 the mono. If the value is 1 it will cancel all the signal common to both
3630 channels. Default is 0.3.
3633 Set level of input signal of original channel. Default is 0.8.
3638 Boost or cut treble (upper) frequencies of the audio using a two-pole
3639 shelving filter with a response similar to that of a standard
3640 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3642 The filter accepts the following options:
3646 Give the gain at whichever is the lower of ~22 kHz and the
3647 Nyquist frequency. Its useful range is about -20 (for a large cut)
3648 to +20 (for a large boost). Beware of clipping when using a positive gain.
3651 Set the filter's central frequency and so can be used
3652 to extend or reduce the frequency range to be boosted or cut.
3653 The default value is @code{3000} Hz.
3656 Set method to specify band-width of filter.
3669 Determine how steep is the filter's shelf transition.
3674 Sinusoidal amplitude modulation.
3676 The filter accepts the following options:
3680 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3681 (20 Hz or lower) will result in a tremolo effect.
3682 This filter may also be used as a ring modulator by specifying
3683 a modulation frequency higher than 20 Hz.
3684 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3687 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3688 Default value is 0.5.
3693 Sinusoidal phase modulation.
3695 The filter accepts the following options:
3699 Modulation frequency in Hertz.
3700 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3703 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3704 Default value is 0.5.
3709 Adjust the input audio volume.
3711 It accepts the following parameters:
3715 Set audio volume expression.
3717 Output values are clipped to the maximum value.
3719 The output audio volume is given by the relation:
3721 @var{output_volume} = @var{volume} * @var{input_volume}
3724 The default value for @var{volume} is "1.0".
3727 This parameter represents the mathematical precision.
3729 It determines which input sample formats will be allowed, which affects the
3730 precision of the volume scaling.
3734 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3736 32-bit floating-point; this limits input sample format to FLT. (default)
3738 64-bit floating-point; this limits input sample format to DBL.
3742 Choose the behaviour on encountering ReplayGain side data in input frames.
3746 Remove ReplayGain side data, ignoring its contents (the default).
3749 Ignore ReplayGain side data, but leave it in the frame.
3752 Prefer the track gain, if present.
3755 Prefer the album gain, if present.
3758 @item replaygain_preamp
3759 Pre-amplification gain in dB to apply to the selected replaygain gain.
3761 Default value for @var{replaygain_preamp} is 0.0.
3764 Set when the volume expression is evaluated.
3766 It accepts the following values:
3769 only evaluate expression once during the filter initialization, or
3770 when the @samp{volume} command is sent
3773 evaluate expression for each incoming frame
3776 Default value is @samp{once}.
3779 The volume expression can contain the following parameters.
3783 frame number (starting at zero)
3786 @item nb_consumed_samples
3787 number of samples consumed by the filter
3789 number of samples in the current frame
3791 original frame position in the file
3797 PTS at start of stream
3799 time at start of stream
3805 last set volume value
3808 Note that when @option{eval} is set to @samp{once} only the
3809 @var{sample_rate} and @var{tb} variables are available, all other
3810 variables will evaluate to NAN.
3812 @subsection Commands
3814 This filter supports the following commands:
3817 Modify the volume expression.
3818 The command accepts the same syntax of the corresponding option.
3820 If the specified expression is not valid, it is kept at its current
3822 @item replaygain_noclip
3823 Prevent clipping by limiting the gain applied.
3825 Default value for @var{replaygain_noclip} is 1.
3829 @subsection Examples
3833 Halve the input audio volume:
3837 volume=volume=-6.0206dB
3840 In all the above example the named key for @option{volume} can be
3841 omitted, for example like in:
3847 Increase input audio power by 6 decibels using fixed-point precision:
3849 volume=volume=6dB:precision=fixed
3853 Fade volume after time 10 with an annihilation period of 5 seconds:
3855 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3859 @section volumedetect
3861 Detect the volume of the input video.
3863 The filter has no parameters. The input is not modified. Statistics about
3864 the volume will be printed in the log when the input stream end is reached.
3866 In particular it will show the mean volume (root mean square), maximum
3867 volume (on a per-sample basis), and the beginning of a histogram of the
3868 registered volume values (from the maximum value to a cumulated 1/1000 of
3871 All volumes are in decibels relative to the maximum PCM value.
3873 @subsection Examples
3875 Here is an excerpt of the output:
3877 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3878 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3879 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3880 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3881 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3882 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3883 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3884 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3885 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3891 The mean square energy is approximately -27 dB, or 10^-2.7.
3893 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3895 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3898 In other words, raising the volume by +4 dB does not cause any clipping,
3899 raising it by +5 dB causes clipping for 6 samples, etc.
3901 @c man end AUDIO FILTERS
3903 @chapter Audio Sources
3904 @c man begin AUDIO SOURCES
3906 Below is a description of the currently available audio sources.
3910 Buffer audio frames, and make them available to the filter chain.
3912 This source is mainly intended for a programmatic use, in particular
3913 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3915 It accepts the following parameters:
3919 The timebase which will be used for timestamps of submitted frames. It must be
3920 either a floating-point number or in @var{numerator}/@var{denominator} form.
3923 The sample rate of the incoming audio buffers.
3926 The sample format of the incoming audio buffers.
3927 Either a sample format name or its corresponding integer representation from
3928 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3930 @item channel_layout
3931 The channel layout of the incoming audio buffers.
3932 Either a channel layout name from channel_layout_map in
3933 @file{libavutil/channel_layout.c} or its corresponding integer representation
3934 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3937 The number of channels of the incoming audio buffers.
3938 If both @var{channels} and @var{channel_layout} are specified, then they
3943 @subsection Examples
3946 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3949 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3950 Since the sample format with name "s16p" corresponds to the number
3951 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3954 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3959 Generate an audio signal specified by an expression.
3961 This source accepts in input one or more expressions (one for each
3962 channel), which are evaluated and used to generate a corresponding
3965 This source accepts the following options:
3969 Set the '|'-separated expressions list for each separate channel. In case the
3970 @option{channel_layout} option is not specified, the selected channel layout
3971 depends on the number of provided expressions. Otherwise the last
3972 specified expression is applied to the remaining output channels.
3974 @item channel_layout, c
3975 Set the channel layout. The number of channels in the specified layout
3976 must be equal to the number of specified expressions.
3979 Set the minimum duration of the sourced audio. See
3980 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3981 for the accepted syntax.
3982 Note that the resulting duration may be greater than the specified
3983 duration, as the generated audio is always cut at the end of a
3986 If not specified, or the expressed duration is negative, the audio is
3987 supposed to be generated forever.
3990 Set the number of samples per channel per each output frame,
3993 @item sample_rate, s
3994 Specify the sample rate, default to 44100.
3997 Each expression in @var{exprs} can contain the following constants:
4001 number of the evaluated sample, starting from 0
4004 time of the evaluated sample expressed in seconds, starting from 0
4011 @subsection Examples
4021 Generate a sin signal with frequency of 440 Hz, set sample rate to
4024 aevalsrc="sin(440*2*PI*t):s=8000"
4028 Generate a two channels signal, specify the channel layout (Front
4029 Center + Back Center) explicitly:
4031 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4035 Generate white noise:
4037 aevalsrc="-2+random(0)"
4041 Generate an amplitude modulated signal:
4043 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4047 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4049 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4056 The null audio source, return unprocessed audio frames. It is mainly useful
4057 as a template and to be employed in analysis / debugging tools, or as
4058 the source for filters which ignore the input data (for example the sox
4061 This source accepts the following options:
4065 @item channel_layout, cl
4067 Specifies the channel layout, and can be either an integer or a string
4068 representing a channel layout. The default value of @var{channel_layout}
4071 Check the channel_layout_map definition in
4072 @file{libavutil/channel_layout.c} for the mapping between strings and
4073 channel layout values.
4075 @item sample_rate, r
4076 Specifies the sample rate, and defaults to 44100.
4079 Set the number of samples per requested frames.
4083 @subsection Examples
4087 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4089 anullsrc=r=48000:cl=4
4093 Do the same operation with a more obvious syntax:
4095 anullsrc=r=48000:cl=mono
4099 All the parameters need to be explicitly defined.
4103 Synthesize a voice utterance using the libflite library.
4105 To enable compilation of this filter you need to configure FFmpeg with
4106 @code{--enable-libflite}.
4108 Note that the flite library is not thread-safe.
4110 The filter accepts the following options:
4115 If set to 1, list the names of the available voices and exit
4116 immediately. Default value is 0.
4119 Set the maximum number of samples per frame. Default value is 512.
4122 Set the filename containing the text to speak.
4125 Set the text to speak.
4128 Set the voice to use for the speech synthesis. Default value is
4129 @code{kal}. See also the @var{list_voices} option.
4132 @subsection Examples
4136 Read from file @file{speech.txt}, and synthesize the text using the
4137 standard flite voice:
4139 flite=textfile=speech.txt
4143 Read the specified text selecting the @code{slt} voice:
4145 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4149 Input text to ffmpeg:
4151 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4155 Make @file{ffplay} speak the specified text, using @code{flite} and
4156 the @code{lavfi} device:
4158 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4162 For more information about libflite, check:
4163 @url{http://www.speech.cs.cmu.edu/flite/}
4167 Generate a noise audio signal.
4169 The filter accepts the following options:
4172 @item sample_rate, r
4173 Specify the sample rate. Default value is 48000 Hz.
4176 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4180 Specify the duration of the generated audio stream. Not specifying this option
4181 results in noise with an infinite length.
4183 @item color, colour, c
4184 Specify the color of noise. Available noise colors are white, pink, and brown.
4185 Default color is white.
4188 Specify a value used to seed the PRNG.
4191 Set the number of samples per each output frame, default is 1024.
4194 @subsection Examples
4199 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4201 anoisesrc=d=60:c=pink:r=44100:a=0.5
4207 Generate an audio signal made of a sine wave with amplitude 1/8.
4209 The audio signal is bit-exact.
4211 The filter accepts the following options:
4216 Set the carrier frequency. Default is 440 Hz.
4218 @item beep_factor, b
4219 Enable a periodic beep every second with frequency @var{beep_factor} times
4220 the carrier frequency. Default is 0, meaning the beep is disabled.
4222 @item sample_rate, r
4223 Specify the sample rate, default is 44100.
4226 Specify the duration of the generated audio stream.
4228 @item samples_per_frame
4229 Set the number of samples per output frame.
4231 The expression can contain the following constants:
4235 The (sequential) number of the output audio frame, starting from 0.
4238 The PTS (Presentation TimeStamp) of the output audio frame,
4239 expressed in @var{TB} units.
4242 The PTS of the output audio frame, expressed in seconds.
4245 The timebase of the output audio frames.
4248 Default is @code{1024}.
4251 @subsection Examples
4256 Generate a simple 440 Hz sine wave:
4262 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4266 sine=frequency=220:beep_factor=4:duration=5
4270 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4273 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4277 @c man end AUDIO SOURCES
4279 @chapter Audio Sinks
4280 @c man begin AUDIO SINKS
4282 Below is a description of the currently available audio sinks.
4284 @section abuffersink
4286 Buffer audio frames, and make them available to the end of filter chain.
4288 This sink is mainly intended for programmatic use, in particular
4289 through the interface defined in @file{libavfilter/buffersink.h}
4290 or the options system.
4292 It accepts a pointer to an AVABufferSinkContext structure, which
4293 defines the incoming buffers' formats, to be passed as the opaque
4294 parameter to @code{avfilter_init_filter} for initialization.
4297 Null audio sink; do absolutely nothing with the input audio. It is
4298 mainly useful as a template and for use in analysis / debugging
4301 @c man end AUDIO SINKS
4303 @chapter Video Filters
4304 @c man begin VIDEO FILTERS
4306 When you configure your FFmpeg build, you can disable any of the
4307 existing filters using @code{--disable-filters}.
4308 The configure output will show the video filters included in your
4311 Below is a description of the currently available video filters.
4313 @section alphaextract
4315 Extract the alpha component from the input as a grayscale video. This
4316 is especially useful with the @var{alphamerge} filter.
4320 Add or replace the alpha component of the primary input with the
4321 grayscale value of a second input. This is intended for use with
4322 @var{alphaextract} to allow the transmission or storage of frame
4323 sequences that have alpha in a format that doesn't support an alpha
4326 For example, to reconstruct full frames from a normal YUV-encoded video
4327 and a separate video created with @var{alphaextract}, you might use:
4329 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4332 Since this filter is designed for reconstruction, it operates on frame
4333 sequences without considering timestamps, and terminates when either
4334 input reaches end of stream. This will cause problems if your encoding
4335 pipeline drops frames. If you're trying to apply an image as an
4336 overlay to a video stream, consider the @var{overlay} filter instead.
4340 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4341 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4342 Substation Alpha) subtitles files.
4344 This filter accepts the following option in addition to the common options from
4345 the @ref{subtitles} filter:
4349 Set the shaping engine
4351 Available values are:
4354 The default libass shaping engine, which is the best available.
4356 Fast, font-agnostic shaper that can do only substitutions
4358 Slower shaper using OpenType for substitutions and positioning
4361 The default is @code{auto}.
4365 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4367 The filter accepts the following options:
4371 Set threshold A for 1st plane. Default is 0.02.
4372 Valid range is 0 to 0.3.
4375 Set threshold B for 1st plane. Default is 0.04.
4376 Valid range is 0 to 5.
4379 Set threshold A for 2nd plane. Default is 0.02.
4380 Valid range is 0 to 0.3.
4383 Set threshold B for 2nd plane. Default is 0.04.
4384 Valid range is 0 to 5.
4387 Set threshold A for 3rd plane. Default is 0.02.
4388 Valid range is 0 to 0.3.
4391 Set threshold B for 3rd plane. Default is 0.04.
4392 Valid range is 0 to 5.
4394 Threshold A is designed to react on abrupt changes in the input signal and
4395 threshold B is designed to react on continuous changes in the input signal.
4398 Set number of frames filter will use for averaging. Default is 33. Must be odd
4399 number in range [5, 129].
4402 Set what planes of frame filter will use for averaging. Default is all.
4407 Apply average blur filter.
4409 The filter accepts the following options:
4413 Set horizontal kernel size.
4416 Set which planes to filter. By default all planes are filtered.
4419 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4420 Default is @code{0}.
4425 Compute the bounding box for the non-black pixels in the input frame
4428 This filter computes the bounding box containing all the pixels with a
4429 luminance value greater than the minimum allowed value.
4430 The parameters describing the bounding box are printed on the filter
4433 The filter accepts the following option:
4437 Set the minimal luminance value. Default is @code{16}.
4440 @section bitplanenoise
4442 Show and measure bit plane noise.
4444 The filter accepts the following options:
4448 Set which plane to analyze. Default is @code{1}.
4451 Filter out noisy pixels from @code{bitplane} set above.
4452 Default is disabled.
4455 @section blackdetect
4457 Detect video intervals that are (almost) completely black. Can be
4458 useful to detect chapter transitions, commercials, or invalid
4459 recordings. Output lines contains the time for the start, end and
4460 duration of the detected black interval expressed in seconds.
4462 In order to display the output lines, you need to set the loglevel at
4463 least to the AV_LOG_INFO value.
4465 The filter accepts the following options:
4468 @item black_min_duration, d
4469 Set the minimum detected black duration expressed in seconds. It must
4470 be a non-negative floating point number.
4472 Default value is 2.0.
4474 @item picture_black_ratio_th, pic_th
4475 Set the threshold for considering a picture "black".
4476 Express the minimum value for the ratio:
4478 @var{nb_black_pixels} / @var{nb_pixels}
4481 for which a picture is considered black.
4482 Default value is 0.98.
4484 @item pixel_black_th, pix_th
4485 Set the threshold for considering a pixel "black".
4487 The threshold expresses the maximum pixel luminance value for which a
4488 pixel is considered "black". The provided value is scaled according to
4489 the following equation:
4491 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4494 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4495 the input video format, the range is [0-255] for YUV full-range
4496 formats and [16-235] for YUV non full-range formats.
4498 Default value is 0.10.
4501 The following example sets the maximum pixel threshold to the minimum
4502 value, and detects only black intervals of 2 or more seconds:
4504 blackdetect=d=2:pix_th=0.00
4509 Detect frames that are (almost) completely black. Can be useful to
4510 detect chapter transitions or commercials. Output lines consist of
4511 the frame number of the detected frame, the percentage of blackness,
4512 the position in the file if known or -1 and the timestamp in seconds.
4514 In order to display the output lines, you need to set the loglevel at
4515 least to the AV_LOG_INFO value.
4517 It accepts the following parameters:
4522 The percentage of the pixels that have to be below the threshold; it defaults to
4525 @item threshold, thresh
4526 The threshold below which a pixel value is considered black; it defaults to
4531 @section blend, tblend
4533 Blend two video frames into each other.
4535 The @code{blend} filter takes two input streams and outputs one
4536 stream, the first input is the "top" layer and second input is
4537 "bottom" layer. By default, the output terminates when the longest input terminates.
4539 The @code{tblend} (time blend) filter takes two consecutive frames
4540 from one single stream, and outputs the result obtained by blending
4541 the new frame on top of the old frame.
4543 A description of the accepted options follows.
4551 Set blend mode for specific pixel component or all pixel components in case
4552 of @var{all_mode}. Default value is @code{normal}.
4554 Available values for component modes are:
4595 Set blend opacity for specific pixel component or all pixel components in case
4596 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4603 Set blend expression for specific pixel component or all pixel components in case
4604 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4606 The expressions can use the following variables:
4610 The sequential number of the filtered frame, starting from @code{0}.
4614 the coordinates of the current sample
4618 the width and height of currently filtered plane
4622 Width and height scale depending on the currently filtered plane. It is the
4623 ratio between the corresponding luma plane number of pixels and the current
4624 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4625 @code{0.5,0.5} for chroma planes.
4628 Time of the current frame, expressed in seconds.
4631 Value of pixel component at current location for first video frame (top layer).
4634 Value of pixel component at current location for second video frame (bottom layer).
4638 Force termination when the shortest input terminates. Default is
4639 @code{0}. This option is only defined for the @code{blend} filter.
4642 Continue applying the last bottom frame after the end of the stream. A value of
4643 @code{0} disable the filter after the last frame of the bottom layer is reached.
4644 Default is @code{1}. This option is only defined for the @code{blend} filter.
4647 @subsection Examples
4651 Apply transition from bottom layer to top layer in first 10 seconds:
4653 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4657 Apply 1x1 checkerboard effect:
4659 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4663 Apply uncover left effect:
4665 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4669 Apply uncover down effect:
4671 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4675 Apply uncover up-left effect:
4677 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4681 Split diagonally video and shows top and bottom layer on each side:
4683 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4687 Display differences between the current and the previous frame:
4689 tblend=all_mode=difference128
4695 Apply a boxblur algorithm to the input video.
4697 It accepts the following parameters:
4701 @item luma_radius, lr
4702 @item luma_power, lp
4703 @item chroma_radius, cr
4704 @item chroma_power, cp
4705 @item alpha_radius, ar
4706 @item alpha_power, ap
4710 A description of the accepted options follows.
4713 @item luma_radius, lr
4714 @item chroma_radius, cr
4715 @item alpha_radius, ar
4716 Set an expression for the box radius in pixels used for blurring the
4717 corresponding input plane.
4719 The radius value must be a non-negative number, and must not be
4720 greater than the value of the expression @code{min(w,h)/2} for the
4721 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4724 Default value for @option{luma_radius} is "2". If not specified,
4725 @option{chroma_radius} and @option{alpha_radius} default to the
4726 corresponding value set for @option{luma_radius}.
4728 The expressions can contain the following constants:
4732 The input width and height in pixels.
4736 The input chroma image width and height in pixels.
4740 The horizontal and vertical chroma subsample values. For example, for the
4741 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4744 @item luma_power, lp
4745 @item chroma_power, cp
4746 @item alpha_power, ap
4747 Specify how many times the boxblur filter is applied to the
4748 corresponding plane.
4750 Default value for @option{luma_power} is 2. If not specified,
4751 @option{chroma_power} and @option{alpha_power} default to the
4752 corresponding value set for @option{luma_power}.
4754 A value of 0 will disable the effect.
4757 @subsection Examples
4761 Apply a boxblur filter with the luma, chroma, and alpha radii
4764 boxblur=luma_radius=2:luma_power=1
4769 Set the luma radius to 2, and alpha and chroma radius to 0:
4771 boxblur=2:1:cr=0:ar=0
4775 Set the luma and chroma radii to a fraction of the video dimension:
4777 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4783 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4784 Deinterlacing Filter").
4786 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4787 interpolation algorithms.
4788 It accepts the following parameters:
4792 The interlacing mode to adopt. It accepts one of the following values:
4796 Output one frame for each frame.
4798 Output one frame for each field.
4801 The default value is @code{send_field}.
4804 The picture field parity assumed for the input interlaced video. It accepts one
4805 of the following values:
4809 Assume the top field is first.
4811 Assume the bottom field is first.
4813 Enable automatic detection of field parity.
4816 The default value is @code{auto}.
4817 If the interlacing is unknown or the decoder does not export this information,
4818 top field first will be assumed.
4821 Specify which frames to deinterlace. Accept one of the following
4826 Deinterlace all frames.
4828 Only deinterlace frames marked as interlaced.
4831 The default value is @code{all}.
4835 YUV colorspace color/chroma keying.
4837 The filter accepts the following options:
4841 The color which will be replaced with transparency.
4844 Similarity percentage with the key color.
4846 0.01 matches only the exact key color, while 1.0 matches everything.
4851 0.0 makes pixels either fully transparent, or not transparent at all.
4853 Higher values result in semi-transparent pixels, with a higher transparency
4854 the more similar the pixels color is to the key color.
4857 Signals that the color passed is already in YUV instead of RGB.
4859 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4860 This can be used to pass exact YUV values as hexadecimal numbers.
4863 @subsection Examples
4867 Make every green pixel in the input image transparent:
4869 ffmpeg -i input.png -vf chromakey=green out.png
4873 Overlay a greenscreen-video on top of a static black background.
4875 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
4881 Display CIE color diagram with pixels overlaid onto it.
4883 The filter accepts the following options:
4898 @item uhdtv, rec2020
4911 Set what gamuts to draw.
4913 See @code{system} option for available values.
4916 Set ciescope size, by default set to 512.
4919 Set intensity used to map input pixel values to CIE diagram.
4922 Set contrast used to draw tongue colors that are out of active color system gamut.
4925 Correct gamma displayed on scope, by default enabled.
4928 Show white point on CIE diagram, by default disabled.
4931 Set input gamma. Used only with XYZ input color space.
4936 Visualize information exported by some codecs.
4938 Some codecs can export information through frames using side-data or other
4939 means. For example, some MPEG based codecs export motion vectors through the
4940 @var{export_mvs} flag in the codec @option{flags2} option.
4942 The filter accepts the following option:
4946 Set motion vectors to visualize.
4948 Available flags for @var{mv} are:
4952 forward predicted MVs of P-frames
4954 forward predicted MVs of B-frames
4956 backward predicted MVs of B-frames
4960 Display quantization parameters using the chroma planes.
4963 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4965 Available flags for @var{mv_type} are:
4969 forward predicted MVs
4971 backward predicted MVs
4974 @item frame_type, ft
4975 Set frame type to visualize motion vectors of.
4977 Available flags for @var{frame_type} are:
4981 intra-coded frames (I-frames)
4983 predicted frames (P-frames)
4985 bi-directionally predicted frames (B-frames)
4989 @subsection Examples
4993 Visualize forward predicted MVs of all frames using @command{ffplay}:
4995 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
4999 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5001 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5005 @section colorbalance
5006 Modify intensity of primary colors (red, green and blue) of input frames.
5008 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5009 regions for the red-cyan, green-magenta or blue-yellow balance.
5011 A positive adjustment value shifts the balance towards the primary color, a negative
5012 value towards the complementary color.
5014 The filter accepts the following options:
5020 Adjust red, green and blue shadows (darkest pixels).
5025 Adjust red, green and blue midtones (medium pixels).
5030 Adjust red, green and blue highlights (brightest pixels).
5032 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5035 @subsection Examples
5039 Add red color cast to shadows:
5046 RGB colorspace color keying.
5048 The filter accepts the following options:
5052 The color which will be replaced with transparency.
5055 Similarity percentage with the key color.
5057 0.01 matches only the exact key color, while 1.0 matches everything.
5062 0.0 makes pixels either fully transparent, or not transparent at all.
5064 Higher values result in semi-transparent pixels, with a higher transparency
5065 the more similar the pixels color is to the key color.
5068 @subsection Examples
5072 Make every green pixel in the input image transparent:
5074 ffmpeg -i input.png -vf colorkey=green out.png
5078 Overlay a greenscreen-video on top of a static background image.
5080 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
5084 @section colorlevels
5086 Adjust video input frames using levels.
5088 The filter accepts the following options:
5095 Adjust red, green, blue and alpha input black point.
5096 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5102 Adjust red, green, blue and alpha input white point.
5103 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5105 Input levels are used to lighten highlights (bright tones), darken shadows
5106 (dark tones), change the balance of bright and dark tones.
5112 Adjust red, green, blue and alpha output black point.
5113 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5119 Adjust red, green, blue and alpha output white point.
5120 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5122 Output levels allows manual selection of a constrained output level range.
5125 @subsection Examples
5129 Make video output darker:
5131 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5137 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5141 Make video output lighter:
5143 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5147 Increase brightness:
5149 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5153 @section colorchannelmixer
5155 Adjust video input frames by re-mixing color channels.
5157 This filter modifies a color channel by adding the values associated to
5158 the other channels of the same pixels. For example if the value to
5159 modify is red, the output value will be:
5161 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5164 The filter accepts the following options:
5171 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5172 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5178 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5179 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5185 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5186 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5192 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5193 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5195 Allowed ranges for options are @code{[-2.0, 2.0]}.
5198 @subsection Examples
5202 Convert source to grayscale:
5204 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5207 Simulate sepia tones:
5209 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5213 @section colormatrix
5215 Convert color matrix.
5217 The filter accepts the following options:
5222 Specify the source and destination color matrix. Both values must be
5225 The accepted values are:
5244 For example to convert from BT.601 to SMPTE-240M, use the command:
5246 colormatrix=bt601:smpte240m
5251 Convert colorspace, transfer characteristics or color primaries.
5253 The filter accepts the following options:
5258 Specify all color properties at once.
5260 The accepted values are:
5290 Specify output colorspace.
5292 The accepted values are:
5301 BT.470BG or BT.601-6 625
5304 SMPTE-170M or BT.601-6 525
5310 BT.2020 with non-constant luminance
5316 Specify output transfer characteristics.
5318 The accepted values are:
5324 Constant gamma of 2.2
5327 Constant gamma of 2.8
5330 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5336 BT.2020 for 10-bits content
5339 BT.2020 for 12-bits content
5345 Specify output color primaries.
5347 The accepted values are:
5356 BT.470BG or BT.601-6 625
5359 SMPTE-170M or BT.601-6 525
5371 Specify output color range.
5373 The accepted values are:
5376 MPEG (restricted) range
5384 Specify output color format.
5386 The accepted values are:
5389 YUV 4:2:0 planar 8-bits
5392 YUV 4:2:0 planar 10-bits
5395 YUV 4:2:0 planar 12-bits
5398 YUV 4:2:2 planar 8-bits
5401 YUV 4:2:2 planar 10-bits
5404 YUV 4:2:2 planar 12-bits
5407 YUV 4:4:4 planar 8-bits
5410 YUV 4:4:4 planar 10-bits
5413 YUV 4:4:4 planar 12-bits
5418 Do a fast conversion, which skips gamma/primary correction. This will take
5419 significantly less CPU, but will be mathematically incorrect. To get output
5420 compatible with that produced by the colormatrix filter, use fast=1.
5423 Specify dithering mode.
5425 The accepted values are:
5431 Floyd-Steinberg dithering
5435 Whitepoint adaptation mode.
5437 The accepted values are:
5440 Bradford whitepoint adaptation
5443 von Kries whitepoint adaptation
5446 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5450 Override all input properties at once. Same accepted values as @ref{all}.
5453 Override input colorspace. Same accepted values as @ref{space}.
5456 Override input color primaries. Same accepted values as @ref{primaries}.
5459 Override input transfer characteristics. Same accepted values as @ref{trc}.
5462 Override input color range. Same accepted values as @ref{range}.
5466 The filter converts the transfer characteristics, color space and color
5467 primaries to the specified user values. The output value, if not specified,
5468 is set to a default value based on the "all" property. If that property is
5469 also not specified, the filter will log an error. The output color range and
5470 format default to the same value as the input color range and format. The
5471 input transfer characteristics, color space, color primaries and color range
5472 should be set on the input data. If any of these are missing, the filter will
5473 log an error and no conversion will take place.
5475 For example to convert the input to SMPTE-240M, use the command:
5477 colorspace=smpte240m
5480 @section convolution
5482 Apply convolution 3x3 or 5x5 filter.
5484 The filter accepts the following options:
5491 Set matrix for each plane.
5492 Matrix is sequence of 9 or 25 signed integers.
5498 Set multiplier for calculated value for each plane.
5504 Set bias for each plane. This value is added to the result of the multiplication.
5505 Useful for making the overall image brighter or darker. Default is 0.0.
5508 @subsection Examples
5514 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"
5520 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"
5526 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"
5532 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"
5538 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"
5544 Copy the input source unchanged to the output. This is mainly useful for
5549 Video filtering on GPU using Apple's CoreImage API on OSX.
5551 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5552 processed by video hardware. However, software-based OpenGL implementations
5553 exist which means there is no guarantee for hardware processing. It depends on
5556 There are many filters and image generators provided by Apple that come with a
5557 large variety of options. The filter has to be referenced by its name along
5560 The coreimage filter accepts the following options:
5563 List all available filters and generators along with all their respective
5564 options as well as possible minimum and maximum values along with the default
5571 Specify all filters by their respective name and options.
5572 Use @var{list_filters} to determine all valid filter names and options.
5573 Numerical options are specified by a float value and are automatically clamped
5574 to their respective value range. Vector and color options have to be specified
5575 by a list of space separated float values. Character escaping has to be done.
5576 A special option name @code{default} is available to use default options for a
5579 It is required to specify either @code{default} or at least one of the filter options.
5580 All omitted options are used with their default values.
5581 The syntax of the filter string is as follows:
5583 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5587 Specify a rectangle where the output of the filter chain is copied into the
5588 input image. It is given by a list of space separated float values:
5590 output_rect=x\ y\ width\ height
5592 If not given, the output rectangle equals the dimensions of the input image.
5593 The output rectangle is automatically cropped at the borders of the input
5594 image. Negative values are valid for each component.
5596 output_rect=25\ 25\ 100\ 100
5600 Several filters can be chained for successive processing without GPU-HOST
5601 transfers allowing for fast processing of complex filter chains.
5602 Currently, only filters with zero (generators) or exactly one (filters) input
5603 image and one output image are supported. Also, transition filters are not yet
5606 Some filters generate output images with additional padding depending on the
5607 respective filter kernel. The padding is automatically removed to ensure the
5608 filter output has the same size as the input image.
5610 For image generators, the size of the output image is determined by the
5611 previous output image of the filter chain or the input image of the whole
5612 filterchain, respectively. The generators do not use the pixel information of
5613 this image to generate their output. However, the generated output is
5614 blended onto this image, resulting in partial or complete coverage of the
5617 The @ref{coreimagesrc} video source can be used for generating input images
5618 which are directly fed into the filter chain. By using it, providing input
5619 images by another video source or an input video is not required.
5621 @subsection Examples
5626 List all filters available:
5628 coreimage=list_filters=true
5632 Use the CIBoxBlur filter with default options to blur an image:
5634 coreimage=filter=CIBoxBlur@@default
5638 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5639 its center at 100x100 and a radius of 50 pixels:
5641 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5645 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5646 given as complete and escaped command-line for Apple's standard bash shell:
5648 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5654 Crop the input video to given dimensions.
5656 It accepts the following parameters:
5660 The width of the output video. It defaults to @code{iw}.
5661 This expression is evaluated only once during the filter
5662 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5665 The height of the output video. It defaults to @code{ih}.
5666 This expression is evaluated only once during the filter
5667 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5670 The horizontal position, in the input video, of the left edge of the output
5671 video. It defaults to @code{(in_w-out_w)/2}.
5672 This expression is evaluated per-frame.
5675 The vertical position, in the input video, of the top edge of the output video.
5676 It defaults to @code{(in_h-out_h)/2}.
5677 This expression is evaluated per-frame.
5680 If set to 1 will force the output display aspect ratio
5681 to be the same of the input, by changing the output sample aspect
5682 ratio. It defaults to 0.
5685 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5686 width/height/x/y as specified and will not be rounded to nearest smaller value.
5690 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5691 expressions containing the following constants:
5696 The computed values for @var{x} and @var{y}. They are evaluated for
5701 The input width and height.
5705 These are the same as @var{in_w} and @var{in_h}.
5709 The output (cropped) width and height.
5713 These are the same as @var{out_w} and @var{out_h}.
5716 same as @var{iw} / @var{ih}
5719 input sample aspect ratio
5722 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5726 horizontal and vertical chroma subsample values. For example for the
5727 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5730 The number of the input frame, starting from 0.
5733 the position in the file of the input frame, NAN if unknown
5736 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5740 The expression for @var{out_w} may depend on the value of @var{out_h},
5741 and the expression for @var{out_h} may depend on @var{out_w}, but they
5742 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5743 evaluated after @var{out_w} and @var{out_h}.
5745 The @var{x} and @var{y} parameters specify the expressions for the
5746 position of the top-left corner of the output (non-cropped) area. They
5747 are evaluated for each frame. If the evaluated value is not valid, it
5748 is approximated to the nearest valid value.
5750 The expression for @var{x} may depend on @var{y}, and the expression
5751 for @var{y} may depend on @var{x}.
5753 @subsection Examples
5757 Crop area with size 100x100 at position (12,34).
5762 Using named options, the example above becomes:
5764 crop=w=100:h=100:x=12:y=34
5768 Crop the central input area with size 100x100:
5774 Crop the central input area with size 2/3 of the input video:
5776 crop=2/3*in_w:2/3*in_h
5780 Crop the input video central square:
5787 Delimit the rectangle with the top-left corner placed at position
5788 100:100 and the right-bottom corner corresponding to the right-bottom
5789 corner of the input image.
5791 crop=in_w-100:in_h-100:100:100
5795 Crop 10 pixels from the left and right borders, and 20 pixels from
5796 the top and bottom borders
5798 crop=in_w-2*10:in_h-2*20
5802 Keep only the bottom right quarter of the input image:
5804 crop=in_w/2:in_h/2:in_w/2:in_h/2
5808 Crop height for getting Greek harmony:
5810 crop=in_w:1/PHI*in_w
5814 Apply trembling effect:
5816 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)
5820 Apply erratic camera effect depending on timestamp:
5822 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)"
5826 Set x depending on the value of y:
5828 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5832 @subsection Commands
5834 This filter supports the following commands:
5840 Set width/height of the output video and the horizontal/vertical position
5842 The command accepts the same syntax of the corresponding option.
5844 If the specified expression is not valid, it is kept at its current
5850 Auto-detect the crop size.
5852 It calculates the necessary cropping parameters and prints the
5853 recommended parameters via the logging system. The detected dimensions
5854 correspond to the non-black area of the input video.
5856 It accepts the following parameters:
5861 Set higher black value threshold, which can be optionally specified
5862 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5863 value greater to the set value is considered non-black. It defaults to 24.
5864 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5865 on the bitdepth of the pixel format.
5868 The value which the width/height should be divisible by. It defaults to
5869 16. The offset is automatically adjusted to center the video. Use 2 to
5870 get only even dimensions (needed for 4:2:2 video). 16 is best when
5871 encoding to most video codecs.
5873 @item reset_count, reset
5874 Set the counter that determines after how many frames cropdetect will
5875 reset the previously detected largest video area and start over to
5876 detect the current optimal crop area. Default value is 0.
5878 This can be useful when channel logos distort the video area. 0
5879 indicates 'never reset', and returns the largest area encountered during
5886 Apply color adjustments using curves.
5888 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5889 component (red, green and blue) has its values defined by @var{N} key points
5890 tied from each other using a smooth curve. The x-axis represents the pixel
5891 values from the input frame, and the y-axis the new pixel values to be set for
5894 By default, a component curve is defined by the two points @var{(0;0)} and
5895 @var{(1;1)}. This creates a straight line where each original pixel value is
5896 "adjusted" to its own value, which means no change to the image.
5898 The filter allows you to redefine these two points and add some more. A new
5899 curve (using a natural cubic spline interpolation) will be define to pass
5900 smoothly through all these new coordinates. The new defined points needs to be
5901 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5902 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5903 the vector spaces, the values will be clipped accordingly.
5905 The filter accepts the following options:
5909 Select one of the available color presets. This option can be used in addition
5910 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5911 options takes priority on the preset values.
5912 Available presets are:
5915 @item color_negative
5918 @item increase_contrast
5920 @item linear_contrast
5921 @item medium_contrast
5923 @item strong_contrast
5926 Default is @code{none}.
5928 Set the master key points. These points will define a second pass mapping. It
5929 is sometimes called a "luminance" or "value" mapping. It can be used with
5930 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5931 post-processing LUT.
5933 Set the key points for the red component.
5935 Set the key points for the green component.
5937 Set the key points for the blue component.
5939 Set the key points for all components (not including master).
5940 Can be used in addition to the other key points component
5941 options. In this case, the unset component(s) will fallback on this
5942 @option{all} setting.
5944 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5946 Save Gnuplot script of the curves in specified file.
5949 To avoid some filtergraph syntax conflicts, each key points list need to be
5950 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5952 @subsection Examples
5956 Increase slightly the middle level of blue:
5958 curves=blue='0/0 0.5/0.58 1/1'
5964 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'
5966 Here we obtain the following coordinates for each components:
5969 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5971 @code{(0;0) (0.50;0.48) (1;1)}
5973 @code{(0;0.22) (0.49;0.44) (1;0.80)}
5977 The previous example can also be achieved with the associated built-in preset:
5979 curves=preset=vintage
5989 Use a Photoshop preset and redefine the points of the green component:
5991 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
5995 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
5996 and @command{gnuplot}:
5998 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
5999 gnuplot -p /tmp/curves.plt
6005 Video data analysis filter.
6007 This filter shows hexadecimal pixel values of part of video.
6009 The filter accepts the following options:
6013 Set output video size.
6016 Set x offset from where to pick pixels.
6019 Set y offset from where to pick pixels.
6022 Set scope mode, can be one of the following:
6025 Draw hexadecimal pixel values with white color on black background.
6028 Draw hexadecimal pixel values with input video pixel color on black
6032 Draw hexadecimal pixel values on color background picked from input video,
6033 the text color is picked in such way so its always visible.
6037 Draw rows and columns numbers on left and top of video.
6040 Set background opacity.
6045 Denoise frames using 2D DCT (frequency domain filtering).
6047 This filter is not designed for real time.
6049 The filter accepts the following options:
6053 Set the noise sigma constant.
6055 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6056 coefficient (absolute value) below this threshold with be dropped.
6058 If you need a more advanced filtering, see @option{expr}.
6060 Default is @code{0}.
6063 Set number overlapping pixels for each block. Since the filter can be slow, you
6064 may want to reduce this value, at the cost of a less effective filter and the
6065 risk of various artefacts.
6067 If the overlapping value doesn't permit processing the whole input width or
6068 height, a warning will be displayed and according borders won't be denoised.
6070 Default value is @var{blocksize}-1, which is the best possible setting.
6073 Set the coefficient factor expression.
6075 For each coefficient of a DCT block, this expression will be evaluated as a
6076 multiplier value for the coefficient.
6078 If this is option is set, the @option{sigma} option will be ignored.
6080 The absolute value of the coefficient can be accessed through the @var{c}
6084 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6085 @var{blocksize}, which is the width and height of the processed blocks.
6087 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6088 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6089 on the speed processing. Also, a larger block size does not necessarily means a
6093 @subsection Examples
6095 Apply a denoise with a @option{sigma} of @code{4.5}:
6100 The same operation can be achieved using the expression system:
6102 dctdnoiz=e='gte(c, 4.5*3)'
6105 Violent denoise using a block size of @code{16x16}:
6112 Remove banding artifacts from input video.
6113 It works by replacing banded pixels with average value of referenced pixels.
6115 The filter accepts the following options:
6122 Set banding detection threshold for each plane. Default is 0.02.
6123 Valid range is 0.00003 to 0.5.
6124 If difference between current pixel and reference pixel is less than threshold,
6125 it will be considered as banded.
6128 Banding detection range in pixels. Default is 16. If positive, random number
6129 in range 0 to set value will be used. If negative, exact absolute value
6131 The range defines square of four pixels around current pixel.
6134 Set direction in radians from which four pixel will be compared. If positive,
6135 random direction from 0 to set direction will be picked. If negative, exact of
6136 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6137 will pick only pixels on same row and -PI/2 will pick only pixels on same
6141 If enabled, current pixel is compared with average value of all four
6142 surrounding pixels. The default is enabled. If disabled current pixel is
6143 compared with all four surrounding pixels. The pixel is considered banded
6144 if only all four differences with surrounding pixels are less than threshold.
6150 Drop duplicated frames at regular intervals.
6152 The filter accepts the following options:
6156 Set the number of frames from which one will be dropped. Setting this to
6157 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6158 Default is @code{5}.
6161 Set the threshold for duplicate detection. If the difference metric for a frame
6162 is less than or equal to this value, then it is declared as duplicate. Default
6166 Set scene change threshold. Default is @code{15}.
6170 Set the size of the x and y-axis blocks used during metric calculations.
6171 Larger blocks give better noise suppression, but also give worse detection of
6172 small movements. Must be a power of two. Default is @code{32}.
6175 Mark main input as a pre-processed input and activate clean source input
6176 stream. This allows the input to be pre-processed with various filters to help
6177 the metrics calculation while keeping the frame selection lossless. When set to
6178 @code{1}, the first stream is for the pre-processed input, and the second
6179 stream is the clean source from where the kept frames are chosen. Default is
6183 Set whether or not chroma is considered in the metric calculations. Default is
6189 Apply deflate effect to the video.
6191 This filter replaces the pixel by the local(3x3) average by taking into account
6192 only values lower than the pixel.
6194 It accepts the following options:
6201 Limit the maximum change for each plane, default is 65535.
6202 If 0, plane will remain unchanged.
6207 Remove judder produced by partially interlaced telecined content.
6209 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6210 source was partially telecined content then the output of @code{pullup,dejudder}
6211 will have a variable frame rate. May change the recorded frame rate of the
6212 container. Aside from that change, this filter will not affect constant frame
6215 The option available in this filter is:
6219 Specify the length of the window over which the judder repeats.
6221 Accepts any integer greater than 1. Useful values are:
6225 If the original was telecined from 24 to 30 fps (Film to NTSC).
6228 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6231 If a mixture of the two.
6234 The default is @samp{4}.
6239 Suppress a TV station logo by a simple interpolation of the surrounding
6240 pixels. Just set a rectangle covering the logo and watch it disappear
6241 (and sometimes something even uglier appear - your mileage may vary).
6243 It accepts the following parameters:
6248 Specify the top left corner coordinates of the logo. They must be
6253 Specify the width and height of the logo to clear. They must be
6257 Specify the thickness of the fuzzy edge of the rectangle (added to
6258 @var{w} and @var{h}). The default value is 1. This option is
6259 deprecated, setting higher values should no longer be necessary and
6263 When set to 1, a green rectangle is drawn on the screen to simplify
6264 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6265 The default value is 0.
6267 The rectangle is drawn on the outermost pixels which will be (partly)
6268 replaced with interpolated values. The values of the next pixels
6269 immediately outside this rectangle in each direction will be used to
6270 compute the interpolated pixel values inside the rectangle.
6274 @subsection Examples
6278 Set a rectangle covering the area with top left corner coordinates 0,0
6279 and size 100x77, and a band of size 10:
6281 delogo=x=0:y=0:w=100:h=77:band=10
6288 Attempt to fix small changes in horizontal and/or vertical shift. This
6289 filter helps remove camera shake from hand-holding a camera, bumping a
6290 tripod, moving on a vehicle, etc.
6292 The filter accepts the following options:
6300 Specify a rectangular area where to limit the search for motion
6302 If desired the search for motion vectors can be limited to a
6303 rectangular area of the frame defined by its top left corner, width
6304 and height. These parameters have the same meaning as the drawbox
6305 filter which can be used to visualise the position of the bounding
6308 This is useful when simultaneous movement of subjects within the frame
6309 might be confused for camera motion by the motion vector search.
6311 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6312 then the full frame is used. This allows later options to be set
6313 without specifying the bounding box for the motion vector search.
6315 Default - search the whole frame.
6319 Specify the maximum extent of movement in x and y directions in the
6320 range 0-64 pixels. Default 16.
6323 Specify how to generate pixels to fill blanks at the edge of the
6324 frame. Available values are:
6327 Fill zeroes at blank locations
6329 Original image at blank locations
6331 Extruded edge value at blank locations
6333 Mirrored edge at blank locations
6335 Default value is @samp{mirror}.
6338 Specify the blocksize to use for motion search. Range 4-128 pixels,
6342 Specify the contrast threshold for blocks. Only blocks with more than
6343 the specified contrast (difference between darkest and lightest
6344 pixels) will be considered. Range 1-255, default 125.
6347 Specify the search strategy. Available values are:
6350 Set exhaustive search
6352 Set less exhaustive search.
6354 Default value is @samp{exhaustive}.
6357 If set then a detailed log of the motion search is written to the
6361 If set to 1, specify using OpenCL capabilities, only available if
6362 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6368 Apply an exact inverse of the telecine operation. It requires a predefined
6369 pattern specified using the pattern option which must be the same as that passed
6370 to the telecine filter.
6372 This filter accepts the following options:
6381 The default value is @code{top}.
6385 A string of numbers representing the pulldown pattern you wish to apply.
6386 The default value is @code{23}.
6389 A number representing position of the first frame with respect to the telecine
6390 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6395 Apply dilation effect to the video.
6397 This filter replaces the pixel by the local(3x3) maximum.
6399 It accepts the following options:
6406 Limit the maximum change for each plane, default is 65535.
6407 If 0, plane will remain unchanged.
6410 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6413 Flags to local 3x3 coordinates maps like this:
6422 Displace pixels as indicated by second and third input stream.
6424 It takes three input streams and outputs one stream, the first input is the
6425 source, and second and third input are displacement maps.
6427 The second input specifies how much to displace pixels along the
6428 x-axis, while the third input specifies how much to displace pixels
6430 If one of displacement map streams terminates, last frame from that
6431 displacement map will be used.
6433 Note that once generated, displacements maps can be reused over and over again.
6435 A description of the accepted options follows.
6439 Set displace behavior for pixels that are out of range.
6441 Available values are:
6444 Missing pixels are replaced by black pixels.
6447 Adjacent pixels will spread out to replace missing pixels.
6450 Out of range pixels are wrapped so they point to pixels of other side.
6452 Default is @samp{smear}.
6456 @subsection Examples
6460 Add ripple effect to rgb input of video size hd720:
6462 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
6466 Add wave effect to rgb input of video size hd720:
6468 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
6474 Draw a colored box on the input image.
6476 It accepts the following parameters:
6481 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6485 The expressions which specify the width and height of the box; if 0 they are interpreted as
6486 the input width and height. It defaults to 0.
6489 Specify the color of the box to write. For the general syntax of this option,
6490 check the "Color" section in the ffmpeg-utils manual. If the special
6491 value @code{invert} is used, the box edge color is the same as the
6492 video with inverted luma.
6495 The expression which sets the thickness of the box edge. Default value is @code{3}.
6497 See below for the list of accepted constants.
6500 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6501 following constants:
6505 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6509 horizontal and vertical chroma subsample values. For example for the
6510 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6514 The input width and height.
6517 The input sample aspect ratio.
6521 The x and y offset coordinates where the box is drawn.
6525 The width and height of the drawn box.
6528 The thickness of the drawn box.
6530 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6531 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6535 @subsection Examples
6539 Draw a black box around the edge of the input image:
6545 Draw a box with color red and an opacity of 50%:
6547 drawbox=10:20:200:60:red@@0.5
6550 The previous example can be specified as:
6552 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6556 Fill the box with pink color:
6558 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6562 Draw a 2-pixel red 2.40:1 mask:
6564 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
6570 Draw a grid on the input image.
6572 It accepts the following parameters:
6577 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6581 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6582 input width and height, respectively, minus @code{thickness}, so image gets
6583 framed. Default to 0.
6586 Specify the color of the grid. For the general syntax of this option,
6587 check the "Color" section in the ffmpeg-utils manual. If the special
6588 value @code{invert} is used, the grid color is the same as the
6589 video with inverted luma.
6592 The expression which sets the thickness of the grid line. Default value is @code{1}.
6594 See below for the list of accepted constants.
6597 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6598 following constants:
6602 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6606 horizontal and vertical chroma subsample values. For example for the
6607 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6611 The input grid cell width and height.
6614 The input sample aspect ratio.
6618 The x and y coordinates of some point of grid intersection (meant to configure offset).
6622 The width and height of the drawn cell.
6625 The thickness of the drawn cell.
6627 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6628 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6632 @subsection Examples
6636 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6638 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6642 Draw a white 3x3 grid with an opacity of 50%:
6644 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6651 Draw a text string or text from a specified file on top of a video, using the
6652 libfreetype library.
6654 To enable compilation of this filter, you need to configure FFmpeg with
6655 @code{--enable-libfreetype}.
6656 To enable default font fallback and the @var{font} option you need to
6657 configure FFmpeg with @code{--enable-libfontconfig}.
6658 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6659 @code{--enable-libfribidi}.
6663 It accepts the following parameters:
6668 Used to draw a box around text using the background color.
6669 The value must be either 1 (enable) or 0 (disable).
6670 The default value of @var{box} is 0.
6673 Set the width of the border to be drawn around the box using @var{boxcolor}.
6674 The default value of @var{boxborderw} is 0.
6677 The color to be used for drawing box around text. For the syntax of this
6678 option, check the "Color" section in the ffmpeg-utils manual.
6680 The default value of @var{boxcolor} is "white".
6683 Set the width of the border to be drawn around the text using @var{bordercolor}.
6684 The default value of @var{borderw} is 0.
6687 Set the color to be used for drawing border around text. For the syntax of this
6688 option, check the "Color" section in the ffmpeg-utils manual.
6690 The default value of @var{bordercolor} is "black".
6693 Select how the @var{text} is expanded. Can be either @code{none},
6694 @code{strftime} (deprecated) or
6695 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6699 If true, check and fix text coords to avoid clipping.
6702 The color to be used for drawing fonts. For the syntax of this option, check
6703 the "Color" section in the ffmpeg-utils manual.
6705 The default value of @var{fontcolor} is "black".
6707 @item fontcolor_expr
6708 String which is expanded the same way as @var{text} to obtain dynamic
6709 @var{fontcolor} value. By default this option has empty value and is not
6710 processed. When this option is set, it overrides @var{fontcolor} option.
6713 The font family to be used for drawing text. By default Sans.
6716 The font file to be used for drawing text. The path must be included.
6717 This parameter is mandatory if the fontconfig support is disabled.
6720 This option does not exist, please see the timeline system
6723 Draw the text applying alpha blending. The value can
6724 be a number between 0.0 and 1.0.
6725 The expression accepts the same variables @var{x, y} as well.
6726 The default value is 1.
6727 Please see @var{fontcolor_expr}.
6730 The font size to be used for drawing text.
6731 The default value of @var{fontsize} is 16.
6734 If set to 1, attempt to shape the text (for example, reverse the order of
6735 right-to-left text and join Arabic characters) before drawing it.
6736 Otherwise, just draw the text exactly as given.
6737 By default 1 (if supported).
6740 The flags to be used for loading the fonts.
6742 The flags map the corresponding flags supported by libfreetype, and are
6743 a combination of the following values:
6750 @item vertical_layout
6751 @item force_autohint
6754 @item ignore_global_advance_width
6756 @item ignore_transform
6762 Default value is "default".
6764 For more information consult the documentation for the FT_LOAD_*
6768 The color to be used for drawing a shadow behind the drawn text. For the
6769 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6771 The default value of @var{shadowcolor} is "black".
6775 The x and y offsets for the text shadow position with respect to the
6776 position of the text. They can be either positive or negative
6777 values. The default value for both is "0".
6780 The starting frame number for the n/frame_num variable. The default value
6784 The size in number of spaces to use for rendering the tab.
6788 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6789 format. It can be used with or without text parameter. @var{timecode_rate}
6790 option must be specified.
6792 @item timecode_rate, rate, r
6793 Set the timecode frame rate (timecode only).
6796 The text string to be drawn. The text must be a sequence of UTF-8
6798 This parameter is mandatory if no file is specified with the parameter
6802 A text file containing text to be drawn. The text must be a sequence
6803 of UTF-8 encoded characters.
6805 This parameter is mandatory if no text string is specified with the
6806 parameter @var{text}.
6808 If both @var{text} and @var{textfile} are specified, an error is thrown.
6811 If set to 1, the @var{textfile} will be reloaded before each frame.
6812 Be sure to update it atomically, or it may be read partially, or even fail.
6816 The expressions which specify the offsets where text will be drawn
6817 within the video frame. They are relative to the top/left border of the
6820 The default value of @var{x} and @var{y} is "0".
6822 See below for the list of accepted constants and functions.
6825 The parameters for @var{x} and @var{y} are expressions containing the
6826 following constants and functions:
6830 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6834 horizontal and vertical chroma subsample values. For example for the
6835 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6838 the height of each text line
6846 @item max_glyph_a, ascent
6847 the maximum distance from the baseline to the highest/upper grid
6848 coordinate used to place a glyph outline point, for all the rendered
6850 It is a positive value, due to the grid's orientation with the Y axis
6853 @item max_glyph_d, descent
6854 the maximum distance from the baseline to the lowest grid coordinate
6855 used to place a glyph outline point, for all the rendered glyphs.
6856 This is a negative value, due to the grid's orientation, with the Y axis
6860 maximum glyph height, that is the maximum height for all the glyphs
6861 contained in the rendered text, it is equivalent to @var{ascent} -
6865 maximum glyph width, that is the maximum width for all the glyphs
6866 contained in the rendered text
6869 the number of input frame, starting from 0
6871 @item rand(min, max)
6872 return a random number included between @var{min} and @var{max}
6875 The input sample aspect ratio.
6878 timestamp expressed in seconds, NAN if the input timestamp is unknown
6881 the height of the rendered text
6884 the width of the rendered text
6888 the x and y offset coordinates where the text is drawn.
6890 These parameters allow the @var{x} and @var{y} expressions to refer
6891 each other, so you can for example specify @code{y=x/dar}.
6894 @anchor{drawtext_expansion}
6895 @subsection Text expansion
6897 If @option{expansion} is set to @code{strftime},
6898 the filter recognizes strftime() sequences in the provided text and
6899 expands them accordingly. Check the documentation of strftime(). This
6900 feature is deprecated.
6902 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6904 If @option{expansion} is set to @code{normal} (which is the default),
6905 the following expansion mechanism is used.
6907 The backslash character @samp{\}, followed by any character, always expands to
6908 the second character.
6910 Sequences of the form @code{%@{...@}} are expanded. The text between the
6911 braces is a function name, possibly followed by arguments separated by ':'.
6912 If the arguments contain special characters or delimiters (':' or '@}'),
6913 they should be escaped.
6915 Note that they probably must also be escaped as the value for the
6916 @option{text} option in the filter argument string and as the filter
6917 argument in the filtergraph description, and possibly also for the shell,
6918 that makes up to four levels of escaping; using a text file avoids these
6921 The following functions are available:
6926 The expression evaluation result.
6928 It must take one argument specifying the expression to be evaluated,
6929 which accepts the same constants and functions as the @var{x} and
6930 @var{y} values. Note that not all constants should be used, for
6931 example the text size is not known when evaluating the expression, so
6932 the constants @var{text_w} and @var{text_h} will have an undefined
6935 @item expr_int_format, eif
6936 Evaluate the expression's value and output as formatted integer.
6938 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6939 The second argument specifies the output format. Allowed values are @samp{x},
6940 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6941 @code{printf} function.
6942 The third parameter is optional and sets the number of positions taken by the output.
6943 It can be used to add padding with zeros from the left.
6946 The time at which the filter is running, expressed in UTC.
6947 It can accept an argument: a strftime() format string.
6950 The time at which the filter is running, expressed in the local time zone.
6951 It can accept an argument: a strftime() format string.
6954 Frame metadata. Takes one or two arguments.
6956 The first argument is mandatory and specifies the metadata key.
6958 The second argument is optional and specifies a default value, used when the
6959 metadata key is not found or empty.
6962 The frame number, starting from 0.
6965 A 1 character description of the current picture type.
6968 The timestamp of the current frame.
6969 It can take up to three arguments.
6971 The first argument is the format of the timestamp; it defaults to @code{flt}
6972 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6973 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6974 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
6975 @code{localtime} stands for the timestamp of the frame formatted as
6976 local time zone time.
6978 The second argument is an offset added to the timestamp.
6980 If the format is set to @code{localtime} or @code{gmtime},
6981 a third argument may be supplied: a strftime() format string.
6982 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
6985 @subsection Examples
6989 Draw "Test Text" with font FreeSerif, using the default values for the
6990 optional parameters.
6993 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
6997 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
6998 and y=50 (counting from the top-left corner of the screen), text is
6999 yellow with a red box around it. Both the text and the box have an
7003 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7004 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7007 Note that the double quotes are not necessary if spaces are not used
7008 within the parameter list.
7011 Show the text at the center of the video frame:
7013 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7017 Show the text at a random position, switching to a new position every 30 seconds:
7019 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)"
7023 Show a text line sliding from right to left in the last row of the video
7024 frame. The file @file{LONG_LINE} is assumed to contain a single line
7027 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7031 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7033 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7037 Draw a single green letter "g", at the center of the input video.
7038 The glyph baseline is placed at half screen height.
7040 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7044 Show text for 1 second every 3 seconds:
7046 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7050 Use fontconfig to set the font. Note that the colons need to be escaped.
7052 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7056 Print the date of a real-time encoding (see strftime(3)):
7058 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7062 Show text fading in and out (appearing/disappearing):
7065 DS=1.0 # display start
7066 DE=10.0 # display end
7067 FID=1.5 # fade in duration
7068 FOD=5 # fade out duration
7069 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 @}"
7074 For more information about libfreetype, check:
7075 @url{http://www.freetype.org/}.
7077 For more information about fontconfig, check:
7078 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7080 For more information about libfribidi, check:
7081 @url{http://fribidi.org/}.
7085 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7087 The filter accepts the following options:
7092 Set low and high threshold values used by the Canny thresholding
7095 The high threshold selects the "strong" edge pixels, which are then
7096 connected through 8-connectivity with the "weak" edge pixels selected
7097 by the low threshold.
7099 @var{low} and @var{high} threshold values must be chosen in the range
7100 [0,1], and @var{low} should be lesser or equal to @var{high}.
7102 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7106 Define the drawing mode.
7110 Draw white/gray wires on black background.
7113 Mix the colors to create a paint/cartoon effect.
7116 Default value is @var{wires}.
7119 @subsection Examples
7123 Standard edge detection with custom values for the hysteresis thresholding:
7125 edgedetect=low=0.1:high=0.4
7129 Painting effect without thresholding:
7131 edgedetect=mode=colormix:high=0
7136 Set brightness, contrast, saturation and approximate gamma adjustment.
7138 The filter accepts the following options:
7142 Set the contrast expression. The value must be a float value in range
7143 @code{-2.0} to @code{2.0}. The default value is "1".
7146 Set the brightness expression. The value must be a float value in
7147 range @code{-1.0} to @code{1.0}. The default value is "0".
7150 Set the saturation expression. The value must be a float in
7151 range @code{0.0} to @code{3.0}. The default value is "1".
7154 Set the gamma expression. The value must be a float in range
7155 @code{0.1} to @code{10.0}. The default value is "1".
7158 Set the gamma expression for red. The value must be a float in
7159 range @code{0.1} to @code{10.0}. The default value is "1".
7162 Set the gamma expression for green. The value must be a float in range
7163 @code{0.1} to @code{10.0}. The default value is "1".
7166 Set the gamma expression for blue. The value must be a float in range
7167 @code{0.1} to @code{10.0}. The default value is "1".
7170 Set the gamma weight expression. It can be used to reduce the effect
7171 of a high gamma value on bright image areas, e.g. keep them from
7172 getting overamplified and just plain white. The value must be a float
7173 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7174 gamma correction all the way down while @code{1.0} leaves it at its
7175 full strength. Default is "1".
7178 Set when the expressions for brightness, contrast, saturation and
7179 gamma expressions are evaluated.
7181 It accepts the following values:
7184 only evaluate expressions once during the filter initialization or
7185 when a command is processed
7188 evaluate expressions for each incoming frame
7191 Default value is @samp{init}.
7194 The expressions accept the following parameters:
7197 frame count of the input frame starting from 0
7200 byte position of the corresponding packet in the input file, NAN if
7204 frame rate of the input video, NAN if the input frame rate is unknown
7207 timestamp expressed in seconds, NAN if the input timestamp is unknown
7210 @subsection Commands
7211 The filter supports the following commands:
7215 Set the contrast expression.
7218 Set the brightness expression.
7221 Set the saturation expression.
7224 Set the gamma expression.
7227 Set the gamma_r expression.
7230 Set gamma_g expression.
7233 Set gamma_b expression.
7236 Set gamma_weight expression.
7238 The command accepts the same syntax of the corresponding option.
7240 If the specified expression is not valid, it is kept at its current
7247 Apply erosion effect to the video.
7249 This filter replaces the pixel by the local(3x3) minimum.
7251 It accepts the following options:
7258 Limit the maximum change for each plane, default is 65535.
7259 If 0, plane will remain unchanged.
7262 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7265 Flags to local 3x3 coordinates maps like this:
7272 @section extractplanes
7274 Extract color channel components from input video stream into
7275 separate grayscale video streams.
7277 The filter accepts the following option:
7281 Set plane(s) to extract.
7283 Available values for planes are:
7294 Choosing planes not available in the input will result in an error.
7295 That means you cannot select @code{r}, @code{g}, @code{b} planes
7296 with @code{y}, @code{u}, @code{v} planes at same time.
7299 @subsection Examples
7303 Extract luma, u and v color channel component from input video frame
7304 into 3 grayscale outputs:
7306 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
7312 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7314 For each input image, the filter will compute the optimal mapping from
7315 the input to the output given the codebook length, that is the number
7316 of distinct output colors.
7318 This filter accepts the following options.
7321 @item codebook_length, l
7322 Set codebook length. The value must be a positive integer, and
7323 represents the number of distinct output colors. Default value is 256.
7326 Set the maximum number of iterations to apply for computing the optimal
7327 mapping. The higher the value the better the result and the higher the
7328 computation time. Default value is 1.
7331 Set a random seed, must be an integer included between 0 and
7332 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7333 will try to use a good random seed on a best effort basis.
7336 Set pal8 output pixel format. This option does not work with codebook
7337 length greater than 256.
7342 Apply a fade-in/out effect to the input video.
7344 It accepts the following parameters:
7348 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7350 Default is @code{in}.
7352 @item start_frame, s
7353 Specify the number of the frame to start applying the fade
7354 effect at. Default is 0.
7357 The number of frames that the fade effect lasts. At the end of the
7358 fade-in effect, the output video will have the same intensity as the input video.
7359 At the end of the fade-out transition, the output video will be filled with the
7360 selected @option{color}.
7364 If set to 1, fade only alpha channel, if one exists on the input.
7367 @item start_time, st
7368 Specify the timestamp (in seconds) of the frame to start to apply the fade
7369 effect. If both start_frame and start_time are specified, the fade will start at
7370 whichever comes last. Default is 0.
7373 The number of seconds for which the fade effect has to last. At the end of the
7374 fade-in effect the output video will have the same intensity as the input video,
7375 at the end of the fade-out transition the output video will be filled with the
7376 selected @option{color}.
7377 If both duration and nb_frames are specified, duration is used. Default is 0
7378 (nb_frames is used by default).
7381 Specify the color of the fade. Default is "black".
7384 @subsection Examples
7388 Fade in the first 30 frames of video:
7393 The command above is equivalent to:
7399 Fade out the last 45 frames of a 200-frame video:
7402 fade=type=out:start_frame=155:nb_frames=45
7406 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7408 fade=in:0:25, fade=out:975:25
7412 Make the first 5 frames yellow, then fade in from frame 5-24:
7414 fade=in:5:20:color=yellow
7418 Fade in alpha over first 25 frames of video:
7420 fade=in:0:25:alpha=1
7424 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7426 fade=t=in:st=5.5:d=0.5
7432 Apply arbitrary expressions to samples in frequency domain
7436 Adjust the dc value (gain) of the luma plane of the image. The filter
7437 accepts an integer value in range @code{0} to @code{1000}. The default
7438 value is set to @code{0}.
7441 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7442 filter accepts an integer value in range @code{0} to @code{1000}. The
7443 default value is set to @code{0}.
7446 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7447 filter accepts an integer value in range @code{0} to @code{1000}. The
7448 default value is set to @code{0}.
7451 Set the frequency domain weight expression for the luma plane.
7454 Set the frequency domain weight expression for the 1st chroma plane.
7457 Set the frequency domain weight expression for the 2nd chroma plane.
7459 The filter accepts the following variables:
7462 The coordinates of the current sample.
7466 The width and height of the image.
7469 @subsection Examples
7475 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7481 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7487 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7493 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7500 Extract a single field from an interlaced image using stride
7501 arithmetic to avoid wasting CPU time. The output frames are marked as
7504 The filter accepts the following options:
7508 Specify whether to extract the top (if the value is @code{0} or
7509 @code{top}) or the bottom field (if the value is @code{1} or
7515 Create new frames by copying the top and bottom fields from surrounding frames
7516 supplied as numbers by the hint file.
7520 Set file containing hints: absolute/relative frame numbers.
7522 There must be one line for each frame in a clip. Each line must contain two
7523 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7524 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7525 is current frame number for @code{absolute} mode or out of [-1, 1] range
7526 for @code{relative} mode. First number tells from which frame to pick up top
7527 field and second number tells from which frame to pick up bottom field.
7529 If optionally followed by @code{+} output frame will be marked as interlaced,
7530 else if followed by @code{-} output frame will be marked as progressive, else
7531 it will be marked same as input frame.
7532 If line starts with @code{#} or @code{;} that line is skipped.
7535 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7538 Example of first several lines of @code{hint} file for @code{relative} mode:
7541 1,0 - # second frame, use third's frame top field and second's frame bottom field
7542 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7559 Field matching filter for inverse telecine. It is meant to reconstruct the
7560 progressive frames from a telecined stream. The filter does not drop duplicated
7561 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7562 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7564 The separation of the field matching and the decimation is notably motivated by
7565 the possibility of inserting a de-interlacing filter fallback between the two.
7566 If the source has mixed telecined and real interlaced content,
7567 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7568 But these remaining combed frames will be marked as interlaced, and thus can be
7569 de-interlaced by a later filter such as @ref{yadif} before decimation.
7571 In addition to the various configuration options, @code{fieldmatch} can take an
7572 optional second stream, activated through the @option{ppsrc} option. If
7573 enabled, the frames reconstruction will be based on the fields and frames from
7574 this second stream. This allows the first input to be pre-processed in order to
7575 help the various algorithms of the filter, while keeping the output lossless
7576 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7577 or brightness/contrast adjustments can help.
7579 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7580 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7581 which @code{fieldmatch} is based on. While the semantic and usage are very
7582 close, some behaviour and options names can differ.
7584 The @ref{decimate} filter currently only works for constant frame rate input.
7585 If your input has mixed telecined (30fps) and progressive content with a lower
7586 framerate like 24fps use the following filterchain to produce the necessary cfr
7587 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7589 The filter accepts the following options:
7593 Specify the assumed field order of the input stream. Available values are:
7597 Auto detect parity (use FFmpeg's internal parity value).
7599 Assume bottom field first.
7601 Assume top field first.
7604 Note that it is sometimes recommended not to trust the parity announced by the
7607 Default value is @var{auto}.
7610 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7611 sense that it won't risk creating jerkiness due to duplicate frames when
7612 possible, but if there are bad edits or blended fields it will end up
7613 outputting combed frames when a good match might actually exist. On the other
7614 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7615 but will almost always find a good frame if there is one. The other values are
7616 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7617 jerkiness and creating duplicate frames versus finding good matches in sections
7618 with bad edits, orphaned fields, blended fields, etc.
7620 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7622 Available values are:
7626 2-way matching (p/c)
7628 2-way matching, and trying 3rd match if still combed (p/c + n)
7630 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7632 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7633 still combed (p/c + n + u/b)
7635 3-way matching (p/c/n)
7637 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7638 detected as combed (p/c/n + u/b)
7641 The parenthesis at the end indicate the matches that would be used for that
7642 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7645 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7648 Default value is @var{pc_n}.
7651 Mark the main input stream as a pre-processed input, and enable the secondary
7652 input stream as the clean source to pick the fields from. See the filter
7653 introduction for more details. It is similar to the @option{clip2} feature from
7656 Default value is @code{0} (disabled).
7659 Set the field to match from. It is recommended to set this to the same value as
7660 @option{order} unless you experience matching failures with that setting. In
7661 certain circumstances changing the field that is used to match from can have a
7662 large impact on matching performance. Available values are:
7666 Automatic (same value as @option{order}).
7668 Match from the bottom field.
7670 Match from the top field.
7673 Default value is @var{auto}.
7676 Set whether or not chroma is included during the match comparisons. In most
7677 cases it is recommended to leave this enabled. You should set this to @code{0}
7678 only if your clip has bad chroma problems such as heavy rainbowing or other
7679 artifacts. Setting this to @code{0} could also be used to speed things up at
7680 the cost of some accuracy.
7682 Default value is @code{1}.
7686 These define an exclusion band which excludes the lines between @option{y0} and
7687 @option{y1} from being included in the field matching decision. An exclusion
7688 band can be used to ignore subtitles, a logo, or other things that may
7689 interfere with the matching. @option{y0} sets the starting scan line and
7690 @option{y1} sets the ending line; all lines in between @option{y0} and
7691 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7692 @option{y0} and @option{y1} to the same value will disable the feature.
7693 @option{y0} and @option{y1} defaults to @code{0}.
7696 Set the scene change detection threshold as a percentage of maximum change on
7697 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7698 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7699 @option{scthresh} is @code{[0.0, 100.0]}.
7701 Default value is @code{12.0}.
7704 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7705 account the combed scores of matches when deciding what match to use as the
7706 final match. Available values are:
7710 No final matching based on combed scores.
7712 Combed scores are only used when a scene change is detected.
7714 Use combed scores all the time.
7717 Default is @var{sc}.
7720 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7721 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7722 Available values are:
7726 No forced calculation.
7728 Force p/c/n calculations.
7730 Force p/c/n/u/b calculations.
7733 Default value is @var{none}.
7736 This is the area combing threshold used for combed frame detection. This
7737 essentially controls how "strong" or "visible" combing must be to be detected.
7738 Larger values mean combing must be more visible and smaller values mean combing
7739 can be less visible or strong and still be detected. Valid settings are from
7740 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7741 be detected as combed). This is basically a pixel difference value. A good
7742 range is @code{[8, 12]}.
7744 Default value is @code{9}.
7747 Sets whether or not chroma is considered in the combed frame decision. Only
7748 disable this if your source has chroma problems (rainbowing, etc.) that are
7749 causing problems for the combed frame detection with chroma enabled. Actually,
7750 using @option{chroma}=@var{0} is usually more reliable, except for the case
7751 where there is chroma only combing in the source.
7753 Default value is @code{0}.
7757 Respectively set the x-axis and y-axis size of the window used during combed
7758 frame detection. This has to do with the size of the area in which
7759 @option{combpel} pixels are required to be detected as combed for a frame to be
7760 declared combed. See the @option{combpel} parameter description for more info.
7761 Possible values are any number that is a power of 2 starting at 4 and going up
7764 Default value is @code{16}.
7767 The number of combed pixels inside any of the @option{blocky} by
7768 @option{blockx} size blocks on the frame for the frame to be detected as
7769 combed. While @option{cthresh} controls how "visible" the combing must be, this
7770 setting controls "how much" combing there must be in any localized area (a
7771 window defined by the @option{blockx} and @option{blocky} settings) on the
7772 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7773 which point no frames will ever be detected as combed). This setting is known
7774 as @option{MI} in TFM/VFM vocabulary.
7776 Default value is @code{80}.
7779 @anchor{p/c/n/u/b meaning}
7780 @subsection p/c/n/u/b meaning
7782 @subsubsection p/c/n
7784 We assume the following telecined stream:
7787 Top fields: 1 2 2 3 4
7788 Bottom fields: 1 2 3 4 4
7791 The numbers correspond to the progressive frame the fields relate to. Here, the
7792 first two frames are progressive, the 3rd and 4th are combed, and so on.
7794 When @code{fieldmatch} is configured to run a matching from bottom
7795 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7800 B 1 2 3 4 4 <-- matching reference
7809 As a result of the field matching, we can see that some frames get duplicated.
7810 To perform a complete inverse telecine, you need to rely on a decimation filter
7811 after this operation. See for instance the @ref{decimate} filter.
7813 The same operation now matching from top fields (@option{field}=@var{top})
7818 T 1 2 2 3 4 <-- matching reference
7828 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7829 basically, they refer to the frame and field of the opposite parity:
7832 @item @var{p} matches the field of the opposite parity in the previous frame
7833 @item @var{c} matches the field of the opposite parity in the current frame
7834 @item @var{n} matches the field of the opposite parity in the next frame
7839 The @var{u} and @var{b} matching are a bit special in the sense that they match
7840 from the opposite parity flag. In the following examples, we assume that we are
7841 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7842 'x' is placed above and below each matched fields.
7844 With bottom matching (@option{field}=@var{bottom}):
7849 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7850 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7858 With top matching (@option{field}=@var{top}):
7863 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7864 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7872 @subsection Examples
7874 Simple IVTC of a top field first telecined stream:
7876 fieldmatch=order=tff:combmatch=none, decimate
7879 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7881 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7886 Transform the field order of the input video.
7888 It accepts the following parameters:
7893 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7894 for bottom field first.
7897 The default value is @samp{tff}.
7899 The transformation is done by shifting the picture content up or down
7900 by one line, and filling the remaining line with appropriate picture content.
7901 This method is consistent with most broadcast field order converters.
7903 If the input video is not flagged as being interlaced, or it is already
7904 flagged as being of the required output field order, then this filter does
7905 not alter the incoming video.
7907 It is very useful when converting to or from PAL DV material,
7908 which is bottom field first.
7912 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7915 @section fifo, afifo
7917 Buffer input images and send them when they are requested.
7919 It is mainly useful when auto-inserted by the libavfilter
7922 It does not take parameters.
7926 Find a rectangular object
7928 It accepts the following options:
7932 Filepath of the object image, needs to be in gray8.
7935 Detection threshold, default is 0.5.
7938 Number of mipmaps, default is 3.
7940 @item xmin, ymin, xmax, ymax
7941 Specifies the rectangle in which to search.
7944 @subsection Examples
7948 Generate a representative palette of a given video using @command{ffmpeg}:
7950 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7956 Cover a rectangular object
7958 It accepts the following options:
7962 Filepath of the optional cover image, needs to be in yuv420.
7967 It accepts the following values:
7970 cover it by the supplied image
7972 cover it by interpolating the surrounding pixels
7975 Default value is @var{blur}.
7978 @subsection Examples
7982 Generate a representative palette of a given video using @command{ffmpeg}:
7984 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7991 Convert the input video to one of the specified pixel formats.
7992 Libavfilter will try to pick one that is suitable as input to
7995 It accepts the following parameters:
7999 A '|'-separated list of pixel format names, such as
8000 "pix_fmts=yuv420p|monow|rgb24".
8004 @subsection Examples
8008 Convert the input video to the @var{yuv420p} format
8010 format=pix_fmts=yuv420p
8013 Convert the input video to any of the formats in the list
8015 format=pix_fmts=yuv420p|yuv444p|yuv410p
8022 Convert the video to specified constant frame rate by duplicating or dropping
8023 frames as necessary.
8025 It accepts the following parameters:
8029 The desired output frame rate. The default is @code{25}.
8034 Possible values are:
8037 zero round towards 0
8041 round towards -infinity
8043 round towards +infinity
8047 The default is @code{near}.
8050 Assume the first PTS should be the given value, in seconds. This allows for
8051 padding/trimming at the start of stream. By default, no assumption is made
8052 about the first frame's expected PTS, so no padding or trimming is done.
8053 For example, this could be set to 0 to pad the beginning with duplicates of
8054 the first frame if a video stream starts after the audio stream or to trim any
8055 frames with a negative PTS.
8059 Alternatively, the options can be specified as a flat string:
8060 @var{fps}[:@var{round}].
8062 See also the @ref{setpts} filter.
8064 @subsection Examples
8068 A typical usage in order to set the fps to 25:
8074 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8076 fps=fps=film:round=near
8082 Pack two different video streams into a stereoscopic video, setting proper
8083 metadata on supported codecs. The two views should have the same size and
8084 framerate and processing will stop when the shorter video ends. Please note
8085 that you may conveniently adjust view properties with the @ref{scale} and
8088 It accepts the following parameters:
8092 The desired packing format. Supported values are:
8097 The views are next to each other (default).
8100 The views are on top of each other.
8103 The views are packed by line.
8106 The views are packed by column.
8109 The views are temporally interleaved.
8118 # Convert left and right views into a frame-sequential video
8119 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8121 # Convert views into a side-by-side video with the same output resolution as the input
8122 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
8127 Change the frame rate by interpolating new video output frames from the source
8130 This filter is not designed to function correctly with interlaced media. If
8131 you wish to change the frame rate of interlaced media then you are required
8132 to deinterlace before this filter and re-interlace after this filter.
8134 A description of the accepted options follows.
8138 Specify the output frames per second. This option can also be specified
8139 as a value alone. The default is @code{50}.
8142 Specify the start of a range where the output frame will be created as a
8143 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8144 the default is @code{15}.
8147 Specify the end of a range where the output frame will be created as a
8148 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8149 the default is @code{240}.
8152 Specify the level at which a scene change is detected as a value between
8153 0 and 100 to indicate a new scene; a low value reflects a low
8154 probability for the current frame to introduce a new scene, while a higher
8155 value means the current frame is more likely to be one.
8156 The default is @code{7}.
8159 Specify flags influencing the filter process.
8161 Available value for @var{flags} is:
8164 @item scene_change_detect, scd
8165 Enable scene change detection using the value of the option @var{scene}.
8166 This flag is enabled by default.
8172 Select one frame every N-th frame.
8174 This filter accepts the following option:
8177 Select frame after every @code{step} frames.
8178 Allowed values are positive integers higher than 0. Default value is @code{1}.
8184 Apply a frei0r effect to the input video.
8186 To enable the compilation of this filter, you need to install the frei0r
8187 header and configure FFmpeg with @code{--enable-frei0r}.
8189 It accepts the following parameters:
8194 The name of the frei0r effect to load. If the environment variable
8195 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8196 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8197 Otherwise, the standard frei0r paths are searched, in this order:
8198 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8199 @file{/usr/lib/frei0r-1/}.
8202 A '|'-separated list of parameters to pass to the frei0r effect.
8206 A frei0r effect parameter can be a boolean (its value is either
8207 "y" or "n"), a double, a color (specified as
8208 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8209 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8210 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8211 @var{X} and @var{Y} are floating point numbers) and/or a string.
8213 The number and types of parameters depend on the loaded effect. If an
8214 effect parameter is not specified, the default value is set.
8216 @subsection Examples
8220 Apply the distort0r effect, setting the first two double parameters:
8222 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8226 Apply the colordistance effect, taking a color as the first parameter:
8228 frei0r=colordistance:0.2/0.3/0.4
8229 frei0r=colordistance:violet
8230 frei0r=colordistance:0x112233
8234 Apply the perspective effect, specifying the top left and top right image
8237 frei0r=perspective:0.2/0.2|0.8/0.2
8241 For more information, see
8242 @url{http://frei0r.dyne.org}
8246 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8248 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8249 processing filter, one of them is performed once per block, not per pixel.
8250 This allows for much higher speed.
8252 The filter accepts the following options:
8256 Set quality. This option defines the number of levels for averaging. It accepts
8257 an integer in the range 4-5. Default value is @code{4}.
8260 Force a constant quantization parameter. It accepts an integer in range 0-63.
8261 If not set, the filter will use the QP from the video stream (if available).
8264 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8265 more details but also more artifacts, while higher values make the image smoother
8266 but also blurrier. Default value is @code{0} − PSNR optimal.
8269 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8270 option may cause flicker since the B-Frames have often larger QP. Default is
8271 @code{0} (not enabled).
8277 Apply Gaussian blur filter.
8279 The filter accepts the following options:
8283 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8286 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8289 Set which planes to filter. By default all planes are filtered.
8292 Set vertical sigma, if negative it will be same as @code{sigma}.
8293 Default is @code{-1}.
8298 The filter accepts the following options:
8302 Set the luminance expression.
8304 Set the chrominance blue expression.
8306 Set the chrominance red expression.
8308 Set the alpha expression.
8310 Set the red expression.
8312 Set the green expression.
8314 Set the blue expression.
8317 The colorspace is selected according to the specified options. If one
8318 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8319 options is specified, the filter will automatically select a YCbCr
8320 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8321 @option{blue_expr} options is specified, it will select an RGB
8324 If one of the chrominance expression is not defined, it falls back on the other
8325 one. If no alpha expression is specified it will evaluate to opaque value.
8326 If none of chrominance expressions are specified, they will evaluate
8327 to the luminance expression.
8329 The expressions can use the following variables and functions:
8333 The sequential number of the filtered frame, starting from @code{0}.
8337 The coordinates of the current sample.
8341 The width and height of the image.
8345 Width and height scale depending on the currently filtered plane. It is the
8346 ratio between the corresponding luma plane number of pixels and the current
8347 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8348 @code{0.5,0.5} for chroma planes.
8351 Time of the current frame, expressed in seconds.
8354 Return the value of the pixel at location (@var{x},@var{y}) of the current
8358 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8362 Return the value of the pixel at location (@var{x},@var{y}) of the
8363 blue-difference chroma plane. Return 0 if there is no such plane.
8366 Return the value of the pixel at location (@var{x},@var{y}) of the
8367 red-difference chroma plane. Return 0 if there is no such plane.
8372 Return the value of the pixel at location (@var{x},@var{y}) of the
8373 red/green/blue component. Return 0 if there is no such component.
8376 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8377 plane. Return 0 if there is no such plane.
8380 For functions, if @var{x} and @var{y} are outside the area, the value will be
8381 automatically clipped to the closer edge.
8383 @subsection Examples
8387 Flip the image horizontally:
8393 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8394 wavelength of 100 pixels:
8396 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8400 Generate a fancy enigmatic moving light:
8402 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
8406 Generate a quick emboss effect:
8408 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8412 Modify RGB components depending on pixel position:
8414 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8418 Create a radial gradient that is the same size as the input (also see
8419 the @ref{vignette} filter):
8421 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8427 Fix the banding artifacts that are sometimes introduced into nearly flat
8428 regions by truncation to 8-bit color depth.
8429 Interpolate the gradients that should go where the bands are, and
8432 It is designed for playback only. Do not use it prior to
8433 lossy compression, because compression tends to lose the dither and
8434 bring back the bands.
8436 It accepts the following parameters:
8441 The maximum amount by which the filter will change any one pixel. This is also
8442 the threshold for detecting nearly flat regions. Acceptable values range from
8443 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8447 The neighborhood to fit the gradient to. A larger radius makes for smoother
8448 gradients, but also prevents the filter from modifying the pixels near detailed
8449 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8450 values will be clipped to the valid range.
8454 Alternatively, the options can be specified as a flat string:
8455 @var{strength}[:@var{radius}]
8457 @subsection Examples
8461 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8467 Specify radius, omitting the strength (which will fall-back to the default
8478 Apply a Hald CLUT to a video stream.
8480 First input is the video stream to process, and second one is the Hald CLUT.
8481 The Hald CLUT input can be a simple picture or a complete video stream.
8483 The filter accepts the following options:
8487 Force termination when the shortest input terminates. Default is @code{0}.
8489 Continue applying the last CLUT after the end of the stream. A value of
8490 @code{0} disable the filter after the last frame of the CLUT is reached.
8491 Default is @code{1}.
8494 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8495 filters share the same internals).
8497 More information about the Hald CLUT can be found on Eskil Steenberg's website
8498 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8500 @subsection Workflow examples
8502 @subsubsection Hald CLUT video stream
8504 Generate an identity Hald CLUT stream altered with various effects:
8506 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
8509 Note: make sure you use a lossless codec.
8511 Then use it with @code{haldclut} to apply it on some random stream:
8513 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8516 The Hald CLUT will be applied to the 10 first seconds (duration of
8517 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8518 to the remaining frames of the @code{mandelbrot} stream.
8520 @subsubsection Hald CLUT with preview
8522 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8523 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8524 biggest possible square starting at the top left of the picture. The remaining
8525 padding pixels (bottom or right) will be ignored. This area can be used to add
8526 a preview of the Hald CLUT.
8528 Typically, the following generated Hald CLUT will be supported by the
8529 @code{haldclut} filter:
8532 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8533 pad=iw+320 [padded_clut];
8534 smptebars=s=320x256, split [a][b];
8535 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8536 [main][b] overlay=W-320" -frames:v 1 clut.png
8539 It contains the original and a preview of the effect of the CLUT: SMPTE color
8540 bars are displayed on the right-top, and below the same color bars processed by
8543 Then, the effect of this Hald CLUT can be visualized with:
8545 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8550 Flip the input video horizontally.
8552 For example, to horizontally flip the input video with @command{ffmpeg}:
8554 ffmpeg -i in.avi -vf "hflip" out.avi
8558 This filter applies a global color histogram equalization on a
8561 It can be used to correct video that has a compressed range of pixel
8562 intensities. The filter redistributes the pixel intensities to
8563 equalize their distribution across the intensity range. It may be
8564 viewed as an "automatically adjusting contrast filter". This filter is
8565 useful only for correcting degraded or poorly captured source
8568 The filter accepts the following options:
8572 Determine the amount of equalization to be applied. As the strength
8573 is reduced, the distribution of pixel intensities more-and-more
8574 approaches that of the input frame. The value must be a float number
8575 in the range [0,1] and defaults to 0.200.
8578 Set the maximum intensity that can generated and scale the output
8579 values appropriately. The strength should be set as desired and then
8580 the intensity can be limited if needed to avoid washing-out. The value
8581 must be a float number in the range [0,1] and defaults to 0.210.
8584 Set the antibanding level. If enabled the filter will randomly vary
8585 the luminance of output pixels by a small amount to avoid banding of
8586 the histogram. Possible values are @code{none}, @code{weak} or
8587 @code{strong}. It defaults to @code{none}.
8592 Compute and draw a color distribution histogram for the input video.
8594 The computed histogram is a representation of the color component
8595 distribution in an image.
8597 Standard histogram displays the color components distribution in an image.
8598 Displays color graph for each color component. Shows distribution of
8599 the Y, U, V, A or R, G, B components, depending on input format, in the
8600 current frame. Below each graph a color component scale meter is shown.
8602 The filter accepts the following options:
8606 Set height of level. Default value is @code{200}.
8607 Allowed range is [50, 2048].
8610 Set height of color scale. Default value is @code{12}.
8611 Allowed range is [0, 40].
8615 It accepts the following values:
8618 Per color component graphs are placed below each other.
8621 Presents information identical to that in the @code{parade}, except
8622 that the graphs representing color components are superimposed directly
8625 Default is @code{parade}.
8628 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8629 Default is @code{linear}.
8632 Set what color components to display.
8633 Default is @code{7}.
8636 Set foreground opacity. Default is @code{0.7}.
8639 Set background opacity. Default is @code{0.5}.
8642 @subsection Examples
8647 Calculate and draw histogram:
8649 ffplay -i input -vf histogram
8657 This is a high precision/quality 3d denoise filter. It aims to reduce
8658 image noise, producing smooth images and making still images really
8659 still. It should enhance compressibility.
8661 It accepts the following optional parameters:
8665 A non-negative floating point number which specifies spatial luma strength.
8668 @item chroma_spatial
8669 A non-negative floating point number which specifies spatial chroma strength.
8670 It defaults to 3.0*@var{luma_spatial}/4.0.
8673 A floating point number which specifies luma temporal strength. It defaults to
8674 6.0*@var{luma_spatial}/4.0.
8677 A floating point number which specifies chroma temporal strength. It defaults to
8678 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8681 @anchor{hwupload_cuda}
8682 @section hwupload_cuda
8684 Upload system memory frames to a CUDA device.
8686 It accepts the following optional parameters:
8690 The number of the CUDA device to use
8695 Apply a high-quality magnification filter designed for pixel art. This filter
8696 was originally created by Maxim Stepin.
8698 It accepts the following option:
8702 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8703 @code{hq3x} and @code{4} for @code{hq4x}.
8704 Default is @code{3}.
8708 Stack input videos horizontally.
8710 All streams must be of same pixel format and of same height.
8712 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8713 to create same output.
8715 The filter accept the following option:
8719 Set number of input streams. Default is 2.
8722 If set to 1, force the output to terminate when the shortest input
8723 terminates. Default value is 0.
8728 Modify the hue and/or the saturation of the input.
8730 It accepts the following parameters:
8734 Specify the hue angle as a number of degrees. It accepts an expression,
8735 and defaults to "0".
8738 Specify the saturation in the [-10,10] range. It accepts an expression and
8742 Specify the hue angle as a number of radians. It accepts an
8743 expression, and defaults to "0".
8746 Specify the brightness in the [-10,10] range. It accepts an expression and
8750 @option{h} and @option{H} are mutually exclusive, and can't be
8751 specified at the same time.
8753 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8754 expressions containing the following constants:
8758 frame count of the input frame starting from 0
8761 presentation timestamp of the input frame expressed in time base units
8764 frame rate of the input video, NAN if the input frame rate is unknown
8767 timestamp expressed in seconds, NAN if the input timestamp is unknown
8770 time base of the input video
8773 @subsection Examples
8777 Set the hue to 90 degrees and the saturation to 1.0:
8783 Same command but expressing the hue in radians:
8789 Rotate hue and make the saturation swing between 0
8790 and 2 over a period of 1 second:
8792 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8796 Apply a 3 seconds saturation fade-in effect starting at 0:
8801 The general fade-in expression can be written as:
8803 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8807 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8809 hue="s=max(0\, min(1\, (8-t)/3))"
8812 The general fade-out expression can be written as:
8814 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8819 @subsection Commands
8821 This filter supports the following commands:
8827 Modify the hue and/or the saturation and/or brightness of the input video.
8828 The command accepts the same syntax of the corresponding option.
8830 If the specified expression is not valid, it is kept at its current
8836 Grow first stream into second stream by connecting components.
8837 This allows to build more robust edge masks.
8839 This filter accepts the following options:
8843 Set which planes will be processed as bitmap, unprocessed planes will be
8844 copied from first stream.
8845 By default value 0xf, all planes will be processed.
8848 Set threshold which is used in filtering. If pixel component value is higher than
8849 this value filter algorithm for connecting components is activated.
8850 By default value is 0.
8855 Detect video interlacing type.
8857 This filter tries to detect if the input frames are interlaced, progressive,
8858 top or bottom field first. It will also try to detect fields that are
8859 repeated between adjacent frames (a sign of telecine).
8861 Single frame detection considers only immediately adjacent frames when classifying each frame.
8862 Multiple frame detection incorporates the classification history of previous frames.
8864 The filter will log these metadata values:
8867 @item single.current_frame
8868 Detected type of current frame using single-frame detection. One of:
8869 ``tff'' (top field first), ``bff'' (bottom field first),
8870 ``progressive'', or ``undetermined''
8873 Cumulative number of frames detected as top field first using single-frame detection.
8876 Cumulative number of frames detected as top field first using multiple-frame detection.
8879 Cumulative number of frames detected as bottom field first using single-frame detection.
8881 @item multiple.current_frame
8882 Detected type of current frame using multiple-frame detection. One of:
8883 ``tff'' (top field first), ``bff'' (bottom field first),
8884 ``progressive'', or ``undetermined''
8887 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8889 @item single.progressive
8890 Cumulative number of frames detected as progressive using single-frame detection.
8892 @item multiple.progressive
8893 Cumulative number of frames detected as progressive using multiple-frame detection.
8895 @item single.undetermined
8896 Cumulative number of frames that could not be classified using single-frame detection.
8898 @item multiple.undetermined
8899 Cumulative number of frames that could not be classified using multiple-frame detection.
8901 @item repeated.current_frame
8902 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8904 @item repeated.neither
8905 Cumulative number of frames with no repeated field.
8908 Cumulative number of frames with the top field repeated from the previous frame's top field.
8910 @item repeated.bottom
8911 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8914 The filter accepts the following options:
8918 Set interlacing threshold.
8920 Set progressive threshold.
8922 Threshold for repeated field detection.
8924 Number of frames after which a given frame's contribution to the
8925 statistics is halved (i.e., it contributes only 0.5 to its
8926 classification). The default of 0 means that all frames seen are given
8927 full weight of 1.0 forever.
8928 @item analyze_interlaced_flag
8929 When this is not 0 then idet will use the specified number of frames to determine
8930 if the interlaced flag is accurate, it will not count undetermined frames.
8931 If the flag is found to be accurate it will be used without any further
8932 computations, if it is found to be inaccurate it will be cleared without any
8933 further computations. This allows inserting the idet filter as a low computational
8934 method to clean up the interlaced flag
8939 Deinterleave or interleave fields.
8941 This filter allows one to process interlaced images fields without
8942 deinterlacing them. Deinterleaving splits the input frame into 2
8943 fields (so called half pictures). Odd lines are moved to the top
8944 half of the output image, even lines to the bottom half.
8945 You can process (filter) them independently and then re-interleave them.
8947 The filter accepts the following options:
8951 @item chroma_mode, c
8953 Available values for @var{luma_mode}, @var{chroma_mode} and
8954 @var{alpha_mode} are:
8960 @item deinterleave, d
8961 Deinterleave fields, placing one above the other.
8964 Interleave fields. Reverse the effect of deinterleaving.
8966 Default value is @code{none}.
8969 @item chroma_swap, cs
8970 @item alpha_swap, as
8971 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
8976 Apply inflate effect to the video.
8978 This filter replaces the pixel by the local(3x3) average by taking into account
8979 only values higher than the pixel.
8981 It accepts the following options:
8988 Limit the maximum change for each plane, default is 65535.
8989 If 0, plane will remain unchanged.
8994 Simple interlacing filter from progressive contents. This interleaves upper (or
8995 lower) lines from odd frames with lower (or upper) lines from even frames,
8996 halving the frame rate and preserving image height.
8999 Original Original New Frame
9000 Frame 'j' Frame 'j+1' (tff)
9001 ========== =========== ==================
9002 Line 0 --------------------> Frame 'j' Line 0
9003 Line 1 Line 1 ----> Frame 'j+1' Line 1
9004 Line 2 ---------------------> Frame 'j' Line 2
9005 Line 3 Line 3 ----> Frame 'j+1' Line 3
9007 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9010 It accepts the following optional parameters:
9014 This determines whether the interlaced frame is taken from the even
9015 (tff - default) or odd (bff) lines of the progressive frame.
9018 Enable (default) or disable the vertical lowpass filter to avoid twitter
9019 interlacing and reduce moire patterns.
9024 Deinterlace input video by applying Donald Graft's adaptive kernel
9025 deinterling. Work on interlaced parts of a video to produce
9028 The description of the accepted parameters follows.
9032 Set the threshold which affects the filter's tolerance when
9033 determining if a pixel line must be processed. It must be an integer
9034 in the range [0,255] and defaults to 10. A value of 0 will result in
9035 applying the process on every pixels.
9038 Paint pixels exceeding the threshold value to white if set to 1.
9042 Set the fields order. Swap fields if set to 1, leave fields alone if
9046 Enable additional sharpening if set to 1. Default is 0.
9049 Enable twoway sharpening if set to 1. Default is 0.
9052 @subsection Examples
9056 Apply default values:
9058 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9062 Enable additional sharpening:
9068 Paint processed pixels in white:
9074 @section lenscorrection
9076 Correct radial lens distortion
9078 This filter can be used to correct for radial distortion as can result from the use
9079 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9080 one can use tools available for example as part of opencv or simply trial-and-error.
9081 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9082 and extract the k1 and k2 coefficients from the resulting matrix.
9084 Note that effectively the same filter is available in the open-source tools Krita and
9085 Digikam from the KDE project.
9087 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9088 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9089 brightness distribution, so you may want to use both filters together in certain
9090 cases, though you will have to take care of ordering, i.e. whether vignetting should
9091 be applied before or after lens correction.
9095 The filter accepts the following options:
9099 Relative x-coordinate of the focal point of the image, and thereby the center of the
9100 distortion. This value has a range [0,1] and is expressed as fractions of the image
9103 Relative y-coordinate of the focal point of the image, and thereby the center of the
9104 distortion. This value has a range [0,1] and is expressed as fractions of the image
9107 Coefficient of the quadratic correction term. 0.5 means no correction.
9109 Coefficient of the double quadratic correction term. 0.5 means no correction.
9112 The formula that generates the correction is:
9114 @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)
9116 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9117 distances from the focal point in the source and target images, respectively.
9123 The filter accepts the following options:
9127 Set the number of loops.
9130 Set maximal size in number of frames.
9133 Set first frame of loop.
9139 Apply a 3D LUT to an input video.
9141 The filter accepts the following options:
9145 Set the 3D LUT file name.
9147 Currently supported formats:
9159 Select interpolation mode.
9161 Available values are:
9165 Use values from the nearest defined point.
9167 Interpolate values using the 8 points defining a cube.
9169 Interpolate values using a tetrahedron.
9173 @section lut, lutrgb, lutyuv
9175 Compute a look-up table for binding each pixel component input value
9176 to an output value, and apply it to the input video.
9178 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9179 to an RGB input video.
9181 These filters accept the following parameters:
9184 set first pixel component expression
9186 set second pixel component expression
9188 set third pixel component expression
9190 set fourth pixel component expression, corresponds to the alpha component
9193 set red component expression
9195 set green component expression
9197 set blue component expression
9199 alpha component expression
9202 set Y/luminance component expression
9204 set U/Cb component expression
9206 set V/Cr component expression
9209 Each of them specifies the expression to use for computing the lookup table for
9210 the corresponding pixel component values.
9212 The exact component associated to each of the @var{c*} options depends on the
9215 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9216 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9218 The expressions can contain the following constants and functions:
9223 The input width and height.
9226 The input value for the pixel component.
9229 The input value, clipped to the @var{minval}-@var{maxval} range.
9232 The maximum value for the pixel component.
9235 The minimum value for the pixel component.
9238 The negated value for the pixel component value, clipped to the
9239 @var{minval}-@var{maxval} range; it corresponds to the expression
9240 "maxval-clipval+minval".
9243 The computed value in @var{val}, clipped to the
9244 @var{minval}-@var{maxval} range.
9246 @item gammaval(gamma)
9247 The computed gamma correction value of the pixel component value,
9248 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9250 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9254 All expressions default to "val".
9256 @subsection Examples
9262 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9263 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9266 The above is the same as:
9268 lutrgb="r=negval:g=negval:b=negval"
9269 lutyuv="y=negval:u=negval:v=negval"
9279 Remove chroma components, turning the video into a graytone image:
9281 lutyuv="u=128:v=128"
9285 Apply a luma burning effect:
9291 Remove green and blue components:
9297 Set a constant alpha channel value on input:
9299 format=rgba,lutrgb=a="maxval-minval/2"
9303 Correct luminance gamma by a factor of 0.5:
9305 lutyuv=y=gammaval(0.5)
9309 Discard least significant bits of luma:
9311 lutyuv=y='bitand(val, 128+64+32)'
9315 Technicolor like effect:
9317 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9323 Compute and apply a lookup table from two video inputs.
9325 This filter accepts the following parameters:
9328 set first pixel component expression
9330 set second pixel component expression
9332 set third pixel component expression
9334 set fourth pixel component expression, corresponds to the alpha component
9337 Each of them specifies the expression to use for computing the lookup table for
9338 the corresponding pixel component values.
9340 The exact component associated to each of the @var{c*} options depends on the
9343 The expressions can contain the following constants:
9348 The input width and height.
9351 The first input value for the pixel component.
9354 The second input value for the pixel component.
9357 The first input video bit depth.
9360 The second input video bit depth.
9363 All expressions default to "x".
9365 @subsection Examples
9369 Highlight differences between two RGB video streams:
9371 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)'
9375 Highlight differences between two YUV video streams:
9377 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)'
9381 @section maskedclamp
9383 Clamp the first input stream with the second input and third input stream.
9385 Returns the value of first stream to be between second input
9386 stream - @code{undershoot} and third input stream + @code{overshoot}.
9388 This filter accepts the following options:
9391 Default value is @code{0}.
9394 Default value is @code{0}.
9397 Set which planes will be processed as bitmap, unprocessed planes will be
9398 copied from first stream.
9399 By default value 0xf, all planes will be processed.
9402 @section maskedmerge
9404 Merge the first input stream with the second input stream using per pixel
9405 weights in the third input stream.
9407 A value of 0 in the third stream pixel component means that pixel component
9408 from first stream is returned unchanged, while maximum value (eg. 255 for
9409 8-bit videos) means that pixel component from second stream is returned
9410 unchanged. Intermediate values define the amount of merging between both
9411 input stream's pixel components.
9413 This filter accepts the following options:
9416 Set which planes will be processed as bitmap, unprocessed planes will be
9417 copied from first stream.
9418 By default value 0xf, all planes will be processed.
9423 Apply motion-compensation deinterlacing.
9425 It needs one field per frame as input and must thus be used together
9426 with yadif=1/3 or equivalent.
9428 This filter accepts the following options:
9431 Set the deinterlacing mode.
9433 It accepts one of the following values:
9438 use iterative motion estimation
9440 like @samp{slow}, but use multiple reference frames.
9442 Default value is @samp{fast}.
9445 Set the picture field parity assumed for the input video. It must be
9446 one of the following values:
9450 assume top field first
9452 assume bottom field first
9455 Default value is @samp{bff}.
9458 Set per-block quantization parameter (QP) used by the internal
9461 Higher values should result in a smoother motion vector field but less
9462 optimal individual vectors. Default value is 1.
9465 @section mergeplanes
9467 Merge color channel components from several video streams.
9469 The filter accepts up to 4 input streams, and merge selected input
9470 planes to the output video.
9472 This filter accepts the following options:
9475 Set input to output plane mapping. Default is @code{0}.
9477 The mappings is specified as a bitmap. It should be specified as a
9478 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9479 mapping for the first plane of the output stream. 'A' sets the number of
9480 the input stream to use (from 0 to 3), and 'a' the plane number of the
9481 corresponding input to use (from 0 to 3). The rest of the mappings is
9482 similar, 'Bb' describes the mapping for the output stream second
9483 plane, 'Cc' describes the mapping for the output stream third plane and
9484 'Dd' describes the mapping for the output stream fourth plane.
9487 Set output pixel format. Default is @code{yuva444p}.
9490 @subsection Examples
9494 Merge three gray video streams of same width and height into single video stream:
9496 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9500 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9502 [a0][a1]mergeplanes=0x00010210:yuva444p
9506 Swap Y and A plane in yuva444p stream:
9508 format=yuva444p,mergeplanes=0x03010200:yuva444p
9512 Swap U and V plane in yuv420p stream:
9514 format=yuv420p,mergeplanes=0x000201:yuv420p
9518 Cast a rgb24 clip to yuv444p:
9520 format=rgb24,mergeplanes=0x000102:yuv444p
9526 Estimate and export motion vectors using block matching algorithms.
9527 Motion vectors are stored in frame side data to be used by other filters.
9529 This filter accepts the following options:
9532 Specify the motion estimation method. Accepts one of the following values:
9536 Exhaustive search algorithm.
9538 Three step search algorithm.
9540 Two dimensional logarithmic search algorithm.
9542 New three step search algorithm.
9544 Four step search algorithm.
9546 Diamond search algorithm.
9548 Hexagon-based search algorithm.
9550 Enhanced predictive zonal search algorithm.
9552 Uneven multi-hexagon search algorithm.
9554 Default value is @samp{esa}.
9557 Macroblock size. Default @code{16}.
9560 Search parameter. Default @code{7}.
9563 @section minterpolate
9565 Convert the video to specified frame rate using motion interpolation.
9567 This filter accepts the following options:
9570 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}.
9573 Motion interpolation mode. Following values are accepted:
9576 Duplicate previous or next frame for interpolating new ones.
9578 Blend source frames. Interpolated frame is mean of previous and next frames.
9580 Motion compensated interpolation. Following options are effective when this mode is selected:
9584 Motion compensation mode. Following values are accepted:
9587 Overlapped block motion compensation.
9589 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
9591 Default mode is @samp{obmc}.
9594 Motion estimation mode. Following values are accepted:
9597 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
9599 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
9601 Default mode is @samp{bilat}.
9604 The algorithm to be used for motion estimation. Following values are accepted:
9607 Exhaustive search algorithm.
9609 Three step search algorithm.
9611 Two dimensional logarithmic search algorithm.
9613 New three step search algorithm.
9615 Four step search algorithm.
9617 Diamond search algorithm.
9619 Hexagon-based search algorithm.
9621 Enhanced predictive zonal search algorithm.
9623 Uneven multi-hexagon search algorithm.
9625 Default algorithm is @samp{epzs}.
9628 Macroblock size. Default @code{16}.
9631 Motion estimation search parameter. Default @code{32}.
9634 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).
9639 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:
9642 Disable scene change detection.
9644 Frame difference. Corresponding pixel values are compared and if it statisfies @var{scd_threshold} scene change is detected.
9646 Default method is @samp{fdiff}.
9649 Scene change detection threshold. Default is @code{5.0}.
9654 Drop frames that do not differ greatly from the previous frame in
9655 order to reduce frame rate.
9657 The main use of this filter is for very-low-bitrate encoding
9658 (e.g. streaming over dialup modem), but it could in theory be used for
9659 fixing movies that were inverse-telecined incorrectly.
9661 A description of the accepted options follows.
9665 Set the maximum number of consecutive frames which can be dropped (if
9666 positive), or the minimum interval between dropped frames (if
9667 negative). If the value is 0, the frame is dropped unregarding the
9668 number of previous sequentially dropped frames.
9675 Set the dropping threshold values.
9677 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9678 represent actual pixel value differences, so a threshold of 64
9679 corresponds to 1 unit of difference for each pixel, or the same spread
9680 out differently over the block.
9682 A frame is a candidate for dropping if no 8x8 blocks differ by more
9683 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9684 meaning the whole image) differ by more than a threshold of @option{lo}.
9686 Default value for @option{hi} is 64*12, default value for @option{lo} is
9687 64*5, and default value for @option{frac} is 0.33.
9695 It accepts an integer in input; if non-zero it negates the
9696 alpha component (if available). The default value in input is 0.
9700 Denoise frames using Non-Local Means algorithm.
9702 Each pixel is adjusted by looking for other pixels with similar contexts. This
9703 context similarity is defined by comparing their surrounding patches of size
9704 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
9707 Note that the research area defines centers for patches, which means some
9708 patches will be made of pixels outside that research area.
9710 The filter accepts the following options.
9714 Set denoising strength.
9720 Same as @option{p} but for chroma planes.
9722 The default value is @var{0} and means automatic.
9728 Same as @option{r} but for chroma planes.
9730 The default value is @var{0} and means automatic.
9735 Deinterlace video using neural network edge directed interpolation.
9737 This filter accepts the following options:
9741 Mandatory option, without binary file filter can not work.
9742 Currently file can be found here:
9743 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9746 Set which frames to deinterlace, by default it is @code{all}.
9747 Can be @code{all} or @code{interlaced}.
9750 Set mode of operation.
9752 Can be one of the following:
9756 Use frame flags, both fields.
9758 Use frame flags, single field.
9762 Use bottom field only.
9764 Use both fields, top first.
9766 Use both fields, bottom first.
9770 Set which planes to process, by default filter process all frames.
9773 Set size of local neighborhood around each pixel, used by the predictor neural
9776 Can be one of the following:
9789 Set the number of neurons in predicctor neural network.
9790 Can be one of the following:
9801 Controls the number of different neural network predictions that are blended
9802 together to compute the final output value. Can be @code{fast}, default or
9806 Set which set of weights to use in the predictor.
9807 Can be one of the following:
9811 weights trained to minimize absolute error
9813 weights trained to minimize squared error
9817 Controls whether or not the prescreener neural network is used to decide
9818 which pixels should be processed by the predictor neural network and which
9819 can be handled by simple cubic interpolation.
9820 The prescreener is trained to know whether cubic interpolation will be
9821 sufficient for a pixel or whether it should be predicted by the predictor nn.
9822 The computational complexity of the prescreener nn is much less than that of
9823 the predictor nn. Since most pixels can be handled by cubic interpolation,
9824 using the prescreener generally results in much faster processing.
9825 The prescreener is pretty accurate, so the difference between using it and not
9826 using it is almost always unnoticeable.
9828 Can be one of the following:
9836 Default is @code{new}.
9839 Set various debugging flags.
9844 Force libavfilter not to use any of the specified pixel formats for the
9845 input to the next filter.
9847 It accepts the following parameters:
9851 A '|'-separated list of pixel format names, such as
9852 apix_fmts=yuv420p|monow|rgb24".
9856 @subsection Examples
9860 Force libavfilter to use a format different from @var{yuv420p} for the
9861 input to the vflip filter:
9863 noformat=pix_fmts=yuv420p,vflip
9867 Convert the input video to any of the formats not contained in the list:
9869 noformat=yuv420p|yuv444p|yuv410p
9875 Add noise on video input frame.
9877 The filter accepts the following options:
9885 Set noise seed for specific pixel component or all pixel components in case
9886 of @var{all_seed}. Default value is @code{123457}.
9888 @item all_strength, alls
9889 @item c0_strength, c0s
9890 @item c1_strength, c1s
9891 @item c2_strength, c2s
9892 @item c3_strength, c3s
9893 Set noise strength for specific pixel component or all pixel components in case
9894 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9896 @item all_flags, allf
9901 Set pixel component flags or set flags for all components if @var{all_flags}.
9902 Available values for component flags are:
9905 averaged temporal noise (smoother)
9907 mix random noise with a (semi)regular pattern
9909 temporal noise (noise pattern changes between frames)
9911 uniform noise (gaussian otherwise)
9915 @subsection Examples
9917 Add temporal and uniform noise to input video:
9919 noise=alls=20:allf=t+u
9924 Pass the video source unchanged to the output.
9927 Optical Character Recognition
9929 This filter uses Tesseract for optical character recognition.
9931 It accepts the following options:
9935 Set datapath to tesseract data. Default is to use whatever was
9936 set at installation.
9939 Set language, default is "eng".
9942 Set character whitelist.
9945 Set character blacklist.
9948 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9952 Apply a video transform using libopencv.
9954 To enable this filter, install the libopencv library and headers and
9955 configure FFmpeg with @code{--enable-libopencv}.
9957 It accepts the following parameters:
9962 The name of the libopencv filter to apply.
9965 The parameters to pass to the libopencv filter. If not specified, the default
9970 Refer to the official libopencv documentation for more precise
9972 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9974 Several libopencv filters are supported; see the following subsections.
9979 Dilate an image by using a specific structuring element.
9980 It corresponds to the libopencv function @code{cvDilate}.
9982 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
9984 @var{struct_el} represents a structuring element, and has the syntax:
9985 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
9987 @var{cols} and @var{rows} represent the number of columns and rows of
9988 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
9989 point, and @var{shape} the shape for the structuring element. @var{shape}
9990 must be "rect", "cross", "ellipse", or "custom".
9992 If the value for @var{shape} is "custom", it must be followed by a
9993 string of the form "=@var{filename}". The file with name
9994 @var{filename} is assumed to represent a binary image, with each
9995 printable character corresponding to a bright pixel. When a custom
9996 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
9997 or columns and rows of the read file are assumed instead.
9999 The default value for @var{struct_el} is "3x3+0x0/rect".
10001 @var{nb_iterations} specifies the number of times the transform is
10002 applied to the image, and defaults to 1.
10006 # Use the default values
10009 # Dilate using a structuring element with a 5x5 cross, iterating two times
10010 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10012 # Read the shape from the file diamond.shape, iterating two times.
10013 # The file diamond.shape may contain a pattern of characters like this
10019 # The specified columns and rows are ignored
10020 # but the anchor point coordinates are not
10021 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10026 Erode an image by using a specific structuring element.
10027 It corresponds to the libopencv function @code{cvErode}.
10029 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10030 with the same syntax and semantics as the @ref{dilate} filter.
10034 Smooth the input video.
10036 The filter takes the following parameters:
10037 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10039 @var{type} is the type of smooth filter to apply, and must be one of
10040 the following values: "blur", "blur_no_scale", "median", "gaussian",
10041 or "bilateral". The default value is "gaussian".
10043 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10044 depend on the smooth type. @var{param1} and
10045 @var{param2} accept integer positive values or 0. @var{param3} and
10046 @var{param4} accept floating point values.
10048 The default value for @var{param1} is 3. The default value for the
10049 other parameters is 0.
10051 These parameters correspond to the parameters assigned to the
10052 libopencv function @code{cvSmooth}.
10057 Overlay one video on top of another.
10059 It takes two inputs and has one output. The first input is the "main"
10060 video on which the second input is overlaid.
10062 It accepts the following parameters:
10064 A description of the accepted options follows.
10069 Set the expression for the x and y coordinates of the overlaid video
10070 on the main video. Default value is "0" for both expressions. In case
10071 the expression is invalid, it is set to a huge value (meaning that the
10072 overlay will not be displayed within the output visible area).
10075 The action to take when EOF is encountered on the secondary input; it accepts
10076 one of the following values:
10080 Repeat the last frame (the default).
10084 Pass the main input through.
10088 Set when the expressions for @option{x}, and @option{y} are evaluated.
10090 It accepts the following values:
10093 only evaluate expressions once during the filter initialization or
10094 when a command is processed
10097 evaluate expressions for each incoming frame
10100 Default value is @samp{frame}.
10103 If set to 1, force the output to terminate when the shortest input
10104 terminates. Default value is 0.
10107 Set the format for the output video.
10109 It accepts the following values:
10112 force YUV420 output
10115 force YUV422 output
10118 force YUV444 output
10124 Default value is @samp{yuv420}.
10126 @item rgb @emph{(deprecated)}
10127 If set to 1, force the filter to accept inputs in the RGB
10128 color space. Default value is 0. This option is deprecated, use
10129 @option{format} instead.
10132 If set to 1, force the filter to draw the last overlay frame over the
10133 main input until the end of the stream. A value of 0 disables this
10134 behavior. Default value is 1.
10137 The @option{x}, and @option{y} expressions can contain the following
10143 The main input width and height.
10147 The overlay input width and height.
10151 The computed values for @var{x} and @var{y}. They are evaluated for
10156 horizontal and vertical chroma subsample values of the output
10157 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
10161 the number of input frame, starting from 0
10164 the position in the file of the input frame, NAN if unknown
10167 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
10171 Note that the @var{n}, @var{pos}, @var{t} variables are available only
10172 when evaluation is done @emph{per frame}, and will evaluate to NAN
10173 when @option{eval} is set to @samp{init}.
10175 Be aware that frames are taken from each input video in timestamp
10176 order, hence, if their initial timestamps differ, it is a good idea
10177 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
10178 have them begin in the same zero timestamp, as the example for
10179 the @var{movie} filter does.
10181 You can chain together more overlays but you should test the
10182 efficiency of such approach.
10184 @subsection Commands
10186 This filter supports the following commands:
10190 Modify the x and y of the overlay input.
10191 The command accepts the same syntax of the corresponding option.
10193 If the specified expression is not valid, it is kept at its current
10197 @subsection Examples
10201 Draw the overlay at 10 pixels from the bottom right corner of the main
10204 overlay=main_w-overlay_w-10:main_h-overlay_h-10
10207 Using named options the example above becomes:
10209 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
10213 Insert a transparent PNG logo in the bottom left corner of the input,
10214 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
10216 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
10220 Insert 2 different transparent PNG logos (second logo on bottom
10221 right corner) using the @command{ffmpeg} tool:
10223 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
10227 Add a transparent color layer on top of the main video; @code{WxH}
10228 must specify the size of the main input to the overlay filter:
10230 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
10234 Play an original video and a filtered version (here with the deshake
10235 filter) side by side using the @command{ffplay} tool:
10237 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
10240 The above command is the same as:
10242 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
10246 Make a sliding overlay appearing from the left to the right top part of the
10247 screen starting since time 2:
10249 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
10253 Compose output by putting two input videos side to side:
10255 ffmpeg -i left.avi -i right.avi -filter_complex "
10256 nullsrc=size=200x100 [background];
10257 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
10258 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
10259 [background][left] overlay=shortest=1 [background+left];
10260 [background+left][right] overlay=shortest=1:x=100 [left+right]
10265 Mask 10-20 seconds of a video by applying the delogo filter to a section
10267 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
10268 -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]'
10273 Chain several overlays in cascade:
10275 nullsrc=s=200x200 [bg];
10276 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
10277 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
10278 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
10279 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
10280 [in3] null, [mid2] overlay=100:100 [out0]
10287 Apply Overcomplete Wavelet denoiser.
10289 The filter accepts the following options:
10295 Larger depth values will denoise lower frequency components more, but
10296 slow down filtering.
10298 Must be an int in the range 8-16, default is @code{8}.
10300 @item luma_strength, ls
10303 Must be a double value in the range 0-1000, default is @code{1.0}.
10305 @item chroma_strength, cs
10306 Set chroma strength.
10308 Must be a double value in the range 0-1000, default is @code{1.0}.
10314 Add paddings to the input image, and place the original input at the
10315 provided @var{x}, @var{y} coordinates.
10317 It accepts the following parameters:
10322 Specify an expression for the size of the output image with the
10323 paddings added. If the value for @var{width} or @var{height} is 0, the
10324 corresponding input size is used for the output.
10326 The @var{width} expression can reference the value set by the
10327 @var{height} expression, and vice versa.
10329 The default value of @var{width} and @var{height} is 0.
10333 Specify the offsets to place the input image at within the padded area,
10334 with respect to the top/left border of the output image.
10336 The @var{x} expression can reference the value set by the @var{y}
10337 expression, and vice versa.
10339 The default value of @var{x} and @var{y} is 0.
10342 Specify the color of the padded area. For the syntax of this option,
10343 check the "Color" section in the ffmpeg-utils manual.
10345 The default value of @var{color} is "black".
10348 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10349 options are expressions containing the following constants:
10354 The input video width and height.
10358 These are the same as @var{in_w} and @var{in_h}.
10362 The output width and height (the size of the padded area), as
10363 specified by the @var{width} and @var{height} expressions.
10367 These are the same as @var{out_w} and @var{out_h}.
10371 The x and y offsets as specified by the @var{x} and @var{y}
10372 expressions, or NAN if not yet specified.
10375 same as @var{iw} / @var{ih}
10378 input sample aspect ratio
10381 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10385 The horizontal and vertical chroma subsample values. For example for the
10386 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10389 @subsection Examples
10393 Add paddings with the color "violet" to the input video. The output video
10394 size is 640x480, and the top-left corner of the input video is placed at
10397 pad=640:480:0:40:violet
10400 The example above is equivalent to the following command:
10402 pad=width=640:height=480:x=0:y=40:color=violet
10406 Pad the input to get an output with dimensions increased by 3/2,
10407 and put the input video at the center of the padded area:
10409 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10413 Pad the input to get a squared output with size equal to the maximum
10414 value between the input width and height, and put the input video at
10415 the center of the padded area:
10417 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10421 Pad the input to get a final w/h ratio of 16:9:
10423 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10427 In case of anamorphic video, in order to set the output display aspect
10428 correctly, it is necessary to use @var{sar} in the expression,
10429 according to the relation:
10431 (ih * X / ih) * sar = output_dar
10432 X = output_dar / sar
10435 Thus the previous example needs to be modified to:
10437 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10441 Double the output size and put the input video in the bottom-right
10442 corner of the output padded area:
10444 pad="2*iw:2*ih:ow-iw:oh-ih"
10448 @anchor{palettegen}
10449 @section palettegen
10451 Generate one palette for a whole video stream.
10453 It accepts the following options:
10457 Set the maximum number of colors to quantize in the palette.
10458 Note: the palette will still contain 256 colors; the unused palette entries
10461 @item reserve_transparent
10462 Create a palette of 255 colors maximum and reserve the last one for
10463 transparency. Reserving the transparency color is useful for GIF optimization.
10464 If not set, the maximum of colors in the palette will be 256. You probably want
10465 to disable this option for a standalone image.
10469 Set statistics mode.
10471 It accepts the following values:
10474 Compute full frame histograms.
10476 Compute histograms only for the part that differs from previous frame. This
10477 might be relevant to give more importance to the moving part of your input if
10478 the background is static.
10480 Compute new histogram for each frame.
10483 Default value is @var{full}.
10486 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10487 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10488 color quantization of the palette. This information is also visible at
10489 @var{info} logging level.
10491 @subsection Examples
10495 Generate a representative palette of a given video using @command{ffmpeg}:
10497 ffmpeg -i input.mkv -vf palettegen palette.png
10501 @section paletteuse
10503 Use a palette to downsample an input video stream.
10505 The filter takes two inputs: one video stream and a palette. The palette must
10506 be a 256 pixels image.
10508 It accepts the following options:
10512 Select dithering mode. Available algorithms are:
10515 Ordered 8x8 bayer dithering (deterministic)
10517 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10518 Note: this dithering is sometimes considered "wrong" and is included as a
10520 @item floyd_steinberg
10521 Floyd and Steingberg dithering (error diffusion)
10523 Frankie Sierra dithering v2 (error diffusion)
10525 Frankie Sierra dithering v2 "Lite" (error diffusion)
10528 Default is @var{sierra2_4a}.
10531 When @var{bayer} dithering is selected, this option defines the scale of the
10532 pattern (how much the crosshatch pattern is visible). A low value means more
10533 visible pattern for less banding, and higher value means less visible pattern
10534 at the cost of more banding.
10536 The option must be an integer value in the range [0,5]. Default is @var{2}.
10539 If set, define the zone to process
10543 Only the changing rectangle will be reprocessed. This is similar to GIF
10544 cropping/offsetting compression mechanism. This option can be useful for speed
10545 if only a part of the image is changing, and has use cases such as limiting the
10546 scope of the error diffusal @option{dither} to the rectangle that bounds the
10547 moving scene (it leads to more deterministic output if the scene doesn't change
10548 much, and as a result less moving noise and better GIF compression).
10551 Default is @var{none}.
10554 Take new palette for each output frame.
10557 @subsection Examples
10561 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10562 using @command{ffmpeg}:
10564 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10568 @section perspective
10570 Correct perspective of video not recorded perpendicular to the screen.
10572 A description of the accepted parameters follows.
10583 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10584 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10585 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10586 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10587 then the corners of the source will be sent to the specified coordinates.
10589 The expressions can use the following variables:
10594 the width and height of video frame.
10598 Output frame count.
10601 @item interpolation
10602 Set interpolation for perspective correction.
10604 It accepts the following values:
10610 Default value is @samp{linear}.
10613 Set interpretation of coordinate options.
10615 It accepts the following values:
10619 Send point in the source specified by the given coordinates to
10620 the corners of the destination.
10622 @item 1, destination
10624 Send the corners of the source to the point in the destination specified
10625 by the given coordinates.
10627 Default value is @samp{source}.
10631 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10633 It accepts the following values:
10636 only evaluate expressions once during the filter initialization or
10637 when a command is processed
10640 evaluate expressions for each incoming frame
10643 Default value is @samp{init}.
10648 Delay interlaced video by one field time so that the field order changes.
10650 The intended use is to fix PAL movies that have been captured with the
10651 opposite field order to the film-to-video transfer.
10653 A description of the accepted parameters follows.
10659 It accepts the following values:
10662 Capture field order top-first, transfer bottom-first.
10663 Filter will delay the bottom field.
10666 Capture field order bottom-first, transfer top-first.
10667 Filter will delay the top field.
10670 Capture and transfer with the same field order. This mode only exists
10671 for the documentation of the other options to refer to, but if you
10672 actually select it, the filter will faithfully do nothing.
10675 Capture field order determined automatically by field flags, transfer
10677 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10678 basis using field flags. If no field information is available,
10679 then this works just like @samp{u}.
10682 Capture unknown or varying, transfer opposite.
10683 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10684 analyzing the images and selecting the alternative that produces best
10685 match between the fields.
10688 Capture top-first, transfer unknown or varying.
10689 Filter selects among @samp{t} and @samp{p} using image analysis.
10692 Capture bottom-first, transfer unknown or varying.
10693 Filter selects among @samp{b} and @samp{p} using image analysis.
10696 Capture determined by field flags, transfer unknown or varying.
10697 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10698 image analysis. If no field information is available, then this works just
10699 like @samp{U}. This is the default mode.
10702 Both capture and transfer unknown or varying.
10703 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10707 @section pixdesctest
10709 Pixel format descriptor test filter, mainly useful for internal
10710 testing. The output video should be equal to the input video.
10714 format=monow, pixdesctest
10717 can be used to test the monowhite pixel format descriptor definition.
10721 Enable the specified chain of postprocessing subfilters using libpostproc. This
10722 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10723 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10724 Each subfilter and some options have a short and a long name that can be used
10725 interchangeably, i.e. dr/dering are the same.
10727 The filters accept the following options:
10731 Set postprocessing subfilters string.
10734 All subfilters share common options to determine their scope:
10738 Honor the quality commands for this subfilter.
10741 Do chrominance filtering, too (default).
10744 Do luminance filtering only (no chrominance).
10747 Do chrominance filtering only (no luminance).
10750 These options can be appended after the subfilter name, separated by a '|'.
10752 Available subfilters are:
10755 @item hb/hdeblock[|difference[|flatness]]
10756 Horizontal deblocking filter
10759 Difference factor where higher values mean more deblocking (default: @code{32}).
10761 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10764 @item vb/vdeblock[|difference[|flatness]]
10765 Vertical deblocking filter
10768 Difference factor where higher values mean more deblocking (default: @code{32}).
10770 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10773 @item ha/hadeblock[|difference[|flatness]]
10774 Accurate horizontal deblocking filter
10777 Difference factor where higher values mean more deblocking (default: @code{32}).
10779 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10782 @item va/vadeblock[|difference[|flatness]]
10783 Accurate vertical deblocking filter
10786 Difference factor where higher values mean more deblocking (default: @code{32}).
10788 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10792 The horizontal and vertical deblocking filters share the difference and
10793 flatness values so you cannot set different horizontal and vertical
10797 @item h1/x1hdeblock
10798 Experimental horizontal deblocking filter
10800 @item v1/x1vdeblock
10801 Experimental vertical deblocking filter
10806 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10809 larger -> stronger filtering
10811 larger -> stronger filtering
10813 larger -> stronger filtering
10816 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10819 Stretch luminance to @code{0-255}.
10822 @item lb/linblenddeint
10823 Linear blend deinterlacing filter that deinterlaces the given block by
10824 filtering all lines with a @code{(1 2 1)} filter.
10826 @item li/linipoldeint
10827 Linear interpolating deinterlacing filter that deinterlaces the given block by
10828 linearly interpolating every second line.
10830 @item ci/cubicipoldeint
10831 Cubic interpolating deinterlacing filter deinterlaces the given block by
10832 cubically interpolating every second line.
10834 @item md/mediandeint
10835 Median deinterlacing filter that deinterlaces the given block by applying a
10836 median filter to every second line.
10838 @item fd/ffmpegdeint
10839 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10840 second line with a @code{(-1 4 2 4 -1)} filter.
10843 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10844 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10846 @item fq/forceQuant[|quantizer]
10847 Overrides the quantizer table from the input with the constant quantizer you
10855 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10858 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10861 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10864 @subsection Examples
10868 Apply horizontal and vertical deblocking, deringing and automatic
10869 brightness/contrast:
10875 Apply default filters without brightness/contrast correction:
10881 Apply default filters and temporal denoiser:
10883 pp=default/tmpnoise|1|2|3
10887 Apply deblocking on luminance only, and switch vertical deblocking on or off
10888 automatically depending on available CPU time:
10895 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10896 similar to spp = 6 with 7 point DCT, where only the center sample is
10899 The filter accepts the following options:
10903 Force a constant quantization parameter. It accepts an integer in range
10904 0 to 63. If not set, the filter will use the QP from the video stream
10908 Set thresholding mode. Available modes are:
10912 Set hard thresholding.
10914 Set soft thresholding (better de-ringing effect, but likely blurrier).
10916 Set medium thresholding (good results, default).
10921 Apply prewitt operator to input video stream.
10923 The filter accepts the following option:
10927 Set which planes will be processed, unprocessed planes will be copied.
10928 By default value 0xf, all planes will be processed.
10931 Set value which will be multiplied with filtered result.
10934 Set value which will be added to filtered result.
10939 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10940 Ratio) between two input videos.
10942 This filter takes in input two input videos, the first input is
10943 considered the "main" source and is passed unchanged to the
10944 output. The second input is used as a "reference" video for computing
10947 Both video inputs must have the same resolution and pixel format for
10948 this filter to work correctly. Also it assumes that both inputs
10949 have the same number of frames, which are compared one by one.
10951 The obtained average PSNR is printed through the logging system.
10953 The filter stores the accumulated MSE (mean squared error) of each
10954 frame, and at the end of the processing it is averaged across all frames
10955 equally, and the following formula is applied to obtain the PSNR:
10958 PSNR = 10*log10(MAX^2/MSE)
10961 Where MAX is the average of the maximum values of each component of the
10964 The description of the accepted parameters follows.
10967 @item stats_file, f
10968 If specified the filter will use the named file to save the PSNR of
10969 each individual frame. When filename equals "-" the data is sent to
10972 @item stats_version
10973 Specifies which version of the stats file format to use. Details of
10974 each format are written below.
10975 Default value is 1.
10977 @item stats_add_max
10978 Determines whether the max value is output to the stats log.
10979 Default value is 0.
10980 Requires stats_version >= 2. If this is set and stats_version < 2,
10981 the filter will return an error.
10984 The file printed if @var{stats_file} is selected, contains a sequence of
10985 key/value pairs of the form @var{key}:@var{value} for each compared
10988 If a @var{stats_version} greater than 1 is specified, a header line precedes
10989 the list of per-frame-pair stats, with key value pairs following the frame
10990 format with the following parameters:
10993 @item psnr_log_version
10994 The version of the log file format. Will match @var{stats_version}.
10997 A comma separated list of the per-frame-pair parameters included in
11001 A description of each shown per-frame-pair parameter follows:
11005 sequential number of the input frame, starting from 1
11008 Mean Square Error pixel-by-pixel average difference of the compared
11009 frames, averaged over all the image components.
11011 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
11012 Mean Square Error pixel-by-pixel average difference of the compared
11013 frames for the component specified by the suffix.
11015 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
11016 Peak Signal to Noise ratio of the compared frames for the component
11017 specified by the suffix.
11019 @item max_avg, max_y, max_u, max_v
11020 Maximum allowed value for each channel, and average over all
11026 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
11027 [main][ref] psnr="stats_file=stats.log" [out]
11030 On this example the input file being processed is compared with the
11031 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
11032 is stored in @file{stats.log}.
11037 Pulldown reversal (inverse telecine) filter, capable of handling mixed
11038 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
11041 The pullup filter is designed to take advantage of future context in making
11042 its decisions. This filter is stateless in the sense that it does not lock
11043 onto a pattern to follow, but it instead looks forward to the following
11044 fields in order to identify matches and rebuild progressive frames.
11046 To produce content with an even framerate, insert the fps filter after
11047 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
11048 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
11050 The filter accepts the following options:
11057 These options set the amount of "junk" to ignore at the left, right, top, and
11058 bottom of the image, respectively. Left and right are in units of 8 pixels,
11059 while top and bottom are in units of 2 lines.
11060 The default is 8 pixels on each side.
11063 Set the strict breaks. Setting this option to 1 will reduce the chances of
11064 filter generating an occasional mismatched frame, but it may also cause an
11065 excessive number of frames to be dropped during high motion sequences.
11066 Conversely, setting it to -1 will make filter match fields more easily.
11067 This may help processing of video where there is slight blurring between
11068 the fields, but may also cause there to be interlaced frames in the output.
11069 Default value is @code{0}.
11072 Set the metric plane to use. It accepts the following values:
11078 Use chroma blue plane.
11081 Use chroma red plane.
11084 This option may be set to use chroma plane instead of the default luma plane
11085 for doing filter's computations. This may improve accuracy on very clean
11086 source material, but more likely will decrease accuracy, especially if there
11087 is chroma noise (rainbow effect) or any grayscale video.
11088 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
11089 load and make pullup usable in realtime on slow machines.
11092 For best results (without duplicated frames in the output file) it is
11093 necessary to change the output frame rate. For example, to inverse
11094 telecine NTSC input:
11096 ffmpeg -i input -vf pullup -r 24000/1001 ...
11101 Change video quantization parameters (QP).
11103 The filter accepts the following option:
11107 Set expression for quantization parameter.
11110 The expression is evaluated through the eval API and can contain, among others,
11111 the following constants:
11115 1 if index is not 129, 0 otherwise.
11118 Sequentional index starting from -129 to 128.
11121 @subsection Examples
11125 Some equation like:
11133 Flush video frames from internal cache of frames into a random order.
11134 No frame is discarded.
11135 Inspired by @ref{frei0r} nervous filter.
11139 Set size in number of frames of internal cache, in range from @code{2} to
11140 @code{512}. Default is @code{30}.
11143 Set seed for random number generator, must be an integer included between
11144 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
11145 less than @code{0}, the filter will try to use a good random seed on a
11151 Read vertical interval timecode (VITC) information from the top lines of a
11154 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
11155 timecode value, if a valid timecode has been detected. Further metadata key
11156 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
11157 timecode data has been found or not.
11159 This filter accepts the following options:
11163 Set the maximum number of lines to scan for VITC data. If the value is set to
11164 @code{-1} the full video frame is scanned. Default is @code{45}.
11167 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
11168 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
11171 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
11172 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
11175 @subsection Examples
11179 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
11180 draw @code{--:--:--:--} as a placeholder:
11182 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
11188 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
11190 Destination pixel at position (X, Y) will be picked from source (x, y) position
11191 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
11192 value for pixel will be used for destination pixel.
11194 Xmap and Ymap input video streams must be of same dimensions. Output video stream
11195 will have Xmap/Ymap video stream dimensions.
11196 Xmap and Ymap input video streams are 16bit depth, single channel.
11198 @section removegrain
11200 The removegrain filter is a spatial denoiser for progressive video.
11204 Set mode for the first plane.
11207 Set mode for the second plane.
11210 Set mode for the third plane.
11213 Set mode for the fourth plane.
11216 Range of mode is from 0 to 24. Description of each mode follows:
11220 Leave input plane unchanged. Default.
11223 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
11226 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
11229 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
11232 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
11233 This is equivalent to a median filter.
11236 Line-sensitive clipping giving the minimal change.
11239 Line-sensitive clipping, intermediate.
11242 Line-sensitive clipping, intermediate.
11245 Line-sensitive clipping, intermediate.
11248 Line-sensitive clipping on a line where the neighbours pixels are the closest.
11251 Replaces the target pixel with the closest neighbour.
11254 [1 2 1] horizontal and vertical kernel blur.
11260 Bob mode, interpolates top field from the line where the neighbours
11261 pixels are the closest.
11264 Bob mode, interpolates bottom field from the line where the neighbours
11265 pixels are the closest.
11268 Bob mode, interpolates top field. Same as 13 but with a more complicated
11269 interpolation formula.
11272 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
11273 interpolation formula.
11276 Clips the pixel with the minimum and maximum of respectively the maximum and
11277 minimum of each pair of opposite neighbour pixels.
11280 Line-sensitive clipping using opposite neighbours whose greatest distance from
11281 the current pixel is minimal.
11284 Replaces the pixel with the average of its 8 neighbours.
11287 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
11290 Clips pixels using the averages of opposite neighbour.
11293 Same as mode 21 but simpler and faster.
11296 Small edge and halo removal, but reputed useless.
11302 @section removelogo
11304 Suppress a TV station logo, using an image file to determine which
11305 pixels comprise the logo. It works by filling in the pixels that
11306 comprise the logo with neighboring pixels.
11308 The filter accepts the following options:
11312 Set the filter bitmap file, which can be any image format supported by
11313 libavformat. The width and height of the image file must match those of the
11314 video stream being processed.
11317 Pixels in the provided bitmap image with a value of zero are not
11318 considered part of the logo, non-zero pixels are considered part of
11319 the logo. If you use white (255) for the logo and black (0) for the
11320 rest, you will be safe. For making the filter bitmap, it is
11321 recommended to take a screen capture of a black frame with the logo
11322 visible, and then using a threshold filter followed by the erode
11323 filter once or twice.
11325 If needed, little splotches can be fixed manually. Remember that if
11326 logo pixels are not covered, the filter quality will be much
11327 reduced. Marking too many pixels as part of the logo does not hurt as
11328 much, but it will increase the amount of blurring needed to cover over
11329 the image and will destroy more information than necessary, and extra
11330 pixels will slow things down on a large logo.
11332 @section repeatfields
11334 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11335 fields based on its value.
11339 Reverse a video clip.
11341 Warning: This filter requires memory to buffer the entire clip, so trimming
11344 @subsection Examples
11348 Take the first 5 seconds of a clip, and reverse it.
11356 Rotate video by an arbitrary angle expressed in radians.
11358 The filter accepts the following options:
11360 A description of the optional parameters follows.
11363 Set an expression for the angle by which to rotate the input video
11364 clockwise, expressed as a number of radians. A negative value will
11365 result in a counter-clockwise rotation. By default it is set to "0".
11367 This expression is evaluated for each frame.
11370 Set the output width expression, default value is "iw".
11371 This expression is evaluated just once during configuration.
11374 Set the output height expression, default value is "ih".
11375 This expression is evaluated just once during configuration.
11378 Enable bilinear interpolation if set to 1, a value of 0 disables
11379 it. Default value is 1.
11382 Set the color used to fill the output area not covered by the rotated
11383 image. For the general syntax of this option, check the "Color" section in the
11384 ffmpeg-utils manual. If the special value "none" is selected then no
11385 background is printed (useful for example if the background is never shown).
11387 Default value is "black".
11390 The expressions for the angle and the output size can contain the
11391 following constants and functions:
11395 sequential number of the input frame, starting from 0. It is always NAN
11396 before the first frame is filtered.
11399 time in seconds of the input frame, it is set to 0 when the filter is
11400 configured. It is always NAN before the first frame is filtered.
11404 horizontal and vertical chroma subsample values. For example for the
11405 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11409 the input video width and height
11413 the output width and height, that is the size of the padded area as
11414 specified by the @var{width} and @var{height} expressions
11418 the minimal width/height required for completely containing the input
11419 video rotated by @var{a} radians.
11421 These are only available when computing the @option{out_w} and
11422 @option{out_h} expressions.
11425 @subsection Examples
11429 Rotate the input by PI/6 radians clockwise:
11435 Rotate the input by PI/6 radians counter-clockwise:
11441 Rotate the input by 45 degrees clockwise:
11447 Apply a constant rotation with period T, starting from an angle of PI/3:
11449 rotate=PI/3+2*PI*t/T
11453 Make the input video rotation oscillating with a period of T
11454 seconds and an amplitude of A radians:
11456 rotate=A*sin(2*PI/T*t)
11460 Rotate the video, output size is chosen so that the whole rotating
11461 input video is always completely contained in the output:
11463 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11467 Rotate the video, reduce the output size so that no background is ever
11470 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11474 @subsection Commands
11476 The filter supports the following commands:
11480 Set the angle expression.
11481 The command accepts the same syntax of the corresponding option.
11483 If the specified expression is not valid, it is kept at its current
11489 Apply Shape Adaptive Blur.
11491 The filter accepts the following options:
11494 @item luma_radius, lr
11495 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11496 value is 1.0. A greater value will result in a more blurred image, and
11497 in slower processing.
11499 @item luma_pre_filter_radius, lpfr
11500 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11503 @item luma_strength, ls
11504 Set luma maximum difference between pixels to still be considered, must
11505 be a value in the 0.1-100.0 range, default value is 1.0.
11507 @item chroma_radius, cr
11508 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11509 greater value will result in a more blurred image, and in slower
11512 @item chroma_pre_filter_radius, cpfr
11513 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11515 @item chroma_strength, cs
11516 Set chroma maximum difference between pixels to still be considered,
11517 must be a value in the -0.9-100.0 range.
11520 Each chroma option value, if not explicitly specified, is set to the
11521 corresponding luma option value.
11526 Scale (resize) the input video, using the libswscale library.
11528 The scale filter forces the output display aspect ratio to be the same
11529 of the input, by changing the output sample aspect ratio.
11531 If the input image format is different from the format requested by
11532 the next filter, the scale filter will convert the input to the
11535 @subsection Options
11536 The filter accepts the following options, or any of the options
11537 supported by the libswscale scaler.
11539 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11540 the complete list of scaler options.
11545 Set the output video dimension expression. Default value is the input
11548 If the value is 0, the input width is used for the output.
11550 If one of the values is -1, the scale filter will use a value that
11551 maintains the aspect ratio of the input image, calculated from the
11552 other specified dimension. If both of them are -1, the input size is
11555 If one of the values is -n with n > 1, the scale filter will also use a value
11556 that maintains the aspect ratio of the input image, calculated from the other
11557 specified dimension. After that it will, however, make sure that the calculated
11558 dimension is divisible by n and adjust the value if necessary.
11560 See below for the list of accepted constants for use in the dimension
11564 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11568 Only evaluate expressions once during the filter initialization or when a command is processed.
11571 Evaluate expressions for each incoming frame.
11575 Default value is @samp{init}.
11579 Set the interlacing mode. It accepts the following values:
11583 Force interlaced aware scaling.
11586 Do not apply interlaced scaling.
11589 Select interlaced aware scaling depending on whether the source frames
11590 are flagged as interlaced or not.
11593 Default value is @samp{0}.
11596 Set libswscale scaling flags. See
11597 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11598 complete list of values. If not explicitly specified the filter applies
11602 @item param0, param1
11603 Set libswscale input parameters for scaling algorithms that need them. See
11604 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11605 complete documentation. If not explicitly specified the filter applies
11611 Set the video size. For the syntax of this option, check the
11612 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11614 @item in_color_matrix
11615 @item out_color_matrix
11616 Set in/output YCbCr color space type.
11618 This allows the autodetected value to be overridden as well as allows forcing
11619 a specific value used for the output and encoder.
11621 If not specified, the color space type depends on the pixel format.
11627 Choose automatically.
11630 Format conforming to International Telecommunication Union (ITU)
11631 Recommendation BT.709.
11634 Set color space conforming to the United States Federal Communications
11635 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11638 Set color space conforming to:
11642 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11645 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11648 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11653 Set color space conforming to SMPTE ST 240:1999.
11658 Set in/output YCbCr sample range.
11660 This allows the autodetected value to be overridden as well as allows forcing
11661 a specific value used for the output and encoder. If not specified, the
11662 range depends on the pixel format. Possible values:
11666 Choose automatically.
11669 Set full range (0-255 in case of 8-bit luma).
11672 Set "MPEG" range (16-235 in case of 8-bit luma).
11675 @item force_original_aspect_ratio
11676 Enable decreasing or increasing output video width or height if necessary to
11677 keep the original aspect ratio. Possible values:
11681 Scale the video as specified and disable this feature.
11684 The output video dimensions will automatically be decreased if needed.
11687 The output video dimensions will automatically be increased if needed.
11691 One useful instance of this option is that when you know a specific device's
11692 maximum allowed resolution, you can use this to limit the output video to
11693 that, while retaining the aspect ratio. For example, device A allows
11694 1280x720 playback, and your video is 1920x800. Using this option (set it to
11695 decrease) and specifying 1280x720 to the command line makes the output
11698 Please note that this is a different thing than specifying -1 for @option{w}
11699 or @option{h}, you still need to specify the output resolution for this option
11704 The values of the @option{w} and @option{h} options are expressions
11705 containing the following constants:
11710 The input width and height
11714 These are the same as @var{in_w} and @var{in_h}.
11718 The output (scaled) width and height
11722 These are the same as @var{out_w} and @var{out_h}
11725 The same as @var{iw} / @var{ih}
11728 input sample aspect ratio
11731 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11735 horizontal and vertical input chroma subsample values. For example for the
11736 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11740 horizontal and vertical output chroma subsample values. For example for the
11741 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11744 @subsection Examples
11748 Scale the input video to a size of 200x100
11753 This is equivalent to:
11764 Specify a size abbreviation for the output size:
11769 which can also be written as:
11775 Scale the input to 2x:
11777 scale=w=2*iw:h=2*ih
11781 The above is the same as:
11783 scale=2*in_w:2*in_h
11787 Scale the input to 2x with forced interlaced scaling:
11789 scale=2*iw:2*ih:interl=1
11793 Scale the input to half size:
11795 scale=w=iw/2:h=ih/2
11799 Increase the width, and set the height to the same size:
11805 Seek Greek harmony:
11812 Increase the height, and set the width to 3/2 of the height:
11814 scale=w=3/2*oh:h=3/5*ih
11818 Increase the size, making the size a multiple of the chroma
11821 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11825 Increase the width to a maximum of 500 pixels,
11826 keeping the same aspect ratio as the input:
11828 scale=w='min(500\, iw*3/2):h=-1'
11832 @subsection Commands
11834 This filter supports the following commands:
11838 Set the output video dimension expression.
11839 The command accepts the same syntax of the corresponding option.
11841 If the specified expression is not valid, it is kept at its current
11847 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11848 format conversion on CUDA video frames. Setting the output width and height
11849 works in the same way as for the @var{scale} filter.
11851 The following additional options are accepted:
11854 The pixel format of the output CUDA frames. If set to the string "same" (the
11855 default), the input format will be kept. Note that automatic format negotiation
11856 and conversion is not yet supported for hardware frames
11859 The interpolation algorithm used for resizing. One of the following:
11866 @item cubic2p_bspline
11867 2-parameter cubic (B=1, C=0)
11869 @item cubic2p_catmullrom
11870 2-parameter cubic (B=0, C=1/2)
11872 @item cubic2p_b05c03
11873 2-parameter cubic (B=1/2, C=3/10)
11885 Scale (resize) the input video, based on a reference video.
11887 See the scale filter for available options, scale2ref supports the same but
11888 uses the reference video instead of the main input as basis.
11890 @subsection Examples
11894 Scale a subtitle stream to match the main video in size before overlaying
11896 'scale2ref[b][a];[a][b]overlay'
11900 @anchor{selectivecolor}
11901 @section selectivecolor
11903 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11904 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11905 by the "purity" of the color (that is, how saturated it already is).
11907 This filter is similar to the Adobe Photoshop Selective Color tool.
11909 The filter accepts the following options:
11912 @item correction_method
11913 Select color correction method.
11915 Available values are:
11918 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11921 Specified adjustments are relative to the original component value.
11923 Default is @code{absolute}.
11925 Adjustments for red pixels (pixels where the red component is the maximum)
11927 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11929 Adjustments for green pixels (pixels where the green component is the maximum)
11931 Adjustments for cyan pixels (pixels where the red component is the minimum)
11933 Adjustments for blue pixels (pixels where the blue component is the maximum)
11935 Adjustments for magenta pixels (pixels where the green component is the minimum)
11937 Adjustments for white pixels (pixels where all components are greater than 128)
11939 Adjustments for all pixels except pure black and pure white
11941 Adjustments for black pixels (pixels where all components are lesser than 128)
11943 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11946 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11947 4 space separated floating point adjustment values in the [-1,1] range,
11948 respectively to adjust the amount of cyan, magenta, yellow and black for the
11949 pixels of its range.
11951 @subsection Examples
11955 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11956 increase magenta by 27% in blue areas:
11958 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11962 Use a Photoshop selective color preset:
11964 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11968 @anchor{separatefields}
11969 @section separatefields
11971 The @code{separatefields} takes a frame-based video input and splits
11972 each frame into its components fields, producing a new half height clip
11973 with twice the frame rate and twice the frame count.
11975 This filter use field-dominance information in frame to decide which
11976 of each pair of fields to place first in the output.
11977 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
11979 @section setdar, setsar
11981 The @code{setdar} filter sets the Display Aspect Ratio for the filter
11984 This is done by changing the specified Sample (aka Pixel) Aspect
11985 Ratio, according to the following equation:
11987 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
11990 Keep in mind that the @code{setdar} filter does not modify the pixel
11991 dimensions of the video frame. Also, the display aspect ratio set by
11992 this filter may be changed by later filters in the filterchain,
11993 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
11996 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
11997 the filter output video.
11999 Note that as a consequence of the application of this filter, the
12000 output display aspect ratio will change according to the equation
12003 Keep in mind that the sample aspect ratio set by the @code{setsar}
12004 filter may be changed by later filters in the filterchain, e.g. if
12005 another "setsar" or a "setdar" filter is applied.
12007 It accepts the following parameters:
12010 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
12011 Set the aspect ratio used by the filter.
12013 The parameter can be a floating point number string, an expression, or
12014 a string of the form @var{num}:@var{den}, where @var{num} and
12015 @var{den} are the numerator and denominator of the aspect ratio. If
12016 the parameter is not specified, it is assumed the value "0".
12017 In case the form "@var{num}:@var{den}" is used, the @code{:} character
12021 Set the maximum integer value to use for expressing numerator and
12022 denominator when reducing the expressed aspect ratio to a rational.
12023 Default value is @code{100}.
12027 The parameter @var{sar} is an expression containing
12028 the following constants:
12032 These are approximated values for the mathematical constants e
12033 (Euler's number), pi (Greek pi), and phi (the golden ratio).
12036 The input width and height.
12039 These are the same as @var{w} / @var{h}.
12042 The input sample aspect ratio.
12045 The input display aspect ratio. It is the same as
12046 (@var{w} / @var{h}) * @var{sar}.
12049 Horizontal and vertical chroma subsample values. For example, for the
12050 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12053 @subsection Examples
12058 To change the display aspect ratio to 16:9, specify one of the following:
12065 To change the sample aspect ratio to 10:11, specify:
12071 To set a display aspect ratio of 16:9, and specify a maximum integer value of
12072 1000 in the aspect ratio reduction, use the command:
12074 setdar=ratio=16/9:max=1000
12082 Force field for the output video frame.
12084 The @code{setfield} filter marks the interlace type field for the
12085 output frames. It does not change the input frame, but only sets the
12086 corresponding property, which affects how the frame is treated by
12087 following filters (e.g. @code{fieldorder} or @code{yadif}).
12089 The filter accepts the following options:
12094 Available values are:
12098 Keep the same field property.
12101 Mark the frame as bottom-field-first.
12104 Mark the frame as top-field-first.
12107 Mark the frame as progressive.
12113 Show a line containing various information for each input video frame.
12114 The input video is not modified.
12116 The shown line contains a sequence of key/value pairs of the form
12117 @var{key}:@var{value}.
12119 The following values are shown in the output:
12123 The (sequential) number of the input frame, starting from 0.
12126 The Presentation TimeStamp of the input frame, expressed as a number of
12127 time base units. The time base unit depends on the filter input pad.
12130 The Presentation TimeStamp of the input frame, expressed as a number of
12134 The position of the frame in the input stream, or -1 if this information is
12135 unavailable and/or meaningless (for example in case of synthetic video).
12138 The pixel format name.
12141 The sample aspect ratio of the input frame, expressed in the form
12142 @var{num}/@var{den}.
12145 The size of the input frame. For the syntax of this option, check the
12146 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12149 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
12150 for bottom field first).
12153 This is 1 if the frame is a key frame, 0 otherwise.
12156 The picture type of the input frame ("I" for an I-frame, "P" for a
12157 P-frame, "B" for a B-frame, or "?" for an unknown type).
12158 Also refer to the documentation of the @code{AVPictureType} enum and of
12159 the @code{av_get_picture_type_char} function defined in
12160 @file{libavutil/avutil.h}.
12163 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
12165 @item plane_checksum
12166 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
12167 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
12170 @section showpalette
12172 Displays the 256 colors palette of each frame. This filter is only relevant for
12173 @var{pal8} pixel format frames.
12175 It accepts the following option:
12179 Set the size of the box used to represent one palette color entry. Default is
12180 @code{30} (for a @code{30x30} pixel box).
12183 @section shuffleframes
12185 Reorder and/or duplicate video frames.
12187 It accepts the following parameters:
12191 Set the destination indexes of input frames.
12192 This is space or '|' separated list of indexes that maps input frames to output
12193 frames. Number of indexes also sets maximal value that each index may have.
12196 The first frame has the index 0. The default is to keep the input unchanged.
12198 @subsection Examples
12202 Swap second and third frame of every three frames of the input:
12204 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
12208 Swap 10th and 1st frame of every ten frames of the input:
12210 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
12214 @section shuffleplanes
12216 Reorder and/or duplicate video planes.
12218 It accepts the following parameters:
12223 The index of the input plane to be used as the first output plane.
12226 The index of the input plane to be used as the second output plane.
12229 The index of the input plane to be used as the third output plane.
12232 The index of the input plane to be used as the fourth output plane.
12236 The first plane has the index 0. The default is to keep the input unchanged.
12238 @subsection Examples
12242 Swap the second and third planes of the input:
12244 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
12248 @anchor{signalstats}
12249 @section signalstats
12250 Evaluate various visual metrics that assist in determining issues associated
12251 with the digitization of analog video media.
12253 By default the filter will log these metadata values:
12257 Display the minimal Y value contained within the input frame. Expressed in
12261 Display the Y value at the 10% percentile within the input frame. Expressed in
12265 Display the average Y value within the input frame. Expressed in range of
12269 Display the Y value at the 90% percentile within the input frame. Expressed in
12273 Display the maximum Y value contained within the input frame. Expressed in
12277 Display the minimal U value contained within the input frame. Expressed in
12281 Display the U value at the 10% percentile within the input frame. Expressed in
12285 Display the average U value within the input frame. Expressed in range of
12289 Display the U value at the 90% percentile within the input frame. Expressed in
12293 Display the maximum U value contained within the input frame. Expressed in
12297 Display the minimal V value contained within the input frame. Expressed in
12301 Display the V value at the 10% percentile within the input frame. Expressed in
12305 Display the average V value within the input frame. Expressed in range of
12309 Display the V value at the 90% percentile within the input frame. Expressed in
12313 Display the maximum V value contained within the input frame. Expressed in
12317 Display the minimal saturation value contained within the input frame.
12318 Expressed in range of [0-~181.02].
12321 Display the saturation value at the 10% percentile within the input frame.
12322 Expressed in range of [0-~181.02].
12325 Display the average saturation value within the input frame. Expressed in range
12329 Display the saturation value at the 90% percentile within the input frame.
12330 Expressed in range of [0-~181.02].
12333 Display the maximum saturation value contained within the input frame.
12334 Expressed in range of [0-~181.02].
12337 Display the median value for hue within the input frame. Expressed in range of
12341 Display the average value for hue within the input frame. Expressed in range of
12345 Display the average of sample value difference between all values of the Y
12346 plane in the current frame and corresponding values of the previous input frame.
12347 Expressed in range of [0-255].
12350 Display the average of sample value difference between all values of the U
12351 plane in the current frame and corresponding values of the previous input frame.
12352 Expressed in range of [0-255].
12355 Display the average of sample value difference between all values of the V
12356 plane in the current frame and corresponding values of the previous input frame.
12357 Expressed in range of [0-255].
12360 Display bit depth of Y plane in current frame.
12361 Expressed in range of [0-16].
12364 Display bit depth of U plane in current frame.
12365 Expressed in range of [0-16].
12368 Display bit depth of V plane in current frame.
12369 Expressed in range of [0-16].
12372 The filter accepts the following options:
12378 @option{stat} specify an additional form of image analysis.
12379 @option{out} output video with the specified type of pixel highlighted.
12381 Both options accept the following values:
12385 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12386 unlike the neighboring pixels of the same field. Examples of temporal outliers
12387 include the results of video dropouts, head clogs, or tape tracking issues.
12390 Identify @var{vertical line repetition}. Vertical line repetition includes
12391 similar rows of pixels within a frame. In born-digital video vertical line
12392 repetition is common, but this pattern is uncommon in video digitized from an
12393 analog source. When it occurs in video that results from the digitization of an
12394 analog source it can indicate concealment from a dropout compensator.
12397 Identify pixels that fall outside of legal broadcast range.
12401 Set the highlight color for the @option{out} option. The default color is
12405 @subsection Examples
12409 Output data of various video metrics:
12411 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12415 Output specific data about the minimum and maximum values of the Y plane per frame:
12417 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12421 Playback video while highlighting pixels that are outside of broadcast range in red.
12423 ffplay example.mov -vf signalstats="out=brng:color=red"
12427 Playback video with signalstats metadata drawn over the frame.
12429 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12432 The contents of signalstat_drawtext.txt used in the command are:
12435 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12436 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12437 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12438 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12446 Blur the input video without impacting the outlines.
12448 It accepts the following options:
12451 @item luma_radius, lr
12452 Set the luma radius. The option value must be a float number in
12453 the range [0.1,5.0] that specifies the variance of the gaussian filter
12454 used to blur the image (slower if larger). Default value is 1.0.
12456 @item luma_strength, ls
12457 Set the luma strength. The option value must be a float number
12458 in the range [-1.0,1.0] that configures the blurring. A value included
12459 in [0.0,1.0] will blur the image whereas a value included in
12460 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12462 @item luma_threshold, lt
12463 Set the luma threshold used as a coefficient to determine
12464 whether a pixel should be blurred or not. The option value must be an
12465 integer in the range [-30,30]. A value of 0 will filter all the image,
12466 a value included in [0,30] will filter flat areas and a value included
12467 in [-30,0] will filter edges. Default value is 0.
12469 @item chroma_radius, cr
12470 Set the chroma radius. The option value must be a float number in
12471 the range [0.1,5.0] that specifies the variance of the gaussian filter
12472 used to blur the image (slower if larger). Default value is 1.0.
12474 @item chroma_strength, cs
12475 Set the chroma strength. The option value must be a float number
12476 in the range [-1.0,1.0] that configures the blurring. A value included
12477 in [0.0,1.0] will blur the image whereas a value included in
12478 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12480 @item chroma_threshold, ct
12481 Set the chroma threshold used as a coefficient to determine
12482 whether a pixel should be blurred or not. The option value must be an
12483 integer in the range [-30,30]. A value of 0 will filter all the image,
12484 a value included in [0,30] will filter flat areas and a value included
12485 in [-30,0] will filter edges. Default value is 0.
12488 If a chroma option is not explicitly set, the corresponding luma value
12493 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12495 This filter takes in input two input videos, the first input is
12496 considered the "main" source and is passed unchanged to the
12497 output. The second input is used as a "reference" video for computing
12500 Both video inputs must have the same resolution and pixel format for
12501 this filter to work correctly. Also it assumes that both inputs
12502 have the same number of frames, which are compared one by one.
12504 The filter stores the calculated SSIM of each frame.
12506 The description of the accepted parameters follows.
12509 @item stats_file, f
12510 If specified the filter will use the named file to save the SSIM of
12511 each individual frame. When filename equals "-" the data is sent to
12515 The file printed if @var{stats_file} is selected, contains a sequence of
12516 key/value pairs of the form @var{key}:@var{value} for each compared
12519 A description of each shown parameter follows:
12523 sequential number of the input frame, starting from 1
12525 @item Y, U, V, R, G, B
12526 SSIM of the compared frames for the component specified by the suffix.
12529 SSIM of the compared frames for the whole frame.
12532 Same as above but in dB representation.
12537 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12538 [main][ref] ssim="stats_file=stats.log" [out]
12541 On this example the input file being processed is compared with the
12542 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12543 is stored in @file{stats.log}.
12545 Another example with both psnr and ssim at same time:
12547 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12552 Convert between different stereoscopic image formats.
12554 The filters accept the following options:
12558 Set stereoscopic image format of input.
12560 Available values for input image formats are:
12563 side by side parallel (left eye left, right eye right)
12566 side by side crosseye (right eye left, left eye right)
12569 side by side parallel with half width resolution
12570 (left eye left, right eye right)
12573 side by side crosseye with half width resolution
12574 (right eye left, left eye right)
12577 above-below (left eye above, right eye below)
12580 above-below (right eye above, left eye below)
12583 above-below with half height resolution
12584 (left eye above, right eye below)
12587 above-below with half height resolution
12588 (right eye above, left eye below)
12591 alternating frames (left eye first, right eye second)
12594 alternating frames (right eye first, left eye second)
12597 interleaved rows (left eye has top row, right eye starts on next row)
12600 interleaved rows (right eye has top row, left eye starts on next row)
12603 interleaved columns, left eye first
12606 interleaved columns, right eye first
12608 Default value is @samp{sbsl}.
12612 Set stereoscopic image format of output.
12616 side by side parallel (left eye left, right eye right)
12619 side by side crosseye (right eye left, left eye right)
12622 side by side parallel with half width resolution
12623 (left eye left, right eye right)
12626 side by side crosseye with half width resolution
12627 (right eye left, left eye right)
12630 above-below (left eye above, right eye below)
12633 above-below (right eye above, left eye below)
12636 above-below with half height resolution
12637 (left eye above, right eye below)
12640 above-below with half height resolution
12641 (right eye above, left eye below)
12644 alternating frames (left eye first, right eye second)
12647 alternating frames (right eye first, left eye second)
12650 interleaved rows (left eye has top row, right eye starts on next row)
12653 interleaved rows (right eye has top row, left eye starts on next row)
12656 anaglyph red/blue gray
12657 (red filter on left eye, blue filter on right eye)
12660 anaglyph red/green gray
12661 (red filter on left eye, green filter on right eye)
12664 anaglyph red/cyan gray
12665 (red filter on left eye, cyan filter on right eye)
12668 anaglyph red/cyan half colored
12669 (red filter on left eye, cyan filter on right eye)
12672 anaglyph red/cyan color
12673 (red filter on left eye, cyan filter on right eye)
12676 anaglyph red/cyan color optimized with the least squares projection of dubois
12677 (red filter on left eye, cyan filter on right eye)
12680 anaglyph green/magenta gray
12681 (green filter on left eye, magenta filter on right eye)
12684 anaglyph green/magenta half colored
12685 (green filter on left eye, magenta filter on right eye)
12688 anaglyph green/magenta colored
12689 (green filter on left eye, magenta filter on right eye)
12692 anaglyph green/magenta color optimized with the least squares projection of dubois
12693 (green filter on left eye, magenta filter on right eye)
12696 anaglyph yellow/blue gray
12697 (yellow filter on left eye, blue filter on right eye)
12700 anaglyph yellow/blue half colored
12701 (yellow filter on left eye, blue filter on right eye)
12704 anaglyph yellow/blue colored
12705 (yellow filter on left eye, blue filter on right eye)
12708 anaglyph yellow/blue color optimized with the least squares projection of dubois
12709 (yellow filter on left eye, blue filter on right eye)
12712 mono output (left eye only)
12715 mono output (right eye only)
12718 checkerboard, left eye first
12721 checkerboard, right eye first
12724 interleaved columns, left eye first
12727 interleaved columns, right eye first
12733 Default value is @samp{arcd}.
12736 @subsection Examples
12740 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12746 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12752 @section streamselect, astreamselect
12753 Select video or audio streams.
12755 The filter accepts the following options:
12759 Set number of inputs. Default is 2.
12762 Set input indexes to remap to outputs.
12765 @subsection Commands
12767 The @code{streamselect} and @code{astreamselect} filter supports the following
12772 Set input indexes to remap to outputs.
12775 @subsection Examples
12779 Select first 5 seconds 1st stream and rest of time 2nd stream:
12781 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12785 Same as above, but for audio:
12787 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12792 Apply sobel operator to input video stream.
12794 The filter accepts the following option:
12798 Set which planes will be processed, unprocessed planes will be copied.
12799 By default value 0xf, all planes will be processed.
12802 Set value which will be multiplied with filtered result.
12805 Set value which will be added to filtered result.
12811 Apply a simple postprocessing filter that compresses and decompresses the image
12812 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12813 and average the results.
12815 The filter accepts the following options:
12819 Set quality. This option defines the number of levels for averaging. It accepts
12820 an integer in the range 0-6. If set to @code{0}, the filter will have no
12821 effect. A value of @code{6} means the higher quality. For each increment of
12822 that value the speed drops by a factor of approximately 2. Default value is
12826 Force a constant quantization parameter. If not set, the filter will use the QP
12827 from the video stream (if available).
12830 Set thresholding mode. Available modes are:
12834 Set hard thresholding (default).
12836 Set soft thresholding (better de-ringing effect, but likely blurrier).
12839 @item use_bframe_qp
12840 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12841 option may cause flicker since the B-Frames have often larger QP. Default is
12842 @code{0} (not enabled).
12848 Draw subtitles on top of input video using the libass library.
12850 To enable compilation of this filter you need to configure FFmpeg with
12851 @code{--enable-libass}. This filter also requires a build with libavcodec and
12852 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12853 Alpha) subtitles format.
12855 The filter accepts the following options:
12859 Set the filename of the subtitle file to read. It must be specified.
12861 @item original_size
12862 Specify the size of the original video, the video for which the ASS file
12863 was composed. For the syntax of this option, check the
12864 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12865 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12866 correctly scale the fonts if the aspect ratio has been changed.
12869 Set a directory path containing fonts that can be used by the filter.
12870 These fonts will be used in addition to whatever the font provider uses.
12873 Set subtitles input character encoding. @code{subtitles} filter only. Only
12874 useful if not UTF-8.
12876 @item stream_index, si
12877 Set subtitles stream index. @code{subtitles} filter only.
12880 Override default style or script info parameters of the subtitles. It accepts a
12881 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12884 If the first key is not specified, it is assumed that the first value
12885 specifies the @option{filename}.
12887 For example, to render the file @file{sub.srt} on top of the input
12888 video, use the command:
12893 which is equivalent to:
12895 subtitles=filename=sub.srt
12898 To render the default subtitles stream from file @file{video.mkv}, use:
12900 subtitles=video.mkv
12903 To render the second subtitles stream from that file, use:
12905 subtitles=video.mkv:si=1
12908 To make the subtitles stream from @file{sub.srt} appear in transparent green
12909 @code{DejaVu Serif}, use:
12911 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12914 @section super2xsai
12916 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12917 Interpolate) pixel art scaling algorithm.
12919 Useful for enlarging pixel art images without reducing sharpness.
12923 Swap two rectangular objects in video.
12925 This filter accepts the following options:
12935 Set 1st rect x coordinate.
12938 Set 1st rect y coordinate.
12941 Set 2nd rect x coordinate.
12944 Set 2nd rect y coordinate.
12946 All expressions are evaluated once for each frame.
12949 The all options are expressions containing the following constants:
12954 The input width and height.
12957 same as @var{w} / @var{h}
12960 input sample aspect ratio
12963 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12966 The number of the input frame, starting from 0.
12969 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12972 the position in the file of the input frame, NAN if unknown
12980 Apply telecine process to the video.
12982 This filter accepts the following options:
12991 The default value is @code{top}.
12995 A string of numbers representing the pulldown pattern you wish to apply.
12996 The default value is @code{23}.
13000 Some typical patterns:
13005 24p: 2332 (preferred)
13012 24p: 222222222223 ("Euro pulldown")
13018 Select the most representative frame in a given sequence of consecutive frames.
13020 The filter accepts the following options:
13024 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
13025 will pick one of them, and then handle the next batch of @var{n} frames until
13026 the end. Default is @code{100}.
13029 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
13030 value will result in a higher memory usage, so a high value is not recommended.
13032 @subsection Examples
13036 Extract one picture each 50 frames:
13042 Complete example of a thumbnail creation with @command{ffmpeg}:
13044 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
13050 Tile several successive frames together.
13052 The filter accepts the following options:
13057 Set the grid size (i.e. the number of lines and columns). For the syntax of
13058 this option, check the
13059 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13062 Set the maximum number of frames to render in the given area. It must be less
13063 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
13064 the area will be used.
13067 Set the outer border margin in pixels.
13070 Set the inner border thickness (i.e. the number of pixels between frames). For
13071 more advanced padding options (such as having different values for the edges),
13072 refer to the pad video filter.
13075 Specify the color of the unused area. For the syntax of this option, check the
13076 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
13080 @subsection Examples
13084 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
13086 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
13088 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
13089 duplicating each output frame to accommodate the originally detected frame
13093 Display @code{5} pictures in an area of @code{3x2} frames,
13094 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
13095 mixed flat and named options:
13097 tile=3x2:nb_frames=5:padding=7:margin=2
13101 @section tinterlace
13103 Perform various types of temporal field interlacing.
13105 Frames are counted starting from 1, so the first input frame is
13108 The filter accepts the following options:
13113 Specify the mode of the interlacing. This option can also be specified
13114 as a value alone. See below for a list of values for this option.
13116 Available values are:
13120 Move odd frames into the upper field, even into the lower field,
13121 generating a double height frame at half frame rate.
13125 Frame 1 Frame 2 Frame 3 Frame 4
13127 11111 22222 33333 44444
13128 11111 22222 33333 44444
13129 11111 22222 33333 44444
13130 11111 22222 33333 44444
13144 Only output odd frames, even frames are dropped, generating a frame with
13145 unchanged height at half frame rate.
13150 Frame 1 Frame 2 Frame 3 Frame 4
13152 11111 22222 33333 44444
13153 11111 22222 33333 44444
13154 11111 22222 33333 44444
13155 11111 22222 33333 44444
13165 Only output even frames, odd frames are dropped, generating a frame with
13166 unchanged height at half frame rate.
13171 Frame 1 Frame 2 Frame 3 Frame 4
13173 11111 22222 33333 44444
13174 11111 22222 33333 44444
13175 11111 22222 33333 44444
13176 11111 22222 33333 44444
13186 Expand each frame to full height, but pad alternate lines with black,
13187 generating a frame with double height at the same input frame rate.
13192 Frame 1 Frame 2 Frame 3 Frame 4
13194 11111 22222 33333 44444
13195 11111 22222 33333 44444
13196 11111 22222 33333 44444
13197 11111 22222 33333 44444
13200 11111 ..... 33333 .....
13201 ..... 22222 ..... 44444
13202 11111 ..... 33333 .....
13203 ..... 22222 ..... 44444
13204 11111 ..... 33333 .....
13205 ..... 22222 ..... 44444
13206 11111 ..... 33333 .....
13207 ..... 22222 ..... 44444
13211 @item interleave_top, 4
13212 Interleave the upper field from odd frames with the lower field from
13213 even frames, generating a frame with unchanged height at half frame rate.
13218 Frame 1 Frame 2 Frame 3 Frame 4
13220 11111<- 22222 33333<- 44444
13221 11111 22222<- 33333 44444<-
13222 11111<- 22222 33333<- 44444
13223 11111 22222<- 33333 44444<-
13233 @item interleave_bottom, 5
13234 Interleave the lower field from odd frames with the upper field from
13235 even frames, generating a frame with unchanged height at half frame rate.
13240 Frame 1 Frame 2 Frame 3 Frame 4
13242 11111 22222<- 33333 44444<-
13243 11111<- 22222 33333<- 44444
13244 11111 22222<- 33333 44444<-
13245 11111<- 22222 33333<- 44444
13255 @item interlacex2, 6
13256 Double frame rate with unchanged height. Frames are inserted each
13257 containing the second temporal field from the previous input frame and
13258 the first temporal field from the next input frame. This mode relies on
13259 the top_field_first flag. Useful for interlaced video displays with no
13260 field synchronisation.
13265 Frame 1 Frame 2 Frame 3 Frame 4
13267 11111 22222 33333 44444
13268 11111 22222 33333 44444
13269 11111 22222 33333 44444
13270 11111 22222 33333 44444
13273 11111 22222 22222 33333 33333 44444 44444
13274 11111 11111 22222 22222 33333 33333 44444
13275 11111 22222 22222 33333 33333 44444 44444
13276 11111 11111 22222 22222 33333 33333 44444
13281 Move odd frames into the upper field, even into the lower field,
13282 generating a double height frame at same frame rate.
13287 Frame 1 Frame 2 Frame 3 Frame 4
13289 11111 22222 33333 44444
13290 11111 22222 33333 44444
13291 11111 22222 33333 44444
13292 11111 22222 33333 44444
13295 11111 33333 33333 55555
13296 22222 22222 44444 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
13307 Numeric values are deprecated but are accepted for backward
13308 compatibility reasons.
13310 Default mode is @code{merge}.
13313 Specify flags influencing the filter process.
13315 Available value for @var{flags} is:
13318 @item low_pass_filter, vlfp
13319 Enable vertical low-pass filtering in the filter.
13320 Vertical low-pass filtering is required when creating an interlaced
13321 destination from a progressive source which contains high-frequency
13322 vertical detail. Filtering will reduce interlace 'twitter' and Moire
13325 Vertical low-pass filtering can only be enabled for @option{mode}
13326 @var{interleave_top} and @var{interleave_bottom}.
13333 Transpose rows with columns in the input video and optionally flip it.
13335 It accepts the following parameters:
13340 Specify the transposition direction.
13342 Can assume the following values:
13344 @item 0, 4, cclock_flip
13345 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
13353 Rotate by 90 degrees clockwise, that is:
13361 Rotate by 90 degrees counterclockwise, that is:
13368 @item 3, 7, clock_flip
13369 Rotate by 90 degrees clockwise and vertically flip, that is:
13377 For values between 4-7, the transposition is only done if the input
13378 video geometry is portrait and not landscape. These values are
13379 deprecated, the @code{passthrough} option should be used instead.
13381 Numerical values are deprecated, and should be dropped in favor of
13382 symbolic constants.
13385 Do not apply the transposition if the input geometry matches the one
13386 specified by the specified value. It accepts the following values:
13389 Always apply transposition.
13391 Preserve portrait geometry (when @var{height} >= @var{width}).
13393 Preserve landscape geometry (when @var{width} >= @var{height}).
13396 Default value is @code{none}.
13399 For example to rotate by 90 degrees clockwise and preserve portrait
13402 transpose=dir=1:passthrough=portrait
13405 The command above can also be specified as:
13407 transpose=1:portrait
13411 Trim the input so that the output contains one continuous subpart of the input.
13413 It accepts the following parameters:
13416 Specify the time of the start of the kept section, i.e. the frame with the
13417 timestamp @var{start} will be the first frame in the output.
13420 Specify the time of the first frame that will be dropped, i.e. the frame
13421 immediately preceding the one with the timestamp @var{end} will be the last
13422 frame in the output.
13425 This is the same as @var{start}, except this option sets the start timestamp
13426 in timebase units instead of seconds.
13429 This is the same as @var{end}, except this option sets the end timestamp
13430 in timebase units instead of seconds.
13433 The maximum duration of the output in seconds.
13436 The number of the first frame that should be passed to the output.
13439 The number of the first frame that should be dropped.
13442 @option{start}, @option{end}, and @option{duration} are expressed as time
13443 duration specifications; see
13444 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13445 for the accepted syntax.
13447 Note that the first two sets of the start/end options and the @option{duration}
13448 option look at the frame timestamp, while the _frame variants simply count the
13449 frames that pass through the filter. Also note that this filter does not modify
13450 the timestamps. If you wish for the output timestamps to start at zero, insert a
13451 setpts filter after the trim filter.
13453 If multiple start or end options are set, this filter tries to be greedy and
13454 keep all the frames that match at least one of the specified constraints. To keep
13455 only the part that matches all the constraints at once, chain multiple trim
13458 The defaults are such that all the input is kept. So it is possible to set e.g.
13459 just the end values to keep everything before the specified time.
13464 Drop everything except the second minute of input:
13466 ffmpeg -i INPUT -vf trim=60:120
13470 Keep only the first second:
13472 ffmpeg -i INPUT -vf trim=duration=1
13481 Sharpen or blur the input video.
13483 It accepts the following parameters:
13486 @item luma_msize_x, lx
13487 Set the luma matrix horizontal size. It must be an odd integer between
13488 3 and 23. The default value is 5.
13490 @item luma_msize_y, ly
13491 Set the luma matrix vertical size. It must be an odd integer between 3
13492 and 23. The default value is 5.
13494 @item luma_amount, la
13495 Set the luma effect strength. It must be a floating point number, reasonable
13496 values lay between -1.5 and 1.5.
13498 Negative values will blur the input video, while positive values will
13499 sharpen it, a value of zero will disable the effect.
13501 Default value is 1.0.
13503 @item chroma_msize_x, cx
13504 Set the chroma matrix horizontal size. It must be an odd integer
13505 between 3 and 23. The default value is 5.
13507 @item chroma_msize_y, cy
13508 Set the chroma matrix vertical size. It must be an odd integer
13509 between 3 and 23. The default value is 5.
13511 @item chroma_amount, ca
13512 Set the chroma effect strength. It must be a floating point number, reasonable
13513 values lay between -1.5 and 1.5.
13515 Negative values will blur the input video, while positive values will
13516 sharpen it, a value of zero will disable the effect.
13518 Default value is 0.0.
13521 If set to 1, specify using OpenCL capabilities, only available if
13522 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13526 All parameters are optional and default to the equivalent of the
13527 string '5:5:1.0:5:5:0.0'.
13529 @subsection Examples
13533 Apply strong luma sharpen effect:
13535 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13539 Apply a strong blur of both luma and chroma parameters:
13541 unsharp=7:7:-2:7:7:-2
13547 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13548 the image at several (or - in the case of @option{quality} level @code{8} - all)
13549 shifts and average the results.
13551 The way this differs from the behavior of spp is that uspp actually encodes &
13552 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13553 DCT similar to MJPEG.
13555 The filter accepts the following options:
13559 Set quality. This option defines the number of levels for averaging. It accepts
13560 an integer in the range 0-8. If set to @code{0}, the filter will have no
13561 effect. A value of @code{8} means the higher quality. For each increment of
13562 that value the speed drops by a factor of approximately 2. Default value is
13566 Force a constant quantization parameter. If not set, the filter will use the QP
13567 from the video stream (if available).
13570 @section vaguedenoiser
13572 Apply a wavelet based denoiser.
13574 It transforms each frame from the video input into the wavelet domain,
13575 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
13576 the obtained coefficients. It does an inverse wavelet transform after.
13577 Due to wavelet properties, it should give a nice smoothed result, and
13578 reduced noise, without blurring picture features.
13580 This filter accepts the following options:
13584 The filtering strength. The higher, the more filtered the video will be.
13585 Hard thresholding can use a higher threshold than soft thresholding
13586 before the video looks overfiltered.
13589 The filtering method the filter will use.
13591 It accepts the following values:
13594 All values under the threshold will be zeroed.
13597 All values under the threshold will be zeroed. All values above will be
13598 reduced by the threshold.
13601 Scales or nullifies coefficients - intermediary between (more) soft and
13602 (less) hard thresholding.
13606 Number of times, the wavelet will decompose the picture. Picture can't
13607 be decomposed beyond a particular point (typically, 8 for a 640x480
13608 frame - as 2^9 = 512 > 480)
13611 Partial of full denoising (limited coefficients shrinking), from 0 to 100.
13614 A list of the planes to process. By default all planes are processed.
13617 @section vectorscope
13619 Display 2 color component values in the two dimensional graph (which is called
13622 This filter accepts the following options:
13626 Set vectorscope mode.
13628 It accepts the following values:
13631 Gray values are displayed on graph, higher brightness means more pixels have
13632 same component color value on location in graph. This is the default mode.
13635 Gray values are displayed on graph. Surrounding pixels values which are not
13636 present in video frame are drawn in gradient of 2 color components which are
13637 set by option @code{x} and @code{y}. The 3rd color component is static.
13640 Actual color components values present in video frame are displayed on graph.
13643 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13644 on graph increases value of another color component, which is luminance by
13645 default values of @code{x} and @code{y}.
13648 Actual colors present in video frame are displayed on graph. If two different
13649 colors map to same position on graph then color with higher value of component
13650 not present in graph is picked.
13653 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13654 component picked from radial gradient.
13658 Set which color component will be represented on X-axis. Default is @code{1}.
13661 Set which color component will be represented on Y-axis. Default is @code{2}.
13664 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13665 of color component which represents frequency of (X, Y) location in graph.
13670 No envelope, this is default.
13673 Instant envelope, even darkest single pixel will be clearly highlighted.
13676 Hold maximum and minimum values presented in graph over time. This way you
13677 can still spot out of range values without constantly looking at vectorscope.
13680 Peak and instant envelope combined together.
13684 Set what kind of graticule to draw.
13692 Set graticule opacity.
13695 Set graticule flags.
13699 Draw graticule for white point.
13702 Draw graticule for black point.
13705 Draw color points short names.
13709 Set background opacity.
13711 @item lthreshold, l
13712 Set low threshold for color component not represented on X or Y axis.
13713 Values lower than this value will be ignored. Default is 0.
13714 Note this value is multiplied with actual max possible value one pixel component
13715 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13718 @item hthreshold, h
13719 Set high threshold for color component not represented on X or Y axis.
13720 Values higher than this value will be ignored. Default is 1.
13721 Note this value is multiplied with actual max possible value one pixel component
13722 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13723 is 0.9 * 255 = 230.
13725 @item colorspace, c
13726 Set what kind of colorspace to use when drawing graticule.
13735 @anchor{vidstabdetect}
13736 @section vidstabdetect
13738 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13739 @ref{vidstabtransform} for pass 2.
13741 This filter generates a file with relative translation and rotation
13742 transform information about subsequent frames, which is then used by
13743 the @ref{vidstabtransform} filter.
13745 To enable compilation of this filter you need to configure FFmpeg with
13746 @code{--enable-libvidstab}.
13748 This filter accepts the following options:
13752 Set the path to the file used to write the transforms information.
13753 Default value is @file{transforms.trf}.
13756 Set how shaky the video is and how quick the camera is. It accepts an
13757 integer in the range 1-10, a value of 1 means little shakiness, a
13758 value of 10 means strong shakiness. Default value is 5.
13761 Set the accuracy of the detection process. It must be a value in the
13762 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13763 accuracy. Default value is 15.
13766 Set stepsize of the search process. The region around minimum is
13767 scanned with 1 pixel resolution. Default value is 6.
13770 Set minimum contrast. Below this value a local measurement field is
13771 discarded. Must be a floating point value in the range 0-1. Default
13775 Set reference frame number for tripod mode.
13777 If enabled, the motion of the frames is compared to a reference frame
13778 in the filtered stream, identified by the specified number. The idea
13779 is to compensate all movements in a more-or-less static scene and keep
13780 the camera view absolutely still.
13782 If set to 0, it is disabled. The frames are counted starting from 1.
13785 Show fields and transforms in the resulting frames. It accepts an
13786 integer in the range 0-2. Default value is 0, which disables any
13790 @subsection Examples
13794 Use default values:
13800 Analyze strongly shaky movie and put the results in file
13801 @file{mytransforms.trf}:
13803 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13807 Visualize the result of internal transformations in the resulting
13810 vidstabdetect=show=1
13814 Analyze a video with medium shakiness using @command{ffmpeg}:
13816 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13820 @anchor{vidstabtransform}
13821 @section vidstabtransform
13823 Video stabilization/deshaking: pass 2 of 2,
13824 see @ref{vidstabdetect} for pass 1.
13826 Read a file with transform information for each frame and
13827 apply/compensate them. Together with the @ref{vidstabdetect}
13828 filter this can be used to deshake videos. See also
13829 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13830 the @ref{unsharp} filter, see below.
13832 To enable compilation of this filter you need to configure FFmpeg with
13833 @code{--enable-libvidstab}.
13835 @subsection Options
13839 Set path to the file used to read the transforms. Default value is
13840 @file{transforms.trf}.
13843 Set the number of frames (value*2 + 1) used for lowpass filtering the
13844 camera movements. Default value is 10.
13846 For example a number of 10 means that 21 frames are used (10 in the
13847 past and 10 in the future) to smoothen the motion in the video. A
13848 larger value leads to a smoother video, but limits the acceleration of
13849 the camera (pan/tilt movements). 0 is a special case where a static
13850 camera is simulated.
13853 Set the camera path optimization algorithm.
13855 Accepted values are:
13858 gaussian kernel low-pass filter on camera motion (default)
13860 averaging on transformations
13864 Set maximal number of pixels to translate frames. Default value is -1,
13868 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13869 value is -1, meaning no limit.
13872 Specify how to deal with borders that may be visible due to movement
13875 Available values are:
13878 keep image information from previous frame (default)
13880 fill the border black
13884 Invert transforms if set to 1. Default value is 0.
13887 Consider transforms as relative to previous frame if set to 1,
13888 absolute if set to 0. Default value is 0.
13891 Set percentage to zoom. A positive value will result in a zoom-in
13892 effect, a negative value in a zoom-out effect. Default value is 0 (no
13896 Set optimal zooming to avoid borders.
13898 Accepted values are:
13903 optimal static zoom value is determined (only very strong movements
13904 will lead to visible borders) (default)
13906 optimal adaptive zoom value is determined (no borders will be
13907 visible), see @option{zoomspeed}
13910 Note that the value given at zoom is added to the one calculated here.
13913 Set percent to zoom maximally each frame (enabled when
13914 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13918 Specify type of interpolation.
13920 Available values are:
13925 linear only horizontal
13927 linear in both directions (default)
13929 cubic in both directions (slow)
13933 Enable virtual tripod mode if set to 1, which is equivalent to
13934 @code{relative=0:smoothing=0}. Default value is 0.
13936 Use also @code{tripod} option of @ref{vidstabdetect}.
13939 Increase log verbosity if set to 1. Also the detected global motions
13940 are written to the temporary file @file{global_motions.trf}. Default
13944 @subsection Examples
13948 Use @command{ffmpeg} for a typical stabilization with default values:
13950 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13953 Note the use of the @ref{unsharp} filter which is always recommended.
13956 Zoom in a bit more and load transform data from a given file:
13958 vidstabtransform=zoom=5:input="mytransforms.trf"
13962 Smoothen the video even more:
13964 vidstabtransform=smoothing=30
13970 Flip the input video vertically.
13972 For example, to vertically flip a video with @command{ffmpeg}:
13974 ffmpeg -i in.avi -vf "vflip" out.avi
13980 Make or reverse a natural vignetting effect.
13982 The filter accepts the following options:
13986 Set lens angle expression as a number of radians.
13988 The value is clipped in the @code{[0,PI/2]} range.
13990 Default value: @code{"PI/5"}
13994 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
13998 Set forward/backward mode.
14000 Available modes are:
14003 The larger the distance from the central point, the darker the image becomes.
14006 The larger the distance from the central point, the brighter the image becomes.
14007 This can be used to reverse a vignette effect, though there is no automatic
14008 detection to extract the lens @option{angle} and other settings (yet). It can
14009 also be used to create a burning effect.
14012 Default value is @samp{forward}.
14015 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
14017 It accepts the following values:
14020 Evaluate expressions only once during the filter initialization.
14023 Evaluate expressions for each incoming frame. This is way slower than the
14024 @samp{init} mode since it requires all the scalers to be re-computed, but it
14025 allows advanced dynamic expressions.
14028 Default value is @samp{init}.
14031 Set dithering to reduce the circular banding effects. Default is @code{1}
14035 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
14036 Setting this value to the SAR of the input will make a rectangular vignetting
14037 following the dimensions of the video.
14039 Default is @code{1/1}.
14042 @subsection Expressions
14044 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
14045 following parameters.
14050 input width and height
14053 the number of input frame, starting from 0
14056 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
14057 @var{TB} units, NAN if undefined
14060 frame rate of the input video, NAN if the input frame rate is unknown
14063 the PTS (Presentation TimeStamp) of the filtered video frame,
14064 expressed in seconds, NAN if undefined
14067 time base of the input video
14071 @subsection Examples
14075 Apply simple strong vignetting effect:
14081 Make a flickering vignetting:
14083 vignette='PI/4+random(1)*PI/50':eval=frame
14089 Stack input videos vertically.
14091 All streams must be of same pixel format and of same width.
14093 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
14094 to create same output.
14096 The filter accept the following option:
14100 Set number of input streams. Default is 2.
14103 If set to 1, force the output to terminate when the shortest input
14104 terminates. Default value is 0.
14109 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
14110 Deinterlacing Filter").
14112 Based on the process described by Martin Weston for BBC R&D, and
14113 implemented based on the de-interlace algorithm written by Jim
14114 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
14115 uses filter coefficients calculated by BBC R&D.
14117 There are two sets of filter coefficients, so called "simple":
14118 and "complex". Which set of filter coefficients is used can
14119 be set by passing an optional parameter:
14123 Set the interlacing filter coefficients. Accepts one of the following values:
14127 Simple filter coefficient set.
14129 More-complex filter coefficient set.
14131 Default value is @samp{complex}.
14134 Specify which frames to deinterlace. Accept one of the following values:
14138 Deinterlace all frames,
14140 Only deinterlace frames marked as interlaced.
14143 Default value is @samp{all}.
14147 Video waveform monitor.
14149 The waveform monitor plots color component intensity. By default luminance
14150 only. Each column of the waveform corresponds to a column of pixels in the
14153 It accepts the following options:
14157 Can be either @code{row}, or @code{column}. Default is @code{column}.
14158 In row mode, the graph on the left side represents color component value 0 and
14159 the right side represents value = 255. In column mode, the top side represents
14160 color component value = 0 and bottom side represents value = 255.
14163 Set intensity. Smaller values are useful to find out how many values of the same
14164 luminance are distributed across input rows/columns.
14165 Default value is @code{0.04}. Allowed range is [0, 1].
14168 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
14169 In mirrored mode, higher values will be represented on the left
14170 side for @code{row} mode and at the top for @code{column} mode. Default is
14171 @code{1} (mirrored).
14175 It accepts the following values:
14178 Presents information identical to that in the @code{parade}, except
14179 that the graphs representing color components are superimposed directly
14182 This display mode makes it easier to spot relative differences or similarities
14183 in overlapping areas of the color components that are supposed to be identical,
14184 such as neutral whites, grays, or blacks.
14187 Display separate graph for the color components side by side in
14188 @code{row} mode or one below the other in @code{column} mode.
14191 Display separate graph for the color components side by side in
14192 @code{column} mode or one below the other in @code{row} mode.
14194 Using this display mode makes it easy to spot color casts in the highlights
14195 and shadows of an image, by comparing the contours of the top and the bottom
14196 graphs of each waveform. Since whites, grays, and blacks are characterized
14197 by exactly equal amounts of red, green, and blue, neutral areas of the picture
14198 should display three waveforms of roughly equal width/height. If not, the
14199 correction is easy to perform by making level adjustments the three waveforms.
14201 Default is @code{stack}.
14203 @item components, c
14204 Set which color components to display. Default is 1, which means only luminance
14205 or red color component if input is in RGB colorspace. If is set for example to
14206 7 it will display all 3 (if) available color components.
14211 No envelope, this is default.
14214 Instant envelope, minimum and maximum values presented in graph will be easily
14215 visible even with small @code{step} value.
14218 Hold minimum and maximum values presented in graph across time. This way you
14219 can still spot out of range values without constantly looking at waveforms.
14222 Peak and instant envelope combined together.
14228 No filtering, this is default.
14231 Luma and chroma combined together.
14234 Similar as above, but shows difference between blue and red chroma.
14237 Displays only chroma.
14240 Displays actual color value on waveform.
14243 Similar as above, but with luma showing frequency of chroma values.
14247 Set which graticule to display.
14251 Do not display graticule.
14254 Display green graticule showing legal broadcast ranges.
14258 Set graticule opacity.
14261 Set graticule flags.
14265 Draw numbers above lines. By default enabled.
14268 Draw dots instead of lines.
14272 Set scale used for displaying graticule.
14279 Default is digital.
14282 Set background opacity.
14287 The @code{weave} takes a field-based video input and join
14288 each two sequential fields into single frame, producing a new double
14289 height clip with half the frame rate and half the frame count.
14291 It accepts the following option:
14295 Set first field. Available values are:
14299 Set the frame as top-field-first.
14302 Set the frame as bottom-field-first.
14306 @subsection Examples
14310 Interlace video using @ref{select} and @ref{separatefields} filter:
14312 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
14317 Apply the xBR high-quality magnification filter which is designed for pixel
14318 art. It follows a set of edge-detection rules, see
14319 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
14321 It accepts the following option:
14325 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
14326 @code{3xBR} and @code{4} for @code{4xBR}.
14327 Default is @code{3}.
14333 Deinterlace the input video ("yadif" means "yet another deinterlacing
14336 It accepts the following parameters:
14342 The interlacing mode to adopt. It accepts one of the following values:
14345 @item 0, send_frame
14346 Output one frame for each frame.
14347 @item 1, send_field
14348 Output one frame for each field.
14349 @item 2, send_frame_nospatial
14350 Like @code{send_frame}, but it skips the spatial interlacing check.
14351 @item 3, send_field_nospatial
14352 Like @code{send_field}, but it skips the spatial interlacing check.
14355 The default value is @code{send_frame}.
14358 The picture field parity assumed for the input interlaced video. It accepts one
14359 of the following values:
14363 Assume the top field is first.
14365 Assume the bottom field is first.
14367 Enable automatic detection of field parity.
14370 The default value is @code{auto}.
14371 If the interlacing is unknown or the decoder does not export this information,
14372 top field first will be assumed.
14375 Specify which frames to deinterlace. Accept one of the following
14380 Deinterlace all frames.
14381 @item 1, interlaced
14382 Only deinterlace frames marked as interlaced.
14385 The default value is @code{all}.
14390 Apply Zoom & Pan effect.
14392 This filter accepts the following options:
14396 Set the zoom expression. Default is 1.
14400 Set the x and y expression. Default is 0.
14403 Set the duration expression in number of frames.
14404 This sets for how many number of frames effect will last for
14405 single input image.
14408 Set the output image size, default is 'hd720'.
14411 Set the output frame rate, default is '25'.
14414 Each expression can contain the following constants:
14433 Output frame count.
14437 Last calculated 'x' and 'y' position from 'x' and 'y' expression
14438 for current input frame.
14442 'x' and 'y' of last output frame of previous input frame or 0 when there was
14443 not yet such frame (first input frame).
14446 Last calculated zoom from 'z' expression for current input frame.
14449 Last calculated zoom of last output frame of previous input frame.
14452 Number of output frames for current input frame. Calculated from 'd' expression
14453 for each input frame.
14456 number of output frames created for previous input frame
14459 Rational number: input width / input height
14462 sample aspect ratio
14465 display aspect ratio
14469 @subsection Examples
14473 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
14475 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
14479 Zoom-in up to 1.5 and pan always at center of picture:
14481 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14485 Same as above but without pausing:
14487 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14492 Scale (resize) the input video, using the z.lib library:
14493 https://github.com/sekrit-twc/zimg.
14495 The zscale filter forces the output display aspect ratio to be the same
14496 as the input, by changing the output sample aspect ratio.
14498 If the input image format is different from the format requested by
14499 the next filter, the zscale filter will convert the input to the
14502 @subsection Options
14503 The filter accepts the following options.
14508 Set the output video dimension expression. Default value is the input
14511 If the @var{width} or @var{w} is 0, the input width is used for the output.
14512 If the @var{height} or @var{h} is 0, the input height is used for the output.
14514 If one of the values is -1, the zscale filter will use a value that
14515 maintains the aspect ratio of the input image, calculated from the
14516 other specified dimension. If both of them are -1, the input size is
14519 If one of the values is -n with n > 1, the zscale filter will also use a value
14520 that maintains the aspect ratio of the input image, calculated from the other
14521 specified dimension. After that it will, however, make sure that the calculated
14522 dimension is divisible by n and adjust the value if necessary.
14524 See below for the list of accepted constants for use in the dimension
14528 Set the video size. For the syntax of this option, check the
14529 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14532 Set the dither type.
14534 Possible values are:
14539 @item error_diffusion
14545 Set the resize filter type.
14547 Possible values are:
14557 Default is bilinear.
14560 Set the color range.
14562 Possible values are:
14569 Default is same as input.
14572 Set the color primaries.
14574 Possible values are:
14584 Default is same as input.
14587 Set the transfer characteristics.
14589 Possible values are:
14600 Default is same as input.
14603 Set the colorspace matrix.
14605 Possible value are:
14616 Default is same as input.
14619 Set the input color range.
14621 Possible values are:
14628 Default is same as input.
14630 @item primariesin, pin
14631 Set the input color primaries.
14633 Possible values are:
14643 Default is same as input.
14645 @item transferin, tin
14646 Set the input transfer characteristics.
14648 Possible values are:
14659 Default is same as input.
14661 @item matrixin, min
14662 Set the input colorspace matrix.
14664 Possible value are:
14676 Set the output chroma location.
14678 Possible values are:
14689 @item chromalin, cin
14690 Set the input chroma location.
14692 Possible values are:
14704 The values of the @option{w} and @option{h} options are expressions
14705 containing the following constants:
14710 The input width and height
14714 These are the same as @var{in_w} and @var{in_h}.
14718 The output (scaled) width and height
14722 These are the same as @var{out_w} and @var{out_h}
14725 The same as @var{iw} / @var{ih}
14728 input sample aspect ratio
14731 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14735 horizontal and vertical input chroma subsample values. For example for the
14736 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14740 horizontal and vertical output chroma subsample values. For example for the
14741 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14747 @c man end VIDEO FILTERS
14749 @chapter Video Sources
14750 @c man begin VIDEO SOURCES
14752 Below is a description of the currently available video sources.
14756 Buffer video frames, and make them available to the filter chain.
14758 This source is mainly intended for a programmatic use, in particular
14759 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14761 It accepts the following parameters:
14766 Specify the size (width and height) of the buffered video frames. For the
14767 syntax of this option, check the
14768 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14771 The input video width.
14774 The input video height.
14777 A string representing the pixel format of the buffered video frames.
14778 It may be a number corresponding to a pixel format, or a pixel format
14782 Specify the timebase assumed by the timestamps of the buffered frames.
14785 Specify the frame rate expected for the video stream.
14787 @item pixel_aspect, sar
14788 The sample (pixel) aspect ratio of the input video.
14791 Specify the optional parameters to be used for the scale filter which
14792 is automatically inserted when an input change is detected in the
14793 input size or format.
14795 @item hw_frames_ctx
14796 When using a hardware pixel format, this should be a reference to an
14797 AVHWFramesContext describing input frames.
14802 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14805 will instruct the source to accept video frames with size 320x240 and
14806 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14807 square pixels (1:1 sample aspect ratio).
14808 Since the pixel format with name "yuv410p" corresponds to the number 6
14809 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14810 this example corresponds to:
14812 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14815 Alternatively, the options can be specified as a flat string, but this
14816 syntax is deprecated:
14818 @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}]
14822 Create a pattern generated by an elementary cellular automaton.
14824 The initial state of the cellular automaton can be defined through the
14825 @option{filename} and @option{pattern} options. If such options are
14826 not specified an initial state is created randomly.
14828 At each new frame a new row in the video is filled with the result of
14829 the cellular automaton next generation. The behavior when the whole
14830 frame is filled is defined by the @option{scroll} option.
14832 This source accepts the following options:
14836 Read the initial cellular automaton state, i.e. the starting row, from
14837 the specified file.
14838 In the file, each non-whitespace character is considered an alive
14839 cell, a newline will terminate the row, and further characters in the
14840 file will be ignored.
14843 Read the initial cellular automaton state, i.e. the starting row, from
14844 the specified string.
14846 Each non-whitespace character in the string is considered an alive
14847 cell, a newline will terminate the row, and further characters in the
14848 string will be ignored.
14851 Set the video rate, that is the number of frames generated per second.
14854 @item random_fill_ratio, ratio
14855 Set the random fill ratio for the initial cellular automaton row. It
14856 is a floating point number value ranging from 0 to 1, defaults to
14859 This option is ignored when a file or a pattern is specified.
14861 @item random_seed, seed
14862 Set the seed for filling randomly the initial row, must be an integer
14863 included between 0 and UINT32_MAX. If not specified, or if explicitly
14864 set to -1, the filter will try to use a good random seed on a best
14868 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14869 Default value is 110.
14872 Set the size of the output video. For the syntax of this option, check the
14873 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14875 If @option{filename} or @option{pattern} is specified, the size is set
14876 by default to the width of the specified initial state row, and the
14877 height is set to @var{width} * PHI.
14879 If @option{size} is set, it must contain the width of the specified
14880 pattern string, and the specified pattern will be centered in the
14883 If a filename or a pattern string is not specified, the size value
14884 defaults to "320x518" (used for a randomly generated initial state).
14887 If set to 1, scroll the output upward when all the rows in the output
14888 have been already filled. If set to 0, the new generated row will be
14889 written over the top row just after the bottom row is filled.
14892 @item start_full, full
14893 If set to 1, completely fill the output with generated rows before
14894 outputting the first frame.
14895 This is the default behavior, for disabling set the value to 0.
14898 If set to 1, stitch the left and right row edges together.
14899 This is the default behavior, for disabling set the value to 0.
14902 @subsection Examples
14906 Read the initial state from @file{pattern}, and specify an output of
14909 cellauto=f=pattern:s=200x400
14913 Generate a random initial row with a width of 200 cells, with a fill
14916 cellauto=ratio=2/3:s=200x200
14920 Create a pattern generated by rule 18 starting by a single alive cell
14921 centered on an initial row with width 100:
14923 cellauto=p=@@:s=100x400:full=0:rule=18
14927 Specify a more elaborated initial pattern:
14929 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14934 @anchor{coreimagesrc}
14935 @section coreimagesrc
14936 Video source generated on GPU using Apple's CoreImage API on OSX.
14938 This video source is a specialized version of the @ref{coreimage} video filter.
14939 Use a core image generator at the beginning of the applied filterchain to
14940 generate the content.
14942 The coreimagesrc video source accepts the following options:
14944 @item list_generators
14945 List all available generators along with all their respective options as well as
14946 possible minimum and maximum values along with the default values.
14948 list_generators=true
14952 Specify the size of the sourced video. For the syntax of this option, check the
14953 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14954 The default value is @code{320x240}.
14957 Specify the frame rate of the sourced video, as the number of frames
14958 generated per second. It has to be a string in the format
14959 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14960 number or a valid video frame rate abbreviation. The default value is
14964 Set the sample aspect ratio of the sourced video.
14967 Set the duration of the sourced video. See
14968 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14969 for the accepted syntax.
14971 If not specified, or the expressed duration is negative, the video is
14972 supposed to be generated forever.
14975 Additionally, all options of the @ref{coreimage} video filter are accepted.
14976 A complete filterchain can be used for further processing of the
14977 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
14978 and examples for details.
14980 @subsection Examples
14985 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
14986 given as complete and escaped command-line for Apple's standard bash shell:
14988 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
14990 This example is equivalent to the QRCode example of @ref{coreimage} without the
14991 need for a nullsrc video source.
14995 @section mandelbrot
14997 Generate a Mandelbrot set fractal, and progressively zoom towards the
14998 point specified with @var{start_x} and @var{start_y}.
15000 This source accepts the following options:
15005 Set the terminal pts value. Default value is 400.
15008 Set the terminal scale value.
15009 Must be a floating point value. Default value is 0.3.
15012 Set the inner coloring mode, that is the algorithm used to draw the
15013 Mandelbrot fractal internal region.
15015 It shall assume one of the following values:
15020 Show time until convergence.
15022 Set color based on point closest to the origin of the iterations.
15027 Default value is @var{mincol}.
15030 Set the bailout value. Default value is 10.0.
15033 Set the maximum of iterations performed by the rendering
15034 algorithm. Default value is 7189.
15037 Set outer coloring mode.
15038 It shall assume one of following values:
15040 @item iteration_count
15041 Set iteration cound mode.
15042 @item normalized_iteration_count
15043 set normalized iteration count mode.
15045 Default value is @var{normalized_iteration_count}.
15048 Set frame rate, expressed as number of frames per second. Default
15052 Set frame size. For the syntax of this option, check the "Video
15053 size" section in the ffmpeg-utils manual. Default value is "640x480".
15056 Set the initial scale value. Default value is 3.0.
15059 Set the initial x position. Must be a floating point value between
15060 -100 and 100. Default value is -0.743643887037158704752191506114774.
15063 Set the initial y position. Must be a floating point value between
15064 -100 and 100. Default value is -0.131825904205311970493132056385139.
15069 Generate various test patterns, as generated by the MPlayer test filter.
15071 The size of the generated video is fixed, and is 256x256.
15072 This source is useful in particular for testing encoding features.
15074 This source accepts the following options:
15079 Specify the frame rate of the sourced video, as the number of frames
15080 generated per second. It has to be a string in the format
15081 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15082 number or a valid video frame rate abbreviation. The default value is
15086 Set the duration of the sourced video. See
15087 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15088 for the accepted syntax.
15090 If not specified, or the expressed duration is negative, the video is
15091 supposed to be generated forever.
15095 Set the number or the name of the test to perform. Supported tests are:
15111 Default value is "all", which will cycle through the list of all tests.
15116 mptestsrc=t=dc_luma
15119 will generate a "dc_luma" test pattern.
15121 @section frei0r_src
15123 Provide a frei0r source.
15125 To enable compilation of this filter you need to install the frei0r
15126 header and configure FFmpeg with @code{--enable-frei0r}.
15128 This source accepts the following parameters:
15133 The size of the video to generate. For the syntax of this option, check the
15134 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15137 The framerate of the generated video. It may be a string of the form
15138 @var{num}/@var{den} or a frame rate abbreviation.
15141 The name to the frei0r source to load. For more information regarding frei0r and
15142 how to set the parameters, read the @ref{frei0r} section in the video filters
15145 @item filter_params
15146 A '|'-separated list of parameters to pass to the frei0r source.
15150 For example, to generate a frei0r partik0l source with size 200x200
15151 and frame rate 10 which is overlaid on the overlay filter main input:
15153 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
15158 Generate a life pattern.
15160 This source is based on a generalization of John Conway's life game.
15162 The sourced input represents a life grid, each pixel represents a cell
15163 which can be in one of two possible states, alive or dead. Every cell
15164 interacts with its eight neighbours, which are the cells that are
15165 horizontally, vertically, or diagonally adjacent.
15167 At each interaction the grid evolves according to the adopted rule,
15168 which specifies the number of neighbor alive cells which will make a
15169 cell stay alive or born. The @option{rule} option allows one to specify
15172 This source accepts the following options:
15176 Set the file from which to read the initial grid state. In the file,
15177 each non-whitespace character is considered an alive cell, and newline
15178 is used to delimit the end of each row.
15180 If this option is not specified, the initial grid is generated
15184 Set the video rate, that is the number of frames generated per second.
15187 @item random_fill_ratio, ratio
15188 Set the random fill ratio for the initial random grid. It is a
15189 floating point number value ranging from 0 to 1, defaults to 1/PHI.
15190 It is ignored when a file is specified.
15192 @item random_seed, seed
15193 Set the seed for filling the initial random grid, must be an integer
15194 included between 0 and UINT32_MAX. If not specified, or if explicitly
15195 set to -1, the filter will try to use a good random seed on a best
15201 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
15202 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
15203 @var{NS} specifies the number of alive neighbor cells which make a
15204 live cell stay alive, and @var{NB} the number of alive neighbor cells
15205 which make a dead cell to become alive (i.e. to "born").
15206 "s" and "b" can be used in place of "S" and "B", respectively.
15208 Alternatively a rule can be specified by an 18-bits integer. The 9
15209 high order bits are used to encode the next cell state if it is alive
15210 for each number of neighbor alive cells, the low order bits specify
15211 the rule for "borning" new cells. Higher order bits encode for an
15212 higher number of neighbor cells.
15213 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
15214 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
15216 Default value is "S23/B3", which is the original Conway's game of life
15217 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
15218 cells, and will born a new cell if there are three alive cells around
15222 Set the size of the output video. For the syntax of this option, check the
15223 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15225 If @option{filename} is specified, the size is set by default to the
15226 same size of the input file. If @option{size} is set, it must contain
15227 the size specified in the input file, and the initial grid defined in
15228 that file is centered in the larger resulting area.
15230 If a filename is not specified, the size value defaults to "320x240"
15231 (used for a randomly generated initial grid).
15234 If set to 1, stitch the left and right grid edges together, and the
15235 top and bottom edges also. Defaults to 1.
15238 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
15239 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
15240 value from 0 to 255.
15243 Set the color of living (or new born) cells.
15246 Set the color of dead cells. If @option{mold} is set, this is the first color
15247 used to represent a dead cell.
15250 Set mold color, for definitely dead and moldy cells.
15252 For the syntax of these 3 color options, check the "Color" section in the
15253 ffmpeg-utils manual.
15256 @subsection Examples
15260 Read a grid from @file{pattern}, and center it on a grid of size
15263 life=f=pattern:s=300x300
15267 Generate a random grid of size 200x200, with a fill ratio of 2/3:
15269 life=ratio=2/3:s=200x200
15273 Specify a custom rule for evolving a randomly generated grid:
15279 Full example with slow death effect (mold) using @command{ffplay}:
15281 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
15288 @anchor{haldclutsrc}
15290 @anchor{rgbtestsrc}
15292 @anchor{smptehdbars}
15295 @anchor{yuvtestsrc}
15296 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
15298 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
15300 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
15302 The @code{color} source provides an uniformly colored input.
15304 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
15305 @ref{haldclut} filter.
15307 The @code{nullsrc} source returns unprocessed video frames. It is
15308 mainly useful to be employed in analysis / debugging tools, or as the
15309 source for filters which ignore the input data.
15311 The @code{rgbtestsrc} source generates an RGB test pattern useful for
15312 detecting RGB vs BGR issues. You should see a red, green and blue
15313 stripe from top to bottom.
15315 The @code{smptebars} source generates a color bars pattern, based on
15316 the SMPTE Engineering Guideline EG 1-1990.
15318 The @code{smptehdbars} source generates a color bars pattern, based on
15319 the SMPTE RP 219-2002.
15321 The @code{testsrc} source generates a test video pattern, showing a
15322 color pattern, a scrolling gradient and a timestamp. This is mainly
15323 intended for testing purposes.
15325 The @code{testsrc2} source is similar to testsrc, but supports more
15326 pixel formats instead of just @code{rgb24}. This allows using it as an
15327 input for other tests without requiring a format conversion.
15329 The @code{yuvtestsrc} source generates an YUV test pattern. You should
15330 see a y, cb and cr stripe from top to bottom.
15332 The sources accept the following parameters:
15337 Specify the color of the source, only available in the @code{color}
15338 source. For the syntax of this option, check the "Color" section in the
15339 ffmpeg-utils manual.
15342 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
15343 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
15344 pixels to be used as identity matrix for 3D lookup tables. Each component is
15345 coded on a @code{1/(N*N)} scale.
15348 Specify the size of the sourced video. For the syntax of this option, check the
15349 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15350 The default value is @code{320x240}.
15352 This option is not available with the @code{haldclutsrc} filter.
15355 Specify the frame rate of the sourced video, as the number of frames
15356 generated per second. It has to be a string in the format
15357 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15358 number or a valid video frame rate abbreviation. The default value is
15362 Set the sample aspect ratio of the sourced video.
15365 Set the duration of the sourced video. See
15366 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15367 for the accepted syntax.
15369 If not specified, or the expressed duration is negative, the video is
15370 supposed to be generated forever.
15373 Set the number of decimals to show in the timestamp, only available in the
15374 @code{testsrc} source.
15376 The displayed timestamp value will correspond to the original
15377 timestamp value multiplied by the power of 10 of the specified
15378 value. Default value is 0.
15381 For example the following:
15383 testsrc=duration=5.3:size=qcif:rate=10
15386 will generate a video with a duration of 5.3 seconds, with size
15387 176x144 and a frame rate of 10 frames per second.
15389 The following graph description will generate a red source
15390 with an opacity of 0.2, with size "qcif" and a frame rate of 10
15393 color=c=red@@0.2:s=qcif:r=10
15396 If the input content is to be ignored, @code{nullsrc} can be used. The
15397 following command generates noise in the luminance plane by employing
15398 the @code{geq} filter:
15400 nullsrc=s=256x256, geq=random(1)*255:128:128
15403 @subsection Commands
15405 The @code{color} source supports the following commands:
15409 Set the color of the created image. Accepts the same syntax of the
15410 corresponding @option{color} option.
15413 @c man end VIDEO SOURCES
15415 @chapter Video Sinks
15416 @c man begin VIDEO SINKS
15418 Below is a description of the currently available video sinks.
15420 @section buffersink
15422 Buffer video frames, and make them available to the end of the filter
15425 This sink is mainly intended for programmatic use, in particular
15426 through the interface defined in @file{libavfilter/buffersink.h}
15427 or the options system.
15429 It accepts a pointer to an AVBufferSinkContext structure, which
15430 defines the incoming buffers' formats, to be passed as the opaque
15431 parameter to @code{avfilter_init_filter} for initialization.
15435 Null video sink: do absolutely nothing with the input video. It is
15436 mainly useful as a template and for use in analysis / debugging
15439 @c man end VIDEO SINKS
15441 @chapter Multimedia Filters
15442 @c man begin MULTIMEDIA FILTERS
15444 Below is a description of the currently available multimedia filters.
15446 @section ahistogram
15448 Convert input audio to a video output, displaying the volume histogram.
15450 The filter accepts the following options:
15454 Specify how histogram is calculated.
15456 It accepts the following values:
15459 Use single histogram for all channels.
15461 Use separate histogram for each channel.
15463 Default is @code{single}.
15466 Set frame rate, expressed as number of frames per second. Default
15470 Specify the video size for the output. For the syntax of this option, check the
15471 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15472 Default value is @code{hd720}.
15477 It accepts the following values:
15488 reverse logarithmic
15490 Default is @code{log}.
15493 Set amplitude scale.
15495 It accepts the following values:
15502 Default is @code{log}.
15505 Set how much frames to accumulate in histogram.
15506 Defauls is 1. Setting this to -1 accumulates all frames.
15509 Set histogram ratio of window height.
15512 Set sonogram sliding.
15514 It accepts the following values:
15517 replace old rows with new ones.
15519 scroll from top to bottom.
15521 Default is @code{replace}.
15524 @section aphasemeter
15526 Convert input audio to a video output, displaying the audio phase.
15528 The filter accepts the following options:
15532 Set the output frame rate. Default value is @code{25}.
15535 Set the video size for the output. For the syntax of this option, check the
15536 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15537 Default value is @code{800x400}.
15542 Specify the red, green, blue contrast. Default values are @code{2},
15543 @code{7} and @code{1}.
15544 Allowed range is @code{[0, 255]}.
15547 Set color which will be used for drawing median phase. If color is
15548 @code{none} which is default, no median phase value will be drawn.
15551 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15552 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15553 The @code{-1} means left and right channels are completely out of phase and
15554 @code{1} means channels are in phase.
15556 @section avectorscope
15558 Convert input audio to a video output, representing the audio vector
15561 The filter is used to measure the difference between channels of stereo
15562 audio stream. A monoaural signal, consisting of identical left and right
15563 signal, results in straight vertical line. Any stereo separation is visible
15564 as a deviation from this line, creating a Lissajous figure.
15565 If the straight (or deviation from it) but horizontal line appears this
15566 indicates that the left and right channels are out of phase.
15568 The filter accepts the following options:
15572 Set the vectorscope mode.
15574 Available values are:
15577 Lissajous rotated by 45 degrees.
15580 Same as above but not rotated.
15583 Shape resembling half of circle.
15586 Default value is @samp{lissajous}.
15589 Set the video size for the output. For the syntax of this option, check the
15590 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15591 Default value is @code{400x400}.
15594 Set the output frame rate. Default value is @code{25}.
15600 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15601 @code{160}, @code{80} and @code{255}.
15602 Allowed range is @code{[0, 255]}.
15608 Specify the red, green, blue and alpha fade. Default values are @code{15},
15609 @code{10}, @code{5} and @code{5}.
15610 Allowed range is @code{[0, 255]}.
15613 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15616 Set the vectorscope drawing mode.
15618 Available values are:
15621 Draw dot for each sample.
15624 Draw line between previous and current sample.
15627 Default value is @samp{dot}.
15630 Specify amplitude scale of audio samples.
15632 Available values are:
15649 @subsection Examples
15653 Complete example using @command{ffplay}:
15655 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15656 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15660 @section bench, abench
15662 Benchmark part of a filtergraph.
15664 The filter accepts the following options:
15668 Start or stop a timer.
15670 Available values are:
15673 Get the current time, set it as frame metadata (using the key
15674 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15677 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15678 the input frame metadata to get the time difference. Time difference, average,
15679 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15680 @code{min}) are then printed. The timestamps are expressed in seconds.
15684 @subsection Examples
15688 Benchmark @ref{selectivecolor} filter:
15690 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15696 Concatenate audio and video streams, joining them together one after the
15699 The filter works on segments of synchronized video and audio streams. All
15700 segments must have the same number of streams of each type, and that will
15701 also be the number of streams at output.
15703 The filter accepts the following options:
15708 Set the number of segments. Default is 2.
15711 Set the number of output video streams, that is also the number of video
15712 streams in each segment. Default is 1.
15715 Set the number of output audio streams, that is also the number of audio
15716 streams in each segment. Default is 0.
15719 Activate unsafe mode: do not fail if segments have a different format.
15723 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15724 @var{a} audio outputs.
15726 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15727 segment, in the same order as the outputs, then the inputs for the second
15730 Related streams do not always have exactly the same duration, for various
15731 reasons including codec frame size or sloppy authoring. For that reason,
15732 related synchronized streams (e.g. a video and its audio track) should be
15733 concatenated at once. The concat filter will use the duration of the longest
15734 stream in each segment (except the last one), and if necessary pad shorter
15735 audio streams with silence.
15737 For this filter to work correctly, all segments must start at timestamp 0.
15739 All corresponding streams must have the same parameters in all segments; the
15740 filtering system will automatically select a common pixel format for video
15741 streams, and a common sample format, sample rate and channel layout for
15742 audio streams, but other settings, such as resolution, must be converted
15743 explicitly by the user.
15745 Different frame rates are acceptable but will result in variable frame rate
15746 at output; be sure to configure the output file to handle it.
15748 @subsection Examples
15752 Concatenate an opening, an episode and an ending, all in bilingual version
15753 (video in stream 0, audio in streams 1 and 2):
15755 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15756 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15757 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15758 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15762 Concatenate two parts, handling audio and video separately, using the
15763 (a)movie sources, and adjusting the resolution:
15765 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15766 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15767 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15769 Note that a desync will happen at the stitch if the audio and video streams
15770 do not have exactly the same duration in the first file.
15774 @section drawgraph, adrawgraph
15776 Draw a graph using input video or audio metadata.
15778 It accepts the following parameters:
15782 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15785 Set 1st foreground color expression.
15788 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15791 Set 2nd foreground color expression.
15794 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15797 Set 3rd foreground color expression.
15800 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15803 Set 4th foreground color expression.
15806 Set minimal value of metadata value.
15809 Set maximal value of metadata value.
15812 Set graph background color. Default is white.
15817 Available values for mode is:
15824 Default is @code{line}.
15829 Available values for slide is:
15832 Draw new frame when right border is reached.
15835 Replace old columns with new ones.
15838 Scroll from right to left.
15841 Scroll from left to right.
15844 Draw single picture.
15847 Default is @code{frame}.
15850 Set size of graph video. For the syntax of this option, check the
15851 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15852 The default value is @code{900x256}.
15854 The foreground color expressions can use the following variables:
15857 Minimal value of metadata value.
15860 Maximal value of metadata value.
15863 Current metadata key value.
15866 The color is defined as 0xAABBGGRR.
15869 Example using metadata from @ref{signalstats} filter:
15871 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15874 Example using metadata from @ref{ebur128} filter:
15876 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15882 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15883 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15884 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15885 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15887 The filter also has a video output (see the @var{video} option) with a real
15888 time graph to observe the loudness evolution. The graphic contains the logged
15889 message mentioned above, so it is not printed anymore when this option is set,
15890 unless the verbose logging is set. The main graphing area contains the
15891 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15892 the momentary loudness (400 milliseconds).
15894 More information about the Loudness Recommendation EBU R128 on
15895 @url{http://tech.ebu.ch/loudness}.
15897 The filter accepts the following options:
15902 Activate the video output. The audio stream is passed unchanged whether this
15903 option is set or no. The video stream will be the first output stream if
15904 activated. Default is @code{0}.
15907 Set the video size. This option is for video only. For the syntax of this
15909 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15910 Default and minimum resolution is @code{640x480}.
15913 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15914 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15915 other integer value between this range is allowed.
15918 Set metadata injection. If set to @code{1}, the audio input will be segmented
15919 into 100ms output frames, each of them containing various loudness information
15920 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15922 Default is @code{0}.
15925 Force the frame logging level.
15927 Available values are:
15930 information logging level
15932 verbose logging level
15935 By default, the logging level is set to @var{info}. If the @option{video} or
15936 the @option{metadata} options are set, it switches to @var{verbose}.
15941 Available modes can be cumulated (the option is a @code{flag} type). Possible
15945 Disable any peak mode (default).
15947 Enable sample-peak mode.
15949 Simple peak mode looking for the higher sample value. It logs a message
15950 for sample-peak (identified by @code{SPK}).
15952 Enable true-peak mode.
15954 If enabled, the peak lookup is done on an over-sampled version of the input
15955 stream for better peak accuracy. It logs a message for true-peak.
15956 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15957 This mode requires a build with @code{libswresample}.
15961 Treat mono input files as "dual mono". If a mono file is intended for playback
15962 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15963 If set to @code{true}, this option will compensate for this effect.
15964 Multi-channel input files are not affected by this option.
15967 Set a specific pan law to be used for the measurement of dual mono files.
15968 This parameter is optional, and has a default value of -3.01dB.
15971 @subsection Examples
15975 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
15977 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
15981 Run an analysis with @command{ffmpeg}:
15983 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
15987 @section interleave, ainterleave
15989 Temporally interleave frames from several inputs.
15991 @code{interleave} works with video inputs, @code{ainterleave} with audio.
15993 These filters read frames from several inputs and send the oldest
15994 queued frame to the output.
15996 Input streams must have well defined, monotonically increasing frame
15999 In order to submit one frame to output, these filters need to enqueue
16000 at least one frame for each input, so they cannot work in case one
16001 input is not yet terminated and will not receive incoming frames.
16003 For example consider the case when one input is a @code{select} filter
16004 which always drops input frames. The @code{interleave} filter will keep
16005 reading from that input, but it will never be able to send new frames
16006 to output until the input sends an end-of-stream signal.
16008 Also, depending on inputs synchronization, the filters will drop
16009 frames in case one input receives more frames than the other ones, and
16010 the queue is already filled.
16012 These filters accept the following options:
16016 Set the number of different inputs, it is 2 by default.
16019 @subsection Examples
16023 Interleave frames belonging to different streams using @command{ffmpeg}:
16025 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
16029 Add flickering blur effect:
16031 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
16035 @section metadata, ametadata
16037 Manipulate frame metadata.
16039 This filter accepts the following options:
16043 Set mode of operation of the filter.
16045 Can be one of the following:
16049 If both @code{value} and @code{key} is set, select frames
16050 which have such metadata. If only @code{key} is set, select
16051 every frame that has such key in metadata.
16054 Add new metadata @code{key} and @code{value}. If key is already available
16058 Modify value of already present key.
16061 If @code{value} is set, delete only keys that have such value.
16062 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
16066 Print key and its value if metadata was found. If @code{key} is not set print all
16067 metadata values available in frame.
16071 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
16074 Set metadata value which will be used. This option is mandatory for
16075 @code{modify} and @code{add} mode.
16078 Which function to use when comparing metadata value and @code{value}.
16080 Can be one of following:
16084 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
16087 Values are interpreted as strings, returns true if metadata value starts with
16088 the @code{value} option string.
16091 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
16094 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
16097 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
16100 Values are interpreted as floats, returns true if expression from option @code{expr}
16105 Set expression which is used when @code{function} is set to @code{expr}.
16106 The expression is evaluated through the eval API and can contain the following
16111 Float representation of @code{value} from metadata key.
16114 Float representation of @code{value} as supplied by user in @code{value} option.
16117 If specified in @code{print} mode, output is written to the named file. Instead of
16118 plain filename any writable url can be specified. Filename ``-'' is a shorthand
16119 for standard output. If @code{file} option is not set, output is written to the log
16120 with AV_LOG_INFO loglevel.
16125 @subsection Examples
16129 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
16132 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
16135 Print silencedetect output to file @file{metadata.txt}.
16137 silencedetect,ametadata=mode=print:file=metadata.txt
16140 Direct all metadata to a pipe with file descriptor 4.
16142 metadata=mode=print:file='pipe\:4'
16146 @section perms, aperms
16148 Set read/write permissions for the output frames.
16150 These filters are mainly aimed at developers to test direct path in the
16151 following filter in the filtergraph.
16153 The filters accept the following options:
16157 Select the permissions mode.
16159 It accepts the following values:
16162 Do nothing. This is the default.
16164 Set all the output frames read-only.
16166 Set all the output frames directly writable.
16168 Make the frame read-only if writable, and writable if read-only.
16170 Set each output frame read-only or writable randomly.
16174 Set the seed for the @var{random} mode, must be an integer included between
16175 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16176 @code{-1}, the filter will try to use a good random seed on a best effort
16180 Note: in case of auto-inserted filter between the permission filter and the
16181 following one, the permission might not be received as expected in that
16182 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
16183 perms/aperms filter can avoid this problem.
16185 @section realtime, arealtime
16187 Slow down filtering to match real time approximatively.
16189 These filters will pause the filtering for a variable amount of time to
16190 match the output rate with the input timestamps.
16191 They are similar to the @option{re} option to @code{ffmpeg}.
16193 They accept the following options:
16197 Time limit for the pauses. Any pause longer than that will be considered
16198 a timestamp discontinuity and reset the timer. Default is 2 seconds.
16202 @section select, aselect
16204 Select frames to pass in output.
16206 This filter accepts the following options:
16211 Set expression, which is evaluated for each input frame.
16213 If the expression is evaluated to zero, the frame is discarded.
16215 If the evaluation result is negative or NaN, the frame is sent to the
16216 first output; otherwise it is sent to the output with index
16217 @code{ceil(val)-1}, assuming that the input index starts from 0.
16219 For example a value of @code{1.2} corresponds to the output with index
16220 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
16223 Set the number of outputs. The output to which to send the selected
16224 frame is based on the result of the evaluation. Default value is 1.
16227 The expression can contain the following constants:
16231 The (sequential) number of the filtered frame, starting from 0.
16234 The (sequential) number of the selected frame, starting from 0.
16236 @item prev_selected_n
16237 The sequential number of the last selected frame. It's NAN if undefined.
16240 The timebase of the input timestamps.
16243 The PTS (Presentation TimeStamp) of the filtered video frame,
16244 expressed in @var{TB} units. It's NAN if undefined.
16247 The PTS of the filtered video frame,
16248 expressed in seconds. It's NAN if undefined.
16251 The PTS of the previously filtered video frame. It's NAN if undefined.
16253 @item prev_selected_pts
16254 The PTS of the last previously filtered video frame. It's NAN if undefined.
16256 @item prev_selected_t
16257 The PTS of the last previously selected video frame. It's NAN if undefined.
16260 The PTS of the first video frame in the video. It's NAN if undefined.
16263 The time of the first video frame in the video. It's NAN if undefined.
16265 @item pict_type @emph{(video only)}
16266 The type of the filtered frame. It can assume one of the following
16278 @item interlace_type @emph{(video only)}
16279 The frame interlace type. It can assume one of the following values:
16282 The frame is progressive (not interlaced).
16284 The frame is top-field-first.
16286 The frame is bottom-field-first.
16289 @item consumed_sample_n @emph{(audio only)}
16290 the number of selected samples before the current frame
16292 @item samples_n @emph{(audio only)}
16293 the number of samples in the current frame
16295 @item sample_rate @emph{(audio only)}
16296 the input sample rate
16299 This is 1 if the filtered frame is a key-frame, 0 otherwise.
16302 the position in the file of the filtered frame, -1 if the information
16303 is not available (e.g. for synthetic video)
16305 @item scene @emph{(video only)}
16306 value between 0 and 1 to indicate a new scene; a low value reflects a low
16307 probability for the current frame to introduce a new scene, while a higher
16308 value means the current frame is more likely to be one (see the example below)
16310 @item concatdec_select
16311 The concat demuxer can select only part of a concat input file by setting an
16312 inpoint and an outpoint, but the output packets may not be entirely contained
16313 in the selected interval. By using this variable, it is possible to skip frames
16314 generated by the concat demuxer which are not exactly contained in the selected
16317 This works by comparing the frame pts against the @var{lavf.concat.start_time}
16318 and the @var{lavf.concat.duration} packet metadata values which are also
16319 present in the decoded frames.
16321 The @var{concatdec_select} variable is -1 if the frame pts is at least
16322 start_time and either the duration metadata is missing or the frame pts is less
16323 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
16326 That basically means that an input frame is selected if its pts is within the
16327 interval set by the concat demuxer.
16331 The default value of the select expression is "1".
16333 @subsection Examples
16337 Select all frames in input:
16342 The example above is the same as:
16354 Select only I-frames:
16356 select='eq(pict_type\,I)'
16360 Select one frame every 100:
16362 select='not(mod(n\,100))'
16366 Select only frames contained in the 10-20 time interval:
16368 select=between(t\,10\,20)
16372 Select only I-frames contained in the 10-20 time interval:
16374 select=between(t\,10\,20)*eq(pict_type\,I)
16378 Select frames with a minimum distance of 10 seconds:
16380 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
16384 Use aselect to select only audio frames with samples number > 100:
16386 aselect='gt(samples_n\,100)'
16390 Create a mosaic of the first scenes:
16392 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
16395 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
16399 Send even and odd frames to separate outputs, and compose them:
16401 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
16405 Select useful frames from an ffconcat file which is using inpoints and
16406 outpoints but where the source files are not intra frame only.
16408 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
16412 @section sendcmd, asendcmd
16414 Send commands to filters in the filtergraph.
16416 These filters read commands to be sent to other filters in the
16419 @code{sendcmd} must be inserted between two video filters,
16420 @code{asendcmd} must be inserted between two audio filters, but apart
16421 from that they act the same way.
16423 The specification of commands can be provided in the filter arguments
16424 with the @var{commands} option, or in a file specified by the
16425 @var{filename} option.
16427 These filters accept the following options:
16430 Set the commands to be read and sent to the other filters.
16432 Set the filename of the commands to be read and sent to the other
16436 @subsection Commands syntax
16438 A commands description consists of a sequence of interval
16439 specifications, comprising a list of commands to be executed when a
16440 particular event related to that interval occurs. The occurring event
16441 is typically the current frame time entering or leaving a given time
16444 An interval is specified by the following syntax:
16446 @var{START}[-@var{END}] @var{COMMANDS};
16449 The time interval is specified by the @var{START} and @var{END} times.
16450 @var{END} is optional and defaults to the maximum time.
16452 The current frame time is considered within the specified interval if
16453 it is included in the interval [@var{START}, @var{END}), that is when
16454 the time is greater or equal to @var{START} and is lesser than
16457 @var{COMMANDS} consists of a sequence of one or more command
16458 specifications, separated by ",", relating to that interval. The
16459 syntax of a command specification is given by:
16461 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
16464 @var{FLAGS} is optional and specifies the type of events relating to
16465 the time interval which enable sending the specified command, and must
16466 be a non-null sequence of identifier flags separated by "+" or "|" and
16467 enclosed between "[" and "]".
16469 The following flags are recognized:
16472 The command is sent when the current frame timestamp enters the
16473 specified interval. In other words, the command is sent when the
16474 previous frame timestamp was not in the given interval, and the
16478 The command is sent when the current frame timestamp leaves the
16479 specified interval. In other words, the command is sent when the
16480 previous frame timestamp was in the given interval, and the
16484 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
16487 @var{TARGET} specifies the target of the command, usually the name of
16488 the filter class or a specific filter instance name.
16490 @var{COMMAND} specifies the name of the command for the target filter.
16492 @var{ARG} is optional and specifies the optional list of argument for
16493 the given @var{COMMAND}.
16495 Between one interval specification and another, whitespaces, or
16496 sequences of characters starting with @code{#} until the end of line,
16497 are ignored and can be used to annotate comments.
16499 A simplified BNF description of the commands specification syntax
16502 @var{COMMAND_FLAG} ::= "enter" | "leave"
16503 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
16504 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
16505 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
16506 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
16507 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
16510 @subsection Examples
16514 Specify audio tempo change at second 4:
16516 asendcmd=c='4.0 atempo tempo 1.5',atempo
16520 Specify a list of drawtext and hue commands in a file.
16522 # show text in the interval 5-10
16523 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
16524 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16526 # desaturate the image in the interval 15-20
16527 15.0-20.0 [enter] hue s 0,
16528 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16530 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16532 # apply an exponential saturation fade-out effect, starting from time 25
16533 25 [enter] hue s exp(25-t)
16536 A filtergraph allowing to read and process the above command list
16537 stored in a file @file{test.cmd}, can be specified with:
16539 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16544 @section setpts, asetpts
16546 Change the PTS (presentation timestamp) of the input frames.
16548 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16550 This filter accepts the following options:
16555 The expression which is evaluated for each frame to construct its timestamp.
16559 The expression is evaluated through the eval API and can contain the following
16564 frame rate, only defined for constant frame-rate video
16567 The presentation timestamp in input
16570 The count of the input frame for video or the number of consumed samples,
16571 not including the current frame for audio, starting from 0.
16573 @item NB_CONSUMED_SAMPLES
16574 The number of consumed samples, not including the current frame (only
16577 @item NB_SAMPLES, S
16578 The number of samples in the current frame (only audio)
16580 @item SAMPLE_RATE, SR
16581 The audio sample rate.
16584 The PTS of the first frame.
16587 the time in seconds of the first frame
16590 State whether the current frame is interlaced.
16593 the time in seconds of the current frame
16596 original position in the file of the frame, or undefined if undefined
16597 for the current frame
16600 The previous input PTS.
16603 previous input time in seconds
16606 The previous output PTS.
16609 previous output time in seconds
16612 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16616 The wallclock (RTC) time at the start of the movie in microseconds.
16619 The timebase of the input timestamps.
16623 @subsection Examples
16627 Start counting PTS from zero
16629 setpts=PTS-STARTPTS
16633 Apply fast motion effect:
16639 Apply slow motion effect:
16645 Set fixed rate of 25 frames per second:
16651 Set fixed rate 25 fps with some jitter:
16653 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16657 Apply an offset of 10 seconds to the input PTS:
16663 Generate timestamps from a "live source" and rebase onto the current timebase:
16665 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16669 Generate timestamps by counting samples:
16676 @section settb, asettb
16678 Set the timebase to use for the output frames timestamps.
16679 It is mainly useful for testing timebase configuration.
16681 It accepts the following parameters:
16686 The expression which is evaluated into the output timebase.
16690 The value for @option{tb} is an arithmetic expression representing a
16691 rational. The expression can contain the constants "AVTB" (the default
16692 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16693 audio only). Default value is "intb".
16695 @subsection Examples
16699 Set the timebase to 1/25:
16705 Set the timebase to 1/10:
16711 Set the timebase to 1001/1000:
16717 Set the timebase to 2*intb:
16723 Set the default timebase value:
16730 Convert input audio to a video output representing frequency spectrum
16731 logarithmically using Brown-Puckette constant Q transform algorithm with
16732 direct frequency domain coefficient calculation (but the transform itself
16733 is not really constant Q, instead the Q factor is actually variable/clamped),
16734 with musical tone scale, from E0 to D#10.
16736 The filter accepts the following options:
16740 Specify the video size for the output. It must be even. For the syntax of this option,
16741 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16742 Default value is @code{1920x1080}.
16745 Set the output frame rate. Default value is @code{25}.
16748 Set the bargraph height. It must be even. Default value is @code{-1} which
16749 computes the bargraph height automatically.
16752 Set the axis height. It must be even. Default value is @code{-1} which computes
16753 the axis height automatically.
16756 Set the sonogram height. It must be even. Default value is @code{-1} which
16757 computes the sonogram height automatically.
16760 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16761 instead. Default value is @code{1}.
16763 @item sono_v, volume
16764 Specify the sonogram volume expression. It can contain variables:
16767 the @var{bar_v} evaluated expression
16768 @item frequency, freq, f
16769 the frequency where it is evaluated
16770 @item timeclamp, tc
16771 the value of @var{timeclamp} option
16775 @item a_weighting(f)
16776 A-weighting of equal loudness
16777 @item b_weighting(f)
16778 B-weighting of equal loudness
16779 @item c_weighting(f)
16780 C-weighting of equal loudness.
16782 Default value is @code{16}.
16784 @item bar_v, volume2
16785 Specify the bargraph volume expression. It can contain variables:
16788 the @var{sono_v} evaluated expression
16789 @item frequency, freq, f
16790 the frequency where it is evaluated
16791 @item timeclamp, tc
16792 the value of @var{timeclamp} option
16796 @item a_weighting(f)
16797 A-weighting of equal loudness
16798 @item b_weighting(f)
16799 B-weighting of equal loudness
16800 @item c_weighting(f)
16801 C-weighting of equal loudness.
16803 Default value is @code{sono_v}.
16805 @item sono_g, gamma
16806 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16807 higher gamma makes the spectrum having more range. Default value is @code{3}.
16808 Acceptable range is @code{[1, 7]}.
16810 @item bar_g, gamma2
16811 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16814 @item timeclamp, tc
16815 Specify the transform timeclamp. At low frequency, there is trade-off between
16816 accuracy in time domain and frequency domain. If timeclamp is lower,
16817 event in time domain is represented more accurately (such as fast bass drum),
16818 otherwise event in frequency domain is represented more accurately
16819 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16822 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16823 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16826 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16827 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16830 This option is deprecated and ignored.
16833 Specify the transform length in time domain. Use this option to control accuracy
16834 trade-off between time domain and frequency domain at every frequency sample.
16835 It can contain variables:
16837 @item frequency, freq, f
16838 the frequency where it is evaluated
16839 @item timeclamp, tc
16840 the value of @var{timeclamp} option.
16842 Default value is @code{384*tc/(384+tc*f)}.
16845 Specify the transform count for every video frame. Default value is @code{6}.
16846 Acceptable range is @code{[1, 30]}.
16849 Specify the transform count for every single pixel. Default value is @code{0},
16850 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16853 Specify font file for use with freetype to draw the axis. If not specified,
16854 use embedded font. Note that drawing with font file or embedded font is not
16855 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16859 Specify font color expression. This is arithmetic expression that should return
16860 integer value 0xRRGGBB. It can contain variables:
16862 @item frequency, freq, f
16863 the frequency where it is evaluated
16864 @item timeclamp, tc
16865 the value of @var{timeclamp} option
16870 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16871 @item r(x), g(x), b(x)
16872 red, green, and blue value of intensity x.
16874 Default value is @code{st(0, (midi(f)-59.5)/12);
16875 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16876 r(1-ld(1)) + b(ld(1))}.
16879 Specify image file to draw the axis. This option override @var{fontfile} and
16880 @var{fontcolor} option.
16883 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16884 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16885 Default value is @code{1}.
16889 @subsection Examples
16893 Playing audio while showing the spectrum:
16895 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16899 Same as above, but with frame rate 30 fps:
16901 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16905 Playing at 1280x720:
16907 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16911 Disable sonogram display:
16917 A1 and its harmonics: A1, A2, (near)E3, A3:
16919 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),
16920 asplit[a][out1]; [a] showcqt [out0]'
16924 Same as above, but with more accuracy in frequency domain:
16926 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),
16927 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16933 bar_v=10:sono_v=bar_v*a_weighting(f)
16937 Custom gamma, now spectrum is linear to the amplitude.
16943 Custom tlength equation:
16945 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)))'
16949 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
16951 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
16955 Custom frequency range with custom axis using image file:
16957 axisfile=myaxis.png:basefreq=40:endfreq=10000
16963 Convert input audio to video output representing the audio power spectrum.
16964 Audio amplitude is on Y-axis while frequency is on X-axis.
16966 The filter accepts the following options:
16970 Specify size of video. For the syntax of this option, check the
16971 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16972 Default is @code{1024x512}.
16976 This set how each frequency bin will be represented.
16978 It accepts the following values:
16984 Default is @code{bar}.
16987 Set amplitude scale.
16989 It accepts the following values:
17003 Default is @code{log}.
17006 Set frequency scale.
17008 It accepts the following values:
17017 Reverse logarithmic scale.
17019 Default is @code{lin}.
17024 It accepts the following values:
17040 Default is @code{w2048}
17043 Set windowing function.
17045 It accepts the following values:
17067 Default is @code{hanning}.
17070 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17071 which means optimal overlap for selected window function will be picked.
17074 Set time averaging. Setting this to 0 will display current maximal peaks.
17075 Default is @code{1}, which means time averaging is disabled.
17078 Specify list of colors separated by space or by '|' which will be used to
17079 draw channel frequencies. Unrecognized or missing colors will be replaced
17083 Set channel display mode.
17085 It accepts the following values:
17090 Default is @code{combined}.
17093 Set minimum amplitude used in @code{log} amplitude scaler.
17097 @anchor{showspectrum}
17098 @section showspectrum
17100 Convert input audio to a video output, representing the audio frequency
17103 The filter accepts the following options:
17107 Specify the video size for the output. For the syntax of this option, check the
17108 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17109 Default value is @code{640x512}.
17112 Specify how the spectrum should slide along the window.
17114 It accepts the following values:
17117 the samples start again on the left when they reach the right
17119 the samples scroll from right to left
17121 frames are only produced when the samples reach the right
17123 the samples scroll from left to right
17126 Default value is @code{replace}.
17129 Specify display mode.
17131 It accepts the following values:
17134 all channels are displayed in the same row
17136 all channels are displayed in separate rows
17139 Default value is @samp{combined}.
17142 Specify display color mode.
17144 It accepts the following values:
17147 each channel is displayed in a separate color
17149 each channel is displayed using the same color scheme
17151 each channel is displayed using the rainbow color scheme
17153 each channel is displayed using the moreland color scheme
17155 each channel is displayed using the nebulae color scheme
17157 each channel is displayed using the fire color scheme
17159 each channel is displayed using the fiery color scheme
17161 each channel is displayed using the fruit color scheme
17163 each channel is displayed using the cool color scheme
17166 Default value is @samp{channel}.
17169 Specify scale used for calculating intensity color values.
17171 It accepts the following values:
17176 square root, default
17187 Default value is @samp{sqrt}.
17190 Set saturation modifier for displayed colors. Negative values provide
17191 alternative color scheme. @code{0} is no saturation at all.
17192 Saturation must be in [-10.0, 10.0] range.
17193 Default value is @code{1}.
17196 Set window function.
17198 It accepts the following values:
17222 Default value is @code{hann}.
17225 Set orientation of time vs frequency axis. Can be @code{vertical} or
17226 @code{horizontal}. Default is @code{vertical}.
17229 Set ratio of overlap window. Default value is @code{0}.
17230 When value is @code{1} overlap is set to recommended size for specific
17231 window function currently used.
17234 Set scale gain for calculating intensity color values.
17235 Default value is @code{1}.
17238 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
17241 Set color rotation, must be in [-1.0, 1.0] range.
17242 Default value is @code{0}.
17245 The usage is very similar to the showwaves filter; see the examples in that
17248 @subsection Examples
17252 Large window with logarithmic color scaling:
17254 showspectrum=s=1280x480:scale=log
17258 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
17260 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17261 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
17265 @section showspectrumpic
17267 Convert input audio to a single video frame, representing the audio frequency
17270 The filter accepts the following options:
17274 Specify the video size for the output. For the syntax of this option, check the
17275 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17276 Default value is @code{4096x2048}.
17279 Specify display mode.
17281 It accepts the following values:
17284 all channels are displayed in the same row
17286 all channels are displayed in separate rows
17288 Default value is @samp{combined}.
17291 Specify display color mode.
17293 It accepts the following values:
17296 each channel is displayed in a separate color
17298 each channel is displayed using the same color scheme
17300 each channel is displayed using the rainbow color scheme
17302 each channel is displayed using the moreland color scheme
17304 each channel is displayed using the nebulae color scheme
17306 each channel is displayed using the fire color scheme
17308 each channel is displayed using the fiery color scheme
17310 each channel is displayed using the fruit color scheme
17312 each channel is displayed using the cool color scheme
17314 Default value is @samp{intensity}.
17317 Specify scale used for calculating intensity color values.
17319 It accepts the following values:
17324 square root, default
17334 Default value is @samp{log}.
17337 Set saturation modifier for displayed colors. Negative values provide
17338 alternative color scheme. @code{0} is no saturation at all.
17339 Saturation must be in [-10.0, 10.0] range.
17340 Default value is @code{1}.
17343 Set window function.
17345 It accepts the following values:
17368 Default value is @code{hann}.
17371 Set orientation of time vs frequency axis. Can be @code{vertical} or
17372 @code{horizontal}. Default is @code{vertical}.
17375 Set scale gain for calculating intensity color values.
17376 Default value is @code{1}.
17379 Draw time and frequency axes and legends. Default is enabled.
17382 Set color rotation, must be in [-1.0, 1.0] range.
17383 Default value is @code{0}.
17386 @subsection Examples
17390 Extract an audio spectrogram of a whole audio track
17391 in a 1024x1024 picture using @command{ffmpeg}:
17393 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
17397 @section showvolume
17399 Convert input audio volume to a video output.
17401 The filter accepts the following options:
17408 Set border width, allowed range is [0, 5]. Default is 1.
17411 Set channel width, allowed range is [80, 8192]. Default is 400.
17414 Set channel height, allowed range is [1, 900]. Default is 20.
17417 Set fade, allowed range is [0.001, 1]. Default is 0.95.
17420 Set volume color expression.
17422 The expression can use the following variables:
17426 Current max volume of channel in dB.
17432 Current channel number, starting from 0.
17436 If set, displays channel names. Default is enabled.
17439 If set, displays volume values. Default is enabled.
17442 Set orientation, can be @code{horizontal} or @code{vertical},
17443 default is @code{horizontal}.
17446 Set step size, allowed range s [0, 5]. Default is 0, which means
17452 Convert input audio to a video output, representing the samples waves.
17454 The filter accepts the following options:
17458 Specify the video size for the output. For the syntax of this option, check the
17459 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17460 Default value is @code{600x240}.
17465 Available values are:
17468 Draw a point for each sample.
17471 Draw a vertical line for each sample.
17474 Draw a point for each sample and a line between them.
17477 Draw a centered vertical line for each sample.
17480 Default value is @code{point}.
17483 Set the number of samples which are printed on the same column. A
17484 larger value will decrease the frame rate. Must be a positive
17485 integer. This option can be set only if the value for @var{rate}
17486 is not explicitly specified.
17489 Set the (approximate) output frame rate. This is done by setting the
17490 option @var{n}. Default value is "25".
17492 @item split_channels
17493 Set if channels should be drawn separately or overlap. Default value is 0.
17496 Set colors separated by '|' which are going to be used for drawing of each channel.
17499 Set amplitude scale.
17501 Available values are:
17519 @subsection Examples
17523 Output the input file audio and the corresponding video representation
17526 amovie=a.mp3,asplit[out0],showwaves[out1]
17530 Create a synthetic signal and show it with showwaves, forcing a
17531 frame rate of 30 frames per second:
17533 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17537 @section showwavespic
17539 Convert input audio to a single video frame, representing the samples waves.
17541 The filter accepts the following options:
17545 Specify the video size for the output. For the syntax of this option, check the
17546 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17547 Default value is @code{600x240}.
17549 @item split_channels
17550 Set if channels should be drawn separately or overlap. Default value is 0.
17553 Set colors separated by '|' which are going to be used for drawing of each channel.
17556 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17560 @subsection Examples
17564 Extract a channel split representation of the wave form of a whole audio track
17565 in a 1024x800 picture using @command{ffmpeg}:
17567 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17571 @section spectrumsynth
17573 Sythesize audio from 2 input video spectrums, first input stream represents
17574 magnitude across time and second represents phase across time.
17575 The filter will transform from frequency domain as displayed in videos back
17576 to time domain as presented in audio output.
17578 This filter is primarly created for reversing processed @ref{showspectrum}
17579 filter outputs, but can synthesize sound from other spectrograms too.
17580 But in such case results are going to be poor if the phase data is not
17581 available, because in such cases phase data need to be recreated, usually
17582 its just recreated from random noise.
17583 For best results use gray only output (@code{channel} color mode in
17584 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17585 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17586 @code{data} option. Inputs videos should generally use @code{fullframe}
17587 slide mode as that saves resources needed for decoding video.
17589 The filter accepts the following options:
17593 Specify sample rate of output audio, the sample rate of audio from which
17594 spectrum was generated may differ.
17597 Set number of channels represented in input video spectrums.
17600 Set scale which was used when generating magnitude input spectrum.
17601 Can be @code{lin} or @code{log}. Default is @code{log}.
17604 Set slide which was used when generating inputs spectrums.
17605 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17606 Default is @code{fullframe}.
17609 Set window function used for resynthesis.
17612 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17613 which means optimal overlap for selected window function will be picked.
17616 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17617 Default is @code{vertical}.
17620 @subsection Examples
17624 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17625 then resynthesize videos back to audio with spectrumsynth:
17627 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
17628 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
17629 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17633 @section split, asplit
17635 Split input into several identical outputs.
17637 @code{asplit} works with audio input, @code{split} with video.
17639 The filter accepts a single parameter which specifies the number of outputs. If
17640 unspecified, it defaults to 2.
17642 @subsection Examples
17646 Create two separate outputs from the same input:
17648 [in] split [out0][out1]
17652 To create 3 or more outputs, you need to specify the number of
17655 [in] asplit=3 [out0][out1][out2]
17659 Create two separate outputs from the same input, one cropped and
17662 [in] split [splitout1][splitout2];
17663 [splitout1] crop=100:100:0:0 [cropout];
17664 [splitout2] pad=200:200:100:100 [padout];
17668 Create 5 copies of the input audio with @command{ffmpeg}:
17670 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17676 Receive commands sent through a libzmq client, and forward them to
17677 filters in the filtergraph.
17679 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17680 must be inserted between two video filters, @code{azmq} between two
17683 To enable these filters you need to install the libzmq library and
17684 headers and configure FFmpeg with @code{--enable-libzmq}.
17686 For more information about libzmq see:
17687 @url{http://www.zeromq.org/}
17689 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17690 receives messages sent through a network interface defined by the
17691 @option{bind_address} option.
17693 The received message must be in the form:
17695 @var{TARGET} @var{COMMAND} [@var{ARG}]
17698 @var{TARGET} specifies the target of the command, usually the name of
17699 the filter class or a specific filter instance name.
17701 @var{COMMAND} specifies the name of the command for the target filter.
17703 @var{ARG} is optional and specifies the optional argument list for the
17704 given @var{COMMAND}.
17706 Upon reception, the message is processed and the corresponding command
17707 is injected into the filtergraph. Depending on the result, the filter
17708 will send a reply to the client, adopting the format:
17710 @var{ERROR_CODE} @var{ERROR_REASON}
17714 @var{MESSAGE} is optional.
17716 @subsection Examples
17718 Look at @file{tools/zmqsend} for an example of a zmq client which can
17719 be used to send commands processed by these filters.
17721 Consider the following filtergraph generated by @command{ffplay}
17723 ffplay -dumpgraph 1 -f lavfi "
17724 color=s=100x100:c=red [l];
17725 color=s=100x100:c=blue [r];
17726 nullsrc=s=200x100, zmq [bg];
17727 [bg][l] overlay [bg+l];
17728 [bg+l][r] overlay=x=100 "
17731 To change the color of the left side of the video, the following
17732 command can be used:
17734 echo Parsed_color_0 c yellow | tools/zmqsend
17737 To change the right side:
17739 echo Parsed_color_1 c pink | tools/zmqsend
17742 @c man end MULTIMEDIA FILTERS
17744 @chapter Multimedia Sources
17745 @c man begin MULTIMEDIA SOURCES
17747 Below is a description of the currently available multimedia sources.
17751 This is the same as @ref{movie} source, except it selects an audio
17757 Read audio and/or video stream(s) from a movie container.
17759 It accepts the following parameters:
17763 The name of the resource to read (not necessarily a file; it can also be a
17764 device or a stream accessed through some protocol).
17766 @item format_name, f
17767 Specifies the format assumed for the movie to read, and can be either
17768 the name of a container or an input device. If not specified, the
17769 format is guessed from @var{movie_name} or by probing.
17771 @item seek_point, sp
17772 Specifies the seek point in seconds. The frames will be output
17773 starting from this seek point. The parameter is evaluated with
17774 @code{av_strtod}, so the numerical value may be suffixed by an IS
17775 postfix. The default value is "0".
17778 Specifies the streams to read. Several streams can be specified,
17779 separated by "+". The source will then have as many outputs, in the
17780 same order. The syntax is explained in the ``Stream specifiers''
17781 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17782 respectively the default (best suited) video and audio stream. Default
17783 is "dv", or "da" if the filter is called as "amovie".
17785 @item stream_index, si
17786 Specifies the index of the video stream to read. If the value is -1,
17787 the most suitable video stream will be automatically selected. The default
17788 value is "-1". Deprecated. If the filter is called "amovie", it will select
17789 audio instead of video.
17792 Specifies how many times to read the stream in sequence.
17793 If the value is less than 1, the stream will be read again and again.
17794 Default value is "1".
17796 Note that when the movie is looped the source timestamps are not
17797 changed, so it will generate non monotonically increasing timestamps.
17799 @item discontinuity
17800 Specifies the time difference between frames above which the point is
17801 considered a timestamp discontinuity which is removed by adjusting the later
17805 It allows overlaying a second video on top of the main input of
17806 a filtergraph, as shown in this graph:
17808 input -----------> deltapts0 --> overlay --> output
17811 movie --> scale--> deltapts1 -------+
17813 @subsection Examples
17817 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17818 on top of the input labelled "in":
17820 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17821 [in] setpts=PTS-STARTPTS [main];
17822 [main][over] overlay=16:16 [out]
17826 Read from a video4linux2 device, and overlay it on top of the input
17829 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17830 [in] setpts=PTS-STARTPTS [main];
17831 [main][over] overlay=16:16 [out]
17835 Read the first video stream and the audio stream with id 0x81 from
17836 dvd.vob; the video is connected to the pad named "video" and the audio is
17837 connected to the pad named "audio":
17839 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17843 @subsection Commands
17845 Both movie and amovie support the following commands:
17848 Perform seek using "av_seek_frame".
17849 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17852 @var{stream_index}: If stream_index is -1, a default
17853 stream is selected, and @var{timestamp} is automatically converted
17854 from AV_TIME_BASE units to the stream specific time_base.
17856 @var{timestamp}: Timestamp in AVStream.time_base units
17857 or, if no stream is specified, in AV_TIME_BASE units.
17859 @var{flags}: Flags which select direction and seeking mode.
17863 Get movie duration in AV_TIME_BASE units.
17867 @c man end MULTIMEDIA SOURCES