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.
522 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
523 the second channel (and any other channels that may be present) unchanged.
531 Apply echoing to the input audio.
533 Echoes are reflected sound and can occur naturally amongst mountains
534 (and sometimes large buildings) when talking or shouting; digital echo
535 effects emulate this behaviour and are often used to help fill out the
536 sound of a single instrument or vocal. The time difference between the
537 original signal and the reflection is the @code{delay}, and the
538 loudness of the reflected signal is the @code{decay}.
539 Multiple echoes can have different delays and decays.
541 A description of the accepted parameters follows.
545 Set input gain of reflected signal. Default is @code{0.6}.
548 Set output gain of reflected signal. Default is @code{0.3}.
551 Set list of time intervals in milliseconds between original signal and reflections
552 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
553 Default is @code{1000}.
556 Set list of loudnesses of reflected signals separated by '|'.
557 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
558 Default is @code{0.5}.
565 Make it sound as if there are twice as many instruments as are actually playing:
567 aecho=0.8:0.88:60:0.4
571 If delay is very short, then it sound like a (metallic) robot playing music:
577 A longer delay will sound like an open air concert in the mountains:
579 aecho=0.8:0.9:1000:0.3
583 Same as above but with one more mountain:
585 aecho=0.8:0.9:1000|1800:0.3|0.25
590 Audio emphasis filter creates or restores material directly taken from LPs or
591 emphased CDs with different filter curves. E.g. to store music on vinyl the
592 signal has to be altered by a filter first to even out the disadvantages of
593 this recording medium.
594 Once the material is played back the inverse filter has to be applied to
595 restore the distortion of the frequency response.
597 The filter accepts the following options:
607 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
608 use @code{production} mode. Default is @code{reproduction} mode.
611 Set filter type. Selects medium. Can be one of the following:
623 select Compact Disc (CD).
629 select 50µs (FM-KF).
631 select 75µs (FM-KF).
637 Modify an audio signal according to the specified expressions.
639 This filter accepts one or more expressions (one for each channel),
640 which are evaluated and used to modify a corresponding audio signal.
642 It accepts the following parameters:
646 Set the '|'-separated expressions list for each separate channel. If
647 the number of input channels is greater than the number of
648 expressions, the last specified expression is used for the remaining
651 @item channel_layout, c
652 Set output channel layout. If not specified, the channel layout is
653 specified by the number of expressions. If set to @samp{same}, it will
654 use by default the same input channel layout.
657 Each expression in @var{exprs} can contain the following constants and functions:
661 channel number of the current expression
664 number of the evaluated sample, starting from 0
670 time of the evaluated sample expressed in seconds
673 @item nb_out_channels
674 input and output number of channels
677 the value of input channel with number @var{CH}
680 Note: this filter is slow. For faster processing you should use a
689 aeval=val(ch)/2:c=same
693 Invert phase of the second channel:
702 Apply fade-in/out effect to input audio.
704 A description of the accepted parameters follows.
708 Specify the effect type, can be either @code{in} for fade-in, or
709 @code{out} for a fade-out effect. Default is @code{in}.
711 @item start_sample, ss
712 Specify the number of the start sample for starting to apply the fade
713 effect. Default is 0.
716 Specify the number of samples for which the fade effect has to last. At
717 the end of the fade-in effect the output audio will have the same
718 volume as the input audio, at the end of the fade-out transition
719 the output audio will be silence. Default is 44100.
722 Specify the start time of the fade effect. Default is 0.
723 The value must be specified as a time duration; see
724 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
725 for the accepted syntax.
726 If set this option is used instead of @var{start_sample}.
729 Specify the duration of the fade effect. See
730 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
731 for the accepted syntax.
732 At the end of the fade-in effect the output audio will have the same
733 volume as the input audio, at the end of the fade-out transition
734 the output audio will be silence.
735 By default the duration is determined by @var{nb_samples}.
736 If set this option is used instead of @var{nb_samples}.
739 Set curve for fade transition.
741 It accepts the following values:
744 select triangular, linear slope (default)
746 select quarter of sine wave
748 select half of sine wave
750 select exponential sine wave
754 select inverted parabola
768 select inverted quarter of sine wave
770 select inverted half of sine wave
772 select double-exponential seat
774 select double-exponential sigmoid
782 Fade in first 15 seconds of audio:
788 Fade out last 25 seconds of a 900 seconds audio:
790 afade=t=out:st=875:d=25
795 Apply arbitrary expressions to samples in frequency domain.
799 Set frequency domain real expression for each separate channel separated
800 by '|'. Default is "1".
801 If the number of input channels is greater than the number of
802 expressions, the last specified expression is used for the remaining
806 Set frequency domain imaginary expression for each separate channel
807 separated by '|'. If not set, @var{real} option is used.
809 Each expression in @var{real} and @var{imag} can contain the following
817 current frequency bin number
820 number of available bins
823 channel number of the current expression
835 It accepts the following values:
851 Default is @code{w4096}
854 Set window function. Default is @code{hann}.
857 Set window overlap. If set to 1, the recommended overlap for selected
858 window function will be picked. Default is @code{0.75}.
865 Leave almost only low frequencies in audio:
867 afftfilt="1-clip((b/nb)*b,0,1)"
874 Set output format constraints for the input audio. The framework will
875 negotiate the most appropriate format to minimize conversions.
877 It accepts the following parameters:
881 A '|'-separated list of requested sample formats.
884 A '|'-separated list of requested sample rates.
886 @item channel_layouts
887 A '|'-separated list of requested channel layouts.
889 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
890 for the required syntax.
893 If a parameter is omitted, all values are allowed.
895 Force the output to either unsigned 8-bit or signed 16-bit stereo
897 aformat=sample_fmts=u8|s16:channel_layouts=stereo
902 A gate is mainly used to reduce lower parts of a signal. This kind of signal
903 processing reduces disturbing noise between useful signals.
905 Gating is done by detecting the volume below a chosen level @var{threshold}
906 and divide it by the factor set with @var{ratio}. The bottom of the noise
907 floor is set via @var{range}. Because an exact manipulation of the signal
908 would cause distortion of the waveform the reduction can be levelled over
909 time. This is done by setting @var{attack} and @var{release}.
911 @var{attack} determines how long the signal has to fall below the threshold
912 before any reduction will occur and @var{release} sets the time the signal
913 has to raise above the threshold to reduce the reduction again.
914 Shorter signals than the chosen attack time will be left untouched.
918 Set input level before filtering.
919 Default is 1. Allowed range is from 0.015625 to 64.
922 Set the level of gain reduction when the signal is below the threshold.
923 Default is 0.06125. Allowed range is from 0 to 1.
926 If a signal rises above this level the gain reduction is released.
927 Default is 0.125. Allowed range is from 0 to 1.
930 Set a ratio about which the signal is reduced.
931 Default is 2. Allowed range is from 1 to 9000.
934 Amount of milliseconds the signal has to rise above the threshold before gain
936 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
939 Amount of milliseconds the signal has to fall below the threshold before the
940 reduction is increased again. Default is 250 milliseconds.
941 Allowed range is from 0.01 to 9000.
944 Set amount of amplification of signal after processing.
945 Default is 1. Allowed range is from 1 to 64.
948 Curve the sharp knee around the threshold to enter gain reduction more softly.
949 Default is 2.828427125. Allowed range is from 1 to 8.
952 Choose if exact signal should be taken for detection or an RMS like one.
953 Default is rms. Can be peak or rms.
956 Choose if the average level between all channels or the louder channel affects
958 Default is average. Can be average or maximum.
963 The limiter prevents input signal from raising over a desired threshold.
964 This limiter uses lookahead technology to prevent your signal from distorting.
965 It means that there is a small delay after signal is processed. Keep in mind
966 that the delay it produces is the attack time you set.
968 The filter accepts the following options:
972 Set input gain. Default is 1.
975 Set output gain. Default is 1.
978 Don't let signals above this level pass the limiter. Default is 1.
981 The limiter will reach its attenuation level in this amount of time in
982 milliseconds. Default is 5 milliseconds.
985 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
986 Default is 50 milliseconds.
989 When gain reduction is always needed ASC takes care of releasing to an
990 average reduction level rather than reaching a reduction of 0 in the release
994 Select how much the release time is affected by ASC, 0 means nearly no changes
995 in release time while 1 produces higher release times.
998 Auto level output signal. Default is enabled.
999 This normalizes audio back to 0dB if enabled.
1002 Depending on picked setting it is recommended to upsample input 2x or 4x times
1003 with @ref{aresample} before applying this filter.
1007 Apply a two-pole all-pass filter with central frequency (in Hz)
1008 @var{frequency}, and filter-width @var{width}.
1009 An all-pass filter changes the audio's frequency to phase relationship
1010 without changing its frequency to amplitude relationship.
1012 The filter accepts the following options:
1016 Set frequency in Hz.
1019 Set method to specify band-width of filter.
1032 Specify the band-width of a filter in width_type units.
1039 The filter accepts the following options:
1043 Set the number of loops.
1046 Set maximal number of samples.
1049 Set first sample of loop.
1055 Merge two or more audio streams into a single multi-channel stream.
1057 The filter accepts the following options:
1062 Set the number of inputs. Default is 2.
1066 If the channel layouts of the inputs are disjoint, and therefore compatible,
1067 the channel layout of the output will be set accordingly and the channels
1068 will be reordered as necessary. If the channel layouts of the inputs are not
1069 disjoint, the output will have all the channels of the first input then all
1070 the channels of the second input, in that order, and the channel layout of
1071 the output will be the default value corresponding to the total number of
1074 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1075 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1076 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1077 first input, b1 is the first channel of the second input).
1079 On the other hand, if both input are in stereo, the output channels will be
1080 in the default order: a1, a2, b1, b2, and the channel layout will be
1081 arbitrarily set to 4.0, which may or may not be the expected value.
1083 All inputs must have the same sample rate, and format.
1085 If inputs do not have the same duration, the output will stop with the
1088 @subsection Examples
1092 Merge two mono files into a stereo stream:
1094 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1098 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1100 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
1106 Mixes multiple audio inputs into a single output.
1108 Note that this filter only supports float samples (the @var{amerge}
1109 and @var{pan} audio filters support many formats). If the @var{amix}
1110 input has integer samples then @ref{aresample} will be automatically
1111 inserted to perform the conversion to float samples.
1115 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1117 will mix 3 input audio streams to a single output with the same duration as the
1118 first input and a dropout transition time of 3 seconds.
1120 It accepts the following parameters:
1124 The number of inputs. If unspecified, it defaults to 2.
1127 How to determine the end-of-stream.
1131 The duration of the longest input. (default)
1134 The duration of the shortest input.
1137 The duration of the first input.
1141 @item dropout_transition
1142 The transition time, in seconds, for volume renormalization when an input
1143 stream ends. The default value is 2 seconds.
1147 @section anequalizer
1149 High-order parametric multiband equalizer for each channel.
1151 It accepts the following parameters:
1155 This option string is in format:
1156 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1157 Each equalizer band is separated by '|'.
1161 Set channel number to which equalization will be applied.
1162 If input doesn't have that channel the entry is ignored.
1165 Set central frequency for band.
1166 If input doesn't have that frequency the entry is ignored.
1169 Set band width in hertz.
1172 Set band gain in dB.
1175 Set filter type for band, optional, can be:
1179 Butterworth, this is default.
1190 With this option activated frequency response of anequalizer is displayed
1194 Set video stream size. Only useful if curves option is activated.
1197 Set max gain that will be displayed. Only useful if curves option is activated.
1198 Setting this to reasonable value allows to display gain which is derived from
1199 neighbour bands which are too close to each other and thus produce higher gain
1200 when both are activated.
1203 Set frequency scale used to draw frequency response in video output.
1204 Can be linear or logarithmic. Default is logarithmic.
1207 Set color for each channel curve which is going to be displayed in video stream.
1208 This is list of color names separated by space or by '|'.
1209 Unrecognised or missing colors will be replaced by white color.
1212 @subsection Examples
1216 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1217 for first 2 channels using Chebyshev type 1 filter:
1219 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1223 @subsection Commands
1225 This filter supports the following commands:
1228 Alter existing filter parameters.
1229 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1231 @var{fN} is existing filter number, starting from 0, if no such filter is available
1233 @var{freq} set new frequency parameter.
1234 @var{width} set new width parameter in herz.
1235 @var{gain} set new gain parameter in dB.
1237 Full filter invocation with asendcmd may look like this:
1238 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1243 Pass the audio source unchanged to the output.
1247 Pad the end of an audio stream with silence.
1249 This can be used together with @command{ffmpeg} @option{-shortest} to
1250 extend audio streams to the same length as the video stream.
1252 A description of the accepted options follows.
1256 Set silence packet size. Default value is 4096.
1259 Set the number of samples of silence to add to the end. After the
1260 value is reached, the stream is terminated. This option is mutually
1261 exclusive with @option{whole_len}.
1264 Set the minimum total number of samples in the output audio stream. If
1265 the value is longer than the input audio length, silence is added to
1266 the end, until the value is reached. This option is mutually exclusive
1267 with @option{pad_len}.
1270 If neither the @option{pad_len} nor the @option{whole_len} option is
1271 set, the filter will add silence to the end of the input stream
1274 @subsection Examples
1278 Add 1024 samples of silence to the end of the input:
1284 Make sure the audio output will contain at least 10000 samples, pad
1285 the input with silence if required:
1287 apad=whole_len=10000
1291 Use @command{ffmpeg} to pad the audio input with silence, so that the
1292 video stream will always result the shortest and will be converted
1293 until the end in the output file when using the @option{shortest}
1296 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1301 Add a phasing effect to the input audio.
1303 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1304 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1306 A description of the accepted parameters follows.
1310 Set input gain. Default is 0.4.
1313 Set output gain. Default is 0.74
1316 Set delay in milliseconds. Default is 3.0.
1319 Set decay. Default is 0.4.
1322 Set modulation speed in Hz. Default is 0.5.
1325 Set modulation type. Default is triangular.
1327 It accepts the following values:
1336 Audio pulsator is something between an autopanner and a tremolo.
1337 But it can produce funny stereo effects as well. Pulsator changes the volume
1338 of the left and right channel based on a LFO (low frequency oscillator) with
1339 different waveforms and shifted phases.
1340 This filter have the ability to define an offset between left and right
1341 channel. An offset of 0 means that both LFO shapes match each other.
1342 The left and right channel are altered equally - a conventional tremolo.
1343 An offset of 50% means that the shape of the right channel is exactly shifted
1344 in phase (or moved backwards about half of the frequency) - pulsator acts as
1345 an autopanner. At 1 both curves match again. Every setting in between moves the
1346 phase shift gapless between all stages and produces some "bypassing" sounds with
1347 sine and triangle waveforms. The more you set the offset near 1 (starting from
1348 the 0.5) the faster the signal passes from the left to the right speaker.
1350 The filter accepts the following options:
1354 Set input gain. By default it is 1. Range is [0.015625 - 64].
1357 Set output gain. By default it is 1. Range is [0.015625 - 64].
1360 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1361 sawup or sawdown. Default is sine.
1364 Set modulation. Define how much of original signal is affected by the LFO.
1367 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1370 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1373 Set pulse width. Default is 1. Allowed range is [0 - 2].
1376 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1379 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1383 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1387 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1388 if timing is set to hz.
1394 Resample the input audio to the specified parameters, using the
1395 libswresample library. If none are specified then the filter will
1396 automatically convert between its input and output.
1398 This filter is also able to stretch/squeeze the audio data to make it match
1399 the timestamps or to inject silence / cut out audio to make it match the
1400 timestamps, do a combination of both or do neither.
1402 The filter accepts the syntax
1403 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1404 expresses a sample rate and @var{resampler_options} is a list of
1405 @var{key}=@var{value} pairs, separated by ":". See the
1406 ffmpeg-resampler manual for the complete list of supported options.
1408 @subsection Examples
1412 Resample the input audio to 44100Hz:
1418 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1419 samples per second compensation:
1421 aresample=async=1000
1427 Reverse an audio clip.
1429 Warning: This filter requires memory to buffer the entire clip, so trimming
1432 @subsection Examples
1436 Take the first 5 seconds of a clip, and reverse it.
1438 atrim=end=5,areverse
1442 @section asetnsamples
1444 Set the number of samples per each output audio frame.
1446 The last output packet may contain a different number of samples, as
1447 the filter will flush all the remaining samples when the input audio
1450 The filter accepts the following options:
1454 @item nb_out_samples, n
1455 Set the number of frames per each output audio frame. The number is
1456 intended as the number of samples @emph{per each channel}.
1457 Default value is 1024.
1460 If set to 1, the filter will pad the last audio frame with zeroes, so
1461 that the last frame will contain the same number of samples as the
1462 previous ones. Default value is 1.
1465 For example, to set the number of per-frame samples to 1234 and
1466 disable padding for the last frame, use:
1468 asetnsamples=n=1234:p=0
1473 Set the sample rate without altering the PCM data.
1474 This will result in a change of speed and pitch.
1476 The filter accepts the following options:
1479 @item sample_rate, r
1480 Set the output sample rate. Default is 44100 Hz.
1485 Show a line containing various information for each input audio frame.
1486 The input audio is not modified.
1488 The shown line contains a sequence of key/value pairs of the form
1489 @var{key}:@var{value}.
1491 The following values are shown in the output:
1495 The (sequential) number of the input frame, starting from 0.
1498 The presentation timestamp of the input frame, in time base units; the time base
1499 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1502 The presentation timestamp of the input frame in seconds.
1505 position of the frame in the input stream, -1 if this information in
1506 unavailable and/or meaningless (for example in case of synthetic audio)
1515 The sample rate for the audio frame.
1518 The number of samples (per channel) in the frame.
1521 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1522 audio, the data is treated as if all the planes were concatenated.
1524 @item plane_checksums
1525 A list of Adler-32 checksums for each data plane.
1531 Display time domain statistical information about the audio channels.
1532 Statistics are calculated and displayed for each audio channel and,
1533 where applicable, an overall figure is also given.
1535 It accepts the following option:
1538 Short window length in seconds, used for peak and trough RMS measurement.
1539 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1543 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1544 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1547 Available keys for each channel are:
1578 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1579 this @code{lavfi.astats.Overall.Peak_count}.
1581 For description what each key means read below.
1584 Set number of frame after which stats are going to be recalculated.
1585 Default is disabled.
1588 A description of each shown parameter follows:
1592 Mean amplitude displacement from zero.
1595 Minimal sample level.
1598 Maximal sample level.
1600 @item Min difference
1601 Minimal difference between two consecutive samples.
1603 @item Max difference
1604 Maximal difference between two consecutive samples.
1606 @item Mean difference
1607 Mean difference between two consecutive samples.
1608 The average of each difference between two consecutive samples.
1612 Standard peak and RMS level measured in dBFS.
1616 Peak and trough values for RMS level measured over a short window.
1619 Standard ratio of peak to RMS level (note: not in dB).
1622 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1623 (i.e. either @var{Min level} or @var{Max level}).
1626 Number of occasions (not the number of samples) that the signal attained either
1627 @var{Min level} or @var{Max level}.
1630 Overall bit depth of audio. Number of bits used for each sample.
1635 Synchronize audio data with timestamps by squeezing/stretching it and/or
1636 dropping samples/adding silence when needed.
1638 This filter is not built by default, please use @ref{aresample} to do squeezing/stretching.
1640 It accepts the following parameters:
1644 Enable stretching/squeezing the data to make it match the timestamps. Disabled
1645 by default. When disabled, time gaps are covered with silence.
1648 The minimum difference between timestamps and audio data (in seconds) to trigger
1649 adding/dropping samples. The default value is 0.1. If you get an imperfect
1650 sync with this filter, try setting this parameter to 0.
1653 The maximum compensation in samples per second. Only relevant with compensate=1.
1654 The default value is 500.
1657 Assume that the first PTS should be this value. The time base is 1 / sample
1658 rate. This allows for padding/trimming at the start of the stream. By default,
1659 no assumption is made about the first frame's expected PTS, so no padding or
1660 trimming is done. For example, this could be set to 0 to pad the beginning with
1661 silence if an audio stream starts after the video stream or to trim any samples
1662 with a negative PTS due to encoder delay.
1670 The filter accepts exactly one parameter, the audio tempo. If not
1671 specified then the filter will assume nominal 1.0 tempo. Tempo must
1672 be in the [0.5, 2.0] range.
1674 @subsection Examples
1678 Slow down audio to 80% tempo:
1684 To speed up audio to 125% tempo:
1692 Trim the input so that the output contains one continuous subpart of the input.
1694 It accepts the following parameters:
1697 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1698 sample with the timestamp @var{start} will be the first sample in the output.
1701 Specify time of the first audio sample that will be dropped, i.e. the
1702 audio sample immediately preceding the one with the timestamp @var{end} will be
1703 the last sample in the output.
1706 Same as @var{start}, except this option sets the start timestamp in samples
1710 Same as @var{end}, except this option sets the end timestamp in samples instead
1714 The maximum duration of the output in seconds.
1717 The number of the first sample that should be output.
1720 The number of the first sample that should be dropped.
1723 @option{start}, @option{end}, and @option{duration} are expressed as time
1724 duration specifications; see
1725 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1727 Note that the first two sets of the start/end options and the @option{duration}
1728 option look at the frame timestamp, while the _sample options simply count the
1729 samples that pass through the filter. So start/end_pts and start/end_sample will
1730 give different results when the timestamps are wrong, inexact or do not start at
1731 zero. Also note that this filter does not modify the timestamps. If you wish
1732 to have the output timestamps start at zero, insert the asetpts filter after the
1735 If multiple start or end options are set, this filter tries to be greedy and
1736 keep all samples that match at least one of the specified constraints. To keep
1737 only the part that matches all the constraints at once, chain multiple atrim
1740 The defaults are such that all the input is kept. So it is possible to set e.g.
1741 just the end values to keep everything before the specified time.
1746 Drop everything except the second minute of input:
1748 ffmpeg -i INPUT -af atrim=60:120
1752 Keep only the first 1000 samples:
1754 ffmpeg -i INPUT -af atrim=end_sample=1000
1761 Apply a two-pole Butterworth band-pass filter with central
1762 frequency @var{frequency}, and (3dB-point) band-width width.
1763 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1764 instead of the default: constant 0dB peak gain.
1765 The filter roll off at 6dB per octave (20dB per decade).
1767 The filter accepts the following options:
1771 Set the filter's central frequency. Default is @code{3000}.
1774 Constant skirt gain if set to 1. Defaults to 0.
1777 Set method to specify band-width of filter.
1790 Specify the band-width of a filter in width_type units.
1795 Apply a two-pole Butterworth band-reject filter with central
1796 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1797 The filter roll off at 6dB per octave (20dB per decade).
1799 The filter accepts the following options:
1803 Set the filter's central frequency. Default is @code{3000}.
1806 Set method to specify band-width of filter.
1819 Specify the band-width of a filter in width_type units.
1824 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1825 shelving filter with a response similar to that of a standard
1826 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1828 The filter accepts the following options:
1832 Give the gain at 0 Hz. Its useful range is about -20
1833 (for a large cut) to +20 (for a large boost).
1834 Beware of clipping when using a positive gain.
1837 Set the filter's central frequency and so can be used
1838 to extend or reduce the frequency range to be boosted or cut.
1839 The default value is @code{100} Hz.
1842 Set method to specify band-width of filter.
1855 Determine how steep is the filter's shelf transition.
1860 Apply a biquad IIR filter with the given coefficients.
1861 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1862 are the numerator and denominator coefficients respectively.
1865 Bauer stereo to binaural transformation, which improves headphone listening of
1866 stereo audio records.
1868 It accepts the following parameters:
1872 Pre-defined crossfeed level.
1876 Default level (fcut=700, feed=50).
1879 Chu Moy circuit (fcut=700, feed=60).
1882 Jan Meier circuit (fcut=650, feed=95).
1887 Cut frequency (in Hz).
1896 Remap input channels to new locations.
1898 It accepts the following parameters:
1900 @item channel_layout
1901 The channel layout of the output stream.
1904 Map channels from input to output. The argument is a '|'-separated list of
1905 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1906 @var{in_channel} form. @var{in_channel} can be either the name of the input
1907 channel (e.g. FL for front left) or its index in the input channel layout.
1908 @var{out_channel} is the name of the output channel or its index in the output
1909 channel layout. If @var{out_channel} is not given then it is implicitly an
1910 index, starting with zero and increasing by one for each mapping.
1913 If no mapping is present, the filter will implicitly map input channels to
1914 output channels, preserving indices.
1916 For example, assuming a 5.1+downmix input MOV file,
1918 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1920 will create an output WAV file tagged as stereo from the downmix channels of
1923 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1925 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
1928 @section channelsplit
1930 Split each channel from an input audio stream into a separate output stream.
1932 It accepts the following parameters:
1934 @item channel_layout
1935 The channel layout of the input stream. The default is "stereo".
1938 For example, assuming a stereo input MP3 file,
1940 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1942 will create an output Matroska file with two audio streams, one containing only
1943 the left channel and the other the right channel.
1945 Split a 5.1 WAV file into per-channel files:
1947 ffmpeg -i in.wav -filter_complex
1948 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1949 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1950 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
1955 Add a chorus effect to the audio.
1957 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
1959 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
1960 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
1961 The modulation depth defines the range the modulated delay is played before or after
1962 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
1963 sound tuned around the original one, like in a chorus where some vocals are slightly
1966 It accepts the following parameters:
1969 Set input gain. Default is 0.4.
1972 Set output gain. Default is 0.4.
1975 Set delays. A typical delay is around 40ms to 60ms.
1987 @subsection Examples
1993 chorus=0.7:0.9:55:0.4:0.25:2
1999 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2003 Fuller sounding chorus with three delays:
2005 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
2010 Compress or expand the audio's dynamic range.
2012 It accepts the following parameters:
2018 A list of times in seconds for each channel over which the instantaneous level
2019 of the input signal is averaged to determine its volume. @var{attacks} refers to
2020 increase of volume and @var{decays} refers to decrease of volume. For most
2021 situations, the attack time (response to the audio getting louder) should be
2022 shorter than the decay time, because the human ear is more sensitive to sudden
2023 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2024 a typical value for decay is 0.8 seconds.
2025 If specified number of attacks & decays is lower than number of channels, the last
2026 set attack/decay will be used for all remaining channels.
2029 A list of points for the transfer function, specified in dB relative to the
2030 maximum possible signal amplitude. Each key points list must be defined using
2031 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2032 @code{x0/y0 x1/y1 x2/y2 ....}
2034 The input values must be in strictly increasing order but the transfer function
2035 does not have to be monotonically rising. The point @code{0/0} is assumed but
2036 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2037 function are @code{-70/-70|-60/-20}.
2040 Set the curve radius in dB for all joints. It defaults to 0.01.
2043 Set the additional gain in dB to be applied at all points on the transfer
2044 function. This allows for easy adjustment of the overall gain.
2048 Set an initial volume, in dB, to be assumed for each channel when filtering
2049 starts. This permits the user to supply a nominal level initially, so that, for
2050 example, a very large gain is not applied to initial signal levels before the
2051 companding has begun to operate. A typical value for audio which is initially
2052 quiet is -90 dB. It defaults to 0.
2055 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2056 delayed before being fed to the volume adjuster. Specifying a delay
2057 approximately equal to the attack/decay times allows the filter to effectively
2058 operate in predictive rather than reactive mode. It defaults to 0.
2062 @subsection Examples
2066 Make music with both quiet and loud passages suitable for listening to in a
2069 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2072 Another example for audio with whisper and explosion parts:
2074 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2078 A noise gate for when the noise is at a lower level than the signal:
2080 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2084 Here is another noise gate, this time for when the noise is at a higher level
2085 than the signal (making it, in some ways, similar to squelch):
2087 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2091 2:1 compression starting at -6dB:
2093 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2097 2:1 compression starting at -9dB:
2099 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2103 2:1 compression starting at -12dB:
2105 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2109 2:1 compression starting at -18dB:
2111 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2115 3:1 compression starting at -15dB:
2117 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2123 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2129 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
2133 Hard limiter at -6dB:
2135 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2139 Hard limiter at -12dB:
2141 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2145 Hard noise gate at -35 dB:
2147 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2153 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2157 @section compensationdelay
2159 Compensation Delay Line is a metric based delay to compensate differing
2160 positions of microphones or speakers.
2162 For example, you have recorded guitar with two microphones placed in
2163 different location. Because the front of sound wave has fixed speed in
2164 normal conditions, the phasing of microphones can vary and depends on
2165 their location and interposition. The best sound mix can be achieved when
2166 these microphones are in phase (synchronized). Note that distance of
2167 ~30 cm between microphones makes one microphone to capture signal in
2168 antiphase to another microphone. That makes the final mix sounding moody.
2169 This filter helps to solve phasing problems by adding different delays
2170 to each microphone track and make them synchronized.
2172 The best result can be reached when you take one track as base and
2173 synchronize other tracks one by one with it.
2174 Remember that synchronization/delay tolerance depends on sample rate, too.
2175 Higher sample rates will give more tolerance.
2177 It accepts the following parameters:
2181 Set millimeters distance. This is compensation distance for fine tuning.
2185 Set cm distance. This is compensation distance for tightening distance setup.
2189 Set meters distance. This is compensation distance for hard distance setup.
2193 Set dry amount. Amount of unprocessed (dry) signal.
2197 Set wet amount. Amount of processed (wet) signal.
2201 Set temperature degree in Celsius. This is the temperature of the environment.
2205 @section crystalizer
2206 Simple algorithm to expand audio dynamic range.
2208 The filter accepts the following options:
2212 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2213 (unchanged sound) to 10.0 (maximum effect).
2216 Enable clipping. By default is enabled.
2220 Apply a DC shift to the audio.
2222 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2223 in the recording chain) from the audio. The effect of a DC offset is reduced
2224 headroom and hence volume. The @ref{astats} filter can be used to determine if
2225 a signal has a DC offset.
2229 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2233 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2234 used to prevent clipping.
2238 Dynamic Audio Normalizer.
2240 This filter applies a certain amount of gain to the input audio in order
2241 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2242 contrast to more "simple" normalization algorithms, the Dynamic Audio
2243 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2244 This allows for applying extra gain to the "quiet" sections of the audio
2245 while avoiding distortions or clipping the "loud" sections. In other words:
2246 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2247 sections, in the sense that the volume of each section is brought to the
2248 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2249 this goal *without* applying "dynamic range compressing". It will retain 100%
2250 of the dynamic range *within* each section of the audio file.
2254 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2255 Default is 500 milliseconds.
2256 The Dynamic Audio Normalizer processes the input audio in small chunks,
2257 referred to as frames. This is required, because a peak magnitude has no
2258 meaning for just a single sample value. Instead, we need to determine the
2259 peak magnitude for a contiguous sequence of sample values. While a "standard"
2260 normalizer would simply use the peak magnitude of the complete file, the
2261 Dynamic Audio Normalizer determines the peak magnitude individually for each
2262 frame. The length of a frame is specified in milliseconds. By default, the
2263 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2264 been found to give good results with most files.
2265 Note that the exact frame length, in number of samples, will be determined
2266 automatically, based on the sampling rate of the individual input audio file.
2269 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2270 number. Default is 31.
2271 Probably the most important parameter of the Dynamic Audio Normalizer is the
2272 @code{window size} of the Gaussian smoothing filter. The filter's window size
2273 is specified in frames, centered around the current frame. For the sake of
2274 simplicity, this must be an odd number. Consequently, the default value of 31
2275 takes into account the current frame, as well as the 15 preceding frames and
2276 the 15 subsequent frames. Using a larger window results in a stronger
2277 smoothing effect and thus in less gain variation, i.e. slower gain
2278 adaptation. Conversely, using a smaller window results in a weaker smoothing
2279 effect and thus in more gain variation, i.e. faster gain adaptation.
2280 In other words, the more you increase this value, the more the Dynamic Audio
2281 Normalizer will behave like a "traditional" normalization filter. On the
2282 contrary, the more you decrease this value, the more the Dynamic Audio
2283 Normalizer will behave like a dynamic range compressor.
2286 Set the target peak value. This specifies the highest permissible magnitude
2287 level for the normalized audio input. This filter will try to approach the
2288 target peak magnitude as closely as possible, but at the same time it also
2289 makes sure that the normalized signal will never exceed the peak magnitude.
2290 A frame's maximum local gain factor is imposed directly by the target peak
2291 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2292 It is not recommended to go above this value.
2295 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2296 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2297 factor for each input frame, i.e. the maximum gain factor that does not
2298 result in clipping or distortion. The maximum gain factor is determined by
2299 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2300 additionally bounds the frame's maximum gain factor by a predetermined
2301 (global) maximum gain factor. This is done in order to avoid excessive gain
2302 factors in "silent" or almost silent frames. By default, the maximum gain
2303 factor is 10.0, For most inputs the default value should be sufficient and
2304 it usually is not recommended to increase this value. Though, for input
2305 with an extremely low overall volume level, it may be necessary to allow even
2306 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2307 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2308 Instead, a "sigmoid" threshold function will be applied. This way, the
2309 gain factors will smoothly approach the threshold value, but never exceed that
2313 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2314 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2315 This means that the maximum local gain factor for each frame is defined
2316 (only) by the frame's highest magnitude sample. This way, the samples can
2317 be amplified as much as possible without exceeding the maximum signal
2318 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2319 Normalizer can also take into account the frame's root mean square,
2320 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2321 determine the power of a time-varying signal. It is therefore considered
2322 that the RMS is a better approximation of the "perceived loudness" than
2323 just looking at the signal's peak magnitude. Consequently, by adjusting all
2324 frames to a constant RMS value, a uniform "perceived loudness" can be
2325 established. If a target RMS value has been specified, a frame's local gain
2326 factor is defined as the factor that would result in exactly that RMS value.
2327 Note, however, that the maximum local gain factor is still restricted by the
2328 frame's highest magnitude sample, in order to prevent clipping.
2331 Enable channels coupling. By default is enabled.
2332 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2333 amount. This means the same gain factor will be applied to all channels, i.e.
2334 the maximum possible gain factor is determined by the "loudest" channel.
2335 However, in some recordings, it may happen that the volume of the different
2336 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2337 In this case, this option can be used to disable the channel coupling. This way,
2338 the gain factor will be determined independently for each channel, depending
2339 only on the individual channel's highest magnitude sample. This allows for
2340 harmonizing the volume of the different channels.
2343 Enable DC bias correction. By default is disabled.
2344 An audio signal (in the time domain) is a sequence of sample values.
2345 In the Dynamic Audio Normalizer these sample values are represented in the
2346 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2347 audio signal, or "waveform", should be centered around the zero point.
2348 That means if we calculate the mean value of all samples in a file, or in a
2349 single frame, then the result should be 0.0 or at least very close to that
2350 value. If, however, there is a significant deviation of the mean value from
2351 0.0, in either positive or negative direction, this is referred to as a
2352 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2353 Audio Normalizer provides optional DC bias correction.
2354 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2355 the mean value, or "DC correction" offset, of each input frame and subtract
2356 that value from all of the frame's sample values which ensures those samples
2357 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2358 boundaries, the DC correction offset values will be interpolated smoothly
2359 between neighbouring frames.
2362 Enable alternative boundary mode. By default is disabled.
2363 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2364 around each frame. This includes the preceding frames as well as the
2365 subsequent frames. However, for the "boundary" frames, located at the very
2366 beginning and at the very end of the audio file, not all neighbouring
2367 frames are available. In particular, for the first few frames in the audio
2368 file, the preceding frames are not known. And, similarly, for the last few
2369 frames in the audio file, the subsequent frames are not known. Thus, the
2370 question arises which gain factors should be assumed for the missing frames
2371 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2372 to deal with this situation. The default boundary mode assumes a gain factor
2373 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2374 "fade out" at the beginning and at the end of the input, respectively.
2377 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2378 By default, the Dynamic Audio Normalizer does not apply "traditional"
2379 compression. This means that signal peaks will not be pruned and thus the
2380 full dynamic range will be retained within each local neighbourhood. However,
2381 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2382 normalization algorithm with a more "traditional" compression.
2383 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2384 (thresholding) function. If (and only if) the compression feature is enabled,
2385 all input frames will be processed by a soft knee thresholding function prior
2386 to the actual normalization process. Put simply, the thresholding function is
2387 going to prune all samples whose magnitude exceeds a certain threshold value.
2388 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2389 value. Instead, the threshold value will be adjusted for each individual
2391 In general, smaller parameters result in stronger compression, and vice versa.
2392 Values below 3.0 are not recommended, because audible distortion may appear.
2397 Make audio easier to listen to on headphones.
2399 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2400 so that when listened to on headphones the stereo image is moved from
2401 inside your head (standard for headphones) to outside and in front of
2402 the listener (standard for speakers).
2408 Apply a two-pole peaking equalisation (EQ) filter. With this
2409 filter, the signal-level at and around a selected frequency can
2410 be increased or decreased, whilst (unlike bandpass and bandreject
2411 filters) that at all other frequencies is unchanged.
2413 In order to produce complex equalisation curves, this filter can
2414 be given several times, each with a different central frequency.
2416 The filter accepts the following options:
2420 Set the filter's central frequency in Hz.
2423 Set method to specify band-width of filter.
2436 Specify the band-width of a filter in width_type units.
2439 Set the required gain or attenuation in dB.
2440 Beware of clipping when using a positive gain.
2443 @subsection Examples
2446 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2448 equalizer=f=1000:width_type=h:width=200:g=-10
2452 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2454 equalizer=f=1000:width_type=q:width=1:g=2,equalizer=f=100:width_type=q:width=2:g=-5
2458 @section extrastereo
2460 Linearly increases the difference between left and right channels which
2461 adds some sort of "live" effect to playback.
2463 The filter accepts the following options:
2467 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2468 (average of both channels), with 1.0 sound will be unchanged, with
2469 -1.0 left and right channels will be swapped.
2472 Enable clipping. By default is enabled.
2475 @section firequalizer
2476 Apply FIR Equalization using arbitrary frequency response.
2478 The filter accepts the following option:
2482 Set gain curve equation (in dB). The expression can contain variables:
2485 the evaluated frequency
2489 channel number, set to 0 when multichannels evaluation is disabled
2491 channel id, see libavutil/channel_layout.h, set to the first channel id when
2492 multichannels evaluation is disabled
2496 channel_layout, see libavutil/channel_layout.h
2501 @item gain_interpolate(f)
2502 interpolate gain on frequency f based on gain_entry
2504 This option is also available as command. Default is @code{gain_interpolate(f)}.
2507 Set gain entry for gain_interpolate function. The expression can
2511 store gain entry at frequency f with value g
2513 This option is also available as command.
2516 Set filter delay in seconds. Higher value means more accurate.
2517 Default is @code{0.01}.
2520 Set filter accuracy in Hz. Lower value means more accurate.
2521 Default is @code{5}.
2524 Set window function. Acceptable values are:
2527 rectangular window, useful when gain curve is already smooth
2529 hann window (default)
2535 3-terms continuous 1st derivative nuttall window
2537 minimum 3-terms discontinuous nuttall window
2539 4-terms continuous 1st derivative nuttall window
2541 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2543 blackman-harris window
2547 If enabled, use fixed number of audio samples. This improves speed when
2548 filtering with large delay. Default is disabled.
2551 Enable multichannels evaluation on gain. Default is disabled.
2554 Enable zero phase mode by substracting timestamp to compensate delay.
2555 Default is disabled.
2558 @subsection Examples
2563 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2566 lowpass at 1000 Hz with gain_entry:
2568 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2571 custom equalization:
2573 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2576 higher delay with zero phase to compensate delay:
2578 firequalizer=delay=0.1:fixed=on:zero_phase=on
2581 lowpass on left channel, highpass on right channel:
2583 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2584 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2589 Apply a flanging effect to the audio.
2591 The filter accepts the following options:
2595 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2598 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2601 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2605 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2606 Default value is 71.
2609 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2612 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2613 Default value is @var{sinusoidal}.
2616 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2617 Default value is 25.
2620 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2621 Default is @var{linear}.
2626 Apply a high-pass filter with 3dB point frequency.
2627 The filter can be either single-pole, or double-pole (the default).
2628 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2630 The filter accepts the following options:
2634 Set frequency in Hz. Default is 3000.
2637 Set number of poles. Default is 2.
2640 Set method to specify band-width of filter.
2653 Specify the band-width of a filter in width_type units.
2654 Applies only to double-pole filter.
2655 The default is 0.707q and gives a Butterworth response.
2660 Join multiple input streams into one multi-channel stream.
2662 It accepts the following parameters:
2666 The number of input streams. It defaults to 2.
2668 @item channel_layout
2669 The desired output channel layout. It defaults to stereo.
2672 Map channels from inputs to output. The argument is a '|'-separated list of
2673 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2674 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2675 can be either the name of the input channel (e.g. FL for front left) or its
2676 index in the specified input stream. @var{out_channel} is the name of the output
2680 The filter will attempt to guess the mappings when they are not specified
2681 explicitly. It does so by first trying to find an unused matching input channel
2682 and if that fails it picks the first unused input channel.
2684 Join 3 inputs (with properly set channel layouts):
2686 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2689 Build a 5.1 output from 6 single-channel streams:
2691 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2692 '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'
2698 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2700 To enable compilation of this filter you need to configure FFmpeg with
2701 @code{--enable-ladspa}.
2705 Specifies the name of LADSPA plugin library to load. If the environment
2706 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2707 each one of the directories specified by the colon separated list in
2708 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2709 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2710 @file{/usr/lib/ladspa/}.
2713 Specifies the plugin within the library. Some libraries contain only
2714 one plugin, but others contain many of them. If this is not set filter
2715 will list all available plugins within the specified library.
2718 Set the '|' separated list of controls which are zero or more floating point
2719 values that determine the behavior of the loaded plugin (for example delay,
2721 Controls need to be defined using the following syntax:
2722 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2723 @var{valuei} is the value set on the @var{i}-th control.
2724 Alternatively they can be also defined using the following syntax:
2725 @var{value0}|@var{value1}|@var{value2}|..., where
2726 @var{valuei} is the value set on the @var{i}-th control.
2727 If @option{controls} is set to @code{help}, all available controls and
2728 their valid ranges are printed.
2730 @item sample_rate, s
2731 Specify the sample rate, default to 44100. Only used if plugin have
2735 Set the number of samples per channel per each output frame, default
2736 is 1024. Only used if plugin have zero inputs.
2739 Set the minimum duration of the sourced audio. See
2740 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2741 for the accepted syntax.
2742 Note that the resulting duration may be greater than the specified duration,
2743 as the generated audio is always cut at the end of a complete frame.
2744 If not specified, or the expressed duration is negative, the audio is
2745 supposed to be generated forever.
2746 Only used if plugin have zero inputs.
2750 @subsection Examples
2754 List all available plugins within amp (LADSPA example plugin) library:
2760 List all available controls and their valid ranges for @code{vcf_notch}
2761 plugin from @code{VCF} library:
2763 ladspa=f=vcf:p=vcf_notch:c=help
2767 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2770 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2774 Add reverberation to the audio using TAP-plugins
2775 (Tom's Audio Processing plugins):
2777 ladspa=file=tap_reverb:tap_reverb
2781 Generate white noise, with 0.2 amplitude:
2783 ladspa=file=cmt:noise_source_white:c=c0=.2
2787 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2788 @code{C* Audio Plugin Suite} (CAPS) library:
2790 ladspa=file=caps:Click:c=c1=20'
2794 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2796 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2800 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2801 @code{SWH Plugins} collection:
2803 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2807 Attenuate low frequencies using Multiband EQ from Steve Harris
2808 @code{SWH Plugins} collection:
2810 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2814 @subsection Commands
2816 This filter supports the following commands:
2819 Modify the @var{N}-th control value.
2821 If the specified value is not valid, it is ignored and prior one is kept.
2826 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2827 Support for both single pass (livestreams, files) and double pass (files) modes.
2828 This algorithm can target IL, LRA, and maximum true peak.
2830 To enable compilation of this filter you need to configure FFmpeg with
2831 @code{--enable-libebur128}.
2833 The filter accepts the following options:
2837 Set integrated loudness target.
2838 Range is -70.0 - -5.0. Default value is -24.0.
2841 Set loudness range target.
2842 Range is 1.0 - 20.0. Default value is 7.0.
2845 Set maximum true peak.
2846 Range is -9.0 - +0.0. Default value is -2.0.
2848 @item measured_I, measured_i
2849 Measured IL of input file.
2850 Range is -99.0 - +0.0.
2852 @item measured_LRA, measured_lra
2853 Measured LRA of input file.
2854 Range is 0.0 - 99.0.
2856 @item measured_TP, measured_tp
2857 Measured true peak of input file.
2858 Range is -99.0 - +99.0.
2860 @item measured_thresh
2861 Measured threshold of input file.
2862 Range is -99.0 - +0.0.
2865 Set offset gain. Gain is applied before the true-peak limiter.
2866 Range is -99.0 - +99.0. Default is +0.0.
2869 Normalize linearly if possible.
2870 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2871 to be specified in order to use this mode.
2872 Options are true or false. Default is true.
2875 Treat mono input files as "dual-mono". If a mono file is intended for playback
2876 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2877 If set to @code{true}, this option will compensate for this effect.
2878 Multi-channel input files are not affected by this option.
2879 Options are true or false. Default is false.
2882 Set print format for stats. Options are summary, json, or none.
2883 Default value is none.
2888 Apply a low-pass filter with 3dB point frequency.
2889 The filter can be either single-pole or double-pole (the default).
2890 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2892 The filter accepts the following options:
2896 Set frequency in Hz. Default is 500.
2899 Set number of poles. Default is 2.
2902 Set method to specify band-width of filter.
2915 Specify the band-width of a filter in width_type units.
2916 Applies only to double-pole filter.
2917 The default is 0.707q and gives a Butterworth response.
2923 Mix channels with specific gain levels. The filter accepts the output
2924 channel layout followed by a set of channels definitions.
2926 This filter is also designed to efficiently remap the channels of an audio
2929 The filter accepts parameters of the form:
2930 "@var{l}|@var{outdef}|@var{outdef}|..."
2934 output channel layout or number of channels
2937 output channel specification, of the form:
2938 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
2941 output channel to define, either a channel name (FL, FR, etc.) or a channel
2942 number (c0, c1, etc.)
2945 multiplicative coefficient for the channel, 1 leaving the volume unchanged
2948 input channel to use, see out_name for details; it is not possible to mix
2949 named and numbered input channels
2952 If the `=' in a channel specification is replaced by `<', then the gains for
2953 that specification will be renormalized so that the total is 1, thus
2954 avoiding clipping noise.
2956 @subsection Mixing examples
2958 For example, if you want to down-mix from stereo to mono, but with a bigger
2959 factor for the left channel:
2961 pan=1c|c0=0.9*c0+0.1*c1
2964 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
2965 7-channels surround:
2967 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
2970 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
2971 that should be preferred (see "-ac" option) unless you have very specific
2974 @subsection Remapping examples
2976 The channel remapping will be effective if, and only if:
2979 @item gain coefficients are zeroes or ones,
2980 @item only one input per channel output,
2983 If all these conditions are satisfied, the filter will notify the user ("Pure
2984 channel mapping detected"), and use an optimized and lossless method to do the
2987 For example, if you have a 5.1 source and want a stereo audio stream by
2988 dropping the extra channels:
2990 pan="stereo| c0=FL | c1=FR"
2993 Given the same source, you can also switch front left and front right channels
2994 and keep the input channel layout:
2996 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
2999 If the input is a stereo audio stream, you can mute the front left channel (and
3000 still keep the stereo channel layout) with:
3005 Still with a stereo audio stream input, you can copy the right channel in both
3006 front left and right:
3008 pan="stereo| c0=FR | c1=FR"
3013 ReplayGain scanner filter. This filter takes an audio stream as an input and
3014 outputs it unchanged.
3015 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3019 Convert the audio sample format, sample rate and channel layout. It is
3020 not meant to be used directly.
3023 Apply time-stretching and pitch-shifting with librubberband.
3025 The filter accepts the following options:
3029 Set tempo scale factor.
3032 Set pitch scale factor.
3035 Set transients detector.
3036 Possible values are:
3045 Possible values are:
3054 Possible values are:
3061 Set processing window size.
3062 Possible values are:
3071 Possible values are:
3078 Enable formant preservation when shift pitching.
3079 Possible values are:
3087 Possible values are:
3096 Possible values are:
3103 @section sidechaincompress
3105 This filter acts like normal compressor but has the ability to compress
3106 detected signal using second input signal.
3107 It needs two input streams and returns one output stream.
3108 First input stream will be processed depending on second stream signal.
3109 The filtered signal then can be filtered with other filters in later stages of
3110 processing. See @ref{pan} and @ref{amerge} filter.
3112 The filter accepts the following options:
3116 Set input gain. Default is 1. Range is between 0.015625 and 64.
3119 If a signal of second stream raises above this level it will affect the gain
3120 reduction of first stream.
3121 By default is 0.125. Range is between 0.00097563 and 1.
3124 Set a ratio about which the signal is reduced. 1:2 means that if the level
3125 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3126 Default is 2. Range is between 1 and 20.
3129 Amount of milliseconds the signal has to rise above the threshold before gain
3130 reduction starts. Default is 20. Range is between 0.01 and 2000.
3133 Amount of milliseconds the signal has to fall below the threshold before
3134 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3137 Set the amount by how much signal will be amplified after processing.
3138 Default is 2. Range is from 1 and 64.
3141 Curve the sharp knee around the threshold to enter gain reduction more softly.
3142 Default is 2.82843. Range is between 1 and 8.
3145 Choose if the @code{average} level between all channels of side-chain stream
3146 or the louder(@code{maximum}) channel of side-chain stream affects the
3147 reduction. Default is @code{average}.
3150 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3151 of @code{rms}. Default is @code{rms} which is mainly smoother.
3154 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3157 How much to use compressed signal in output. Default is 1.
3158 Range is between 0 and 1.
3161 @subsection Examples
3165 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3166 depending on the signal of 2nd input and later compressed signal to be
3167 merged with 2nd input:
3169 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3173 @section sidechaingate
3175 A sidechain gate acts like a normal (wideband) gate but has the ability to
3176 filter the detected signal before sending it to the gain reduction stage.
3177 Normally a gate uses the full range signal to detect a level above the
3179 For example: If you cut all lower frequencies from your sidechain signal
3180 the gate will decrease the volume of your track only if not enough highs
3181 appear. With this technique you are able to reduce the resonation of a
3182 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3184 It needs two input streams and returns one output stream.
3185 First input stream will be processed depending on second stream signal.
3187 The filter accepts the following options:
3191 Set input level before filtering.
3192 Default is 1. Allowed range is from 0.015625 to 64.
3195 Set the level of gain reduction when the signal is below the threshold.
3196 Default is 0.06125. Allowed range is from 0 to 1.
3199 If a signal rises above this level the gain reduction is released.
3200 Default is 0.125. Allowed range is from 0 to 1.
3203 Set a ratio about which the signal is reduced.
3204 Default is 2. Allowed range is from 1 to 9000.
3207 Amount of milliseconds the signal has to rise above the threshold before gain
3209 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3212 Amount of milliseconds the signal has to fall below the threshold before the
3213 reduction is increased again. Default is 250 milliseconds.
3214 Allowed range is from 0.01 to 9000.
3217 Set amount of amplification of signal after processing.
3218 Default is 1. Allowed range is from 1 to 64.
3221 Curve the sharp knee around the threshold to enter gain reduction more softly.
3222 Default is 2.828427125. Allowed range is from 1 to 8.
3225 Choose if exact signal should be taken for detection or an RMS like one.
3226 Default is rms. Can be peak or rms.
3229 Choose if the average level between all channels or the louder channel affects
3231 Default is average. Can be average or maximum.
3234 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3237 @section silencedetect
3239 Detect silence in an audio stream.
3241 This filter logs a message when it detects that the input audio volume is less
3242 or equal to a noise tolerance value for a duration greater or equal to the
3243 minimum detected noise duration.
3245 The printed times and duration are expressed in seconds.
3247 The filter accepts the following options:
3251 Set silence duration until notification (default is 2 seconds).
3254 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3255 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3258 @subsection Examples
3262 Detect 5 seconds of silence with -50dB noise tolerance:
3264 silencedetect=n=-50dB:d=5
3268 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3269 tolerance in @file{silence.mp3}:
3271 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3275 @section silenceremove
3277 Remove silence from the beginning, middle or end of the audio.
3279 The filter accepts the following options:
3283 This value is used to indicate if audio should be trimmed at beginning of
3284 the audio. A value of zero indicates no silence should be trimmed from the
3285 beginning. When specifying a non-zero value, it trims audio up until it
3286 finds non-silence. Normally, when trimming silence from beginning of audio
3287 the @var{start_periods} will be @code{1} but it can be increased to higher
3288 values to trim all audio up to specific count of non-silence periods.
3289 Default value is @code{0}.
3291 @item start_duration
3292 Specify the amount of time that non-silence must be detected before it stops
3293 trimming audio. By increasing the duration, bursts of noises can be treated
3294 as silence and trimmed off. Default value is @code{0}.
3296 @item start_threshold
3297 This indicates what sample value should be treated as silence. For digital
3298 audio, a value of @code{0} may be fine but for audio recorded from analog,
3299 you may wish to increase the value to account for background noise.
3300 Can be specified in dB (in case "dB" is appended to the specified value)
3301 or amplitude ratio. Default value is @code{0}.
3304 Set the count for trimming silence from the end of audio.
3305 To remove silence from the middle of a file, specify a @var{stop_periods}
3306 that is negative. This value is then treated as a positive value and is
3307 used to indicate the effect should restart processing as specified by
3308 @var{start_periods}, making it suitable for removing periods of silence
3309 in the middle of the audio.
3310 Default value is @code{0}.
3313 Specify a duration of silence that must exist before audio is not copied any
3314 more. By specifying a higher duration, silence that is wanted can be left in
3316 Default value is @code{0}.
3318 @item stop_threshold
3319 This is the same as @option{start_threshold} but for trimming silence from
3321 Can be specified in dB (in case "dB" is appended to the specified value)
3322 or amplitude ratio. Default value is @code{0}.
3325 This indicate that @var{stop_duration} length of audio should be left intact
3326 at the beginning of each period of silence.
3327 For example, if you want to remove long pauses between words but do not want
3328 to remove the pauses completely. Default value is @code{0}.
3331 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3332 and works better with digital silence which is exactly 0.
3333 Default value is @code{rms}.
3336 Set ratio used to calculate size of window for detecting silence.
3337 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3340 @subsection Examples
3344 The following example shows how this filter can be used to start a recording
3345 that does not contain the delay at the start which usually occurs between
3346 pressing the record button and the start of the performance:
3348 silenceremove=1:5:0.02
3352 Trim all silence encountered from beginning to end where there is more than 1
3353 second of silence in audio:
3355 silenceremove=0:0:0:-1:1:-90dB
3361 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3362 loudspeakers around the user for binaural listening via headphones (audio
3363 formats up to 9 channels supported).
3364 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3365 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3366 Austrian Academy of Sciences.
3368 To enable compilation of this filter you need to configure FFmpeg with
3369 @code{--enable-netcdf}.
3371 The filter accepts the following options:
3375 Set the SOFA file used for rendering.
3378 Set gain applied to audio. Value is in dB. Default is 0.
3381 Set rotation of virtual loudspeakers in deg. Default is 0.
3384 Set elevation of virtual speakers in deg. Default is 0.
3387 Set distance in meters between loudspeakers and the listener with near-field
3388 HRTFs. Default is 1.
3391 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3392 processing audio in time domain which is slow.
3393 @var{freq} is processing audio in frequency domain which is fast.
3394 Default is @var{freq}.
3397 Set custom positions of virtual loudspeakers. Syntax for this option is:
3398 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3399 Each virtual loudspeaker is described with short channel name following with
3400 azimuth and elevation in degreees.
3401 Each virtual loudspeaker description is separated by '|'.
3402 For example to override front left and front right channel positions use:
3403 'speakers=FL 45 15|FR 345 15'.
3404 Descriptions with unrecognised channel names are ignored.
3407 @subsection Examples
3411 Using ClubFritz6 sofa file:
3413 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3417 Using ClubFritz12 sofa file and bigger radius with small rotation:
3419 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3423 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3424 and also with custom gain:
3426 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3430 @section stereotools
3432 This filter has some handy utilities to manage stereo signals, for converting
3433 M/S stereo recordings to L/R signal while having control over the parameters
3434 or spreading the stereo image of master track.
3436 The filter accepts the following options:
3440 Set input level before filtering for both channels. Defaults is 1.
3441 Allowed range is from 0.015625 to 64.
3444 Set output level after filtering for both channels. Defaults is 1.
3445 Allowed range is from 0.015625 to 64.
3448 Set input balance between both channels. Default is 0.
3449 Allowed range is from -1 to 1.
3452 Set output balance between both channels. Default is 0.
3453 Allowed range is from -1 to 1.
3456 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3457 clipping. Disabled by default.
3460 Mute the left channel. Disabled by default.
3463 Mute the right channel. Disabled by default.
3466 Change the phase of the left channel. Disabled by default.
3469 Change the phase of the right channel. Disabled by default.
3472 Set stereo mode. Available values are:
3476 Left/Right to Left/Right, this is default.
3479 Left/Right to Mid/Side.
3482 Mid/Side to Left/Right.
3485 Left/Right to Left/Left.
3488 Left/Right to Right/Right.
3491 Left/Right to Left + Right.
3494 Left/Right to Right/Left.
3498 Set level of side signal. Default is 1.
3499 Allowed range is from 0.015625 to 64.
3502 Set balance of side signal. Default is 0.
3503 Allowed range is from -1 to 1.
3506 Set level of the middle signal. Default is 1.
3507 Allowed range is from 0.015625 to 64.
3510 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3513 Set stereo base between mono and inversed channels. Default is 0.
3514 Allowed range is from -1 to 1.
3517 Set delay in milliseconds how much to delay left from right channel and
3518 vice versa. Default is 0. Allowed range is from -20 to 20.
3521 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3524 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3527 @subsection Examples
3531 Apply karaoke like effect:
3533 stereotools=mlev=0.015625
3537 Convert M/S signal to L/R:
3539 "stereotools=mode=ms>lr"
3543 @section stereowiden
3545 This filter enhance the stereo effect by suppressing signal common to both
3546 channels and by delaying the signal of left into right and vice versa,
3547 thereby widening the stereo effect.
3549 The filter accepts the following options:
3553 Time in milliseconds of the delay of left signal into right and vice versa.
3554 Default is 20 milliseconds.
3557 Amount of gain in delayed signal into right and vice versa. Gives a delay
3558 effect of left signal in right output and vice versa which gives widening
3559 effect. Default is 0.3.
3562 Cross feed of left into right with inverted phase. This helps in suppressing
3563 the mono. If the value is 1 it will cancel all the signal common to both
3564 channels. Default is 0.3.
3567 Set level of input signal of original channel. Default is 0.8.
3572 Boost or cut treble (upper) frequencies of the audio using a two-pole
3573 shelving filter with a response similar to that of a standard
3574 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3576 The filter accepts the following options:
3580 Give the gain at whichever is the lower of ~22 kHz and the
3581 Nyquist frequency. Its useful range is about -20 (for a large cut)
3582 to +20 (for a large boost). Beware of clipping when using a positive gain.
3585 Set the filter's central frequency and so can be used
3586 to extend or reduce the frequency range to be boosted or cut.
3587 The default value is @code{3000} Hz.
3590 Set method to specify band-width of filter.
3603 Determine how steep is the filter's shelf transition.
3608 Sinusoidal amplitude modulation.
3610 The filter accepts the following options:
3614 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3615 (20 Hz or lower) will result in a tremolo effect.
3616 This filter may also be used as a ring modulator by specifying
3617 a modulation frequency higher than 20 Hz.
3618 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3621 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3622 Default value is 0.5.
3627 Sinusoidal phase modulation.
3629 The filter accepts the following options:
3633 Modulation frequency in Hertz.
3634 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3637 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3638 Default value is 0.5.
3643 Adjust the input audio volume.
3645 It accepts the following parameters:
3649 Set audio volume expression.
3651 Output values are clipped to the maximum value.
3653 The output audio volume is given by the relation:
3655 @var{output_volume} = @var{volume} * @var{input_volume}
3658 The default value for @var{volume} is "1.0".
3661 This parameter represents the mathematical precision.
3663 It determines which input sample formats will be allowed, which affects the
3664 precision of the volume scaling.
3668 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3670 32-bit floating-point; this limits input sample format to FLT. (default)
3672 64-bit floating-point; this limits input sample format to DBL.
3676 Choose the behaviour on encountering ReplayGain side data in input frames.
3680 Remove ReplayGain side data, ignoring its contents (the default).
3683 Ignore ReplayGain side data, but leave it in the frame.
3686 Prefer the track gain, if present.
3689 Prefer the album gain, if present.
3692 @item replaygain_preamp
3693 Pre-amplification gain in dB to apply to the selected replaygain gain.
3695 Default value for @var{replaygain_preamp} is 0.0.
3698 Set when the volume expression is evaluated.
3700 It accepts the following values:
3703 only evaluate expression once during the filter initialization, or
3704 when the @samp{volume} command is sent
3707 evaluate expression for each incoming frame
3710 Default value is @samp{once}.
3713 The volume expression can contain the following parameters.
3717 frame number (starting at zero)
3720 @item nb_consumed_samples
3721 number of samples consumed by the filter
3723 number of samples in the current frame
3725 original frame position in the file
3731 PTS at start of stream
3733 time at start of stream
3739 last set volume value
3742 Note that when @option{eval} is set to @samp{once} only the
3743 @var{sample_rate} and @var{tb} variables are available, all other
3744 variables will evaluate to NAN.
3746 @subsection Commands
3748 This filter supports the following commands:
3751 Modify the volume expression.
3752 The command accepts the same syntax of the corresponding option.
3754 If the specified expression is not valid, it is kept at its current
3756 @item replaygain_noclip
3757 Prevent clipping by limiting the gain applied.
3759 Default value for @var{replaygain_noclip} is 1.
3763 @subsection Examples
3767 Halve the input audio volume:
3771 volume=volume=-6.0206dB
3774 In all the above example the named key for @option{volume} can be
3775 omitted, for example like in:
3781 Increase input audio power by 6 decibels using fixed-point precision:
3783 volume=volume=6dB:precision=fixed
3787 Fade volume after time 10 with an annihilation period of 5 seconds:
3789 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3793 @section volumedetect
3795 Detect the volume of the input video.
3797 The filter has no parameters. The input is not modified. Statistics about
3798 the volume will be printed in the log when the input stream end is reached.
3800 In particular it will show the mean volume (root mean square), maximum
3801 volume (on a per-sample basis), and the beginning of a histogram of the
3802 registered volume values (from the maximum value to a cumulated 1/1000 of
3805 All volumes are in decibels relative to the maximum PCM value.
3807 @subsection Examples
3809 Here is an excerpt of the output:
3811 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3812 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3813 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3814 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3815 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3816 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3817 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3818 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3819 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3825 The mean square energy is approximately -27 dB, or 10^-2.7.
3827 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3829 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3832 In other words, raising the volume by +4 dB does not cause any clipping,
3833 raising it by +5 dB causes clipping for 6 samples, etc.
3835 @c man end AUDIO FILTERS
3837 @chapter Audio Sources
3838 @c man begin AUDIO SOURCES
3840 Below is a description of the currently available audio sources.
3844 Buffer audio frames, and make them available to the filter chain.
3846 This source is mainly intended for a programmatic use, in particular
3847 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3849 It accepts the following parameters:
3853 The timebase which will be used for timestamps of submitted frames. It must be
3854 either a floating-point number or in @var{numerator}/@var{denominator} form.
3857 The sample rate of the incoming audio buffers.
3860 The sample format of the incoming audio buffers.
3861 Either a sample format name or its corresponding integer representation from
3862 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3864 @item channel_layout
3865 The channel layout of the incoming audio buffers.
3866 Either a channel layout name from channel_layout_map in
3867 @file{libavutil/channel_layout.c} or its corresponding integer representation
3868 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3871 The number of channels of the incoming audio buffers.
3872 If both @var{channels} and @var{channel_layout} are specified, then they
3877 @subsection Examples
3880 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3883 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3884 Since the sample format with name "s16p" corresponds to the number
3885 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3888 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3893 Generate an audio signal specified by an expression.
3895 This source accepts in input one or more expressions (one for each
3896 channel), which are evaluated and used to generate a corresponding
3899 This source accepts the following options:
3903 Set the '|'-separated expressions list for each separate channel. In case the
3904 @option{channel_layout} option is not specified, the selected channel layout
3905 depends on the number of provided expressions. Otherwise the last
3906 specified expression is applied to the remaining output channels.
3908 @item channel_layout, c
3909 Set the channel layout. The number of channels in the specified layout
3910 must be equal to the number of specified expressions.
3913 Set the minimum duration of the sourced audio. See
3914 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3915 for the accepted syntax.
3916 Note that the resulting duration may be greater than the specified
3917 duration, as the generated audio is always cut at the end of a
3920 If not specified, or the expressed duration is negative, the audio is
3921 supposed to be generated forever.
3924 Set the number of samples per channel per each output frame,
3927 @item sample_rate, s
3928 Specify the sample rate, default to 44100.
3931 Each expression in @var{exprs} can contain the following constants:
3935 number of the evaluated sample, starting from 0
3938 time of the evaluated sample expressed in seconds, starting from 0
3945 @subsection Examples
3955 Generate a sin signal with frequency of 440 Hz, set sample rate to
3958 aevalsrc="sin(440*2*PI*t):s=8000"
3962 Generate a two channels signal, specify the channel layout (Front
3963 Center + Back Center) explicitly:
3965 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
3969 Generate white noise:
3971 aevalsrc="-2+random(0)"
3975 Generate an amplitude modulated signal:
3977 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
3981 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
3983 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
3990 The null audio source, return unprocessed audio frames. It is mainly useful
3991 as a template and to be employed in analysis / debugging tools, or as
3992 the source for filters which ignore the input data (for example the sox
3995 This source accepts the following options:
3999 @item channel_layout, cl
4001 Specifies the channel layout, and can be either an integer or a string
4002 representing a channel layout. The default value of @var{channel_layout}
4005 Check the channel_layout_map definition in
4006 @file{libavutil/channel_layout.c} for the mapping between strings and
4007 channel layout values.
4009 @item sample_rate, r
4010 Specifies the sample rate, and defaults to 44100.
4013 Set the number of samples per requested frames.
4017 @subsection Examples
4021 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4023 anullsrc=r=48000:cl=4
4027 Do the same operation with a more obvious syntax:
4029 anullsrc=r=48000:cl=mono
4033 All the parameters need to be explicitly defined.
4037 Synthesize a voice utterance using the libflite library.
4039 To enable compilation of this filter you need to configure FFmpeg with
4040 @code{--enable-libflite}.
4042 Note that the flite library is not thread-safe.
4044 The filter accepts the following options:
4049 If set to 1, list the names of the available voices and exit
4050 immediately. Default value is 0.
4053 Set the maximum number of samples per frame. Default value is 512.
4056 Set the filename containing the text to speak.
4059 Set the text to speak.
4062 Set the voice to use for the speech synthesis. Default value is
4063 @code{kal}. See also the @var{list_voices} option.
4066 @subsection Examples
4070 Read from file @file{speech.txt}, and synthesize the text using the
4071 standard flite voice:
4073 flite=textfile=speech.txt
4077 Read the specified text selecting the @code{slt} voice:
4079 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4083 Input text to ffmpeg:
4085 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4089 Make @file{ffplay} speak the specified text, using @code{flite} and
4090 the @code{lavfi} device:
4092 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4096 For more information about libflite, check:
4097 @url{http://www.speech.cs.cmu.edu/flite/}
4101 Generate a noise audio signal.
4103 The filter accepts the following options:
4106 @item sample_rate, r
4107 Specify the sample rate. Default value is 48000 Hz.
4110 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4114 Specify the duration of the generated audio stream. Not specifying this option
4115 results in noise with an infinite length.
4117 @item color, colour, c
4118 Specify the color of noise. Available noise colors are white, pink, and brown.
4119 Default color is white.
4122 Specify a value used to seed the PRNG.
4125 Set the number of samples per each output frame, default is 1024.
4128 @subsection Examples
4133 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4135 anoisesrc=d=60:c=pink:r=44100:a=0.5
4141 Generate an audio signal made of a sine wave with amplitude 1/8.
4143 The audio signal is bit-exact.
4145 The filter accepts the following options:
4150 Set the carrier frequency. Default is 440 Hz.
4152 @item beep_factor, b
4153 Enable a periodic beep every second with frequency @var{beep_factor} times
4154 the carrier frequency. Default is 0, meaning the beep is disabled.
4156 @item sample_rate, r
4157 Specify the sample rate, default is 44100.
4160 Specify the duration of the generated audio stream.
4162 @item samples_per_frame
4163 Set the number of samples per output frame.
4165 The expression can contain the following constants:
4169 The (sequential) number of the output audio frame, starting from 0.
4172 The PTS (Presentation TimeStamp) of the output audio frame,
4173 expressed in @var{TB} units.
4176 The PTS of the output audio frame, expressed in seconds.
4179 The timebase of the output audio frames.
4182 Default is @code{1024}.
4185 @subsection Examples
4190 Generate a simple 440 Hz sine wave:
4196 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4200 sine=frequency=220:beep_factor=4:duration=5
4204 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4207 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4211 @c man end AUDIO SOURCES
4213 @chapter Audio Sinks
4214 @c man begin AUDIO SINKS
4216 Below is a description of the currently available audio sinks.
4218 @section abuffersink
4220 Buffer audio frames, and make them available to the end of filter chain.
4222 This sink is mainly intended for programmatic use, in particular
4223 through the interface defined in @file{libavfilter/buffersink.h}
4224 or the options system.
4226 It accepts a pointer to an AVABufferSinkContext structure, which
4227 defines the incoming buffers' formats, to be passed as the opaque
4228 parameter to @code{avfilter_init_filter} for initialization.
4231 Null audio sink; do absolutely nothing with the input audio. It is
4232 mainly useful as a template and for use in analysis / debugging
4235 @c man end AUDIO SINKS
4237 @chapter Video Filters
4238 @c man begin VIDEO FILTERS
4240 When you configure your FFmpeg build, you can disable any of the
4241 existing filters using @code{--disable-filters}.
4242 The configure output will show the video filters included in your
4245 Below is a description of the currently available video filters.
4247 @section alphaextract
4249 Extract the alpha component from the input as a grayscale video. This
4250 is especially useful with the @var{alphamerge} filter.
4254 Add or replace the alpha component of the primary input with the
4255 grayscale value of a second input. This is intended for use with
4256 @var{alphaextract} to allow the transmission or storage of frame
4257 sequences that have alpha in a format that doesn't support an alpha
4260 For example, to reconstruct full frames from a normal YUV-encoded video
4261 and a separate video created with @var{alphaextract}, you might use:
4263 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4266 Since this filter is designed for reconstruction, it operates on frame
4267 sequences without considering timestamps, and terminates when either
4268 input reaches end of stream. This will cause problems if your encoding
4269 pipeline drops frames. If you're trying to apply an image as an
4270 overlay to a video stream, consider the @var{overlay} filter instead.
4274 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4275 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4276 Substation Alpha) subtitles files.
4278 This filter accepts the following option in addition to the common options from
4279 the @ref{subtitles} filter:
4283 Set the shaping engine
4285 Available values are:
4288 The default libass shaping engine, which is the best available.
4290 Fast, font-agnostic shaper that can do only substitutions
4292 Slower shaper using OpenType for substitutions and positioning
4295 The default is @code{auto}.
4299 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4301 The filter accepts the following options:
4305 Set threshold A for 1st plane. Default is 0.02.
4306 Valid range is 0 to 0.3.
4309 Set threshold B for 1st plane. Default is 0.04.
4310 Valid range is 0 to 5.
4313 Set threshold A for 2nd plane. Default is 0.02.
4314 Valid range is 0 to 0.3.
4317 Set threshold B for 2nd plane. Default is 0.04.
4318 Valid range is 0 to 5.
4321 Set threshold A for 3rd plane. Default is 0.02.
4322 Valid range is 0 to 0.3.
4325 Set threshold B for 3rd plane. Default is 0.04.
4326 Valid range is 0 to 5.
4328 Threshold A is designed to react on abrupt changes in the input signal and
4329 threshold B is designed to react on continuous changes in the input signal.
4332 Set number of frames filter will use for averaging. Default is 33. Must be odd
4333 number in range [5, 129].
4338 Compute the bounding box for the non-black pixels in the input frame
4341 This filter computes the bounding box containing all the pixels with a
4342 luminance value greater than the minimum allowed value.
4343 The parameters describing the bounding box are printed on the filter
4346 The filter accepts the following option:
4350 Set the minimal luminance value. Default is @code{16}.
4353 @section blackdetect
4355 Detect video intervals that are (almost) completely black. Can be
4356 useful to detect chapter transitions, commercials, or invalid
4357 recordings. Output lines contains the time for the start, end and
4358 duration of the detected black interval expressed in seconds.
4360 In order to display the output lines, you need to set the loglevel at
4361 least to the AV_LOG_INFO value.
4363 The filter accepts the following options:
4366 @item black_min_duration, d
4367 Set the minimum detected black duration expressed in seconds. It must
4368 be a non-negative floating point number.
4370 Default value is 2.0.
4372 @item picture_black_ratio_th, pic_th
4373 Set the threshold for considering a picture "black".
4374 Express the minimum value for the ratio:
4376 @var{nb_black_pixels} / @var{nb_pixels}
4379 for which a picture is considered black.
4380 Default value is 0.98.
4382 @item pixel_black_th, pix_th
4383 Set the threshold for considering a pixel "black".
4385 The threshold expresses the maximum pixel luminance value for which a
4386 pixel is considered "black". The provided value is scaled according to
4387 the following equation:
4389 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4392 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4393 the input video format, the range is [0-255] for YUV full-range
4394 formats and [16-235] for YUV non full-range formats.
4396 Default value is 0.10.
4399 The following example sets the maximum pixel threshold to the minimum
4400 value, and detects only black intervals of 2 or more seconds:
4402 blackdetect=d=2:pix_th=0.00
4407 Detect frames that are (almost) completely black. Can be useful to
4408 detect chapter transitions or commercials. Output lines consist of
4409 the frame number of the detected frame, the percentage of blackness,
4410 the position in the file if known or -1 and the timestamp in seconds.
4412 In order to display the output lines, you need to set the loglevel at
4413 least to the AV_LOG_INFO value.
4415 It accepts the following parameters:
4420 The percentage of the pixels that have to be below the threshold; it defaults to
4423 @item threshold, thresh
4424 The threshold below which a pixel value is considered black; it defaults to
4429 @section blend, tblend
4431 Blend two video frames into each other.
4433 The @code{blend} filter takes two input streams and outputs one
4434 stream, the first input is the "top" layer and second input is
4435 "bottom" layer. Output terminates when shortest input terminates.
4437 The @code{tblend} (time blend) filter takes two consecutive frames
4438 from one single stream, and outputs the result obtained by blending
4439 the new frame on top of the old frame.
4441 A description of the accepted options follows.
4449 Set blend mode for specific pixel component or all pixel components in case
4450 of @var{all_mode}. Default value is @code{normal}.
4452 Available values for component modes are:
4493 Set blend opacity for specific pixel component or all pixel components in case
4494 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4501 Set blend expression for specific pixel component or all pixel components in case
4502 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4504 The expressions can use the following variables:
4508 The sequential number of the filtered frame, starting from @code{0}.
4512 the coordinates of the current sample
4516 the width and height of currently filtered plane
4520 Width and height scale depending on the currently filtered plane. It is the
4521 ratio between the corresponding luma plane number of pixels and the current
4522 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4523 @code{0.5,0.5} for chroma planes.
4526 Time of the current frame, expressed in seconds.
4529 Value of pixel component at current location for first video frame (top layer).
4532 Value of pixel component at current location for second video frame (bottom layer).
4536 Force termination when the shortest input terminates. Default is
4537 @code{0}. This option is only defined for the @code{blend} filter.
4540 Continue applying the last bottom frame after the end of the stream. A value of
4541 @code{0} disable the filter after the last frame of the bottom layer is reached.
4542 Default is @code{1}. This option is only defined for the @code{blend} filter.
4545 @subsection Examples
4549 Apply transition from bottom layer to top layer in first 10 seconds:
4551 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4555 Apply 1x1 checkerboard effect:
4557 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4561 Apply uncover left effect:
4563 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4567 Apply uncover down effect:
4569 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4573 Apply uncover up-left effect:
4575 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4579 Split diagonally video and shows top and bottom layer on each side:
4581 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4585 Display differences between the current and the previous frame:
4587 tblend=all_mode=difference128
4593 Apply a boxblur algorithm to the input video.
4595 It accepts the following parameters:
4599 @item luma_radius, lr
4600 @item luma_power, lp
4601 @item chroma_radius, cr
4602 @item chroma_power, cp
4603 @item alpha_radius, ar
4604 @item alpha_power, ap
4608 A description of the accepted options follows.
4611 @item luma_radius, lr
4612 @item chroma_radius, cr
4613 @item alpha_radius, ar
4614 Set an expression for the box radius in pixels used for blurring the
4615 corresponding input plane.
4617 The radius value must be a non-negative number, and must not be
4618 greater than the value of the expression @code{min(w,h)/2} for the
4619 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4622 Default value for @option{luma_radius} is "2". If not specified,
4623 @option{chroma_radius} and @option{alpha_radius} default to the
4624 corresponding value set for @option{luma_radius}.
4626 The expressions can contain the following constants:
4630 The input width and height in pixels.
4634 The input chroma image width and height in pixels.
4638 The horizontal and vertical chroma subsample values. For example, for the
4639 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4642 @item luma_power, lp
4643 @item chroma_power, cp
4644 @item alpha_power, ap
4645 Specify how many times the boxblur filter is applied to the
4646 corresponding plane.
4648 Default value for @option{luma_power} is 2. If not specified,
4649 @option{chroma_power} and @option{alpha_power} default to the
4650 corresponding value set for @option{luma_power}.
4652 A value of 0 will disable the effect.
4655 @subsection Examples
4659 Apply a boxblur filter with the luma, chroma, and alpha radii
4662 boxblur=luma_radius=2:luma_power=1
4667 Set the luma radius to 2, and alpha and chroma radius to 0:
4669 boxblur=2:1:cr=0:ar=0
4673 Set the luma and chroma radii to a fraction of the video dimension:
4675 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4681 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4682 Deinterlacing Filter").
4684 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4685 interpolation algorithms.
4686 It accepts the following parameters:
4690 The interlacing mode to adopt. It accepts one of the following values:
4694 Output one frame for each frame.
4696 Output one frame for each field.
4699 The default value is @code{send_field}.
4702 The picture field parity assumed for the input interlaced video. It accepts one
4703 of the following values:
4707 Assume the top field is first.
4709 Assume the bottom field is first.
4711 Enable automatic detection of field parity.
4714 The default value is @code{auto}.
4715 If the interlacing is unknown or the decoder does not export this information,
4716 top field first will be assumed.
4719 Specify which frames to deinterlace. Accept one of the following
4724 Deinterlace all frames.
4726 Only deinterlace frames marked as interlaced.
4729 The default value is @code{all}.
4733 YUV colorspace color/chroma keying.
4735 The filter accepts the following options:
4739 The color which will be replaced with transparency.
4742 Similarity percentage with the key color.
4744 0.01 matches only the exact key color, while 1.0 matches everything.
4749 0.0 makes pixels either fully transparent, or not transparent at all.
4751 Higher values result in semi-transparent pixels, with a higher transparency
4752 the more similar the pixels color is to the key color.
4755 Signals that the color passed is already in YUV instead of RGB.
4757 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4758 This can be used to pass exact YUV values as hexadecimal numbers.
4761 @subsection Examples
4765 Make every green pixel in the input image transparent:
4767 ffmpeg -i input.png -vf chromakey=green out.png
4771 Overlay a greenscreen-video on top of a static black background.
4773 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
4779 Display CIE color diagram with pixels overlaid onto it.
4781 The filter accepts the following options:
4796 @item uhdtv, rec2020
4809 Set what gamuts to draw.
4811 See @code{system} option for available values.
4814 Set ciescope size, by default set to 512.
4817 Set intensity used to map input pixel values to CIE diagram.
4820 Set contrast used to draw tongue colors that are out of active color system gamut.
4823 Correct gamma displayed on scope, by default enabled.
4826 Show white point on CIE diagram, by default disabled.
4829 Set input gamma. Used only with XYZ input color space.
4834 Visualize information exported by some codecs.
4836 Some codecs can export information through frames using side-data or other
4837 means. For example, some MPEG based codecs export motion vectors through the
4838 @var{export_mvs} flag in the codec @option{flags2} option.
4840 The filter accepts the following option:
4844 Set motion vectors to visualize.
4846 Available flags for @var{mv} are:
4850 forward predicted MVs of P-frames
4852 forward predicted MVs of B-frames
4854 backward predicted MVs of B-frames
4858 Display quantization parameters using the chroma planes.
4861 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4863 Available flags for @var{mv_type} are:
4867 forward predicted MVs
4869 backward predicted MVs
4872 @item frame_type, ft
4873 Set frame type to visualize motion vectors of.
4875 Available flags for @var{frame_type} are:
4879 intra-coded frames (I-frames)
4881 predicted frames (P-frames)
4883 bi-directionally predicted frames (B-frames)
4887 @subsection Examples
4891 Visualize forward predicted MVs of all frames using @command{ffplay}:
4893 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
4897 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
4899 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
4903 @section colorbalance
4904 Modify intensity of primary colors (red, green and blue) of input frames.
4906 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
4907 regions for the red-cyan, green-magenta or blue-yellow balance.
4909 A positive adjustment value shifts the balance towards the primary color, a negative
4910 value towards the complementary color.
4912 The filter accepts the following options:
4918 Adjust red, green and blue shadows (darkest pixels).
4923 Adjust red, green and blue midtones (medium pixels).
4928 Adjust red, green and blue highlights (brightest pixels).
4930 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
4933 @subsection Examples
4937 Add red color cast to shadows:
4944 RGB colorspace color keying.
4946 The filter accepts the following options:
4950 The color which will be replaced with transparency.
4953 Similarity percentage with the key color.
4955 0.01 matches only the exact key color, while 1.0 matches everything.
4960 0.0 makes pixels either fully transparent, or not transparent at all.
4962 Higher values result in semi-transparent pixels, with a higher transparency
4963 the more similar the pixels color is to the key color.
4966 @subsection Examples
4970 Make every green pixel in the input image transparent:
4972 ffmpeg -i input.png -vf colorkey=green out.png
4976 Overlay a greenscreen-video on top of a static background image.
4978 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
4982 @section colorlevels
4984 Adjust video input frames using levels.
4986 The filter accepts the following options:
4993 Adjust red, green, blue and alpha input black point.
4994 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5000 Adjust red, green, blue and alpha input white point.
5001 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5003 Input levels are used to lighten highlights (bright tones), darken shadows
5004 (dark tones), change the balance of bright and dark tones.
5010 Adjust red, green, blue and alpha output black point.
5011 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5017 Adjust red, green, blue and alpha output white point.
5018 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5020 Output levels allows manual selection of a constrained output level range.
5023 @subsection Examples
5027 Make video output darker:
5029 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5035 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5039 Make video output lighter:
5041 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5045 Increase brightness:
5047 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5051 @section colorchannelmixer
5053 Adjust video input frames by re-mixing color channels.
5055 This filter modifies a color channel by adding the values associated to
5056 the other channels of the same pixels. For example if the value to
5057 modify is red, the output value will be:
5059 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5062 The filter accepts the following options:
5069 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5070 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5076 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5077 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5083 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5084 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5090 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5091 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5093 Allowed ranges for options are @code{[-2.0, 2.0]}.
5096 @subsection Examples
5100 Convert source to grayscale:
5102 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5105 Simulate sepia tones:
5107 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5111 @section colormatrix
5113 Convert color matrix.
5115 The filter accepts the following options:
5120 Specify the source and destination color matrix. Both values must be
5123 The accepted values are:
5142 For example to convert from BT.601 to SMPTE-240M, use the command:
5144 colormatrix=bt601:smpte240m
5149 Convert colorspace, transfer characteristics or color primaries.
5151 The filter accepts the following options:
5155 Specify all color properties at once.
5157 The accepted values are:
5186 Specify output colorspace.
5188 The accepted values are:
5197 BT.470BG or BT.601-6 625
5200 SMPTE-170M or BT.601-6 525
5206 BT.2020 with non-constant luminance
5211 Specify output transfer characteristics.
5213 The accepted values are:
5219 Constant gamma of 2.2
5222 Constant gamma of 2.8
5225 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5231 BT.2020 for 10-bits content
5234 BT.2020 for 12-bits content
5239 Specify output color primaries.
5241 The accepted values are:
5250 BT.470BG or BT.601-6 625
5253 SMPTE-170M or BT.601-6 525
5264 Specify output color range.
5266 The accepted values are:
5269 MPEG (restricted) range
5277 Specify output color format.
5279 The accepted values are:
5282 YUV 4:2:0 planar 8-bits
5285 YUV 4:2:0 planar 10-bits
5288 YUV 4:2:0 planar 12-bits
5291 YUV 4:2:2 planar 8-bits
5294 YUV 4:2:2 planar 10-bits
5297 YUV 4:2:2 planar 12-bits
5300 YUV 4:4:4 planar 8-bits
5303 YUV 4:4:4 planar 10-bits
5306 YUV 4:4:4 planar 12-bits
5311 Do a fast conversion, which skips gamma/primary correction. This will take
5312 significantly less CPU, but will be mathematically incorrect. To get output
5313 compatible with that produced by the colormatrix filter, use fast=1.
5316 Specify dithering mode.
5318 The accepted values are:
5324 Floyd-Steinberg dithering
5328 Whitepoint adaptation mode.
5330 The accepted values are:
5333 Bradford whitepoint adaptation
5336 von Kries whitepoint adaptation
5339 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5344 The filter converts the transfer characteristics, color space and color
5345 primaries to the specified user values. The output value, if not specified,
5346 is set to a default value based on the "all" property. If that property is
5347 also not specified, the filter will log an error. The output color range and
5348 format default to the same value as the input color range and format. The
5349 input transfer characteristics, color space, color primaries and color range
5350 should be set on the input data. If any of these are missing, the filter will
5351 log an error and no conversion will take place.
5353 For example to convert the input to SMPTE-240M, use the command:
5355 colorspace=smpte240m
5358 @section convolution
5360 Apply convolution 3x3 or 5x5 filter.
5362 The filter accepts the following options:
5369 Set matrix for each plane.
5370 Matrix is sequence of 9 or 25 signed integers.
5376 Set multiplier for calculated value for each plane.
5382 Set bias for each plane. This value is added to the result of the multiplication.
5383 Useful for making the overall image brighter or darker. Default is 0.0.
5386 @subsection Examples
5392 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"
5398 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"
5404 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"
5410 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"
5416 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"
5422 Copy the input source unchanged to the output. This is mainly useful for
5427 Video filtering on GPU using Apple's CoreImage API on OSX.
5429 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5430 processed by video hardware. However, software-based OpenGL implementations
5431 exist which means there is no guarantee for hardware processing. It depends on
5434 There are many filters and image generators provided by Apple that come with a
5435 large variety of options. The filter has to be referenced by its name along
5438 The coreimage filter accepts the following options:
5441 List all available filters and generators along with all their respective
5442 options as well as possible minimum and maximum values along with the default
5449 Specify all filters by their respective name and options.
5450 Use @var{list_filters} to determine all valid filter names and options.
5451 Numerical options are specified by a float value and are automatically clamped
5452 to their respective value range. Vector and color options have to be specified
5453 by a list of space separated float values. Character escaping has to be done.
5454 A special option name @code{default} is available to use default options for a
5457 It is required to specify either @code{default} or at least one of the filter options.
5458 All omitted options are used with their default values.
5459 The syntax of the filter string is as follows:
5461 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5465 Specify a rectangle where the output of the filter chain is copied into the
5466 input image. It is given by a list of space separated float values:
5468 output_rect=x\ y\ width\ height
5470 If not given, the output rectangle equals the dimensions of the input image.
5471 The output rectangle is automatically cropped at the borders of the input
5472 image. Negative values are valid for each component.
5474 output_rect=25\ 25\ 100\ 100
5478 Several filters can be chained for successive processing without GPU-HOST
5479 transfers allowing for fast processing of complex filter chains.
5480 Currently, only filters with zero (generators) or exactly one (filters) input
5481 image and one output image are supported. Also, transition filters are not yet
5484 Some filters generate output images with additional padding depending on the
5485 respective filter kernel. The padding is automatically removed to ensure the
5486 filter output has the same size as the input image.
5488 For image generators, the size of the output image is determined by the
5489 previous output image of the filter chain or the input image of the whole
5490 filterchain, respectively. The generators do not use the pixel information of
5491 this image to generate their output. However, the generated output is
5492 blended onto this image, resulting in partial or complete coverage of the
5495 The @ref{coreimagesrc} video source can be used for generating input images
5496 which are directly fed into the filter chain. By using it, providing input
5497 images by another video source or an input video is not required.
5499 @subsection Examples
5504 List all filters available:
5506 coreimage=list_filters=true
5510 Use the CIBoxBlur filter with default options to blur an image:
5512 coreimage=filter=CIBoxBlur@@default
5516 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5517 its center at 100x100 and a radius of 50 pixels:
5519 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5523 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5524 given as complete and escaped command-line for Apple's standard bash shell:
5526 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5532 Crop the input video to given dimensions.
5534 It accepts the following parameters:
5538 The width of the output video. It defaults to @code{iw}.
5539 This expression is evaluated only once during the filter
5540 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5543 The height of the output video. It defaults to @code{ih}.
5544 This expression is evaluated only once during the filter
5545 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5548 The horizontal position, in the input video, of the left edge of the output
5549 video. It defaults to @code{(in_w-out_w)/2}.
5550 This expression is evaluated per-frame.
5553 The vertical position, in the input video, of the top edge of the output video.
5554 It defaults to @code{(in_h-out_h)/2}.
5555 This expression is evaluated per-frame.
5558 If set to 1 will force the output display aspect ratio
5559 to be the same of the input, by changing the output sample aspect
5560 ratio. It defaults to 0.
5563 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5564 expressions containing the following constants:
5569 The computed values for @var{x} and @var{y}. They are evaluated for
5574 The input width and height.
5578 These are the same as @var{in_w} and @var{in_h}.
5582 The output (cropped) width and height.
5586 These are the same as @var{out_w} and @var{out_h}.
5589 same as @var{iw} / @var{ih}
5592 input sample aspect ratio
5595 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5599 horizontal and vertical chroma subsample values. For example for the
5600 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5603 The number of the input frame, starting from 0.
5606 the position in the file of the input frame, NAN if unknown
5609 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5613 The expression for @var{out_w} may depend on the value of @var{out_h},
5614 and the expression for @var{out_h} may depend on @var{out_w}, but they
5615 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5616 evaluated after @var{out_w} and @var{out_h}.
5618 The @var{x} and @var{y} parameters specify the expressions for the
5619 position of the top-left corner of the output (non-cropped) area. They
5620 are evaluated for each frame. If the evaluated value is not valid, it
5621 is approximated to the nearest valid value.
5623 The expression for @var{x} may depend on @var{y}, and the expression
5624 for @var{y} may depend on @var{x}.
5626 @subsection Examples
5630 Crop area with size 100x100 at position (12,34).
5635 Using named options, the example above becomes:
5637 crop=w=100:h=100:x=12:y=34
5641 Crop the central input area with size 100x100:
5647 Crop the central input area with size 2/3 of the input video:
5649 crop=2/3*in_w:2/3*in_h
5653 Crop the input video central square:
5660 Delimit the rectangle with the top-left corner placed at position
5661 100:100 and the right-bottom corner corresponding to the right-bottom
5662 corner of the input image.
5664 crop=in_w-100:in_h-100:100:100
5668 Crop 10 pixels from the left and right borders, and 20 pixels from
5669 the top and bottom borders
5671 crop=in_w-2*10:in_h-2*20
5675 Keep only the bottom right quarter of the input image:
5677 crop=in_w/2:in_h/2:in_w/2:in_h/2
5681 Crop height for getting Greek harmony:
5683 crop=in_w:1/PHI*in_w
5687 Apply trembling effect:
5689 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)
5693 Apply erratic camera effect depending on timestamp:
5695 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)"
5699 Set x depending on the value of y:
5701 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5705 @subsection Commands
5707 This filter supports the following commands:
5713 Set width/height of the output video and the horizontal/vertical position
5715 The command accepts the same syntax of the corresponding option.
5717 If the specified expression is not valid, it is kept at its current
5723 Auto-detect the crop size.
5725 It calculates the necessary cropping parameters and prints the
5726 recommended parameters via the logging system. The detected dimensions
5727 correspond to the non-black area of the input video.
5729 It accepts the following parameters:
5734 Set higher black value threshold, which can be optionally specified
5735 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5736 value greater to the set value is considered non-black. It defaults to 24.
5737 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5738 on the bitdepth of the pixel format.
5741 The value which the width/height should be divisible by. It defaults to
5742 16. The offset is automatically adjusted to center the video. Use 2 to
5743 get only even dimensions (needed for 4:2:2 video). 16 is best when
5744 encoding to most video codecs.
5746 @item reset_count, reset
5747 Set the counter that determines after how many frames cropdetect will
5748 reset the previously detected largest video area and start over to
5749 detect the current optimal crop area. Default value is 0.
5751 This can be useful when channel logos distort the video area. 0
5752 indicates 'never reset', and returns the largest area encountered during
5759 Apply color adjustments using curves.
5761 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5762 component (red, green and blue) has its values defined by @var{N} key points
5763 tied from each other using a smooth curve. The x-axis represents the pixel
5764 values from the input frame, and the y-axis the new pixel values to be set for
5767 By default, a component curve is defined by the two points @var{(0;0)} and
5768 @var{(1;1)}. This creates a straight line where each original pixel value is
5769 "adjusted" to its own value, which means no change to the image.
5771 The filter allows you to redefine these two points and add some more. A new
5772 curve (using a natural cubic spline interpolation) will be define to pass
5773 smoothly through all these new coordinates. The new defined points needs to be
5774 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5775 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5776 the vector spaces, the values will be clipped accordingly.
5778 The filter accepts the following options:
5782 Select one of the available color presets. This option can be used in addition
5783 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5784 options takes priority on the preset values.
5785 Available presets are:
5788 @item color_negative
5791 @item increase_contrast
5793 @item linear_contrast
5794 @item medium_contrast
5796 @item strong_contrast
5799 Default is @code{none}.
5801 Set the master key points. These points will define a second pass mapping. It
5802 is sometimes called a "luminance" or "value" mapping. It can be used with
5803 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5804 post-processing LUT.
5806 Set the key points for the red component.
5808 Set the key points for the green component.
5810 Set the key points for the blue component.
5812 Set the key points for all components (not including master).
5813 Can be used in addition to the other key points component
5814 options. In this case, the unset component(s) will fallback on this
5815 @option{all} setting.
5817 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5819 Save Gnuplot script of the curves in specified file.
5822 To avoid some filtergraph syntax conflicts, each key points list need to be
5823 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5825 @subsection Examples
5829 Increase slightly the middle level of blue:
5831 curves=blue='0/0 0.5/0.58 1/1'
5837 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'
5839 Here we obtain the following coordinates for each components:
5842 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5844 @code{(0;0) (0.50;0.48) (1;1)}
5846 @code{(0;0.22) (0.49;0.44) (1;0.80)}
5850 The previous example can also be achieved with the associated built-in preset:
5852 curves=preset=vintage
5862 Use a Photoshop preset and redefine the points of the green component:
5864 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
5868 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
5869 and @command{gnuplot}:
5871 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
5872 gnuplot -p /tmp/curves.plt
5878 Video data analysis filter.
5880 This filter shows hexadecimal pixel values of part of video.
5882 The filter accepts the following options:
5886 Set output video size.
5889 Set x offset from where to pick pixels.
5892 Set y offset from where to pick pixels.
5895 Set scope mode, can be one of the following:
5898 Draw hexadecimal pixel values with white color on black background.
5901 Draw hexadecimal pixel values with input video pixel color on black
5905 Draw hexadecimal pixel values on color background picked from input video,
5906 the text color is picked in such way so its always visible.
5910 Draw rows and columns numbers on left and top of video.
5915 Denoise frames using 2D DCT (frequency domain filtering).
5917 This filter is not designed for real time.
5919 The filter accepts the following options:
5923 Set the noise sigma constant.
5925 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
5926 coefficient (absolute value) below this threshold with be dropped.
5928 If you need a more advanced filtering, see @option{expr}.
5930 Default is @code{0}.
5933 Set number overlapping pixels for each block. Since the filter can be slow, you
5934 may want to reduce this value, at the cost of a less effective filter and the
5935 risk of various artefacts.
5937 If the overlapping value doesn't permit processing the whole input width or
5938 height, a warning will be displayed and according borders won't be denoised.
5940 Default value is @var{blocksize}-1, which is the best possible setting.
5943 Set the coefficient factor expression.
5945 For each coefficient of a DCT block, this expression will be evaluated as a
5946 multiplier value for the coefficient.
5948 If this is option is set, the @option{sigma} option will be ignored.
5950 The absolute value of the coefficient can be accessed through the @var{c}
5954 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
5955 @var{blocksize}, which is the width and height of the processed blocks.
5957 The default value is @var{3} (8x8) and can be raised to @var{4} for a
5958 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
5959 on the speed processing. Also, a larger block size does not necessarily means a
5963 @subsection Examples
5965 Apply a denoise with a @option{sigma} of @code{4.5}:
5970 The same operation can be achieved using the expression system:
5972 dctdnoiz=e='gte(c, 4.5*3)'
5975 Violent denoise using a block size of @code{16x16}:
5982 Remove banding artifacts from input video.
5983 It works by replacing banded pixels with average value of referenced pixels.
5985 The filter accepts the following options:
5992 Set banding detection threshold for each plane. Default is 0.02.
5993 Valid range is 0.00003 to 0.5.
5994 If difference between current pixel and reference pixel is less than threshold,
5995 it will be considered as banded.
5998 Banding detection range in pixels. Default is 16. If positive, random number
5999 in range 0 to set value will be used. If negative, exact absolute value
6001 The range defines square of four pixels around current pixel.
6004 Set direction in radians from which four pixel will be compared. If positive,
6005 random direction from 0 to set direction will be picked. If negative, exact of
6006 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6007 will pick only pixels on same row and -PI/2 will pick only pixels on same
6011 If enabled, current pixel is compared with average value of all four
6012 surrounding pixels. The default is enabled. If disabled current pixel is
6013 compared with all four surrounding pixels. The pixel is considered banded
6014 if only all four differences with surrounding pixels are less than threshold.
6020 Drop duplicated frames at regular intervals.
6022 The filter accepts the following options:
6026 Set the number of frames from which one will be dropped. Setting this to
6027 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6028 Default is @code{5}.
6031 Set the threshold for duplicate detection. If the difference metric for a frame
6032 is less than or equal to this value, then it is declared as duplicate. Default
6036 Set scene change threshold. Default is @code{15}.
6040 Set the size of the x and y-axis blocks used during metric calculations.
6041 Larger blocks give better noise suppression, but also give worse detection of
6042 small movements. Must be a power of two. Default is @code{32}.
6045 Mark main input as a pre-processed input and activate clean source input
6046 stream. This allows the input to be pre-processed with various filters to help
6047 the metrics calculation while keeping the frame selection lossless. When set to
6048 @code{1}, the first stream is for the pre-processed input, and the second
6049 stream is the clean source from where the kept frames are chosen. Default is
6053 Set whether or not chroma is considered in the metric calculations. Default is
6059 Apply deflate effect to the video.
6061 This filter replaces the pixel by the local(3x3) average by taking into account
6062 only values lower than the pixel.
6064 It accepts the following options:
6071 Limit the maximum change for each plane, default is 65535.
6072 If 0, plane will remain unchanged.
6077 Remove judder produced by partially interlaced telecined content.
6079 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6080 source was partially telecined content then the output of @code{pullup,dejudder}
6081 will have a variable frame rate. May change the recorded frame rate of the
6082 container. Aside from that change, this filter will not affect constant frame
6085 The option available in this filter is:
6089 Specify the length of the window over which the judder repeats.
6091 Accepts any integer greater than 1. Useful values are:
6095 If the original was telecined from 24 to 30 fps (Film to NTSC).
6098 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6101 If a mixture of the two.
6104 The default is @samp{4}.
6109 Suppress a TV station logo by a simple interpolation of the surrounding
6110 pixels. Just set a rectangle covering the logo and watch it disappear
6111 (and sometimes something even uglier appear - your mileage may vary).
6113 It accepts the following parameters:
6118 Specify the top left corner coordinates of the logo. They must be
6123 Specify the width and height of the logo to clear. They must be
6127 Specify the thickness of the fuzzy edge of the rectangle (added to
6128 @var{w} and @var{h}). The default value is 1. This option is
6129 deprecated, setting higher values should no longer be necessary and
6133 When set to 1, a green rectangle is drawn on the screen to simplify
6134 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6135 The default value is 0.
6137 The rectangle is drawn on the outermost pixels which will be (partly)
6138 replaced with interpolated values. The values of the next pixels
6139 immediately outside this rectangle in each direction will be used to
6140 compute the interpolated pixel values inside the rectangle.
6144 @subsection Examples
6148 Set a rectangle covering the area with top left corner coordinates 0,0
6149 and size 100x77, and a band of size 10:
6151 delogo=x=0:y=0:w=100:h=77:band=10
6158 Attempt to fix small changes in horizontal and/or vertical shift. This
6159 filter helps remove camera shake from hand-holding a camera, bumping a
6160 tripod, moving on a vehicle, etc.
6162 The filter accepts the following options:
6170 Specify a rectangular area where to limit the search for motion
6172 If desired the search for motion vectors can be limited to a
6173 rectangular area of the frame defined by its top left corner, width
6174 and height. These parameters have the same meaning as the drawbox
6175 filter which can be used to visualise the position of the bounding
6178 This is useful when simultaneous movement of subjects within the frame
6179 might be confused for camera motion by the motion vector search.
6181 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6182 then the full frame is used. This allows later options to be set
6183 without specifying the bounding box for the motion vector search.
6185 Default - search the whole frame.
6189 Specify the maximum extent of movement in x and y directions in the
6190 range 0-64 pixels. Default 16.
6193 Specify how to generate pixels to fill blanks at the edge of the
6194 frame. Available values are:
6197 Fill zeroes at blank locations
6199 Original image at blank locations
6201 Extruded edge value at blank locations
6203 Mirrored edge at blank locations
6205 Default value is @samp{mirror}.
6208 Specify the blocksize to use for motion search. Range 4-128 pixels,
6212 Specify the contrast threshold for blocks. Only blocks with more than
6213 the specified contrast (difference between darkest and lightest
6214 pixels) will be considered. Range 1-255, default 125.
6217 Specify the search strategy. Available values are:
6220 Set exhaustive search
6222 Set less exhaustive search.
6224 Default value is @samp{exhaustive}.
6227 If set then a detailed log of the motion search is written to the
6231 If set to 1, specify using OpenCL capabilities, only available if
6232 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6238 Apply an exact inverse of the telecine operation. It requires a predefined
6239 pattern specified using the pattern option which must be the same as that passed
6240 to the telecine filter.
6242 This filter accepts the following options:
6251 The default value is @code{top}.
6255 A string of numbers representing the pulldown pattern you wish to apply.
6256 The default value is @code{23}.
6259 A number representing position of the first frame with respect to the telecine
6260 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6265 Apply dilation effect to the video.
6267 This filter replaces the pixel by the local(3x3) maximum.
6269 It accepts the following options:
6276 Limit the maximum change for each plane, default is 65535.
6277 If 0, plane will remain unchanged.
6280 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6283 Flags to local 3x3 coordinates maps like this:
6292 Displace pixels as indicated by second and third input stream.
6294 It takes three input streams and outputs one stream, the first input is the
6295 source, and second and third input are displacement maps.
6297 The second input specifies how much to displace pixels along the
6298 x-axis, while the third input specifies how much to displace pixels
6300 If one of displacement map streams terminates, last frame from that
6301 displacement map will be used.
6303 Note that once generated, displacements maps can be reused over and over again.
6305 A description of the accepted options follows.
6309 Set displace behavior for pixels that are out of range.
6311 Available values are:
6314 Missing pixels are replaced by black pixels.
6317 Adjacent pixels will spread out to replace missing pixels.
6320 Out of range pixels are wrapped so they point to pixels of other side.
6322 Default is @samp{smear}.
6326 @subsection Examples
6330 Add ripple effect to rgb input of video size hd720:
6332 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
6336 Add wave effect to rgb input of video size hd720:
6338 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
6344 Draw a colored box on the input image.
6346 It accepts the following parameters:
6351 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6355 The expressions which specify the width and height of the box; if 0 they are interpreted as
6356 the input width and height. It defaults to 0.
6359 Specify the color of the box to write. For the general syntax of this option,
6360 check the "Color" section in the ffmpeg-utils manual. If the special
6361 value @code{invert} is used, the box edge color is the same as the
6362 video with inverted luma.
6365 The expression which sets the thickness of the box edge. Default value is @code{3}.
6367 See below for the list of accepted constants.
6370 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6371 following constants:
6375 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6379 horizontal and vertical chroma subsample values. For example for the
6380 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6384 The input width and height.
6387 The input sample aspect ratio.
6391 The x and y offset coordinates where the box is drawn.
6395 The width and height of the drawn box.
6398 The thickness of the drawn box.
6400 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6401 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6405 @subsection Examples
6409 Draw a black box around the edge of the input image:
6415 Draw a box with color red and an opacity of 50%:
6417 drawbox=10:20:200:60:red@@0.5
6420 The previous example can be specified as:
6422 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6426 Fill the box with pink color:
6428 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6432 Draw a 2-pixel red 2.40:1 mask:
6434 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
6440 Draw a grid on the input image.
6442 It accepts the following parameters:
6447 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6451 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6452 input width and height, respectively, minus @code{thickness}, so image gets
6453 framed. Default to 0.
6456 Specify the color of the grid. For the general syntax of this option,
6457 check the "Color" section in the ffmpeg-utils manual. If the special
6458 value @code{invert} is used, the grid color is the same as the
6459 video with inverted luma.
6462 The expression which sets the thickness of the grid line. Default value is @code{1}.
6464 See below for the list of accepted constants.
6467 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6468 following constants:
6472 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6476 horizontal and vertical chroma subsample values. For example for the
6477 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6481 The input grid cell width and height.
6484 The input sample aspect ratio.
6488 The x and y coordinates of some point of grid intersection (meant to configure offset).
6492 The width and height of the drawn cell.
6495 The thickness of the drawn cell.
6497 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6498 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6502 @subsection Examples
6506 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6508 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6512 Draw a white 3x3 grid with an opacity of 50%:
6514 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6521 Draw a text string or text from a specified file on top of a video, using the
6522 libfreetype library.
6524 To enable compilation of this filter, you need to configure FFmpeg with
6525 @code{--enable-libfreetype}.
6526 To enable default font fallback and the @var{font} option you need to
6527 configure FFmpeg with @code{--enable-libfontconfig}.
6528 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6529 @code{--enable-libfribidi}.
6533 It accepts the following parameters:
6538 Used to draw a box around text using the background color.
6539 The value must be either 1 (enable) or 0 (disable).
6540 The default value of @var{box} is 0.
6543 Set the width of the border to be drawn around the box using @var{boxcolor}.
6544 The default value of @var{boxborderw} is 0.
6547 The color to be used for drawing box around text. For the syntax of this
6548 option, check the "Color" section in the ffmpeg-utils manual.
6550 The default value of @var{boxcolor} is "white".
6553 Set the width of the border to be drawn around the text using @var{bordercolor}.
6554 The default value of @var{borderw} is 0.
6557 Set the color to be used for drawing border around text. For the syntax of this
6558 option, check the "Color" section in the ffmpeg-utils manual.
6560 The default value of @var{bordercolor} is "black".
6563 Select how the @var{text} is expanded. Can be either @code{none},
6564 @code{strftime} (deprecated) or
6565 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6569 If true, check and fix text coords to avoid clipping.
6572 The color to be used for drawing fonts. For the syntax of this option, check
6573 the "Color" section in the ffmpeg-utils manual.
6575 The default value of @var{fontcolor} is "black".
6577 @item fontcolor_expr
6578 String which is expanded the same way as @var{text} to obtain dynamic
6579 @var{fontcolor} value. By default this option has empty value and is not
6580 processed. When this option is set, it overrides @var{fontcolor} option.
6583 The font family to be used for drawing text. By default Sans.
6586 The font file to be used for drawing text. The path must be included.
6587 This parameter is mandatory if the fontconfig support is disabled.
6590 This option does not exist, please see the timeline system
6593 Draw the text applying alpha blending. The value can
6594 be either a number between 0.0 and 1.0
6595 The expression accepts the same variables @var{x, y} do.
6596 The default value is 1.
6597 Please see fontcolor_expr
6600 The font size to be used for drawing text.
6601 The default value of @var{fontsize} is 16.
6604 If set to 1, attempt to shape the text (for example, reverse the order of
6605 right-to-left text and join Arabic characters) before drawing it.
6606 Otherwise, just draw the text exactly as given.
6607 By default 1 (if supported).
6610 The flags to be used for loading the fonts.
6612 The flags map the corresponding flags supported by libfreetype, and are
6613 a combination of the following values:
6620 @item vertical_layout
6621 @item force_autohint
6624 @item ignore_global_advance_width
6626 @item ignore_transform
6632 Default value is "default".
6634 For more information consult the documentation for the FT_LOAD_*
6638 The color to be used for drawing a shadow behind the drawn text. For the
6639 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6641 The default value of @var{shadowcolor} is "black".
6645 The x and y offsets for the text shadow position with respect to the
6646 position of the text. They can be either positive or negative
6647 values. The default value for both is "0".
6650 The starting frame number for the n/frame_num variable. The default value
6654 The size in number of spaces to use for rendering the tab.
6658 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6659 format. It can be used with or without text parameter. @var{timecode_rate}
6660 option must be specified.
6662 @item timecode_rate, rate, r
6663 Set the timecode frame rate (timecode only).
6666 The text string to be drawn. The text must be a sequence of UTF-8
6668 This parameter is mandatory if no file is specified with the parameter
6672 A text file containing text to be drawn. The text must be a sequence
6673 of UTF-8 encoded characters.
6675 This parameter is mandatory if no text string is specified with the
6676 parameter @var{text}.
6678 If both @var{text} and @var{textfile} are specified, an error is thrown.
6681 If set to 1, the @var{textfile} will be reloaded before each frame.
6682 Be sure to update it atomically, or it may be read partially, or even fail.
6686 The expressions which specify the offsets where text will be drawn
6687 within the video frame. They are relative to the top/left border of the
6690 The default value of @var{x} and @var{y} is "0".
6692 See below for the list of accepted constants and functions.
6695 The parameters for @var{x} and @var{y} are expressions containing the
6696 following constants and functions:
6700 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6704 horizontal and vertical chroma subsample values. For example for the
6705 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6708 the height of each text line
6716 @item max_glyph_a, ascent
6717 the maximum distance from the baseline to the highest/upper grid
6718 coordinate used to place a glyph outline point, for all the rendered
6720 It is a positive value, due to the grid's orientation with the Y axis
6723 @item max_glyph_d, descent
6724 the maximum distance from the baseline to the lowest grid coordinate
6725 used to place a glyph outline point, for all the rendered glyphs.
6726 This is a negative value, due to the grid's orientation, with the Y axis
6730 maximum glyph height, that is the maximum height for all the glyphs
6731 contained in the rendered text, it is equivalent to @var{ascent} -
6735 maximum glyph width, that is the maximum width for all the glyphs
6736 contained in the rendered text
6739 the number of input frame, starting from 0
6741 @item rand(min, max)
6742 return a random number included between @var{min} and @var{max}
6745 The input sample aspect ratio.
6748 timestamp expressed in seconds, NAN if the input timestamp is unknown
6751 the height of the rendered text
6754 the width of the rendered text
6758 the x and y offset coordinates where the text is drawn.
6760 These parameters allow the @var{x} and @var{y} expressions to refer
6761 each other, so you can for example specify @code{y=x/dar}.
6764 @anchor{drawtext_expansion}
6765 @subsection Text expansion
6767 If @option{expansion} is set to @code{strftime},
6768 the filter recognizes strftime() sequences in the provided text and
6769 expands them accordingly. Check the documentation of strftime(). This
6770 feature is deprecated.
6772 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6774 If @option{expansion} is set to @code{normal} (which is the default),
6775 the following expansion mechanism is used.
6777 The backslash character @samp{\}, followed by any character, always expands to
6778 the second character.
6780 Sequence of the form @code{%@{...@}} are expanded. The text between the
6781 braces is a function name, possibly followed by arguments separated by ':'.
6782 If the arguments contain special characters or delimiters (':' or '@}'),
6783 they should be escaped.
6785 Note that they probably must also be escaped as the value for the
6786 @option{text} option in the filter argument string and as the filter
6787 argument in the filtergraph description, and possibly also for the shell,
6788 that makes up to four levels of escaping; using a text file avoids these
6791 The following functions are available:
6796 The expression evaluation result.
6798 It must take one argument specifying the expression to be evaluated,
6799 which accepts the same constants and functions as the @var{x} and
6800 @var{y} values. Note that not all constants should be used, for
6801 example the text size is not known when evaluating the expression, so
6802 the constants @var{text_w} and @var{text_h} will have an undefined
6805 @item expr_int_format, eif
6806 Evaluate the expression's value and output as formatted integer.
6808 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6809 The second argument specifies the output format. Allowed values are @samp{x},
6810 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6811 @code{printf} function.
6812 The third parameter is optional and sets the number of positions taken by the output.
6813 It can be used to add padding with zeros from the left.
6816 The time at which the filter is running, expressed in UTC.
6817 It can accept an argument: a strftime() format string.
6820 The time at which the filter is running, expressed in the local time zone.
6821 It can accept an argument: a strftime() format string.
6824 Frame metadata. Takes one or two arguments.
6826 The first argument is mandatory and specifies the metadata key.
6828 The second argument is optional and specifies a default value, used when the
6829 metadata key is not found or empty.
6832 The frame number, starting from 0.
6835 A 1 character description of the current picture type.
6838 The timestamp of the current frame.
6839 It can take up to three arguments.
6841 The first argument is the format of the timestamp; it defaults to @code{flt}
6842 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6843 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6844 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
6845 @code{localtime} stands for the timestamp of the frame formatted as
6846 local time zone time.
6848 The second argument is an offset added to the timestamp.
6850 If the format is set to @code{localtime} or @code{gmtime},
6851 a third argument may be supplied: a strftime() format string.
6852 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
6855 @subsection Examples
6859 Draw "Test Text" with font FreeSerif, using the default values for the
6860 optional parameters.
6863 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
6867 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
6868 and y=50 (counting from the top-left corner of the screen), text is
6869 yellow with a red box around it. Both the text and the box have an
6873 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
6874 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
6877 Note that the double quotes are not necessary if spaces are not used
6878 within the parameter list.
6881 Show the text at the center of the video frame:
6883 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
6887 Show the text at a random position, switching to a new position every 30 seconds:
6889 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)"
6893 Show a text line sliding from right to left in the last row of the video
6894 frame. The file @file{LONG_LINE} is assumed to contain a single line
6897 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
6901 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
6903 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
6907 Draw a single green letter "g", at the center of the input video.
6908 The glyph baseline is placed at half screen height.
6910 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
6914 Show text for 1 second every 3 seconds:
6916 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
6920 Use fontconfig to set the font. Note that the colons need to be escaped.
6922 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
6926 Print the date of a real-time encoding (see strftime(3)):
6928 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
6932 Show text fading in and out (appearing/disappearing):
6935 DS=1.0 # display start
6936 DE=10.0 # display end
6937 FID=1.5 # fade in duration
6938 FOD=5 # fade out duration
6939 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 @}"
6944 For more information about libfreetype, check:
6945 @url{http://www.freetype.org/}.
6947 For more information about fontconfig, check:
6948 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
6950 For more information about libfribidi, check:
6951 @url{http://fribidi.org/}.
6955 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
6957 The filter accepts the following options:
6962 Set low and high threshold values used by the Canny thresholding
6965 The high threshold selects the "strong" edge pixels, which are then
6966 connected through 8-connectivity with the "weak" edge pixels selected
6967 by the low threshold.
6969 @var{low} and @var{high} threshold values must be chosen in the range
6970 [0,1], and @var{low} should be lesser or equal to @var{high}.
6972 Default value for @var{low} is @code{20/255}, and default value for @var{high}
6976 Define the drawing mode.
6980 Draw white/gray wires on black background.
6983 Mix the colors to create a paint/cartoon effect.
6986 Default value is @var{wires}.
6989 @subsection Examples
6993 Standard edge detection with custom values for the hysteresis thresholding:
6995 edgedetect=low=0.1:high=0.4
6999 Painting effect without thresholding:
7001 edgedetect=mode=colormix:high=0
7006 Set brightness, contrast, saturation and approximate gamma adjustment.
7008 The filter accepts the following options:
7012 Set the contrast expression. The value must be a float value in range
7013 @code{-2.0} to @code{2.0}. The default value is "1".
7016 Set the brightness expression. The value must be a float value in
7017 range @code{-1.0} to @code{1.0}. The default value is "0".
7020 Set the saturation expression. The value must be a float in
7021 range @code{0.0} to @code{3.0}. The default value is "1".
7024 Set the gamma expression. The value must be a float in range
7025 @code{0.1} to @code{10.0}. The default value is "1".
7028 Set the gamma expression for red. The value must be a float in
7029 range @code{0.1} to @code{10.0}. The default value is "1".
7032 Set the gamma expression for green. The value must be a float in range
7033 @code{0.1} to @code{10.0}. The default value is "1".
7036 Set the gamma expression for blue. The value must be a float in range
7037 @code{0.1} to @code{10.0}. The default value is "1".
7040 Set the gamma weight expression. It can be used to reduce the effect
7041 of a high gamma value on bright image areas, e.g. keep them from
7042 getting overamplified and just plain white. The value must be a float
7043 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7044 gamma correction all the way down while @code{1.0} leaves it at its
7045 full strength. Default is "1".
7048 Set when the expressions for brightness, contrast, saturation and
7049 gamma expressions are evaluated.
7051 It accepts the following values:
7054 only evaluate expressions once during the filter initialization or
7055 when a command is processed
7058 evaluate expressions for each incoming frame
7061 Default value is @samp{init}.
7064 The expressions accept the following parameters:
7067 frame count of the input frame starting from 0
7070 byte position of the corresponding packet in the input file, NAN if
7074 frame rate of the input video, NAN if the input frame rate is unknown
7077 timestamp expressed in seconds, NAN if the input timestamp is unknown
7080 @subsection Commands
7081 The filter supports the following commands:
7085 Set the contrast expression.
7088 Set the brightness expression.
7091 Set the saturation expression.
7094 Set the gamma expression.
7097 Set the gamma_r expression.
7100 Set gamma_g expression.
7103 Set gamma_b expression.
7106 Set gamma_weight expression.
7108 The command accepts the same syntax of the corresponding option.
7110 If the specified expression is not valid, it is kept at its current
7117 Apply erosion effect to the video.
7119 This filter replaces the pixel by the local(3x3) minimum.
7121 It accepts the following options:
7128 Limit the maximum change for each plane, default is 65535.
7129 If 0, plane will remain unchanged.
7132 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7135 Flags to local 3x3 coordinates maps like this:
7142 @section extractplanes
7144 Extract color channel components from input video stream into
7145 separate grayscale video streams.
7147 The filter accepts the following option:
7151 Set plane(s) to extract.
7153 Available values for planes are:
7164 Choosing planes not available in the input will result in an error.
7165 That means you cannot select @code{r}, @code{g}, @code{b} planes
7166 with @code{y}, @code{u}, @code{v} planes at same time.
7169 @subsection Examples
7173 Extract luma, u and v color channel component from input video frame
7174 into 3 grayscale outputs:
7176 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
7182 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7184 For each input image, the filter will compute the optimal mapping from
7185 the input to the output given the codebook length, that is the number
7186 of distinct output colors.
7188 This filter accepts the following options.
7191 @item codebook_length, l
7192 Set codebook length. The value must be a positive integer, and
7193 represents the number of distinct output colors. Default value is 256.
7196 Set the maximum number of iterations to apply for computing the optimal
7197 mapping. The higher the value the better the result and the higher the
7198 computation time. Default value is 1.
7201 Set a random seed, must be an integer included between 0 and
7202 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7203 will try to use a good random seed on a best effort basis.
7206 Set pal8 output pixel format. This option does not work with codebook
7207 length greater than 256.
7212 Apply a fade-in/out effect to the input video.
7214 It accepts the following parameters:
7218 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7220 Default is @code{in}.
7222 @item start_frame, s
7223 Specify the number of the frame to start applying the fade
7224 effect at. Default is 0.
7227 The number of frames that the fade effect lasts. At the end of the
7228 fade-in effect, the output video will have the same intensity as the input video.
7229 At the end of the fade-out transition, the output video will be filled with the
7230 selected @option{color}.
7234 If set to 1, fade only alpha channel, if one exists on the input.
7237 @item start_time, st
7238 Specify the timestamp (in seconds) of the frame to start to apply the fade
7239 effect. If both start_frame and start_time are specified, the fade will start at
7240 whichever comes last. Default is 0.
7243 The number of seconds for which the fade effect has to last. At the end of the
7244 fade-in effect the output video will have the same intensity as the input video,
7245 at the end of the fade-out transition the output video will be filled with the
7246 selected @option{color}.
7247 If both duration and nb_frames are specified, duration is used. Default is 0
7248 (nb_frames is used by default).
7251 Specify the color of the fade. Default is "black".
7254 @subsection Examples
7258 Fade in the first 30 frames of video:
7263 The command above is equivalent to:
7269 Fade out the last 45 frames of a 200-frame video:
7272 fade=type=out:start_frame=155:nb_frames=45
7276 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7278 fade=in:0:25, fade=out:975:25
7282 Make the first 5 frames yellow, then fade in from frame 5-24:
7284 fade=in:5:20:color=yellow
7288 Fade in alpha over first 25 frames of video:
7290 fade=in:0:25:alpha=1
7294 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7296 fade=t=in:st=5.5:d=0.5
7302 Apply arbitrary expressions to samples in frequency domain
7306 Adjust the dc value (gain) of the luma plane of the image. The filter
7307 accepts an integer value in range @code{0} to @code{1000}. The default
7308 value is set to @code{0}.
7311 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7312 filter accepts an integer value in range @code{0} to @code{1000}. The
7313 default value is set to @code{0}.
7316 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7317 filter accepts an integer value in range @code{0} to @code{1000}. The
7318 default value is set to @code{0}.
7321 Set the frequency domain weight expression for the luma plane.
7324 Set the frequency domain weight expression for the 1st chroma plane.
7327 Set the frequency domain weight expression for the 2nd chroma plane.
7329 The filter accepts the following variables:
7332 The coordinates of the current sample.
7336 The width and height of the image.
7339 @subsection Examples
7345 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7351 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7357 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7363 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7370 Extract a single field from an interlaced image using stride
7371 arithmetic to avoid wasting CPU time. The output frames are marked as
7374 The filter accepts the following options:
7378 Specify whether to extract the top (if the value is @code{0} or
7379 @code{top}) or the bottom field (if the value is @code{1} or
7385 Create new frames by copying the top and bottom fields from surrounding frames
7386 supplied as numbers by the hint file.
7390 Set file containing hints: absolute/relative frame numbers.
7392 There must be one line for each frame in a clip. Each line must contain two
7393 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7394 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7395 is current frame number for @code{absolute} mode or out of [-1, 1] range
7396 for @code{relative} mode. First number tells from which frame to pick up top
7397 field and second number tells from which frame to pick up bottom field.
7399 If optionally followed by @code{+} output frame will be marked as interlaced,
7400 else if followed by @code{-} output frame will be marked as progressive, else
7401 it will be marked same as input frame.
7402 If line starts with @code{#} or @code{;} that line is skipped.
7405 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7408 Example of first several lines of @code{hint} file for @code{relative} mode:
7411 1,0 - # second frame, use third's frame top field and second's frame bottom field
7412 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7429 Field matching filter for inverse telecine. It is meant to reconstruct the
7430 progressive frames from a telecined stream. The filter does not drop duplicated
7431 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7432 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7434 The separation of the field matching and the decimation is notably motivated by
7435 the possibility of inserting a de-interlacing filter fallback between the two.
7436 If the source has mixed telecined and real interlaced content,
7437 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7438 But these remaining combed frames will be marked as interlaced, and thus can be
7439 de-interlaced by a later filter such as @ref{yadif} before decimation.
7441 In addition to the various configuration options, @code{fieldmatch} can take an
7442 optional second stream, activated through the @option{ppsrc} option. If
7443 enabled, the frames reconstruction will be based on the fields and frames from
7444 this second stream. This allows the first input to be pre-processed in order to
7445 help the various algorithms of the filter, while keeping the output lossless
7446 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7447 or brightness/contrast adjustments can help.
7449 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7450 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7451 which @code{fieldmatch} is based on. While the semantic and usage are very
7452 close, some behaviour and options names can differ.
7454 The @ref{decimate} filter currently only works for constant frame rate input.
7455 If your input has mixed telecined (30fps) and progressive content with a lower
7456 framerate like 24fps use the following filterchain to produce the necessary cfr
7457 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7459 The filter accepts the following options:
7463 Specify the assumed field order of the input stream. Available values are:
7467 Auto detect parity (use FFmpeg's internal parity value).
7469 Assume bottom field first.
7471 Assume top field first.
7474 Note that it is sometimes recommended not to trust the parity announced by the
7477 Default value is @var{auto}.
7480 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7481 sense that it won't risk creating jerkiness due to duplicate frames when
7482 possible, but if there are bad edits or blended fields it will end up
7483 outputting combed frames when a good match might actually exist. On the other
7484 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7485 but will almost always find a good frame if there is one. The other values are
7486 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7487 jerkiness and creating duplicate frames versus finding good matches in sections
7488 with bad edits, orphaned fields, blended fields, etc.
7490 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7492 Available values are:
7496 2-way matching (p/c)
7498 2-way matching, and trying 3rd match if still combed (p/c + n)
7500 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7502 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7503 still combed (p/c + n + u/b)
7505 3-way matching (p/c/n)
7507 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7508 detected as combed (p/c/n + u/b)
7511 The parenthesis at the end indicate the matches that would be used for that
7512 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7515 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7518 Default value is @var{pc_n}.
7521 Mark the main input stream as a pre-processed input, and enable the secondary
7522 input stream as the clean source to pick the fields from. See the filter
7523 introduction for more details. It is similar to the @option{clip2} feature from
7526 Default value is @code{0} (disabled).
7529 Set the field to match from. It is recommended to set this to the same value as
7530 @option{order} unless you experience matching failures with that setting. In
7531 certain circumstances changing the field that is used to match from can have a
7532 large impact on matching performance. Available values are:
7536 Automatic (same value as @option{order}).
7538 Match from the bottom field.
7540 Match from the top field.
7543 Default value is @var{auto}.
7546 Set whether or not chroma is included during the match comparisons. In most
7547 cases it is recommended to leave this enabled. You should set this to @code{0}
7548 only if your clip has bad chroma problems such as heavy rainbowing or other
7549 artifacts. Setting this to @code{0} could also be used to speed things up at
7550 the cost of some accuracy.
7552 Default value is @code{1}.
7556 These define an exclusion band which excludes the lines between @option{y0} and
7557 @option{y1} from being included in the field matching decision. An exclusion
7558 band can be used to ignore subtitles, a logo, or other things that may
7559 interfere with the matching. @option{y0} sets the starting scan line and
7560 @option{y1} sets the ending line; all lines in between @option{y0} and
7561 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7562 @option{y0} and @option{y1} to the same value will disable the feature.
7563 @option{y0} and @option{y1} defaults to @code{0}.
7566 Set the scene change detection threshold as a percentage of maximum change on
7567 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7568 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7569 @option{scthresh} is @code{[0.0, 100.0]}.
7571 Default value is @code{12.0}.
7574 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7575 account the combed scores of matches when deciding what match to use as the
7576 final match. Available values are:
7580 No final matching based on combed scores.
7582 Combed scores are only used when a scene change is detected.
7584 Use combed scores all the time.
7587 Default is @var{sc}.
7590 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7591 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7592 Available values are:
7596 No forced calculation.
7598 Force p/c/n calculations.
7600 Force p/c/n/u/b calculations.
7603 Default value is @var{none}.
7606 This is the area combing threshold used for combed frame detection. This
7607 essentially controls how "strong" or "visible" combing must be to be detected.
7608 Larger values mean combing must be more visible and smaller values mean combing
7609 can be less visible or strong and still be detected. Valid settings are from
7610 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7611 be detected as combed). This is basically a pixel difference value. A good
7612 range is @code{[8, 12]}.
7614 Default value is @code{9}.
7617 Sets whether or not chroma is considered in the combed frame decision. Only
7618 disable this if your source has chroma problems (rainbowing, etc.) that are
7619 causing problems for the combed frame detection with chroma enabled. Actually,
7620 using @option{chroma}=@var{0} is usually more reliable, except for the case
7621 where there is chroma only combing in the source.
7623 Default value is @code{0}.
7627 Respectively set the x-axis and y-axis size of the window used during combed
7628 frame detection. This has to do with the size of the area in which
7629 @option{combpel} pixels are required to be detected as combed for a frame to be
7630 declared combed. See the @option{combpel} parameter description for more info.
7631 Possible values are any number that is a power of 2 starting at 4 and going up
7634 Default value is @code{16}.
7637 The number of combed pixels inside any of the @option{blocky} by
7638 @option{blockx} size blocks on the frame for the frame to be detected as
7639 combed. While @option{cthresh} controls how "visible" the combing must be, this
7640 setting controls "how much" combing there must be in any localized area (a
7641 window defined by the @option{blockx} and @option{blocky} settings) on the
7642 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7643 which point no frames will ever be detected as combed). This setting is known
7644 as @option{MI} in TFM/VFM vocabulary.
7646 Default value is @code{80}.
7649 @anchor{p/c/n/u/b meaning}
7650 @subsection p/c/n/u/b meaning
7652 @subsubsection p/c/n
7654 We assume the following telecined stream:
7657 Top fields: 1 2 2 3 4
7658 Bottom fields: 1 2 3 4 4
7661 The numbers correspond to the progressive frame the fields relate to. Here, the
7662 first two frames are progressive, the 3rd and 4th are combed, and so on.
7664 When @code{fieldmatch} is configured to run a matching from bottom
7665 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7670 B 1 2 3 4 4 <-- matching reference
7679 As a result of the field matching, we can see that some frames get duplicated.
7680 To perform a complete inverse telecine, you need to rely on a decimation filter
7681 after this operation. See for instance the @ref{decimate} filter.
7683 The same operation now matching from top fields (@option{field}=@var{top})
7688 T 1 2 2 3 4 <-- matching reference
7698 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7699 basically, they refer to the frame and field of the opposite parity:
7702 @item @var{p} matches the field of the opposite parity in the previous frame
7703 @item @var{c} matches the field of the opposite parity in the current frame
7704 @item @var{n} matches the field of the opposite parity in the next frame
7709 The @var{u} and @var{b} matching are a bit special in the sense that they match
7710 from the opposite parity flag. In the following examples, we assume that we are
7711 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7712 'x' is placed above and below each matched fields.
7714 With bottom matching (@option{field}=@var{bottom}):
7719 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7720 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7728 With top matching (@option{field}=@var{top}):
7733 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7734 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7742 @subsection Examples
7744 Simple IVTC of a top field first telecined stream:
7746 fieldmatch=order=tff:combmatch=none, decimate
7749 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7751 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7756 Transform the field order of the input video.
7758 It accepts the following parameters:
7763 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7764 for bottom field first.
7767 The default value is @samp{tff}.
7769 The transformation is done by shifting the picture content up or down
7770 by one line, and filling the remaining line with appropriate picture content.
7771 This method is consistent with most broadcast field order converters.
7773 If the input video is not flagged as being interlaced, or it is already
7774 flagged as being of the required output field order, then this filter does
7775 not alter the incoming video.
7777 It is very useful when converting to or from PAL DV material,
7778 which is bottom field first.
7782 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7785 @section fifo, afifo
7787 Buffer input images and send them when they are requested.
7789 It is mainly useful when auto-inserted by the libavfilter
7792 It does not take parameters.
7796 Find a rectangular object
7798 It accepts the following options:
7802 Filepath of the object image, needs to be in gray8.
7805 Detection threshold, default is 0.5.
7808 Number of mipmaps, default is 3.
7810 @item xmin, ymin, xmax, ymax
7811 Specifies the rectangle in which to search.
7814 @subsection Examples
7818 Generate a representative palette of a given video using @command{ffmpeg}:
7820 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7826 Cover a rectangular object
7828 It accepts the following options:
7832 Filepath of the optional cover image, needs to be in yuv420.
7837 It accepts the following values:
7840 cover it by the supplied image
7842 cover it by interpolating the surrounding pixels
7845 Default value is @var{blur}.
7848 @subsection Examples
7852 Generate a representative palette of a given video using @command{ffmpeg}:
7854 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7861 Convert the input video to one of the specified pixel formats.
7862 Libavfilter will try to pick one that is suitable as input to
7865 It accepts the following parameters:
7869 A '|'-separated list of pixel format names, such as
7870 "pix_fmts=yuv420p|monow|rgb24".
7874 @subsection Examples
7878 Convert the input video to the @var{yuv420p} format
7880 format=pix_fmts=yuv420p
7883 Convert the input video to any of the formats in the list
7885 format=pix_fmts=yuv420p|yuv444p|yuv410p
7892 Convert the video to specified constant frame rate by duplicating or dropping
7893 frames as necessary.
7895 It accepts the following parameters:
7899 The desired output frame rate. The default is @code{25}.
7904 Possible values are:
7907 zero round towards 0
7911 round towards -infinity
7913 round towards +infinity
7917 The default is @code{near}.
7920 Assume the first PTS should be the given value, in seconds. This allows for
7921 padding/trimming at the start of stream. By default, no assumption is made
7922 about the first frame's expected PTS, so no padding or trimming is done.
7923 For example, this could be set to 0 to pad the beginning with duplicates of
7924 the first frame if a video stream starts after the audio stream or to trim any
7925 frames with a negative PTS.
7929 Alternatively, the options can be specified as a flat string:
7930 @var{fps}[:@var{round}].
7932 See also the @ref{setpts} filter.
7934 @subsection Examples
7938 A typical usage in order to set the fps to 25:
7944 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
7946 fps=fps=film:round=near
7952 Pack two different video streams into a stereoscopic video, setting proper
7953 metadata on supported codecs. The two views should have the same size and
7954 framerate and processing will stop when the shorter video ends. Please note
7955 that you may conveniently adjust view properties with the @ref{scale} and
7958 It accepts the following parameters:
7962 The desired packing format. Supported values are:
7967 The views are next to each other (default).
7970 The views are on top of each other.
7973 The views are packed by line.
7976 The views are packed by column.
7979 The views are temporally interleaved.
7988 # Convert left and right views into a frame-sequential video
7989 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
7991 # Convert views into a side-by-side video with the same output resolution as the input
7992 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
7997 Change the frame rate by interpolating new video output frames from the source
8000 This filter is not designed to function correctly with interlaced media. If
8001 you wish to change the frame rate of interlaced media then you are required
8002 to deinterlace before this filter and re-interlace after this filter.
8004 A description of the accepted options follows.
8008 Specify the output frames per second. This option can also be specified
8009 as a value alone. The default is @code{50}.
8012 Specify the start of a range where the output frame will be created as a
8013 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8014 the default is @code{15}.
8017 Specify the end of a range where the output frame will be created as a
8018 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8019 the default is @code{240}.
8022 Specify the level at which a scene change is detected as a value between
8023 0 and 100 to indicate a new scene; a low value reflects a low
8024 probability for the current frame to introduce a new scene, while a higher
8025 value means the current frame is more likely to be one.
8026 The default is @code{7}.
8029 Specify flags influencing the filter process.
8031 Available value for @var{flags} is:
8034 @item scene_change_detect, scd
8035 Enable scene change detection using the value of the option @var{scene}.
8036 This flag is enabled by default.
8042 Select one frame every N-th frame.
8044 This filter accepts the following option:
8047 Select frame after every @code{step} frames.
8048 Allowed values are positive integers higher than 0. Default value is @code{1}.
8054 Apply a frei0r effect to the input video.
8056 To enable the compilation of this filter, you need to install the frei0r
8057 header and configure FFmpeg with @code{--enable-frei0r}.
8059 It accepts the following parameters:
8064 The name of the frei0r effect to load. If the environment variable
8065 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8066 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8067 Otherwise, the standard frei0r paths are searched, in this order:
8068 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8069 @file{/usr/lib/frei0r-1/}.
8072 A '|'-separated list of parameters to pass to the frei0r effect.
8076 A frei0r effect parameter can be a boolean (its value is either
8077 "y" or "n"), a double, a color (specified as
8078 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8079 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8080 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8081 @var{X} and @var{Y} are floating point numbers) and/or a string.
8083 The number and types of parameters depend on the loaded effect. If an
8084 effect parameter is not specified, the default value is set.
8086 @subsection Examples
8090 Apply the distort0r effect, setting the first two double parameters:
8092 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8096 Apply the colordistance effect, taking a color as the first parameter:
8098 frei0r=colordistance:0.2/0.3/0.4
8099 frei0r=colordistance:violet
8100 frei0r=colordistance:0x112233
8104 Apply the perspective effect, specifying the top left and top right image
8107 frei0r=perspective:0.2/0.2|0.8/0.2
8111 For more information, see
8112 @url{http://frei0r.dyne.org}
8116 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8118 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8119 processing filter, one of them is performed once per block, not per pixel.
8120 This allows for much higher speed.
8122 The filter accepts the following options:
8126 Set quality. This option defines the number of levels for averaging. It accepts
8127 an integer in the range 4-5. Default value is @code{4}.
8130 Force a constant quantization parameter. It accepts an integer in range 0-63.
8131 If not set, the filter will use the QP from the video stream (if available).
8134 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8135 more details but also more artifacts, while higher values make the image smoother
8136 but also blurrier. Default value is @code{0} − PSNR optimal.
8139 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8140 option may cause flicker since the B-Frames have often larger QP. Default is
8141 @code{0} (not enabled).
8147 The filter accepts the following options:
8151 Set the luminance expression.
8153 Set the chrominance blue expression.
8155 Set the chrominance red expression.
8157 Set the alpha expression.
8159 Set the red expression.
8161 Set the green expression.
8163 Set the blue expression.
8166 The colorspace is selected according to the specified options. If one
8167 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8168 options is specified, the filter will automatically select a YCbCr
8169 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8170 @option{blue_expr} options is specified, it will select an RGB
8173 If one of the chrominance expression is not defined, it falls back on the other
8174 one. If no alpha expression is specified it will evaluate to opaque value.
8175 If none of chrominance expressions are specified, they will evaluate
8176 to the luminance expression.
8178 The expressions can use the following variables and functions:
8182 The sequential number of the filtered frame, starting from @code{0}.
8186 The coordinates of the current sample.
8190 The width and height of the image.
8194 Width and height scale depending on the currently filtered plane. It is the
8195 ratio between the corresponding luma plane number of pixels and the current
8196 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8197 @code{0.5,0.5} for chroma planes.
8200 Time of the current frame, expressed in seconds.
8203 Return the value of the pixel at location (@var{x},@var{y}) of the current
8207 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8211 Return the value of the pixel at location (@var{x},@var{y}) of the
8212 blue-difference chroma plane. Return 0 if there is no such plane.
8215 Return the value of the pixel at location (@var{x},@var{y}) of the
8216 red-difference chroma plane. Return 0 if there is no such plane.
8221 Return the value of the pixel at location (@var{x},@var{y}) of the
8222 red/green/blue component. Return 0 if there is no such component.
8225 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8226 plane. Return 0 if there is no such plane.
8229 For functions, if @var{x} and @var{y} are outside the area, the value will be
8230 automatically clipped to the closer edge.
8232 @subsection Examples
8236 Flip the image horizontally:
8242 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8243 wavelength of 100 pixels:
8245 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8249 Generate a fancy enigmatic moving light:
8251 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
8255 Generate a quick emboss effect:
8257 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8261 Modify RGB components depending on pixel position:
8263 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8267 Create a radial gradient that is the same size as the input (also see
8268 the @ref{vignette} filter):
8270 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8276 Fix the banding artifacts that are sometimes introduced into nearly flat
8277 regions by truncation to 8-bit color depth.
8278 Interpolate the gradients that should go where the bands are, and
8281 It is designed for playback only. Do not use it prior to
8282 lossy compression, because compression tends to lose the dither and
8283 bring back the bands.
8285 It accepts the following parameters:
8290 The maximum amount by which the filter will change any one pixel. This is also
8291 the threshold for detecting nearly flat regions. Acceptable values range from
8292 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8296 The neighborhood to fit the gradient to. A larger radius makes for smoother
8297 gradients, but also prevents the filter from modifying the pixels near detailed
8298 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8299 values will be clipped to the valid range.
8303 Alternatively, the options can be specified as a flat string:
8304 @var{strength}[:@var{radius}]
8306 @subsection Examples
8310 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8316 Specify radius, omitting the strength (which will fall-back to the default
8327 Apply a Hald CLUT to a video stream.
8329 First input is the video stream to process, and second one is the Hald CLUT.
8330 The Hald CLUT input can be a simple picture or a complete video stream.
8332 The filter accepts the following options:
8336 Force termination when the shortest input terminates. Default is @code{0}.
8338 Continue applying the last CLUT after the end of the stream. A value of
8339 @code{0} disable the filter after the last frame of the CLUT is reached.
8340 Default is @code{1}.
8343 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8344 filters share the same internals).
8346 More information about the Hald CLUT can be found on Eskil Steenberg's website
8347 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8349 @subsection Workflow examples
8351 @subsubsection Hald CLUT video stream
8353 Generate an identity Hald CLUT stream altered with various effects:
8355 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
8358 Note: make sure you use a lossless codec.
8360 Then use it with @code{haldclut} to apply it on some random stream:
8362 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8365 The Hald CLUT will be applied to the 10 first seconds (duration of
8366 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8367 to the remaining frames of the @code{mandelbrot} stream.
8369 @subsubsection Hald CLUT with preview
8371 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8372 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8373 biggest possible square starting at the top left of the picture. The remaining
8374 padding pixels (bottom or right) will be ignored. This area can be used to add
8375 a preview of the Hald CLUT.
8377 Typically, the following generated Hald CLUT will be supported by the
8378 @code{haldclut} filter:
8381 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8382 pad=iw+320 [padded_clut];
8383 smptebars=s=320x256, split [a][b];
8384 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8385 [main][b] overlay=W-320" -frames:v 1 clut.png
8388 It contains the original and a preview of the effect of the CLUT: SMPTE color
8389 bars are displayed on the right-top, and below the same color bars processed by
8392 Then, the effect of this Hald CLUT can be visualized with:
8394 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8399 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
8400 embedded HDCD codes is expanded into a 20-bit PCM stream.
8402 The filter supports the Peak Extend and Low-level Gain Adjustment features
8403 of HDCD, and detects the Transient Filter flag.
8406 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
8409 When using the filter with wav, note the default encoding for wav is 16-bit,
8410 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
8411 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
8413 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
8414 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
8417 The filter accepts the following options:
8420 @item process_stereo
8421 Process the stereo channels together. If target_gain does not match between
8422 channels, consider it invalid and use the last valid target_gain.
8425 Always extend peaks above -3dBFS even if PE isn't signaled.
8428 Replace audio with a solid tone and adjust the amplitude to signal some
8429 specific aspect of the decoding process. The output file can be loaded in
8430 an audio editor alongside the original to aid analysis.
8432 @code{analyze_mode=pe:force_pe=1} can be used to see all samples above the PE level.
8439 Gain adjustment level at each sample
8441 Samples where peak extend occurs
8443 Samples where the code detect timer is active
8445 Samples where the target gain does not match between channels
8451 Flip the input video horizontally.
8453 For example, to horizontally flip the input video with @command{ffmpeg}:
8455 ffmpeg -i in.avi -vf "hflip" out.avi
8459 This filter applies a global color histogram equalization on a
8462 It can be used to correct video that has a compressed range of pixel
8463 intensities. The filter redistributes the pixel intensities to
8464 equalize their distribution across the intensity range. It may be
8465 viewed as an "automatically adjusting contrast filter". This filter is
8466 useful only for correcting degraded or poorly captured source
8469 The filter accepts the following options:
8473 Determine the amount of equalization to be applied. As the strength
8474 is reduced, the distribution of pixel intensities more-and-more
8475 approaches that of the input frame. The value must be a float number
8476 in the range [0,1] and defaults to 0.200.
8479 Set the maximum intensity that can generated and scale the output
8480 values appropriately. The strength should be set as desired and then
8481 the intensity can be limited if needed to avoid washing-out. The value
8482 must be a float number in the range [0,1] and defaults to 0.210.
8485 Set the antibanding level. If enabled the filter will randomly vary
8486 the luminance of output pixels by a small amount to avoid banding of
8487 the histogram. Possible values are @code{none}, @code{weak} or
8488 @code{strong}. It defaults to @code{none}.
8493 Compute and draw a color distribution histogram for the input video.
8495 The computed histogram is a representation of the color component
8496 distribution in an image.
8498 Standard histogram displays the color components distribution in an image.
8499 Displays color graph for each color component. Shows distribution of
8500 the Y, U, V, A or R, G, B components, depending on input format, in the
8501 current frame. Below each graph a color component scale meter is shown.
8503 The filter accepts the following options:
8507 Set height of level. Default value is @code{200}.
8508 Allowed range is [50, 2048].
8511 Set height of color scale. Default value is @code{12}.
8512 Allowed range is [0, 40].
8516 It accepts the following values:
8519 Per color component graphs are placed below each other.
8522 Presents information identical to that in the @code{parade}, except
8523 that the graphs representing color components are superimposed directly
8526 Default is @code{parade}.
8529 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8530 Default is @code{linear}.
8533 Set what color components to display.
8534 Default is @code{7}.
8537 @subsection Examples
8542 Calculate and draw histogram:
8544 ffplay -i input -vf histogram
8552 This is a high precision/quality 3d denoise filter. It aims to reduce
8553 image noise, producing smooth images and making still images really
8554 still. It should enhance compressibility.
8556 It accepts the following optional parameters:
8560 A non-negative floating point number which specifies spatial luma strength.
8563 @item chroma_spatial
8564 A non-negative floating point number which specifies spatial chroma strength.
8565 It defaults to 3.0*@var{luma_spatial}/4.0.
8568 A floating point number which specifies luma temporal strength. It defaults to
8569 6.0*@var{luma_spatial}/4.0.
8572 A floating point number which specifies chroma temporal strength. It defaults to
8573 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8576 @anchor{hwupload_cuda}
8577 @section hwupload_cuda
8579 Upload system memory frames to a CUDA device.
8581 It accepts the following optional parameters:
8585 The number of the CUDA device to use
8590 Apply a high-quality magnification filter designed for pixel art. This filter
8591 was originally created by Maxim Stepin.
8593 It accepts the following option:
8597 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8598 @code{hq3x} and @code{4} for @code{hq4x}.
8599 Default is @code{3}.
8603 Stack input videos horizontally.
8605 All streams must be of same pixel format and of same height.
8607 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8608 to create same output.
8610 The filter accept the following option:
8614 Set number of input streams. Default is 2.
8617 If set to 1, force the output to terminate when the shortest input
8618 terminates. Default value is 0.
8623 Modify the hue and/or the saturation of the input.
8625 It accepts the following parameters:
8629 Specify the hue angle as a number of degrees. It accepts an expression,
8630 and defaults to "0".
8633 Specify the saturation in the [-10,10] range. It accepts an expression and
8637 Specify the hue angle as a number of radians. It accepts an
8638 expression, and defaults to "0".
8641 Specify the brightness in the [-10,10] range. It accepts an expression and
8645 @option{h} and @option{H} are mutually exclusive, and can't be
8646 specified at the same time.
8648 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8649 expressions containing the following constants:
8653 frame count of the input frame starting from 0
8656 presentation timestamp of the input frame expressed in time base units
8659 frame rate of the input video, NAN if the input frame rate is unknown
8662 timestamp expressed in seconds, NAN if the input timestamp is unknown
8665 time base of the input video
8668 @subsection Examples
8672 Set the hue to 90 degrees and the saturation to 1.0:
8678 Same command but expressing the hue in radians:
8684 Rotate hue and make the saturation swing between 0
8685 and 2 over a period of 1 second:
8687 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8691 Apply a 3 seconds saturation fade-in effect starting at 0:
8696 The general fade-in expression can be written as:
8698 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8702 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8704 hue="s=max(0\, min(1\, (8-t)/3))"
8707 The general fade-out expression can be written as:
8709 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8714 @subsection Commands
8716 This filter supports the following commands:
8722 Modify the hue and/or the saturation and/or brightness of the input video.
8723 The command accepts the same syntax of the corresponding option.
8725 If the specified expression is not valid, it is kept at its current
8731 Detect video interlacing type.
8733 This filter tries to detect if the input frames as interlaced, progressive,
8734 top or bottom field first. It will also try and detect fields that are
8735 repeated between adjacent frames (a sign of telecine).
8737 Single frame detection considers only immediately adjacent frames when classifying each frame.
8738 Multiple frame detection incorporates the classification history of previous frames.
8740 The filter will log these metadata values:
8743 @item single.current_frame
8744 Detected type of current frame using single-frame detection. One of:
8745 ``tff'' (top field first), ``bff'' (bottom field first),
8746 ``progressive'', or ``undetermined''
8749 Cumulative number of frames detected as top field first using single-frame detection.
8752 Cumulative number of frames detected as top field first using multiple-frame detection.
8755 Cumulative number of frames detected as bottom field first using single-frame detection.
8757 @item multiple.current_frame
8758 Detected type of current frame using multiple-frame detection. One of:
8759 ``tff'' (top field first), ``bff'' (bottom field first),
8760 ``progressive'', or ``undetermined''
8763 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8765 @item single.progressive
8766 Cumulative number of frames detected as progressive using single-frame detection.
8768 @item multiple.progressive
8769 Cumulative number of frames detected as progressive using multiple-frame detection.
8771 @item single.undetermined
8772 Cumulative number of frames that could not be classified using single-frame detection.
8774 @item multiple.undetermined
8775 Cumulative number of frames that could not be classified using multiple-frame detection.
8777 @item repeated.current_frame
8778 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8780 @item repeated.neither
8781 Cumulative number of frames with no repeated field.
8784 Cumulative number of frames with the top field repeated from the previous frame's top field.
8786 @item repeated.bottom
8787 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8790 The filter accepts the following options:
8794 Set interlacing threshold.
8796 Set progressive threshold.
8798 Threshold for repeated field detection.
8800 Number of frames after which a given frame's contribution to the
8801 statistics is halved (i.e., it contributes only 0.5 to it's
8802 classification). The default of 0 means that all frames seen are given
8803 full weight of 1.0 forever.
8804 @item analyze_interlaced_flag
8805 When this is not 0 then idet will use the specified number of frames to determine
8806 if the interlaced flag is accurate, it will not count undetermined frames.
8807 If the flag is found to be accurate it will be used without any further
8808 computations, if it is found to be inaccurate it will be cleared without any
8809 further computations. This allows inserting the idet filter as a low computational
8810 method to clean up the interlaced flag
8815 Deinterleave or interleave fields.
8817 This filter allows one to process interlaced images fields without
8818 deinterlacing them. Deinterleaving splits the input frame into 2
8819 fields (so called half pictures). Odd lines are moved to the top
8820 half of the output image, even lines to the bottom half.
8821 You can process (filter) them independently and then re-interleave them.
8823 The filter accepts the following options:
8827 @item chroma_mode, c
8829 Available values for @var{luma_mode}, @var{chroma_mode} and
8830 @var{alpha_mode} are:
8836 @item deinterleave, d
8837 Deinterleave fields, placing one above the other.
8840 Interleave fields. Reverse the effect of deinterleaving.
8842 Default value is @code{none}.
8845 @item chroma_swap, cs
8846 @item alpha_swap, as
8847 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
8852 Apply inflate effect to the video.
8854 This filter replaces the pixel by the local(3x3) average by taking into account
8855 only values higher than the pixel.
8857 It accepts the following options:
8864 Limit the maximum change for each plane, default is 65535.
8865 If 0, plane will remain unchanged.
8870 Simple interlacing filter from progressive contents. This interleaves upper (or
8871 lower) lines from odd frames with lower (or upper) lines from even frames,
8872 halving the frame rate and preserving image height.
8875 Original Original New Frame
8876 Frame 'j' Frame 'j+1' (tff)
8877 ========== =========== ==================
8878 Line 0 --------------------> Frame 'j' Line 0
8879 Line 1 Line 1 ----> Frame 'j+1' Line 1
8880 Line 2 ---------------------> Frame 'j' Line 2
8881 Line 3 Line 3 ----> Frame 'j+1' Line 3
8883 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
8886 It accepts the following optional parameters:
8890 This determines whether the interlaced frame is taken from the even
8891 (tff - default) or odd (bff) lines of the progressive frame.
8894 Enable (default) or disable the vertical lowpass filter to avoid twitter
8895 interlacing and reduce moire patterns.
8900 Deinterlace input video by applying Donald Graft's adaptive kernel
8901 deinterling. Work on interlaced parts of a video to produce
8904 The description of the accepted parameters follows.
8908 Set the threshold which affects the filter's tolerance when
8909 determining if a pixel line must be processed. It must be an integer
8910 in the range [0,255] and defaults to 10. A value of 0 will result in
8911 applying the process on every pixels.
8914 Paint pixels exceeding the threshold value to white if set to 1.
8918 Set the fields order. Swap fields if set to 1, leave fields alone if
8922 Enable additional sharpening if set to 1. Default is 0.
8925 Enable twoway sharpening if set to 1. Default is 0.
8928 @subsection Examples
8932 Apply default values:
8934 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
8938 Enable additional sharpening:
8944 Paint processed pixels in white:
8950 @section lenscorrection
8952 Correct radial lens distortion
8954 This filter can be used to correct for radial distortion as can result from the use
8955 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
8956 one can use tools available for example as part of opencv or simply trial-and-error.
8957 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
8958 and extract the k1 and k2 coefficients from the resulting matrix.
8960 Note that effectively the same filter is available in the open-source tools Krita and
8961 Digikam from the KDE project.
8963 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
8964 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
8965 brightness distribution, so you may want to use both filters together in certain
8966 cases, though you will have to take care of ordering, i.e. whether vignetting should
8967 be applied before or after lens correction.
8971 The filter accepts the following options:
8975 Relative x-coordinate of the focal point of the image, and thereby the center of the
8976 distortion. This value has a range [0,1] and is expressed as fractions of the image
8979 Relative y-coordinate of the focal point of the image, and thereby the center of the
8980 distortion. This value has a range [0,1] and is expressed as fractions of the image
8983 Coefficient of the quadratic correction term. 0.5 means no correction.
8985 Coefficient of the double quadratic correction term. 0.5 means no correction.
8988 The formula that generates the correction is:
8990 @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)
8992 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
8993 distances from the focal point in the source and target images, respectively.
8999 The filter accepts the following options:
9003 Set the number of loops.
9006 Set maximal size in number of frames.
9009 Set first frame of loop.
9015 Apply a 3D LUT to an input video.
9017 The filter accepts the following options:
9021 Set the 3D LUT file name.
9023 Currently supported formats:
9035 Select interpolation mode.
9037 Available values are:
9041 Use values from the nearest defined point.
9043 Interpolate values using the 8 points defining a cube.
9045 Interpolate values using a tetrahedron.
9049 @section lut, lutrgb, lutyuv
9051 Compute a look-up table for binding each pixel component input value
9052 to an output value, and apply it to the input video.
9054 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9055 to an RGB input video.
9057 These filters accept the following parameters:
9060 set first pixel component expression
9062 set second pixel component expression
9064 set third pixel component expression
9066 set fourth pixel component expression, corresponds to the alpha component
9069 set red component expression
9071 set green component expression
9073 set blue component expression
9075 alpha component expression
9078 set Y/luminance component expression
9080 set U/Cb component expression
9082 set V/Cr component expression
9085 Each of them specifies the expression to use for computing the lookup table for
9086 the corresponding pixel component values.
9088 The exact component associated to each of the @var{c*} options depends on the
9091 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9092 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9094 The expressions can contain the following constants and functions:
9099 The input width and height.
9102 The input value for the pixel component.
9105 The input value, clipped to the @var{minval}-@var{maxval} range.
9108 The maximum value for the pixel component.
9111 The minimum value for the pixel component.
9114 The negated value for the pixel component value, clipped to the
9115 @var{minval}-@var{maxval} range; it corresponds to the expression
9116 "maxval-clipval+minval".
9119 The computed value in @var{val}, clipped to the
9120 @var{minval}-@var{maxval} range.
9122 @item gammaval(gamma)
9123 The computed gamma correction value of the pixel component value,
9124 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9126 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9130 All expressions default to "val".
9132 @subsection Examples
9138 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9139 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9142 The above is the same as:
9144 lutrgb="r=negval:g=negval:b=negval"
9145 lutyuv="y=negval:u=negval:v=negval"
9155 Remove chroma components, turning the video into a graytone image:
9157 lutyuv="u=128:v=128"
9161 Apply a luma burning effect:
9167 Remove green and blue components:
9173 Set a constant alpha channel value on input:
9175 format=rgba,lutrgb=a="maxval-minval/2"
9179 Correct luminance gamma by a factor of 0.5:
9181 lutyuv=y=gammaval(0.5)
9185 Discard least significant bits of luma:
9187 lutyuv=y='bitand(val, 128+64+32)'
9191 Technicolor like effect:
9193 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9197 @section maskedmerge
9199 Merge the first input stream with the second input stream using per pixel
9200 weights in the third input stream.
9202 A value of 0 in the third stream pixel component means that pixel component
9203 from first stream is returned unchanged, while maximum value (eg. 255 for
9204 8-bit videos) means that pixel component from second stream is returned
9205 unchanged. Intermediate values define the amount of merging between both
9206 input stream's pixel components.
9208 This filter accepts the following options:
9211 Set which planes will be processed as bitmap, unprocessed planes will be
9212 copied from first stream.
9213 By default value 0xf, all planes will be processed.
9218 Apply motion-compensation deinterlacing.
9220 It needs one field per frame as input and must thus be used together
9221 with yadif=1/3 or equivalent.
9223 This filter accepts the following options:
9226 Set the deinterlacing mode.
9228 It accepts one of the following values:
9233 use iterative motion estimation
9235 like @samp{slow}, but use multiple reference frames.
9237 Default value is @samp{fast}.
9240 Set the picture field parity assumed for the input video. It must be
9241 one of the following values:
9245 assume top field first
9247 assume bottom field first
9250 Default value is @samp{bff}.
9253 Set per-block quantization parameter (QP) used by the internal
9256 Higher values should result in a smoother motion vector field but less
9257 optimal individual vectors. Default value is 1.
9260 @section mergeplanes
9262 Merge color channel components from several video streams.
9264 The filter accepts up to 4 input streams, and merge selected input
9265 planes to the output video.
9267 This filter accepts the following options:
9270 Set input to output plane mapping. Default is @code{0}.
9272 The mappings is specified as a bitmap. It should be specified as a
9273 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9274 mapping for the first plane of the output stream. 'A' sets the number of
9275 the input stream to use (from 0 to 3), and 'a' the plane number of the
9276 corresponding input to use (from 0 to 3). The rest of the mappings is
9277 similar, 'Bb' describes the mapping for the output stream second
9278 plane, 'Cc' describes the mapping for the output stream third plane and
9279 'Dd' describes the mapping for the output stream fourth plane.
9282 Set output pixel format. Default is @code{yuva444p}.
9285 @subsection Examples
9289 Merge three gray video streams of same width and height into single video stream:
9291 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9295 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9297 [a0][a1]mergeplanes=0x00010210:yuva444p
9301 Swap Y and A plane in yuva444p stream:
9303 format=yuva444p,mergeplanes=0x03010200:yuva444p
9307 Swap U and V plane in yuv420p stream:
9309 format=yuv420p,mergeplanes=0x000201:yuv420p
9313 Cast a rgb24 clip to yuv444p:
9315 format=rgb24,mergeplanes=0x000102:yuv444p
9321 Drop frames that do not differ greatly from the previous frame in
9322 order to reduce frame rate.
9324 The main use of this filter is for very-low-bitrate encoding
9325 (e.g. streaming over dialup modem), but it could in theory be used for
9326 fixing movies that were inverse-telecined incorrectly.
9328 A description of the accepted options follows.
9332 Set the maximum number of consecutive frames which can be dropped (if
9333 positive), or the minimum interval between dropped frames (if
9334 negative). If the value is 0, the frame is dropped unregarding the
9335 number of previous sequentially dropped frames.
9342 Set the dropping threshold values.
9344 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9345 represent actual pixel value differences, so a threshold of 64
9346 corresponds to 1 unit of difference for each pixel, or the same spread
9347 out differently over the block.
9349 A frame is a candidate for dropping if no 8x8 blocks differ by more
9350 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9351 meaning the whole image) differ by more than a threshold of @option{lo}.
9353 Default value for @option{hi} is 64*12, default value for @option{lo} is
9354 64*5, and default value for @option{frac} is 0.33.
9362 It accepts an integer in input; if non-zero it negates the
9363 alpha component (if available). The default value in input is 0.
9367 Deinterlace video using neural network edge directed interpolation.
9369 This filter accepts the following options:
9373 Mandatory option, without binary file filter can not work.
9374 Currently file can be found here:
9375 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9378 Set which frames to deinterlace, by default it is @code{all}.
9379 Can be @code{all} or @code{interlaced}.
9382 Set mode of operation.
9384 Can be one of the following:
9388 Use frame flags, both fields.
9390 Use frame flags, single field.
9394 Use bottom field only.
9396 Use both fields, top first.
9398 Use both fields, bottom first.
9402 Set which planes to process, by default filter process all frames.
9405 Set size of local neighborhood around each pixel, used by the predictor neural
9408 Can be one of the following:
9421 Set the number of neurons in predicctor neural network.
9422 Can be one of the following:
9433 Controls the number of different neural network predictions that are blended
9434 together to compute the final output value. Can be @code{fast}, default or
9438 Set which set of weights to use in the predictor.
9439 Can be one of the following:
9443 weights trained to minimize absolute error
9445 weights trained to minimize squared error
9449 Controls whether or not the prescreener neural network is used to decide
9450 which pixels should be processed by the predictor neural network and which
9451 can be handled by simple cubic interpolation.
9452 The prescreener is trained to know whether cubic interpolation will be
9453 sufficient for a pixel or whether it should be predicted by the predictor nn.
9454 The computational complexity of the prescreener nn is much less than that of
9455 the predictor nn. Since most pixels can be handled by cubic interpolation,
9456 using the prescreener generally results in much faster processing.
9457 The prescreener is pretty accurate, so the difference between using it and not
9458 using it is almost always unnoticeable.
9460 Can be one of the following:
9468 Default is @code{new}.
9471 Set various debugging flags.
9476 Force libavfilter not to use any of the specified pixel formats for the
9477 input to the next filter.
9479 It accepts the following parameters:
9483 A '|'-separated list of pixel format names, such as
9484 apix_fmts=yuv420p|monow|rgb24".
9488 @subsection Examples
9492 Force libavfilter to use a format different from @var{yuv420p} for the
9493 input to the vflip filter:
9495 noformat=pix_fmts=yuv420p,vflip
9499 Convert the input video to any of the formats not contained in the list:
9501 noformat=yuv420p|yuv444p|yuv410p
9507 Add noise on video input frame.
9509 The filter accepts the following options:
9517 Set noise seed for specific pixel component or all pixel components in case
9518 of @var{all_seed}. Default value is @code{123457}.
9520 @item all_strength, alls
9521 @item c0_strength, c0s
9522 @item c1_strength, c1s
9523 @item c2_strength, c2s
9524 @item c3_strength, c3s
9525 Set noise strength for specific pixel component or all pixel components in case
9526 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9528 @item all_flags, allf
9533 Set pixel component flags or set flags for all components if @var{all_flags}.
9534 Available values for component flags are:
9537 averaged temporal noise (smoother)
9539 mix random noise with a (semi)regular pattern
9541 temporal noise (noise pattern changes between frames)
9543 uniform noise (gaussian otherwise)
9547 @subsection Examples
9549 Add temporal and uniform noise to input video:
9551 noise=alls=20:allf=t+u
9556 Pass the video source unchanged to the output.
9559 Optical Character Recognition
9561 This filter uses Tesseract for optical character recognition.
9563 It accepts the following options:
9567 Set datapath to tesseract data. Default is to use whatever was
9568 set at installation.
9571 Set language, default is "eng".
9574 Set character whitelist.
9577 Set character blacklist.
9580 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9584 Apply a video transform using libopencv.
9586 To enable this filter, install the libopencv library and headers and
9587 configure FFmpeg with @code{--enable-libopencv}.
9589 It accepts the following parameters:
9594 The name of the libopencv filter to apply.
9597 The parameters to pass to the libopencv filter. If not specified, the default
9602 Refer to the official libopencv documentation for more precise
9604 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9606 Several libopencv filters are supported; see the following subsections.
9611 Dilate an image by using a specific structuring element.
9612 It corresponds to the libopencv function @code{cvDilate}.
9614 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
9616 @var{struct_el} represents a structuring element, and has the syntax:
9617 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
9619 @var{cols} and @var{rows} represent the number of columns and rows of
9620 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
9621 point, and @var{shape} the shape for the structuring element. @var{shape}
9622 must be "rect", "cross", "ellipse", or "custom".
9624 If the value for @var{shape} is "custom", it must be followed by a
9625 string of the form "=@var{filename}". The file with name
9626 @var{filename} is assumed to represent a binary image, with each
9627 printable character corresponding to a bright pixel. When a custom
9628 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
9629 or columns and rows of the read file are assumed instead.
9631 The default value for @var{struct_el} is "3x3+0x0/rect".
9633 @var{nb_iterations} specifies the number of times the transform is
9634 applied to the image, and defaults to 1.
9638 # Use the default values
9641 # Dilate using a structuring element with a 5x5 cross, iterating two times
9642 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
9644 # Read the shape from the file diamond.shape, iterating two times.
9645 # The file diamond.shape may contain a pattern of characters like this
9651 # The specified columns and rows are ignored
9652 # but the anchor point coordinates are not
9653 ocv=dilate:0x0+2x2/custom=diamond.shape|2
9658 Erode an image by using a specific structuring element.
9659 It corresponds to the libopencv function @code{cvErode}.
9661 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
9662 with the same syntax and semantics as the @ref{dilate} filter.
9666 Smooth the input video.
9668 The filter takes the following parameters:
9669 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
9671 @var{type} is the type of smooth filter to apply, and must be one of
9672 the following values: "blur", "blur_no_scale", "median", "gaussian",
9673 or "bilateral". The default value is "gaussian".
9675 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
9676 depend on the smooth type. @var{param1} and
9677 @var{param2} accept integer positive values or 0. @var{param3} and
9678 @var{param4} accept floating point values.
9680 The default value for @var{param1} is 3. The default value for the
9681 other parameters is 0.
9683 These parameters correspond to the parameters assigned to the
9684 libopencv function @code{cvSmooth}.
9689 Overlay one video on top of another.
9691 It takes two inputs and has one output. The first input is the "main"
9692 video on which the second input is overlaid.
9694 It accepts the following parameters:
9696 A description of the accepted options follows.
9701 Set the expression for the x and y coordinates of the overlaid video
9702 on the main video. Default value is "0" for both expressions. In case
9703 the expression is invalid, it is set to a huge value (meaning that the
9704 overlay will not be displayed within the output visible area).
9707 The action to take when EOF is encountered on the secondary input; it accepts
9708 one of the following values:
9712 Repeat the last frame (the default).
9716 Pass the main input through.
9720 Set when the expressions for @option{x}, and @option{y} are evaluated.
9722 It accepts the following values:
9725 only evaluate expressions once during the filter initialization or
9726 when a command is processed
9729 evaluate expressions for each incoming frame
9732 Default value is @samp{frame}.
9735 If set to 1, force the output to terminate when the shortest input
9736 terminates. Default value is 0.
9739 Set the format for the output video.
9741 It accepts the following values:
9756 Default value is @samp{yuv420}.
9758 @item rgb @emph{(deprecated)}
9759 If set to 1, force the filter to accept inputs in the RGB
9760 color space. Default value is 0. This option is deprecated, use
9761 @option{format} instead.
9764 If set to 1, force the filter to draw the last overlay frame over the
9765 main input until the end of the stream. A value of 0 disables this
9766 behavior. Default value is 1.
9769 The @option{x}, and @option{y} expressions can contain the following
9775 The main input width and height.
9779 The overlay input width and height.
9783 The computed values for @var{x} and @var{y}. They are evaluated for
9788 horizontal and vertical chroma subsample values of the output
9789 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
9793 the number of input frame, starting from 0
9796 the position in the file of the input frame, NAN if unknown
9799 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
9803 Note that the @var{n}, @var{pos}, @var{t} variables are available only
9804 when evaluation is done @emph{per frame}, and will evaluate to NAN
9805 when @option{eval} is set to @samp{init}.
9807 Be aware that frames are taken from each input video in timestamp
9808 order, hence, if their initial timestamps differ, it is a good idea
9809 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
9810 have them begin in the same zero timestamp, as the example for
9811 the @var{movie} filter does.
9813 You can chain together more overlays but you should test the
9814 efficiency of such approach.
9816 @subsection Commands
9818 This filter supports the following commands:
9822 Modify the x and y of the overlay input.
9823 The command accepts the same syntax of the corresponding option.
9825 If the specified expression is not valid, it is kept at its current
9829 @subsection Examples
9833 Draw the overlay at 10 pixels from the bottom right corner of the main
9836 overlay=main_w-overlay_w-10:main_h-overlay_h-10
9839 Using named options the example above becomes:
9841 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
9845 Insert a transparent PNG logo in the bottom left corner of the input,
9846 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
9848 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
9852 Insert 2 different transparent PNG logos (second logo on bottom
9853 right corner) using the @command{ffmpeg} tool:
9855 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
9859 Add a transparent color layer on top of the main video; @code{WxH}
9860 must specify the size of the main input to the overlay filter:
9862 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
9866 Play an original video and a filtered version (here with the deshake
9867 filter) side by side using the @command{ffplay} tool:
9869 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
9872 The above command is the same as:
9874 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
9878 Make a sliding overlay appearing from the left to the right top part of the
9879 screen starting since time 2:
9881 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
9885 Compose output by putting two input videos side to side:
9887 ffmpeg -i left.avi -i right.avi -filter_complex "
9888 nullsrc=size=200x100 [background];
9889 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
9890 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
9891 [background][left] overlay=shortest=1 [background+left];
9892 [background+left][right] overlay=shortest=1:x=100 [left+right]
9897 Mask 10-20 seconds of a video by applying the delogo filter to a section
9899 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
9900 -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]'
9905 Chain several overlays in cascade:
9907 nullsrc=s=200x200 [bg];
9908 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
9909 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
9910 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
9911 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
9912 [in3] null, [mid2] overlay=100:100 [out0]
9919 Apply Overcomplete Wavelet denoiser.
9921 The filter accepts the following options:
9927 Larger depth values will denoise lower frequency components more, but
9928 slow down filtering.
9930 Must be an int in the range 8-16, default is @code{8}.
9932 @item luma_strength, ls
9935 Must be a double value in the range 0-1000, default is @code{1.0}.
9937 @item chroma_strength, cs
9938 Set chroma strength.
9940 Must be a double value in the range 0-1000, default is @code{1.0}.
9946 Add paddings to the input image, and place the original input at the
9947 provided @var{x}, @var{y} coordinates.
9949 It accepts the following parameters:
9954 Specify an expression for the size of the output image with the
9955 paddings added. If the value for @var{width} or @var{height} is 0, the
9956 corresponding input size is used for the output.
9958 The @var{width} expression can reference the value set by the
9959 @var{height} expression, and vice versa.
9961 The default value of @var{width} and @var{height} is 0.
9965 Specify the offsets to place the input image at within the padded area,
9966 with respect to the top/left border of the output image.
9968 The @var{x} expression can reference the value set by the @var{y}
9969 expression, and vice versa.
9971 The default value of @var{x} and @var{y} is 0.
9974 Specify the color of the padded area. For the syntax of this option,
9975 check the "Color" section in the ffmpeg-utils manual.
9977 The default value of @var{color} is "black".
9980 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
9981 options are expressions containing the following constants:
9986 The input video width and height.
9990 These are the same as @var{in_w} and @var{in_h}.
9994 The output width and height (the size of the padded area), as
9995 specified by the @var{width} and @var{height} expressions.
9999 These are the same as @var{out_w} and @var{out_h}.
10003 The x and y offsets as specified by the @var{x} and @var{y}
10004 expressions, or NAN if not yet specified.
10007 same as @var{iw} / @var{ih}
10010 input sample aspect ratio
10013 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10017 The horizontal and vertical chroma subsample values. For example for the
10018 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10021 @subsection Examples
10025 Add paddings with the color "violet" to the input video. The output video
10026 size is 640x480, and the top-left corner of the input video is placed at
10029 pad=640:480:0:40:violet
10032 The example above is equivalent to the following command:
10034 pad=width=640:height=480:x=0:y=40:color=violet
10038 Pad the input to get an output with dimensions increased by 3/2,
10039 and put the input video at the center of the padded area:
10041 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10045 Pad the input to get a squared output with size equal to the maximum
10046 value between the input width and height, and put the input video at
10047 the center of the padded area:
10049 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10053 Pad the input to get a final w/h ratio of 16:9:
10055 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10059 In case of anamorphic video, in order to set the output display aspect
10060 correctly, it is necessary to use @var{sar} in the expression,
10061 according to the relation:
10063 (ih * X / ih) * sar = output_dar
10064 X = output_dar / sar
10067 Thus the previous example needs to be modified to:
10069 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10073 Double the output size and put the input video in the bottom-right
10074 corner of the output padded area:
10076 pad="2*iw:2*ih:ow-iw:oh-ih"
10080 @anchor{palettegen}
10081 @section palettegen
10083 Generate one palette for a whole video stream.
10085 It accepts the following options:
10089 Set the maximum number of colors to quantize in the palette.
10090 Note: the palette will still contain 256 colors; the unused palette entries
10093 @item reserve_transparent
10094 Create a palette of 255 colors maximum and reserve the last one for
10095 transparency. Reserving the transparency color is useful for GIF optimization.
10096 If not set, the maximum of colors in the palette will be 256. You probably want
10097 to disable this option for a standalone image.
10101 Set statistics mode.
10103 It accepts the following values:
10106 Compute full frame histograms.
10108 Compute histograms only for the part that differs from previous frame. This
10109 might be relevant to give more importance to the moving part of your input if
10110 the background is static.
10113 Default value is @var{full}.
10116 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10117 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10118 color quantization of the palette. This information is also visible at
10119 @var{info} logging level.
10121 @subsection Examples
10125 Generate a representative palette of a given video using @command{ffmpeg}:
10127 ffmpeg -i input.mkv -vf palettegen palette.png
10131 @section paletteuse
10133 Use a palette to downsample an input video stream.
10135 The filter takes two inputs: one video stream and a palette. The palette must
10136 be a 256 pixels image.
10138 It accepts the following options:
10142 Select dithering mode. Available algorithms are:
10145 Ordered 8x8 bayer dithering (deterministic)
10147 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10148 Note: this dithering is sometimes considered "wrong" and is included as a
10150 @item floyd_steinberg
10151 Floyd and Steingberg dithering (error diffusion)
10153 Frankie Sierra dithering v2 (error diffusion)
10155 Frankie Sierra dithering v2 "Lite" (error diffusion)
10158 Default is @var{sierra2_4a}.
10161 When @var{bayer} dithering is selected, this option defines the scale of the
10162 pattern (how much the crosshatch pattern is visible). A low value means more
10163 visible pattern for less banding, and higher value means less visible pattern
10164 at the cost of more banding.
10166 The option must be an integer value in the range [0,5]. Default is @var{2}.
10169 If set, define the zone to process
10173 Only the changing rectangle will be reprocessed. This is similar to GIF
10174 cropping/offsetting compression mechanism. This option can be useful for speed
10175 if only a part of the image is changing, and has use cases such as limiting the
10176 scope of the error diffusal @option{dither} to the rectangle that bounds the
10177 moving scene (it leads to more deterministic output if the scene doesn't change
10178 much, and as a result less moving noise and better GIF compression).
10181 Default is @var{none}.
10184 @subsection Examples
10188 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10189 using @command{ffmpeg}:
10191 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10195 @section perspective
10197 Correct perspective of video not recorded perpendicular to the screen.
10199 A description of the accepted parameters follows.
10210 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10211 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10212 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10213 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10214 then the corners of the source will be sent to the specified coordinates.
10216 The expressions can use the following variables:
10221 the width and height of video frame.
10225 Output frame count.
10228 @item interpolation
10229 Set interpolation for perspective correction.
10231 It accepts the following values:
10237 Default value is @samp{linear}.
10240 Set interpretation of coordinate options.
10242 It accepts the following values:
10246 Send point in the source specified by the given coordinates to
10247 the corners of the destination.
10249 @item 1, destination
10251 Send the corners of the source to the point in the destination specified
10252 by the given coordinates.
10254 Default value is @samp{source}.
10258 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10260 It accepts the following values:
10263 only evaluate expressions once during the filter initialization or
10264 when a command is processed
10267 evaluate expressions for each incoming frame
10270 Default value is @samp{init}.
10275 Delay interlaced video by one field time so that the field order changes.
10277 The intended use is to fix PAL movies that have been captured with the
10278 opposite field order to the film-to-video transfer.
10280 A description of the accepted parameters follows.
10286 It accepts the following values:
10289 Capture field order top-first, transfer bottom-first.
10290 Filter will delay the bottom field.
10293 Capture field order bottom-first, transfer top-first.
10294 Filter will delay the top field.
10297 Capture and transfer with the same field order. This mode only exists
10298 for the documentation of the other options to refer to, but if you
10299 actually select it, the filter will faithfully do nothing.
10302 Capture field order determined automatically by field flags, transfer
10304 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10305 basis using field flags. If no field information is available,
10306 then this works just like @samp{u}.
10309 Capture unknown or varying, transfer opposite.
10310 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10311 analyzing the images and selecting the alternative that produces best
10312 match between the fields.
10315 Capture top-first, transfer unknown or varying.
10316 Filter selects among @samp{t} and @samp{p} using image analysis.
10319 Capture bottom-first, transfer unknown or varying.
10320 Filter selects among @samp{b} and @samp{p} using image analysis.
10323 Capture determined by field flags, transfer unknown or varying.
10324 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10325 image analysis. If no field information is available, then this works just
10326 like @samp{U}. This is the default mode.
10329 Both capture and transfer unknown or varying.
10330 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10334 @section pixdesctest
10336 Pixel format descriptor test filter, mainly useful for internal
10337 testing. The output video should be equal to the input video.
10341 format=monow, pixdesctest
10344 can be used to test the monowhite pixel format descriptor definition.
10348 Enable the specified chain of postprocessing subfilters using libpostproc. This
10349 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10350 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10351 Each subfilter and some options have a short and a long name that can be used
10352 interchangeably, i.e. dr/dering are the same.
10354 The filters accept the following options:
10358 Set postprocessing subfilters string.
10361 All subfilters share common options to determine their scope:
10365 Honor the quality commands for this subfilter.
10368 Do chrominance filtering, too (default).
10371 Do luminance filtering only (no chrominance).
10374 Do chrominance filtering only (no luminance).
10377 These options can be appended after the subfilter name, separated by a '|'.
10379 Available subfilters are:
10382 @item hb/hdeblock[|difference[|flatness]]
10383 Horizontal deblocking filter
10386 Difference factor where higher values mean more deblocking (default: @code{32}).
10388 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10391 @item vb/vdeblock[|difference[|flatness]]
10392 Vertical deblocking filter
10395 Difference factor where higher values mean more deblocking (default: @code{32}).
10397 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10400 @item ha/hadeblock[|difference[|flatness]]
10401 Accurate horizontal deblocking filter
10404 Difference factor where higher values mean more deblocking (default: @code{32}).
10406 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10409 @item va/vadeblock[|difference[|flatness]]
10410 Accurate vertical deblocking filter
10413 Difference factor where higher values mean more deblocking (default: @code{32}).
10415 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10419 The horizontal and vertical deblocking filters share the difference and
10420 flatness values so you cannot set different horizontal and vertical
10424 @item h1/x1hdeblock
10425 Experimental horizontal deblocking filter
10427 @item v1/x1vdeblock
10428 Experimental vertical deblocking filter
10433 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10436 larger -> stronger filtering
10438 larger -> stronger filtering
10440 larger -> stronger filtering
10443 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10446 Stretch luminance to @code{0-255}.
10449 @item lb/linblenddeint
10450 Linear blend deinterlacing filter that deinterlaces the given block by
10451 filtering all lines with a @code{(1 2 1)} filter.
10453 @item li/linipoldeint
10454 Linear interpolating deinterlacing filter that deinterlaces the given block by
10455 linearly interpolating every second line.
10457 @item ci/cubicipoldeint
10458 Cubic interpolating deinterlacing filter deinterlaces the given block by
10459 cubically interpolating every second line.
10461 @item md/mediandeint
10462 Median deinterlacing filter that deinterlaces the given block by applying a
10463 median filter to every second line.
10465 @item fd/ffmpegdeint
10466 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10467 second line with a @code{(-1 4 2 4 -1)} filter.
10470 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10471 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10473 @item fq/forceQuant[|quantizer]
10474 Overrides the quantizer table from the input with the constant quantizer you
10482 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10485 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10488 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10491 @subsection Examples
10495 Apply horizontal and vertical deblocking, deringing and automatic
10496 brightness/contrast:
10502 Apply default filters without brightness/contrast correction:
10508 Apply default filters and temporal denoiser:
10510 pp=default/tmpnoise|1|2|3
10514 Apply deblocking on luminance only, and switch vertical deblocking on or off
10515 automatically depending on available CPU time:
10522 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10523 similar to spp = 6 with 7 point DCT, where only the center sample is
10526 The filter accepts the following options:
10530 Force a constant quantization parameter. It accepts an integer in range
10531 0 to 63. If not set, the filter will use the QP from the video stream
10535 Set thresholding mode. Available modes are:
10539 Set hard thresholding.
10541 Set soft thresholding (better de-ringing effect, but likely blurrier).
10543 Set medium thresholding (good results, default).
10549 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10550 Ratio) between two input videos.
10552 This filter takes in input two input videos, the first input is
10553 considered the "main" source and is passed unchanged to the
10554 output. The second input is used as a "reference" video for computing
10557 Both video inputs must have the same resolution and pixel format for
10558 this filter to work correctly. Also it assumes that both inputs
10559 have the same number of frames, which are compared one by one.
10561 The obtained average PSNR is printed through the logging system.
10563 The filter stores the accumulated MSE (mean squared error) of each
10564 frame, and at the end of the processing it is averaged across all frames
10565 equally, and the following formula is applied to obtain the PSNR:
10568 PSNR = 10*log10(MAX^2/MSE)
10571 Where MAX is the average of the maximum values of each component of the
10574 The description of the accepted parameters follows.
10577 @item stats_file, f
10578 If specified the filter will use the named file to save the PSNR of
10579 each individual frame. When filename equals "-" the data is sent to
10582 @item stats_version
10583 Specifies which version of the stats file format to use. Details of
10584 each format are written below.
10585 Default value is 1.
10588 The file printed if @var{stats_file} is selected, contains a sequence of
10589 key/value pairs of the form @var{key}:@var{value} for each compared
10592 If a @var{stats_version} greater than 1 is specified, a header line precedes
10593 the list of per-frame-pair stats, with key value pairs following the frame
10594 format with the following parameters:
10597 @item psnr_log_version
10598 The version of the log file format. Will match @var{stats_version}.
10601 A comma separated list of the per-frame-pair parameters included in
10605 A description of each shown per-frame-pair parameter follows:
10609 sequential number of the input frame, starting from 1
10612 Mean Square Error pixel-by-pixel average difference of the compared
10613 frames, averaged over all the image components.
10615 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
10616 Mean Square Error pixel-by-pixel average difference of the compared
10617 frames for the component specified by the suffix.
10619 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
10620 Peak Signal to Noise ratio of the compared frames for the component
10621 specified by the suffix.
10626 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
10627 [main][ref] psnr="stats_file=stats.log" [out]
10630 On this example the input file being processed is compared with the
10631 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
10632 is stored in @file{stats.log}.
10637 Pulldown reversal (inverse telecine) filter, capable of handling mixed
10638 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
10641 The pullup filter is designed to take advantage of future context in making
10642 its decisions. This filter is stateless in the sense that it does not lock
10643 onto a pattern to follow, but it instead looks forward to the following
10644 fields in order to identify matches and rebuild progressive frames.
10646 To produce content with an even framerate, insert the fps filter after
10647 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
10648 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
10650 The filter accepts the following options:
10657 These options set the amount of "junk" to ignore at the left, right, top, and
10658 bottom of the image, respectively. Left and right are in units of 8 pixels,
10659 while top and bottom are in units of 2 lines.
10660 The default is 8 pixels on each side.
10663 Set the strict breaks. Setting this option to 1 will reduce the chances of
10664 filter generating an occasional mismatched frame, but it may also cause an
10665 excessive number of frames to be dropped during high motion sequences.
10666 Conversely, setting it to -1 will make filter match fields more easily.
10667 This may help processing of video where there is slight blurring between
10668 the fields, but may also cause there to be interlaced frames in the output.
10669 Default value is @code{0}.
10672 Set the metric plane to use. It accepts the following values:
10678 Use chroma blue plane.
10681 Use chroma red plane.
10684 This option may be set to use chroma plane instead of the default luma plane
10685 for doing filter's computations. This may improve accuracy on very clean
10686 source material, but more likely will decrease accuracy, especially if there
10687 is chroma noise (rainbow effect) or any grayscale video.
10688 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
10689 load and make pullup usable in realtime on slow machines.
10692 For best results (without duplicated frames in the output file) it is
10693 necessary to change the output frame rate. For example, to inverse
10694 telecine NTSC input:
10696 ffmpeg -i input -vf pullup -r 24000/1001 ...
10701 Change video quantization parameters (QP).
10703 The filter accepts the following option:
10707 Set expression for quantization parameter.
10710 The expression is evaluated through the eval API and can contain, among others,
10711 the following constants:
10715 1 if index is not 129, 0 otherwise.
10718 Sequentional index starting from -129 to 128.
10721 @subsection Examples
10725 Some equation like:
10733 Flush video frames from internal cache of frames into a random order.
10734 No frame is discarded.
10735 Inspired by @ref{frei0r} nervous filter.
10739 Set size in number of frames of internal cache, in range from @code{2} to
10740 @code{512}. Default is @code{30}.
10743 Set seed for random number generator, must be an integer included between
10744 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
10745 less than @code{0}, the filter will try to use a good random seed on a
10751 Read vertical interval timecode (VITC) information from the top lines of a
10754 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
10755 timecode value, if a valid timecode has been detected. Further metadata key
10756 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
10757 timecode data has been found or not.
10759 This filter accepts the following options:
10763 Set the maximum number of lines to scan for VITC data. If the value is set to
10764 @code{-1} the full video frame is scanned. Default is @code{45}.
10767 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
10768 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
10771 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
10772 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
10775 @subsection Examples
10779 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
10780 draw @code{--:--:--:--} as a placeholder:
10782 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
10788 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
10790 Destination pixel at position (X, Y) will be picked from source (x, y) position
10791 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
10792 value for pixel will be used for destination pixel.
10794 Xmap and Ymap input video streams must be of same dimensions. Output video stream
10795 will have Xmap/Ymap video stream dimensions.
10796 Xmap and Ymap input video streams are 16bit depth, single channel.
10798 @section removegrain
10800 The removegrain filter is a spatial denoiser for progressive video.
10804 Set mode for the first plane.
10807 Set mode for the second plane.
10810 Set mode for the third plane.
10813 Set mode for the fourth plane.
10816 Range of mode is from 0 to 24. Description of each mode follows:
10820 Leave input plane unchanged. Default.
10823 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
10826 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
10829 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
10832 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
10833 This is equivalent to a median filter.
10836 Line-sensitive clipping giving the minimal change.
10839 Line-sensitive clipping, intermediate.
10842 Line-sensitive clipping, intermediate.
10845 Line-sensitive clipping, intermediate.
10848 Line-sensitive clipping on a line where the neighbours pixels are the closest.
10851 Replaces the target pixel with the closest neighbour.
10854 [1 2 1] horizontal and vertical kernel blur.
10860 Bob mode, interpolates top field from the line where the neighbours
10861 pixels are the closest.
10864 Bob mode, interpolates bottom field from the line where the neighbours
10865 pixels are the closest.
10868 Bob mode, interpolates top field. Same as 13 but with a more complicated
10869 interpolation formula.
10872 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
10873 interpolation formula.
10876 Clips the pixel with the minimum and maximum of respectively the maximum and
10877 minimum of each pair of opposite neighbour pixels.
10880 Line-sensitive clipping using opposite neighbours whose greatest distance from
10881 the current pixel is minimal.
10884 Replaces the pixel with the average of its 8 neighbours.
10887 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
10890 Clips pixels using the averages of opposite neighbour.
10893 Same as mode 21 but simpler and faster.
10896 Small edge and halo removal, but reputed useless.
10902 @section removelogo
10904 Suppress a TV station logo, using an image file to determine which
10905 pixels comprise the logo. It works by filling in the pixels that
10906 comprise the logo with neighboring pixels.
10908 The filter accepts the following options:
10912 Set the filter bitmap file, which can be any image format supported by
10913 libavformat. The width and height of the image file must match those of the
10914 video stream being processed.
10917 Pixels in the provided bitmap image with a value of zero are not
10918 considered part of the logo, non-zero pixels are considered part of
10919 the logo. If you use white (255) for the logo and black (0) for the
10920 rest, you will be safe. For making the filter bitmap, it is
10921 recommended to take a screen capture of a black frame with the logo
10922 visible, and then using a threshold filter followed by the erode
10923 filter once or twice.
10925 If needed, little splotches can be fixed manually. Remember that if
10926 logo pixels are not covered, the filter quality will be much
10927 reduced. Marking too many pixels as part of the logo does not hurt as
10928 much, but it will increase the amount of blurring needed to cover over
10929 the image and will destroy more information than necessary, and extra
10930 pixels will slow things down on a large logo.
10932 @section repeatfields
10934 This filter uses the repeat_field flag from the Video ES headers and hard repeats
10935 fields based on its value.
10939 Reverse a video clip.
10941 Warning: This filter requires memory to buffer the entire clip, so trimming
10944 @subsection Examples
10948 Take the first 5 seconds of a clip, and reverse it.
10956 Rotate video by an arbitrary angle expressed in radians.
10958 The filter accepts the following options:
10960 A description of the optional parameters follows.
10963 Set an expression for the angle by which to rotate the input video
10964 clockwise, expressed as a number of radians. A negative value will
10965 result in a counter-clockwise rotation. By default it is set to "0".
10967 This expression is evaluated for each frame.
10970 Set the output width expression, default value is "iw".
10971 This expression is evaluated just once during configuration.
10974 Set the output height expression, default value is "ih".
10975 This expression is evaluated just once during configuration.
10978 Enable bilinear interpolation if set to 1, a value of 0 disables
10979 it. Default value is 1.
10982 Set the color used to fill the output area not covered by the rotated
10983 image. For the general syntax of this option, check the "Color" section in the
10984 ffmpeg-utils manual. If the special value "none" is selected then no
10985 background is printed (useful for example if the background is never shown).
10987 Default value is "black".
10990 The expressions for the angle and the output size can contain the
10991 following constants and functions:
10995 sequential number of the input frame, starting from 0. It is always NAN
10996 before the first frame is filtered.
10999 time in seconds of the input frame, it is set to 0 when the filter is
11000 configured. It is always NAN before the first frame is filtered.
11004 horizontal and vertical chroma subsample values. For example for the
11005 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11009 the input video width and height
11013 the output width and height, that is the size of the padded area as
11014 specified by the @var{width} and @var{height} expressions
11018 the minimal width/height required for completely containing the input
11019 video rotated by @var{a} radians.
11021 These are only available when computing the @option{out_w} and
11022 @option{out_h} expressions.
11025 @subsection Examples
11029 Rotate the input by PI/6 radians clockwise:
11035 Rotate the input by PI/6 radians counter-clockwise:
11041 Rotate the input by 45 degrees clockwise:
11047 Apply a constant rotation with period T, starting from an angle of PI/3:
11049 rotate=PI/3+2*PI*t/T
11053 Make the input video rotation oscillating with a period of T
11054 seconds and an amplitude of A radians:
11056 rotate=A*sin(2*PI/T*t)
11060 Rotate the video, output size is chosen so that the whole rotating
11061 input video is always completely contained in the output:
11063 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11067 Rotate the video, reduce the output size so that no background is ever
11070 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11074 @subsection Commands
11076 The filter supports the following commands:
11080 Set the angle expression.
11081 The command accepts the same syntax of the corresponding option.
11083 If the specified expression is not valid, it is kept at its current
11089 Apply Shape Adaptive Blur.
11091 The filter accepts the following options:
11094 @item luma_radius, lr
11095 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11096 value is 1.0. A greater value will result in a more blurred image, and
11097 in slower processing.
11099 @item luma_pre_filter_radius, lpfr
11100 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11103 @item luma_strength, ls
11104 Set luma maximum difference between pixels to still be considered, must
11105 be a value in the 0.1-100.0 range, default value is 1.0.
11107 @item chroma_radius, cr
11108 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11109 greater value will result in a more blurred image, and in slower
11112 @item chroma_pre_filter_radius, cpfr
11113 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11115 @item chroma_strength, cs
11116 Set chroma maximum difference between pixels to still be considered,
11117 must be a value in the -0.9-100.0 range.
11120 Each chroma option value, if not explicitly specified, is set to the
11121 corresponding luma option value.
11126 Scale (resize) the input video, using the libswscale library.
11128 The scale filter forces the output display aspect ratio to be the same
11129 of the input, by changing the output sample aspect ratio.
11131 If the input image format is different from the format requested by
11132 the next filter, the scale filter will convert the input to the
11135 @subsection Options
11136 The filter accepts the following options, or any of the options
11137 supported by the libswscale scaler.
11139 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11140 the complete list of scaler options.
11145 Set the output video dimension expression. Default value is the input
11148 If the value is 0, the input width is used for the output.
11150 If one of the values is -1, the scale filter will use a value that
11151 maintains the aspect ratio of the input image, calculated from the
11152 other specified dimension. If both of them are -1, the input size is
11155 If one of the values is -n with n > 1, the scale filter will also use a value
11156 that maintains the aspect ratio of the input image, calculated from the other
11157 specified dimension. After that it will, however, make sure that the calculated
11158 dimension is divisible by n and adjust the value if necessary.
11160 See below for the list of accepted constants for use in the dimension
11164 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11168 Only evaluate expressions once during the filter initialization or when a command is processed.
11171 Evaluate expressions for each incoming frame.
11175 Default value is @samp{init}.
11179 Set the interlacing mode. It accepts the following values:
11183 Force interlaced aware scaling.
11186 Do not apply interlaced scaling.
11189 Select interlaced aware scaling depending on whether the source frames
11190 are flagged as interlaced or not.
11193 Default value is @samp{0}.
11196 Set libswscale scaling flags. See
11197 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11198 complete list of values. If not explicitly specified the filter applies
11202 @item param0, param1
11203 Set libswscale input parameters for scaling algorithms that need them. See
11204 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11205 complete documentation. If not explicitly specified the filter applies
11211 Set the video size. For the syntax of this option, check the
11212 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11214 @item in_color_matrix
11215 @item out_color_matrix
11216 Set in/output YCbCr color space type.
11218 This allows the autodetected value to be overridden as well as allows forcing
11219 a specific value used for the output and encoder.
11221 If not specified, the color space type depends on the pixel format.
11227 Choose automatically.
11230 Format conforming to International Telecommunication Union (ITU)
11231 Recommendation BT.709.
11234 Set color space conforming to the United States Federal Communications
11235 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11238 Set color space conforming to:
11242 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11245 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11248 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11253 Set color space conforming to SMPTE ST 240:1999.
11258 Set in/output YCbCr sample range.
11260 This allows the autodetected value to be overridden as well as allows forcing
11261 a specific value used for the output and encoder. If not specified, the
11262 range depends on the pixel format. Possible values:
11266 Choose automatically.
11269 Set full range (0-255 in case of 8-bit luma).
11272 Set "MPEG" range (16-235 in case of 8-bit luma).
11275 @item force_original_aspect_ratio
11276 Enable decreasing or increasing output video width or height if necessary to
11277 keep the original aspect ratio. Possible values:
11281 Scale the video as specified and disable this feature.
11284 The output video dimensions will automatically be decreased if needed.
11287 The output video dimensions will automatically be increased if needed.
11291 One useful instance of this option is that when you know a specific device's
11292 maximum allowed resolution, you can use this to limit the output video to
11293 that, while retaining the aspect ratio. For example, device A allows
11294 1280x720 playback, and your video is 1920x800. Using this option (set it to
11295 decrease) and specifying 1280x720 to the command line makes the output
11298 Please note that this is a different thing than specifying -1 for @option{w}
11299 or @option{h}, you still need to specify the output resolution for this option
11304 The values of the @option{w} and @option{h} options are expressions
11305 containing the following constants:
11310 The input width and height
11314 These are the same as @var{in_w} and @var{in_h}.
11318 The output (scaled) width and height
11322 These are the same as @var{out_w} and @var{out_h}
11325 The same as @var{iw} / @var{ih}
11328 input sample aspect ratio
11331 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11335 horizontal and vertical input chroma subsample values. For example for the
11336 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11340 horizontal and vertical output chroma subsample values. For example for the
11341 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11344 @subsection Examples
11348 Scale the input video to a size of 200x100
11353 This is equivalent to:
11364 Specify a size abbreviation for the output size:
11369 which can also be written as:
11375 Scale the input to 2x:
11377 scale=w=2*iw:h=2*ih
11381 The above is the same as:
11383 scale=2*in_w:2*in_h
11387 Scale the input to 2x with forced interlaced scaling:
11389 scale=2*iw:2*ih:interl=1
11393 Scale the input to half size:
11395 scale=w=iw/2:h=ih/2
11399 Increase the width, and set the height to the same size:
11405 Seek Greek harmony:
11412 Increase the height, and set the width to 3/2 of the height:
11414 scale=w=3/2*oh:h=3/5*ih
11418 Increase the size, making the size a multiple of the chroma
11421 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11425 Increase the width to a maximum of 500 pixels,
11426 keeping the same aspect ratio as the input:
11428 scale=w='min(500\, iw*3/2):h=-1'
11432 @subsection Commands
11434 This filter supports the following commands:
11438 Set the output video dimension expression.
11439 The command accepts the same syntax of the corresponding option.
11441 If the specified expression is not valid, it is kept at its current
11447 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11448 format conversion on CUDA video frames. Setting the output width and height
11449 works in the same way as for the @var{scale} filter.
11451 The following additional options are accepted:
11454 The pixel format of the output CUDA frames. If set to the string "same" (the
11455 default), the input format will be kept. Note that automatic format negotiation
11456 and conversion is not yet supported for hardware frames
11459 The interpolation algorithm used for resizing. One of the following:
11466 @item cubic2p_bspline
11467 2-parameter cubic (B=1, C=0)
11469 @item cubic2p_catmullrom
11470 2-parameter cubic (B=0, C=1/2)
11472 @item cubic2p_b05c03
11473 2-parameter cubic (B=1/2, C=3/10)
11485 Scale (resize) the input video, based on a reference video.
11487 See the scale filter for available options, scale2ref supports the same but
11488 uses the reference video instead of the main input as basis.
11490 @subsection Examples
11494 Scale a subtitle stream to match the main video in size before overlaying
11496 'scale2ref[b][a];[a][b]overlay'
11500 @anchor{selectivecolor}
11501 @section selectivecolor
11503 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11504 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11505 by the "purity" of the color (that is, how saturated it already is).
11507 This filter is similar to the Adobe Photoshop Selective Color tool.
11509 The filter accepts the following options:
11512 @item correction_method
11513 Select color correction method.
11515 Available values are:
11518 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11521 Specified adjustments are relative to the original component value.
11523 Default is @code{absolute}.
11525 Adjustments for red pixels (pixels where the red component is the maximum)
11527 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11529 Adjustments for green pixels (pixels where the green component is the maximum)
11531 Adjustments for cyan pixels (pixels where the red component is the minimum)
11533 Adjustments for blue pixels (pixels where the blue component is the maximum)
11535 Adjustments for magenta pixels (pixels where the green component is the minimum)
11537 Adjustments for white pixels (pixels where all components are greater than 128)
11539 Adjustments for all pixels except pure black and pure white
11541 Adjustments for black pixels (pixels where all components are lesser than 128)
11543 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11546 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11547 4 space separated floating point adjustment values in the [-1,1] range,
11548 respectively to adjust the amount of cyan, magenta, yellow and black for the
11549 pixels of its range.
11551 @subsection Examples
11555 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11556 increase magenta by 27% in blue areas:
11558 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11562 Use a Photoshop selective color preset:
11564 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11568 @section separatefields
11570 The @code{separatefields} takes a frame-based video input and splits
11571 each frame into its components fields, producing a new half height clip
11572 with twice the frame rate and twice the frame count.
11574 This filter use field-dominance information in frame to decide which
11575 of each pair of fields to place first in the output.
11576 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
11578 @section setdar, setsar
11580 The @code{setdar} filter sets the Display Aspect Ratio for the filter
11583 This is done by changing the specified Sample (aka Pixel) Aspect
11584 Ratio, according to the following equation:
11586 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
11589 Keep in mind that the @code{setdar} filter does not modify the pixel
11590 dimensions of the video frame. Also, the display aspect ratio set by
11591 this filter may be changed by later filters in the filterchain,
11592 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
11595 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
11596 the filter output video.
11598 Note that as a consequence of the application of this filter, the
11599 output display aspect ratio will change according to the equation
11602 Keep in mind that the sample aspect ratio set by the @code{setsar}
11603 filter may be changed by later filters in the filterchain, e.g. if
11604 another "setsar" or a "setdar" filter is applied.
11606 It accepts the following parameters:
11609 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
11610 Set the aspect ratio used by the filter.
11612 The parameter can be a floating point number string, an expression, or
11613 a string of the form @var{num}:@var{den}, where @var{num} and
11614 @var{den} are the numerator and denominator of the aspect ratio. If
11615 the parameter is not specified, it is assumed the value "0".
11616 In case the form "@var{num}:@var{den}" is used, the @code{:} character
11620 Set the maximum integer value to use for expressing numerator and
11621 denominator when reducing the expressed aspect ratio to a rational.
11622 Default value is @code{100}.
11626 The parameter @var{sar} is an expression containing
11627 the following constants:
11631 These are approximated values for the mathematical constants e
11632 (Euler's number), pi (Greek pi), and phi (the golden ratio).
11635 The input width and height.
11638 These are the same as @var{w} / @var{h}.
11641 The input sample aspect ratio.
11644 The input display aspect ratio. It is the same as
11645 (@var{w} / @var{h}) * @var{sar}.
11648 Horizontal and vertical chroma subsample values. For example, for the
11649 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11652 @subsection Examples
11657 To change the display aspect ratio to 16:9, specify one of the following:
11664 To change the sample aspect ratio to 10:11, specify:
11670 To set a display aspect ratio of 16:9, and specify a maximum integer value of
11671 1000 in the aspect ratio reduction, use the command:
11673 setdar=ratio=16/9:max=1000
11681 Force field for the output video frame.
11683 The @code{setfield} filter marks the interlace type field for the
11684 output frames. It does not change the input frame, but only sets the
11685 corresponding property, which affects how the frame is treated by
11686 following filters (e.g. @code{fieldorder} or @code{yadif}).
11688 The filter accepts the following options:
11693 Available values are:
11697 Keep the same field property.
11700 Mark the frame as bottom-field-first.
11703 Mark the frame as top-field-first.
11706 Mark the frame as progressive.
11712 Show a line containing various information for each input video frame.
11713 The input video is not modified.
11715 The shown line contains a sequence of key/value pairs of the form
11716 @var{key}:@var{value}.
11718 The following values are shown in the output:
11722 The (sequential) number of the input frame, starting from 0.
11725 The Presentation TimeStamp of the input frame, expressed as a number of
11726 time base units. The time base unit depends on the filter input pad.
11729 The Presentation TimeStamp of the input frame, expressed as a number of
11733 The position of the frame in the input stream, or -1 if this information is
11734 unavailable and/or meaningless (for example in case of synthetic video).
11737 The pixel format name.
11740 The sample aspect ratio of the input frame, expressed in the form
11741 @var{num}/@var{den}.
11744 The size of the input frame. For the syntax of this option, check the
11745 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11748 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
11749 for bottom field first).
11752 This is 1 if the frame is a key frame, 0 otherwise.
11755 The picture type of the input frame ("I" for an I-frame, "P" for a
11756 P-frame, "B" for a B-frame, or "?" for an unknown type).
11757 Also refer to the documentation of the @code{AVPictureType} enum and of
11758 the @code{av_get_picture_type_char} function defined in
11759 @file{libavutil/avutil.h}.
11762 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
11764 @item plane_checksum
11765 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
11766 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
11769 @section showpalette
11771 Displays the 256 colors palette of each frame. This filter is only relevant for
11772 @var{pal8} pixel format frames.
11774 It accepts the following option:
11778 Set the size of the box used to represent one palette color entry. Default is
11779 @code{30} (for a @code{30x30} pixel box).
11782 @section shuffleframes
11784 Reorder and/or duplicate video frames.
11786 It accepts the following parameters:
11790 Set the destination indexes of input frames.
11791 This is space or '|' separated list of indexes that maps input frames to output
11792 frames. Number of indexes also sets maximal value that each index may have.
11795 The first frame has the index 0. The default is to keep the input unchanged.
11797 Swap second and third frame of every three frames of the input:
11799 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
11802 @section shuffleplanes
11804 Reorder and/or duplicate video planes.
11806 It accepts the following parameters:
11811 The index of the input plane to be used as the first output plane.
11814 The index of the input plane to be used as the second output plane.
11817 The index of the input plane to be used as the third output plane.
11820 The index of the input plane to be used as the fourth output plane.
11824 The first plane has the index 0. The default is to keep the input unchanged.
11826 Swap the second and third planes of the input:
11828 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
11831 @anchor{signalstats}
11832 @section signalstats
11833 Evaluate various visual metrics that assist in determining issues associated
11834 with the digitization of analog video media.
11836 By default the filter will log these metadata values:
11840 Display the minimal Y value contained within the input frame. Expressed in
11844 Display the Y value at the 10% percentile within the input frame. Expressed in
11848 Display the average Y value within the input frame. Expressed in range of
11852 Display the Y value at the 90% percentile within the input frame. Expressed in
11856 Display the maximum Y value contained within the input frame. Expressed in
11860 Display the minimal U value contained within the input frame. Expressed in
11864 Display the U value at the 10% percentile within the input frame. Expressed in
11868 Display the average U value within the input frame. Expressed in range of
11872 Display the U value at the 90% percentile within the input frame. Expressed in
11876 Display the maximum U value contained within the input frame. Expressed in
11880 Display the minimal V value contained within the input frame. Expressed in
11884 Display the V value at the 10% percentile within the input frame. Expressed in
11888 Display the average V value within the input frame. Expressed in range of
11892 Display the V value at the 90% percentile within the input frame. Expressed in
11896 Display the maximum V value contained within the input frame. Expressed in
11900 Display the minimal saturation value contained within the input frame.
11901 Expressed in range of [0-~181.02].
11904 Display the saturation value at the 10% percentile within the input frame.
11905 Expressed in range of [0-~181.02].
11908 Display the average saturation value within the input frame. Expressed in range
11912 Display the saturation value at the 90% percentile within the input frame.
11913 Expressed in range of [0-~181.02].
11916 Display the maximum saturation value contained within the input frame.
11917 Expressed in range of [0-~181.02].
11920 Display the median value for hue within the input frame. Expressed in range of
11924 Display the average value for hue within the input frame. Expressed in range of
11928 Display the average of sample value difference between all values of the Y
11929 plane in the current frame and corresponding values of the previous input frame.
11930 Expressed in range of [0-255].
11933 Display the average of sample value difference between all values of the U
11934 plane in the current frame and corresponding values of the previous input frame.
11935 Expressed in range of [0-255].
11938 Display the average of sample value difference between all values of the V
11939 plane in the current frame and corresponding values of the previous input frame.
11940 Expressed in range of [0-255].
11943 The filter accepts the following options:
11949 @option{stat} specify an additional form of image analysis.
11950 @option{out} output video with the specified type of pixel highlighted.
11952 Both options accept the following values:
11956 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
11957 unlike the neighboring pixels of the same field. Examples of temporal outliers
11958 include the results of video dropouts, head clogs, or tape tracking issues.
11961 Identify @var{vertical line repetition}. Vertical line repetition includes
11962 similar rows of pixels within a frame. In born-digital video vertical line
11963 repetition is common, but this pattern is uncommon in video digitized from an
11964 analog source. When it occurs in video that results from the digitization of an
11965 analog source it can indicate concealment from a dropout compensator.
11968 Identify pixels that fall outside of legal broadcast range.
11972 Set the highlight color for the @option{out} option. The default color is
11976 @subsection Examples
11980 Output data of various video metrics:
11982 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
11986 Output specific data about the minimum and maximum values of the Y plane per frame:
11988 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
11992 Playback video while highlighting pixels that are outside of broadcast range in red.
11994 ffplay example.mov -vf signalstats="out=brng:color=red"
11998 Playback video with signalstats metadata drawn over the frame.
12000 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12003 The contents of signalstat_drawtext.txt used in the command are:
12006 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12007 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12008 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12009 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12017 Blur the input video without impacting the outlines.
12019 It accepts the following options:
12022 @item luma_radius, lr
12023 Set the luma radius. The option value must be a float number in
12024 the range [0.1,5.0] that specifies the variance of the gaussian filter
12025 used to blur the image (slower if larger). Default value is 1.0.
12027 @item luma_strength, ls
12028 Set the luma strength. The option value must be a float number
12029 in the range [-1.0,1.0] that configures the blurring. A value included
12030 in [0.0,1.0] will blur the image whereas a value included in
12031 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12033 @item luma_threshold, lt
12034 Set the luma threshold used as a coefficient to determine
12035 whether a pixel should be blurred or not. The option value must be an
12036 integer in the range [-30,30]. A value of 0 will filter all the image,
12037 a value included in [0,30] will filter flat areas and a value included
12038 in [-30,0] will filter edges. Default value is 0.
12040 @item chroma_radius, cr
12041 Set the chroma radius. The option value must be a float number in
12042 the range [0.1,5.0] that specifies the variance of the gaussian filter
12043 used to blur the image (slower if larger). Default value is 1.0.
12045 @item chroma_strength, cs
12046 Set the chroma strength. The option value must be a float number
12047 in the range [-1.0,1.0] that configures the blurring. A value included
12048 in [0.0,1.0] will blur the image whereas a value included in
12049 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12051 @item chroma_threshold, ct
12052 Set the chroma threshold used as a coefficient to determine
12053 whether a pixel should be blurred or not. The option value must be an
12054 integer in the range [-30,30]. A value of 0 will filter all the image,
12055 a value included in [0,30] will filter flat areas and a value included
12056 in [-30,0] will filter edges. Default value is 0.
12059 If a chroma option is not explicitly set, the corresponding luma value
12064 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12066 This filter takes in input two input videos, the first input is
12067 considered the "main" source and is passed unchanged to the
12068 output. The second input is used as a "reference" video for computing
12071 Both video inputs must have the same resolution and pixel format for
12072 this filter to work correctly. Also it assumes that both inputs
12073 have the same number of frames, which are compared one by one.
12075 The filter stores the calculated SSIM of each frame.
12077 The description of the accepted parameters follows.
12080 @item stats_file, f
12081 If specified the filter will use the named file to save the SSIM of
12082 each individual frame. When filename equals "-" the data is sent to
12086 The file printed if @var{stats_file} is selected, contains a sequence of
12087 key/value pairs of the form @var{key}:@var{value} for each compared
12090 A description of each shown parameter follows:
12094 sequential number of the input frame, starting from 1
12096 @item Y, U, V, R, G, B
12097 SSIM of the compared frames for the component specified by the suffix.
12100 SSIM of the compared frames for the whole frame.
12103 Same as above but in dB representation.
12108 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12109 [main][ref] ssim="stats_file=stats.log" [out]
12112 On this example the input file being processed is compared with the
12113 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12114 is stored in @file{stats.log}.
12116 Another example with both psnr and ssim at same time:
12118 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12123 Convert between different stereoscopic image formats.
12125 The filters accept the following options:
12129 Set stereoscopic image format of input.
12131 Available values for input image formats are:
12134 side by side parallel (left eye left, right eye right)
12137 side by side crosseye (right eye left, left eye right)
12140 side by side parallel with half width resolution
12141 (left eye left, right eye right)
12144 side by side crosseye with half width resolution
12145 (right eye left, left eye right)
12148 above-below (left eye above, right eye below)
12151 above-below (right eye above, left eye below)
12154 above-below with half height resolution
12155 (left eye above, right eye below)
12158 above-below with half height resolution
12159 (right eye above, left eye below)
12162 alternating frames (left eye first, right eye second)
12165 alternating frames (right eye first, left eye second)
12168 interleaved rows (left eye has top row, right eye starts on next row)
12171 interleaved rows (right eye has top row, left eye starts on next row)
12174 interleaved columns, left eye first
12177 interleaved columns, right eye first
12179 Default value is @samp{sbsl}.
12183 Set stereoscopic image format of output.
12187 side by side parallel (left eye left, right eye right)
12190 side by side crosseye (right eye left, left eye right)
12193 side by side parallel with half width resolution
12194 (left eye left, right eye right)
12197 side by side crosseye with half width resolution
12198 (right eye left, left eye right)
12201 above-below (left eye above, right eye below)
12204 above-below (right eye above, left eye below)
12207 above-below with half height resolution
12208 (left eye above, right eye below)
12211 above-below with half height resolution
12212 (right eye above, left eye below)
12215 alternating frames (left eye first, right eye second)
12218 alternating frames (right eye first, left eye second)
12221 interleaved rows (left eye has top row, right eye starts on next row)
12224 interleaved rows (right eye has top row, left eye starts on next row)
12227 anaglyph red/blue gray
12228 (red filter on left eye, blue filter on right eye)
12231 anaglyph red/green gray
12232 (red filter on left eye, green filter on right eye)
12235 anaglyph red/cyan gray
12236 (red filter on left eye, cyan filter on right eye)
12239 anaglyph red/cyan half colored
12240 (red filter on left eye, cyan filter on right eye)
12243 anaglyph red/cyan color
12244 (red filter on left eye, cyan filter on right eye)
12247 anaglyph red/cyan color optimized with the least squares projection of dubois
12248 (red filter on left eye, cyan filter on right eye)
12251 anaglyph green/magenta gray
12252 (green filter on left eye, magenta filter on right eye)
12255 anaglyph green/magenta half colored
12256 (green filter on left eye, magenta filter on right eye)
12259 anaglyph green/magenta colored
12260 (green filter on left eye, magenta filter on right eye)
12263 anaglyph green/magenta color optimized with the least squares projection of dubois
12264 (green filter on left eye, magenta filter on right eye)
12267 anaglyph yellow/blue gray
12268 (yellow filter on left eye, blue filter on right eye)
12271 anaglyph yellow/blue half colored
12272 (yellow filter on left eye, blue filter on right eye)
12275 anaglyph yellow/blue colored
12276 (yellow filter on left eye, blue filter on right eye)
12279 anaglyph yellow/blue color optimized with the least squares projection of dubois
12280 (yellow filter on left eye, blue filter on right eye)
12283 mono output (left eye only)
12286 mono output (right eye only)
12289 checkerboard, left eye first
12292 checkerboard, right eye first
12295 interleaved columns, left eye first
12298 interleaved columns, right eye first
12304 Default value is @samp{arcd}.
12307 @subsection Examples
12311 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12317 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12323 @section streamselect, astreamselect
12324 Select video or audio streams.
12326 The filter accepts the following options:
12330 Set number of inputs. Default is 2.
12333 Set input indexes to remap to outputs.
12336 @subsection Commands
12338 The @code{streamselect} and @code{astreamselect} filter supports the following
12343 Set input indexes to remap to outputs.
12346 @subsection Examples
12350 Select first 5 seconds 1st stream and rest of time 2nd stream:
12352 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12356 Same as above, but for audio:
12358 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12365 Apply a simple postprocessing filter that compresses and decompresses the image
12366 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12367 and average the results.
12369 The filter accepts the following options:
12373 Set quality. This option defines the number of levels for averaging. It accepts
12374 an integer in the range 0-6. If set to @code{0}, the filter will have no
12375 effect. A value of @code{6} means the higher quality. For each increment of
12376 that value the speed drops by a factor of approximately 2. Default value is
12380 Force a constant quantization parameter. If not set, the filter will use the QP
12381 from the video stream (if available).
12384 Set thresholding mode. Available modes are:
12388 Set hard thresholding (default).
12390 Set soft thresholding (better de-ringing effect, but likely blurrier).
12393 @item use_bframe_qp
12394 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12395 option may cause flicker since the B-Frames have often larger QP. Default is
12396 @code{0} (not enabled).
12402 Draw subtitles on top of input video using the libass library.
12404 To enable compilation of this filter you need to configure FFmpeg with
12405 @code{--enable-libass}. This filter also requires a build with libavcodec and
12406 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12407 Alpha) subtitles format.
12409 The filter accepts the following options:
12413 Set the filename of the subtitle file to read. It must be specified.
12415 @item original_size
12416 Specify the size of the original video, the video for which the ASS file
12417 was composed. For the syntax of this option, check the
12418 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12419 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12420 correctly scale the fonts if the aspect ratio has been changed.
12423 Set a directory path containing fonts that can be used by the filter.
12424 These fonts will be used in addition to whatever the font provider uses.
12427 Set subtitles input character encoding. @code{subtitles} filter only. Only
12428 useful if not UTF-8.
12430 @item stream_index, si
12431 Set subtitles stream index. @code{subtitles} filter only.
12434 Override default style or script info parameters of the subtitles. It accepts a
12435 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12438 If the first key is not specified, it is assumed that the first value
12439 specifies the @option{filename}.
12441 For example, to render the file @file{sub.srt} on top of the input
12442 video, use the command:
12447 which is equivalent to:
12449 subtitles=filename=sub.srt
12452 To render the default subtitles stream from file @file{video.mkv}, use:
12454 subtitles=video.mkv
12457 To render the second subtitles stream from that file, use:
12459 subtitles=video.mkv:si=1
12462 To make the subtitles stream from @file{sub.srt} appear in transparent green
12463 @code{DejaVu Serif}, use:
12465 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12468 @section super2xsai
12470 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12471 Interpolate) pixel art scaling algorithm.
12473 Useful for enlarging pixel art images without reducing sharpness.
12477 Swap two rectangular objects in video.
12479 This filter accepts the following options:
12489 Set 1st rect x coordinate.
12492 Set 1st rect y coordinate.
12495 Set 2nd rect x coordinate.
12498 Set 2nd rect y coordinate.
12500 All expressions are evaluated once for each frame.
12503 The all options are expressions containing the following constants:
12508 The input width and height.
12511 same as @var{w} / @var{h}
12514 input sample aspect ratio
12517 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12520 The number of the input frame, starting from 0.
12523 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12526 the position in the file of the input frame, NAN if unknown
12534 Apply telecine process to the video.
12536 This filter accepts the following options:
12545 The default value is @code{top}.
12549 A string of numbers representing the pulldown pattern you wish to apply.
12550 The default value is @code{23}.
12554 Some typical patterns:
12559 24p: 2332 (preferred)
12566 24p: 222222222223 ("Euro pulldown")
12572 Select the most representative frame in a given sequence of consecutive frames.
12574 The filter accepts the following options:
12578 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
12579 will pick one of them, and then handle the next batch of @var{n} frames until
12580 the end. Default is @code{100}.
12583 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
12584 value will result in a higher memory usage, so a high value is not recommended.
12586 @subsection Examples
12590 Extract one picture each 50 frames:
12596 Complete example of a thumbnail creation with @command{ffmpeg}:
12598 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
12604 Tile several successive frames together.
12606 The filter accepts the following options:
12611 Set the grid size (i.e. the number of lines and columns). For the syntax of
12612 this option, check the
12613 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12616 Set the maximum number of frames to render in the given area. It must be less
12617 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
12618 the area will be used.
12621 Set the outer border margin in pixels.
12624 Set the inner border thickness (i.e. the number of pixels between frames). For
12625 more advanced padding options (such as having different values for the edges),
12626 refer to the pad video filter.
12629 Specify the color of the unused area. For the syntax of this option, check the
12630 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
12634 @subsection Examples
12638 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
12640 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
12642 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
12643 duplicating each output frame to accommodate the originally detected frame
12647 Display @code{5} pictures in an area of @code{3x2} frames,
12648 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
12649 mixed flat and named options:
12651 tile=3x2:nb_frames=5:padding=7:margin=2
12655 @section tinterlace
12657 Perform various types of temporal field interlacing.
12659 Frames are counted starting from 1, so the first input frame is
12662 The filter accepts the following options:
12667 Specify the mode of the interlacing. This option can also be specified
12668 as a value alone. See below for a list of values for this option.
12670 Available values are:
12674 Move odd frames into the upper field, even into the lower field,
12675 generating a double height frame at half frame rate.
12679 Frame 1 Frame 2 Frame 3 Frame 4
12681 11111 22222 33333 44444
12682 11111 22222 33333 44444
12683 11111 22222 33333 44444
12684 11111 22222 33333 44444
12698 Only output odd frames, even frames are dropped, generating a frame with
12699 unchanged height at half frame rate.
12704 Frame 1 Frame 2 Frame 3 Frame 4
12706 11111 22222 33333 44444
12707 11111 22222 33333 44444
12708 11111 22222 33333 44444
12709 11111 22222 33333 44444
12719 Only output even frames, odd frames are dropped, generating a frame with
12720 unchanged height at half frame rate.
12725 Frame 1 Frame 2 Frame 3 Frame 4
12727 11111 22222 33333 44444
12728 11111 22222 33333 44444
12729 11111 22222 33333 44444
12730 11111 22222 33333 44444
12740 Expand each frame to full height, but pad alternate lines with black,
12741 generating a frame with double height at the same input frame rate.
12746 Frame 1 Frame 2 Frame 3 Frame 4
12748 11111 22222 33333 44444
12749 11111 22222 33333 44444
12750 11111 22222 33333 44444
12751 11111 22222 33333 44444
12754 11111 ..... 33333 .....
12755 ..... 22222 ..... 44444
12756 11111 ..... 33333 .....
12757 ..... 22222 ..... 44444
12758 11111 ..... 33333 .....
12759 ..... 22222 ..... 44444
12760 11111 ..... 33333 .....
12761 ..... 22222 ..... 44444
12765 @item interleave_top, 4
12766 Interleave the upper field from odd frames with the lower field from
12767 even frames, generating a frame with unchanged height at half frame rate.
12772 Frame 1 Frame 2 Frame 3 Frame 4
12774 11111<- 22222 33333<- 44444
12775 11111 22222<- 33333 44444<-
12776 11111<- 22222 33333<- 44444
12777 11111 22222<- 33333 44444<-
12787 @item interleave_bottom, 5
12788 Interleave the lower field from odd frames with the upper field from
12789 even frames, generating a frame with unchanged height at half frame rate.
12794 Frame 1 Frame 2 Frame 3 Frame 4
12796 11111 22222<- 33333 44444<-
12797 11111<- 22222 33333<- 44444
12798 11111 22222<- 33333 44444<-
12799 11111<- 22222 33333<- 44444
12809 @item interlacex2, 6
12810 Double frame rate with unchanged height. Frames are inserted each
12811 containing the second temporal field from the previous input frame and
12812 the first temporal field from the next input frame. This mode relies on
12813 the top_field_first flag. Useful for interlaced video displays with no
12814 field synchronisation.
12819 Frame 1 Frame 2 Frame 3 Frame 4
12821 11111 22222 33333 44444
12822 11111 22222 33333 44444
12823 11111 22222 33333 44444
12824 11111 22222 33333 44444
12827 11111 22222 22222 33333 33333 44444 44444
12828 11111 11111 22222 22222 33333 33333 44444
12829 11111 22222 22222 33333 33333 44444 44444
12830 11111 11111 22222 22222 33333 33333 44444
12835 Move odd frames into the upper field, even into the lower field,
12836 generating a double height frame at same frame rate.
12841 Frame 1 Frame 2 Frame 3 Frame 4
12843 11111 22222 33333 44444
12844 11111 22222 33333 44444
12845 11111 22222 33333 44444
12846 11111 22222 33333 44444
12849 11111 33333 33333 55555
12850 22222 22222 44444 44444
12851 11111 33333 33333 55555
12852 22222 22222 44444 44444
12853 11111 33333 33333 55555
12854 22222 22222 44444 44444
12855 11111 33333 33333 55555
12856 22222 22222 44444 44444
12861 Numeric values are deprecated but are accepted for backward
12862 compatibility reasons.
12864 Default mode is @code{merge}.
12867 Specify flags influencing the filter process.
12869 Available value for @var{flags} is:
12872 @item low_pass_filter, vlfp
12873 Enable vertical low-pass filtering in the filter.
12874 Vertical low-pass filtering is required when creating an interlaced
12875 destination from a progressive source which contains high-frequency
12876 vertical detail. Filtering will reduce interlace 'twitter' and Moire
12879 Vertical low-pass filtering can only be enabled for @option{mode}
12880 @var{interleave_top} and @var{interleave_bottom}.
12887 Transpose rows with columns in the input video and optionally flip it.
12889 It accepts the following parameters:
12894 Specify the transposition direction.
12896 Can assume the following values:
12898 @item 0, 4, cclock_flip
12899 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
12907 Rotate by 90 degrees clockwise, that is:
12915 Rotate by 90 degrees counterclockwise, that is:
12922 @item 3, 7, clock_flip
12923 Rotate by 90 degrees clockwise and vertically flip, that is:
12931 For values between 4-7, the transposition is only done if the input
12932 video geometry is portrait and not landscape. These values are
12933 deprecated, the @code{passthrough} option should be used instead.
12935 Numerical values are deprecated, and should be dropped in favor of
12936 symbolic constants.
12939 Do not apply the transposition if the input geometry matches the one
12940 specified by the specified value. It accepts the following values:
12943 Always apply transposition.
12945 Preserve portrait geometry (when @var{height} >= @var{width}).
12947 Preserve landscape geometry (when @var{width} >= @var{height}).
12950 Default value is @code{none}.
12953 For example to rotate by 90 degrees clockwise and preserve portrait
12956 transpose=dir=1:passthrough=portrait
12959 The command above can also be specified as:
12961 transpose=1:portrait
12965 Trim the input so that the output contains one continuous subpart of the input.
12967 It accepts the following parameters:
12970 Specify the time of the start of the kept section, i.e. the frame with the
12971 timestamp @var{start} will be the first frame in the output.
12974 Specify the time of the first frame that will be dropped, i.e. the frame
12975 immediately preceding the one with the timestamp @var{end} will be the last
12976 frame in the output.
12979 This is the same as @var{start}, except this option sets the start timestamp
12980 in timebase units instead of seconds.
12983 This is the same as @var{end}, except this option sets the end timestamp
12984 in timebase units instead of seconds.
12987 The maximum duration of the output in seconds.
12990 The number of the first frame that should be passed to the output.
12993 The number of the first frame that should be dropped.
12996 @option{start}, @option{end}, and @option{duration} are expressed as time
12997 duration specifications; see
12998 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
12999 for the accepted syntax.
13001 Note that the first two sets of the start/end options and the @option{duration}
13002 option look at the frame timestamp, while the _frame variants simply count the
13003 frames that pass through the filter. Also note that this filter does not modify
13004 the timestamps. If you wish for the output timestamps to start at zero, insert a
13005 setpts filter after the trim filter.
13007 If multiple start or end options are set, this filter tries to be greedy and
13008 keep all the frames that match at least one of the specified constraints. To keep
13009 only the part that matches all the constraints at once, chain multiple trim
13012 The defaults are such that all the input is kept. So it is possible to set e.g.
13013 just the end values to keep everything before the specified time.
13018 Drop everything except the second minute of input:
13020 ffmpeg -i INPUT -vf trim=60:120
13024 Keep only the first second:
13026 ffmpeg -i INPUT -vf trim=duration=1
13035 Sharpen or blur the input video.
13037 It accepts the following parameters:
13040 @item luma_msize_x, lx
13041 Set the luma matrix horizontal size. It must be an odd integer between
13042 3 and 63. The default value is 5.
13044 @item luma_msize_y, ly
13045 Set the luma matrix vertical size. It must be an odd integer between 3
13046 and 63. The default value is 5.
13048 @item luma_amount, la
13049 Set the luma effect strength. It must be a floating point number, reasonable
13050 values lay between -1.5 and 1.5.
13052 Negative values will blur the input video, while positive values will
13053 sharpen it, a value of zero will disable the effect.
13055 Default value is 1.0.
13057 @item chroma_msize_x, cx
13058 Set the chroma matrix horizontal size. It must be an odd integer
13059 between 3 and 63. The default value is 5.
13061 @item chroma_msize_y, cy
13062 Set the chroma matrix vertical size. It must be an odd integer
13063 between 3 and 63. The default value is 5.
13065 @item chroma_amount, ca
13066 Set the chroma effect strength. It must be a floating point number, reasonable
13067 values lay between -1.5 and 1.5.
13069 Negative values will blur the input video, while positive values will
13070 sharpen it, a value of zero will disable the effect.
13072 Default value is 0.0.
13075 If set to 1, specify using OpenCL capabilities, only available if
13076 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13080 All parameters are optional and default to the equivalent of the
13081 string '5:5:1.0:5:5:0.0'.
13083 @subsection Examples
13087 Apply strong luma sharpen effect:
13089 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13093 Apply a strong blur of both luma and chroma parameters:
13095 unsharp=7:7:-2:7:7:-2
13101 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13102 the image at several (or - in the case of @option{quality} level @code{8} - all)
13103 shifts and average the results.
13105 The way this differs from the behavior of spp is that uspp actually encodes &
13106 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13107 DCT similar to MJPEG.
13109 The filter accepts the following options:
13113 Set quality. This option defines the number of levels for averaging. It accepts
13114 an integer in the range 0-8. If set to @code{0}, the filter will have no
13115 effect. A value of @code{8} means the higher quality. For each increment of
13116 that value the speed drops by a factor of approximately 2. Default value is
13120 Force a constant quantization parameter. If not set, the filter will use the QP
13121 from the video stream (if available).
13124 @section vectorscope
13126 Display 2 color component values in the two dimensional graph (which is called
13129 This filter accepts the following options:
13133 Set vectorscope mode.
13135 It accepts the following values:
13138 Gray values are displayed on graph, higher brightness means more pixels have
13139 same component color value on location in graph. This is the default mode.
13142 Gray values are displayed on graph. Surrounding pixels values which are not
13143 present in video frame are drawn in gradient of 2 color components which are
13144 set by option @code{x} and @code{y}. The 3rd color component is static.
13147 Actual color components values present in video frame are displayed on graph.
13150 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13151 on graph increases value of another color component, which is luminance by
13152 default values of @code{x} and @code{y}.
13155 Actual colors present in video frame are displayed on graph. If two different
13156 colors map to same position on graph then color with higher value of component
13157 not present in graph is picked.
13160 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13161 component picked from radial gradient.
13165 Set which color component will be represented on X-axis. Default is @code{1}.
13168 Set which color component will be represented on Y-axis. Default is @code{2}.
13171 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13172 of color component which represents frequency of (X, Y) location in graph.
13177 No envelope, this is default.
13180 Instant envelope, even darkest single pixel will be clearly highlighted.
13183 Hold maximum and minimum values presented in graph over time. This way you
13184 can still spot out of range values without constantly looking at vectorscope.
13187 Peak and instant envelope combined together.
13191 Set what kind of graticule to draw.
13199 Set graticule opacity.
13202 Set graticule flags.
13206 Draw graticule for white point.
13209 Draw graticule for black point.
13212 Draw color points short names.
13216 Set background opacity.
13218 @item lthreshold, l
13219 Set low threshold for color component not represented on X or Y axis.
13220 Values lower than this value will be ignored. Default is 0.
13221 Note this value is multiplied with actual max possible value one pixel component
13222 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13225 @item hthreshold, h
13226 Set high threshold for color component not represented on X or Y axis.
13227 Values higher than this value will be ignored. Default is 1.
13228 Note this value is multiplied with actual max possible value one pixel component
13229 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13230 is 0.9 * 255 = 230.
13232 @item colorspace, c
13233 Set what kind of colorspace to use when drawing graticule.
13242 @anchor{vidstabdetect}
13243 @section vidstabdetect
13245 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13246 @ref{vidstabtransform} for pass 2.
13248 This filter generates a file with relative translation and rotation
13249 transform information about subsequent frames, which is then used by
13250 the @ref{vidstabtransform} filter.
13252 To enable compilation of this filter you need to configure FFmpeg with
13253 @code{--enable-libvidstab}.
13255 This filter accepts the following options:
13259 Set the path to the file used to write the transforms information.
13260 Default value is @file{transforms.trf}.
13263 Set how shaky the video is and how quick the camera is. It accepts an
13264 integer in the range 1-10, a value of 1 means little shakiness, a
13265 value of 10 means strong shakiness. Default value is 5.
13268 Set the accuracy of the detection process. It must be a value in the
13269 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13270 accuracy. Default value is 15.
13273 Set stepsize of the search process. The region around minimum is
13274 scanned with 1 pixel resolution. Default value is 6.
13277 Set minimum contrast. Below this value a local measurement field is
13278 discarded. Must be a floating point value in the range 0-1. Default
13282 Set reference frame number for tripod mode.
13284 If enabled, the motion of the frames is compared to a reference frame
13285 in the filtered stream, identified by the specified number. The idea
13286 is to compensate all movements in a more-or-less static scene and keep
13287 the camera view absolutely still.
13289 If set to 0, it is disabled. The frames are counted starting from 1.
13292 Show fields and transforms in the resulting frames. It accepts an
13293 integer in the range 0-2. Default value is 0, which disables any
13297 @subsection Examples
13301 Use default values:
13307 Analyze strongly shaky movie and put the results in file
13308 @file{mytransforms.trf}:
13310 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13314 Visualize the result of internal transformations in the resulting
13317 vidstabdetect=show=1
13321 Analyze a video with medium shakiness using @command{ffmpeg}:
13323 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13327 @anchor{vidstabtransform}
13328 @section vidstabtransform
13330 Video stabilization/deshaking: pass 2 of 2,
13331 see @ref{vidstabdetect} for pass 1.
13333 Read a file with transform information for each frame and
13334 apply/compensate them. Together with the @ref{vidstabdetect}
13335 filter this can be used to deshake videos. See also
13336 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13337 the @ref{unsharp} filter, see below.
13339 To enable compilation of this filter you need to configure FFmpeg with
13340 @code{--enable-libvidstab}.
13342 @subsection Options
13346 Set path to the file used to read the transforms. Default value is
13347 @file{transforms.trf}.
13350 Set the number of frames (value*2 + 1) used for lowpass filtering the
13351 camera movements. Default value is 10.
13353 For example a number of 10 means that 21 frames are used (10 in the
13354 past and 10 in the future) to smoothen the motion in the video. A
13355 larger value leads to a smoother video, but limits the acceleration of
13356 the camera (pan/tilt movements). 0 is a special case where a static
13357 camera is simulated.
13360 Set the camera path optimization algorithm.
13362 Accepted values are:
13365 gaussian kernel low-pass filter on camera motion (default)
13367 averaging on transformations
13371 Set maximal number of pixels to translate frames. Default value is -1,
13375 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13376 value is -1, meaning no limit.
13379 Specify how to deal with borders that may be visible due to movement
13382 Available values are:
13385 keep image information from previous frame (default)
13387 fill the border black
13391 Invert transforms if set to 1. Default value is 0.
13394 Consider transforms as relative to previous frame if set to 1,
13395 absolute if set to 0. Default value is 0.
13398 Set percentage to zoom. A positive value will result in a zoom-in
13399 effect, a negative value in a zoom-out effect. Default value is 0 (no
13403 Set optimal zooming to avoid borders.
13405 Accepted values are:
13410 optimal static zoom value is determined (only very strong movements
13411 will lead to visible borders) (default)
13413 optimal adaptive zoom value is determined (no borders will be
13414 visible), see @option{zoomspeed}
13417 Note that the value given at zoom is added to the one calculated here.
13420 Set percent to zoom maximally each frame (enabled when
13421 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13425 Specify type of interpolation.
13427 Available values are:
13432 linear only horizontal
13434 linear in both directions (default)
13436 cubic in both directions (slow)
13440 Enable virtual tripod mode if set to 1, which is equivalent to
13441 @code{relative=0:smoothing=0}. Default value is 0.
13443 Use also @code{tripod} option of @ref{vidstabdetect}.
13446 Increase log verbosity if set to 1. Also the detected global motions
13447 are written to the temporary file @file{global_motions.trf}. Default
13451 @subsection Examples
13455 Use @command{ffmpeg} for a typical stabilization with default values:
13457 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13460 Note the use of the @ref{unsharp} filter which is always recommended.
13463 Zoom in a bit more and load transform data from a given file:
13465 vidstabtransform=zoom=5:input="mytransforms.trf"
13469 Smoothen the video even more:
13471 vidstabtransform=smoothing=30
13477 Flip the input video vertically.
13479 For example, to vertically flip a video with @command{ffmpeg}:
13481 ffmpeg -i in.avi -vf "vflip" out.avi
13487 Make or reverse a natural vignetting effect.
13489 The filter accepts the following options:
13493 Set lens angle expression as a number of radians.
13495 The value is clipped in the @code{[0,PI/2]} range.
13497 Default value: @code{"PI/5"}
13501 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
13505 Set forward/backward mode.
13507 Available modes are:
13510 The larger the distance from the central point, the darker the image becomes.
13513 The larger the distance from the central point, the brighter the image becomes.
13514 This can be used to reverse a vignette effect, though there is no automatic
13515 detection to extract the lens @option{angle} and other settings (yet). It can
13516 also be used to create a burning effect.
13519 Default value is @samp{forward}.
13522 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
13524 It accepts the following values:
13527 Evaluate expressions only once during the filter initialization.
13530 Evaluate expressions for each incoming frame. This is way slower than the
13531 @samp{init} mode since it requires all the scalers to be re-computed, but it
13532 allows advanced dynamic expressions.
13535 Default value is @samp{init}.
13538 Set dithering to reduce the circular banding effects. Default is @code{1}
13542 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
13543 Setting this value to the SAR of the input will make a rectangular vignetting
13544 following the dimensions of the video.
13546 Default is @code{1/1}.
13549 @subsection Expressions
13551 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
13552 following parameters.
13557 input width and height
13560 the number of input frame, starting from 0
13563 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
13564 @var{TB} units, NAN if undefined
13567 frame rate of the input video, NAN if the input frame rate is unknown
13570 the PTS (Presentation TimeStamp) of the filtered video frame,
13571 expressed in seconds, NAN if undefined
13574 time base of the input video
13578 @subsection Examples
13582 Apply simple strong vignetting effect:
13588 Make a flickering vignetting:
13590 vignette='PI/4+random(1)*PI/50':eval=frame
13596 Stack input videos vertically.
13598 All streams must be of same pixel format and of same width.
13600 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
13601 to create same output.
13603 The filter accept the following option:
13607 Set number of input streams. Default is 2.
13610 If set to 1, force the output to terminate when the shortest input
13611 terminates. Default value is 0.
13616 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
13617 Deinterlacing Filter").
13619 Based on the process described by Martin Weston for BBC R&D, and
13620 implemented based on the de-interlace algorithm written by Jim
13621 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
13622 uses filter coefficients calculated by BBC R&D.
13624 There are two sets of filter coefficients, so called "simple":
13625 and "complex". Which set of filter coefficients is used can
13626 be set by passing an optional parameter:
13630 Set the interlacing filter coefficients. Accepts one of the following values:
13634 Simple filter coefficient set.
13636 More-complex filter coefficient set.
13638 Default value is @samp{complex}.
13641 Specify which frames to deinterlace. Accept one of the following values:
13645 Deinterlace all frames,
13647 Only deinterlace frames marked as interlaced.
13650 Default value is @samp{all}.
13654 Video waveform monitor.
13656 The waveform monitor plots color component intensity. By default luminance
13657 only. Each column of the waveform corresponds to a column of pixels in the
13660 It accepts the following options:
13664 Can be either @code{row}, or @code{column}. Default is @code{column}.
13665 In row mode, the graph on the left side represents color component value 0 and
13666 the right side represents value = 255. In column mode, the top side represents
13667 color component value = 0 and bottom side represents value = 255.
13670 Set intensity. Smaller values are useful to find out how many values of the same
13671 luminance are distributed across input rows/columns.
13672 Default value is @code{0.04}. Allowed range is [0, 1].
13675 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
13676 In mirrored mode, higher values will be represented on the left
13677 side for @code{row} mode and at the top for @code{column} mode. Default is
13678 @code{1} (mirrored).
13682 It accepts the following values:
13685 Presents information identical to that in the @code{parade}, except
13686 that the graphs representing color components are superimposed directly
13689 This display mode makes it easier to spot relative differences or similarities
13690 in overlapping areas of the color components that are supposed to be identical,
13691 such as neutral whites, grays, or blacks.
13694 Display separate graph for the color components side by side in
13695 @code{row} mode or one below the other in @code{column} mode.
13698 Display separate graph for the color components side by side in
13699 @code{column} mode or one below the other in @code{row} mode.
13701 Using this display mode makes it easy to spot color casts in the highlights
13702 and shadows of an image, by comparing the contours of the top and the bottom
13703 graphs of each waveform. Since whites, grays, and blacks are characterized
13704 by exactly equal amounts of red, green, and blue, neutral areas of the picture
13705 should display three waveforms of roughly equal width/height. If not, the
13706 correction is easy to perform by making level adjustments the three waveforms.
13708 Default is @code{stack}.
13710 @item components, c
13711 Set which color components to display. Default is 1, which means only luminance
13712 or red color component if input is in RGB colorspace. If is set for example to
13713 7 it will display all 3 (if) available color components.
13718 No envelope, this is default.
13721 Instant envelope, minimum and maximum values presented in graph will be easily
13722 visible even with small @code{step} value.
13725 Hold minimum and maximum values presented in graph across time. This way you
13726 can still spot out of range values without constantly looking at waveforms.
13729 Peak and instant envelope combined together.
13735 No filtering, this is default.
13738 Luma and chroma combined together.
13741 Similar as above, but shows difference between blue and red chroma.
13744 Displays only chroma.
13747 Displays actual color value on waveform.
13750 Similar as above, but with luma showing frequency of chroma values.
13754 Set which graticule to display.
13758 Do not display graticule.
13761 Display green graticule showing legal broadcast ranges.
13765 Set graticule opacity.
13768 Set graticule flags.
13772 Draw numbers above lines. By default enabled.
13775 Draw dots instead of lines.
13779 Set scale used for displaying graticule.
13786 Default is digital.
13790 Apply the xBR high-quality magnification filter which is designed for pixel
13791 art. It follows a set of edge-detection rules, see
13792 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
13794 It accepts the following option:
13798 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
13799 @code{3xBR} and @code{4} for @code{4xBR}.
13800 Default is @code{3}.
13806 Deinterlace the input video ("yadif" means "yet another deinterlacing
13809 It accepts the following parameters:
13815 The interlacing mode to adopt. It accepts one of the following values:
13818 @item 0, send_frame
13819 Output one frame for each frame.
13820 @item 1, send_field
13821 Output one frame for each field.
13822 @item 2, send_frame_nospatial
13823 Like @code{send_frame}, but it skips the spatial interlacing check.
13824 @item 3, send_field_nospatial
13825 Like @code{send_field}, but it skips the spatial interlacing check.
13828 The default value is @code{send_frame}.
13831 The picture field parity assumed for the input interlaced video. It accepts one
13832 of the following values:
13836 Assume the top field is first.
13838 Assume the bottom field is first.
13840 Enable automatic detection of field parity.
13843 The default value is @code{auto}.
13844 If the interlacing is unknown or the decoder does not export this information,
13845 top field first will be assumed.
13848 Specify which frames to deinterlace. Accept one of the following
13853 Deinterlace all frames.
13854 @item 1, interlaced
13855 Only deinterlace frames marked as interlaced.
13858 The default value is @code{all}.
13863 Apply Zoom & Pan effect.
13865 This filter accepts the following options:
13869 Set the zoom expression. Default is 1.
13873 Set the x and y expression. Default is 0.
13876 Set the duration expression in number of frames.
13877 This sets for how many number of frames effect will last for
13878 single input image.
13881 Set the output image size, default is 'hd720'.
13884 Set the output frame rate, default is '25'.
13887 Each expression can contain the following constants:
13906 Output frame count.
13910 Last calculated 'x' and 'y' position from 'x' and 'y' expression
13911 for current input frame.
13915 'x' and 'y' of last output frame of previous input frame or 0 when there was
13916 not yet such frame (first input frame).
13919 Last calculated zoom from 'z' expression for current input frame.
13922 Last calculated zoom of last output frame of previous input frame.
13925 Number of output frames for current input frame. Calculated from 'd' expression
13926 for each input frame.
13929 number of output frames created for previous input frame
13932 Rational number: input width / input height
13935 sample aspect ratio
13938 display aspect ratio
13942 @subsection Examples
13946 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
13948 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
13952 Zoom-in up to 1.5 and pan always at center of picture:
13954 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
13959 Scale (resize) the input video, using the z.lib library:
13960 https://github.com/sekrit-twc/zimg.
13962 The zscale filter forces the output display aspect ratio to be the same
13963 as the input, by changing the output sample aspect ratio.
13965 If the input image format is different from the format requested by
13966 the next filter, the zscale filter will convert the input to the
13969 @subsection Options
13970 The filter accepts the following options.
13975 Set the output video dimension expression. Default value is the input
13978 If the @var{width} or @var{w} is 0, the input width is used for the output.
13979 If the @var{height} or @var{h} is 0, the input height is used for the output.
13981 If one of the values is -1, the zscale filter will use a value that
13982 maintains the aspect ratio of the input image, calculated from the
13983 other specified dimension. If both of them are -1, the input size is
13986 If one of the values is -n with n > 1, the zscale filter will also use a value
13987 that maintains the aspect ratio of the input image, calculated from the other
13988 specified dimension. After that it will, however, make sure that the calculated
13989 dimension is divisible by n and adjust the value if necessary.
13991 See below for the list of accepted constants for use in the dimension
13995 Set the video size. For the syntax of this option, check the
13996 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13999 Set the dither type.
14001 Possible values are:
14006 @item error_diffusion
14012 Set the resize filter type.
14014 Possible values are:
14024 Default is bilinear.
14027 Set the color range.
14029 Possible values are:
14036 Default is same as input.
14039 Set the color primaries.
14041 Possible values are:
14051 Default is same as input.
14054 Set the transfer characteristics.
14056 Possible values are:
14067 Default is same as input.
14070 Set the colorspace matrix.
14072 Possible value are:
14083 Default is same as input.
14086 Set the input color range.
14088 Possible values are:
14095 Default is same as input.
14097 @item primariesin, pin
14098 Set the input color primaries.
14100 Possible values are:
14110 Default is same as input.
14112 @item transferin, tin
14113 Set the input transfer characteristics.
14115 Possible values are:
14126 Default is same as input.
14128 @item matrixin, min
14129 Set the input colorspace matrix.
14131 Possible value are:
14143 The values of the @option{w} and @option{h} options are expressions
14144 containing the following constants:
14149 The input width and height
14153 These are the same as @var{in_w} and @var{in_h}.
14157 The output (scaled) width and height
14161 These are the same as @var{out_w} and @var{out_h}
14164 The same as @var{iw} / @var{ih}
14167 input sample aspect ratio
14170 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14174 horizontal and vertical input chroma subsample values. For example for the
14175 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14179 horizontal and vertical output chroma subsample values. For example for the
14180 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14186 @c man end VIDEO FILTERS
14188 @chapter Video Sources
14189 @c man begin VIDEO SOURCES
14191 Below is a description of the currently available video sources.
14195 Buffer video frames, and make them available to the filter chain.
14197 This source is mainly intended for a programmatic use, in particular
14198 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14200 It accepts the following parameters:
14205 Specify the size (width and height) of the buffered video frames. For the
14206 syntax of this option, check the
14207 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14210 The input video width.
14213 The input video height.
14216 A string representing the pixel format of the buffered video frames.
14217 It may be a number corresponding to a pixel format, or a pixel format
14221 Specify the timebase assumed by the timestamps of the buffered frames.
14224 Specify the frame rate expected for the video stream.
14226 @item pixel_aspect, sar
14227 The sample (pixel) aspect ratio of the input video.
14230 Specify the optional parameters to be used for the scale filter which
14231 is automatically inserted when an input change is detected in the
14232 input size or format.
14234 @item hw_frames_ctx
14235 When using a hardware pixel format, this should be a reference to an
14236 AVHWFramesContext describing input frames.
14241 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14244 will instruct the source to accept video frames with size 320x240 and
14245 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14246 square pixels (1:1 sample aspect ratio).
14247 Since the pixel format with name "yuv410p" corresponds to the number 6
14248 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14249 this example corresponds to:
14251 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14254 Alternatively, the options can be specified as a flat string, but this
14255 syntax is deprecated:
14257 @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}]
14261 Create a pattern generated by an elementary cellular automaton.
14263 The initial state of the cellular automaton can be defined through the
14264 @option{filename}, and @option{pattern} options. If such options are
14265 not specified an initial state is created randomly.
14267 At each new frame a new row in the video is filled with the result of
14268 the cellular automaton next generation. The behavior when the whole
14269 frame is filled is defined by the @option{scroll} option.
14271 This source accepts the following options:
14275 Read the initial cellular automaton state, i.e. the starting row, from
14276 the specified file.
14277 In the file, each non-whitespace character is considered an alive
14278 cell, a newline will terminate the row, and further characters in the
14279 file will be ignored.
14282 Read the initial cellular automaton state, i.e. the starting row, from
14283 the specified string.
14285 Each non-whitespace character in the string is considered an alive
14286 cell, a newline will terminate the row, and further characters in the
14287 string will be ignored.
14290 Set the video rate, that is the number of frames generated per second.
14293 @item random_fill_ratio, ratio
14294 Set the random fill ratio for the initial cellular automaton row. It
14295 is a floating point number value ranging from 0 to 1, defaults to
14298 This option is ignored when a file or a pattern is specified.
14300 @item random_seed, seed
14301 Set the seed for filling randomly the initial row, must be an integer
14302 included between 0 and UINT32_MAX. If not specified, or if explicitly
14303 set to -1, the filter will try to use a good random seed on a best
14307 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14308 Default value is 110.
14311 Set the size of the output video. For the syntax of this option, check the
14312 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14314 If @option{filename} or @option{pattern} is specified, the size is set
14315 by default to the width of the specified initial state row, and the
14316 height is set to @var{width} * PHI.
14318 If @option{size} is set, it must contain the width of the specified
14319 pattern string, and the specified pattern will be centered in the
14322 If a filename or a pattern string is not specified, the size value
14323 defaults to "320x518" (used for a randomly generated initial state).
14326 If set to 1, scroll the output upward when all the rows in the output
14327 have been already filled. If set to 0, the new generated row will be
14328 written over the top row just after the bottom row is filled.
14331 @item start_full, full
14332 If set to 1, completely fill the output with generated rows before
14333 outputting the first frame.
14334 This is the default behavior, for disabling set the value to 0.
14337 If set to 1, stitch the left and right row edges together.
14338 This is the default behavior, for disabling set the value to 0.
14341 @subsection Examples
14345 Read the initial state from @file{pattern}, and specify an output of
14348 cellauto=f=pattern:s=200x400
14352 Generate a random initial row with a width of 200 cells, with a fill
14355 cellauto=ratio=2/3:s=200x200
14359 Create a pattern generated by rule 18 starting by a single alive cell
14360 centered on an initial row with width 100:
14362 cellauto=p=@@:s=100x400:full=0:rule=18
14366 Specify a more elaborated initial pattern:
14368 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14373 @anchor{coreimagesrc}
14374 @section coreimagesrc
14375 Video source generated on GPU using Apple's CoreImage API on OSX.
14377 This video source is a specialized version of the @ref{coreimage} video filter.
14378 Use a core image generator at the beginning of the applied filterchain to
14379 generate the content.
14381 The coreimagesrc video source accepts the following options:
14383 @item list_generators
14384 List all available generators along with all their respective options as well as
14385 possible minimum and maximum values along with the default values.
14387 list_generators=true
14391 Specify the size of the sourced video. For the syntax of this option, check the
14392 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14393 The default value is @code{320x240}.
14396 Specify the frame rate of the sourced video, as the number of frames
14397 generated per second. It has to be a string in the format
14398 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14399 number or a valid video frame rate abbreviation. The default value is
14403 Set the sample aspect ratio of the sourced video.
14406 Set the duration of the sourced video. See
14407 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14408 for the accepted syntax.
14410 If not specified, or the expressed duration is negative, the video is
14411 supposed to be generated forever.
14414 Additionally, all options of the @ref{coreimage} video filter are accepted.
14415 A complete filterchain can be used for further processing of the
14416 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
14417 and examples for details.
14419 @subsection Examples
14424 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
14425 given as complete and escaped command-line for Apple's standard bash shell:
14427 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
14429 This example is equivalent to the QRCode example of @ref{coreimage} without the
14430 need for a nullsrc video source.
14434 @section mandelbrot
14436 Generate a Mandelbrot set fractal, and progressively zoom towards the
14437 point specified with @var{start_x} and @var{start_y}.
14439 This source accepts the following options:
14444 Set the terminal pts value. Default value is 400.
14447 Set the terminal scale value.
14448 Must be a floating point value. Default value is 0.3.
14451 Set the inner coloring mode, that is the algorithm used to draw the
14452 Mandelbrot fractal internal region.
14454 It shall assume one of the following values:
14459 Show time until convergence.
14461 Set color based on point closest to the origin of the iterations.
14466 Default value is @var{mincol}.
14469 Set the bailout value. Default value is 10.0.
14472 Set the maximum of iterations performed by the rendering
14473 algorithm. Default value is 7189.
14476 Set outer coloring mode.
14477 It shall assume one of following values:
14479 @item iteration_count
14480 Set iteration cound mode.
14481 @item normalized_iteration_count
14482 set normalized iteration count mode.
14484 Default value is @var{normalized_iteration_count}.
14487 Set frame rate, expressed as number of frames per second. Default
14491 Set frame size. For the syntax of this option, check the "Video
14492 size" section in the ffmpeg-utils manual. Default value is "640x480".
14495 Set the initial scale value. Default value is 3.0.
14498 Set the initial x position. Must be a floating point value between
14499 -100 and 100. Default value is -0.743643887037158704752191506114774.
14502 Set the initial y position. Must be a floating point value between
14503 -100 and 100. Default value is -0.131825904205311970493132056385139.
14508 Generate various test patterns, as generated by the MPlayer test filter.
14510 The size of the generated video is fixed, and is 256x256.
14511 This source is useful in particular for testing encoding features.
14513 This source accepts the following options:
14518 Specify the frame rate of the sourced video, as the number of frames
14519 generated per second. It has to be a string in the format
14520 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14521 number or a valid video frame rate abbreviation. The default value is
14525 Set the duration of the sourced video. See
14526 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14527 for the accepted syntax.
14529 If not specified, or the expressed duration is negative, the video is
14530 supposed to be generated forever.
14534 Set the number or the name of the test to perform. Supported tests are:
14550 Default value is "all", which will cycle through the list of all tests.
14555 mptestsrc=t=dc_luma
14558 will generate a "dc_luma" test pattern.
14560 @section frei0r_src
14562 Provide a frei0r source.
14564 To enable compilation of this filter you need to install the frei0r
14565 header and configure FFmpeg with @code{--enable-frei0r}.
14567 This source accepts the following parameters:
14572 The size of the video to generate. For the syntax of this option, check the
14573 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14576 The framerate of the generated video. It may be a string of the form
14577 @var{num}/@var{den} or a frame rate abbreviation.
14580 The name to the frei0r source to load. For more information regarding frei0r and
14581 how to set the parameters, read the @ref{frei0r} section in the video filters
14584 @item filter_params
14585 A '|'-separated list of parameters to pass to the frei0r source.
14589 For example, to generate a frei0r partik0l source with size 200x200
14590 and frame rate 10 which is overlaid on the overlay filter main input:
14592 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
14597 Generate a life pattern.
14599 This source is based on a generalization of John Conway's life game.
14601 The sourced input represents a life grid, each pixel represents a cell
14602 which can be in one of two possible states, alive or dead. Every cell
14603 interacts with its eight neighbours, which are the cells that are
14604 horizontally, vertically, or diagonally adjacent.
14606 At each interaction the grid evolves according to the adopted rule,
14607 which specifies the number of neighbor alive cells which will make a
14608 cell stay alive or born. The @option{rule} option allows one to specify
14611 This source accepts the following options:
14615 Set the file from which to read the initial grid state. In the file,
14616 each non-whitespace character is considered an alive cell, and newline
14617 is used to delimit the end of each row.
14619 If this option is not specified, the initial grid is generated
14623 Set the video rate, that is the number of frames generated per second.
14626 @item random_fill_ratio, ratio
14627 Set the random fill ratio for the initial random grid. It is a
14628 floating point number value ranging from 0 to 1, defaults to 1/PHI.
14629 It is ignored when a file is specified.
14631 @item random_seed, seed
14632 Set the seed for filling the initial random grid, must be an integer
14633 included between 0 and UINT32_MAX. If not specified, or if explicitly
14634 set to -1, the filter will try to use a good random seed on a best
14640 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
14641 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
14642 @var{NS} specifies the number of alive neighbor cells which make a
14643 live cell stay alive, and @var{NB} the number of alive neighbor cells
14644 which make a dead cell to become alive (i.e. to "born").
14645 "s" and "b" can be used in place of "S" and "B", respectively.
14647 Alternatively a rule can be specified by an 18-bits integer. The 9
14648 high order bits are used to encode the next cell state if it is alive
14649 for each number of neighbor alive cells, the low order bits specify
14650 the rule for "borning" new cells. Higher order bits encode for an
14651 higher number of neighbor cells.
14652 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
14653 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
14655 Default value is "S23/B3", which is the original Conway's game of life
14656 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
14657 cells, and will born a new cell if there are three alive cells around
14661 Set the size of the output video. For the syntax of this option, check the
14662 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14664 If @option{filename} is specified, the size is set by default to the
14665 same size of the input file. If @option{size} is set, it must contain
14666 the size specified in the input file, and the initial grid defined in
14667 that file is centered in the larger resulting area.
14669 If a filename is not specified, the size value defaults to "320x240"
14670 (used for a randomly generated initial grid).
14673 If set to 1, stitch the left and right grid edges together, and the
14674 top and bottom edges also. Defaults to 1.
14677 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
14678 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
14679 value from 0 to 255.
14682 Set the color of living (or new born) cells.
14685 Set the color of dead cells. If @option{mold} is set, this is the first color
14686 used to represent a dead cell.
14689 Set mold color, for definitely dead and moldy cells.
14691 For the syntax of these 3 color options, check the "Color" section in the
14692 ffmpeg-utils manual.
14695 @subsection Examples
14699 Read a grid from @file{pattern}, and center it on a grid of size
14702 life=f=pattern:s=300x300
14706 Generate a random grid of size 200x200, with a fill ratio of 2/3:
14708 life=ratio=2/3:s=200x200
14712 Specify a custom rule for evolving a randomly generated grid:
14718 Full example with slow death effect (mold) using @command{ffplay}:
14720 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
14727 @anchor{haldclutsrc}
14729 @anchor{rgbtestsrc}
14731 @anchor{smptehdbars}
14734 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2
14736 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
14738 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
14740 The @code{color} source provides an uniformly colored input.
14742 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
14743 @ref{haldclut} filter.
14745 The @code{nullsrc} source returns unprocessed video frames. It is
14746 mainly useful to be employed in analysis / debugging tools, or as the
14747 source for filters which ignore the input data.
14749 The @code{rgbtestsrc} source generates an RGB test pattern useful for
14750 detecting RGB vs BGR issues. You should see a red, green and blue
14751 stripe from top to bottom.
14753 The @code{smptebars} source generates a color bars pattern, based on
14754 the SMPTE Engineering Guideline EG 1-1990.
14756 The @code{smptehdbars} source generates a color bars pattern, based on
14757 the SMPTE RP 219-2002.
14759 The @code{testsrc} source generates a test video pattern, showing a
14760 color pattern, a scrolling gradient and a timestamp. This is mainly
14761 intended for testing purposes.
14763 The @code{testsrc2} source is similar to testsrc, but supports more
14764 pixel formats instead of just @code{rgb24}. This allows using it as an
14765 input for other tests without requiring a format conversion.
14767 The sources accept the following parameters:
14772 Specify the color of the source, only available in the @code{color}
14773 source. For the syntax of this option, check the "Color" section in the
14774 ffmpeg-utils manual.
14777 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
14778 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
14779 pixels to be used as identity matrix for 3D lookup tables. Each component is
14780 coded on a @code{1/(N*N)} scale.
14783 Specify the size of the sourced video. For the syntax of this option, check the
14784 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14785 The default value is @code{320x240}.
14787 This option is not available with the @code{haldclutsrc} filter.
14790 Specify the frame rate of the sourced video, as the number of frames
14791 generated per second. It has to be a string in the format
14792 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14793 number or a valid video frame rate abbreviation. The default value is
14797 Set the sample aspect ratio of the sourced video.
14800 Set the duration of the sourced video. See
14801 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14802 for the accepted syntax.
14804 If not specified, or the expressed duration is negative, the video is
14805 supposed to be generated forever.
14808 Set the number of decimals to show in the timestamp, only available in the
14809 @code{testsrc} source.
14811 The displayed timestamp value will correspond to the original
14812 timestamp value multiplied by the power of 10 of the specified
14813 value. Default value is 0.
14816 For example the following:
14818 testsrc=duration=5.3:size=qcif:rate=10
14821 will generate a video with a duration of 5.3 seconds, with size
14822 176x144 and a frame rate of 10 frames per second.
14824 The following graph description will generate a red source
14825 with an opacity of 0.2, with size "qcif" and a frame rate of 10
14828 color=c=red@@0.2:s=qcif:r=10
14831 If the input content is to be ignored, @code{nullsrc} can be used. The
14832 following command generates noise in the luminance plane by employing
14833 the @code{geq} filter:
14835 nullsrc=s=256x256, geq=random(1)*255:128:128
14838 @subsection Commands
14840 The @code{color} source supports the following commands:
14844 Set the color of the created image. Accepts the same syntax of the
14845 corresponding @option{color} option.
14848 @c man end VIDEO SOURCES
14850 @chapter Video Sinks
14851 @c man begin VIDEO SINKS
14853 Below is a description of the currently available video sinks.
14855 @section buffersink
14857 Buffer video frames, and make them available to the end of the filter
14860 This sink is mainly intended for programmatic use, in particular
14861 through the interface defined in @file{libavfilter/buffersink.h}
14862 or the options system.
14864 It accepts a pointer to an AVBufferSinkContext structure, which
14865 defines the incoming buffers' formats, to be passed as the opaque
14866 parameter to @code{avfilter_init_filter} for initialization.
14870 Null video sink: do absolutely nothing with the input video. It is
14871 mainly useful as a template and for use in analysis / debugging
14874 @c man end VIDEO SINKS
14876 @chapter Multimedia Filters
14877 @c man begin MULTIMEDIA FILTERS
14879 Below is a description of the currently available multimedia filters.
14881 @section ahistogram
14883 Convert input audio to a video output, displaying the volume histogram.
14885 The filter accepts the following options:
14889 Specify how histogram is calculated.
14891 It accepts the following values:
14894 Use single histogram for all channels.
14896 Use separate histogram for each channel.
14898 Default is @code{single}.
14901 Set frame rate, expressed as number of frames per second. Default
14905 Specify the video size for the output. For the syntax of this option, check the
14906 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14907 Default value is @code{hd720}.
14912 It accepts the following values:
14923 reverse logarithmic
14925 Default is @code{log}.
14928 Set amplitude scale.
14930 It accepts the following values:
14937 Default is @code{log}.
14940 Set how much frames to accumulate in histogram.
14941 Defauls is 1. Setting this to -1 accumulates all frames.
14944 Set histogram ratio of window height.
14947 Set sonogram sliding.
14949 It accepts the following values:
14952 replace old rows with new ones.
14954 scroll from top to bottom.
14956 Default is @code{replace}.
14959 @section aphasemeter
14961 Convert input audio to a video output, displaying the audio phase.
14963 The filter accepts the following options:
14967 Set the output frame rate. Default value is @code{25}.
14970 Set the video size for the output. For the syntax of this option, check the
14971 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14972 Default value is @code{800x400}.
14977 Specify the red, green, blue contrast. Default values are @code{2},
14978 @code{7} and @code{1}.
14979 Allowed range is @code{[0, 255]}.
14982 Set color which will be used for drawing median phase. If color is
14983 @code{none} which is default, no median phase value will be drawn.
14986 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
14987 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
14988 The @code{-1} means left and right channels are completely out of phase and
14989 @code{1} means channels are in phase.
14991 @section avectorscope
14993 Convert input audio to a video output, representing the audio vector
14996 The filter is used to measure the difference between channels of stereo
14997 audio stream. A monoaural signal, consisting of identical left and right
14998 signal, results in straight vertical line. Any stereo separation is visible
14999 as a deviation from this line, creating a Lissajous figure.
15000 If the straight (or deviation from it) but horizontal line appears this
15001 indicates that the left and right channels are out of phase.
15003 The filter accepts the following options:
15007 Set the vectorscope mode.
15009 Available values are:
15012 Lissajous rotated by 45 degrees.
15015 Same as above but not rotated.
15018 Shape resembling half of circle.
15021 Default value is @samp{lissajous}.
15024 Set the video size for the output. For the syntax of this option, check the
15025 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15026 Default value is @code{400x400}.
15029 Set the output frame rate. Default value is @code{25}.
15035 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15036 @code{160}, @code{80} and @code{255}.
15037 Allowed range is @code{[0, 255]}.
15043 Specify the red, green, blue and alpha fade. Default values are @code{15},
15044 @code{10}, @code{5} and @code{5}.
15045 Allowed range is @code{[0, 255]}.
15048 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15051 Set the vectorscope drawing mode.
15053 Available values are:
15056 Draw dot for each sample.
15059 Draw line between previous and current sample.
15062 Default value is @samp{dot}.
15065 @subsection Examples
15069 Complete example using @command{ffplay}:
15071 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15072 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15076 @section bench, abench
15078 Benchmark part of a filtergraph.
15080 The filter accepts the following options:
15084 Start or stop a timer.
15086 Available values are:
15089 Get the current time, set it as frame metadata (using the key
15090 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15093 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15094 the input frame metadata to get the time difference. Time difference, average,
15095 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15096 @code{min}) are then printed. The timestamps are expressed in seconds.
15100 @subsection Examples
15104 Benchmark @ref{selectivecolor} filter:
15106 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15112 Concatenate audio and video streams, joining them together one after the
15115 The filter works on segments of synchronized video and audio streams. All
15116 segments must have the same number of streams of each type, and that will
15117 also be the number of streams at output.
15119 The filter accepts the following options:
15124 Set the number of segments. Default is 2.
15127 Set the number of output video streams, that is also the number of video
15128 streams in each segment. Default is 1.
15131 Set the number of output audio streams, that is also the number of audio
15132 streams in each segment. Default is 0.
15135 Activate unsafe mode: do not fail if segments have a different format.
15139 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15140 @var{a} audio outputs.
15142 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15143 segment, in the same order as the outputs, then the inputs for the second
15146 Related streams do not always have exactly the same duration, for various
15147 reasons including codec frame size or sloppy authoring. For that reason,
15148 related synchronized streams (e.g. a video and its audio track) should be
15149 concatenated at once. The concat filter will use the duration of the longest
15150 stream in each segment (except the last one), and if necessary pad shorter
15151 audio streams with silence.
15153 For this filter to work correctly, all segments must start at timestamp 0.
15155 All corresponding streams must have the same parameters in all segments; the
15156 filtering system will automatically select a common pixel format for video
15157 streams, and a common sample format, sample rate and channel layout for
15158 audio streams, but other settings, such as resolution, must be converted
15159 explicitly by the user.
15161 Different frame rates are acceptable but will result in variable frame rate
15162 at output; be sure to configure the output file to handle it.
15164 @subsection Examples
15168 Concatenate an opening, an episode and an ending, all in bilingual version
15169 (video in stream 0, audio in streams 1 and 2):
15171 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15172 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15173 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15174 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15178 Concatenate two parts, handling audio and video separately, using the
15179 (a)movie sources, and adjusting the resolution:
15181 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15182 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15183 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15185 Note that a desync will happen at the stitch if the audio and video streams
15186 do not have exactly the same duration in the first file.
15190 @section drawgraph, adrawgraph
15192 Draw a graph using input video or audio metadata.
15194 It accepts the following parameters:
15198 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15201 Set 1st foreground color expression.
15204 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15207 Set 2nd foreground color expression.
15210 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15213 Set 3rd foreground color expression.
15216 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15219 Set 4th foreground color expression.
15222 Set minimal value of metadata value.
15225 Set maximal value of metadata value.
15228 Set graph background color. Default is white.
15233 Available values for mode is:
15240 Default is @code{line}.
15245 Available values for slide is:
15248 Draw new frame when right border is reached.
15251 Replace old columns with new ones.
15254 Scroll from right to left.
15257 Scroll from left to right.
15260 Draw single picture.
15263 Default is @code{frame}.
15266 Set size of graph video. For the syntax of this option, check the
15267 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15268 The default value is @code{900x256}.
15270 The foreground color expressions can use the following variables:
15273 Minimal value of metadata value.
15276 Maximal value of metadata value.
15279 Current metadata key value.
15282 The color is defined as 0xAABBGGRR.
15285 Example using metadata from @ref{signalstats} filter:
15287 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15290 Example using metadata from @ref{ebur128} filter:
15292 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15298 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15299 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15300 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15301 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15303 The filter also has a video output (see the @var{video} option) with a real
15304 time graph to observe the loudness evolution. The graphic contains the logged
15305 message mentioned above, so it is not printed anymore when this option is set,
15306 unless the verbose logging is set. The main graphing area contains the
15307 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15308 the momentary loudness (400 milliseconds).
15310 More information about the Loudness Recommendation EBU R128 on
15311 @url{http://tech.ebu.ch/loudness}.
15313 The filter accepts the following options:
15318 Activate the video output. The audio stream is passed unchanged whether this
15319 option is set or no. The video stream will be the first output stream if
15320 activated. Default is @code{0}.
15323 Set the video size. This option is for video only. For the syntax of this
15325 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15326 Default and minimum resolution is @code{640x480}.
15329 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15330 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15331 other integer value between this range is allowed.
15334 Set metadata injection. If set to @code{1}, the audio input will be segmented
15335 into 100ms output frames, each of them containing various loudness information
15336 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15338 Default is @code{0}.
15341 Force the frame logging level.
15343 Available values are:
15346 information logging level
15348 verbose logging level
15351 By default, the logging level is set to @var{info}. If the @option{video} or
15352 the @option{metadata} options are set, it switches to @var{verbose}.
15357 Available modes can be cumulated (the option is a @code{flag} type). Possible
15361 Disable any peak mode (default).
15363 Enable sample-peak mode.
15365 Simple peak mode looking for the higher sample value. It logs a message
15366 for sample-peak (identified by @code{SPK}).
15368 Enable true-peak mode.
15370 If enabled, the peak lookup is done on an over-sampled version of the input
15371 stream for better peak accuracy. It logs a message for true-peak.
15372 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15373 This mode requires a build with @code{libswresample}.
15377 Treat mono input files as "dual mono". If a mono file is intended for playback
15378 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15379 If set to @code{true}, this option will compensate for this effect.
15380 Multi-channel input files are not affected by this option.
15383 Set a specific pan law to be used for the measurement of dual mono files.
15384 This parameter is optional, and has a default value of -3.01dB.
15387 @subsection Examples
15391 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
15393 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
15397 Run an analysis with @command{ffmpeg}:
15399 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
15403 @section interleave, ainterleave
15405 Temporally interleave frames from several inputs.
15407 @code{interleave} works with video inputs, @code{ainterleave} with audio.
15409 These filters read frames from several inputs and send the oldest
15410 queued frame to the output.
15412 Input streams must have a well defined, monotonically increasing frame
15415 In order to submit one frame to output, these filters need to enqueue
15416 at least one frame for each input, so they cannot work in case one
15417 input is not yet terminated and will not receive incoming frames.
15419 For example consider the case when one input is a @code{select} filter
15420 which always drop input frames. The @code{interleave} filter will keep
15421 reading from that input, but it will never be able to send new frames
15422 to output until the input will send an end-of-stream signal.
15424 Also, depending on inputs synchronization, the filters will drop
15425 frames in case one input receives more frames than the other ones, and
15426 the queue is already filled.
15428 These filters accept the following options:
15432 Set the number of different inputs, it is 2 by default.
15435 @subsection Examples
15439 Interleave frames belonging to different streams using @command{ffmpeg}:
15441 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
15445 Add flickering blur effect:
15447 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
15451 @section metadata, ametadata
15453 Manipulate frame metadata.
15455 This filter accepts the following options:
15459 Set mode of operation of the filter.
15461 Can be one of the following:
15465 If both @code{value} and @code{key} is set, select frames
15466 which have such metadata. If only @code{key} is set, select
15467 every frame that has such key in metadata.
15470 Add new metadata @code{key} and @code{value}. If key is already available
15474 Modify value of already present key.
15477 If @code{value} is set, delete only keys that have such value.
15478 Otherwise, delete key.
15481 Print key and its value if metadata was found. If @code{key} is not set print all
15482 metadata values available in frame.
15486 Set key used with all modes. Must be set for all modes except @code{print}.
15489 Set metadata value which will be used. This option is mandatory for
15490 @code{modify} and @code{add} mode.
15493 Which function to use when comparing metadata value and @code{value}.
15495 Can be one of following:
15499 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
15502 Values are interpreted as strings, returns true if metadata value starts with
15503 the @code{value} option string.
15506 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
15509 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
15512 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
15515 Values are interpreted as floats, returns true if expression from option @code{expr}
15520 Set expression which is used when @code{function} is set to @code{expr}.
15521 The expression is evaluated through the eval API and can contain the following
15526 Float representation of @code{value} from metadata key.
15529 Float representation of @code{value} as supplied by user in @code{value} option.
15532 If specified in @code{print} mode, output is written to the named file. Instead of
15533 plain filename any writable url can be specified. Filename ``-'' is a shorthand
15534 for standard output. If @code{file} option is not set, output is written to the log
15535 with AV_LOG_INFO loglevel.
15540 @subsection Examples
15544 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
15547 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
15550 Print silencedetect output to file @file{metadata.txt}.
15552 silencedetect,ametadata=mode=print:file=metadata.txt
15555 Direct all metadata to a pipe with file descriptor 4.
15557 metadata=mode=print:file='pipe\:4'
15561 @section perms, aperms
15563 Set read/write permissions for the output frames.
15565 These filters are mainly aimed at developers to test direct path in the
15566 following filter in the filtergraph.
15568 The filters accept the following options:
15572 Select the permissions mode.
15574 It accepts the following values:
15577 Do nothing. This is the default.
15579 Set all the output frames read-only.
15581 Set all the output frames directly writable.
15583 Make the frame read-only if writable, and writable if read-only.
15585 Set each output frame read-only or writable randomly.
15589 Set the seed for the @var{random} mode, must be an integer included between
15590 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
15591 @code{-1}, the filter will try to use a good random seed on a best effort
15595 Note: in case of auto-inserted filter between the permission filter and the
15596 following one, the permission might not be received as expected in that
15597 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
15598 perms/aperms filter can avoid this problem.
15600 @section realtime, arealtime
15602 Slow down filtering to match real time approximatively.
15604 These filters will pause the filtering for a variable amount of time to
15605 match the output rate with the input timestamps.
15606 They are similar to the @option{re} option to @code{ffmpeg}.
15608 They accept the following options:
15612 Time limit for the pauses. Any pause longer than that will be considered
15613 a timestamp discontinuity and reset the timer. Default is 2 seconds.
15616 @section select, aselect
15618 Select frames to pass in output.
15620 This filter accepts the following options:
15625 Set expression, which is evaluated for each input frame.
15627 If the expression is evaluated to zero, the frame is discarded.
15629 If the evaluation result is negative or NaN, the frame is sent to the
15630 first output; otherwise it is sent to the output with index
15631 @code{ceil(val)-1}, assuming that the input index starts from 0.
15633 For example a value of @code{1.2} corresponds to the output with index
15634 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
15637 Set the number of outputs. The output to which to send the selected
15638 frame is based on the result of the evaluation. Default value is 1.
15641 The expression can contain the following constants:
15645 The (sequential) number of the filtered frame, starting from 0.
15648 The (sequential) number of the selected frame, starting from 0.
15650 @item prev_selected_n
15651 The sequential number of the last selected frame. It's NAN if undefined.
15654 The timebase of the input timestamps.
15657 The PTS (Presentation TimeStamp) of the filtered video frame,
15658 expressed in @var{TB} units. It's NAN if undefined.
15661 The PTS of the filtered video frame,
15662 expressed in seconds. It's NAN if undefined.
15665 The PTS of the previously filtered video frame. It's NAN if undefined.
15667 @item prev_selected_pts
15668 The PTS of the last previously filtered video frame. It's NAN if undefined.
15670 @item prev_selected_t
15671 The PTS of the last previously selected video frame. It's NAN if undefined.
15674 The PTS of the first video frame in the video. It's NAN if undefined.
15677 The time of the first video frame in the video. It's NAN if undefined.
15679 @item pict_type @emph{(video only)}
15680 The type of the filtered frame. It can assume one of the following
15692 @item interlace_type @emph{(video only)}
15693 The frame interlace type. It can assume one of the following values:
15696 The frame is progressive (not interlaced).
15698 The frame is top-field-first.
15700 The frame is bottom-field-first.
15703 @item consumed_sample_n @emph{(audio only)}
15704 the number of selected samples before the current frame
15706 @item samples_n @emph{(audio only)}
15707 the number of samples in the current frame
15709 @item sample_rate @emph{(audio only)}
15710 the input sample rate
15713 This is 1 if the filtered frame is a key-frame, 0 otherwise.
15716 the position in the file of the filtered frame, -1 if the information
15717 is not available (e.g. for synthetic video)
15719 @item scene @emph{(video only)}
15720 value between 0 and 1 to indicate a new scene; a low value reflects a low
15721 probability for the current frame to introduce a new scene, while a higher
15722 value means the current frame is more likely to be one (see the example below)
15724 @item concatdec_select
15725 The concat demuxer can select only part of a concat input file by setting an
15726 inpoint and an outpoint, but the output packets may not be entirely contained
15727 in the selected interval. By using this variable, it is possible to skip frames
15728 generated by the concat demuxer which are not exactly contained in the selected
15731 This works by comparing the frame pts against the @var{lavf.concat.start_time}
15732 and the @var{lavf.concat.duration} packet metadata values which are also
15733 present in the decoded frames.
15735 The @var{concatdec_select} variable is -1 if the frame pts is at least
15736 start_time and either the duration metadata is missing or the frame pts is less
15737 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
15740 That basically means that an input frame is selected if its pts is within the
15741 interval set by the concat demuxer.
15745 The default value of the select expression is "1".
15747 @subsection Examples
15751 Select all frames in input:
15756 The example above is the same as:
15768 Select only I-frames:
15770 select='eq(pict_type\,I)'
15774 Select one frame every 100:
15776 select='not(mod(n\,100))'
15780 Select only frames contained in the 10-20 time interval:
15782 select=between(t\,10\,20)
15786 Select only I-frames contained in the 10-20 time interval:
15788 select=between(t\,10\,20)*eq(pict_type\,I)
15792 Select frames with a minimum distance of 10 seconds:
15794 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
15798 Use aselect to select only audio frames with samples number > 100:
15800 aselect='gt(samples_n\,100)'
15804 Create a mosaic of the first scenes:
15806 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
15809 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
15813 Send even and odd frames to separate outputs, and compose them:
15815 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
15819 Select useful frames from an ffconcat file which is using inpoints and
15820 outpoints but where the source files are not intra frame only.
15822 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
15826 @section sendcmd, asendcmd
15828 Send commands to filters in the filtergraph.
15830 These filters read commands to be sent to other filters in the
15833 @code{sendcmd} must be inserted between two video filters,
15834 @code{asendcmd} must be inserted between two audio filters, but apart
15835 from that they act the same way.
15837 The specification of commands can be provided in the filter arguments
15838 with the @var{commands} option, or in a file specified by the
15839 @var{filename} option.
15841 These filters accept the following options:
15844 Set the commands to be read and sent to the other filters.
15846 Set the filename of the commands to be read and sent to the other
15850 @subsection Commands syntax
15852 A commands description consists of a sequence of interval
15853 specifications, comprising a list of commands to be executed when a
15854 particular event related to that interval occurs. The occurring event
15855 is typically the current frame time entering or leaving a given time
15858 An interval is specified by the following syntax:
15860 @var{START}[-@var{END}] @var{COMMANDS};
15863 The time interval is specified by the @var{START} and @var{END} times.
15864 @var{END} is optional and defaults to the maximum time.
15866 The current frame time is considered within the specified interval if
15867 it is included in the interval [@var{START}, @var{END}), that is when
15868 the time is greater or equal to @var{START} and is lesser than
15871 @var{COMMANDS} consists of a sequence of one or more command
15872 specifications, separated by ",", relating to that interval. The
15873 syntax of a command specification is given by:
15875 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
15878 @var{FLAGS} is optional and specifies the type of events relating to
15879 the time interval which enable sending the specified command, and must
15880 be a non-null sequence of identifier flags separated by "+" or "|" and
15881 enclosed between "[" and "]".
15883 The following flags are recognized:
15886 The command is sent when the current frame timestamp enters the
15887 specified interval. In other words, the command is sent when the
15888 previous frame timestamp was not in the given interval, and the
15892 The command is sent when the current frame timestamp leaves the
15893 specified interval. In other words, the command is sent when the
15894 previous frame timestamp was in the given interval, and the
15898 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
15901 @var{TARGET} specifies the target of the command, usually the name of
15902 the filter class or a specific filter instance name.
15904 @var{COMMAND} specifies the name of the command for the target filter.
15906 @var{ARG} is optional and specifies the optional list of argument for
15907 the given @var{COMMAND}.
15909 Between one interval specification and another, whitespaces, or
15910 sequences of characters starting with @code{#} until the end of line,
15911 are ignored and can be used to annotate comments.
15913 A simplified BNF description of the commands specification syntax
15916 @var{COMMAND_FLAG} ::= "enter" | "leave"
15917 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
15918 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
15919 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
15920 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
15921 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
15924 @subsection Examples
15928 Specify audio tempo change at second 4:
15930 asendcmd=c='4.0 atempo tempo 1.5',atempo
15934 Specify a list of drawtext and hue commands in a file.
15936 # show text in the interval 5-10
15937 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
15938 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
15940 # desaturate the image in the interval 15-20
15941 15.0-20.0 [enter] hue s 0,
15942 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
15944 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
15946 # apply an exponential saturation fade-out effect, starting from time 25
15947 25 [enter] hue s exp(25-t)
15950 A filtergraph allowing to read and process the above command list
15951 stored in a file @file{test.cmd}, can be specified with:
15953 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
15958 @section setpts, asetpts
15960 Change the PTS (presentation timestamp) of the input frames.
15962 @code{setpts} works on video frames, @code{asetpts} on audio frames.
15964 This filter accepts the following options:
15969 The expression which is evaluated for each frame to construct its timestamp.
15973 The expression is evaluated through the eval API and can contain the following
15978 frame rate, only defined for constant frame-rate video
15981 The presentation timestamp in input
15984 The count of the input frame for video or the number of consumed samples,
15985 not including the current frame for audio, starting from 0.
15987 @item NB_CONSUMED_SAMPLES
15988 The number of consumed samples, not including the current frame (only
15991 @item NB_SAMPLES, S
15992 The number of samples in the current frame (only audio)
15994 @item SAMPLE_RATE, SR
15995 The audio sample rate.
15998 The PTS of the first frame.
16001 the time in seconds of the first frame
16004 State whether the current frame is interlaced.
16007 the time in seconds of the current frame
16010 original position in the file of the frame, or undefined if undefined
16011 for the current frame
16014 The previous input PTS.
16017 previous input time in seconds
16020 The previous output PTS.
16023 previous output time in seconds
16026 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16030 The wallclock (RTC) time at the start of the movie in microseconds.
16033 The timebase of the input timestamps.
16037 @subsection Examples
16041 Start counting PTS from zero
16043 setpts=PTS-STARTPTS
16047 Apply fast motion effect:
16053 Apply slow motion effect:
16059 Set fixed rate of 25 frames per second:
16065 Set fixed rate 25 fps with some jitter:
16067 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16071 Apply an offset of 10 seconds to the input PTS:
16077 Generate timestamps from a "live source" and rebase onto the current timebase:
16079 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16083 Generate timestamps by counting samples:
16090 @section settb, asettb
16092 Set the timebase to use for the output frames timestamps.
16093 It is mainly useful for testing timebase configuration.
16095 It accepts the following parameters:
16100 The expression which is evaluated into the output timebase.
16104 The value for @option{tb} is an arithmetic expression representing a
16105 rational. The expression can contain the constants "AVTB" (the default
16106 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16107 audio only). Default value is "intb".
16109 @subsection Examples
16113 Set the timebase to 1/25:
16119 Set the timebase to 1/10:
16125 Set the timebase to 1001/1000:
16131 Set the timebase to 2*intb:
16137 Set the default timebase value:
16144 Convert input audio to a video output representing frequency spectrum
16145 logarithmically using Brown-Puckette constant Q transform algorithm with
16146 direct frequency domain coefficient calculation (but the transform itself
16147 is not really constant Q, instead the Q factor is actually variable/clamped),
16148 with musical tone scale, from E0 to D#10.
16150 The filter accepts the following options:
16154 Specify the video size for the output. It must be even. For the syntax of this option,
16155 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16156 Default value is @code{1920x1080}.
16159 Set the output frame rate. Default value is @code{25}.
16162 Set the bargraph height. It must be even. Default value is @code{-1} which
16163 computes the bargraph height automatically.
16166 Set the axis height. It must be even. Default value is @code{-1} which computes
16167 the axis height automatically.
16170 Set the sonogram height. It must be even. Default value is @code{-1} which
16171 computes the sonogram height automatically.
16174 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16175 instead. Default value is @code{1}.
16177 @item sono_v, volume
16178 Specify the sonogram volume expression. It can contain variables:
16181 the @var{bar_v} evaluated expression
16182 @item frequency, freq, f
16183 the frequency where it is evaluated
16184 @item timeclamp, tc
16185 the value of @var{timeclamp} option
16189 @item a_weighting(f)
16190 A-weighting of equal loudness
16191 @item b_weighting(f)
16192 B-weighting of equal loudness
16193 @item c_weighting(f)
16194 C-weighting of equal loudness.
16196 Default value is @code{16}.
16198 @item bar_v, volume2
16199 Specify the bargraph volume expression. It can contain variables:
16202 the @var{sono_v} evaluated expression
16203 @item frequency, freq, f
16204 the frequency where it is evaluated
16205 @item timeclamp, tc
16206 the value of @var{timeclamp} option
16210 @item a_weighting(f)
16211 A-weighting of equal loudness
16212 @item b_weighting(f)
16213 B-weighting of equal loudness
16214 @item c_weighting(f)
16215 C-weighting of equal loudness.
16217 Default value is @code{sono_v}.
16219 @item sono_g, gamma
16220 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16221 higher gamma makes the spectrum having more range. Default value is @code{3}.
16222 Acceptable range is @code{[1, 7]}.
16224 @item bar_g, gamma2
16225 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16228 @item timeclamp, tc
16229 Specify the transform timeclamp. At low frequency, there is trade-off between
16230 accuracy in time domain and frequency domain. If timeclamp is lower,
16231 event in time domain is represented more accurately (such as fast bass drum),
16232 otherwise event in frequency domain is represented more accurately
16233 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16236 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16237 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16240 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16241 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16244 This option is deprecated and ignored.
16247 Specify the transform length in time domain. Use this option to control accuracy
16248 trade-off between time domain and frequency domain at every frequency sample.
16249 It can contain variables:
16251 @item frequency, freq, f
16252 the frequency where it is evaluated
16253 @item timeclamp, tc
16254 the value of @var{timeclamp} option.
16256 Default value is @code{384*tc/(384+tc*f)}.
16259 Specify the transform count for every video frame. Default value is @code{6}.
16260 Acceptable range is @code{[1, 30]}.
16263 Specify the transform count for every single pixel. Default value is @code{0},
16264 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16267 Specify font file for use with freetype to draw the axis. If not specified,
16268 use embedded font. Note that drawing with font file or embedded font is not
16269 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16273 Specify font color expression. This is arithmetic expression that should return
16274 integer value 0xRRGGBB. It can contain variables:
16276 @item frequency, freq, f
16277 the frequency where it is evaluated
16278 @item timeclamp, tc
16279 the value of @var{timeclamp} option
16284 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16285 @item r(x), g(x), b(x)
16286 red, green, and blue value of intensity x.
16288 Default value is @code{st(0, (midi(f)-59.5)/12);
16289 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16290 r(1-ld(1)) + b(ld(1))}.
16293 Specify image file to draw the axis. This option override @var{fontfile} and
16294 @var{fontcolor} option.
16297 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16298 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16299 Default value is @code{1}.
16303 @subsection Examples
16307 Playing audio while showing the spectrum:
16309 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16313 Same as above, but with frame rate 30 fps:
16315 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16319 Playing at 1280x720:
16321 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16325 Disable sonogram display:
16331 A1 and its harmonics: A1, A2, (near)E3, A3:
16333 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),
16334 asplit[a][out1]; [a] showcqt [out0]'
16338 Same as above, but with more accuracy in frequency domain:
16340 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),
16341 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16347 bar_v=10:sono_v=bar_v*a_weighting(f)
16351 Custom gamma, now spectrum is linear to the amplitude.
16357 Custom tlength equation:
16359 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)))'
16363 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
16365 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
16369 Custom frequency range with custom axis using image file:
16371 axisfile=myaxis.png:basefreq=40:endfreq=10000
16377 Convert input audio to video output representing the audio power spectrum.
16378 Audio amplitude is on Y-axis while frequency is on X-axis.
16380 The filter accepts the following options:
16384 Specify size of video. For the syntax of this option, check the
16385 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16386 Default is @code{1024x512}.
16390 This set how each frequency bin will be represented.
16392 It accepts the following values:
16398 Default is @code{bar}.
16401 Set amplitude scale.
16403 It accepts the following values:
16417 Default is @code{log}.
16420 Set frequency scale.
16422 It accepts the following values:
16431 Reverse logarithmic scale.
16433 Default is @code{lin}.
16438 It accepts the following values:
16454 Default is @code{w2048}
16457 Set windowing function.
16459 It accepts the following values:
16477 Default is @code{hanning}.
16480 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
16481 which means optimal overlap for selected window function will be picked.
16484 Set time averaging. Setting this to 0 will display current maximal peaks.
16485 Default is @code{1}, which means time averaging is disabled.
16488 Specify list of colors separated by space or by '|' which will be used to
16489 draw channel frequencies. Unrecognized or missing colors will be replaced
16493 Set channel display mode.
16495 It accepts the following values:
16500 Default is @code{combined}.
16504 @anchor{showspectrum}
16505 @section showspectrum
16507 Convert input audio to a video output, representing the audio frequency
16510 The filter accepts the following options:
16514 Specify the video size for the output. For the syntax of this option, check the
16515 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16516 Default value is @code{640x512}.
16519 Specify how the spectrum should slide along the window.
16521 It accepts the following values:
16524 the samples start again on the left when they reach the right
16526 the samples scroll from right to left
16528 the samples scroll from left to right
16530 frames are only produced when the samples reach the right
16533 Default value is @code{replace}.
16536 Specify display mode.
16538 It accepts the following values:
16541 all channels are displayed in the same row
16543 all channels are displayed in separate rows
16546 Default value is @samp{combined}.
16549 Specify display color mode.
16551 It accepts the following values:
16554 each channel is displayed in a separate color
16556 each channel is displayed using the same color scheme
16558 each channel is displayed using the rainbow color scheme
16560 each channel is displayed using the moreland color scheme
16562 each channel is displayed using the nebulae color scheme
16564 each channel is displayed using the fire color scheme
16566 each channel is displayed using the fiery color scheme
16568 each channel is displayed using the fruit color scheme
16570 each channel is displayed using the cool color scheme
16573 Default value is @samp{channel}.
16576 Specify scale used for calculating intensity color values.
16578 It accepts the following values:
16583 square root, default
16594 Default value is @samp{sqrt}.
16597 Set saturation modifier for displayed colors. Negative values provide
16598 alternative color scheme. @code{0} is no saturation at all.
16599 Saturation must be in [-10.0, 10.0] range.
16600 Default value is @code{1}.
16603 Set window function.
16605 It accepts the following values:
16625 Default value is @code{hann}.
16628 Set orientation of time vs frequency axis. Can be @code{vertical} or
16629 @code{horizontal}. Default is @code{vertical}.
16632 Set ratio of overlap window. Default value is @code{0}.
16633 When value is @code{1} overlap is set to recommended size for specific
16634 window function currently used.
16637 Set scale gain for calculating intensity color values.
16638 Default value is @code{1}.
16641 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
16644 Set color rotation, must be in [-1.0, 1.0] range.
16645 Default value is @code{0}.
16648 The usage is very similar to the showwaves filter; see the examples in that
16651 @subsection Examples
16655 Large window with logarithmic color scaling:
16657 showspectrum=s=1280x480:scale=log
16661 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
16663 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
16664 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
16668 @section showspectrumpic
16670 Convert input audio to a single video frame, representing the audio frequency
16673 The filter accepts the following options:
16677 Specify the video size for the output. For the syntax of this option, check the
16678 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16679 Default value is @code{4096x2048}.
16682 Specify display mode.
16684 It accepts the following values:
16687 all channels are displayed in the same row
16689 all channels are displayed in separate rows
16691 Default value is @samp{combined}.
16694 Specify display color mode.
16696 It accepts the following values:
16699 each channel is displayed in a separate color
16701 each channel is displayed using the same color scheme
16703 each channel is displayed using the rainbow color scheme
16705 each channel is displayed using the moreland color scheme
16707 each channel is displayed using the nebulae color scheme
16709 each channel is displayed using the fire color scheme
16711 each channel is displayed using the fiery color scheme
16713 each channel is displayed using the fruit color scheme
16715 each channel is displayed using the cool color scheme
16717 Default value is @samp{intensity}.
16720 Specify scale used for calculating intensity color values.
16722 It accepts the following values:
16727 square root, default
16737 Default value is @samp{log}.
16740 Set saturation modifier for displayed colors. Negative values provide
16741 alternative color scheme. @code{0} is no saturation at all.
16742 Saturation must be in [-10.0, 10.0] range.
16743 Default value is @code{1}.
16746 Set window function.
16748 It accepts the following values:
16767 Default value is @code{hann}.
16770 Set orientation of time vs frequency axis. Can be @code{vertical} or
16771 @code{horizontal}. Default is @code{vertical}.
16774 Set scale gain for calculating intensity color values.
16775 Default value is @code{1}.
16778 Draw time and frequency axes and legends. Default is enabled.
16781 Set color rotation, must be in [-1.0, 1.0] range.
16782 Default value is @code{0}.
16785 @subsection Examples
16789 Extract an audio spectrogram of a whole audio track
16790 in a 1024x1024 picture using @command{ffmpeg}:
16792 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
16796 @section showvolume
16798 Convert input audio volume to a video output.
16800 The filter accepts the following options:
16807 Set border width, allowed range is [0, 5]. Default is 1.
16810 Set channel width, allowed range is [80, 8192]. Default is 400.
16813 Set channel height, allowed range is [1, 900]. Default is 20.
16816 Set fade, allowed range is [0.001, 1]. Default is 0.95.
16819 Set volume color expression.
16821 The expression can use the following variables:
16825 Current max volume of channel in dB.
16828 Current channel number, starting from 0.
16832 If set, displays channel names. Default is enabled.
16835 If set, displays volume values. Default is enabled.
16838 Set orientation, can be @code{horizontal} or @code{vertical},
16839 default is @code{horizontal}.
16842 Set step size, allowed range s [0, 5]. Default is 0, which means
16848 Convert input audio to a video output, representing the samples waves.
16850 The filter accepts the following options:
16854 Specify the video size for the output. For the syntax of this option, check the
16855 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16856 Default value is @code{600x240}.
16861 Available values are:
16864 Draw a point for each sample.
16867 Draw a vertical line for each sample.
16870 Draw a point for each sample and a line between them.
16873 Draw a centered vertical line for each sample.
16876 Default value is @code{point}.
16879 Set the number of samples which are printed on the same column. A
16880 larger value will decrease the frame rate. Must be a positive
16881 integer. This option can be set only if the value for @var{rate}
16882 is not explicitly specified.
16885 Set the (approximate) output frame rate. This is done by setting the
16886 option @var{n}. Default value is "25".
16888 @item split_channels
16889 Set if channels should be drawn separately or overlap. Default value is 0.
16892 Set colors separated by '|' which are going to be used for drawing of each channel.
16895 Set amplitude scale.
16897 Available values are:
16915 @subsection Examples
16919 Output the input file audio and the corresponding video representation
16922 amovie=a.mp3,asplit[out0],showwaves[out1]
16926 Create a synthetic signal and show it with showwaves, forcing a
16927 frame rate of 30 frames per second:
16929 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
16933 @section showwavespic
16935 Convert input audio to a single video frame, representing the samples waves.
16937 The filter accepts the following options:
16941 Specify the video size for the output. For the syntax of this option, check the
16942 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16943 Default value is @code{600x240}.
16945 @item split_channels
16946 Set if channels should be drawn separately or overlap. Default value is 0.
16949 Set colors separated by '|' which are going to be used for drawing of each channel.
16952 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
16956 @subsection Examples
16960 Extract a channel split representation of the wave form of a whole audio track
16961 in a 1024x800 picture using @command{ffmpeg}:
16963 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
16967 @section spectrumsynth
16969 Sythesize audio from 2 input video spectrums, first input stream represents
16970 magnitude across time and second represents phase across time.
16971 The filter will transform from frequency domain as displayed in videos back
16972 to time domain as presented in audio output.
16974 This filter is primarly created for reversing processed @ref{showspectrum}
16975 filter outputs, but can synthesize sound from other spectrograms too.
16976 But in such case results are going to be poor if the phase data is not
16977 available, because in such cases phase data need to be recreated, usually
16978 its just recreated from random noise.
16979 For best results use gray only output (@code{channel} color mode in
16980 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
16981 @code{lin} scale for phase video. To produce phase, for 2nd video, use
16982 @code{data} option. Inputs videos should generally use @code{fullframe}
16983 slide mode as that saves resources needed for decoding video.
16985 The filter accepts the following options:
16989 Specify sample rate of output audio, the sample rate of audio from which
16990 spectrum was generated may differ.
16993 Set number of channels represented in input video spectrums.
16996 Set scale which was used when generating magnitude input spectrum.
16997 Can be @code{lin} or @code{log}. Default is @code{log}.
17000 Set slide which was used when generating inputs spectrums.
17001 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17002 Default is @code{fullframe}.
17005 Set window function used for resynthesis.
17008 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17009 which means optimal overlap for selected window function will be picked.
17012 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17013 Default is @code{vertical}.
17016 @subsection Examples
17020 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17021 then resynthesize videos back to audio with spectrumsynth:
17023 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
17024 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
17025 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17029 @section split, asplit
17031 Split input into several identical outputs.
17033 @code{asplit} works with audio input, @code{split} with video.
17035 The filter accepts a single parameter which specifies the number of outputs. If
17036 unspecified, it defaults to 2.
17038 @subsection Examples
17042 Create two separate outputs from the same input:
17044 [in] split [out0][out1]
17048 To create 3 or more outputs, you need to specify the number of
17051 [in] asplit=3 [out0][out1][out2]
17055 Create two separate outputs from the same input, one cropped and
17058 [in] split [splitout1][splitout2];
17059 [splitout1] crop=100:100:0:0 [cropout];
17060 [splitout2] pad=200:200:100:100 [padout];
17064 Create 5 copies of the input audio with @command{ffmpeg}:
17066 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17072 Receive commands sent through a libzmq client, and forward them to
17073 filters in the filtergraph.
17075 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17076 must be inserted between two video filters, @code{azmq} between two
17079 To enable these filters you need to install the libzmq library and
17080 headers and configure FFmpeg with @code{--enable-libzmq}.
17082 For more information about libzmq see:
17083 @url{http://www.zeromq.org/}
17085 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17086 receives messages sent through a network interface defined by the
17087 @option{bind_address} option.
17089 The received message must be in the form:
17091 @var{TARGET} @var{COMMAND} [@var{ARG}]
17094 @var{TARGET} specifies the target of the command, usually the name of
17095 the filter class or a specific filter instance name.
17097 @var{COMMAND} specifies the name of the command for the target filter.
17099 @var{ARG} is optional and specifies the optional argument list for the
17100 given @var{COMMAND}.
17102 Upon reception, the message is processed and the corresponding command
17103 is injected into the filtergraph. Depending on the result, the filter
17104 will send a reply to the client, adopting the format:
17106 @var{ERROR_CODE} @var{ERROR_REASON}
17110 @var{MESSAGE} is optional.
17112 @subsection Examples
17114 Look at @file{tools/zmqsend} for an example of a zmq client which can
17115 be used to send commands processed by these filters.
17117 Consider the following filtergraph generated by @command{ffplay}
17119 ffplay -dumpgraph 1 -f lavfi "
17120 color=s=100x100:c=red [l];
17121 color=s=100x100:c=blue [r];
17122 nullsrc=s=200x100, zmq [bg];
17123 [bg][l] overlay [bg+l];
17124 [bg+l][r] overlay=x=100 "
17127 To change the color of the left side of the video, the following
17128 command can be used:
17130 echo Parsed_color_0 c yellow | tools/zmqsend
17133 To change the right side:
17135 echo Parsed_color_1 c pink | tools/zmqsend
17138 @c man end MULTIMEDIA FILTERS
17140 @chapter Multimedia Sources
17141 @c man begin MULTIMEDIA SOURCES
17143 Below is a description of the currently available multimedia sources.
17147 This is the same as @ref{movie} source, except it selects an audio
17153 Read audio and/or video stream(s) from a movie container.
17155 It accepts the following parameters:
17159 The name of the resource to read (not necessarily a file; it can also be a
17160 device or a stream accessed through some protocol).
17162 @item format_name, f
17163 Specifies the format assumed for the movie to read, and can be either
17164 the name of a container or an input device. If not specified, the
17165 format is guessed from @var{movie_name} or by probing.
17167 @item seek_point, sp
17168 Specifies the seek point in seconds. The frames will be output
17169 starting from this seek point. The parameter is evaluated with
17170 @code{av_strtod}, so the numerical value may be suffixed by an IS
17171 postfix. The default value is "0".
17174 Specifies the streams to read. Several streams can be specified,
17175 separated by "+". The source will then have as many outputs, in the
17176 same order. The syntax is explained in the ``Stream specifiers''
17177 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17178 respectively the default (best suited) video and audio stream. Default
17179 is "dv", or "da" if the filter is called as "amovie".
17181 @item stream_index, si
17182 Specifies the index of the video stream to read. If the value is -1,
17183 the most suitable video stream will be automatically selected. The default
17184 value is "-1". Deprecated. If the filter is called "amovie", it will select
17185 audio instead of video.
17188 Specifies how many times to read the stream in sequence.
17189 If the value is less than 1, the stream will be read again and again.
17190 Default value is "1".
17192 Note that when the movie is looped the source timestamps are not
17193 changed, so it will generate non monotonically increasing timestamps.
17195 @item discontinuity
17196 Specifies the time difference between frames above which the point is
17197 considered a timestamp discontinuity which is removed by adjusting the later
17201 It allows overlaying a second video on top of the main input of
17202 a filtergraph, as shown in this graph:
17204 input -----------> deltapts0 --> overlay --> output
17207 movie --> scale--> deltapts1 -------+
17209 @subsection Examples
17213 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17214 on top of the input labelled "in":
17216 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17217 [in] setpts=PTS-STARTPTS [main];
17218 [main][over] overlay=16:16 [out]
17222 Read from a video4linux2 device, and overlay it on top of the input
17225 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17226 [in] setpts=PTS-STARTPTS [main];
17227 [main][over] overlay=16:16 [out]
17231 Read the first video stream and the audio stream with id 0x81 from
17232 dvd.vob; the video is connected to the pad named "video" and the audio is
17233 connected to the pad named "audio":
17235 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17239 @subsection Commands
17241 Both movie and amovie support the following commands:
17244 Perform seek using "av_seek_frame".
17245 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17248 @var{stream_index}: If stream_index is -1, a default
17249 stream is selected, and @var{timestamp} is automatically converted
17250 from AV_TIME_BASE units to the stream specific time_base.
17252 @var{timestamp}: Timestamp in AVStream.time_base units
17253 or, if no stream is specified, in AV_TIME_BASE units.
17255 @var{flags}: Flags which select direction and seeking mode.
17259 Get movie duration in AV_TIME_BASE units.
17263 @c man end MULTIMEDIA SOURCES