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 continuous values instead of discrete
454 Additionally it has a D/C offset which results in different crushing of
455 the lower and the upper half of the signal.
456 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
458 Another feature of this filter is the logarithmic mode.
459 This setting switches from linear distances between bits to logarithmic ones.
460 The result is a much more "natural" sounding crusher which doesn't gate low
461 signals for example. The human ear has a logarithmic perception, too
462 so this kind of crushing is much more pleasant.
463 Logarithmic crushing is also able to get anti-aliased.
465 The filter accepts the following options:
481 Can be linear: @code{lin} or logarithmic: @code{log}.
490 Set sample reduction.
493 Enable LFO. By default disabled.
504 Delay one or more audio channels.
506 Samples in delayed channel are filled with silence.
508 The filter accepts the following option:
512 Set list of delays in milliseconds for each channel separated by '|'.
513 At least one delay greater than 0 should be provided.
514 Unused delays will be silently ignored. If number of given delays is
515 smaller than number of channels all remaining channels will not be delayed.
516 If you want to delay exact number of samples, append 'S' to number.
523 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
524 the second channel (and any other channels that may be present) unchanged.
530 Delay second channel by 500 samples, the third channel by 700 samples and leave
531 the first channel (and any other channels that may be present) unchanged.
539 Apply echoing to the input audio.
541 Echoes are reflected sound and can occur naturally amongst mountains
542 (and sometimes large buildings) when talking or shouting; digital echo
543 effects emulate this behaviour and are often used to help fill out the
544 sound of a single instrument or vocal. The time difference between the
545 original signal and the reflection is the @code{delay}, and the
546 loudness of the reflected signal is the @code{decay}.
547 Multiple echoes can have different delays and decays.
549 A description of the accepted parameters follows.
553 Set input gain of reflected signal. Default is @code{0.6}.
556 Set output gain of reflected signal. Default is @code{0.3}.
559 Set list of time intervals in milliseconds between original signal and reflections
560 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
561 Default is @code{1000}.
564 Set list of loudnesses of reflected signals separated by '|'.
565 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
566 Default is @code{0.5}.
573 Make it sound as if there are twice as many instruments as are actually playing:
575 aecho=0.8:0.88:60:0.4
579 If delay is very short, then it sound like a (metallic) robot playing music:
585 A longer delay will sound like an open air concert in the mountains:
587 aecho=0.8:0.9:1000:0.3
591 Same as above but with one more mountain:
593 aecho=0.8:0.9:1000|1800:0.3|0.25
598 Audio emphasis filter creates or restores material directly taken from LPs or
599 emphased CDs with different filter curves. E.g. to store music on vinyl the
600 signal has to be altered by a filter first to even out the disadvantages of
601 this recording medium.
602 Once the material is played back the inverse filter has to be applied to
603 restore the distortion of the frequency response.
605 The filter accepts the following options:
615 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
616 use @code{production} mode. Default is @code{reproduction} mode.
619 Set filter type. Selects medium. Can be one of the following:
631 select Compact Disc (CD).
637 select 50µs (FM-KF).
639 select 75µs (FM-KF).
645 Modify an audio signal according to the specified expressions.
647 This filter accepts one or more expressions (one for each channel),
648 which are evaluated and used to modify a corresponding audio signal.
650 It accepts the following parameters:
654 Set the '|'-separated expressions list for each separate channel. If
655 the number of input channels is greater than the number of
656 expressions, the last specified expression is used for the remaining
659 @item channel_layout, c
660 Set output channel layout. If not specified, the channel layout is
661 specified by the number of expressions. If set to @samp{same}, it will
662 use by default the same input channel layout.
665 Each expression in @var{exprs} can contain the following constants and functions:
669 channel number of the current expression
672 number of the evaluated sample, starting from 0
678 time of the evaluated sample expressed in seconds
681 @item nb_out_channels
682 input and output number of channels
685 the value of input channel with number @var{CH}
688 Note: this filter is slow. For faster processing you should use a
697 aeval=val(ch)/2:c=same
701 Invert phase of the second channel:
710 Apply fade-in/out effect to input audio.
712 A description of the accepted parameters follows.
716 Specify the effect type, can be either @code{in} for fade-in, or
717 @code{out} for a fade-out effect. Default is @code{in}.
719 @item start_sample, ss
720 Specify the number of the start sample for starting to apply the fade
721 effect. Default is 0.
724 Specify the number of samples for which the fade effect has to last. At
725 the end of the fade-in effect the output audio will have the same
726 volume as the input audio, at the end of the fade-out transition
727 the output audio will be silence. Default is 44100.
730 Specify the start time of the fade effect. Default is 0.
731 The value must be specified as a time duration; see
732 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
733 for the accepted syntax.
734 If set this option is used instead of @var{start_sample}.
737 Specify the duration of the fade effect. See
738 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
739 for the accepted syntax.
740 At the end of the fade-in effect the output audio will have the same
741 volume as the input audio, at the end of the fade-out transition
742 the output audio will be silence.
743 By default the duration is determined by @var{nb_samples}.
744 If set this option is used instead of @var{nb_samples}.
747 Set curve for fade transition.
749 It accepts the following values:
752 select triangular, linear slope (default)
754 select quarter of sine wave
756 select half of sine wave
758 select exponential sine wave
762 select inverted parabola
776 select inverted quarter of sine wave
778 select inverted half of sine wave
780 select double-exponential seat
782 select double-exponential sigmoid
790 Fade in first 15 seconds of audio:
796 Fade out last 25 seconds of a 900 seconds audio:
798 afade=t=out:st=875:d=25
803 Apply arbitrary expressions to samples in frequency domain.
807 Set frequency domain real expression for each separate channel separated
808 by '|'. Default is "1".
809 If the number of input channels is greater than the number of
810 expressions, the last specified expression is used for the remaining
814 Set frequency domain imaginary expression for each separate channel
815 separated by '|'. If not set, @var{real} option is used.
817 Each expression in @var{real} and @var{imag} can contain the following
825 current frequency bin number
828 number of available bins
831 channel number of the current expression
843 It accepts the following values:
859 Default is @code{w4096}
862 Set window function. Default is @code{hann}.
865 Set window overlap. If set to 1, the recommended overlap for selected
866 window function will be picked. Default is @code{0.75}.
873 Leave almost only low frequencies in audio:
875 afftfilt="1-clip((b/nb)*b,0,1)"
882 Set output format constraints for the input audio. The framework will
883 negotiate the most appropriate format to minimize conversions.
885 It accepts the following parameters:
889 A '|'-separated list of requested sample formats.
892 A '|'-separated list of requested sample rates.
894 @item channel_layouts
895 A '|'-separated list of requested channel layouts.
897 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
898 for the required syntax.
901 If a parameter is omitted, all values are allowed.
903 Force the output to either unsigned 8-bit or signed 16-bit stereo
905 aformat=sample_fmts=u8|s16:channel_layouts=stereo
910 A gate is mainly used to reduce lower parts of a signal. This kind of signal
911 processing reduces disturbing noise between useful signals.
913 Gating is done by detecting the volume below a chosen level @var{threshold}
914 and dividing it by the factor set with @var{ratio}. The bottom of the noise
915 floor is set via @var{range}. Because an exact manipulation of the signal
916 would cause distortion of the waveform the reduction can be levelled over
917 time. This is done by setting @var{attack} and @var{release}.
919 @var{attack} determines how long the signal has to fall below the threshold
920 before any reduction will occur and @var{release} sets the time the signal
921 has to rise above the threshold to reduce the reduction again.
922 Shorter signals than the chosen attack time will be left untouched.
926 Set input level before filtering.
927 Default is 1. Allowed range is from 0.015625 to 64.
930 Set the level of gain reduction when the signal is below the threshold.
931 Default is 0.06125. Allowed range is from 0 to 1.
934 If a signal rises above this level the gain reduction is released.
935 Default is 0.125. Allowed range is from 0 to 1.
938 Set a ratio by which the signal is reduced.
939 Default is 2. Allowed range is from 1 to 9000.
942 Amount of milliseconds the signal has to rise above the threshold before gain
944 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
947 Amount of milliseconds the signal has to fall below the threshold before the
948 reduction is increased again. Default is 250 milliseconds.
949 Allowed range is from 0.01 to 9000.
952 Set amount of amplification of signal after processing.
953 Default is 1. Allowed range is from 1 to 64.
956 Curve the sharp knee around the threshold to enter gain reduction more softly.
957 Default is 2.828427125. Allowed range is from 1 to 8.
960 Choose if exact signal should be taken for detection or an RMS like one.
961 Default is @code{rms}. Can be @code{peak} or @code{rms}.
964 Choose if the average level between all channels or the louder channel affects
966 Default is @code{average}. Can be @code{average} or @code{maximum}.
971 The limiter prevents an input signal from rising over a desired threshold.
972 This limiter uses lookahead technology to prevent your signal from distorting.
973 It means that there is a small delay after the signal is processed. Keep in mind
974 that the delay it produces is the attack time you set.
976 The filter accepts the following options:
980 Set input gain. Default is 1.
983 Set output gain. Default is 1.
986 Don't let signals above this level pass the limiter. Default is 1.
989 The limiter will reach its attenuation level in this amount of time in
990 milliseconds. Default is 5 milliseconds.
993 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
994 Default is 50 milliseconds.
997 When gain reduction is always needed ASC takes care of releasing to an
998 average reduction level rather than reaching a reduction of 0 in the release
1002 Select how much the release time is affected by ASC, 0 means nearly no changes
1003 in release time while 1 produces higher release times.
1006 Auto level output signal. Default is enabled.
1007 This normalizes audio back to 0dB if enabled.
1010 Depending on picked setting it is recommended to upsample input 2x or 4x times
1011 with @ref{aresample} before applying this filter.
1015 Apply a two-pole all-pass filter with central frequency (in Hz)
1016 @var{frequency}, and filter-width @var{width}.
1017 An all-pass filter changes the audio's frequency to phase relationship
1018 without changing its frequency to amplitude relationship.
1020 The filter accepts the following options:
1024 Set frequency in Hz.
1027 Set method to specify band-width of filter.
1040 Specify the band-width of a filter in width_type units.
1047 The filter accepts the following options:
1051 Set the number of loops.
1054 Set maximal number of samples.
1057 Set first sample of loop.
1063 Merge two or more audio streams into a single multi-channel stream.
1065 The filter accepts the following options:
1070 Set the number of inputs. Default is 2.
1074 If the channel layouts of the inputs are disjoint, and therefore compatible,
1075 the channel layout of the output will be set accordingly and the channels
1076 will be reordered as necessary. If the channel layouts of the inputs are not
1077 disjoint, the output will have all the channels of the first input then all
1078 the channels of the second input, in that order, and the channel layout of
1079 the output will be the default value corresponding to the total number of
1082 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1083 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1084 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1085 first input, b1 is the first channel of the second input).
1087 On the other hand, if both input are in stereo, the output channels will be
1088 in the default order: a1, a2, b1, b2, and the channel layout will be
1089 arbitrarily set to 4.0, which may or may not be the expected value.
1091 All inputs must have the same sample rate, and format.
1093 If inputs do not have the same duration, the output will stop with the
1096 @subsection Examples
1100 Merge two mono files into a stereo stream:
1102 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1106 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1108 ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
1114 Mixes multiple audio inputs into a single output.
1116 Note that this filter only supports float samples (the @var{amerge}
1117 and @var{pan} audio filters support many formats). If the @var{amix}
1118 input has integer samples then @ref{aresample} will be automatically
1119 inserted to perform the conversion to float samples.
1123 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1125 will mix 3 input audio streams to a single output with the same duration as the
1126 first input and a dropout transition time of 3 seconds.
1128 It accepts the following parameters:
1132 The number of inputs. If unspecified, it defaults to 2.
1135 How to determine the end-of-stream.
1139 The duration of the longest input. (default)
1142 The duration of the shortest input.
1145 The duration of the first input.
1149 @item dropout_transition
1150 The transition time, in seconds, for volume renormalization when an input
1151 stream ends. The default value is 2 seconds.
1155 @section anequalizer
1157 High-order parametric multiband equalizer for each channel.
1159 It accepts the following parameters:
1163 This option string is in format:
1164 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1165 Each equalizer band is separated by '|'.
1169 Set channel number to which equalization will be applied.
1170 If input doesn't have that channel the entry is ignored.
1173 Set central frequency for band.
1174 If input doesn't have that frequency the entry is ignored.
1177 Set band width in hertz.
1180 Set band gain in dB.
1183 Set filter type for band, optional, can be:
1187 Butterworth, this is default.
1198 With this option activated frequency response of anequalizer is displayed
1202 Set video stream size. Only useful if curves option is activated.
1205 Set max gain that will be displayed. Only useful if curves option is activated.
1206 Setting this to a reasonable value makes it possible to display gain which is derived from
1207 neighbour bands which are too close to each other and thus produce higher gain
1208 when both are activated.
1211 Set frequency scale used to draw frequency response in video output.
1212 Can be linear or logarithmic. Default is logarithmic.
1215 Set color for each channel curve which is going to be displayed in video stream.
1216 This is list of color names separated by space or by '|'.
1217 Unrecognised or missing colors will be replaced by white color.
1220 @subsection Examples
1224 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1225 for first 2 channels using Chebyshev type 1 filter:
1227 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1231 @subsection Commands
1233 This filter supports the following commands:
1236 Alter existing filter parameters.
1237 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1239 @var{fN} is existing filter number, starting from 0, if no such filter is available
1241 @var{freq} set new frequency parameter.
1242 @var{width} set new width parameter in herz.
1243 @var{gain} set new gain parameter in dB.
1245 Full filter invocation with asendcmd may look like this:
1246 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1251 Pass the audio source unchanged to the output.
1255 Pad the end of an audio stream with silence.
1257 This can be used together with @command{ffmpeg} @option{-shortest} to
1258 extend audio streams to the same length as the video stream.
1260 A description of the accepted options follows.
1264 Set silence packet size. Default value is 4096.
1267 Set the number of samples of silence to add to the end. After the
1268 value is reached, the stream is terminated. This option is mutually
1269 exclusive with @option{whole_len}.
1272 Set the minimum total number of samples in the output audio stream. If
1273 the value is longer than the input audio length, silence is added to
1274 the end, until the value is reached. This option is mutually exclusive
1275 with @option{pad_len}.
1278 If neither the @option{pad_len} nor the @option{whole_len} option is
1279 set, the filter will add silence to the end of the input stream
1282 @subsection Examples
1286 Add 1024 samples of silence to the end of the input:
1292 Make sure the audio output will contain at least 10000 samples, pad
1293 the input with silence if required:
1295 apad=whole_len=10000
1299 Use @command{ffmpeg} to pad the audio input with silence, so that the
1300 video stream will always result the shortest and will be converted
1301 until the end in the output file when using the @option{shortest}
1304 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1309 Add a phasing effect to the input audio.
1311 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1312 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1314 A description of the accepted parameters follows.
1318 Set input gain. Default is 0.4.
1321 Set output gain. Default is 0.74
1324 Set delay in milliseconds. Default is 3.0.
1327 Set decay. Default is 0.4.
1330 Set modulation speed in Hz. Default is 0.5.
1333 Set modulation type. Default is triangular.
1335 It accepts the following values:
1344 Audio pulsator is something between an autopanner and a tremolo.
1345 But it can produce funny stereo effects as well. Pulsator changes the volume
1346 of the left and right channel based on a LFO (low frequency oscillator) with
1347 different waveforms and shifted phases.
1348 This filter have the ability to define an offset between left and right
1349 channel. An offset of 0 means that both LFO shapes match each other.
1350 The left and right channel are altered equally - a conventional tremolo.
1351 An offset of 50% means that the shape of the right channel is exactly shifted
1352 in phase (or moved backwards about half of the frequency) - pulsator acts as
1353 an autopanner. At 1 both curves match again. Every setting in between moves the
1354 phase shift gapless between all stages and produces some "bypassing" sounds with
1355 sine and triangle waveforms. The more you set the offset near 1 (starting from
1356 the 0.5) the faster the signal passes from the left to the right speaker.
1358 The filter accepts the following options:
1362 Set input gain. By default it is 1. Range is [0.015625 - 64].
1365 Set output gain. By default it is 1. Range is [0.015625 - 64].
1368 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1369 sawup or sawdown. Default is sine.
1372 Set modulation. Define how much of original signal is affected by the LFO.
1375 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1378 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1381 Set pulse width. Default is 1. Allowed range is [0 - 2].
1384 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1387 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1391 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1395 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1396 if timing is set to hz.
1402 Resample the input audio to the specified parameters, using the
1403 libswresample library. If none are specified then the filter will
1404 automatically convert between its input and output.
1406 This filter is also able to stretch/squeeze the audio data to make it match
1407 the timestamps or to inject silence / cut out audio to make it match the
1408 timestamps, do a combination of both or do neither.
1410 The filter accepts the syntax
1411 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1412 expresses a sample rate and @var{resampler_options} is a list of
1413 @var{key}=@var{value} pairs, separated by ":". See the
1414 ffmpeg-resampler manual for the complete list of supported options.
1416 @subsection Examples
1420 Resample the input audio to 44100Hz:
1426 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1427 samples per second compensation:
1429 aresample=async=1000
1435 Reverse an audio clip.
1437 Warning: This filter requires memory to buffer the entire clip, so trimming
1440 @subsection Examples
1444 Take the first 5 seconds of a clip, and reverse it.
1446 atrim=end=5,areverse
1450 @section asetnsamples
1452 Set the number of samples per each output audio frame.
1454 The last output packet may contain a different number of samples, as
1455 the filter will flush all the remaining samples when the input audio
1458 The filter accepts the following options:
1462 @item nb_out_samples, n
1463 Set the number of frames per each output audio frame. The number is
1464 intended as the number of samples @emph{per each channel}.
1465 Default value is 1024.
1468 If set to 1, the filter will pad the last audio frame with zeroes, so
1469 that the last frame will contain the same number of samples as the
1470 previous ones. Default value is 1.
1473 For example, to set the number of per-frame samples to 1234 and
1474 disable padding for the last frame, use:
1476 asetnsamples=n=1234:p=0
1481 Set the sample rate without altering the PCM data.
1482 This will result in a change of speed and pitch.
1484 The filter accepts the following options:
1487 @item sample_rate, r
1488 Set the output sample rate. Default is 44100 Hz.
1493 Show a line containing various information for each input audio frame.
1494 The input audio is not modified.
1496 The shown line contains a sequence of key/value pairs of the form
1497 @var{key}:@var{value}.
1499 The following values are shown in the output:
1503 The (sequential) number of the input frame, starting from 0.
1506 The presentation timestamp of the input frame, in time base units; the time base
1507 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1510 The presentation timestamp of the input frame in seconds.
1513 position of the frame in the input stream, -1 if this information in
1514 unavailable and/or meaningless (for example in case of synthetic audio)
1523 The sample rate for the audio frame.
1526 The number of samples (per channel) in the frame.
1529 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1530 audio, the data is treated as if all the planes were concatenated.
1532 @item plane_checksums
1533 A list of Adler-32 checksums for each data plane.
1539 Display time domain statistical information about the audio channels.
1540 Statistics are calculated and displayed for each audio channel and,
1541 where applicable, an overall figure is also given.
1543 It accepts the following option:
1546 Short window length in seconds, used for peak and trough RMS measurement.
1547 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1551 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1552 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1555 Available keys for each channel are:
1586 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1587 this @code{lavfi.astats.Overall.Peak_count}.
1589 For description what each key means read below.
1592 Set number of frame after which stats are going to be recalculated.
1593 Default is disabled.
1596 A description of each shown parameter follows:
1600 Mean amplitude displacement from zero.
1603 Minimal sample level.
1606 Maximal sample level.
1608 @item Min difference
1609 Minimal difference between two consecutive samples.
1611 @item Max difference
1612 Maximal difference between two consecutive samples.
1614 @item Mean difference
1615 Mean difference between two consecutive samples.
1616 The average of each difference between two consecutive samples.
1620 Standard peak and RMS level measured in dBFS.
1624 Peak and trough values for RMS level measured over a short window.
1627 Standard ratio of peak to RMS level (note: not in dB).
1630 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1631 (i.e. either @var{Min level} or @var{Max level}).
1634 Number of occasions (not the number of samples) that the signal attained either
1635 @var{Min level} or @var{Max level}.
1638 Overall bit depth of audio. Number of bits used for each sample.
1643 Synchronize audio data with timestamps by squeezing/stretching it and/or
1644 dropping samples/adding silence when needed.
1646 This filter is not built by default, please use @ref{aresample} to do squeezing/stretching.
1648 It accepts the following parameters:
1652 Enable stretching/squeezing the data to make it match the timestamps. Disabled
1653 by default. When disabled, time gaps are covered with silence.
1656 The minimum difference between timestamps and audio data (in seconds) to trigger
1657 adding/dropping samples. The default value is 0.1. If you get an imperfect
1658 sync with this filter, try setting this parameter to 0.
1661 The maximum compensation in samples per second. Only relevant with compensate=1.
1662 The default value is 500.
1665 Assume that the first PTS should be this value. The time base is 1 / sample
1666 rate. This allows for padding/trimming at the start of the stream. By default,
1667 no assumption is made about the first frame's expected PTS, so no padding or
1668 trimming is done. For example, this could be set to 0 to pad the beginning with
1669 silence if an audio stream starts after the video stream or to trim any samples
1670 with a negative PTS due to encoder delay.
1678 The filter accepts exactly one parameter, the audio tempo. If not
1679 specified then the filter will assume nominal 1.0 tempo. Tempo must
1680 be in the [0.5, 2.0] range.
1682 @subsection Examples
1686 Slow down audio to 80% tempo:
1692 To speed up audio to 125% tempo:
1700 Trim the input so that the output contains one continuous subpart of the input.
1702 It accepts the following parameters:
1705 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1706 sample with the timestamp @var{start} will be the first sample in the output.
1709 Specify time of the first audio sample that will be dropped, i.e. the
1710 audio sample immediately preceding the one with the timestamp @var{end} will be
1711 the last sample in the output.
1714 Same as @var{start}, except this option sets the start timestamp in samples
1718 Same as @var{end}, except this option sets the end timestamp in samples instead
1722 The maximum duration of the output in seconds.
1725 The number of the first sample that should be output.
1728 The number of the first sample that should be dropped.
1731 @option{start}, @option{end}, and @option{duration} are expressed as time
1732 duration specifications; see
1733 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1735 Note that the first two sets of the start/end options and the @option{duration}
1736 option look at the frame timestamp, while the _sample options simply count the
1737 samples that pass through the filter. So start/end_pts and start/end_sample will
1738 give different results when the timestamps are wrong, inexact or do not start at
1739 zero. Also note that this filter does not modify the timestamps. If you wish
1740 to have the output timestamps start at zero, insert the asetpts filter after the
1743 If multiple start or end options are set, this filter tries to be greedy and
1744 keep all samples that match at least one of the specified constraints. To keep
1745 only the part that matches all the constraints at once, chain multiple atrim
1748 The defaults are such that all the input is kept. So it is possible to set e.g.
1749 just the end values to keep everything before the specified time.
1754 Drop everything except the second minute of input:
1756 ffmpeg -i INPUT -af atrim=60:120
1760 Keep only the first 1000 samples:
1762 ffmpeg -i INPUT -af atrim=end_sample=1000
1769 Apply a two-pole Butterworth band-pass filter with central
1770 frequency @var{frequency}, and (3dB-point) band-width width.
1771 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1772 instead of the default: constant 0dB peak gain.
1773 The filter roll off at 6dB per octave (20dB per decade).
1775 The filter accepts the following options:
1779 Set the filter's central frequency. Default is @code{3000}.
1782 Constant skirt gain if set to 1. Defaults to 0.
1785 Set method to specify band-width of filter.
1798 Specify the band-width of a filter in width_type units.
1803 Apply a two-pole Butterworth band-reject filter with central
1804 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1805 The filter roll off at 6dB per octave (20dB per decade).
1807 The filter accepts the following options:
1811 Set the filter's central frequency. Default is @code{3000}.
1814 Set method to specify band-width of filter.
1827 Specify the band-width of a filter in width_type units.
1832 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1833 shelving filter with a response similar to that of a standard
1834 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1836 The filter accepts the following options:
1840 Give the gain at 0 Hz. Its useful range is about -20
1841 (for a large cut) to +20 (for a large boost).
1842 Beware of clipping when using a positive gain.
1845 Set the filter's central frequency and so can be used
1846 to extend or reduce the frequency range to be boosted or cut.
1847 The default value is @code{100} Hz.
1850 Set method to specify band-width of filter.
1863 Determine how steep is the filter's shelf transition.
1868 Apply a biquad IIR filter with the given coefficients.
1869 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1870 are the numerator and denominator coefficients respectively.
1873 Bauer stereo to binaural transformation, which improves headphone listening of
1874 stereo audio records.
1876 It accepts the following parameters:
1880 Pre-defined crossfeed level.
1884 Default level (fcut=700, feed=50).
1887 Chu Moy circuit (fcut=700, feed=60).
1890 Jan Meier circuit (fcut=650, feed=95).
1895 Cut frequency (in Hz).
1904 Remap input channels to new locations.
1906 It accepts the following parameters:
1908 @item channel_layout
1909 The channel layout of the output stream.
1912 Map channels from input to output. The argument is a '|'-separated list of
1913 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1914 @var{in_channel} form. @var{in_channel} can be either the name of the input
1915 channel (e.g. FL for front left) or its index in the input channel layout.
1916 @var{out_channel} is the name of the output channel or its index in the output
1917 channel layout. If @var{out_channel} is not given then it is implicitly an
1918 index, starting with zero and increasing by one for each mapping.
1921 If no mapping is present, the filter will implicitly map input channels to
1922 output channels, preserving indices.
1924 For example, assuming a 5.1+downmix input MOV file,
1926 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1928 will create an output WAV file tagged as stereo from the downmix channels of
1931 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1933 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
1936 @section channelsplit
1938 Split each channel from an input audio stream into a separate output stream.
1940 It accepts the following parameters:
1942 @item channel_layout
1943 The channel layout of the input stream. The default is "stereo".
1946 For example, assuming a stereo input MP3 file,
1948 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1950 will create an output Matroska file with two audio streams, one containing only
1951 the left channel and the other the right channel.
1953 Split a 5.1 WAV file into per-channel files:
1955 ffmpeg -i in.wav -filter_complex
1956 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1957 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1958 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
1963 Add a chorus effect to the audio.
1965 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
1967 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
1968 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
1969 The modulation depth defines the range the modulated delay is played before or after
1970 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
1971 sound tuned around the original one, like in a chorus where some vocals are slightly
1974 It accepts the following parameters:
1977 Set input gain. Default is 0.4.
1980 Set output gain. Default is 0.4.
1983 Set delays. A typical delay is around 40ms to 60ms.
1995 @subsection Examples
2001 chorus=0.7:0.9:55:0.4:0.25:2
2007 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2011 Fuller sounding chorus with three delays:
2013 chorus=0.5:0.9:50|60|40:0.4|0.32|0.3:0.25|0.4|0.3:2|2.3|1.3
2018 Compress or expand the audio's dynamic range.
2020 It accepts the following parameters:
2026 A list of times in seconds for each channel over which the instantaneous level
2027 of the input signal is averaged to determine its volume. @var{attacks} refers to
2028 increase of volume and @var{decays} refers to decrease of volume. For most
2029 situations, the attack time (response to the audio getting louder) should be
2030 shorter than the decay time, because the human ear is more sensitive to sudden
2031 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2032 a typical value for decay is 0.8 seconds.
2033 If specified number of attacks & decays is lower than number of channels, the last
2034 set attack/decay will be used for all remaining channels.
2037 A list of points for the transfer function, specified in dB relative to the
2038 maximum possible signal amplitude. Each key points list must be defined using
2039 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2040 @code{x0/y0 x1/y1 x2/y2 ....}
2042 The input values must be in strictly increasing order but the transfer function
2043 does not have to be monotonically rising. The point @code{0/0} is assumed but
2044 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2045 function are @code{-70/-70|-60/-20}.
2048 Set the curve radius in dB for all joints. It defaults to 0.01.
2051 Set the additional gain in dB to be applied at all points on the transfer
2052 function. This allows for easy adjustment of the overall gain.
2056 Set an initial volume, in dB, to be assumed for each channel when filtering
2057 starts. This permits the user to supply a nominal level initially, so that, for
2058 example, a very large gain is not applied to initial signal levels before the
2059 companding has begun to operate. A typical value for audio which is initially
2060 quiet is -90 dB. It defaults to 0.
2063 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2064 delayed before being fed to the volume adjuster. Specifying a delay
2065 approximately equal to the attack/decay times allows the filter to effectively
2066 operate in predictive rather than reactive mode. It defaults to 0.
2070 @subsection Examples
2074 Make music with both quiet and loud passages suitable for listening to in a
2077 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2080 Another example for audio with whisper and explosion parts:
2082 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2086 A noise gate for when the noise is at a lower level than the signal:
2088 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2092 Here is another noise gate, this time for when the noise is at a higher level
2093 than the signal (making it, in some ways, similar to squelch):
2095 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2099 2:1 compression starting at -6dB:
2101 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2105 2:1 compression starting at -9dB:
2107 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2111 2:1 compression starting at -12dB:
2113 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2117 2:1 compression starting at -18dB:
2119 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2123 3:1 compression starting at -15dB:
2125 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2131 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2137 compand=attacks=0:points=-80/-169|-54/-80|-49.5/-64.6|-41.1/-41.1|-25.8/-15|-10.8/-4.5|0/0|20/8.3
2141 Hard limiter at -6dB:
2143 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2147 Hard limiter at -12dB:
2149 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2153 Hard noise gate at -35 dB:
2155 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2161 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2165 @section compensationdelay
2167 Compensation Delay Line is a metric based delay to compensate differing
2168 positions of microphones or speakers.
2170 For example, you have recorded guitar with two microphones placed in
2171 different location. Because the front of sound wave has fixed speed in
2172 normal conditions, the phasing of microphones can vary and depends on
2173 their location and interposition. The best sound mix can be achieved when
2174 these microphones are in phase (synchronized). Note that distance of
2175 ~30 cm between microphones makes one microphone to capture signal in
2176 antiphase to another microphone. That makes the final mix sounding moody.
2177 This filter helps to solve phasing problems by adding different delays
2178 to each microphone track and make them synchronized.
2180 The best result can be reached when you take one track as base and
2181 synchronize other tracks one by one with it.
2182 Remember that synchronization/delay tolerance depends on sample rate, too.
2183 Higher sample rates will give more tolerance.
2185 It accepts the following parameters:
2189 Set millimeters distance. This is compensation distance for fine tuning.
2193 Set cm distance. This is compensation distance for tightening distance setup.
2197 Set meters distance. This is compensation distance for hard distance setup.
2201 Set dry amount. Amount of unprocessed (dry) signal.
2205 Set wet amount. Amount of processed (wet) signal.
2209 Set temperature degree in Celsius. This is the temperature of the environment.
2213 @section crystalizer
2214 Simple algorithm to expand audio dynamic range.
2216 The filter accepts the following options:
2220 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2221 (unchanged sound) to 10.0 (maximum effect).
2224 Enable clipping. By default is enabled.
2228 Apply a DC shift to the audio.
2230 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2231 in the recording chain) from the audio. The effect of a DC offset is reduced
2232 headroom and hence volume. The @ref{astats} filter can be used to determine if
2233 a signal has a DC offset.
2237 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2241 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2242 used to prevent clipping.
2246 Dynamic Audio Normalizer.
2248 This filter applies a certain amount of gain to the input audio in order
2249 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2250 contrast to more "simple" normalization algorithms, the Dynamic Audio
2251 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2252 This allows for applying extra gain to the "quiet" sections of the audio
2253 while avoiding distortions or clipping the "loud" sections. In other words:
2254 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2255 sections, in the sense that the volume of each section is brought to the
2256 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2257 this goal *without* applying "dynamic range compressing". It will retain 100%
2258 of the dynamic range *within* each section of the audio file.
2262 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2263 Default is 500 milliseconds.
2264 The Dynamic Audio Normalizer processes the input audio in small chunks,
2265 referred to as frames. This is required, because a peak magnitude has no
2266 meaning for just a single sample value. Instead, we need to determine the
2267 peak magnitude for a contiguous sequence of sample values. While a "standard"
2268 normalizer would simply use the peak magnitude of the complete file, the
2269 Dynamic Audio Normalizer determines the peak magnitude individually for each
2270 frame. The length of a frame is specified in milliseconds. By default, the
2271 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2272 been found to give good results with most files.
2273 Note that the exact frame length, in number of samples, will be determined
2274 automatically, based on the sampling rate of the individual input audio file.
2277 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2278 number. Default is 31.
2279 Probably the most important parameter of the Dynamic Audio Normalizer is the
2280 @code{window size} of the Gaussian smoothing filter. The filter's window size
2281 is specified in frames, centered around the current frame. For the sake of
2282 simplicity, this must be an odd number. Consequently, the default value of 31
2283 takes into account the current frame, as well as the 15 preceding frames and
2284 the 15 subsequent frames. Using a larger window results in a stronger
2285 smoothing effect and thus in less gain variation, i.e. slower gain
2286 adaptation. Conversely, using a smaller window results in a weaker smoothing
2287 effect and thus in more gain variation, i.e. faster gain adaptation.
2288 In other words, the more you increase this value, the more the Dynamic Audio
2289 Normalizer will behave like a "traditional" normalization filter. On the
2290 contrary, the more you decrease this value, the more the Dynamic Audio
2291 Normalizer will behave like a dynamic range compressor.
2294 Set the target peak value. This specifies the highest permissible magnitude
2295 level for the normalized audio input. This filter will try to approach the
2296 target peak magnitude as closely as possible, but at the same time it also
2297 makes sure that the normalized signal will never exceed the peak magnitude.
2298 A frame's maximum local gain factor is imposed directly by the target peak
2299 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2300 It is not recommended to go above this value.
2303 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2304 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2305 factor for each input frame, i.e. the maximum gain factor that does not
2306 result in clipping or distortion. The maximum gain factor is determined by
2307 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2308 additionally bounds the frame's maximum gain factor by a predetermined
2309 (global) maximum gain factor. This is done in order to avoid excessive gain
2310 factors in "silent" or almost silent frames. By default, the maximum gain
2311 factor is 10.0, For most inputs the default value should be sufficient and
2312 it usually is not recommended to increase this value. Though, for input
2313 with an extremely low overall volume level, it may be necessary to allow even
2314 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2315 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2316 Instead, a "sigmoid" threshold function will be applied. This way, the
2317 gain factors will smoothly approach the threshold value, but never exceed that
2321 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2322 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2323 This means that the maximum local gain factor for each frame is defined
2324 (only) by the frame's highest magnitude sample. This way, the samples can
2325 be amplified as much as possible without exceeding the maximum signal
2326 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2327 Normalizer can also take into account the frame's root mean square,
2328 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2329 determine the power of a time-varying signal. It is therefore considered
2330 that the RMS is a better approximation of the "perceived loudness" than
2331 just looking at the signal's peak magnitude. Consequently, by adjusting all
2332 frames to a constant RMS value, a uniform "perceived loudness" can be
2333 established. If a target RMS value has been specified, a frame's local gain
2334 factor is defined as the factor that would result in exactly that RMS value.
2335 Note, however, that the maximum local gain factor is still restricted by the
2336 frame's highest magnitude sample, in order to prevent clipping.
2339 Enable channels coupling. By default is enabled.
2340 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2341 amount. This means the same gain factor will be applied to all channels, i.e.
2342 the maximum possible gain factor is determined by the "loudest" channel.
2343 However, in some recordings, it may happen that the volume of the different
2344 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2345 In this case, this option can be used to disable the channel coupling. This way,
2346 the gain factor will be determined independently for each channel, depending
2347 only on the individual channel's highest magnitude sample. This allows for
2348 harmonizing the volume of the different channels.
2351 Enable DC bias correction. By default is disabled.
2352 An audio signal (in the time domain) is a sequence of sample values.
2353 In the Dynamic Audio Normalizer these sample values are represented in the
2354 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2355 audio signal, or "waveform", should be centered around the zero point.
2356 That means if we calculate the mean value of all samples in a file, or in a
2357 single frame, then the result should be 0.0 or at least very close to that
2358 value. If, however, there is a significant deviation of the mean value from
2359 0.0, in either positive or negative direction, this is referred to as a
2360 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2361 Audio Normalizer provides optional DC bias correction.
2362 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2363 the mean value, or "DC correction" offset, of each input frame and subtract
2364 that value from all of the frame's sample values which ensures those samples
2365 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2366 boundaries, the DC correction offset values will be interpolated smoothly
2367 between neighbouring frames.
2370 Enable alternative boundary mode. By default is disabled.
2371 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2372 around each frame. This includes the preceding frames as well as the
2373 subsequent frames. However, for the "boundary" frames, located at the very
2374 beginning and at the very end of the audio file, not all neighbouring
2375 frames are available. In particular, for the first few frames in the audio
2376 file, the preceding frames are not known. And, similarly, for the last few
2377 frames in the audio file, the subsequent frames are not known. Thus, the
2378 question arises which gain factors should be assumed for the missing frames
2379 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2380 to deal with this situation. The default boundary mode assumes a gain factor
2381 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2382 "fade out" at the beginning and at the end of the input, respectively.
2385 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2386 By default, the Dynamic Audio Normalizer does not apply "traditional"
2387 compression. This means that signal peaks will not be pruned and thus the
2388 full dynamic range will be retained within each local neighbourhood. However,
2389 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2390 normalization algorithm with a more "traditional" compression.
2391 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2392 (thresholding) function. If (and only if) the compression feature is enabled,
2393 all input frames will be processed by a soft knee thresholding function prior
2394 to the actual normalization process. Put simply, the thresholding function is
2395 going to prune all samples whose magnitude exceeds a certain threshold value.
2396 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2397 value. Instead, the threshold value will be adjusted for each individual
2399 In general, smaller parameters result in stronger compression, and vice versa.
2400 Values below 3.0 are not recommended, because audible distortion may appear.
2405 Make audio easier to listen to on headphones.
2407 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2408 so that when listened to on headphones the stereo image is moved from
2409 inside your head (standard for headphones) to outside and in front of
2410 the listener (standard for speakers).
2416 Apply a two-pole peaking equalisation (EQ) filter. With this
2417 filter, the signal-level at and around a selected frequency can
2418 be increased or decreased, whilst (unlike bandpass and bandreject
2419 filters) that at all other frequencies is unchanged.
2421 In order to produce complex equalisation curves, this filter can
2422 be given several times, each with a different central frequency.
2424 The filter accepts the following options:
2428 Set the filter's central frequency in Hz.
2431 Set method to specify band-width of filter.
2444 Specify the band-width of a filter in width_type units.
2447 Set the required gain or attenuation in dB.
2448 Beware of clipping when using a positive gain.
2451 @subsection Examples
2454 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2456 equalizer=f=1000:width_type=h:width=200:g=-10
2460 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2462 equalizer=f=1000:width_type=q:width=1:g=2,equalizer=f=100:width_type=q:width=2:g=-5
2466 @section extrastereo
2468 Linearly increases the difference between left and right channels which
2469 adds some sort of "live" effect to playback.
2471 The filter accepts the following options:
2475 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2476 (average of both channels), with 1.0 sound will be unchanged, with
2477 -1.0 left and right channels will be swapped.
2480 Enable clipping. By default is enabled.
2483 @section firequalizer
2484 Apply FIR Equalization using arbitrary frequency response.
2486 The filter accepts the following option:
2490 Set gain curve equation (in dB). The expression can contain variables:
2493 the evaluated frequency
2497 channel number, set to 0 when multichannels evaluation is disabled
2499 channel id, see libavutil/channel_layout.h, set to the first channel id when
2500 multichannels evaluation is disabled
2504 channel_layout, see libavutil/channel_layout.h
2509 @item gain_interpolate(f)
2510 interpolate gain on frequency f based on gain_entry
2511 @item cubic_interpolate(f)
2512 same as gain_interpolate, but smoother
2514 This option is also available as command. Default is @code{gain_interpolate(f)}.
2517 Set gain entry for gain_interpolate function. The expression can
2521 store gain entry at frequency f with value g
2523 This option is also available as command.
2526 Set filter delay in seconds. Higher value means more accurate.
2527 Default is @code{0.01}.
2530 Set filter accuracy in Hz. Lower value means more accurate.
2531 Default is @code{5}.
2534 Set window function. Acceptable values are:
2537 rectangular window, useful when gain curve is already smooth
2539 hann window (default)
2545 3-terms continuous 1st derivative nuttall window
2547 minimum 3-terms discontinuous nuttall window
2549 4-terms continuous 1st derivative nuttall window
2551 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2553 blackman-harris window
2559 If enabled, use fixed number of audio samples. This improves speed when
2560 filtering with large delay. Default is disabled.
2563 Enable multichannels evaluation on gain. Default is disabled.
2566 Enable zero phase mode by subtracting timestamp to compensate delay.
2567 Default is disabled.
2570 Set scale used by gain. Acceptable values are:
2573 linear frequency, linear gain
2575 linear frequency, logarithmic (in dB) gain (default)
2577 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2579 logarithmic frequency, logarithmic gain
2583 Set file for dumping, suitable for gnuplot.
2586 Set scale for dumpfile. Acceptable values are same with scale option.
2590 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2591 Default is disabled.
2594 @subsection Examples
2599 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2602 lowpass at 1000 Hz with gain_entry:
2604 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2607 custom equalization:
2609 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2612 higher delay with zero phase to compensate delay:
2614 firequalizer=delay=0.1:fixed=on:zero_phase=on
2617 lowpass on left channel, highpass on right channel:
2619 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2620 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2625 Apply a flanging effect to the audio.
2627 The filter accepts the following options:
2631 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2634 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2637 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2641 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2642 Default value is 71.
2645 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2648 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2649 Default value is @var{sinusoidal}.
2652 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2653 Default value is 25.
2656 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2657 Default is @var{linear}.
2662 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2663 embedded HDCD codes is expanded into a 20-bit PCM stream.
2665 The filter supports the Peak Extend and Low-level Gain Adjustment features
2666 of HDCD, and detects the Transient Filter flag.
2669 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2672 When using the filter with wav, note the default encoding for wav is 16-bit,
2673 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2674 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2676 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2677 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2680 The filter accepts the following options:
2683 @item disable_autoconvert
2684 Disable any automatic format conversion or resampling in the filter graph.
2686 @item process_stereo
2687 Process the stereo channels together. If target_gain does not match between
2688 channels, consider it invalid and use the last valid target_gain.
2691 Set the code detect timer period in ms.
2694 Always extend peaks above -3dBFS even if PE isn't signaled.
2697 Replace audio with a solid tone and adjust the amplitude to signal some
2698 specific aspect of the decoding process. The output file can be loaded in
2699 an audio editor alongside the original to aid analysis.
2701 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2708 Gain adjustment level at each sample
2710 Samples where peak extend occurs
2712 Samples where the code detect timer is active
2714 Samples where the target gain does not match between channels
2720 Apply a high-pass filter with 3dB point frequency.
2721 The filter can be either single-pole, or double-pole (the default).
2722 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2724 The filter accepts the following options:
2728 Set frequency in Hz. Default is 3000.
2731 Set number of poles. Default is 2.
2734 Set method to specify band-width of filter.
2747 Specify the band-width of a filter in width_type units.
2748 Applies only to double-pole filter.
2749 The default is 0.707q and gives a Butterworth response.
2754 Join multiple input streams into one multi-channel stream.
2756 It accepts the following parameters:
2760 The number of input streams. It defaults to 2.
2762 @item channel_layout
2763 The desired output channel layout. It defaults to stereo.
2766 Map channels from inputs to output. The argument is a '|'-separated list of
2767 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2768 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2769 can be either the name of the input channel (e.g. FL for front left) or its
2770 index in the specified input stream. @var{out_channel} is the name of the output
2774 The filter will attempt to guess the mappings when they are not specified
2775 explicitly. It does so by first trying to find an unused matching input channel
2776 and if that fails it picks the first unused input channel.
2778 Join 3 inputs (with properly set channel layouts):
2780 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2783 Build a 5.1 output from 6 single-channel streams:
2785 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2786 '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'
2792 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2794 To enable compilation of this filter you need to configure FFmpeg with
2795 @code{--enable-ladspa}.
2799 Specifies the name of LADSPA plugin library to load. If the environment
2800 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2801 each one of the directories specified by the colon separated list in
2802 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2803 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2804 @file{/usr/lib/ladspa/}.
2807 Specifies the plugin within the library. Some libraries contain only
2808 one plugin, but others contain many of them. If this is not set filter
2809 will list all available plugins within the specified library.
2812 Set the '|' separated list of controls which are zero or more floating point
2813 values that determine the behavior of the loaded plugin (for example delay,
2815 Controls need to be defined using the following syntax:
2816 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2817 @var{valuei} is the value set on the @var{i}-th control.
2818 Alternatively they can be also defined using the following syntax:
2819 @var{value0}|@var{value1}|@var{value2}|..., where
2820 @var{valuei} is the value set on the @var{i}-th control.
2821 If @option{controls} is set to @code{help}, all available controls and
2822 their valid ranges are printed.
2824 @item sample_rate, s
2825 Specify the sample rate, default to 44100. Only used if plugin have
2829 Set the number of samples per channel per each output frame, default
2830 is 1024. Only used if plugin have zero inputs.
2833 Set the minimum duration of the sourced audio. See
2834 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2835 for the accepted syntax.
2836 Note that the resulting duration may be greater than the specified duration,
2837 as the generated audio is always cut at the end of a complete frame.
2838 If not specified, or the expressed duration is negative, the audio is
2839 supposed to be generated forever.
2840 Only used if plugin have zero inputs.
2844 @subsection Examples
2848 List all available plugins within amp (LADSPA example plugin) library:
2854 List all available controls and their valid ranges for @code{vcf_notch}
2855 plugin from @code{VCF} library:
2857 ladspa=f=vcf:p=vcf_notch:c=help
2861 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2864 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2868 Add reverberation to the audio using TAP-plugins
2869 (Tom's Audio Processing plugins):
2871 ladspa=file=tap_reverb:tap_reverb
2875 Generate white noise, with 0.2 amplitude:
2877 ladspa=file=cmt:noise_source_white:c=c0=.2
2881 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2882 @code{C* Audio Plugin Suite} (CAPS) library:
2884 ladspa=file=caps:Click:c=c1=20'
2888 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2890 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2894 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2895 @code{SWH Plugins} collection:
2897 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2901 Attenuate low frequencies using Multiband EQ from Steve Harris
2902 @code{SWH Plugins} collection:
2904 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2908 @subsection Commands
2910 This filter supports the following commands:
2913 Modify the @var{N}-th control value.
2915 If the specified value is not valid, it is ignored and prior one is kept.
2920 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2921 Support for both single pass (livestreams, files) and double pass (files) modes.
2922 This algorithm can target IL, LRA, and maximum true peak.
2924 The filter accepts the following options:
2928 Set integrated loudness target.
2929 Range is -70.0 - -5.0. Default value is -24.0.
2932 Set loudness range target.
2933 Range is 1.0 - 20.0. Default value is 7.0.
2936 Set maximum true peak.
2937 Range is -9.0 - +0.0. Default value is -2.0.
2939 @item measured_I, measured_i
2940 Measured IL of input file.
2941 Range is -99.0 - +0.0.
2943 @item measured_LRA, measured_lra
2944 Measured LRA of input file.
2945 Range is 0.0 - 99.0.
2947 @item measured_TP, measured_tp
2948 Measured true peak of input file.
2949 Range is -99.0 - +99.0.
2951 @item measured_thresh
2952 Measured threshold of input file.
2953 Range is -99.0 - +0.0.
2956 Set offset gain. Gain is applied before the true-peak limiter.
2957 Range is -99.0 - +99.0. Default is +0.0.
2960 Normalize linearly if possible.
2961 measured_I, measured_LRA, measured_TP, and measured_thresh must also
2962 to be specified in order to use this mode.
2963 Options are true or false. Default is true.
2966 Treat mono input files as "dual-mono". If a mono file is intended for playback
2967 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
2968 If set to @code{true}, this option will compensate for this effect.
2969 Multi-channel input files are not affected by this option.
2970 Options are true or false. Default is false.
2973 Set print format for stats. Options are summary, json, or none.
2974 Default value is none.
2979 Apply a low-pass filter with 3dB point frequency.
2980 The filter can be either single-pole or double-pole (the default).
2981 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2983 The filter accepts the following options:
2987 Set frequency in Hz. Default is 500.
2990 Set number of poles. Default is 2.
2993 Set method to specify band-width of filter.
3006 Specify the band-width of a filter in width_type units.
3007 Applies only to double-pole filter.
3008 The default is 0.707q and gives a Butterworth response.
3014 Mix channels with specific gain levels. The filter accepts the output
3015 channel layout followed by a set of channels definitions.
3017 This filter is also designed to efficiently remap the channels of an audio
3020 The filter accepts parameters of the form:
3021 "@var{l}|@var{outdef}|@var{outdef}|..."
3025 output channel layout or number of channels
3028 output channel specification, of the form:
3029 "@var{out_name}=[@var{gain}*]@var{in_name}[+[@var{gain}*]@var{in_name}...]"
3032 output channel to define, either a channel name (FL, FR, etc.) or a channel
3033 number (c0, c1, etc.)
3036 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3039 input channel to use, see out_name for details; it is not possible to mix
3040 named and numbered input channels
3043 If the `=' in a channel specification is replaced by `<', then the gains for
3044 that specification will be renormalized so that the total is 1, thus
3045 avoiding clipping noise.
3047 @subsection Mixing examples
3049 For example, if you want to down-mix from stereo to mono, but with a bigger
3050 factor for the left channel:
3052 pan=1c|c0=0.9*c0+0.1*c1
3055 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3056 7-channels surround:
3058 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3061 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3062 that should be preferred (see "-ac" option) unless you have very specific
3065 @subsection Remapping examples
3067 The channel remapping will be effective if, and only if:
3070 @item gain coefficients are zeroes or ones,
3071 @item only one input per channel output,
3074 If all these conditions are satisfied, the filter will notify the user ("Pure
3075 channel mapping detected"), and use an optimized and lossless method to do the
3078 For example, if you have a 5.1 source and want a stereo audio stream by
3079 dropping the extra channels:
3081 pan="stereo| c0=FL | c1=FR"
3084 Given the same source, you can also switch front left and front right channels
3085 and keep the input channel layout:
3087 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3090 If the input is a stereo audio stream, you can mute the front left channel (and
3091 still keep the stereo channel layout) with:
3096 Still with a stereo audio stream input, you can copy the right channel in both
3097 front left and right:
3099 pan="stereo| c0=FR | c1=FR"
3104 ReplayGain scanner filter. This filter takes an audio stream as an input and
3105 outputs it unchanged.
3106 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3110 Convert the audio sample format, sample rate and channel layout. It is
3111 not meant to be used directly.
3114 Apply time-stretching and pitch-shifting with librubberband.
3116 The filter accepts the following options:
3120 Set tempo scale factor.
3123 Set pitch scale factor.
3126 Set transients detector.
3127 Possible values are:
3136 Possible values are:
3145 Possible values are:
3152 Set processing window size.
3153 Possible values are:
3162 Possible values are:
3169 Enable formant preservation when shift pitching.
3170 Possible values are:
3178 Possible values are:
3187 Possible values are:
3194 @section sidechaincompress
3196 This filter acts like normal compressor but has the ability to compress
3197 detected signal using second input signal.
3198 It needs two input streams and returns one output stream.
3199 First input stream will be processed depending on second stream signal.
3200 The filtered signal then can be filtered with other filters in later stages of
3201 processing. See @ref{pan} and @ref{amerge} filter.
3203 The filter accepts the following options:
3207 Set input gain. Default is 1. Range is between 0.015625 and 64.
3210 If a signal of second stream raises above this level it will affect the gain
3211 reduction of first stream.
3212 By default is 0.125. Range is between 0.00097563 and 1.
3215 Set a ratio about which the signal is reduced. 1:2 means that if the level
3216 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3217 Default is 2. Range is between 1 and 20.
3220 Amount of milliseconds the signal has to rise above the threshold before gain
3221 reduction starts. Default is 20. Range is between 0.01 and 2000.
3224 Amount of milliseconds the signal has to fall below the threshold before
3225 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3228 Set the amount by how much signal will be amplified after processing.
3229 Default is 2. Range is from 1 and 64.
3232 Curve the sharp knee around the threshold to enter gain reduction more softly.
3233 Default is 2.82843. Range is between 1 and 8.
3236 Choose if the @code{average} level between all channels of side-chain stream
3237 or the louder(@code{maximum}) channel of side-chain stream affects the
3238 reduction. Default is @code{average}.
3241 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3242 of @code{rms}. Default is @code{rms} which is mainly smoother.
3245 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3248 How much to use compressed signal in output. Default is 1.
3249 Range is between 0 and 1.
3252 @subsection Examples
3256 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3257 depending on the signal of 2nd input and later compressed signal to be
3258 merged with 2nd input:
3260 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3264 @section sidechaingate
3266 A sidechain gate acts like a normal (wideband) gate but has the ability to
3267 filter the detected signal before sending it to the gain reduction stage.
3268 Normally a gate uses the full range signal to detect a level above the
3270 For example: If you cut all lower frequencies from your sidechain signal
3271 the gate will decrease the volume of your track only if not enough highs
3272 appear. With this technique you are able to reduce the resonation of a
3273 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3275 It needs two input streams and returns one output stream.
3276 First input stream will be processed depending on second stream signal.
3278 The filter accepts the following options:
3282 Set input level before filtering.
3283 Default is 1. Allowed range is from 0.015625 to 64.
3286 Set the level of gain reduction when the signal is below the threshold.
3287 Default is 0.06125. Allowed range is from 0 to 1.
3290 If a signal rises above this level the gain reduction is released.
3291 Default is 0.125. Allowed range is from 0 to 1.
3294 Set a ratio about which the signal is reduced.
3295 Default is 2. Allowed range is from 1 to 9000.
3298 Amount of milliseconds the signal has to rise above the threshold before gain
3300 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3303 Amount of milliseconds the signal has to fall below the threshold before the
3304 reduction is increased again. Default is 250 milliseconds.
3305 Allowed range is from 0.01 to 9000.
3308 Set amount of amplification of signal after processing.
3309 Default is 1. Allowed range is from 1 to 64.
3312 Curve the sharp knee around the threshold to enter gain reduction more softly.
3313 Default is 2.828427125. Allowed range is from 1 to 8.
3316 Choose if exact signal should be taken for detection or an RMS like one.
3317 Default is rms. Can be peak or rms.
3320 Choose if the average level between all channels or the louder channel affects
3322 Default is average. Can be average or maximum.
3325 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3328 @section silencedetect
3330 Detect silence in an audio stream.
3332 This filter logs a message when it detects that the input audio volume is less
3333 or equal to a noise tolerance value for a duration greater or equal to the
3334 minimum detected noise duration.
3336 The printed times and duration are expressed in seconds.
3338 The filter accepts the following options:
3342 Set silence duration until notification (default is 2 seconds).
3345 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3346 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3349 @subsection Examples
3353 Detect 5 seconds of silence with -50dB noise tolerance:
3355 silencedetect=n=-50dB:d=5
3359 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3360 tolerance in @file{silence.mp3}:
3362 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3366 @section silenceremove
3368 Remove silence from the beginning, middle or end of the audio.
3370 The filter accepts the following options:
3374 This value is used to indicate if audio should be trimmed at beginning of
3375 the audio. A value of zero indicates no silence should be trimmed from the
3376 beginning. When specifying a non-zero value, it trims audio up until it
3377 finds non-silence. Normally, when trimming silence from beginning of audio
3378 the @var{start_periods} will be @code{1} but it can be increased to higher
3379 values to trim all audio up to specific count of non-silence periods.
3380 Default value is @code{0}.
3382 @item start_duration
3383 Specify the amount of time that non-silence must be detected before it stops
3384 trimming audio. By increasing the duration, bursts of noises can be treated
3385 as silence and trimmed off. Default value is @code{0}.
3387 @item start_threshold
3388 This indicates what sample value should be treated as silence. For digital
3389 audio, a value of @code{0} may be fine but for audio recorded from analog,
3390 you may wish to increase the value to account for background noise.
3391 Can be specified in dB (in case "dB" is appended to the specified value)
3392 or amplitude ratio. Default value is @code{0}.
3395 Set the count for trimming silence from the end of audio.
3396 To remove silence from the middle of a file, specify a @var{stop_periods}
3397 that is negative. This value is then treated as a positive value and is
3398 used to indicate the effect should restart processing as specified by
3399 @var{start_periods}, making it suitable for removing periods of silence
3400 in the middle of the audio.
3401 Default value is @code{0}.
3404 Specify a duration of silence that must exist before audio is not copied any
3405 more. By specifying a higher duration, silence that is wanted can be left in
3407 Default value is @code{0}.
3409 @item stop_threshold
3410 This is the same as @option{start_threshold} but for trimming silence from
3412 Can be specified in dB (in case "dB" is appended to the specified value)
3413 or amplitude ratio. Default value is @code{0}.
3416 This indicates that @var{stop_duration} length of audio should be left intact
3417 at the beginning of each period of silence.
3418 For example, if you want to remove long pauses between words but do not want
3419 to remove the pauses completely. Default value is @code{0}.
3422 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3423 and works better with digital silence which is exactly 0.
3424 Default value is @code{rms}.
3427 Set ratio used to calculate size of window for detecting silence.
3428 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3431 @subsection Examples
3435 The following example shows how this filter can be used to start a recording
3436 that does not contain the delay at the start which usually occurs between
3437 pressing the record button and the start of the performance:
3439 silenceremove=1:5:0.02
3443 Trim all silence encountered from beginning to end where there is more than 1
3444 second of silence in audio:
3446 silenceremove=0:0:0:-1:1:-90dB
3452 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3453 loudspeakers around the user for binaural listening via headphones (audio
3454 formats up to 9 channels supported).
3455 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3456 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3457 Austrian Academy of Sciences.
3459 To enable compilation of this filter you need to configure FFmpeg with
3460 @code{--enable-netcdf}.
3462 The filter accepts the following options:
3466 Set the SOFA file used for rendering.
3469 Set gain applied to audio. Value is in dB. Default is 0.
3472 Set rotation of virtual loudspeakers in deg. Default is 0.
3475 Set elevation of virtual speakers in deg. Default is 0.
3478 Set distance in meters between loudspeakers and the listener with near-field
3479 HRTFs. Default is 1.
3482 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3483 processing audio in time domain which is slow.
3484 @var{freq} is processing audio in frequency domain which is fast.
3485 Default is @var{freq}.
3488 Set custom positions of virtual loudspeakers. Syntax for this option is:
3489 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3490 Each virtual loudspeaker is described with short channel name following with
3491 azimuth and elevation in degreees.
3492 Each virtual loudspeaker description is separated by '|'.
3493 For example to override front left and front right channel positions use:
3494 'speakers=FL 45 15|FR 345 15'.
3495 Descriptions with unrecognised channel names are ignored.
3498 @subsection Examples
3502 Using ClubFritz6 sofa file:
3504 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3508 Using ClubFritz12 sofa file and bigger radius with small rotation:
3510 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3514 Similar as above but with custom speaker positions for front left, front right, rear left and rear right
3515 and also with custom gain:
3517 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|RL 135|RR 225:gain=28"
3521 @section stereotools
3523 This filter has some handy utilities to manage stereo signals, for converting
3524 M/S stereo recordings to L/R signal while having control over the parameters
3525 or spreading the stereo image of master track.
3527 The filter accepts the following options:
3531 Set input level before filtering for both channels. Defaults is 1.
3532 Allowed range is from 0.015625 to 64.
3535 Set output level after filtering for both channels. Defaults is 1.
3536 Allowed range is from 0.015625 to 64.
3539 Set input balance between both channels. Default is 0.
3540 Allowed range is from -1 to 1.
3543 Set output balance between both channels. Default is 0.
3544 Allowed range is from -1 to 1.
3547 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3548 clipping. Disabled by default.
3551 Mute the left channel. Disabled by default.
3554 Mute the right channel. Disabled by default.
3557 Change the phase of the left channel. Disabled by default.
3560 Change the phase of the right channel. Disabled by default.
3563 Set stereo mode. Available values are:
3567 Left/Right to Left/Right, this is default.
3570 Left/Right to Mid/Side.
3573 Mid/Side to Left/Right.
3576 Left/Right to Left/Left.
3579 Left/Right to Right/Right.
3582 Left/Right to Left + Right.
3585 Left/Right to Right/Left.
3589 Set level of side signal. Default is 1.
3590 Allowed range is from 0.015625 to 64.
3593 Set balance of side signal. Default is 0.
3594 Allowed range is from -1 to 1.
3597 Set level of the middle signal. Default is 1.
3598 Allowed range is from 0.015625 to 64.
3601 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3604 Set stereo base between mono and inversed channels. Default is 0.
3605 Allowed range is from -1 to 1.
3608 Set delay in milliseconds how much to delay left from right channel and
3609 vice versa. Default is 0. Allowed range is from -20 to 20.
3612 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3615 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3618 @subsection Examples
3622 Apply karaoke like effect:
3624 stereotools=mlev=0.015625
3628 Convert M/S signal to L/R:
3630 "stereotools=mode=ms>lr"
3634 @section stereowiden
3636 This filter enhance the stereo effect by suppressing signal common to both
3637 channels and by delaying the signal of left into right and vice versa,
3638 thereby widening the stereo effect.
3640 The filter accepts the following options:
3644 Time in milliseconds of the delay of left signal into right and vice versa.
3645 Default is 20 milliseconds.
3648 Amount of gain in delayed signal into right and vice versa. Gives a delay
3649 effect of left signal in right output and vice versa which gives widening
3650 effect. Default is 0.3.
3653 Cross feed of left into right with inverted phase. This helps in suppressing
3654 the mono. If the value is 1 it will cancel all the signal common to both
3655 channels. Default is 0.3.
3658 Set level of input signal of original channel. Default is 0.8.
3663 Boost or cut treble (upper) frequencies of the audio using a two-pole
3664 shelving filter with a response similar to that of a standard
3665 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3667 The filter accepts the following options:
3671 Give the gain at whichever is the lower of ~22 kHz and the
3672 Nyquist frequency. Its useful range is about -20 (for a large cut)
3673 to +20 (for a large boost). Beware of clipping when using a positive gain.
3676 Set the filter's central frequency and so can be used
3677 to extend or reduce the frequency range to be boosted or cut.
3678 The default value is @code{3000} Hz.
3681 Set method to specify band-width of filter.
3694 Determine how steep is the filter's shelf transition.
3699 Sinusoidal amplitude modulation.
3701 The filter accepts the following options:
3705 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3706 (20 Hz or lower) will result in a tremolo effect.
3707 This filter may also be used as a ring modulator by specifying
3708 a modulation frequency higher than 20 Hz.
3709 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3712 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3713 Default value is 0.5.
3718 Sinusoidal phase modulation.
3720 The filter accepts the following options:
3724 Modulation frequency in Hertz.
3725 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3728 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3729 Default value is 0.5.
3734 Adjust the input audio volume.
3736 It accepts the following parameters:
3740 Set audio volume expression.
3742 Output values are clipped to the maximum value.
3744 The output audio volume is given by the relation:
3746 @var{output_volume} = @var{volume} * @var{input_volume}
3749 The default value for @var{volume} is "1.0".
3752 This parameter represents the mathematical precision.
3754 It determines which input sample formats will be allowed, which affects the
3755 precision of the volume scaling.
3759 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3761 32-bit floating-point; this limits input sample format to FLT. (default)
3763 64-bit floating-point; this limits input sample format to DBL.
3767 Choose the behaviour on encountering ReplayGain side data in input frames.
3771 Remove ReplayGain side data, ignoring its contents (the default).
3774 Ignore ReplayGain side data, but leave it in the frame.
3777 Prefer the track gain, if present.
3780 Prefer the album gain, if present.
3783 @item replaygain_preamp
3784 Pre-amplification gain in dB to apply to the selected replaygain gain.
3786 Default value for @var{replaygain_preamp} is 0.0.
3789 Set when the volume expression is evaluated.
3791 It accepts the following values:
3794 only evaluate expression once during the filter initialization, or
3795 when the @samp{volume} command is sent
3798 evaluate expression for each incoming frame
3801 Default value is @samp{once}.
3804 The volume expression can contain the following parameters.
3808 frame number (starting at zero)
3811 @item nb_consumed_samples
3812 number of samples consumed by the filter
3814 number of samples in the current frame
3816 original frame position in the file
3822 PTS at start of stream
3824 time at start of stream
3830 last set volume value
3833 Note that when @option{eval} is set to @samp{once} only the
3834 @var{sample_rate} and @var{tb} variables are available, all other
3835 variables will evaluate to NAN.
3837 @subsection Commands
3839 This filter supports the following commands:
3842 Modify the volume expression.
3843 The command accepts the same syntax of the corresponding option.
3845 If the specified expression is not valid, it is kept at its current
3847 @item replaygain_noclip
3848 Prevent clipping by limiting the gain applied.
3850 Default value for @var{replaygain_noclip} is 1.
3854 @subsection Examples
3858 Halve the input audio volume:
3862 volume=volume=-6.0206dB
3865 In all the above example the named key for @option{volume} can be
3866 omitted, for example like in:
3872 Increase input audio power by 6 decibels using fixed-point precision:
3874 volume=volume=6dB:precision=fixed
3878 Fade volume after time 10 with an annihilation period of 5 seconds:
3880 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3884 @section volumedetect
3886 Detect the volume of the input video.
3888 The filter has no parameters. The input is not modified. Statistics about
3889 the volume will be printed in the log when the input stream end is reached.
3891 In particular it will show the mean volume (root mean square), maximum
3892 volume (on a per-sample basis), and the beginning of a histogram of the
3893 registered volume values (from the maximum value to a cumulated 1/1000 of
3896 All volumes are in decibels relative to the maximum PCM value.
3898 @subsection Examples
3900 Here is an excerpt of the output:
3902 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3903 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3904 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3905 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3906 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3907 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3908 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3909 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3910 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3916 The mean square energy is approximately -27 dB, or 10^-2.7.
3918 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3920 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3923 In other words, raising the volume by +4 dB does not cause any clipping,
3924 raising it by +5 dB causes clipping for 6 samples, etc.
3926 @c man end AUDIO FILTERS
3928 @chapter Audio Sources
3929 @c man begin AUDIO SOURCES
3931 Below is a description of the currently available audio sources.
3935 Buffer audio frames, and make them available to the filter chain.
3937 This source is mainly intended for a programmatic use, in particular
3938 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3940 It accepts the following parameters:
3944 The timebase which will be used for timestamps of submitted frames. It must be
3945 either a floating-point number or in @var{numerator}/@var{denominator} form.
3948 The sample rate of the incoming audio buffers.
3951 The sample format of the incoming audio buffers.
3952 Either a sample format name or its corresponding integer representation from
3953 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
3955 @item channel_layout
3956 The channel layout of the incoming audio buffers.
3957 Either a channel layout name from channel_layout_map in
3958 @file{libavutil/channel_layout.c} or its corresponding integer representation
3959 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
3962 The number of channels of the incoming audio buffers.
3963 If both @var{channels} and @var{channel_layout} are specified, then they
3968 @subsection Examples
3971 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
3974 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
3975 Since the sample format with name "s16p" corresponds to the number
3976 6 and the "stereo" channel layout corresponds to the value 0x3, this is
3979 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
3984 Generate an audio signal specified by an expression.
3986 This source accepts in input one or more expressions (one for each
3987 channel), which are evaluated and used to generate a corresponding
3990 This source accepts the following options:
3994 Set the '|'-separated expressions list for each separate channel. In case the
3995 @option{channel_layout} option is not specified, the selected channel layout
3996 depends on the number of provided expressions. Otherwise the last
3997 specified expression is applied to the remaining output channels.
3999 @item channel_layout, c
4000 Set the channel layout. The number of channels in the specified layout
4001 must be equal to the number of specified expressions.
4004 Set the minimum duration of the sourced audio. See
4005 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4006 for the accepted syntax.
4007 Note that the resulting duration may be greater than the specified
4008 duration, as the generated audio is always cut at the end of a
4011 If not specified, or the expressed duration is negative, the audio is
4012 supposed to be generated forever.
4015 Set the number of samples per channel per each output frame,
4018 @item sample_rate, s
4019 Specify the sample rate, default to 44100.
4022 Each expression in @var{exprs} can contain the following constants:
4026 number of the evaluated sample, starting from 0
4029 time of the evaluated sample expressed in seconds, starting from 0
4036 @subsection Examples
4046 Generate a sin signal with frequency of 440 Hz, set sample rate to
4049 aevalsrc="sin(440*2*PI*t):s=8000"
4053 Generate a two channels signal, specify the channel layout (Front
4054 Center + Back Center) explicitly:
4056 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4060 Generate white noise:
4062 aevalsrc="-2+random(0)"
4066 Generate an amplitude modulated signal:
4068 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4072 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4074 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4081 The null audio source, return unprocessed audio frames. It is mainly useful
4082 as a template and to be employed in analysis / debugging tools, or as
4083 the source for filters which ignore the input data (for example the sox
4086 This source accepts the following options:
4090 @item channel_layout, cl
4092 Specifies the channel layout, and can be either an integer or a string
4093 representing a channel layout. The default value of @var{channel_layout}
4096 Check the channel_layout_map definition in
4097 @file{libavutil/channel_layout.c} for the mapping between strings and
4098 channel layout values.
4100 @item sample_rate, r
4101 Specifies the sample rate, and defaults to 44100.
4104 Set the number of samples per requested frames.
4108 @subsection Examples
4112 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4114 anullsrc=r=48000:cl=4
4118 Do the same operation with a more obvious syntax:
4120 anullsrc=r=48000:cl=mono
4124 All the parameters need to be explicitly defined.
4128 Synthesize a voice utterance using the libflite library.
4130 To enable compilation of this filter you need to configure FFmpeg with
4131 @code{--enable-libflite}.
4133 Note that the flite library is not thread-safe.
4135 The filter accepts the following options:
4140 If set to 1, list the names of the available voices and exit
4141 immediately. Default value is 0.
4144 Set the maximum number of samples per frame. Default value is 512.
4147 Set the filename containing the text to speak.
4150 Set the text to speak.
4153 Set the voice to use for the speech synthesis. Default value is
4154 @code{kal}. See also the @var{list_voices} option.
4157 @subsection Examples
4161 Read from file @file{speech.txt}, and synthesize the text using the
4162 standard flite voice:
4164 flite=textfile=speech.txt
4168 Read the specified text selecting the @code{slt} voice:
4170 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4174 Input text to ffmpeg:
4176 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4180 Make @file{ffplay} speak the specified text, using @code{flite} and
4181 the @code{lavfi} device:
4183 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4187 For more information about libflite, check:
4188 @url{http://www.speech.cs.cmu.edu/flite/}
4192 Generate a noise audio signal.
4194 The filter accepts the following options:
4197 @item sample_rate, r
4198 Specify the sample rate. Default value is 48000 Hz.
4201 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4205 Specify the duration of the generated audio stream. Not specifying this option
4206 results in noise with an infinite length.
4208 @item color, colour, c
4209 Specify the color of noise. Available noise colors are white, pink, and brown.
4210 Default color is white.
4213 Specify a value used to seed the PRNG.
4216 Set the number of samples per each output frame, default is 1024.
4219 @subsection Examples
4224 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4226 anoisesrc=d=60:c=pink:r=44100:a=0.5
4232 Generate an audio signal made of a sine wave with amplitude 1/8.
4234 The audio signal is bit-exact.
4236 The filter accepts the following options:
4241 Set the carrier frequency. Default is 440 Hz.
4243 @item beep_factor, b
4244 Enable a periodic beep every second with frequency @var{beep_factor} times
4245 the carrier frequency. Default is 0, meaning the beep is disabled.
4247 @item sample_rate, r
4248 Specify the sample rate, default is 44100.
4251 Specify the duration of the generated audio stream.
4253 @item samples_per_frame
4254 Set the number of samples per output frame.
4256 The expression can contain the following constants:
4260 The (sequential) number of the output audio frame, starting from 0.
4263 The PTS (Presentation TimeStamp) of the output audio frame,
4264 expressed in @var{TB} units.
4267 The PTS of the output audio frame, expressed in seconds.
4270 The timebase of the output audio frames.
4273 Default is @code{1024}.
4276 @subsection Examples
4281 Generate a simple 440 Hz sine wave:
4287 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4291 sine=frequency=220:beep_factor=4:duration=5
4295 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4298 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4302 @c man end AUDIO SOURCES
4304 @chapter Audio Sinks
4305 @c man begin AUDIO SINKS
4307 Below is a description of the currently available audio sinks.
4309 @section abuffersink
4311 Buffer audio frames, and make them available to the end of filter chain.
4313 This sink is mainly intended for programmatic use, in particular
4314 through the interface defined in @file{libavfilter/buffersink.h}
4315 or the options system.
4317 It accepts a pointer to an AVABufferSinkContext structure, which
4318 defines the incoming buffers' formats, to be passed as the opaque
4319 parameter to @code{avfilter_init_filter} for initialization.
4322 Null audio sink; do absolutely nothing with the input audio. It is
4323 mainly useful as a template and for use in analysis / debugging
4326 @c man end AUDIO SINKS
4328 @chapter Video Filters
4329 @c man begin VIDEO FILTERS
4331 When you configure your FFmpeg build, you can disable any of the
4332 existing filters using @code{--disable-filters}.
4333 The configure output will show the video filters included in your
4336 Below is a description of the currently available video filters.
4338 @section alphaextract
4340 Extract the alpha component from the input as a grayscale video. This
4341 is especially useful with the @var{alphamerge} filter.
4345 Add or replace the alpha component of the primary input with the
4346 grayscale value of a second input. This is intended for use with
4347 @var{alphaextract} to allow the transmission or storage of frame
4348 sequences that have alpha in a format that doesn't support an alpha
4351 For example, to reconstruct full frames from a normal YUV-encoded video
4352 and a separate video created with @var{alphaextract}, you might use:
4354 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4357 Since this filter is designed for reconstruction, it operates on frame
4358 sequences without considering timestamps, and terminates when either
4359 input reaches end of stream. This will cause problems if your encoding
4360 pipeline drops frames. If you're trying to apply an image as an
4361 overlay to a video stream, consider the @var{overlay} filter instead.
4365 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4366 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4367 Substation Alpha) subtitles files.
4369 This filter accepts the following option in addition to the common options from
4370 the @ref{subtitles} filter:
4374 Set the shaping engine
4376 Available values are:
4379 The default libass shaping engine, which is the best available.
4381 Fast, font-agnostic shaper that can do only substitutions
4383 Slower shaper using OpenType for substitutions and positioning
4386 The default is @code{auto}.
4390 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4392 The filter accepts the following options:
4396 Set threshold A for 1st plane. Default is 0.02.
4397 Valid range is 0 to 0.3.
4400 Set threshold B for 1st plane. Default is 0.04.
4401 Valid range is 0 to 5.
4404 Set threshold A for 2nd plane. Default is 0.02.
4405 Valid range is 0 to 0.3.
4408 Set threshold B for 2nd plane. Default is 0.04.
4409 Valid range is 0 to 5.
4412 Set threshold A for 3rd plane. Default is 0.02.
4413 Valid range is 0 to 0.3.
4416 Set threshold B for 3rd plane. Default is 0.04.
4417 Valid range is 0 to 5.
4419 Threshold A is designed to react on abrupt changes in the input signal and
4420 threshold B is designed to react on continuous changes in the input signal.
4423 Set number of frames filter will use for averaging. Default is 33. Must be odd
4424 number in range [5, 129].
4427 Set what planes of frame filter will use for averaging. Default is all.
4432 Apply average blur filter.
4434 The filter accepts the following options:
4438 Set horizontal kernel size.
4441 Set which planes to filter. By default all planes are filtered.
4444 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4445 Default is @code{0}.
4450 Compute the bounding box for the non-black pixels in the input frame
4453 This filter computes the bounding box containing all the pixels with a
4454 luminance value greater than the minimum allowed value.
4455 The parameters describing the bounding box are printed on the filter
4458 The filter accepts the following option:
4462 Set the minimal luminance value. Default is @code{16}.
4465 @section bitplanenoise
4467 Show and measure bit plane noise.
4469 The filter accepts the following options:
4473 Set which plane to analyze. Default is @code{1}.
4476 Filter out noisy pixels from @code{bitplane} set above.
4477 Default is disabled.
4480 @section blackdetect
4482 Detect video intervals that are (almost) completely black. Can be
4483 useful to detect chapter transitions, commercials, or invalid
4484 recordings. Output lines contains the time for the start, end and
4485 duration of the detected black interval expressed in seconds.
4487 In order to display the output lines, you need to set the loglevel at
4488 least to the AV_LOG_INFO value.
4490 The filter accepts the following options:
4493 @item black_min_duration, d
4494 Set the minimum detected black duration expressed in seconds. It must
4495 be a non-negative floating point number.
4497 Default value is 2.0.
4499 @item picture_black_ratio_th, pic_th
4500 Set the threshold for considering a picture "black".
4501 Express the minimum value for the ratio:
4503 @var{nb_black_pixels} / @var{nb_pixels}
4506 for which a picture is considered black.
4507 Default value is 0.98.
4509 @item pixel_black_th, pix_th
4510 Set the threshold for considering a pixel "black".
4512 The threshold expresses the maximum pixel luminance value for which a
4513 pixel is considered "black". The provided value is scaled according to
4514 the following equation:
4516 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4519 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4520 the input video format, the range is [0-255] for YUV full-range
4521 formats and [16-235] for YUV non full-range formats.
4523 Default value is 0.10.
4526 The following example sets the maximum pixel threshold to the minimum
4527 value, and detects only black intervals of 2 or more seconds:
4529 blackdetect=d=2:pix_th=0.00
4534 Detect frames that are (almost) completely black. Can be useful to
4535 detect chapter transitions or commercials. Output lines consist of
4536 the frame number of the detected frame, the percentage of blackness,
4537 the position in the file if known or -1 and the timestamp in seconds.
4539 In order to display the output lines, you need to set the loglevel at
4540 least to the AV_LOG_INFO value.
4542 It accepts the following parameters:
4547 The percentage of the pixels that have to be below the threshold; it defaults to
4550 @item threshold, thresh
4551 The threshold below which a pixel value is considered black; it defaults to
4556 @section blend, tblend
4558 Blend two video frames into each other.
4560 The @code{blend} filter takes two input streams and outputs one
4561 stream, the first input is the "top" layer and second input is
4562 "bottom" layer. By default, the output terminates when the longest input terminates.
4564 The @code{tblend} (time blend) filter takes two consecutive frames
4565 from one single stream, and outputs the result obtained by blending
4566 the new frame on top of the old frame.
4568 A description of the accepted options follows.
4576 Set blend mode for specific pixel component or all pixel components in case
4577 of @var{all_mode}. Default value is @code{normal}.
4579 Available values for component modes are:
4620 Set blend opacity for specific pixel component or all pixel components in case
4621 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4628 Set blend expression for specific pixel component or all pixel components in case
4629 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4631 The expressions can use the following variables:
4635 The sequential number of the filtered frame, starting from @code{0}.
4639 the coordinates of the current sample
4643 the width and height of currently filtered plane
4647 Width and height scale depending on the currently filtered plane. It is the
4648 ratio between the corresponding luma plane number of pixels and the current
4649 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4650 @code{0.5,0.5} for chroma planes.
4653 Time of the current frame, expressed in seconds.
4656 Value of pixel component at current location for first video frame (top layer).
4659 Value of pixel component at current location for second video frame (bottom layer).
4663 Force termination when the shortest input terminates. Default is
4664 @code{0}. This option is only defined for the @code{blend} filter.
4667 Continue applying the last bottom frame after the end of the stream. A value of
4668 @code{0} disable the filter after the last frame of the bottom layer is reached.
4669 Default is @code{1}. This option is only defined for the @code{blend} filter.
4672 @subsection Examples
4676 Apply transition from bottom layer to top layer in first 10 seconds:
4678 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4682 Apply 1x1 checkerboard effect:
4684 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4688 Apply uncover left effect:
4690 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4694 Apply uncover down effect:
4696 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4700 Apply uncover up-left effect:
4702 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4706 Split diagonally video and shows top and bottom layer on each side:
4708 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4712 Display differences between the current and the previous frame:
4714 tblend=all_mode=difference128
4720 Apply a boxblur algorithm to the input video.
4722 It accepts the following parameters:
4726 @item luma_radius, lr
4727 @item luma_power, lp
4728 @item chroma_radius, cr
4729 @item chroma_power, cp
4730 @item alpha_radius, ar
4731 @item alpha_power, ap
4735 A description of the accepted options follows.
4738 @item luma_radius, lr
4739 @item chroma_radius, cr
4740 @item alpha_radius, ar
4741 Set an expression for the box radius in pixels used for blurring the
4742 corresponding input plane.
4744 The radius value must be a non-negative number, and must not be
4745 greater than the value of the expression @code{min(w,h)/2} for the
4746 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4749 Default value for @option{luma_radius} is "2". If not specified,
4750 @option{chroma_radius} and @option{alpha_radius} default to the
4751 corresponding value set for @option{luma_radius}.
4753 The expressions can contain the following constants:
4757 The input width and height in pixels.
4761 The input chroma image width and height in pixels.
4765 The horizontal and vertical chroma subsample values. For example, for the
4766 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4769 @item luma_power, lp
4770 @item chroma_power, cp
4771 @item alpha_power, ap
4772 Specify how many times the boxblur filter is applied to the
4773 corresponding plane.
4775 Default value for @option{luma_power} is 2. If not specified,
4776 @option{chroma_power} and @option{alpha_power} default to the
4777 corresponding value set for @option{luma_power}.
4779 A value of 0 will disable the effect.
4782 @subsection Examples
4786 Apply a boxblur filter with the luma, chroma, and alpha radii
4789 boxblur=luma_radius=2:luma_power=1
4794 Set the luma radius to 2, and alpha and chroma radius to 0:
4796 boxblur=2:1:cr=0:ar=0
4800 Set the luma and chroma radii to a fraction of the video dimension:
4802 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4808 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4809 Deinterlacing Filter").
4811 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4812 interpolation algorithms.
4813 It accepts the following parameters:
4817 The interlacing mode to adopt. It accepts one of the following values:
4821 Output one frame for each frame.
4823 Output one frame for each field.
4826 The default value is @code{send_field}.
4829 The picture field parity assumed for the input interlaced video. It accepts one
4830 of the following values:
4834 Assume the top field is first.
4836 Assume the bottom field is first.
4838 Enable automatic detection of field parity.
4841 The default value is @code{auto}.
4842 If the interlacing is unknown or the decoder does not export this information,
4843 top field first will be assumed.
4846 Specify which frames to deinterlace. Accept one of the following
4851 Deinterlace all frames.
4853 Only deinterlace frames marked as interlaced.
4856 The default value is @code{all}.
4860 YUV colorspace color/chroma keying.
4862 The filter accepts the following options:
4866 The color which will be replaced with transparency.
4869 Similarity percentage with the key color.
4871 0.01 matches only the exact key color, while 1.0 matches everything.
4876 0.0 makes pixels either fully transparent, or not transparent at all.
4878 Higher values result in semi-transparent pixels, with a higher transparency
4879 the more similar the pixels color is to the key color.
4882 Signals that the color passed is already in YUV instead of RGB.
4884 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4885 This can be used to pass exact YUV values as hexadecimal numbers.
4888 @subsection Examples
4892 Make every green pixel in the input image transparent:
4894 ffmpeg -i input.png -vf chromakey=green out.png
4898 Overlay a greenscreen-video on top of a static black background.
4900 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
4906 Display CIE color diagram with pixels overlaid onto it.
4908 The filter accepts the following options:
4923 @item uhdtv, rec2020
4936 Set what gamuts to draw.
4938 See @code{system} option for available values.
4941 Set ciescope size, by default set to 512.
4944 Set intensity used to map input pixel values to CIE diagram.
4947 Set contrast used to draw tongue colors that are out of active color system gamut.
4950 Correct gamma displayed on scope, by default enabled.
4953 Show white point on CIE diagram, by default disabled.
4956 Set input gamma. Used only with XYZ input color space.
4961 Visualize information exported by some codecs.
4963 Some codecs can export information through frames using side-data or other
4964 means. For example, some MPEG based codecs export motion vectors through the
4965 @var{export_mvs} flag in the codec @option{flags2} option.
4967 The filter accepts the following option:
4971 Set motion vectors to visualize.
4973 Available flags for @var{mv} are:
4977 forward predicted MVs of P-frames
4979 forward predicted MVs of B-frames
4981 backward predicted MVs of B-frames
4985 Display quantization parameters using the chroma planes.
4988 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
4990 Available flags for @var{mv_type} are:
4994 forward predicted MVs
4996 backward predicted MVs
4999 @item frame_type, ft
5000 Set frame type to visualize motion vectors of.
5002 Available flags for @var{frame_type} are:
5006 intra-coded frames (I-frames)
5008 predicted frames (P-frames)
5010 bi-directionally predicted frames (B-frames)
5014 @subsection Examples
5018 Visualize forward predicted MVs of all frames using @command{ffplay}:
5020 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5024 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5026 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5030 @section colorbalance
5031 Modify intensity of primary colors (red, green and blue) of input frames.
5033 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5034 regions for the red-cyan, green-magenta or blue-yellow balance.
5036 A positive adjustment value shifts the balance towards the primary color, a negative
5037 value towards the complementary color.
5039 The filter accepts the following options:
5045 Adjust red, green and blue shadows (darkest pixels).
5050 Adjust red, green and blue midtones (medium pixels).
5055 Adjust red, green and blue highlights (brightest pixels).
5057 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5060 @subsection Examples
5064 Add red color cast to shadows:
5071 RGB colorspace color keying.
5073 The filter accepts the following options:
5077 The color which will be replaced with transparency.
5080 Similarity percentage with the key color.
5082 0.01 matches only the exact key color, while 1.0 matches everything.
5087 0.0 makes pixels either fully transparent, or not transparent at all.
5089 Higher values result in semi-transparent pixels, with a higher transparency
5090 the more similar the pixels color is to the key color.
5093 @subsection Examples
5097 Make every green pixel in the input image transparent:
5099 ffmpeg -i input.png -vf colorkey=green out.png
5103 Overlay a greenscreen-video on top of a static background image.
5105 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
5109 @section colorlevels
5111 Adjust video input frames using levels.
5113 The filter accepts the following options:
5120 Adjust red, green, blue and alpha input black point.
5121 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5127 Adjust red, green, blue and alpha input white point.
5128 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5130 Input levels are used to lighten highlights (bright tones), darken shadows
5131 (dark tones), change the balance of bright and dark tones.
5137 Adjust red, green, blue and alpha output black point.
5138 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5144 Adjust red, green, blue and alpha output white point.
5145 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5147 Output levels allows manual selection of a constrained output level range.
5150 @subsection Examples
5154 Make video output darker:
5156 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5162 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5166 Make video output lighter:
5168 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5172 Increase brightness:
5174 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5178 @section colorchannelmixer
5180 Adjust video input frames by re-mixing color channels.
5182 This filter modifies a color channel by adding the values associated to
5183 the other channels of the same pixels. For example if the value to
5184 modify is red, the output value will be:
5186 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5189 The filter accepts the following options:
5196 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5197 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5203 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5204 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5210 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5211 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5217 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5218 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5220 Allowed ranges for options are @code{[-2.0, 2.0]}.
5223 @subsection Examples
5227 Convert source to grayscale:
5229 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5232 Simulate sepia tones:
5234 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5238 @section colormatrix
5240 Convert color matrix.
5242 The filter accepts the following options:
5247 Specify the source and destination color matrix. Both values must be
5250 The accepted values are:
5269 For example to convert from BT.601 to SMPTE-240M, use the command:
5271 colormatrix=bt601:smpte240m
5276 Convert colorspace, transfer characteristics or color primaries.
5278 The filter accepts the following options:
5283 Specify all color properties at once.
5285 The accepted values are:
5315 Specify output colorspace.
5317 The accepted values are:
5326 BT.470BG or BT.601-6 625
5329 SMPTE-170M or BT.601-6 525
5335 BT.2020 with non-constant luminance
5341 Specify output transfer characteristics.
5343 The accepted values are:
5349 Constant gamma of 2.2
5352 Constant gamma of 2.8
5355 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5361 BT.2020 for 10-bits content
5364 BT.2020 for 12-bits content
5370 Specify output color primaries.
5372 The accepted values are:
5381 BT.470BG or BT.601-6 625
5384 SMPTE-170M or BT.601-6 525
5396 Specify output color range.
5398 The accepted values are:
5401 MPEG (restricted) range
5409 Specify output color format.
5411 The accepted values are:
5414 YUV 4:2:0 planar 8-bits
5417 YUV 4:2:0 planar 10-bits
5420 YUV 4:2:0 planar 12-bits
5423 YUV 4:2:2 planar 8-bits
5426 YUV 4:2:2 planar 10-bits
5429 YUV 4:2:2 planar 12-bits
5432 YUV 4:4:4 planar 8-bits
5435 YUV 4:4:4 planar 10-bits
5438 YUV 4:4:4 planar 12-bits
5443 Do a fast conversion, which skips gamma/primary correction. This will take
5444 significantly less CPU, but will be mathematically incorrect. To get output
5445 compatible with that produced by the colormatrix filter, use fast=1.
5448 Specify dithering mode.
5450 The accepted values are:
5456 Floyd-Steinberg dithering
5460 Whitepoint adaptation mode.
5462 The accepted values are:
5465 Bradford whitepoint adaptation
5468 von Kries whitepoint adaptation
5471 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5475 Override all input properties at once. Same accepted values as @ref{all}.
5478 Override input colorspace. Same accepted values as @ref{space}.
5481 Override input color primaries. Same accepted values as @ref{primaries}.
5484 Override input transfer characteristics. Same accepted values as @ref{trc}.
5487 Override input color range. Same accepted values as @ref{range}.
5491 The filter converts the transfer characteristics, color space and color
5492 primaries to the specified user values. The output value, if not specified,
5493 is set to a default value based on the "all" property. If that property is
5494 also not specified, the filter will log an error. The output color range and
5495 format default to the same value as the input color range and format. The
5496 input transfer characteristics, color space, color primaries and color range
5497 should be set on the input data. If any of these are missing, the filter will
5498 log an error and no conversion will take place.
5500 For example to convert the input to SMPTE-240M, use the command:
5502 colorspace=smpte240m
5505 @section convolution
5507 Apply convolution 3x3 or 5x5 filter.
5509 The filter accepts the following options:
5516 Set matrix for each plane.
5517 Matrix is sequence of 9 or 25 signed integers.
5523 Set multiplier for calculated value for each plane.
5529 Set bias for each plane. This value is added to the result of the multiplication.
5530 Useful for making the overall image brighter or darker. Default is 0.0.
5533 @subsection Examples
5539 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"
5545 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"
5551 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"
5557 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"
5563 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"
5569 Copy the input source unchanged to the output. This is mainly useful for
5574 Video filtering on GPU using Apple's CoreImage API on OSX.
5576 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5577 processed by video hardware. However, software-based OpenGL implementations
5578 exist which means there is no guarantee for hardware processing. It depends on
5581 There are many filters and image generators provided by Apple that come with a
5582 large variety of options. The filter has to be referenced by its name along
5585 The coreimage filter accepts the following options:
5588 List all available filters and generators along with all their respective
5589 options as well as possible minimum and maximum values along with the default
5596 Specify all filters by their respective name and options.
5597 Use @var{list_filters} to determine all valid filter names and options.
5598 Numerical options are specified by a float value and are automatically clamped
5599 to their respective value range. Vector and color options have to be specified
5600 by a list of space separated float values. Character escaping has to be done.
5601 A special option name @code{default} is available to use default options for a
5604 It is required to specify either @code{default} or at least one of the filter options.
5605 All omitted options are used with their default values.
5606 The syntax of the filter string is as follows:
5608 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5612 Specify a rectangle where the output of the filter chain is copied into the
5613 input image. It is given by a list of space separated float values:
5615 output_rect=x\ y\ width\ height
5617 If not given, the output rectangle equals the dimensions of the input image.
5618 The output rectangle is automatically cropped at the borders of the input
5619 image. Negative values are valid for each component.
5621 output_rect=25\ 25\ 100\ 100
5625 Several filters can be chained for successive processing without GPU-HOST
5626 transfers allowing for fast processing of complex filter chains.
5627 Currently, only filters with zero (generators) or exactly one (filters) input
5628 image and one output image are supported. Also, transition filters are not yet
5631 Some filters generate output images with additional padding depending on the
5632 respective filter kernel. The padding is automatically removed to ensure the
5633 filter output has the same size as the input image.
5635 For image generators, the size of the output image is determined by the
5636 previous output image of the filter chain or the input image of the whole
5637 filterchain, respectively. The generators do not use the pixel information of
5638 this image to generate their output. However, the generated output is
5639 blended onto this image, resulting in partial or complete coverage of the
5642 The @ref{coreimagesrc} video source can be used for generating input images
5643 which are directly fed into the filter chain. By using it, providing input
5644 images by another video source or an input video is not required.
5646 @subsection Examples
5651 List all filters available:
5653 coreimage=list_filters=true
5657 Use the CIBoxBlur filter with default options to blur an image:
5659 coreimage=filter=CIBoxBlur@@default
5663 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5664 its center at 100x100 and a radius of 50 pixels:
5666 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5670 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5671 given as complete and escaped command-line for Apple's standard bash shell:
5673 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5679 Crop the input video to given dimensions.
5681 It accepts the following parameters:
5685 The width of the output video. It defaults to @code{iw}.
5686 This expression is evaluated only once during the filter
5687 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5690 The height of the output video. It defaults to @code{ih}.
5691 This expression is evaluated only once during the filter
5692 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5695 The horizontal position, in the input video, of the left edge of the output
5696 video. It defaults to @code{(in_w-out_w)/2}.
5697 This expression is evaluated per-frame.
5700 The vertical position, in the input video, of the top edge of the output video.
5701 It defaults to @code{(in_h-out_h)/2}.
5702 This expression is evaluated per-frame.
5705 If set to 1 will force the output display aspect ratio
5706 to be the same of the input, by changing the output sample aspect
5707 ratio. It defaults to 0.
5710 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5711 width/height/x/y as specified and will not be rounded to nearest smaller value.
5715 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5716 expressions containing the following constants:
5721 The computed values for @var{x} and @var{y}. They are evaluated for
5726 The input width and height.
5730 These are the same as @var{in_w} and @var{in_h}.
5734 The output (cropped) width and height.
5738 These are the same as @var{out_w} and @var{out_h}.
5741 same as @var{iw} / @var{ih}
5744 input sample aspect ratio
5747 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5751 horizontal and vertical chroma subsample values. For example for the
5752 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5755 The number of the input frame, starting from 0.
5758 the position in the file of the input frame, NAN if unknown
5761 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5765 The expression for @var{out_w} may depend on the value of @var{out_h},
5766 and the expression for @var{out_h} may depend on @var{out_w}, but they
5767 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5768 evaluated after @var{out_w} and @var{out_h}.
5770 The @var{x} and @var{y} parameters specify the expressions for the
5771 position of the top-left corner of the output (non-cropped) area. They
5772 are evaluated for each frame. If the evaluated value is not valid, it
5773 is approximated to the nearest valid value.
5775 The expression for @var{x} may depend on @var{y}, and the expression
5776 for @var{y} may depend on @var{x}.
5778 @subsection Examples
5782 Crop area with size 100x100 at position (12,34).
5787 Using named options, the example above becomes:
5789 crop=w=100:h=100:x=12:y=34
5793 Crop the central input area with size 100x100:
5799 Crop the central input area with size 2/3 of the input video:
5801 crop=2/3*in_w:2/3*in_h
5805 Crop the input video central square:
5812 Delimit the rectangle with the top-left corner placed at position
5813 100:100 and the right-bottom corner corresponding to the right-bottom
5814 corner of the input image.
5816 crop=in_w-100:in_h-100:100:100
5820 Crop 10 pixels from the left and right borders, and 20 pixels from
5821 the top and bottom borders
5823 crop=in_w-2*10:in_h-2*20
5827 Keep only the bottom right quarter of the input image:
5829 crop=in_w/2:in_h/2:in_w/2:in_h/2
5833 Crop height for getting Greek harmony:
5835 crop=in_w:1/PHI*in_w
5839 Apply trembling effect:
5841 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)
5845 Apply erratic camera effect depending on timestamp:
5847 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)"
5851 Set x depending on the value of y:
5853 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5857 @subsection Commands
5859 This filter supports the following commands:
5865 Set width/height of the output video and the horizontal/vertical position
5867 The command accepts the same syntax of the corresponding option.
5869 If the specified expression is not valid, it is kept at its current
5875 Auto-detect the crop size.
5877 It calculates the necessary cropping parameters and prints the
5878 recommended parameters via the logging system. The detected dimensions
5879 correspond to the non-black area of the input video.
5881 It accepts the following parameters:
5886 Set higher black value threshold, which can be optionally specified
5887 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5888 value greater to the set value is considered non-black. It defaults to 24.
5889 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5890 on the bitdepth of the pixel format.
5893 The value which the width/height should be divisible by. It defaults to
5894 16. The offset is automatically adjusted to center the video. Use 2 to
5895 get only even dimensions (needed for 4:2:2 video). 16 is best when
5896 encoding to most video codecs.
5898 @item reset_count, reset
5899 Set the counter that determines after how many frames cropdetect will
5900 reset the previously detected largest video area and start over to
5901 detect the current optimal crop area. Default value is 0.
5903 This can be useful when channel logos distort the video area. 0
5904 indicates 'never reset', and returns the largest area encountered during
5911 Apply color adjustments using curves.
5913 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
5914 component (red, green and blue) has its values defined by @var{N} key points
5915 tied from each other using a smooth curve. The x-axis represents the pixel
5916 values from the input frame, and the y-axis the new pixel values to be set for
5919 By default, a component curve is defined by the two points @var{(0;0)} and
5920 @var{(1;1)}. This creates a straight line where each original pixel value is
5921 "adjusted" to its own value, which means no change to the image.
5923 The filter allows you to redefine these two points and add some more. A new
5924 curve (using a natural cubic spline interpolation) will be define to pass
5925 smoothly through all these new coordinates. The new defined points needs to be
5926 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
5927 be in the @var{[0;1]} interval. If the computed curves happened to go outside
5928 the vector spaces, the values will be clipped accordingly.
5930 The filter accepts the following options:
5934 Select one of the available color presets. This option can be used in addition
5935 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
5936 options takes priority on the preset values.
5937 Available presets are:
5940 @item color_negative
5943 @item increase_contrast
5945 @item linear_contrast
5946 @item medium_contrast
5948 @item strong_contrast
5951 Default is @code{none}.
5953 Set the master key points. These points will define a second pass mapping. It
5954 is sometimes called a "luminance" or "value" mapping. It can be used with
5955 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
5956 post-processing LUT.
5958 Set the key points for the red component.
5960 Set the key points for the green component.
5962 Set the key points for the blue component.
5964 Set the key points for all components (not including master).
5965 Can be used in addition to the other key points component
5966 options. In this case, the unset component(s) will fallback on this
5967 @option{all} setting.
5969 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
5971 Save Gnuplot script of the curves in specified file.
5974 To avoid some filtergraph syntax conflicts, each key points list need to be
5975 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
5977 @subsection Examples
5981 Increase slightly the middle level of blue:
5983 curves=blue='0/0 0.5/0.58 1/1'
5989 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'
5991 Here we obtain the following coordinates for each components:
5994 @code{(0;0.11) (0.42;0.51) (1;0.95)}
5996 @code{(0;0) (0.50;0.48) (1;1)}
5998 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6002 The previous example can also be achieved with the associated built-in preset:
6004 curves=preset=vintage
6014 Use a Photoshop preset and redefine the points of the green component:
6016 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6020 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6021 and @command{gnuplot}:
6023 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6024 gnuplot -p /tmp/curves.plt
6030 Video data analysis filter.
6032 This filter shows hexadecimal pixel values of part of video.
6034 The filter accepts the following options:
6038 Set output video size.
6041 Set x offset from where to pick pixels.
6044 Set y offset from where to pick pixels.
6047 Set scope mode, can be one of the following:
6050 Draw hexadecimal pixel values with white color on black background.
6053 Draw hexadecimal pixel values with input video pixel color on black
6057 Draw hexadecimal pixel values on color background picked from input video,
6058 the text color is picked in such way so its always visible.
6062 Draw rows and columns numbers on left and top of video.
6065 Set background opacity.
6070 Denoise frames using 2D DCT (frequency domain filtering).
6072 This filter is not designed for real time.
6074 The filter accepts the following options:
6078 Set the noise sigma constant.
6080 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6081 coefficient (absolute value) below this threshold with be dropped.
6083 If you need a more advanced filtering, see @option{expr}.
6085 Default is @code{0}.
6088 Set number overlapping pixels for each block. Since the filter can be slow, you
6089 may want to reduce this value, at the cost of a less effective filter and the
6090 risk of various artefacts.
6092 If the overlapping value doesn't permit processing the whole input width or
6093 height, a warning will be displayed and according borders won't be denoised.
6095 Default value is @var{blocksize}-1, which is the best possible setting.
6098 Set the coefficient factor expression.
6100 For each coefficient of a DCT block, this expression will be evaluated as a
6101 multiplier value for the coefficient.
6103 If this is option is set, the @option{sigma} option will be ignored.
6105 The absolute value of the coefficient can be accessed through the @var{c}
6109 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6110 @var{blocksize}, which is the width and height of the processed blocks.
6112 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6113 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6114 on the speed processing. Also, a larger block size does not necessarily means a
6118 @subsection Examples
6120 Apply a denoise with a @option{sigma} of @code{4.5}:
6125 The same operation can be achieved using the expression system:
6127 dctdnoiz=e='gte(c, 4.5*3)'
6130 Violent denoise using a block size of @code{16x16}:
6137 Remove banding artifacts from input video.
6138 It works by replacing banded pixels with average value of referenced pixels.
6140 The filter accepts the following options:
6147 Set banding detection threshold for each plane. Default is 0.02.
6148 Valid range is 0.00003 to 0.5.
6149 If difference between current pixel and reference pixel is less than threshold,
6150 it will be considered as banded.
6153 Banding detection range in pixels. Default is 16. If positive, random number
6154 in range 0 to set value will be used. If negative, exact absolute value
6156 The range defines square of four pixels around current pixel.
6159 Set direction in radians from which four pixel will be compared. If positive,
6160 random direction from 0 to set direction will be picked. If negative, exact of
6161 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6162 will pick only pixels on same row and -PI/2 will pick only pixels on same
6166 If enabled, current pixel is compared with average value of all four
6167 surrounding pixels. The default is enabled. If disabled current pixel is
6168 compared with all four surrounding pixels. The pixel is considered banded
6169 if only all four differences with surrounding pixels are less than threshold.
6175 Drop duplicated frames at regular intervals.
6177 The filter accepts the following options:
6181 Set the number of frames from which one will be dropped. Setting this to
6182 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6183 Default is @code{5}.
6186 Set the threshold for duplicate detection. If the difference metric for a frame
6187 is less than or equal to this value, then it is declared as duplicate. Default
6191 Set scene change threshold. Default is @code{15}.
6195 Set the size of the x and y-axis blocks used during metric calculations.
6196 Larger blocks give better noise suppression, but also give worse detection of
6197 small movements. Must be a power of two. Default is @code{32}.
6200 Mark main input as a pre-processed input and activate clean source input
6201 stream. This allows the input to be pre-processed with various filters to help
6202 the metrics calculation while keeping the frame selection lossless. When set to
6203 @code{1}, the first stream is for the pre-processed input, and the second
6204 stream is the clean source from where the kept frames are chosen. Default is
6208 Set whether or not chroma is considered in the metric calculations. Default is
6214 Apply deflate effect to the video.
6216 This filter replaces the pixel by the local(3x3) average by taking into account
6217 only values lower than the pixel.
6219 It accepts the following options:
6226 Limit the maximum change for each plane, default is 65535.
6227 If 0, plane will remain unchanged.
6232 Remove judder produced by partially interlaced telecined content.
6234 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6235 source was partially telecined content then the output of @code{pullup,dejudder}
6236 will have a variable frame rate. May change the recorded frame rate of the
6237 container. Aside from that change, this filter will not affect constant frame
6240 The option available in this filter is:
6244 Specify the length of the window over which the judder repeats.
6246 Accepts any integer greater than 1. Useful values are:
6250 If the original was telecined from 24 to 30 fps (Film to NTSC).
6253 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6256 If a mixture of the two.
6259 The default is @samp{4}.
6264 Suppress a TV station logo by a simple interpolation of the surrounding
6265 pixels. Just set a rectangle covering the logo and watch it disappear
6266 (and sometimes something even uglier appear - your mileage may vary).
6268 It accepts the following parameters:
6273 Specify the top left corner coordinates of the logo. They must be
6278 Specify the width and height of the logo to clear. They must be
6282 Specify the thickness of the fuzzy edge of the rectangle (added to
6283 @var{w} and @var{h}). The default value is 1. This option is
6284 deprecated, setting higher values should no longer be necessary and
6288 When set to 1, a green rectangle is drawn on the screen to simplify
6289 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6290 The default value is 0.
6292 The rectangle is drawn on the outermost pixels which will be (partly)
6293 replaced with interpolated values. The values of the next pixels
6294 immediately outside this rectangle in each direction will be used to
6295 compute the interpolated pixel values inside the rectangle.
6299 @subsection Examples
6303 Set a rectangle covering the area with top left corner coordinates 0,0
6304 and size 100x77, and a band of size 10:
6306 delogo=x=0:y=0:w=100:h=77:band=10
6313 Attempt to fix small changes in horizontal and/or vertical shift. This
6314 filter helps remove camera shake from hand-holding a camera, bumping a
6315 tripod, moving on a vehicle, etc.
6317 The filter accepts the following options:
6325 Specify a rectangular area where to limit the search for motion
6327 If desired the search for motion vectors can be limited to a
6328 rectangular area of the frame defined by its top left corner, width
6329 and height. These parameters have the same meaning as the drawbox
6330 filter which can be used to visualise the position of the bounding
6333 This is useful when simultaneous movement of subjects within the frame
6334 might be confused for camera motion by the motion vector search.
6336 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6337 then the full frame is used. This allows later options to be set
6338 without specifying the bounding box for the motion vector search.
6340 Default - search the whole frame.
6344 Specify the maximum extent of movement in x and y directions in the
6345 range 0-64 pixels. Default 16.
6348 Specify how to generate pixels to fill blanks at the edge of the
6349 frame. Available values are:
6352 Fill zeroes at blank locations
6354 Original image at blank locations
6356 Extruded edge value at blank locations
6358 Mirrored edge at blank locations
6360 Default value is @samp{mirror}.
6363 Specify the blocksize to use for motion search. Range 4-128 pixels,
6367 Specify the contrast threshold for blocks. Only blocks with more than
6368 the specified contrast (difference between darkest and lightest
6369 pixels) will be considered. Range 1-255, default 125.
6372 Specify the search strategy. Available values are:
6375 Set exhaustive search
6377 Set less exhaustive search.
6379 Default value is @samp{exhaustive}.
6382 If set then a detailed log of the motion search is written to the
6386 If set to 1, specify using OpenCL capabilities, only available if
6387 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6393 Apply an exact inverse of the telecine operation. It requires a predefined
6394 pattern specified using the pattern option which must be the same as that passed
6395 to the telecine filter.
6397 This filter accepts the following options:
6406 The default value is @code{top}.
6410 A string of numbers representing the pulldown pattern you wish to apply.
6411 The default value is @code{23}.
6414 A number representing position of the first frame with respect to the telecine
6415 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6420 Apply dilation effect to the video.
6422 This filter replaces the pixel by the local(3x3) maximum.
6424 It accepts the following options:
6431 Limit the maximum change for each plane, default is 65535.
6432 If 0, plane will remain unchanged.
6435 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6438 Flags to local 3x3 coordinates maps like this:
6447 Displace pixels as indicated by second and third input stream.
6449 It takes three input streams and outputs one stream, the first input is the
6450 source, and second and third input are displacement maps.
6452 The second input specifies how much to displace pixels along the
6453 x-axis, while the third input specifies how much to displace pixels
6455 If one of displacement map streams terminates, last frame from that
6456 displacement map will be used.
6458 Note that once generated, displacements maps can be reused over and over again.
6460 A description of the accepted options follows.
6464 Set displace behavior for pixels that are out of range.
6466 Available values are:
6469 Missing pixels are replaced by black pixels.
6472 Adjacent pixels will spread out to replace missing pixels.
6475 Out of range pixels are wrapped so they point to pixels of other side.
6477 Default is @samp{smear}.
6481 @subsection Examples
6485 Add ripple effect to rgb input of video size hd720:
6487 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
6491 Add wave effect to rgb input of video size hd720:
6493 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
6499 Draw a colored box on the input image.
6501 It accepts the following parameters:
6506 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6510 The expressions which specify the width and height of the box; if 0 they are interpreted as
6511 the input width and height. It defaults to 0.
6514 Specify the color of the box to write. For the general syntax of this option,
6515 check the "Color" section in the ffmpeg-utils manual. If the special
6516 value @code{invert} is used, the box edge color is the same as the
6517 video with inverted luma.
6520 The expression which sets the thickness of the box edge. Default value is @code{3}.
6522 See below for the list of accepted constants.
6525 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6526 following constants:
6530 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6534 horizontal and vertical chroma subsample values. For example for the
6535 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6539 The input width and height.
6542 The input sample aspect ratio.
6546 The x and y offset coordinates where the box is drawn.
6550 The width and height of the drawn box.
6553 The thickness of the drawn box.
6555 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6556 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6560 @subsection Examples
6564 Draw a black box around the edge of the input image:
6570 Draw a box with color red and an opacity of 50%:
6572 drawbox=10:20:200:60:red@@0.5
6575 The previous example can be specified as:
6577 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6581 Fill the box with pink color:
6583 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6587 Draw a 2-pixel red 2.40:1 mask:
6589 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
6595 Draw a grid on the input image.
6597 It accepts the following parameters:
6602 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6606 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6607 input width and height, respectively, minus @code{thickness}, so image gets
6608 framed. Default to 0.
6611 Specify the color of the grid. For the general syntax of this option,
6612 check the "Color" section in the ffmpeg-utils manual. If the special
6613 value @code{invert} is used, the grid color is the same as the
6614 video with inverted luma.
6617 The expression which sets the thickness of the grid line. Default value is @code{1}.
6619 See below for the list of accepted constants.
6622 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6623 following constants:
6627 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6631 horizontal and vertical chroma subsample values. For example for the
6632 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6636 The input grid cell width and height.
6639 The input sample aspect ratio.
6643 The x and y coordinates of some point of grid intersection (meant to configure offset).
6647 The width and height of the drawn cell.
6650 The thickness of the drawn cell.
6652 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6653 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6657 @subsection Examples
6661 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6663 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6667 Draw a white 3x3 grid with an opacity of 50%:
6669 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6676 Draw a text string or text from a specified file on top of a video, using the
6677 libfreetype library.
6679 To enable compilation of this filter, you need to configure FFmpeg with
6680 @code{--enable-libfreetype}.
6681 To enable default font fallback and the @var{font} option you need to
6682 configure FFmpeg with @code{--enable-libfontconfig}.
6683 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6684 @code{--enable-libfribidi}.
6688 It accepts the following parameters:
6693 Used to draw a box around text using the background color.
6694 The value must be either 1 (enable) or 0 (disable).
6695 The default value of @var{box} is 0.
6698 Set the width of the border to be drawn around the box using @var{boxcolor}.
6699 The default value of @var{boxborderw} is 0.
6702 The color to be used for drawing box around text. For the syntax of this
6703 option, check the "Color" section in the ffmpeg-utils manual.
6705 The default value of @var{boxcolor} is "white".
6708 Set the width of the border to be drawn around the text using @var{bordercolor}.
6709 The default value of @var{borderw} is 0.
6712 Set the color to be used for drawing border around text. For the syntax of this
6713 option, check the "Color" section in the ffmpeg-utils manual.
6715 The default value of @var{bordercolor} is "black".
6718 Select how the @var{text} is expanded. Can be either @code{none},
6719 @code{strftime} (deprecated) or
6720 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6724 If true, check and fix text coords to avoid clipping.
6727 The color to be used for drawing fonts. For the syntax of this option, check
6728 the "Color" section in the ffmpeg-utils manual.
6730 The default value of @var{fontcolor} is "black".
6732 @item fontcolor_expr
6733 String which is expanded the same way as @var{text} to obtain dynamic
6734 @var{fontcolor} value. By default this option has empty value and is not
6735 processed. When this option is set, it overrides @var{fontcolor} option.
6738 The font family to be used for drawing text. By default Sans.
6741 The font file to be used for drawing text. The path must be included.
6742 This parameter is mandatory if the fontconfig support is disabled.
6745 This option does not exist, please see the timeline system
6748 Draw the text applying alpha blending. The value can
6749 be a number between 0.0 and 1.0.
6750 The expression accepts the same variables @var{x, y} as well.
6751 The default value is 1.
6752 Please see @var{fontcolor_expr}.
6755 The font size to be used for drawing text.
6756 The default value of @var{fontsize} is 16.
6759 If set to 1, attempt to shape the text (for example, reverse the order of
6760 right-to-left text and join Arabic characters) before drawing it.
6761 Otherwise, just draw the text exactly as given.
6762 By default 1 (if supported).
6765 The flags to be used for loading the fonts.
6767 The flags map the corresponding flags supported by libfreetype, and are
6768 a combination of the following values:
6775 @item vertical_layout
6776 @item force_autohint
6779 @item ignore_global_advance_width
6781 @item ignore_transform
6787 Default value is "default".
6789 For more information consult the documentation for the FT_LOAD_*
6793 The color to be used for drawing a shadow behind the drawn text. For the
6794 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6796 The default value of @var{shadowcolor} is "black".
6800 The x and y offsets for the text shadow position with respect to the
6801 position of the text. They can be either positive or negative
6802 values. The default value for both is "0".
6805 The starting frame number for the n/frame_num variable. The default value
6809 The size in number of spaces to use for rendering the tab.
6813 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6814 format. It can be used with or without text parameter. @var{timecode_rate}
6815 option must be specified.
6817 @item timecode_rate, rate, r
6818 Set the timecode frame rate (timecode only).
6821 The text string to be drawn. The text must be a sequence of UTF-8
6823 This parameter is mandatory if no file is specified with the parameter
6827 A text file containing text to be drawn. The text must be a sequence
6828 of UTF-8 encoded characters.
6830 This parameter is mandatory if no text string is specified with the
6831 parameter @var{text}.
6833 If both @var{text} and @var{textfile} are specified, an error is thrown.
6836 If set to 1, the @var{textfile} will be reloaded before each frame.
6837 Be sure to update it atomically, or it may be read partially, or even fail.
6841 The expressions which specify the offsets where text will be drawn
6842 within the video frame. They are relative to the top/left border of the
6845 The default value of @var{x} and @var{y} is "0".
6847 See below for the list of accepted constants and functions.
6850 The parameters for @var{x} and @var{y} are expressions containing the
6851 following constants and functions:
6855 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
6859 horizontal and vertical chroma subsample values. For example for the
6860 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6863 the height of each text line
6871 @item max_glyph_a, ascent
6872 the maximum distance from the baseline to the highest/upper grid
6873 coordinate used to place a glyph outline point, for all the rendered
6875 It is a positive value, due to the grid's orientation with the Y axis
6878 @item max_glyph_d, descent
6879 the maximum distance from the baseline to the lowest grid coordinate
6880 used to place a glyph outline point, for all the rendered glyphs.
6881 This is a negative value, due to the grid's orientation, with the Y axis
6885 maximum glyph height, that is the maximum height for all the glyphs
6886 contained in the rendered text, it is equivalent to @var{ascent} -
6890 maximum glyph width, that is the maximum width for all the glyphs
6891 contained in the rendered text
6894 the number of input frame, starting from 0
6896 @item rand(min, max)
6897 return a random number included between @var{min} and @var{max}
6900 The input sample aspect ratio.
6903 timestamp expressed in seconds, NAN if the input timestamp is unknown
6906 the height of the rendered text
6909 the width of the rendered text
6913 the x and y offset coordinates where the text is drawn.
6915 These parameters allow the @var{x} and @var{y} expressions to refer
6916 each other, so you can for example specify @code{y=x/dar}.
6919 @anchor{drawtext_expansion}
6920 @subsection Text expansion
6922 If @option{expansion} is set to @code{strftime},
6923 the filter recognizes strftime() sequences in the provided text and
6924 expands them accordingly. Check the documentation of strftime(). This
6925 feature is deprecated.
6927 If @option{expansion} is set to @code{none}, the text is printed verbatim.
6929 If @option{expansion} is set to @code{normal} (which is the default),
6930 the following expansion mechanism is used.
6932 The backslash character @samp{\}, followed by any character, always expands to
6933 the second character.
6935 Sequences of the form @code{%@{...@}} are expanded. The text between the
6936 braces is a function name, possibly followed by arguments separated by ':'.
6937 If the arguments contain special characters or delimiters (':' or '@}'),
6938 they should be escaped.
6940 Note that they probably must also be escaped as the value for the
6941 @option{text} option in the filter argument string and as the filter
6942 argument in the filtergraph description, and possibly also for the shell,
6943 that makes up to four levels of escaping; using a text file avoids these
6946 The following functions are available:
6951 The expression evaluation result.
6953 It must take one argument specifying the expression to be evaluated,
6954 which accepts the same constants and functions as the @var{x} and
6955 @var{y} values. Note that not all constants should be used, for
6956 example the text size is not known when evaluating the expression, so
6957 the constants @var{text_w} and @var{text_h} will have an undefined
6960 @item expr_int_format, eif
6961 Evaluate the expression's value and output as formatted integer.
6963 The first argument is the expression to be evaluated, just as for the @var{expr} function.
6964 The second argument specifies the output format. Allowed values are @samp{x},
6965 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
6966 @code{printf} function.
6967 The third parameter is optional and sets the number of positions taken by the output.
6968 It can be used to add padding with zeros from the left.
6971 The time at which the filter is running, expressed in UTC.
6972 It can accept an argument: a strftime() format string.
6975 The time at which the filter is running, expressed in the local time zone.
6976 It can accept an argument: a strftime() format string.
6979 Frame metadata. Takes one or two arguments.
6981 The first argument is mandatory and specifies the metadata key.
6983 The second argument is optional and specifies a default value, used when the
6984 metadata key is not found or empty.
6987 The frame number, starting from 0.
6990 A 1 character description of the current picture type.
6993 The timestamp of the current frame.
6994 It can take up to three arguments.
6996 The first argument is the format of the timestamp; it defaults to @code{flt}
6997 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
6998 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
6999 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7000 @code{localtime} stands for the timestamp of the frame formatted as
7001 local time zone time.
7003 The second argument is an offset added to the timestamp.
7005 If the format is set to @code{localtime} or @code{gmtime},
7006 a third argument may be supplied: a strftime() format string.
7007 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7010 @subsection Examples
7014 Draw "Test Text" with font FreeSerif, using the default values for the
7015 optional parameters.
7018 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7022 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7023 and y=50 (counting from the top-left corner of the screen), text is
7024 yellow with a red box around it. Both the text and the box have an
7028 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7029 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7032 Note that the double quotes are not necessary if spaces are not used
7033 within the parameter list.
7036 Show the text at the center of the video frame:
7038 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7042 Show the text at a random position, switching to a new position every 30 seconds:
7044 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)"
7048 Show a text line sliding from right to left in the last row of the video
7049 frame. The file @file{LONG_LINE} is assumed to contain a single line
7052 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7056 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7058 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7062 Draw a single green letter "g", at the center of the input video.
7063 The glyph baseline is placed at half screen height.
7065 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7069 Show text for 1 second every 3 seconds:
7071 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7075 Use fontconfig to set the font. Note that the colons need to be escaped.
7077 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7081 Print the date of a real-time encoding (see strftime(3)):
7083 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7087 Show text fading in and out (appearing/disappearing):
7090 DS=1.0 # display start
7091 DE=10.0 # display end
7092 FID=1.5 # fade in duration
7093 FOD=5 # fade out duration
7094 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 @}"
7099 For more information about libfreetype, check:
7100 @url{http://www.freetype.org/}.
7102 For more information about fontconfig, check:
7103 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7105 For more information about libfribidi, check:
7106 @url{http://fribidi.org/}.
7110 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7112 The filter accepts the following options:
7117 Set low and high threshold values used by the Canny thresholding
7120 The high threshold selects the "strong" edge pixels, which are then
7121 connected through 8-connectivity with the "weak" edge pixels selected
7122 by the low threshold.
7124 @var{low} and @var{high} threshold values must be chosen in the range
7125 [0,1], and @var{low} should be lesser or equal to @var{high}.
7127 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7131 Define the drawing mode.
7135 Draw white/gray wires on black background.
7138 Mix the colors to create a paint/cartoon effect.
7141 Default value is @var{wires}.
7144 @subsection Examples
7148 Standard edge detection with custom values for the hysteresis thresholding:
7150 edgedetect=low=0.1:high=0.4
7154 Painting effect without thresholding:
7156 edgedetect=mode=colormix:high=0
7161 Set brightness, contrast, saturation and approximate gamma adjustment.
7163 The filter accepts the following options:
7167 Set the contrast expression. The value must be a float value in range
7168 @code{-2.0} to @code{2.0}. The default value is "1".
7171 Set the brightness expression. The value must be a float value in
7172 range @code{-1.0} to @code{1.0}. The default value is "0".
7175 Set the saturation expression. The value must be a float in
7176 range @code{0.0} to @code{3.0}. The default value is "1".
7179 Set the gamma expression. The value must be a float in range
7180 @code{0.1} to @code{10.0}. The default value is "1".
7183 Set the gamma expression for red. The value must be a float in
7184 range @code{0.1} to @code{10.0}. The default value is "1".
7187 Set the gamma expression for green. The value must be a float in range
7188 @code{0.1} to @code{10.0}. The default value is "1".
7191 Set the gamma expression for blue. The value must be a float in range
7192 @code{0.1} to @code{10.0}. The default value is "1".
7195 Set the gamma weight expression. It can be used to reduce the effect
7196 of a high gamma value on bright image areas, e.g. keep them from
7197 getting overamplified and just plain white. The value must be a float
7198 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7199 gamma correction all the way down while @code{1.0} leaves it at its
7200 full strength. Default is "1".
7203 Set when the expressions for brightness, contrast, saturation and
7204 gamma expressions are evaluated.
7206 It accepts the following values:
7209 only evaluate expressions once during the filter initialization or
7210 when a command is processed
7213 evaluate expressions for each incoming frame
7216 Default value is @samp{init}.
7219 The expressions accept the following parameters:
7222 frame count of the input frame starting from 0
7225 byte position of the corresponding packet in the input file, NAN if
7229 frame rate of the input video, NAN if the input frame rate is unknown
7232 timestamp expressed in seconds, NAN if the input timestamp is unknown
7235 @subsection Commands
7236 The filter supports the following commands:
7240 Set the contrast expression.
7243 Set the brightness expression.
7246 Set the saturation expression.
7249 Set the gamma expression.
7252 Set the gamma_r expression.
7255 Set gamma_g expression.
7258 Set gamma_b expression.
7261 Set gamma_weight expression.
7263 The command accepts the same syntax of the corresponding option.
7265 If the specified expression is not valid, it is kept at its current
7272 Apply erosion effect to the video.
7274 This filter replaces the pixel by the local(3x3) minimum.
7276 It accepts the following options:
7283 Limit the maximum change for each plane, default is 65535.
7284 If 0, plane will remain unchanged.
7287 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7290 Flags to local 3x3 coordinates maps like this:
7297 @section extractplanes
7299 Extract color channel components from input video stream into
7300 separate grayscale video streams.
7302 The filter accepts the following option:
7306 Set plane(s) to extract.
7308 Available values for planes are:
7319 Choosing planes not available in the input will result in an error.
7320 That means you cannot select @code{r}, @code{g}, @code{b} planes
7321 with @code{y}, @code{u}, @code{v} planes at same time.
7324 @subsection Examples
7328 Extract luma, u and v color channel component from input video frame
7329 into 3 grayscale outputs:
7331 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
7337 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7339 For each input image, the filter will compute the optimal mapping from
7340 the input to the output given the codebook length, that is the number
7341 of distinct output colors.
7343 This filter accepts the following options.
7346 @item codebook_length, l
7347 Set codebook length. The value must be a positive integer, and
7348 represents the number of distinct output colors. Default value is 256.
7351 Set the maximum number of iterations to apply for computing the optimal
7352 mapping. The higher the value the better the result and the higher the
7353 computation time. Default value is 1.
7356 Set a random seed, must be an integer included between 0 and
7357 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7358 will try to use a good random seed on a best effort basis.
7361 Set pal8 output pixel format. This option does not work with codebook
7362 length greater than 256.
7367 Apply a fade-in/out effect to the input video.
7369 It accepts the following parameters:
7373 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7375 Default is @code{in}.
7377 @item start_frame, s
7378 Specify the number of the frame to start applying the fade
7379 effect at. Default is 0.
7382 The number of frames that the fade effect lasts. At the end of the
7383 fade-in effect, the output video will have the same intensity as the input video.
7384 At the end of the fade-out transition, the output video will be filled with the
7385 selected @option{color}.
7389 If set to 1, fade only alpha channel, if one exists on the input.
7392 @item start_time, st
7393 Specify the timestamp (in seconds) of the frame to start to apply the fade
7394 effect. If both start_frame and start_time are specified, the fade will start at
7395 whichever comes last. Default is 0.
7398 The number of seconds for which the fade effect has to last. At the end of the
7399 fade-in effect the output video will have the same intensity as the input video,
7400 at the end of the fade-out transition the output video will be filled with the
7401 selected @option{color}.
7402 If both duration and nb_frames are specified, duration is used. Default is 0
7403 (nb_frames is used by default).
7406 Specify the color of the fade. Default is "black".
7409 @subsection Examples
7413 Fade in the first 30 frames of video:
7418 The command above is equivalent to:
7424 Fade out the last 45 frames of a 200-frame video:
7427 fade=type=out:start_frame=155:nb_frames=45
7431 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7433 fade=in:0:25, fade=out:975:25
7437 Make the first 5 frames yellow, then fade in from frame 5-24:
7439 fade=in:5:20:color=yellow
7443 Fade in alpha over first 25 frames of video:
7445 fade=in:0:25:alpha=1
7449 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7451 fade=t=in:st=5.5:d=0.5
7457 Apply arbitrary expressions to samples in frequency domain
7461 Adjust the dc value (gain) of the luma plane of the image. The filter
7462 accepts an integer value in range @code{0} to @code{1000}. The default
7463 value is set to @code{0}.
7466 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7467 filter accepts an integer value in range @code{0} to @code{1000}. The
7468 default value is set to @code{0}.
7471 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7472 filter accepts an integer value in range @code{0} to @code{1000}. The
7473 default value is set to @code{0}.
7476 Set the frequency domain weight expression for the luma plane.
7479 Set the frequency domain weight expression for the 1st chroma plane.
7482 Set the frequency domain weight expression for the 2nd chroma plane.
7484 The filter accepts the following variables:
7487 The coordinates of the current sample.
7491 The width and height of the image.
7494 @subsection Examples
7500 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7506 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7512 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7518 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7525 Extract a single field from an interlaced image using stride
7526 arithmetic to avoid wasting CPU time. The output frames are marked as
7529 The filter accepts the following options:
7533 Specify whether to extract the top (if the value is @code{0} or
7534 @code{top}) or the bottom field (if the value is @code{1} or
7540 Create new frames by copying the top and bottom fields from surrounding frames
7541 supplied as numbers by the hint file.
7545 Set file containing hints: absolute/relative frame numbers.
7547 There must be one line for each frame in a clip. Each line must contain two
7548 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7549 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7550 is current frame number for @code{absolute} mode or out of [-1, 1] range
7551 for @code{relative} mode. First number tells from which frame to pick up top
7552 field and second number tells from which frame to pick up bottom field.
7554 If optionally followed by @code{+} output frame will be marked as interlaced,
7555 else if followed by @code{-} output frame will be marked as progressive, else
7556 it will be marked same as input frame.
7557 If line starts with @code{#} or @code{;} that line is skipped.
7560 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7563 Example of first several lines of @code{hint} file for @code{relative} mode:
7566 1,0 - # second frame, use third's frame top field and second's frame bottom field
7567 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7584 Field matching filter for inverse telecine. It is meant to reconstruct the
7585 progressive frames from a telecined stream. The filter does not drop duplicated
7586 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7587 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7589 The separation of the field matching and the decimation is notably motivated by
7590 the possibility of inserting a de-interlacing filter fallback between the two.
7591 If the source has mixed telecined and real interlaced content,
7592 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7593 But these remaining combed frames will be marked as interlaced, and thus can be
7594 de-interlaced by a later filter such as @ref{yadif} before decimation.
7596 In addition to the various configuration options, @code{fieldmatch} can take an
7597 optional second stream, activated through the @option{ppsrc} option. If
7598 enabled, the frames reconstruction will be based on the fields and frames from
7599 this second stream. This allows the first input to be pre-processed in order to
7600 help the various algorithms of the filter, while keeping the output lossless
7601 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7602 or brightness/contrast adjustments can help.
7604 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7605 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7606 which @code{fieldmatch} is based on. While the semantic and usage are very
7607 close, some behaviour and options names can differ.
7609 The @ref{decimate} filter currently only works for constant frame rate input.
7610 If your input has mixed telecined (30fps) and progressive content with a lower
7611 framerate like 24fps use the following filterchain to produce the necessary cfr
7612 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7614 The filter accepts the following options:
7618 Specify the assumed field order of the input stream. Available values are:
7622 Auto detect parity (use FFmpeg's internal parity value).
7624 Assume bottom field first.
7626 Assume top field first.
7629 Note that it is sometimes recommended not to trust the parity announced by the
7632 Default value is @var{auto}.
7635 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7636 sense that it won't risk creating jerkiness due to duplicate frames when
7637 possible, but if there are bad edits or blended fields it will end up
7638 outputting combed frames when a good match might actually exist. On the other
7639 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7640 but will almost always find a good frame if there is one. The other values are
7641 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7642 jerkiness and creating duplicate frames versus finding good matches in sections
7643 with bad edits, orphaned fields, blended fields, etc.
7645 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7647 Available values are:
7651 2-way matching (p/c)
7653 2-way matching, and trying 3rd match if still combed (p/c + n)
7655 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7657 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7658 still combed (p/c + n + u/b)
7660 3-way matching (p/c/n)
7662 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7663 detected as combed (p/c/n + u/b)
7666 The parenthesis at the end indicate the matches that would be used for that
7667 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7670 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7673 Default value is @var{pc_n}.
7676 Mark the main input stream as a pre-processed input, and enable the secondary
7677 input stream as the clean source to pick the fields from. See the filter
7678 introduction for more details. It is similar to the @option{clip2} feature from
7681 Default value is @code{0} (disabled).
7684 Set the field to match from. It is recommended to set this to the same value as
7685 @option{order} unless you experience matching failures with that setting. In
7686 certain circumstances changing the field that is used to match from can have a
7687 large impact on matching performance. Available values are:
7691 Automatic (same value as @option{order}).
7693 Match from the bottom field.
7695 Match from the top field.
7698 Default value is @var{auto}.
7701 Set whether or not chroma is included during the match comparisons. In most
7702 cases it is recommended to leave this enabled. You should set this to @code{0}
7703 only if your clip has bad chroma problems such as heavy rainbowing or other
7704 artifacts. Setting this to @code{0} could also be used to speed things up at
7705 the cost of some accuracy.
7707 Default value is @code{1}.
7711 These define an exclusion band which excludes the lines between @option{y0} and
7712 @option{y1} from being included in the field matching decision. An exclusion
7713 band can be used to ignore subtitles, a logo, or other things that may
7714 interfere with the matching. @option{y0} sets the starting scan line and
7715 @option{y1} sets the ending line; all lines in between @option{y0} and
7716 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7717 @option{y0} and @option{y1} to the same value will disable the feature.
7718 @option{y0} and @option{y1} defaults to @code{0}.
7721 Set the scene change detection threshold as a percentage of maximum change on
7722 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7723 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7724 @option{scthresh} is @code{[0.0, 100.0]}.
7726 Default value is @code{12.0}.
7729 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7730 account the combed scores of matches when deciding what match to use as the
7731 final match. Available values are:
7735 No final matching based on combed scores.
7737 Combed scores are only used when a scene change is detected.
7739 Use combed scores all the time.
7742 Default is @var{sc}.
7745 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7746 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7747 Available values are:
7751 No forced calculation.
7753 Force p/c/n calculations.
7755 Force p/c/n/u/b calculations.
7758 Default value is @var{none}.
7761 This is the area combing threshold used for combed frame detection. This
7762 essentially controls how "strong" or "visible" combing must be to be detected.
7763 Larger values mean combing must be more visible and smaller values mean combing
7764 can be less visible or strong and still be detected. Valid settings are from
7765 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7766 be detected as combed). This is basically a pixel difference value. A good
7767 range is @code{[8, 12]}.
7769 Default value is @code{9}.
7772 Sets whether or not chroma is considered in the combed frame decision. Only
7773 disable this if your source has chroma problems (rainbowing, etc.) that are
7774 causing problems for the combed frame detection with chroma enabled. Actually,
7775 using @option{chroma}=@var{0} is usually more reliable, except for the case
7776 where there is chroma only combing in the source.
7778 Default value is @code{0}.
7782 Respectively set the x-axis and y-axis size of the window used during combed
7783 frame detection. This has to do with the size of the area in which
7784 @option{combpel} pixels are required to be detected as combed for a frame to be
7785 declared combed. See the @option{combpel} parameter description for more info.
7786 Possible values are any number that is a power of 2 starting at 4 and going up
7789 Default value is @code{16}.
7792 The number of combed pixels inside any of the @option{blocky} by
7793 @option{blockx} size blocks on the frame for the frame to be detected as
7794 combed. While @option{cthresh} controls how "visible" the combing must be, this
7795 setting controls "how much" combing there must be in any localized area (a
7796 window defined by the @option{blockx} and @option{blocky} settings) on the
7797 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7798 which point no frames will ever be detected as combed). This setting is known
7799 as @option{MI} in TFM/VFM vocabulary.
7801 Default value is @code{80}.
7804 @anchor{p/c/n/u/b meaning}
7805 @subsection p/c/n/u/b meaning
7807 @subsubsection p/c/n
7809 We assume the following telecined stream:
7812 Top fields: 1 2 2 3 4
7813 Bottom fields: 1 2 3 4 4
7816 The numbers correspond to the progressive frame the fields relate to. Here, the
7817 first two frames are progressive, the 3rd and 4th are combed, and so on.
7819 When @code{fieldmatch} is configured to run a matching from bottom
7820 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7825 B 1 2 3 4 4 <-- matching reference
7834 As a result of the field matching, we can see that some frames get duplicated.
7835 To perform a complete inverse telecine, you need to rely on a decimation filter
7836 after this operation. See for instance the @ref{decimate} filter.
7838 The same operation now matching from top fields (@option{field}=@var{top})
7843 T 1 2 2 3 4 <-- matching reference
7853 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
7854 basically, they refer to the frame and field of the opposite parity:
7857 @item @var{p} matches the field of the opposite parity in the previous frame
7858 @item @var{c} matches the field of the opposite parity in the current frame
7859 @item @var{n} matches the field of the opposite parity in the next frame
7864 The @var{u} and @var{b} matching are a bit special in the sense that they match
7865 from the opposite parity flag. In the following examples, we assume that we are
7866 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
7867 'x' is placed above and below each matched fields.
7869 With bottom matching (@option{field}=@var{bottom}):
7874 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7875 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7883 With top matching (@option{field}=@var{top}):
7888 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
7889 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
7897 @subsection Examples
7899 Simple IVTC of a top field first telecined stream:
7901 fieldmatch=order=tff:combmatch=none, decimate
7904 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
7906 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
7911 Transform the field order of the input video.
7913 It accepts the following parameters:
7918 The output field order. Valid values are @var{tff} for top field first or @var{bff}
7919 for bottom field first.
7922 The default value is @samp{tff}.
7924 The transformation is done by shifting the picture content up or down
7925 by one line, and filling the remaining line with appropriate picture content.
7926 This method is consistent with most broadcast field order converters.
7928 If the input video is not flagged as being interlaced, or it is already
7929 flagged as being of the required output field order, then this filter does
7930 not alter the incoming video.
7932 It is very useful when converting to or from PAL DV material,
7933 which is bottom field first.
7937 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
7940 @section fifo, afifo
7942 Buffer input images and send them when they are requested.
7944 It is mainly useful when auto-inserted by the libavfilter
7947 It does not take parameters.
7951 Find a rectangular object
7953 It accepts the following options:
7957 Filepath of the object image, needs to be in gray8.
7960 Detection threshold, default is 0.5.
7963 Number of mipmaps, default is 3.
7965 @item xmin, ymin, xmax, ymax
7966 Specifies the rectangle in which to search.
7969 @subsection Examples
7973 Generate a representative palette of a given video using @command{ffmpeg}:
7975 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
7981 Cover a rectangular object
7983 It accepts the following options:
7987 Filepath of the optional cover image, needs to be in yuv420.
7992 It accepts the following values:
7995 cover it by the supplied image
7997 cover it by interpolating the surrounding pixels
8000 Default value is @var{blur}.
8003 @subsection Examples
8007 Generate a representative palette of a given video using @command{ffmpeg}:
8009 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8016 Convert the input video to one of the specified pixel formats.
8017 Libavfilter will try to pick one that is suitable as input to
8020 It accepts the following parameters:
8024 A '|'-separated list of pixel format names, such as
8025 "pix_fmts=yuv420p|monow|rgb24".
8029 @subsection Examples
8033 Convert the input video to the @var{yuv420p} format
8035 format=pix_fmts=yuv420p
8038 Convert the input video to any of the formats in the list
8040 format=pix_fmts=yuv420p|yuv444p|yuv410p
8047 Convert the video to specified constant frame rate by duplicating or dropping
8048 frames as necessary.
8050 It accepts the following parameters:
8054 The desired output frame rate. The default is @code{25}.
8059 Possible values are:
8062 zero round towards 0
8066 round towards -infinity
8068 round towards +infinity
8072 The default is @code{near}.
8075 Assume the first PTS should be the given value, in seconds. This allows for
8076 padding/trimming at the start of stream. By default, no assumption is made
8077 about the first frame's expected PTS, so no padding or trimming is done.
8078 For example, this could be set to 0 to pad the beginning with duplicates of
8079 the first frame if a video stream starts after the audio stream or to trim any
8080 frames with a negative PTS.
8084 Alternatively, the options can be specified as a flat string:
8085 @var{fps}[:@var{round}].
8087 See also the @ref{setpts} filter.
8089 @subsection Examples
8093 A typical usage in order to set the fps to 25:
8099 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8101 fps=fps=film:round=near
8107 Pack two different video streams into a stereoscopic video, setting proper
8108 metadata on supported codecs. The two views should have the same size and
8109 framerate and processing will stop when the shorter video ends. Please note
8110 that you may conveniently adjust view properties with the @ref{scale} and
8113 It accepts the following parameters:
8117 The desired packing format. Supported values are:
8122 The views are next to each other (default).
8125 The views are on top of each other.
8128 The views are packed by line.
8131 The views are packed by column.
8134 The views are temporally interleaved.
8143 # Convert left and right views into a frame-sequential video
8144 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8146 # Convert views into a side-by-side video with the same output resolution as the input
8147 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
8152 Change the frame rate by interpolating new video output frames from the source
8155 This filter is not designed to function correctly with interlaced media. If
8156 you wish to change the frame rate of interlaced media then you are required
8157 to deinterlace before this filter and re-interlace after this filter.
8159 A description of the accepted options follows.
8163 Specify the output frames per second. This option can also be specified
8164 as a value alone. The default is @code{50}.
8167 Specify the start of a range where the output frame will be created as a
8168 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8169 the default is @code{15}.
8172 Specify the end of a range where the output frame will be created as a
8173 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8174 the default is @code{240}.
8177 Specify the level at which a scene change is detected as a value between
8178 0 and 100 to indicate a new scene; a low value reflects a low
8179 probability for the current frame to introduce a new scene, while a higher
8180 value means the current frame is more likely to be one.
8181 The default is @code{7}.
8184 Specify flags influencing the filter process.
8186 Available value for @var{flags} is:
8189 @item scene_change_detect, scd
8190 Enable scene change detection using the value of the option @var{scene}.
8191 This flag is enabled by default.
8197 Select one frame every N-th frame.
8199 This filter accepts the following option:
8202 Select frame after every @code{step} frames.
8203 Allowed values are positive integers higher than 0. Default value is @code{1}.
8209 Apply a frei0r effect to the input video.
8211 To enable the compilation of this filter, you need to install the frei0r
8212 header and configure FFmpeg with @code{--enable-frei0r}.
8214 It accepts the following parameters:
8219 The name of the frei0r effect to load. If the environment variable
8220 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8221 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8222 Otherwise, the standard frei0r paths are searched, in this order:
8223 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8224 @file{/usr/lib/frei0r-1/}.
8227 A '|'-separated list of parameters to pass to the frei0r effect.
8231 A frei0r effect parameter can be a boolean (its value is either
8232 "y" or "n"), a double, a color (specified as
8233 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8234 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8235 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8236 @var{X} and @var{Y} are floating point numbers) and/or a string.
8238 The number and types of parameters depend on the loaded effect. If an
8239 effect parameter is not specified, the default value is set.
8241 @subsection Examples
8245 Apply the distort0r effect, setting the first two double parameters:
8247 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8251 Apply the colordistance effect, taking a color as the first parameter:
8253 frei0r=colordistance:0.2/0.3/0.4
8254 frei0r=colordistance:violet
8255 frei0r=colordistance:0x112233
8259 Apply the perspective effect, specifying the top left and top right image
8262 frei0r=perspective:0.2/0.2|0.8/0.2
8266 For more information, see
8267 @url{http://frei0r.dyne.org}
8271 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8273 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8274 processing filter, one of them is performed once per block, not per pixel.
8275 This allows for much higher speed.
8277 The filter accepts the following options:
8281 Set quality. This option defines the number of levels for averaging. It accepts
8282 an integer in the range 4-5. Default value is @code{4}.
8285 Force a constant quantization parameter. It accepts an integer in range 0-63.
8286 If not set, the filter will use the QP from the video stream (if available).
8289 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8290 more details but also more artifacts, while higher values make the image smoother
8291 but also blurrier. Default value is @code{0} − PSNR optimal.
8294 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8295 option may cause flicker since the B-Frames have often larger QP. Default is
8296 @code{0} (not enabled).
8302 Apply Gaussian blur filter.
8304 The filter accepts the following options:
8308 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8311 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8314 Set which planes to filter. By default all planes are filtered.
8317 Set vertical sigma, if negative it will be same as @code{sigma}.
8318 Default is @code{-1}.
8323 The filter accepts the following options:
8327 Set the luminance expression.
8329 Set the chrominance blue expression.
8331 Set the chrominance red expression.
8333 Set the alpha expression.
8335 Set the red expression.
8337 Set the green expression.
8339 Set the blue expression.
8342 The colorspace is selected according to the specified options. If one
8343 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8344 options is specified, the filter will automatically select a YCbCr
8345 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8346 @option{blue_expr} options is specified, it will select an RGB
8349 If one of the chrominance expression is not defined, it falls back on the other
8350 one. If no alpha expression is specified it will evaluate to opaque value.
8351 If none of chrominance expressions are specified, they will evaluate
8352 to the luminance expression.
8354 The expressions can use the following variables and functions:
8358 The sequential number of the filtered frame, starting from @code{0}.
8362 The coordinates of the current sample.
8366 The width and height of the image.
8370 Width and height scale depending on the currently filtered plane. It is the
8371 ratio between the corresponding luma plane number of pixels and the current
8372 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8373 @code{0.5,0.5} for chroma planes.
8376 Time of the current frame, expressed in seconds.
8379 Return the value of the pixel at location (@var{x},@var{y}) of the current
8383 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8387 Return the value of the pixel at location (@var{x},@var{y}) of the
8388 blue-difference chroma plane. Return 0 if there is no such plane.
8391 Return the value of the pixel at location (@var{x},@var{y}) of the
8392 red-difference chroma plane. Return 0 if there is no such plane.
8397 Return the value of the pixel at location (@var{x},@var{y}) of the
8398 red/green/blue component. Return 0 if there is no such component.
8401 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8402 plane. Return 0 if there is no such plane.
8405 For functions, if @var{x} and @var{y} are outside the area, the value will be
8406 automatically clipped to the closer edge.
8408 @subsection Examples
8412 Flip the image horizontally:
8418 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8419 wavelength of 100 pixels:
8421 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8425 Generate a fancy enigmatic moving light:
8427 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
8431 Generate a quick emboss effect:
8433 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8437 Modify RGB components depending on pixel position:
8439 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8443 Create a radial gradient that is the same size as the input (also see
8444 the @ref{vignette} filter):
8446 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8452 Fix the banding artifacts that are sometimes introduced into nearly flat
8453 regions by truncation to 8-bit color depth.
8454 Interpolate the gradients that should go where the bands are, and
8457 It is designed for playback only. Do not use it prior to
8458 lossy compression, because compression tends to lose the dither and
8459 bring back the bands.
8461 It accepts the following parameters:
8466 The maximum amount by which the filter will change any one pixel. This is also
8467 the threshold for detecting nearly flat regions. Acceptable values range from
8468 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8472 The neighborhood to fit the gradient to. A larger radius makes for smoother
8473 gradients, but also prevents the filter from modifying the pixels near detailed
8474 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8475 values will be clipped to the valid range.
8479 Alternatively, the options can be specified as a flat string:
8480 @var{strength}[:@var{radius}]
8482 @subsection Examples
8486 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8492 Specify radius, omitting the strength (which will fall-back to the default
8503 Apply a Hald CLUT to a video stream.
8505 First input is the video stream to process, and second one is the Hald CLUT.
8506 The Hald CLUT input can be a simple picture or a complete video stream.
8508 The filter accepts the following options:
8512 Force termination when the shortest input terminates. Default is @code{0}.
8514 Continue applying the last CLUT after the end of the stream. A value of
8515 @code{0} disable the filter after the last frame of the CLUT is reached.
8516 Default is @code{1}.
8519 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8520 filters share the same internals).
8522 More information about the Hald CLUT can be found on Eskil Steenberg's website
8523 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8525 @subsection Workflow examples
8527 @subsubsection Hald CLUT video stream
8529 Generate an identity Hald CLUT stream altered with various effects:
8531 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
8534 Note: make sure you use a lossless codec.
8536 Then use it with @code{haldclut} to apply it on some random stream:
8538 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8541 The Hald CLUT will be applied to the 10 first seconds (duration of
8542 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8543 to the remaining frames of the @code{mandelbrot} stream.
8545 @subsubsection Hald CLUT with preview
8547 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8548 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8549 biggest possible square starting at the top left of the picture. The remaining
8550 padding pixels (bottom or right) will be ignored. This area can be used to add
8551 a preview of the Hald CLUT.
8553 Typically, the following generated Hald CLUT will be supported by the
8554 @code{haldclut} filter:
8557 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8558 pad=iw+320 [padded_clut];
8559 smptebars=s=320x256, split [a][b];
8560 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8561 [main][b] overlay=W-320" -frames:v 1 clut.png
8564 It contains the original and a preview of the effect of the CLUT: SMPTE color
8565 bars are displayed on the right-top, and below the same color bars processed by
8568 Then, the effect of this Hald CLUT can be visualized with:
8570 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8575 Flip the input video horizontally.
8577 For example, to horizontally flip the input video with @command{ffmpeg}:
8579 ffmpeg -i in.avi -vf "hflip" out.avi
8583 This filter applies a global color histogram equalization on a
8586 It can be used to correct video that has a compressed range of pixel
8587 intensities. The filter redistributes the pixel intensities to
8588 equalize their distribution across the intensity range. It may be
8589 viewed as an "automatically adjusting contrast filter". This filter is
8590 useful only for correcting degraded or poorly captured source
8593 The filter accepts the following options:
8597 Determine the amount of equalization to be applied. As the strength
8598 is reduced, the distribution of pixel intensities more-and-more
8599 approaches that of the input frame. The value must be a float number
8600 in the range [0,1] and defaults to 0.200.
8603 Set the maximum intensity that can generated and scale the output
8604 values appropriately. The strength should be set as desired and then
8605 the intensity can be limited if needed to avoid washing-out. The value
8606 must be a float number in the range [0,1] and defaults to 0.210.
8609 Set the antibanding level. If enabled the filter will randomly vary
8610 the luminance of output pixels by a small amount to avoid banding of
8611 the histogram. Possible values are @code{none}, @code{weak} or
8612 @code{strong}. It defaults to @code{none}.
8617 Compute and draw a color distribution histogram for the input video.
8619 The computed histogram is a representation of the color component
8620 distribution in an image.
8622 Standard histogram displays the color components distribution in an image.
8623 Displays color graph for each color component. Shows distribution of
8624 the Y, U, V, A or R, G, B components, depending on input format, in the
8625 current frame. Below each graph a color component scale meter is shown.
8627 The filter accepts the following options:
8631 Set height of level. Default value is @code{200}.
8632 Allowed range is [50, 2048].
8635 Set height of color scale. Default value is @code{12}.
8636 Allowed range is [0, 40].
8640 It accepts the following values:
8643 Per color component graphs are placed below each other.
8646 Presents information identical to that in the @code{parade}, except
8647 that the graphs representing color components are superimposed directly
8650 Default is @code{parade}.
8653 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8654 Default is @code{linear}.
8657 Set what color components to display.
8658 Default is @code{7}.
8661 Set foreground opacity. Default is @code{0.7}.
8664 Set background opacity. Default is @code{0.5}.
8667 @subsection Examples
8672 Calculate and draw histogram:
8674 ffplay -i input -vf histogram
8682 This is a high precision/quality 3d denoise filter. It aims to reduce
8683 image noise, producing smooth images and making still images really
8684 still. It should enhance compressibility.
8686 It accepts the following optional parameters:
8690 A non-negative floating point number which specifies spatial luma strength.
8693 @item chroma_spatial
8694 A non-negative floating point number which specifies spatial chroma strength.
8695 It defaults to 3.0*@var{luma_spatial}/4.0.
8698 A floating point number which specifies luma temporal strength. It defaults to
8699 6.0*@var{luma_spatial}/4.0.
8702 A floating point number which specifies chroma temporal strength. It defaults to
8703 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8706 @anchor{hwupload_cuda}
8707 @section hwupload_cuda
8709 Upload system memory frames to a CUDA device.
8711 It accepts the following optional parameters:
8715 The number of the CUDA device to use
8720 Apply a high-quality magnification filter designed for pixel art. This filter
8721 was originally created by Maxim Stepin.
8723 It accepts the following option:
8727 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8728 @code{hq3x} and @code{4} for @code{hq4x}.
8729 Default is @code{3}.
8733 Stack input videos horizontally.
8735 All streams must be of same pixel format and of same height.
8737 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8738 to create same output.
8740 The filter accept the following option:
8744 Set number of input streams. Default is 2.
8747 If set to 1, force the output to terminate when the shortest input
8748 terminates. Default value is 0.
8753 Modify the hue and/or the saturation of the input.
8755 It accepts the following parameters:
8759 Specify the hue angle as a number of degrees. It accepts an expression,
8760 and defaults to "0".
8763 Specify the saturation in the [-10,10] range. It accepts an expression and
8767 Specify the hue angle as a number of radians. It accepts an
8768 expression, and defaults to "0".
8771 Specify the brightness in the [-10,10] range. It accepts an expression and
8775 @option{h} and @option{H} are mutually exclusive, and can't be
8776 specified at the same time.
8778 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8779 expressions containing the following constants:
8783 frame count of the input frame starting from 0
8786 presentation timestamp of the input frame expressed in time base units
8789 frame rate of the input video, NAN if the input frame rate is unknown
8792 timestamp expressed in seconds, NAN if the input timestamp is unknown
8795 time base of the input video
8798 @subsection Examples
8802 Set the hue to 90 degrees and the saturation to 1.0:
8808 Same command but expressing the hue in radians:
8814 Rotate hue and make the saturation swing between 0
8815 and 2 over a period of 1 second:
8817 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8821 Apply a 3 seconds saturation fade-in effect starting at 0:
8826 The general fade-in expression can be written as:
8828 hue="s=min(0\, max((t-START)/DURATION\, 1))"
8832 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
8834 hue="s=max(0\, min(1\, (8-t)/3))"
8837 The general fade-out expression can be written as:
8839 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
8844 @subsection Commands
8846 This filter supports the following commands:
8852 Modify the hue and/or the saturation and/or brightness of the input video.
8853 The command accepts the same syntax of the corresponding option.
8855 If the specified expression is not valid, it is kept at its current
8861 Grow first stream into second stream by connecting components.
8862 This makes it possible to build more robust edge masks.
8864 This filter accepts the following options:
8868 Set which planes will be processed as bitmap, unprocessed planes will be
8869 copied from first stream.
8870 By default value 0xf, all planes will be processed.
8873 Set threshold which is used in filtering. If pixel component value is higher than
8874 this value filter algorithm for connecting components is activated.
8875 By default value is 0.
8880 Detect video interlacing type.
8882 This filter tries to detect if the input frames are interlaced, progressive,
8883 top or bottom field first. It will also try to detect fields that are
8884 repeated between adjacent frames (a sign of telecine).
8886 Single frame detection considers only immediately adjacent frames when classifying each frame.
8887 Multiple frame detection incorporates the classification history of previous frames.
8889 The filter will log these metadata values:
8892 @item single.current_frame
8893 Detected type of current frame using single-frame detection. One of:
8894 ``tff'' (top field first), ``bff'' (bottom field first),
8895 ``progressive'', or ``undetermined''
8898 Cumulative number of frames detected as top field first using single-frame detection.
8901 Cumulative number of frames detected as top field first using multiple-frame detection.
8904 Cumulative number of frames detected as bottom field first using single-frame detection.
8906 @item multiple.current_frame
8907 Detected type of current frame using multiple-frame detection. One of:
8908 ``tff'' (top field first), ``bff'' (bottom field first),
8909 ``progressive'', or ``undetermined''
8912 Cumulative number of frames detected as bottom field first using multiple-frame detection.
8914 @item single.progressive
8915 Cumulative number of frames detected as progressive using single-frame detection.
8917 @item multiple.progressive
8918 Cumulative number of frames detected as progressive using multiple-frame detection.
8920 @item single.undetermined
8921 Cumulative number of frames that could not be classified using single-frame detection.
8923 @item multiple.undetermined
8924 Cumulative number of frames that could not be classified using multiple-frame detection.
8926 @item repeated.current_frame
8927 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
8929 @item repeated.neither
8930 Cumulative number of frames with no repeated field.
8933 Cumulative number of frames with the top field repeated from the previous frame's top field.
8935 @item repeated.bottom
8936 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
8939 The filter accepts the following options:
8943 Set interlacing threshold.
8945 Set progressive threshold.
8947 Threshold for repeated field detection.
8949 Number of frames after which a given frame's contribution to the
8950 statistics is halved (i.e., it contributes only 0.5 to its
8951 classification). The default of 0 means that all frames seen are given
8952 full weight of 1.0 forever.
8953 @item analyze_interlaced_flag
8954 When this is not 0 then idet will use the specified number of frames to determine
8955 if the interlaced flag is accurate, it will not count undetermined frames.
8956 If the flag is found to be accurate it will be used without any further
8957 computations, if it is found to be inaccurate it will be cleared without any
8958 further computations. This allows inserting the idet filter as a low computational
8959 method to clean up the interlaced flag
8964 Deinterleave or interleave fields.
8966 This filter allows one to process interlaced images fields without
8967 deinterlacing them. Deinterleaving splits the input frame into 2
8968 fields (so called half pictures). Odd lines are moved to the top
8969 half of the output image, even lines to the bottom half.
8970 You can process (filter) them independently and then re-interleave them.
8972 The filter accepts the following options:
8976 @item chroma_mode, c
8978 Available values for @var{luma_mode}, @var{chroma_mode} and
8979 @var{alpha_mode} are:
8985 @item deinterleave, d
8986 Deinterleave fields, placing one above the other.
8989 Interleave fields. Reverse the effect of deinterleaving.
8991 Default value is @code{none}.
8994 @item chroma_swap, cs
8995 @item alpha_swap, as
8996 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
9001 Apply inflate effect to the video.
9003 This filter replaces the pixel by the local(3x3) average by taking into account
9004 only values higher than the pixel.
9006 It accepts the following options:
9013 Limit the maximum change for each plane, default is 65535.
9014 If 0, plane will remain unchanged.
9019 Simple interlacing filter from progressive contents. This interleaves upper (or
9020 lower) lines from odd frames with lower (or upper) lines from even frames,
9021 halving the frame rate and preserving image height.
9024 Original Original New Frame
9025 Frame 'j' Frame 'j+1' (tff)
9026 ========== =========== ==================
9027 Line 0 --------------------> Frame 'j' Line 0
9028 Line 1 Line 1 ----> Frame 'j+1' Line 1
9029 Line 2 ---------------------> Frame 'j' Line 2
9030 Line 3 Line 3 ----> Frame 'j+1' Line 3
9032 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9035 It accepts the following optional parameters:
9039 This determines whether the interlaced frame is taken from the even
9040 (tff - default) or odd (bff) lines of the progressive frame.
9043 Enable (default) or disable the vertical lowpass filter to avoid twitter
9044 interlacing and reduce moire patterns.
9049 Deinterlace input video by applying Donald Graft's adaptive kernel
9050 deinterling. Work on interlaced parts of a video to produce
9053 The description of the accepted parameters follows.
9057 Set the threshold which affects the filter's tolerance when
9058 determining if a pixel line must be processed. It must be an integer
9059 in the range [0,255] and defaults to 10. A value of 0 will result in
9060 applying the process on every pixels.
9063 Paint pixels exceeding the threshold value to white if set to 1.
9067 Set the fields order. Swap fields if set to 1, leave fields alone if
9071 Enable additional sharpening if set to 1. Default is 0.
9074 Enable twoway sharpening if set to 1. Default is 0.
9077 @subsection Examples
9081 Apply default values:
9083 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9087 Enable additional sharpening:
9093 Paint processed pixels in white:
9099 @section lenscorrection
9101 Correct radial lens distortion
9103 This filter can be used to correct for radial distortion as can result from the use
9104 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9105 one can use tools available for example as part of opencv or simply trial-and-error.
9106 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9107 and extract the k1 and k2 coefficients from the resulting matrix.
9109 Note that effectively the same filter is available in the open-source tools Krita and
9110 Digikam from the KDE project.
9112 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9113 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9114 brightness distribution, so you may want to use both filters together in certain
9115 cases, though you will have to take care of ordering, i.e. whether vignetting should
9116 be applied before or after lens correction.
9120 The filter accepts the following options:
9124 Relative x-coordinate of the focal point of the image, and thereby the center of the
9125 distortion. This value has a range [0,1] and is expressed as fractions of the image
9128 Relative y-coordinate of the focal point of the image, and thereby the center of the
9129 distortion. This value has a range [0,1] and is expressed as fractions of the image
9132 Coefficient of the quadratic correction term. 0.5 means no correction.
9134 Coefficient of the double quadratic correction term. 0.5 means no correction.
9137 The formula that generates the correction is:
9139 @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)
9141 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9142 distances from the focal point in the source and target images, respectively.
9148 The filter accepts the following options:
9152 Set the number of loops.
9155 Set maximal size in number of frames.
9158 Set first frame of loop.
9164 Apply a 3D LUT to an input video.
9166 The filter accepts the following options:
9170 Set the 3D LUT file name.
9172 Currently supported formats:
9184 Select interpolation mode.
9186 Available values are:
9190 Use values from the nearest defined point.
9192 Interpolate values using the 8 points defining a cube.
9194 Interpolate values using a tetrahedron.
9198 @section lut, lutrgb, lutyuv
9200 Compute a look-up table for binding each pixel component input value
9201 to an output value, and apply it to the input video.
9203 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9204 to an RGB input video.
9206 These filters accept the following parameters:
9209 set first pixel component expression
9211 set second pixel component expression
9213 set third pixel component expression
9215 set fourth pixel component expression, corresponds to the alpha component
9218 set red component expression
9220 set green component expression
9222 set blue component expression
9224 alpha component expression
9227 set Y/luminance component expression
9229 set U/Cb component expression
9231 set V/Cr component expression
9234 Each of them specifies the expression to use for computing the lookup table for
9235 the corresponding pixel component values.
9237 The exact component associated to each of the @var{c*} options depends on the
9240 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9241 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9243 The expressions can contain the following constants and functions:
9248 The input width and height.
9251 The input value for the pixel component.
9254 The input value, clipped to the @var{minval}-@var{maxval} range.
9257 The maximum value for the pixel component.
9260 The minimum value for the pixel component.
9263 The negated value for the pixel component value, clipped to the
9264 @var{minval}-@var{maxval} range; it corresponds to the expression
9265 "maxval-clipval+minval".
9268 The computed value in @var{val}, clipped to the
9269 @var{minval}-@var{maxval} range.
9271 @item gammaval(gamma)
9272 The computed gamma correction value of the pixel component value,
9273 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9275 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9279 All expressions default to "val".
9281 @subsection Examples
9287 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9288 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9291 The above is the same as:
9293 lutrgb="r=negval:g=negval:b=negval"
9294 lutyuv="y=negval:u=negval:v=negval"
9304 Remove chroma components, turning the video into a graytone image:
9306 lutyuv="u=128:v=128"
9310 Apply a luma burning effect:
9316 Remove green and blue components:
9322 Set a constant alpha channel value on input:
9324 format=rgba,lutrgb=a="maxval-minval/2"
9328 Correct luminance gamma by a factor of 0.5:
9330 lutyuv=y=gammaval(0.5)
9334 Discard least significant bits of luma:
9336 lutyuv=y='bitand(val, 128+64+32)'
9340 Technicolor like effect:
9342 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9348 Compute and apply a lookup table from two video inputs.
9350 This filter accepts the following parameters:
9353 set first pixel component expression
9355 set second pixel component expression
9357 set third pixel component expression
9359 set fourth pixel component expression, corresponds to the alpha component
9362 Each of them specifies the expression to use for computing the lookup table for
9363 the corresponding pixel component values.
9365 The exact component associated to each of the @var{c*} options depends on the
9368 The expressions can contain the following constants:
9373 The input width and height.
9376 The first input value for the pixel component.
9379 The second input value for the pixel component.
9382 The first input video bit depth.
9385 The second input video bit depth.
9388 All expressions default to "x".
9390 @subsection Examples
9394 Highlight differences between two RGB video streams:
9396 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1)'
9400 Highlight differences between two YUV video streams:
9402 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1)'
9406 @section maskedclamp
9408 Clamp the first input stream with the second input and third input stream.
9410 Returns the value of first stream to be between second input
9411 stream - @code{undershoot} and third input stream + @code{overshoot}.
9413 This filter accepts the following options:
9416 Default value is @code{0}.
9419 Default value is @code{0}.
9422 Set which planes will be processed as bitmap, unprocessed planes will be
9423 copied from first stream.
9424 By default value 0xf, all planes will be processed.
9427 @section maskedmerge
9429 Merge the first input stream with the second input stream using per pixel
9430 weights in the third input stream.
9432 A value of 0 in the third stream pixel component means that pixel component
9433 from first stream is returned unchanged, while maximum value (eg. 255 for
9434 8-bit videos) means that pixel component from second stream is returned
9435 unchanged. Intermediate values define the amount of merging between both
9436 input stream's pixel components.
9438 This filter accepts the following options:
9441 Set which planes will be processed as bitmap, unprocessed planes will be
9442 copied from first stream.
9443 By default value 0xf, all planes will be processed.
9448 Apply motion-compensation deinterlacing.
9450 It needs one field per frame as input and must thus be used together
9451 with yadif=1/3 or equivalent.
9453 This filter accepts the following options:
9456 Set the deinterlacing mode.
9458 It accepts one of the following values:
9463 use iterative motion estimation
9465 like @samp{slow}, but use multiple reference frames.
9467 Default value is @samp{fast}.
9470 Set the picture field parity assumed for the input video. It must be
9471 one of the following values:
9475 assume top field first
9477 assume bottom field first
9480 Default value is @samp{bff}.
9483 Set per-block quantization parameter (QP) used by the internal
9486 Higher values should result in a smoother motion vector field but less
9487 optimal individual vectors. Default value is 1.
9490 @section mergeplanes
9492 Merge color channel components from several video streams.
9494 The filter accepts up to 4 input streams, and merge selected input
9495 planes to the output video.
9497 This filter accepts the following options:
9500 Set input to output plane mapping. Default is @code{0}.
9502 The mappings is specified as a bitmap. It should be specified as a
9503 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9504 mapping for the first plane of the output stream. 'A' sets the number of
9505 the input stream to use (from 0 to 3), and 'a' the plane number of the
9506 corresponding input to use (from 0 to 3). The rest of the mappings is
9507 similar, 'Bb' describes the mapping for the output stream second
9508 plane, 'Cc' describes the mapping for the output stream third plane and
9509 'Dd' describes the mapping for the output stream fourth plane.
9512 Set output pixel format. Default is @code{yuva444p}.
9515 @subsection Examples
9519 Merge three gray video streams of same width and height into single video stream:
9521 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9525 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9527 [a0][a1]mergeplanes=0x00010210:yuva444p
9531 Swap Y and A plane in yuva444p stream:
9533 format=yuva444p,mergeplanes=0x03010200:yuva444p
9537 Swap U and V plane in yuv420p stream:
9539 format=yuv420p,mergeplanes=0x000201:yuv420p
9543 Cast a rgb24 clip to yuv444p:
9545 format=rgb24,mergeplanes=0x000102:yuv444p
9551 Estimate and export motion vectors using block matching algorithms.
9552 Motion vectors are stored in frame side data to be used by other filters.
9554 This filter accepts the following options:
9557 Specify the motion estimation method. Accepts one of the following values:
9561 Exhaustive search algorithm.
9563 Three step search algorithm.
9565 Two dimensional logarithmic search algorithm.
9567 New three step search algorithm.
9569 Four step search algorithm.
9571 Diamond search algorithm.
9573 Hexagon-based search algorithm.
9575 Enhanced predictive zonal search algorithm.
9577 Uneven multi-hexagon search algorithm.
9579 Default value is @samp{esa}.
9582 Macroblock size. Default @code{16}.
9585 Search parameter. Default @code{7}.
9588 @section minterpolate
9590 Convert the video to specified frame rate using motion interpolation.
9592 This filter accepts the following options:
9595 Specify the output frame rate. This can be rational e.g. @code{60000/1001}. Frames are dropped if @var{fps} is lower than source fps. Default @code{60}.
9598 Motion interpolation mode. Following values are accepted:
9601 Duplicate previous or next frame for interpolating new ones.
9603 Blend source frames. Interpolated frame is mean of previous and next frames.
9605 Motion compensated interpolation. Following options are effective when this mode is selected:
9609 Motion compensation mode. Following values are accepted:
9612 Overlapped block motion compensation.
9614 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
9616 Default mode is @samp{obmc}.
9619 Motion estimation mode. Following values are accepted:
9622 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
9624 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
9626 Default mode is @samp{bilat}.
9629 The algorithm to be used for motion estimation. Following values are accepted:
9632 Exhaustive search algorithm.
9634 Three step search algorithm.
9636 Two dimensional logarithmic search algorithm.
9638 New three step search algorithm.
9640 Four step search algorithm.
9642 Diamond search algorithm.
9644 Hexagon-based search algorithm.
9646 Enhanced predictive zonal search algorithm.
9648 Uneven multi-hexagon search algorithm.
9650 Default algorithm is @samp{epzs}.
9653 Macroblock size. Default @code{16}.
9656 Motion estimation search parameter. Default @code{32}.
9659 Enable variable-size block motion compensation. Motion estimation is applied with smaller block sizes at object boundaries in order to make the them less blur. Default is @code{0} (disabled).
9664 Scene change detection method. Scene change leads motion vectors to be in random direction. Scene change detection replace interpolated frames by duplicate ones. May not be needed for other modes. Following values are accepted:
9667 Disable scene change detection.
9669 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
9671 Default method is @samp{fdiff}.
9674 Scene change detection threshold. Default is @code{5.0}.
9679 Drop frames that do not differ greatly from the previous frame in
9680 order to reduce frame rate.
9682 The main use of this filter is for very-low-bitrate encoding
9683 (e.g. streaming over dialup modem), but it could in theory be used for
9684 fixing movies that were inverse-telecined incorrectly.
9686 A description of the accepted options follows.
9690 Set the maximum number of consecutive frames which can be dropped (if
9691 positive), or the minimum interval between dropped frames (if
9692 negative). If the value is 0, the frame is dropped unregarding the
9693 number of previous sequentially dropped frames.
9700 Set the dropping threshold values.
9702 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9703 represent actual pixel value differences, so a threshold of 64
9704 corresponds to 1 unit of difference for each pixel, or the same spread
9705 out differently over the block.
9707 A frame is a candidate for dropping if no 8x8 blocks differ by more
9708 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9709 meaning the whole image) differ by more than a threshold of @option{lo}.
9711 Default value for @option{hi} is 64*12, default value for @option{lo} is
9712 64*5, and default value for @option{frac} is 0.33.
9720 It accepts an integer in input; if non-zero it negates the
9721 alpha component (if available). The default value in input is 0.
9725 Denoise frames using Non-Local Means algorithm.
9727 Each pixel is adjusted by looking for other pixels with similar contexts. This
9728 context similarity is defined by comparing their surrounding patches of size
9729 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
9732 Note that the research area defines centers for patches, which means some
9733 patches will be made of pixels outside that research area.
9735 The filter accepts the following options.
9739 Set denoising strength.
9745 Same as @option{p} but for chroma planes.
9747 The default value is @var{0} and means automatic.
9753 Same as @option{r} but for chroma planes.
9755 The default value is @var{0} and means automatic.
9760 Deinterlace video using neural network edge directed interpolation.
9762 This filter accepts the following options:
9766 Mandatory option, without binary file filter can not work.
9767 Currently file can be found here:
9768 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9771 Set which frames to deinterlace, by default it is @code{all}.
9772 Can be @code{all} or @code{interlaced}.
9775 Set mode of operation.
9777 Can be one of the following:
9781 Use frame flags, both fields.
9783 Use frame flags, single field.
9787 Use bottom field only.
9789 Use both fields, top first.
9791 Use both fields, bottom first.
9795 Set which planes to process, by default filter process all frames.
9798 Set size of local neighborhood around each pixel, used by the predictor neural
9801 Can be one of the following:
9814 Set the number of neurons in predicctor neural network.
9815 Can be one of the following:
9826 Controls the number of different neural network predictions that are blended
9827 together to compute the final output value. Can be @code{fast}, default or
9831 Set which set of weights to use in the predictor.
9832 Can be one of the following:
9836 weights trained to minimize absolute error
9838 weights trained to minimize squared error
9842 Controls whether or not the prescreener neural network is used to decide
9843 which pixels should be processed by the predictor neural network and which
9844 can be handled by simple cubic interpolation.
9845 The prescreener is trained to know whether cubic interpolation will be
9846 sufficient for a pixel or whether it should be predicted by the predictor nn.
9847 The computational complexity of the prescreener nn is much less than that of
9848 the predictor nn. Since most pixels can be handled by cubic interpolation,
9849 using the prescreener generally results in much faster processing.
9850 The prescreener is pretty accurate, so the difference between using it and not
9851 using it is almost always unnoticeable.
9853 Can be one of the following:
9861 Default is @code{new}.
9864 Set various debugging flags.
9869 Force libavfilter not to use any of the specified pixel formats for the
9870 input to the next filter.
9872 It accepts the following parameters:
9876 A '|'-separated list of pixel format names, such as
9877 apix_fmts=yuv420p|monow|rgb24".
9881 @subsection Examples
9885 Force libavfilter to use a format different from @var{yuv420p} for the
9886 input to the vflip filter:
9888 noformat=pix_fmts=yuv420p,vflip
9892 Convert the input video to any of the formats not contained in the list:
9894 noformat=yuv420p|yuv444p|yuv410p
9900 Add noise on video input frame.
9902 The filter accepts the following options:
9910 Set noise seed for specific pixel component or all pixel components in case
9911 of @var{all_seed}. Default value is @code{123457}.
9913 @item all_strength, alls
9914 @item c0_strength, c0s
9915 @item c1_strength, c1s
9916 @item c2_strength, c2s
9917 @item c3_strength, c3s
9918 Set noise strength for specific pixel component or all pixel components in case
9919 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
9921 @item all_flags, allf
9926 Set pixel component flags or set flags for all components if @var{all_flags}.
9927 Available values for component flags are:
9930 averaged temporal noise (smoother)
9932 mix random noise with a (semi)regular pattern
9934 temporal noise (noise pattern changes between frames)
9936 uniform noise (gaussian otherwise)
9940 @subsection Examples
9942 Add temporal and uniform noise to input video:
9944 noise=alls=20:allf=t+u
9949 Pass the video source unchanged to the output.
9952 Optical Character Recognition
9954 This filter uses Tesseract for optical character recognition.
9956 It accepts the following options:
9960 Set datapath to tesseract data. Default is to use whatever was
9961 set at installation.
9964 Set language, default is "eng".
9967 Set character whitelist.
9970 Set character blacklist.
9973 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
9977 Apply a video transform using libopencv.
9979 To enable this filter, install the libopencv library and headers and
9980 configure FFmpeg with @code{--enable-libopencv}.
9982 It accepts the following parameters:
9987 The name of the libopencv filter to apply.
9990 The parameters to pass to the libopencv filter. If not specified, the default
9995 Refer to the official libopencv documentation for more precise
9997 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
9999 Several libopencv filters are supported; see the following subsections.
10004 Dilate an image by using a specific structuring element.
10005 It corresponds to the libopencv function @code{cvDilate}.
10007 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
10009 @var{struct_el} represents a structuring element, and has the syntax:
10010 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
10012 @var{cols} and @var{rows} represent the number of columns and rows of
10013 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
10014 point, and @var{shape} the shape for the structuring element. @var{shape}
10015 must be "rect", "cross", "ellipse", or "custom".
10017 If the value for @var{shape} is "custom", it must be followed by a
10018 string of the form "=@var{filename}". The file with name
10019 @var{filename} is assumed to represent a binary image, with each
10020 printable character corresponding to a bright pixel. When a custom
10021 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
10022 or columns and rows of the read file are assumed instead.
10024 The default value for @var{struct_el} is "3x3+0x0/rect".
10026 @var{nb_iterations} specifies the number of times the transform is
10027 applied to the image, and defaults to 1.
10031 # Use the default values
10034 # Dilate using a structuring element with a 5x5 cross, iterating two times
10035 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10037 # Read the shape from the file diamond.shape, iterating two times.
10038 # The file diamond.shape may contain a pattern of characters like this
10044 # The specified columns and rows are ignored
10045 # but the anchor point coordinates are not
10046 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10051 Erode an image by using a specific structuring element.
10052 It corresponds to the libopencv function @code{cvErode}.
10054 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10055 with the same syntax and semantics as the @ref{dilate} filter.
10059 Smooth the input video.
10061 The filter takes the following parameters:
10062 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10064 @var{type} is the type of smooth filter to apply, and must be one of
10065 the following values: "blur", "blur_no_scale", "median", "gaussian",
10066 or "bilateral". The default value is "gaussian".
10068 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10069 depend on the smooth type. @var{param1} and
10070 @var{param2} accept integer positive values or 0. @var{param3} and
10071 @var{param4} accept floating point values.
10073 The default value for @var{param1} is 3. The default value for the
10074 other parameters is 0.
10076 These parameters correspond to the parameters assigned to the
10077 libopencv function @code{cvSmooth}.
10082 Overlay one video on top of another.
10084 It takes two inputs and has one output. The first input is the "main"
10085 video on which the second input is overlaid.
10087 It accepts the following parameters:
10089 A description of the accepted options follows.
10094 Set the expression for the x and y coordinates of the overlaid video
10095 on the main video. Default value is "0" for both expressions. In case
10096 the expression is invalid, it is set to a huge value (meaning that the
10097 overlay will not be displayed within the output visible area).
10100 The action to take when EOF is encountered on the secondary input; it accepts
10101 one of the following values:
10105 Repeat the last frame (the default).
10109 Pass the main input through.
10113 Set when the expressions for @option{x}, and @option{y} are evaluated.
10115 It accepts the following values:
10118 only evaluate expressions once during the filter initialization or
10119 when a command is processed
10122 evaluate expressions for each incoming frame
10125 Default value is @samp{frame}.
10128 If set to 1, force the output to terminate when the shortest input
10129 terminates. Default value is 0.
10132 Set the format for the output video.
10134 It accepts the following values:
10137 force YUV420 output
10140 force YUV422 output
10143 force YUV444 output
10149 Default value is @samp{yuv420}.
10151 @item rgb @emph{(deprecated)}
10152 If set to 1, force the filter to accept inputs in the RGB
10153 color space. Default value is 0. This option is deprecated, use
10154 @option{format} instead.
10157 If set to 1, force the filter to draw the last overlay frame over the
10158 main input until the end of the stream. A value of 0 disables this
10159 behavior. Default value is 1.
10162 The @option{x}, and @option{y} expressions can contain the following
10168 The main input width and height.
10172 The overlay input width and height.
10176 The computed values for @var{x} and @var{y}. They are evaluated for
10181 horizontal and vertical chroma subsample values of the output
10182 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
10186 the number of input frame, starting from 0
10189 the position in the file of the input frame, NAN if unknown
10192 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
10196 Note that the @var{n}, @var{pos}, @var{t} variables are available only
10197 when evaluation is done @emph{per frame}, and will evaluate to NAN
10198 when @option{eval} is set to @samp{init}.
10200 Be aware that frames are taken from each input video in timestamp
10201 order, hence, if their initial timestamps differ, it is a good idea
10202 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
10203 have them begin in the same zero timestamp, as the example for
10204 the @var{movie} filter does.
10206 You can chain together more overlays but you should test the
10207 efficiency of such approach.
10209 @subsection Commands
10211 This filter supports the following commands:
10215 Modify the x and y of the overlay input.
10216 The command accepts the same syntax of the corresponding option.
10218 If the specified expression is not valid, it is kept at its current
10222 @subsection Examples
10226 Draw the overlay at 10 pixels from the bottom right corner of the main
10229 overlay=main_w-overlay_w-10:main_h-overlay_h-10
10232 Using named options the example above becomes:
10234 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
10238 Insert a transparent PNG logo in the bottom left corner of the input,
10239 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
10241 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
10245 Insert 2 different transparent PNG logos (second logo on bottom
10246 right corner) using the @command{ffmpeg} tool:
10248 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
10252 Add a transparent color layer on top of the main video; @code{WxH}
10253 must specify the size of the main input to the overlay filter:
10255 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
10259 Play an original video and a filtered version (here with the deshake
10260 filter) side by side using the @command{ffplay} tool:
10262 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
10265 The above command is the same as:
10267 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
10271 Make a sliding overlay appearing from the left to the right top part of the
10272 screen starting since time 2:
10274 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
10278 Compose output by putting two input videos side to side:
10280 ffmpeg -i left.avi -i right.avi -filter_complex "
10281 nullsrc=size=200x100 [background];
10282 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
10283 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
10284 [background][left] overlay=shortest=1 [background+left];
10285 [background+left][right] overlay=shortest=1:x=100 [left+right]
10290 Mask 10-20 seconds of a video by applying the delogo filter to a section
10292 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
10293 -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]'
10298 Chain several overlays in cascade:
10300 nullsrc=s=200x200 [bg];
10301 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
10302 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
10303 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
10304 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
10305 [in3] null, [mid2] overlay=100:100 [out0]
10312 Apply Overcomplete Wavelet denoiser.
10314 The filter accepts the following options:
10320 Larger depth values will denoise lower frequency components more, but
10321 slow down filtering.
10323 Must be an int in the range 8-16, default is @code{8}.
10325 @item luma_strength, ls
10328 Must be a double value in the range 0-1000, default is @code{1.0}.
10330 @item chroma_strength, cs
10331 Set chroma strength.
10333 Must be a double value in the range 0-1000, default is @code{1.0}.
10339 Add paddings to the input image, and place the original input at the
10340 provided @var{x}, @var{y} coordinates.
10342 It accepts the following parameters:
10347 Specify an expression for the size of the output image with the
10348 paddings added. If the value for @var{width} or @var{height} is 0, the
10349 corresponding input size is used for the output.
10351 The @var{width} expression can reference the value set by the
10352 @var{height} expression, and vice versa.
10354 The default value of @var{width} and @var{height} is 0.
10358 Specify the offsets to place the input image at within the padded area,
10359 with respect to the top/left border of the output image.
10361 The @var{x} expression can reference the value set by the @var{y}
10362 expression, and vice versa.
10364 The default value of @var{x} and @var{y} is 0.
10367 Specify the color of the padded area. For the syntax of this option,
10368 check the "Color" section in the ffmpeg-utils manual.
10370 The default value of @var{color} is "black".
10373 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10374 options are expressions containing the following constants:
10379 The input video width and height.
10383 These are the same as @var{in_w} and @var{in_h}.
10387 The output width and height (the size of the padded area), as
10388 specified by the @var{width} and @var{height} expressions.
10392 These are the same as @var{out_w} and @var{out_h}.
10396 The x and y offsets as specified by the @var{x} and @var{y}
10397 expressions, or NAN if not yet specified.
10400 same as @var{iw} / @var{ih}
10403 input sample aspect ratio
10406 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10410 The horizontal and vertical chroma subsample values. For example for the
10411 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10414 @subsection Examples
10418 Add paddings with the color "violet" to the input video. The output video
10419 size is 640x480, and the top-left corner of the input video is placed at
10422 pad=640:480:0:40:violet
10425 The example above is equivalent to the following command:
10427 pad=width=640:height=480:x=0:y=40:color=violet
10431 Pad the input to get an output with dimensions increased by 3/2,
10432 and put the input video at the center of the padded area:
10434 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10438 Pad the input to get a squared output with size equal to the maximum
10439 value between the input width and height, and put the input video at
10440 the center of the padded area:
10442 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10446 Pad the input to get a final w/h ratio of 16:9:
10448 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10452 In case of anamorphic video, in order to set the output display aspect
10453 correctly, it is necessary to use @var{sar} in the expression,
10454 according to the relation:
10456 (ih * X / ih) * sar = output_dar
10457 X = output_dar / sar
10460 Thus the previous example needs to be modified to:
10462 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10466 Double the output size and put the input video in the bottom-right
10467 corner of the output padded area:
10469 pad="2*iw:2*ih:ow-iw:oh-ih"
10473 @anchor{palettegen}
10474 @section palettegen
10476 Generate one palette for a whole video stream.
10478 It accepts the following options:
10482 Set the maximum number of colors to quantize in the palette.
10483 Note: the palette will still contain 256 colors; the unused palette entries
10486 @item reserve_transparent
10487 Create a palette of 255 colors maximum and reserve the last one for
10488 transparency. Reserving the transparency color is useful for GIF optimization.
10489 If not set, the maximum of colors in the palette will be 256. You probably want
10490 to disable this option for a standalone image.
10494 Set statistics mode.
10496 It accepts the following values:
10499 Compute full frame histograms.
10501 Compute histograms only for the part that differs from previous frame. This
10502 might be relevant to give more importance to the moving part of your input if
10503 the background is static.
10505 Compute new histogram for each frame.
10508 Default value is @var{full}.
10511 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10512 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10513 color quantization of the palette. This information is also visible at
10514 @var{info} logging level.
10516 @subsection Examples
10520 Generate a representative palette of a given video using @command{ffmpeg}:
10522 ffmpeg -i input.mkv -vf palettegen palette.png
10526 @section paletteuse
10528 Use a palette to downsample an input video stream.
10530 The filter takes two inputs: one video stream and a palette. The palette must
10531 be a 256 pixels image.
10533 It accepts the following options:
10537 Select dithering mode. Available algorithms are:
10540 Ordered 8x8 bayer dithering (deterministic)
10542 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10543 Note: this dithering is sometimes considered "wrong" and is included as a
10545 @item floyd_steinberg
10546 Floyd and Steingberg dithering (error diffusion)
10548 Frankie Sierra dithering v2 (error diffusion)
10550 Frankie Sierra dithering v2 "Lite" (error diffusion)
10553 Default is @var{sierra2_4a}.
10556 When @var{bayer} dithering is selected, this option defines the scale of the
10557 pattern (how much the crosshatch pattern is visible). A low value means more
10558 visible pattern for less banding, and higher value means less visible pattern
10559 at the cost of more banding.
10561 The option must be an integer value in the range [0,5]. Default is @var{2}.
10564 If set, define the zone to process
10568 Only the changing rectangle will be reprocessed. This is similar to GIF
10569 cropping/offsetting compression mechanism. This option can be useful for speed
10570 if only a part of the image is changing, and has use cases such as limiting the
10571 scope of the error diffusal @option{dither} to the rectangle that bounds the
10572 moving scene (it leads to more deterministic output if the scene doesn't change
10573 much, and as a result less moving noise and better GIF compression).
10576 Default is @var{none}.
10579 Take new palette for each output frame.
10582 @subsection Examples
10586 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10587 using @command{ffmpeg}:
10589 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10593 @section perspective
10595 Correct perspective of video not recorded perpendicular to the screen.
10597 A description of the accepted parameters follows.
10608 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10609 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10610 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10611 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10612 then the corners of the source will be sent to the specified coordinates.
10614 The expressions can use the following variables:
10619 the width and height of video frame.
10623 Output frame count.
10626 @item interpolation
10627 Set interpolation for perspective correction.
10629 It accepts the following values:
10635 Default value is @samp{linear}.
10638 Set interpretation of coordinate options.
10640 It accepts the following values:
10644 Send point in the source specified by the given coordinates to
10645 the corners of the destination.
10647 @item 1, destination
10649 Send the corners of the source to the point in the destination specified
10650 by the given coordinates.
10652 Default value is @samp{source}.
10656 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10658 It accepts the following values:
10661 only evaluate expressions once during the filter initialization or
10662 when a command is processed
10665 evaluate expressions for each incoming frame
10668 Default value is @samp{init}.
10673 Delay interlaced video by one field time so that the field order changes.
10675 The intended use is to fix PAL movies that have been captured with the
10676 opposite field order to the film-to-video transfer.
10678 A description of the accepted parameters follows.
10684 It accepts the following values:
10687 Capture field order top-first, transfer bottom-first.
10688 Filter will delay the bottom field.
10691 Capture field order bottom-first, transfer top-first.
10692 Filter will delay the top field.
10695 Capture and transfer with the same field order. This mode only exists
10696 for the documentation of the other options to refer to, but if you
10697 actually select it, the filter will faithfully do nothing.
10700 Capture field order determined automatically by field flags, transfer
10702 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
10703 basis using field flags. If no field information is available,
10704 then this works just like @samp{u}.
10707 Capture unknown or varying, transfer opposite.
10708 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
10709 analyzing the images and selecting the alternative that produces best
10710 match between the fields.
10713 Capture top-first, transfer unknown or varying.
10714 Filter selects among @samp{t} and @samp{p} using image analysis.
10717 Capture bottom-first, transfer unknown or varying.
10718 Filter selects among @samp{b} and @samp{p} using image analysis.
10721 Capture determined by field flags, transfer unknown or varying.
10722 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
10723 image analysis. If no field information is available, then this works just
10724 like @samp{U}. This is the default mode.
10727 Both capture and transfer unknown or varying.
10728 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
10732 @section pixdesctest
10734 Pixel format descriptor test filter, mainly useful for internal
10735 testing. The output video should be equal to the input video.
10739 format=monow, pixdesctest
10742 can be used to test the monowhite pixel format descriptor definition.
10746 Enable the specified chain of postprocessing subfilters using libpostproc. This
10747 library should be automatically selected with a GPL build (@code{--enable-gpl}).
10748 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
10749 Each subfilter and some options have a short and a long name that can be used
10750 interchangeably, i.e. dr/dering are the same.
10752 The filters accept the following options:
10756 Set postprocessing subfilters string.
10759 All subfilters share common options to determine their scope:
10763 Honor the quality commands for this subfilter.
10766 Do chrominance filtering, too (default).
10769 Do luminance filtering only (no chrominance).
10772 Do chrominance filtering only (no luminance).
10775 These options can be appended after the subfilter name, separated by a '|'.
10777 Available subfilters are:
10780 @item hb/hdeblock[|difference[|flatness]]
10781 Horizontal deblocking filter
10784 Difference factor where higher values mean more deblocking (default: @code{32}).
10786 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10789 @item vb/vdeblock[|difference[|flatness]]
10790 Vertical deblocking filter
10793 Difference factor where higher values mean more deblocking (default: @code{32}).
10795 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10798 @item ha/hadeblock[|difference[|flatness]]
10799 Accurate horizontal deblocking filter
10802 Difference factor where higher values mean more deblocking (default: @code{32}).
10804 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10807 @item va/vadeblock[|difference[|flatness]]
10808 Accurate vertical deblocking filter
10811 Difference factor where higher values mean more deblocking (default: @code{32}).
10813 Flatness threshold where lower values mean more deblocking (default: @code{39}).
10817 The horizontal and vertical deblocking filters share the difference and
10818 flatness values so you cannot set different horizontal and vertical
10822 @item h1/x1hdeblock
10823 Experimental horizontal deblocking filter
10825 @item v1/x1vdeblock
10826 Experimental vertical deblocking filter
10831 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
10834 larger -> stronger filtering
10836 larger -> stronger filtering
10838 larger -> stronger filtering
10841 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
10844 Stretch luminance to @code{0-255}.
10847 @item lb/linblenddeint
10848 Linear blend deinterlacing filter that deinterlaces the given block by
10849 filtering all lines with a @code{(1 2 1)} filter.
10851 @item li/linipoldeint
10852 Linear interpolating deinterlacing filter that deinterlaces the given block by
10853 linearly interpolating every second line.
10855 @item ci/cubicipoldeint
10856 Cubic interpolating deinterlacing filter deinterlaces the given block by
10857 cubically interpolating every second line.
10859 @item md/mediandeint
10860 Median deinterlacing filter that deinterlaces the given block by applying a
10861 median filter to every second line.
10863 @item fd/ffmpegdeint
10864 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
10865 second line with a @code{(-1 4 2 4 -1)} filter.
10868 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
10869 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
10871 @item fq/forceQuant[|quantizer]
10872 Overrides the quantizer table from the input with the constant quantizer you
10880 Default pp filter combination (@code{hb|a,vb|a,dr|a})
10883 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
10886 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
10889 @subsection Examples
10893 Apply horizontal and vertical deblocking, deringing and automatic
10894 brightness/contrast:
10900 Apply default filters without brightness/contrast correction:
10906 Apply default filters and temporal denoiser:
10908 pp=default/tmpnoise|1|2|3
10912 Apply deblocking on luminance only, and switch vertical deblocking on or off
10913 automatically depending on available CPU time:
10920 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
10921 similar to spp = 6 with 7 point DCT, where only the center sample is
10924 The filter accepts the following options:
10928 Force a constant quantization parameter. It accepts an integer in range
10929 0 to 63. If not set, the filter will use the QP from the video stream
10933 Set thresholding mode. Available modes are:
10937 Set hard thresholding.
10939 Set soft thresholding (better de-ringing effect, but likely blurrier).
10941 Set medium thresholding (good results, default).
10946 Apply prewitt operator to input video stream.
10948 The filter accepts the following option:
10952 Set which planes will be processed, unprocessed planes will be copied.
10953 By default value 0xf, all planes will be processed.
10956 Set value which will be multiplied with filtered result.
10959 Set value which will be added to filtered result.
10964 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
10965 Ratio) between two input videos.
10967 This filter takes in input two input videos, the first input is
10968 considered the "main" source and is passed unchanged to the
10969 output. The second input is used as a "reference" video for computing
10972 Both video inputs must have the same resolution and pixel format for
10973 this filter to work correctly. Also it assumes that both inputs
10974 have the same number of frames, which are compared one by one.
10976 The obtained average PSNR is printed through the logging system.
10978 The filter stores the accumulated MSE (mean squared error) of each
10979 frame, and at the end of the processing it is averaged across all frames
10980 equally, and the following formula is applied to obtain the PSNR:
10983 PSNR = 10*log10(MAX^2/MSE)
10986 Where MAX is the average of the maximum values of each component of the
10989 The description of the accepted parameters follows.
10992 @item stats_file, f
10993 If specified the filter will use the named file to save the PSNR of
10994 each individual frame. When filename equals "-" the data is sent to
10997 @item stats_version
10998 Specifies which version of the stats file format to use. Details of
10999 each format are written below.
11000 Default value is 1.
11002 @item stats_add_max
11003 Determines whether the max value is output to the stats log.
11004 Default value is 0.
11005 Requires stats_version >= 2. If this is set and stats_version < 2,
11006 the filter will return an error.
11009 The file printed if @var{stats_file} is selected, contains a sequence of
11010 key/value pairs of the form @var{key}:@var{value} for each compared
11013 If a @var{stats_version} greater than 1 is specified, a header line precedes
11014 the list of per-frame-pair stats, with key value pairs following the frame
11015 format with the following parameters:
11018 @item psnr_log_version
11019 The version of the log file format. Will match @var{stats_version}.
11022 A comma separated list of the per-frame-pair parameters included in
11026 A description of each shown per-frame-pair parameter follows:
11030 sequential number of the input frame, starting from 1
11033 Mean Square Error pixel-by-pixel average difference of the compared
11034 frames, averaged over all the image components.
11036 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
11037 Mean Square Error pixel-by-pixel average difference of the compared
11038 frames for the component specified by the suffix.
11040 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
11041 Peak Signal to Noise ratio of the compared frames for the component
11042 specified by the suffix.
11044 @item max_avg, max_y, max_u, max_v
11045 Maximum allowed value for each channel, and average over all
11051 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
11052 [main][ref] psnr="stats_file=stats.log" [out]
11055 On this example the input file being processed is compared with the
11056 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
11057 is stored in @file{stats.log}.
11062 Pulldown reversal (inverse telecine) filter, capable of handling mixed
11063 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
11066 The pullup filter is designed to take advantage of future context in making
11067 its decisions. This filter is stateless in the sense that it does not lock
11068 onto a pattern to follow, but it instead looks forward to the following
11069 fields in order to identify matches and rebuild progressive frames.
11071 To produce content with an even framerate, insert the fps filter after
11072 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
11073 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
11075 The filter accepts the following options:
11082 These options set the amount of "junk" to ignore at the left, right, top, and
11083 bottom of the image, respectively. Left and right are in units of 8 pixels,
11084 while top and bottom are in units of 2 lines.
11085 The default is 8 pixels on each side.
11088 Set the strict breaks. Setting this option to 1 will reduce the chances of
11089 filter generating an occasional mismatched frame, but it may also cause an
11090 excessive number of frames to be dropped during high motion sequences.
11091 Conversely, setting it to -1 will make filter match fields more easily.
11092 This may help processing of video where there is slight blurring between
11093 the fields, but may also cause there to be interlaced frames in the output.
11094 Default value is @code{0}.
11097 Set the metric plane to use. It accepts the following values:
11103 Use chroma blue plane.
11106 Use chroma red plane.
11109 This option may be set to use chroma plane instead of the default luma plane
11110 for doing filter's computations. This may improve accuracy on very clean
11111 source material, but more likely will decrease accuracy, especially if there
11112 is chroma noise (rainbow effect) or any grayscale video.
11113 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
11114 load and make pullup usable in realtime on slow machines.
11117 For best results (without duplicated frames in the output file) it is
11118 necessary to change the output frame rate. For example, to inverse
11119 telecine NTSC input:
11121 ffmpeg -i input -vf pullup -r 24000/1001 ...
11126 Change video quantization parameters (QP).
11128 The filter accepts the following option:
11132 Set expression for quantization parameter.
11135 The expression is evaluated through the eval API and can contain, among others,
11136 the following constants:
11140 1 if index is not 129, 0 otherwise.
11143 Sequentional index starting from -129 to 128.
11146 @subsection Examples
11150 Some equation like:
11158 Flush video frames from internal cache of frames into a random order.
11159 No frame is discarded.
11160 Inspired by @ref{frei0r} nervous filter.
11164 Set size in number of frames of internal cache, in range from @code{2} to
11165 @code{512}. Default is @code{30}.
11168 Set seed for random number generator, must be an integer included between
11169 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
11170 less than @code{0}, the filter will try to use a good random seed on a
11176 Read vertical interval timecode (VITC) information from the top lines of a
11179 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
11180 timecode value, if a valid timecode has been detected. Further metadata key
11181 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
11182 timecode data has been found or not.
11184 This filter accepts the following options:
11188 Set the maximum number of lines to scan for VITC data. If the value is set to
11189 @code{-1} the full video frame is scanned. Default is @code{45}.
11192 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
11193 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
11196 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
11197 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
11200 @subsection Examples
11204 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
11205 draw @code{--:--:--:--} as a placeholder:
11207 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
11213 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
11215 Destination pixel at position (X, Y) will be picked from source (x, y) position
11216 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
11217 value for pixel will be used for destination pixel.
11219 Xmap and Ymap input video streams must be of same dimensions. Output video stream
11220 will have Xmap/Ymap video stream dimensions.
11221 Xmap and Ymap input video streams are 16bit depth, single channel.
11223 @section removegrain
11225 The removegrain filter is a spatial denoiser for progressive video.
11229 Set mode for the first plane.
11232 Set mode for the second plane.
11235 Set mode for the third plane.
11238 Set mode for the fourth plane.
11241 Range of mode is from 0 to 24. Description of each mode follows:
11245 Leave input plane unchanged. Default.
11248 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
11251 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
11254 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
11257 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
11258 This is equivalent to a median filter.
11261 Line-sensitive clipping giving the minimal change.
11264 Line-sensitive clipping, intermediate.
11267 Line-sensitive clipping, intermediate.
11270 Line-sensitive clipping, intermediate.
11273 Line-sensitive clipping on a line where the neighbours pixels are the closest.
11276 Replaces the target pixel with the closest neighbour.
11279 [1 2 1] horizontal and vertical kernel blur.
11285 Bob mode, interpolates top field from the line where the neighbours
11286 pixels are the closest.
11289 Bob mode, interpolates bottom field from the line where the neighbours
11290 pixels are the closest.
11293 Bob mode, interpolates top field. Same as 13 but with a more complicated
11294 interpolation formula.
11297 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
11298 interpolation formula.
11301 Clips the pixel with the minimum and maximum of respectively the maximum and
11302 minimum of each pair of opposite neighbour pixels.
11305 Line-sensitive clipping using opposite neighbours whose greatest distance from
11306 the current pixel is minimal.
11309 Replaces the pixel with the average of its 8 neighbours.
11312 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
11315 Clips pixels using the averages of opposite neighbour.
11318 Same as mode 21 but simpler and faster.
11321 Small edge and halo removal, but reputed useless.
11327 @section removelogo
11329 Suppress a TV station logo, using an image file to determine which
11330 pixels comprise the logo. It works by filling in the pixels that
11331 comprise the logo with neighboring pixels.
11333 The filter accepts the following options:
11337 Set the filter bitmap file, which can be any image format supported by
11338 libavformat. The width and height of the image file must match those of the
11339 video stream being processed.
11342 Pixels in the provided bitmap image with a value of zero are not
11343 considered part of the logo, non-zero pixels are considered part of
11344 the logo. If you use white (255) for the logo and black (0) for the
11345 rest, you will be safe. For making the filter bitmap, it is
11346 recommended to take a screen capture of a black frame with the logo
11347 visible, and then using a threshold filter followed by the erode
11348 filter once or twice.
11350 If needed, little splotches can be fixed manually. Remember that if
11351 logo pixels are not covered, the filter quality will be much
11352 reduced. Marking too many pixels as part of the logo does not hurt as
11353 much, but it will increase the amount of blurring needed to cover over
11354 the image and will destroy more information than necessary, and extra
11355 pixels will slow things down on a large logo.
11357 @section repeatfields
11359 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11360 fields based on its value.
11364 Reverse a video clip.
11366 Warning: This filter requires memory to buffer the entire clip, so trimming
11369 @subsection Examples
11373 Take the first 5 seconds of a clip, and reverse it.
11381 Rotate video by an arbitrary angle expressed in radians.
11383 The filter accepts the following options:
11385 A description of the optional parameters follows.
11388 Set an expression for the angle by which to rotate the input video
11389 clockwise, expressed as a number of radians. A negative value will
11390 result in a counter-clockwise rotation. By default it is set to "0".
11392 This expression is evaluated for each frame.
11395 Set the output width expression, default value is "iw".
11396 This expression is evaluated just once during configuration.
11399 Set the output height expression, default value is "ih".
11400 This expression is evaluated just once during configuration.
11403 Enable bilinear interpolation if set to 1, a value of 0 disables
11404 it. Default value is 1.
11407 Set the color used to fill the output area not covered by the rotated
11408 image. For the general syntax of this option, check the "Color" section in the
11409 ffmpeg-utils manual. If the special value "none" is selected then no
11410 background is printed (useful for example if the background is never shown).
11412 Default value is "black".
11415 The expressions for the angle and the output size can contain the
11416 following constants and functions:
11420 sequential number of the input frame, starting from 0. It is always NAN
11421 before the first frame is filtered.
11424 time in seconds of the input frame, it is set to 0 when the filter is
11425 configured. It is always NAN before the first frame is filtered.
11429 horizontal and vertical chroma subsample values. For example for the
11430 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11434 the input video width and height
11438 the output width and height, that is the size of the padded area as
11439 specified by the @var{width} and @var{height} expressions
11443 the minimal width/height required for completely containing the input
11444 video rotated by @var{a} radians.
11446 These are only available when computing the @option{out_w} and
11447 @option{out_h} expressions.
11450 @subsection Examples
11454 Rotate the input by PI/6 radians clockwise:
11460 Rotate the input by PI/6 radians counter-clockwise:
11466 Rotate the input by 45 degrees clockwise:
11472 Apply a constant rotation with period T, starting from an angle of PI/3:
11474 rotate=PI/3+2*PI*t/T
11478 Make the input video rotation oscillating with a period of T
11479 seconds and an amplitude of A radians:
11481 rotate=A*sin(2*PI/T*t)
11485 Rotate the video, output size is chosen so that the whole rotating
11486 input video is always completely contained in the output:
11488 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11492 Rotate the video, reduce the output size so that no background is ever
11495 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11499 @subsection Commands
11501 The filter supports the following commands:
11505 Set the angle expression.
11506 The command accepts the same syntax of the corresponding option.
11508 If the specified expression is not valid, it is kept at its current
11514 Apply Shape Adaptive Blur.
11516 The filter accepts the following options:
11519 @item luma_radius, lr
11520 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11521 value is 1.0. A greater value will result in a more blurred image, and
11522 in slower processing.
11524 @item luma_pre_filter_radius, lpfr
11525 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11528 @item luma_strength, ls
11529 Set luma maximum difference between pixels to still be considered, must
11530 be a value in the 0.1-100.0 range, default value is 1.0.
11532 @item chroma_radius, cr
11533 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11534 greater value will result in a more blurred image, and in slower
11537 @item chroma_pre_filter_radius, cpfr
11538 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11540 @item chroma_strength, cs
11541 Set chroma maximum difference between pixels to still be considered,
11542 must be a value in the -0.9-100.0 range.
11545 Each chroma option value, if not explicitly specified, is set to the
11546 corresponding luma option value.
11551 Scale (resize) the input video, using the libswscale library.
11553 The scale filter forces the output display aspect ratio to be the same
11554 of the input, by changing the output sample aspect ratio.
11556 If the input image format is different from the format requested by
11557 the next filter, the scale filter will convert the input to the
11560 @subsection Options
11561 The filter accepts the following options, or any of the options
11562 supported by the libswscale scaler.
11564 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
11565 the complete list of scaler options.
11570 Set the output video dimension expression. Default value is the input
11573 If the value is 0, the input width is used for the output.
11575 If one of the values is -1, the scale filter will use a value that
11576 maintains the aspect ratio of the input image, calculated from the
11577 other specified dimension. If both of them are -1, the input size is
11580 If one of the values is -n with n > 1, the scale filter will also use a value
11581 that maintains the aspect ratio of the input image, calculated from the other
11582 specified dimension. After that it will, however, make sure that the calculated
11583 dimension is divisible by n and adjust the value if necessary.
11585 See below for the list of accepted constants for use in the dimension
11589 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
11593 Only evaluate expressions once during the filter initialization or when a command is processed.
11596 Evaluate expressions for each incoming frame.
11600 Default value is @samp{init}.
11604 Set the interlacing mode. It accepts the following values:
11608 Force interlaced aware scaling.
11611 Do not apply interlaced scaling.
11614 Select interlaced aware scaling depending on whether the source frames
11615 are flagged as interlaced or not.
11618 Default value is @samp{0}.
11621 Set libswscale scaling flags. See
11622 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11623 complete list of values. If not explicitly specified the filter applies
11627 @item param0, param1
11628 Set libswscale input parameters for scaling algorithms that need them. See
11629 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
11630 complete documentation. If not explicitly specified the filter applies
11636 Set the video size. For the syntax of this option, check the
11637 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
11639 @item in_color_matrix
11640 @item out_color_matrix
11641 Set in/output YCbCr color space type.
11643 This allows the autodetected value to be overridden as well as allows forcing
11644 a specific value used for the output and encoder.
11646 If not specified, the color space type depends on the pixel format.
11652 Choose automatically.
11655 Format conforming to International Telecommunication Union (ITU)
11656 Recommendation BT.709.
11659 Set color space conforming to the United States Federal Communications
11660 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
11663 Set color space conforming to:
11667 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
11670 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
11673 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
11678 Set color space conforming to SMPTE ST 240:1999.
11683 Set in/output YCbCr sample range.
11685 This allows the autodetected value to be overridden as well as allows forcing
11686 a specific value used for the output and encoder. If not specified, the
11687 range depends on the pixel format. Possible values:
11691 Choose automatically.
11694 Set full range (0-255 in case of 8-bit luma).
11697 Set "MPEG" range (16-235 in case of 8-bit luma).
11700 @item force_original_aspect_ratio
11701 Enable decreasing or increasing output video width or height if necessary to
11702 keep the original aspect ratio. Possible values:
11706 Scale the video as specified and disable this feature.
11709 The output video dimensions will automatically be decreased if needed.
11712 The output video dimensions will automatically be increased if needed.
11716 One useful instance of this option is that when you know a specific device's
11717 maximum allowed resolution, you can use this to limit the output video to
11718 that, while retaining the aspect ratio. For example, device A allows
11719 1280x720 playback, and your video is 1920x800. Using this option (set it to
11720 decrease) and specifying 1280x720 to the command line makes the output
11723 Please note that this is a different thing than specifying -1 for @option{w}
11724 or @option{h}, you still need to specify the output resolution for this option
11729 The values of the @option{w} and @option{h} options are expressions
11730 containing the following constants:
11735 The input width and height
11739 These are the same as @var{in_w} and @var{in_h}.
11743 The output (scaled) width and height
11747 These are the same as @var{out_w} and @var{out_h}
11750 The same as @var{iw} / @var{ih}
11753 input sample aspect ratio
11756 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
11760 horizontal and vertical input chroma subsample values. For example for the
11761 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11765 horizontal and vertical output chroma subsample values. For example for the
11766 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11769 @subsection Examples
11773 Scale the input video to a size of 200x100
11778 This is equivalent to:
11789 Specify a size abbreviation for the output size:
11794 which can also be written as:
11800 Scale the input to 2x:
11802 scale=w=2*iw:h=2*ih
11806 The above is the same as:
11808 scale=2*in_w:2*in_h
11812 Scale the input to 2x with forced interlaced scaling:
11814 scale=2*iw:2*ih:interl=1
11818 Scale the input to half size:
11820 scale=w=iw/2:h=ih/2
11824 Increase the width, and set the height to the same size:
11830 Seek Greek harmony:
11837 Increase the height, and set the width to 3/2 of the height:
11839 scale=w=3/2*oh:h=3/5*ih
11843 Increase the size, making the size a multiple of the chroma
11846 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
11850 Increase the width to a maximum of 500 pixels,
11851 keeping the same aspect ratio as the input:
11853 scale=w='min(500\, iw*3/2):h=-1'
11857 @subsection Commands
11859 This filter supports the following commands:
11863 Set the output video dimension expression.
11864 The command accepts the same syntax of the corresponding option.
11866 If the specified expression is not valid, it is kept at its current
11872 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
11873 format conversion on CUDA video frames. Setting the output width and height
11874 works in the same way as for the @var{scale} filter.
11876 The following additional options are accepted:
11879 The pixel format of the output CUDA frames. If set to the string "same" (the
11880 default), the input format will be kept. Note that automatic format negotiation
11881 and conversion is not yet supported for hardware frames
11884 The interpolation algorithm used for resizing. One of the following:
11891 @item cubic2p_bspline
11892 2-parameter cubic (B=1, C=0)
11894 @item cubic2p_catmullrom
11895 2-parameter cubic (B=0, C=1/2)
11897 @item cubic2p_b05c03
11898 2-parameter cubic (B=1/2, C=3/10)
11910 Scale (resize) the input video, based on a reference video.
11912 See the scale filter for available options, scale2ref supports the same but
11913 uses the reference video instead of the main input as basis.
11915 @subsection Examples
11919 Scale a subtitle stream to match the main video in size before overlaying
11921 'scale2ref[b][a];[a][b]overlay'
11925 @anchor{selectivecolor}
11926 @section selectivecolor
11928 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
11929 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
11930 by the "purity" of the color (that is, how saturated it already is).
11932 This filter is similar to the Adobe Photoshop Selective Color tool.
11934 The filter accepts the following options:
11937 @item correction_method
11938 Select color correction method.
11940 Available values are:
11943 Specified adjustments are applied "as-is" (added/subtracted to original pixel
11946 Specified adjustments are relative to the original component value.
11948 Default is @code{absolute}.
11950 Adjustments for red pixels (pixels where the red component is the maximum)
11952 Adjustments for yellow pixels (pixels where the blue component is the minimum)
11954 Adjustments for green pixels (pixels where the green component is the maximum)
11956 Adjustments for cyan pixels (pixels where the red component is the minimum)
11958 Adjustments for blue pixels (pixels where the blue component is the maximum)
11960 Adjustments for magenta pixels (pixels where the green component is the minimum)
11962 Adjustments for white pixels (pixels where all components are greater than 128)
11964 Adjustments for all pixels except pure black and pure white
11966 Adjustments for black pixels (pixels where all components are lesser than 128)
11968 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
11971 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
11972 4 space separated floating point adjustment values in the [-1,1] range,
11973 respectively to adjust the amount of cyan, magenta, yellow and black for the
11974 pixels of its range.
11976 @subsection Examples
11980 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
11981 increase magenta by 27% in blue areas:
11983 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
11987 Use a Photoshop selective color preset:
11989 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
11993 @anchor{separatefields}
11994 @section separatefields
11996 The @code{separatefields} takes a frame-based video input and splits
11997 each frame into its components fields, producing a new half height clip
11998 with twice the frame rate and twice the frame count.
12000 This filter use field-dominance information in frame to decide which
12001 of each pair of fields to place first in the output.
12002 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
12004 @section setdar, setsar
12006 The @code{setdar} filter sets the Display Aspect Ratio for the filter
12009 This is done by changing the specified Sample (aka Pixel) Aspect
12010 Ratio, according to the following equation:
12012 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
12015 Keep in mind that the @code{setdar} filter does not modify the pixel
12016 dimensions of the video frame. Also, the display aspect ratio set by
12017 this filter may be changed by later filters in the filterchain,
12018 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
12021 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
12022 the filter output video.
12024 Note that as a consequence of the application of this filter, the
12025 output display aspect ratio will change according to the equation
12028 Keep in mind that the sample aspect ratio set by the @code{setsar}
12029 filter may be changed by later filters in the filterchain, e.g. if
12030 another "setsar" or a "setdar" filter is applied.
12032 It accepts the following parameters:
12035 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
12036 Set the aspect ratio used by the filter.
12038 The parameter can be a floating point number string, an expression, or
12039 a string of the form @var{num}:@var{den}, where @var{num} and
12040 @var{den} are the numerator and denominator of the aspect ratio. If
12041 the parameter is not specified, it is assumed the value "0".
12042 In case the form "@var{num}:@var{den}" is used, the @code{:} character
12046 Set the maximum integer value to use for expressing numerator and
12047 denominator when reducing the expressed aspect ratio to a rational.
12048 Default value is @code{100}.
12052 The parameter @var{sar} is an expression containing
12053 the following constants:
12057 These are approximated values for the mathematical constants e
12058 (Euler's number), pi (Greek pi), and phi (the golden ratio).
12061 The input width and height.
12064 These are the same as @var{w} / @var{h}.
12067 The input sample aspect ratio.
12070 The input display aspect ratio. It is the same as
12071 (@var{w} / @var{h}) * @var{sar}.
12074 Horizontal and vertical chroma subsample values. For example, for the
12075 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12078 @subsection Examples
12083 To change the display aspect ratio to 16:9, specify one of the following:
12090 To change the sample aspect ratio to 10:11, specify:
12096 To set a display aspect ratio of 16:9, and specify a maximum integer value of
12097 1000 in the aspect ratio reduction, use the command:
12099 setdar=ratio=16/9:max=1000
12107 Force field for the output video frame.
12109 The @code{setfield} filter marks the interlace type field for the
12110 output frames. It does not change the input frame, but only sets the
12111 corresponding property, which affects how the frame is treated by
12112 following filters (e.g. @code{fieldorder} or @code{yadif}).
12114 The filter accepts the following options:
12119 Available values are:
12123 Keep the same field property.
12126 Mark the frame as bottom-field-first.
12129 Mark the frame as top-field-first.
12132 Mark the frame as progressive.
12138 Show a line containing various information for each input video frame.
12139 The input video is not modified.
12141 The shown line contains a sequence of key/value pairs of the form
12142 @var{key}:@var{value}.
12144 The following values are shown in the output:
12148 The (sequential) number of the input frame, starting from 0.
12151 The Presentation TimeStamp of the input frame, expressed as a number of
12152 time base units. The time base unit depends on the filter input pad.
12155 The Presentation TimeStamp of the input frame, expressed as a number of
12159 The position of the frame in the input stream, or -1 if this information is
12160 unavailable and/or meaningless (for example in case of synthetic video).
12163 The pixel format name.
12166 The sample aspect ratio of the input frame, expressed in the form
12167 @var{num}/@var{den}.
12170 The size of the input frame. For the syntax of this option, check the
12171 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12174 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
12175 for bottom field first).
12178 This is 1 if the frame is a key frame, 0 otherwise.
12181 The picture type of the input frame ("I" for an I-frame, "P" for a
12182 P-frame, "B" for a B-frame, or "?" for an unknown type).
12183 Also refer to the documentation of the @code{AVPictureType} enum and of
12184 the @code{av_get_picture_type_char} function defined in
12185 @file{libavutil/avutil.h}.
12188 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
12190 @item plane_checksum
12191 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
12192 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
12195 @section showpalette
12197 Displays the 256 colors palette of each frame. This filter is only relevant for
12198 @var{pal8} pixel format frames.
12200 It accepts the following option:
12204 Set the size of the box used to represent one palette color entry. Default is
12205 @code{30} (for a @code{30x30} pixel box).
12208 @section shuffleframes
12210 Reorder and/or duplicate video frames.
12212 It accepts the following parameters:
12216 Set the destination indexes of input frames.
12217 This is space or '|' separated list of indexes that maps input frames to output
12218 frames. Number of indexes also sets maximal value that each index may have.
12221 The first frame has the index 0. The default is to keep the input unchanged.
12223 @subsection Examples
12227 Swap second and third frame of every three frames of the input:
12229 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
12233 Swap 10th and 1st frame of every ten frames of the input:
12235 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
12239 @section shuffleplanes
12241 Reorder and/or duplicate video planes.
12243 It accepts the following parameters:
12248 The index of the input plane to be used as the first output plane.
12251 The index of the input plane to be used as the second output plane.
12254 The index of the input plane to be used as the third output plane.
12257 The index of the input plane to be used as the fourth output plane.
12261 The first plane has the index 0. The default is to keep the input unchanged.
12263 @subsection Examples
12267 Swap the second and third planes of the input:
12269 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
12273 @anchor{signalstats}
12274 @section signalstats
12275 Evaluate various visual metrics that assist in determining issues associated
12276 with the digitization of analog video media.
12278 By default the filter will log these metadata values:
12282 Display the minimal Y value contained within the input frame. Expressed in
12286 Display the Y value at the 10% percentile within the input frame. Expressed in
12290 Display the average Y value within the input frame. Expressed in range of
12294 Display the Y value at the 90% percentile within the input frame. Expressed in
12298 Display the maximum Y value contained within the input frame. Expressed in
12302 Display the minimal U value contained within the input frame. Expressed in
12306 Display the U value at the 10% percentile within the input frame. Expressed in
12310 Display the average U value within the input frame. Expressed in range of
12314 Display the U value at the 90% percentile within the input frame. Expressed in
12318 Display the maximum U value contained within the input frame. Expressed in
12322 Display the minimal V value contained within the input frame. Expressed in
12326 Display the V value at the 10% percentile within the input frame. Expressed in
12330 Display the average V value within the input frame. Expressed in range of
12334 Display the V value at the 90% percentile within the input frame. Expressed in
12338 Display the maximum V value contained within the input frame. Expressed in
12342 Display the minimal saturation value contained within the input frame.
12343 Expressed in range of [0-~181.02].
12346 Display the saturation value at the 10% percentile within the input frame.
12347 Expressed in range of [0-~181.02].
12350 Display the average saturation value within the input frame. Expressed in range
12354 Display the saturation value at the 90% percentile within the input frame.
12355 Expressed in range of [0-~181.02].
12358 Display the maximum saturation value contained within the input frame.
12359 Expressed in range of [0-~181.02].
12362 Display the median value for hue within the input frame. Expressed in range of
12366 Display the average value for hue within the input frame. Expressed in range of
12370 Display the average of sample value difference between all values of the Y
12371 plane in the current frame and corresponding values of the previous input frame.
12372 Expressed in range of [0-255].
12375 Display the average of sample value difference between all values of the U
12376 plane in the current frame and corresponding values of the previous input frame.
12377 Expressed in range of [0-255].
12380 Display the average of sample value difference between all values of the V
12381 plane in the current frame and corresponding values of the previous input frame.
12382 Expressed in range of [0-255].
12385 Display bit depth of Y plane in current frame.
12386 Expressed in range of [0-16].
12389 Display bit depth of U plane in current frame.
12390 Expressed in range of [0-16].
12393 Display bit depth of V plane in current frame.
12394 Expressed in range of [0-16].
12397 The filter accepts the following options:
12403 @option{stat} specify an additional form of image analysis.
12404 @option{out} output video with the specified type of pixel highlighted.
12406 Both options accept the following values:
12410 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12411 unlike the neighboring pixels of the same field. Examples of temporal outliers
12412 include the results of video dropouts, head clogs, or tape tracking issues.
12415 Identify @var{vertical line repetition}. Vertical line repetition includes
12416 similar rows of pixels within a frame. In born-digital video vertical line
12417 repetition is common, but this pattern is uncommon in video digitized from an
12418 analog source. When it occurs in video that results from the digitization of an
12419 analog source it can indicate concealment from a dropout compensator.
12422 Identify pixels that fall outside of legal broadcast range.
12426 Set the highlight color for the @option{out} option. The default color is
12430 @subsection Examples
12434 Output data of various video metrics:
12436 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12440 Output specific data about the minimum and maximum values of the Y plane per frame:
12442 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12446 Playback video while highlighting pixels that are outside of broadcast range in red.
12448 ffplay example.mov -vf signalstats="out=brng:color=red"
12452 Playback video with signalstats metadata drawn over the frame.
12454 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12457 The contents of signalstat_drawtext.txt used in the command are:
12460 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12461 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12462 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12463 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12471 Blur the input video without impacting the outlines.
12473 It accepts the following options:
12476 @item luma_radius, lr
12477 Set the luma radius. The option value must be a float number in
12478 the range [0.1,5.0] that specifies the variance of the gaussian filter
12479 used to blur the image (slower if larger). Default value is 1.0.
12481 @item luma_strength, ls
12482 Set the luma strength. The option value must be a float number
12483 in the range [-1.0,1.0] that configures the blurring. A value included
12484 in [0.0,1.0] will blur the image whereas a value included in
12485 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12487 @item luma_threshold, lt
12488 Set the luma threshold used as a coefficient to determine
12489 whether a pixel should be blurred or not. The option value must be an
12490 integer in the range [-30,30]. A value of 0 will filter all the image,
12491 a value included in [0,30] will filter flat areas and a value included
12492 in [-30,0] will filter edges. Default value is 0.
12494 @item chroma_radius, cr
12495 Set the chroma radius. The option value must be a float number in
12496 the range [0.1,5.0] that specifies the variance of the gaussian filter
12497 used to blur the image (slower if larger). Default value is 1.0.
12499 @item chroma_strength, cs
12500 Set the chroma strength. The option value must be a float number
12501 in the range [-1.0,1.0] that configures the blurring. A value included
12502 in [0.0,1.0] will blur the image whereas a value included in
12503 [-1.0,0.0] will sharpen the image. Default value is 1.0.
12505 @item chroma_threshold, ct
12506 Set the chroma threshold used as a coefficient to determine
12507 whether a pixel should be blurred or not. The option value must be an
12508 integer in the range [-30,30]. A value of 0 will filter all the image,
12509 a value included in [0,30] will filter flat areas and a value included
12510 in [-30,0] will filter edges. Default value is 0.
12513 If a chroma option is not explicitly set, the corresponding luma value
12518 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
12520 This filter takes in input two input videos, the first input is
12521 considered the "main" source and is passed unchanged to the
12522 output. The second input is used as a "reference" video for computing
12525 Both video inputs must have the same resolution and pixel format for
12526 this filter to work correctly. Also it assumes that both inputs
12527 have the same number of frames, which are compared one by one.
12529 The filter stores the calculated SSIM of each frame.
12531 The description of the accepted parameters follows.
12534 @item stats_file, f
12535 If specified the filter will use the named file to save the SSIM of
12536 each individual frame. When filename equals "-" the data is sent to
12540 The file printed if @var{stats_file} is selected, contains a sequence of
12541 key/value pairs of the form @var{key}:@var{value} for each compared
12544 A description of each shown parameter follows:
12548 sequential number of the input frame, starting from 1
12550 @item Y, U, V, R, G, B
12551 SSIM of the compared frames for the component specified by the suffix.
12554 SSIM of the compared frames for the whole frame.
12557 Same as above but in dB representation.
12562 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
12563 [main][ref] ssim="stats_file=stats.log" [out]
12566 On this example the input file being processed is compared with the
12567 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
12568 is stored in @file{stats.log}.
12570 Another example with both psnr and ssim at same time:
12572 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
12577 Convert between different stereoscopic image formats.
12579 The filters accept the following options:
12583 Set stereoscopic image format of input.
12585 Available values for input image formats are:
12588 side by side parallel (left eye left, right eye right)
12591 side by side crosseye (right eye left, left eye right)
12594 side by side parallel with half width resolution
12595 (left eye left, right eye right)
12598 side by side crosseye with half width resolution
12599 (right eye left, left eye right)
12602 above-below (left eye above, right eye below)
12605 above-below (right eye above, left eye below)
12608 above-below with half height resolution
12609 (left eye above, right eye below)
12612 above-below with half height resolution
12613 (right eye above, left eye below)
12616 alternating frames (left eye first, right eye second)
12619 alternating frames (right eye first, left eye second)
12622 interleaved rows (left eye has top row, right eye starts on next row)
12625 interleaved rows (right eye has top row, left eye starts on next row)
12628 interleaved columns, left eye first
12631 interleaved columns, right eye first
12633 Default value is @samp{sbsl}.
12637 Set stereoscopic image format of output.
12641 side by side parallel (left eye left, right eye right)
12644 side by side crosseye (right eye left, left eye right)
12647 side by side parallel with half width resolution
12648 (left eye left, right eye right)
12651 side by side crosseye with half width resolution
12652 (right eye left, left eye right)
12655 above-below (left eye above, right eye below)
12658 above-below (right eye above, left eye below)
12661 above-below with half height resolution
12662 (left eye above, right eye below)
12665 above-below with half height resolution
12666 (right eye above, left eye below)
12669 alternating frames (left eye first, right eye second)
12672 alternating frames (right eye first, left eye second)
12675 interleaved rows (left eye has top row, right eye starts on next row)
12678 interleaved rows (right eye has top row, left eye starts on next row)
12681 anaglyph red/blue gray
12682 (red filter on left eye, blue filter on right eye)
12685 anaglyph red/green gray
12686 (red filter on left eye, green filter on right eye)
12689 anaglyph red/cyan gray
12690 (red filter on left eye, cyan filter on right eye)
12693 anaglyph red/cyan half colored
12694 (red filter on left eye, cyan filter on right eye)
12697 anaglyph red/cyan color
12698 (red filter on left eye, cyan filter on right eye)
12701 anaglyph red/cyan color optimized with the least squares projection of dubois
12702 (red filter on left eye, cyan filter on right eye)
12705 anaglyph green/magenta gray
12706 (green filter on left eye, magenta filter on right eye)
12709 anaglyph green/magenta half colored
12710 (green filter on left eye, magenta filter on right eye)
12713 anaglyph green/magenta colored
12714 (green filter on left eye, magenta filter on right eye)
12717 anaglyph green/magenta color optimized with the least squares projection of dubois
12718 (green filter on left eye, magenta filter on right eye)
12721 anaglyph yellow/blue gray
12722 (yellow filter on left eye, blue filter on right eye)
12725 anaglyph yellow/blue half colored
12726 (yellow filter on left eye, blue filter on right eye)
12729 anaglyph yellow/blue colored
12730 (yellow filter on left eye, blue filter on right eye)
12733 anaglyph yellow/blue color optimized with the least squares projection of dubois
12734 (yellow filter on left eye, blue filter on right eye)
12737 mono output (left eye only)
12740 mono output (right eye only)
12743 checkerboard, left eye first
12746 checkerboard, right eye first
12749 interleaved columns, left eye first
12752 interleaved columns, right eye first
12758 Default value is @samp{arcd}.
12761 @subsection Examples
12765 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
12771 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
12777 @section streamselect, astreamselect
12778 Select video or audio streams.
12780 The filter accepts the following options:
12784 Set number of inputs. Default is 2.
12787 Set input indexes to remap to outputs.
12790 @subsection Commands
12792 The @code{streamselect} and @code{astreamselect} filter supports the following
12797 Set input indexes to remap to outputs.
12800 @subsection Examples
12804 Select first 5 seconds 1st stream and rest of time 2nd stream:
12806 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
12810 Same as above, but for audio:
12812 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
12817 Apply sobel operator to input video stream.
12819 The filter accepts the following option:
12823 Set which planes will be processed, unprocessed planes will be copied.
12824 By default value 0xf, all planes will be processed.
12827 Set value which will be multiplied with filtered result.
12830 Set value which will be added to filtered result.
12836 Apply a simple postprocessing filter that compresses and decompresses the image
12837 at several (or - in the case of @option{quality} level @code{6} - all) shifts
12838 and average the results.
12840 The filter accepts the following options:
12844 Set quality. This option defines the number of levels for averaging. It accepts
12845 an integer in the range 0-6. If set to @code{0}, the filter will have no
12846 effect. A value of @code{6} means the higher quality. For each increment of
12847 that value the speed drops by a factor of approximately 2. Default value is
12851 Force a constant quantization parameter. If not set, the filter will use the QP
12852 from the video stream (if available).
12855 Set thresholding mode. Available modes are:
12859 Set hard thresholding (default).
12861 Set soft thresholding (better de-ringing effect, but likely blurrier).
12864 @item use_bframe_qp
12865 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
12866 option may cause flicker since the B-Frames have often larger QP. Default is
12867 @code{0} (not enabled).
12873 Draw subtitles on top of input video using the libass library.
12875 To enable compilation of this filter you need to configure FFmpeg with
12876 @code{--enable-libass}. This filter also requires a build with libavcodec and
12877 libavformat to convert the passed subtitles file to ASS (Advanced Substation
12878 Alpha) subtitles format.
12880 The filter accepts the following options:
12884 Set the filename of the subtitle file to read. It must be specified.
12886 @item original_size
12887 Specify the size of the original video, the video for which the ASS file
12888 was composed. For the syntax of this option, check the
12889 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12890 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
12891 correctly scale the fonts if the aspect ratio has been changed.
12894 Set a directory path containing fonts that can be used by the filter.
12895 These fonts will be used in addition to whatever the font provider uses.
12898 Set subtitles input character encoding. @code{subtitles} filter only. Only
12899 useful if not UTF-8.
12901 @item stream_index, si
12902 Set subtitles stream index. @code{subtitles} filter only.
12905 Override default style or script info parameters of the subtitles. It accepts a
12906 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
12909 If the first key is not specified, it is assumed that the first value
12910 specifies the @option{filename}.
12912 For example, to render the file @file{sub.srt} on top of the input
12913 video, use the command:
12918 which is equivalent to:
12920 subtitles=filename=sub.srt
12923 To render the default subtitles stream from file @file{video.mkv}, use:
12925 subtitles=video.mkv
12928 To render the second subtitles stream from that file, use:
12930 subtitles=video.mkv:si=1
12933 To make the subtitles stream from @file{sub.srt} appear in transparent green
12934 @code{DejaVu Serif}, use:
12936 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
12939 @section super2xsai
12941 Scale the input by 2x and smooth using the Super2xSaI (Scale and
12942 Interpolate) pixel art scaling algorithm.
12944 Useful for enlarging pixel art images without reducing sharpness.
12948 Swap two rectangular objects in video.
12950 This filter accepts the following options:
12960 Set 1st rect x coordinate.
12963 Set 1st rect y coordinate.
12966 Set 2nd rect x coordinate.
12969 Set 2nd rect y coordinate.
12971 All expressions are evaluated once for each frame.
12974 The all options are expressions containing the following constants:
12979 The input width and height.
12982 same as @var{w} / @var{h}
12985 input sample aspect ratio
12988 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
12991 The number of the input frame, starting from 0.
12994 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
12997 the position in the file of the input frame, NAN if unknown
13005 Apply telecine process to the video.
13007 This filter accepts the following options:
13016 The default value is @code{top}.
13020 A string of numbers representing the pulldown pattern you wish to apply.
13021 The default value is @code{23}.
13025 Some typical patterns:
13030 24p: 2332 (preferred)
13037 24p: 222222222223 ("Euro pulldown")
13043 Select the most representative frame in a given sequence of consecutive frames.
13045 The filter accepts the following options:
13049 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
13050 will pick one of them, and then handle the next batch of @var{n} frames until
13051 the end. Default is @code{100}.
13054 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
13055 value will result in a higher memory usage, so a high value is not recommended.
13057 @subsection Examples
13061 Extract one picture each 50 frames:
13067 Complete example of a thumbnail creation with @command{ffmpeg}:
13069 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
13075 Tile several successive frames together.
13077 The filter accepts the following options:
13082 Set the grid size (i.e. the number of lines and columns). For the syntax of
13083 this option, check the
13084 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13087 Set the maximum number of frames to render in the given area. It must be less
13088 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
13089 the area will be used.
13092 Set the outer border margin in pixels.
13095 Set the inner border thickness (i.e. the number of pixels between frames). For
13096 more advanced padding options (such as having different values for the edges),
13097 refer to the pad video filter.
13100 Specify the color of the unused area. For the syntax of this option, check the
13101 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
13105 @subsection Examples
13109 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
13111 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
13113 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
13114 duplicating each output frame to accommodate the originally detected frame
13118 Display @code{5} pictures in an area of @code{3x2} frames,
13119 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
13120 mixed flat and named options:
13122 tile=3x2:nb_frames=5:padding=7:margin=2
13126 @section tinterlace
13128 Perform various types of temporal field interlacing.
13130 Frames are counted starting from 1, so the first input frame is
13133 The filter accepts the following options:
13138 Specify the mode of the interlacing. This option can also be specified
13139 as a value alone. See below for a list of values for this option.
13141 Available values are:
13145 Move odd frames into the upper field, even into the lower field,
13146 generating a double height frame at half frame rate.
13150 Frame 1 Frame 2 Frame 3 Frame 4
13152 11111 22222 33333 44444
13153 11111 22222 33333 44444
13154 11111 22222 33333 44444
13155 11111 22222 33333 44444
13169 Only output odd frames, even frames are dropped, generating a frame with
13170 unchanged height at half frame rate.
13175 Frame 1 Frame 2 Frame 3 Frame 4
13177 11111 22222 33333 44444
13178 11111 22222 33333 44444
13179 11111 22222 33333 44444
13180 11111 22222 33333 44444
13190 Only output even frames, odd frames are dropped, generating a frame with
13191 unchanged height at half frame rate.
13196 Frame 1 Frame 2 Frame 3 Frame 4
13198 11111 22222 33333 44444
13199 11111 22222 33333 44444
13200 11111 22222 33333 44444
13201 11111 22222 33333 44444
13211 Expand each frame to full height, but pad alternate lines with black,
13212 generating a frame with double height at the same input frame rate.
13217 Frame 1 Frame 2 Frame 3 Frame 4
13219 11111 22222 33333 44444
13220 11111 22222 33333 44444
13221 11111 22222 33333 44444
13222 11111 22222 33333 44444
13225 11111 ..... 33333 .....
13226 ..... 22222 ..... 44444
13227 11111 ..... 33333 .....
13228 ..... 22222 ..... 44444
13229 11111 ..... 33333 .....
13230 ..... 22222 ..... 44444
13231 11111 ..... 33333 .....
13232 ..... 22222 ..... 44444
13236 @item interleave_top, 4
13237 Interleave the upper field from odd frames with the lower field from
13238 even frames, generating a frame with unchanged height at half frame rate.
13243 Frame 1 Frame 2 Frame 3 Frame 4
13245 11111<- 22222 33333<- 44444
13246 11111 22222<- 33333 44444<-
13247 11111<- 22222 33333<- 44444
13248 11111 22222<- 33333 44444<-
13258 @item interleave_bottom, 5
13259 Interleave the lower field from odd frames with the upper field from
13260 even frames, generating a frame with unchanged height at half frame rate.
13265 Frame 1 Frame 2 Frame 3 Frame 4
13267 11111 22222<- 33333 44444<-
13268 11111<- 22222 33333<- 44444
13269 11111 22222<- 33333 44444<-
13270 11111<- 22222 33333<- 44444
13280 @item interlacex2, 6
13281 Double frame rate with unchanged height. Frames are inserted each
13282 containing the second temporal field from the previous input frame and
13283 the first temporal field from the next input frame. This mode relies on
13284 the top_field_first flag. Useful for interlaced video displays with no
13285 field synchronisation.
13290 Frame 1 Frame 2 Frame 3 Frame 4
13292 11111 22222 33333 44444
13293 11111 22222 33333 44444
13294 11111 22222 33333 44444
13295 11111 22222 33333 44444
13298 11111 22222 22222 33333 33333 44444 44444
13299 11111 11111 22222 22222 33333 33333 44444
13300 11111 22222 22222 33333 33333 44444 44444
13301 11111 11111 22222 22222 33333 33333 44444
13306 Move odd frames into the upper field, even into the lower field,
13307 generating a double height frame at same frame rate.
13312 Frame 1 Frame 2 Frame 3 Frame 4
13314 11111 22222 33333 44444
13315 11111 22222 33333 44444
13316 11111 22222 33333 44444
13317 11111 22222 33333 44444
13320 11111 33333 33333 55555
13321 22222 22222 44444 44444
13322 11111 33333 33333 55555
13323 22222 22222 44444 44444
13324 11111 33333 33333 55555
13325 22222 22222 44444 44444
13326 11111 33333 33333 55555
13327 22222 22222 44444 44444
13332 Numeric values are deprecated but are accepted for backward
13333 compatibility reasons.
13335 Default mode is @code{merge}.
13338 Specify flags influencing the filter process.
13340 Available value for @var{flags} is:
13343 @item low_pass_filter, vlfp
13344 Enable vertical low-pass filtering in the filter.
13345 Vertical low-pass filtering is required when creating an interlaced
13346 destination from a progressive source which contains high-frequency
13347 vertical detail. Filtering will reduce interlace 'twitter' and Moire
13350 Vertical low-pass filtering can only be enabled for @option{mode}
13351 @var{interleave_top} and @var{interleave_bottom}.
13358 Transpose rows with columns in the input video and optionally flip it.
13360 It accepts the following parameters:
13365 Specify the transposition direction.
13367 Can assume the following values:
13369 @item 0, 4, cclock_flip
13370 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
13378 Rotate by 90 degrees clockwise, that is:
13386 Rotate by 90 degrees counterclockwise, that is:
13393 @item 3, 7, clock_flip
13394 Rotate by 90 degrees clockwise and vertically flip, that is:
13402 For values between 4-7, the transposition is only done if the input
13403 video geometry is portrait and not landscape. These values are
13404 deprecated, the @code{passthrough} option should be used instead.
13406 Numerical values are deprecated, and should be dropped in favor of
13407 symbolic constants.
13410 Do not apply the transposition if the input geometry matches the one
13411 specified by the specified value. It accepts the following values:
13414 Always apply transposition.
13416 Preserve portrait geometry (when @var{height} >= @var{width}).
13418 Preserve landscape geometry (when @var{width} >= @var{height}).
13421 Default value is @code{none}.
13424 For example to rotate by 90 degrees clockwise and preserve portrait
13427 transpose=dir=1:passthrough=portrait
13430 The command above can also be specified as:
13432 transpose=1:portrait
13436 Trim the input so that the output contains one continuous subpart of the input.
13438 It accepts the following parameters:
13441 Specify the time of the start of the kept section, i.e. the frame with the
13442 timestamp @var{start} will be the first frame in the output.
13445 Specify the time of the first frame that will be dropped, i.e. the frame
13446 immediately preceding the one with the timestamp @var{end} will be the last
13447 frame in the output.
13450 This is the same as @var{start}, except this option sets the start timestamp
13451 in timebase units instead of seconds.
13454 This is the same as @var{end}, except this option sets the end timestamp
13455 in timebase units instead of seconds.
13458 The maximum duration of the output in seconds.
13461 The number of the first frame that should be passed to the output.
13464 The number of the first frame that should be dropped.
13467 @option{start}, @option{end}, and @option{duration} are expressed as time
13468 duration specifications; see
13469 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
13470 for the accepted syntax.
13472 Note that the first two sets of the start/end options and the @option{duration}
13473 option look at the frame timestamp, while the _frame variants simply count the
13474 frames that pass through the filter. Also note that this filter does not modify
13475 the timestamps. If you wish for the output timestamps to start at zero, insert a
13476 setpts filter after the trim filter.
13478 If multiple start or end options are set, this filter tries to be greedy and
13479 keep all the frames that match at least one of the specified constraints. To keep
13480 only the part that matches all the constraints at once, chain multiple trim
13483 The defaults are such that all the input is kept. So it is possible to set e.g.
13484 just the end values to keep everything before the specified time.
13489 Drop everything except the second minute of input:
13491 ffmpeg -i INPUT -vf trim=60:120
13495 Keep only the first second:
13497 ffmpeg -i INPUT -vf trim=duration=1
13506 Sharpen or blur the input video.
13508 It accepts the following parameters:
13511 @item luma_msize_x, lx
13512 Set the luma matrix horizontal size. It must be an odd integer between
13513 3 and 23. The default value is 5.
13515 @item luma_msize_y, ly
13516 Set the luma matrix vertical size. It must be an odd integer between 3
13517 and 23. The default value is 5.
13519 @item luma_amount, la
13520 Set the luma effect strength. It must be a floating point number, reasonable
13521 values lay between -1.5 and 1.5.
13523 Negative values will blur the input video, while positive values will
13524 sharpen it, a value of zero will disable the effect.
13526 Default value is 1.0.
13528 @item chroma_msize_x, cx
13529 Set the chroma matrix horizontal size. It must be an odd integer
13530 between 3 and 23. The default value is 5.
13532 @item chroma_msize_y, cy
13533 Set the chroma matrix vertical size. It must be an odd integer
13534 between 3 and 23. The default value is 5.
13536 @item chroma_amount, ca
13537 Set the chroma effect strength. It must be a floating point number, reasonable
13538 values lay between -1.5 and 1.5.
13540 Negative values will blur the input video, while positive values will
13541 sharpen it, a value of zero will disable the effect.
13543 Default value is 0.0.
13546 If set to 1, specify using OpenCL capabilities, only available if
13547 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
13551 All parameters are optional and default to the equivalent of the
13552 string '5:5:1.0:5:5:0.0'.
13554 @subsection Examples
13558 Apply strong luma sharpen effect:
13560 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
13564 Apply a strong blur of both luma and chroma parameters:
13566 unsharp=7:7:-2:7:7:-2
13572 Apply ultra slow/simple postprocessing filter that compresses and decompresses
13573 the image at several (or - in the case of @option{quality} level @code{8} - all)
13574 shifts and average the results.
13576 The way this differs from the behavior of spp is that uspp actually encodes &
13577 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
13578 DCT similar to MJPEG.
13580 The filter accepts the following options:
13584 Set quality. This option defines the number of levels for averaging. It accepts
13585 an integer in the range 0-8. If set to @code{0}, the filter will have no
13586 effect. A value of @code{8} means the higher quality. For each increment of
13587 that value the speed drops by a factor of approximately 2. Default value is
13591 Force a constant quantization parameter. If not set, the filter will use the QP
13592 from the video stream (if available).
13595 @section vaguedenoiser
13597 Apply a wavelet based denoiser.
13599 It transforms each frame from the video input into the wavelet domain,
13600 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
13601 the obtained coefficients. It does an inverse wavelet transform after.
13602 Due to wavelet properties, it should give a nice smoothed result, and
13603 reduced noise, without blurring picture features.
13605 This filter accepts the following options:
13609 The filtering strength. The higher, the more filtered the video will be.
13610 Hard thresholding can use a higher threshold than soft thresholding
13611 before the video looks overfiltered.
13614 The filtering method the filter will use.
13616 It accepts the following values:
13619 All values under the threshold will be zeroed.
13622 All values under the threshold will be zeroed. All values above will be
13623 reduced by the threshold.
13626 Scales or nullifies coefficients - intermediary between (more) soft and
13627 (less) hard thresholding.
13631 Number of times, the wavelet will decompose the picture. Picture can't
13632 be decomposed beyond a particular point (typically, 8 for a 640x480
13633 frame - as 2^9 = 512 > 480)
13636 Partial of full denoising (limited coefficients shrinking), from 0 to 100.
13639 A list of the planes to process. By default all planes are processed.
13642 @section vectorscope
13644 Display 2 color component values in the two dimensional graph (which is called
13647 This filter accepts the following options:
13651 Set vectorscope mode.
13653 It accepts the following values:
13656 Gray values are displayed on graph, higher brightness means more pixels have
13657 same component color value on location in graph. This is the default mode.
13660 Gray values are displayed on graph. Surrounding pixels values which are not
13661 present in video frame are drawn in gradient of 2 color components which are
13662 set by option @code{x} and @code{y}. The 3rd color component is static.
13665 Actual color components values present in video frame are displayed on graph.
13668 Similar as color2 but higher frequency of same values @code{x} and @code{y}
13669 on graph increases value of another color component, which is luminance by
13670 default values of @code{x} and @code{y}.
13673 Actual colors present in video frame are displayed on graph. If two different
13674 colors map to same position on graph then color with higher value of component
13675 not present in graph is picked.
13678 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
13679 component picked from radial gradient.
13683 Set which color component will be represented on X-axis. Default is @code{1}.
13686 Set which color component will be represented on Y-axis. Default is @code{2}.
13689 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
13690 of color component which represents frequency of (X, Y) location in graph.
13695 No envelope, this is default.
13698 Instant envelope, even darkest single pixel will be clearly highlighted.
13701 Hold maximum and minimum values presented in graph over time. This way you
13702 can still spot out of range values without constantly looking at vectorscope.
13705 Peak and instant envelope combined together.
13709 Set what kind of graticule to draw.
13717 Set graticule opacity.
13720 Set graticule flags.
13724 Draw graticule for white point.
13727 Draw graticule for black point.
13730 Draw color points short names.
13734 Set background opacity.
13736 @item lthreshold, l
13737 Set low threshold for color component not represented on X or Y axis.
13738 Values lower than this value will be ignored. Default is 0.
13739 Note this value is multiplied with actual max possible value one pixel component
13740 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
13743 @item hthreshold, h
13744 Set high threshold for color component not represented on X or Y axis.
13745 Values higher than this value will be ignored. Default is 1.
13746 Note this value is multiplied with actual max possible value one pixel component
13747 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
13748 is 0.9 * 255 = 230.
13750 @item colorspace, c
13751 Set what kind of colorspace to use when drawing graticule.
13760 @anchor{vidstabdetect}
13761 @section vidstabdetect
13763 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
13764 @ref{vidstabtransform} for pass 2.
13766 This filter generates a file with relative translation and rotation
13767 transform information about subsequent frames, which is then used by
13768 the @ref{vidstabtransform} filter.
13770 To enable compilation of this filter you need to configure FFmpeg with
13771 @code{--enable-libvidstab}.
13773 This filter accepts the following options:
13777 Set the path to the file used to write the transforms information.
13778 Default value is @file{transforms.trf}.
13781 Set how shaky the video is and how quick the camera is. It accepts an
13782 integer in the range 1-10, a value of 1 means little shakiness, a
13783 value of 10 means strong shakiness. Default value is 5.
13786 Set the accuracy of the detection process. It must be a value in the
13787 range 1-15. A value of 1 means low accuracy, a value of 15 means high
13788 accuracy. Default value is 15.
13791 Set stepsize of the search process. The region around minimum is
13792 scanned with 1 pixel resolution. Default value is 6.
13795 Set minimum contrast. Below this value a local measurement field is
13796 discarded. Must be a floating point value in the range 0-1. Default
13800 Set reference frame number for tripod mode.
13802 If enabled, the motion of the frames is compared to a reference frame
13803 in the filtered stream, identified by the specified number. The idea
13804 is to compensate all movements in a more-or-less static scene and keep
13805 the camera view absolutely still.
13807 If set to 0, it is disabled. The frames are counted starting from 1.
13810 Show fields and transforms in the resulting frames. It accepts an
13811 integer in the range 0-2. Default value is 0, which disables any
13815 @subsection Examples
13819 Use default values:
13825 Analyze strongly shaky movie and put the results in file
13826 @file{mytransforms.trf}:
13828 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
13832 Visualize the result of internal transformations in the resulting
13835 vidstabdetect=show=1
13839 Analyze a video with medium shakiness using @command{ffmpeg}:
13841 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
13845 @anchor{vidstabtransform}
13846 @section vidstabtransform
13848 Video stabilization/deshaking: pass 2 of 2,
13849 see @ref{vidstabdetect} for pass 1.
13851 Read a file with transform information for each frame and
13852 apply/compensate them. Together with the @ref{vidstabdetect}
13853 filter this can be used to deshake videos. See also
13854 @url{http://public.hronopik.de/vid.stab}. It is important to also use
13855 the @ref{unsharp} filter, see below.
13857 To enable compilation of this filter you need to configure FFmpeg with
13858 @code{--enable-libvidstab}.
13860 @subsection Options
13864 Set path to the file used to read the transforms. Default value is
13865 @file{transforms.trf}.
13868 Set the number of frames (value*2 + 1) used for lowpass filtering the
13869 camera movements. Default value is 10.
13871 For example a number of 10 means that 21 frames are used (10 in the
13872 past and 10 in the future) to smoothen the motion in the video. A
13873 larger value leads to a smoother video, but limits the acceleration of
13874 the camera (pan/tilt movements). 0 is a special case where a static
13875 camera is simulated.
13878 Set the camera path optimization algorithm.
13880 Accepted values are:
13883 gaussian kernel low-pass filter on camera motion (default)
13885 averaging on transformations
13889 Set maximal number of pixels to translate frames. Default value is -1,
13893 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
13894 value is -1, meaning no limit.
13897 Specify how to deal with borders that may be visible due to movement
13900 Available values are:
13903 keep image information from previous frame (default)
13905 fill the border black
13909 Invert transforms if set to 1. Default value is 0.
13912 Consider transforms as relative to previous frame if set to 1,
13913 absolute if set to 0. Default value is 0.
13916 Set percentage to zoom. A positive value will result in a zoom-in
13917 effect, a negative value in a zoom-out effect. Default value is 0 (no
13921 Set optimal zooming to avoid borders.
13923 Accepted values are:
13928 optimal static zoom value is determined (only very strong movements
13929 will lead to visible borders) (default)
13931 optimal adaptive zoom value is determined (no borders will be
13932 visible), see @option{zoomspeed}
13935 Note that the value given at zoom is added to the one calculated here.
13938 Set percent to zoom maximally each frame (enabled when
13939 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
13943 Specify type of interpolation.
13945 Available values are:
13950 linear only horizontal
13952 linear in both directions (default)
13954 cubic in both directions (slow)
13958 Enable virtual tripod mode if set to 1, which is equivalent to
13959 @code{relative=0:smoothing=0}. Default value is 0.
13961 Use also @code{tripod} option of @ref{vidstabdetect}.
13964 Increase log verbosity if set to 1. Also the detected global motions
13965 are written to the temporary file @file{global_motions.trf}. Default
13969 @subsection Examples
13973 Use @command{ffmpeg} for a typical stabilization with default values:
13975 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
13978 Note the use of the @ref{unsharp} filter which is always recommended.
13981 Zoom in a bit more and load transform data from a given file:
13983 vidstabtransform=zoom=5:input="mytransforms.trf"
13987 Smoothen the video even more:
13989 vidstabtransform=smoothing=30
13995 Flip the input video vertically.
13997 For example, to vertically flip a video with @command{ffmpeg}:
13999 ffmpeg -i in.avi -vf "vflip" out.avi
14005 Make or reverse a natural vignetting effect.
14007 The filter accepts the following options:
14011 Set lens angle expression as a number of radians.
14013 The value is clipped in the @code{[0,PI/2]} range.
14015 Default value: @code{"PI/5"}
14019 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
14023 Set forward/backward mode.
14025 Available modes are:
14028 The larger the distance from the central point, the darker the image becomes.
14031 The larger the distance from the central point, the brighter the image becomes.
14032 This can be used to reverse a vignette effect, though there is no automatic
14033 detection to extract the lens @option{angle} and other settings (yet). It can
14034 also be used to create a burning effect.
14037 Default value is @samp{forward}.
14040 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
14042 It accepts the following values:
14045 Evaluate expressions only once during the filter initialization.
14048 Evaluate expressions for each incoming frame. This is way slower than the
14049 @samp{init} mode since it requires all the scalers to be re-computed, but it
14050 allows advanced dynamic expressions.
14053 Default value is @samp{init}.
14056 Set dithering to reduce the circular banding effects. Default is @code{1}
14060 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
14061 Setting this value to the SAR of the input will make a rectangular vignetting
14062 following the dimensions of the video.
14064 Default is @code{1/1}.
14067 @subsection Expressions
14069 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
14070 following parameters.
14075 input width and height
14078 the number of input frame, starting from 0
14081 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
14082 @var{TB} units, NAN if undefined
14085 frame rate of the input video, NAN if the input frame rate is unknown
14088 the PTS (Presentation TimeStamp) of the filtered video frame,
14089 expressed in seconds, NAN if undefined
14092 time base of the input video
14096 @subsection Examples
14100 Apply simple strong vignetting effect:
14106 Make a flickering vignetting:
14108 vignette='PI/4+random(1)*PI/50':eval=frame
14114 Stack input videos vertically.
14116 All streams must be of same pixel format and of same width.
14118 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
14119 to create same output.
14121 The filter accept the following option:
14125 Set number of input streams. Default is 2.
14128 If set to 1, force the output to terminate when the shortest input
14129 terminates. Default value is 0.
14134 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
14135 Deinterlacing Filter").
14137 Based on the process described by Martin Weston for BBC R&D, and
14138 implemented based on the de-interlace algorithm written by Jim
14139 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
14140 uses filter coefficients calculated by BBC R&D.
14142 There are two sets of filter coefficients, so called "simple":
14143 and "complex". Which set of filter coefficients is used can
14144 be set by passing an optional parameter:
14148 Set the interlacing filter coefficients. Accepts one of the following values:
14152 Simple filter coefficient set.
14154 More-complex filter coefficient set.
14156 Default value is @samp{complex}.
14159 Specify which frames to deinterlace. Accept one of the following values:
14163 Deinterlace all frames,
14165 Only deinterlace frames marked as interlaced.
14168 Default value is @samp{all}.
14172 Video waveform monitor.
14174 The waveform monitor plots color component intensity. By default luminance
14175 only. Each column of the waveform corresponds to a column of pixels in the
14178 It accepts the following options:
14182 Can be either @code{row}, or @code{column}. Default is @code{column}.
14183 In row mode, the graph on the left side represents color component value 0 and
14184 the right side represents value = 255. In column mode, the top side represents
14185 color component value = 0 and bottom side represents value = 255.
14188 Set intensity. Smaller values are useful to find out how many values of the same
14189 luminance are distributed across input rows/columns.
14190 Default value is @code{0.04}. Allowed range is [0, 1].
14193 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
14194 In mirrored mode, higher values will be represented on the left
14195 side for @code{row} mode and at the top for @code{column} mode. Default is
14196 @code{1} (mirrored).
14200 It accepts the following values:
14203 Presents information identical to that in the @code{parade}, except
14204 that the graphs representing color components are superimposed directly
14207 This display mode makes it easier to spot relative differences or similarities
14208 in overlapping areas of the color components that are supposed to be identical,
14209 such as neutral whites, grays, or blacks.
14212 Display separate graph for the color components side by side in
14213 @code{row} mode or one below the other in @code{column} mode.
14216 Display separate graph for the color components side by side in
14217 @code{column} mode or one below the other in @code{row} mode.
14219 Using this display mode makes it easy to spot color casts in the highlights
14220 and shadows of an image, by comparing the contours of the top and the bottom
14221 graphs of each waveform. Since whites, grays, and blacks are characterized
14222 by exactly equal amounts of red, green, and blue, neutral areas of the picture
14223 should display three waveforms of roughly equal width/height. If not, the
14224 correction is easy to perform by making level adjustments the three waveforms.
14226 Default is @code{stack}.
14228 @item components, c
14229 Set which color components to display. Default is 1, which means only luminance
14230 or red color component if input is in RGB colorspace. If is set for example to
14231 7 it will display all 3 (if) available color components.
14236 No envelope, this is default.
14239 Instant envelope, minimum and maximum values presented in graph will be easily
14240 visible even with small @code{step} value.
14243 Hold minimum and maximum values presented in graph across time. This way you
14244 can still spot out of range values without constantly looking at waveforms.
14247 Peak and instant envelope combined together.
14253 No filtering, this is default.
14256 Luma and chroma combined together.
14259 Similar as above, but shows difference between blue and red chroma.
14262 Displays only chroma.
14265 Displays actual color value on waveform.
14268 Similar as above, but with luma showing frequency of chroma values.
14272 Set which graticule to display.
14276 Do not display graticule.
14279 Display green graticule showing legal broadcast ranges.
14283 Set graticule opacity.
14286 Set graticule flags.
14290 Draw numbers above lines. By default enabled.
14293 Draw dots instead of lines.
14297 Set scale used for displaying graticule.
14304 Default is digital.
14307 Set background opacity.
14312 The @code{weave} takes a field-based video input and join
14313 each two sequential fields into single frame, producing a new double
14314 height clip with half the frame rate and half the frame count.
14316 It accepts the following option:
14320 Set first field. Available values are:
14324 Set the frame as top-field-first.
14327 Set the frame as bottom-field-first.
14331 @subsection Examples
14335 Interlace video using @ref{select} and @ref{separatefields} filter:
14337 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
14342 Apply the xBR high-quality magnification filter which is designed for pixel
14343 art. It follows a set of edge-detection rules, see
14344 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
14346 It accepts the following option:
14350 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
14351 @code{3xBR} and @code{4} for @code{4xBR}.
14352 Default is @code{3}.
14358 Deinterlace the input video ("yadif" means "yet another deinterlacing
14361 It accepts the following parameters:
14367 The interlacing mode to adopt. It accepts one of the following values:
14370 @item 0, send_frame
14371 Output one frame for each frame.
14372 @item 1, send_field
14373 Output one frame for each field.
14374 @item 2, send_frame_nospatial
14375 Like @code{send_frame}, but it skips the spatial interlacing check.
14376 @item 3, send_field_nospatial
14377 Like @code{send_field}, but it skips the spatial interlacing check.
14380 The default value is @code{send_frame}.
14383 The picture field parity assumed for the input interlaced video. It accepts one
14384 of the following values:
14388 Assume the top field is first.
14390 Assume the bottom field is first.
14392 Enable automatic detection of field parity.
14395 The default value is @code{auto}.
14396 If the interlacing is unknown or the decoder does not export this information,
14397 top field first will be assumed.
14400 Specify which frames to deinterlace. Accept one of the following
14405 Deinterlace all frames.
14406 @item 1, interlaced
14407 Only deinterlace frames marked as interlaced.
14410 The default value is @code{all}.
14415 Apply Zoom & Pan effect.
14417 This filter accepts the following options:
14421 Set the zoom expression. Default is 1.
14425 Set the x and y expression. Default is 0.
14428 Set the duration expression in number of frames.
14429 This sets for how many number of frames effect will last for
14430 single input image.
14433 Set the output image size, default is 'hd720'.
14436 Set the output frame rate, default is '25'.
14439 Each expression can contain the following constants:
14458 Output frame count.
14462 Last calculated 'x' and 'y' position from 'x' and 'y' expression
14463 for current input frame.
14467 'x' and 'y' of last output frame of previous input frame or 0 when there was
14468 not yet such frame (first input frame).
14471 Last calculated zoom from 'z' expression for current input frame.
14474 Last calculated zoom of last output frame of previous input frame.
14477 Number of output frames for current input frame. Calculated from 'd' expression
14478 for each input frame.
14481 number of output frames created for previous input frame
14484 Rational number: input width / input height
14487 sample aspect ratio
14490 display aspect ratio
14494 @subsection Examples
14498 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
14500 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
14504 Zoom-in up to 1.5 and pan always at center of picture:
14506 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14510 Same as above but without pausing:
14512 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
14517 Scale (resize) the input video, using the z.lib library:
14518 https://github.com/sekrit-twc/zimg.
14520 The zscale filter forces the output display aspect ratio to be the same
14521 as the input, by changing the output sample aspect ratio.
14523 If the input image format is different from the format requested by
14524 the next filter, the zscale filter will convert the input to the
14527 @subsection Options
14528 The filter accepts the following options.
14533 Set the output video dimension expression. Default value is the input
14536 If the @var{width} or @var{w} is 0, the input width is used for the output.
14537 If the @var{height} or @var{h} is 0, the input height is used for the output.
14539 If one of the values is -1, the zscale filter will use a value that
14540 maintains the aspect ratio of the input image, calculated from the
14541 other specified dimension. If both of them are -1, the input size is
14544 If one of the values is -n with n > 1, the zscale filter will also use a value
14545 that maintains the aspect ratio of the input image, calculated from the other
14546 specified dimension. After that it will, however, make sure that the calculated
14547 dimension is divisible by n and adjust the value if necessary.
14549 See below for the list of accepted constants for use in the dimension
14553 Set the video size. For the syntax of this option, check the
14554 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14557 Set the dither type.
14559 Possible values are:
14564 @item error_diffusion
14570 Set the resize filter type.
14572 Possible values are:
14582 Default is bilinear.
14585 Set the color range.
14587 Possible values are:
14594 Default is same as input.
14597 Set the color primaries.
14599 Possible values are:
14609 Default is same as input.
14612 Set the transfer characteristics.
14614 Possible values are:
14625 Default is same as input.
14628 Set the colorspace matrix.
14630 Possible value are:
14641 Default is same as input.
14644 Set the input color range.
14646 Possible values are:
14653 Default is same as input.
14655 @item primariesin, pin
14656 Set the input color primaries.
14658 Possible values are:
14668 Default is same as input.
14670 @item transferin, tin
14671 Set the input transfer characteristics.
14673 Possible values are:
14684 Default is same as input.
14686 @item matrixin, min
14687 Set the input colorspace matrix.
14689 Possible value are:
14701 Set the output chroma location.
14703 Possible values are:
14714 @item chromalin, cin
14715 Set the input chroma location.
14717 Possible values are:
14729 The values of the @option{w} and @option{h} options are expressions
14730 containing the following constants:
14735 The input width and height
14739 These are the same as @var{in_w} and @var{in_h}.
14743 The output (scaled) width and height
14747 These are the same as @var{out_w} and @var{out_h}
14750 The same as @var{iw} / @var{ih}
14753 input sample aspect ratio
14756 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14760 horizontal and vertical input chroma subsample values. For example for the
14761 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14765 horizontal and vertical output chroma subsample values. For example for the
14766 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14772 @c man end VIDEO FILTERS
14774 @chapter Video Sources
14775 @c man begin VIDEO SOURCES
14777 Below is a description of the currently available video sources.
14781 Buffer video frames, and make them available to the filter chain.
14783 This source is mainly intended for a programmatic use, in particular
14784 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
14786 It accepts the following parameters:
14791 Specify the size (width and height) of the buffered video frames. For the
14792 syntax of this option, check the
14793 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14796 The input video width.
14799 The input video height.
14802 A string representing the pixel format of the buffered video frames.
14803 It may be a number corresponding to a pixel format, or a pixel format
14807 Specify the timebase assumed by the timestamps of the buffered frames.
14810 Specify the frame rate expected for the video stream.
14812 @item pixel_aspect, sar
14813 The sample (pixel) aspect ratio of the input video.
14816 Specify the optional parameters to be used for the scale filter which
14817 is automatically inserted when an input change is detected in the
14818 input size or format.
14820 @item hw_frames_ctx
14821 When using a hardware pixel format, this should be a reference to an
14822 AVHWFramesContext describing input frames.
14827 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
14830 will instruct the source to accept video frames with size 320x240 and
14831 with format "yuv410p", assuming 1/24 as the timestamps timebase and
14832 square pixels (1:1 sample aspect ratio).
14833 Since the pixel format with name "yuv410p" corresponds to the number 6
14834 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
14835 this example corresponds to:
14837 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
14840 Alternatively, the options can be specified as a flat string, but this
14841 syntax is deprecated:
14843 @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}]
14847 Create a pattern generated by an elementary cellular automaton.
14849 The initial state of the cellular automaton can be defined through the
14850 @option{filename} and @option{pattern} options. If such options are
14851 not specified an initial state is created randomly.
14853 At each new frame a new row in the video is filled with the result of
14854 the cellular automaton next generation. The behavior when the whole
14855 frame is filled is defined by the @option{scroll} option.
14857 This source accepts the following options:
14861 Read the initial cellular automaton state, i.e. the starting row, from
14862 the specified file.
14863 In the file, each non-whitespace character is considered an alive
14864 cell, a newline will terminate the row, and further characters in the
14865 file will be ignored.
14868 Read the initial cellular automaton state, i.e. the starting row, from
14869 the specified string.
14871 Each non-whitespace character in the string is considered an alive
14872 cell, a newline will terminate the row, and further characters in the
14873 string will be ignored.
14876 Set the video rate, that is the number of frames generated per second.
14879 @item random_fill_ratio, ratio
14880 Set the random fill ratio for the initial cellular automaton row. It
14881 is a floating point number value ranging from 0 to 1, defaults to
14884 This option is ignored when a file or a pattern is specified.
14886 @item random_seed, seed
14887 Set the seed for filling randomly the initial row, must be an integer
14888 included between 0 and UINT32_MAX. If not specified, or if explicitly
14889 set to -1, the filter will try to use a good random seed on a best
14893 Set the cellular automaton rule, it is a number ranging from 0 to 255.
14894 Default value is 110.
14897 Set the size of the output video. For the syntax of this option, check the
14898 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14900 If @option{filename} or @option{pattern} is specified, the size is set
14901 by default to the width of the specified initial state row, and the
14902 height is set to @var{width} * PHI.
14904 If @option{size} is set, it must contain the width of the specified
14905 pattern string, and the specified pattern will be centered in the
14908 If a filename or a pattern string is not specified, the size value
14909 defaults to "320x518" (used for a randomly generated initial state).
14912 If set to 1, scroll the output upward when all the rows in the output
14913 have been already filled. If set to 0, the new generated row will be
14914 written over the top row just after the bottom row is filled.
14917 @item start_full, full
14918 If set to 1, completely fill the output with generated rows before
14919 outputting the first frame.
14920 This is the default behavior, for disabling set the value to 0.
14923 If set to 1, stitch the left and right row edges together.
14924 This is the default behavior, for disabling set the value to 0.
14927 @subsection Examples
14931 Read the initial state from @file{pattern}, and specify an output of
14934 cellauto=f=pattern:s=200x400
14938 Generate a random initial row with a width of 200 cells, with a fill
14941 cellauto=ratio=2/3:s=200x200
14945 Create a pattern generated by rule 18 starting by a single alive cell
14946 centered on an initial row with width 100:
14948 cellauto=p=@@:s=100x400:full=0:rule=18
14952 Specify a more elaborated initial pattern:
14954 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
14959 @anchor{coreimagesrc}
14960 @section coreimagesrc
14961 Video source generated on GPU using Apple's CoreImage API on OSX.
14963 This video source is a specialized version of the @ref{coreimage} video filter.
14964 Use a core image generator at the beginning of the applied filterchain to
14965 generate the content.
14967 The coreimagesrc video source accepts the following options:
14969 @item list_generators
14970 List all available generators along with all their respective options as well as
14971 possible minimum and maximum values along with the default values.
14973 list_generators=true
14977 Specify the size of the sourced video. For the syntax of this option, check the
14978 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14979 The default value is @code{320x240}.
14982 Specify the frame rate of the sourced video, as the number of frames
14983 generated per second. It has to be a string in the format
14984 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
14985 number or a valid video frame rate abbreviation. The default value is
14989 Set the sample aspect ratio of the sourced video.
14992 Set the duration of the sourced video. See
14993 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14994 for the accepted syntax.
14996 If not specified, or the expressed duration is negative, the video is
14997 supposed to be generated forever.
15000 Additionally, all options of the @ref{coreimage} video filter are accepted.
15001 A complete filterchain can be used for further processing of the
15002 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
15003 and examples for details.
15005 @subsection Examples
15010 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
15011 given as complete and escaped command-line for Apple's standard bash shell:
15013 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
15015 This example is equivalent to the QRCode example of @ref{coreimage} without the
15016 need for a nullsrc video source.
15020 @section mandelbrot
15022 Generate a Mandelbrot set fractal, and progressively zoom towards the
15023 point specified with @var{start_x} and @var{start_y}.
15025 This source accepts the following options:
15030 Set the terminal pts value. Default value is 400.
15033 Set the terminal scale value.
15034 Must be a floating point value. Default value is 0.3.
15037 Set the inner coloring mode, that is the algorithm used to draw the
15038 Mandelbrot fractal internal region.
15040 It shall assume one of the following values:
15045 Show time until convergence.
15047 Set color based on point closest to the origin of the iterations.
15052 Default value is @var{mincol}.
15055 Set the bailout value. Default value is 10.0.
15058 Set the maximum of iterations performed by the rendering
15059 algorithm. Default value is 7189.
15062 Set outer coloring mode.
15063 It shall assume one of following values:
15065 @item iteration_count
15066 Set iteration cound mode.
15067 @item normalized_iteration_count
15068 set normalized iteration count mode.
15070 Default value is @var{normalized_iteration_count}.
15073 Set frame rate, expressed as number of frames per second. Default
15077 Set frame size. For the syntax of this option, check the "Video
15078 size" section in the ffmpeg-utils manual. Default value is "640x480".
15081 Set the initial scale value. Default value is 3.0.
15084 Set the initial x position. Must be a floating point value between
15085 -100 and 100. Default value is -0.743643887037158704752191506114774.
15088 Set the initial y position. Must be a floating point value between
15089 -100 and 100. Default value is -0.131825904205311970493132056385139.
15094 Generate various test patterns, as generated by the MPlayer test filter.
15096 The size of the generated video is fixed, and is 256x256.
15097 This source is useful in particular for testing encoding features.
15099 This source accepts the following options:
15104 Specify the frame rate of the sourced video, as the number of frames
15105 generated per second. It has to be a string in the format
15106 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15107 number or a valid video frame rate abbreviation. The default value is
15111 Set the duration of the sourced video. See
15112 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15113 for the accepted syntax.
15115 If not specified, or the expressed duration is negative, the video is
15116 supposed to be generated forever.
15120 Set the number or the name of the test to perform. Supported tests are:
15136 Default value is "all", which will cycle through the list of all tests.
15141 mptestsrc=t=dc_luma
15144 will generate a "dc_luma" test pattern.
15146 @section frei0r_src
15148 Provide a frei0r source.
15150 To enable compilation of this filter you need to install the frei0r
15151 header and configure FFmpeg with @code{--enable-frei0r}.
15153 This source accepts the following parameters:
15158 The size of the video to generate. For the syntax of this option, check the
15159 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15162 The framerate of the generated video. It may be a string of the form
15163 @var{num}/@var{den} or a frame rate abbreviation.
15166 The name to the frei0r source to load. For more information regarding frei0r and
15167 how to set the parameters, read the @ref{frei0r} section in the video filters
15170 @item filter_params
15171 A '|'-separated list of parameters to pass to the frei0r source.
15175 For example, to generate a frei0r partik0l source with size 200x200
15176 and frame rate 10 which is overlaid on the overlay filter main input:
15178 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
15183 Generate a life pattern.
15185 This source is based on a generalization of John Conway's life game.
15187 The sourced input represents a life grid, each pixel represents a cell
15188 which can be in one of two possible states, alive or dead. Every cell
15189 interacts with its eight neighbours, which are the cells that are
15190 horizontally, vertically, or diagonally adjacent.
15192 At each interaction the grid evolves according to the adopted rule,
15193 which specifies the number of neighbor alive cells which will make a
15194 cell stay alive or born. The @option{rule} option allows one to specify
15197 This source accepts the following options:
15201 Set the file from which to read the initial grid state. In the file,
15202 each non-whitespace character is considered an alive cell, and newline
15203 is used to delimit the end of each row.
15205 If this option is not specified, the initial grid is generated
15209 Set the video rate, that is the number of frames generated per second.
15212 @item random_fill_ratio, ratio
15213 Set the random fill ratio for the initial random grid. It is a
15214 floating point number value ranging from 0 to 1, defaults to 1/PHI.
15215 It is ignored when a file is specified.
15217 @item random_seed, seed
15218 Set the seed for filling the initial random grid, must be an integer
15219 included between 0 and UINT32_MAX. If not specified, or if explicitly
15220 set to -1, the filter will try to use a good random seed on a best
15226 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
15227 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
15228 @var{NS} specifies the number of alive neighbor cells which make a
15229 live cell stay alive, and @var{NB} the number of alive neighbor cells
15230 which make a dead cell to become alive (i.e. to "born").
15231 "s" and "b" can be used in place of "S" and "B", respectively.
15233 Alternatively a rule can be specified by an 18-bits integer. The 9
15234 high order bits are used to encode the next cell state if it is alive
15235 for each number of neighbor alive cells, the low order bits specify
15236 the rule for "borning" new cells. Higher order bits encode for an
15237 higher number of neighbor cells.
15238 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
15239 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
15241 Default value is "S23/B3", which is the original Conway's game of life
15242 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
15243 cells, and will born a new cell if there are three alive cells around
15247 Set the size of the output video. For the syntax of this option, check the
15248 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15250 If @option{filename} is specified, the size is set by default to the
15251 same size of the input file. If @option{size} is set, it must contain
15252 the size specified in the input file, and the initial grid defined in
15253 that file is centered in the larger resulting area.
15255 If a filename is not specified, the size value defaults to "320x240"
15256 (used for a randomly generated initial grid).
15259 If set to 1, stitch the left and right grid edges together, and the
15260 top and bottom edges also. Defaults to 1.
15263 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
15264 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
15265 value from 0 to 255.
15268 Set the color of living (or new born) cells.
15271 Set the color of dead cells. If @option{mold} is set, this is the first color
15272 used to represent a dead cell.
15275 Set mold color, for definitely dead and moldy cells.
15277 For the syntax of these 3 color options, check the "Color" section in the
15278 ffmpeg-utils manual.
15281 @subsection Examples
15285 Read a grid from @file{pattern}, and center it on a grid of size
15288 life=f=pattern:s=300x300
15292 Generate a random grid of size 200x200, with a fill ratio of 2/3:
15294 life=ratio=2/3:s=200x200
15298 Specify a custom rule for evolving a randomly generated grid:
15304 Full example with slow death effect (mold) using @command{ffplay}:
15306 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
15313 @anchor{haldclutsrc}
15315 @anchor{rgbtestsrc}
15317 @anchor{smptehdbars}
15320 @anchor{yuvtestsrc}
15321 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
15323 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
15325 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
15327 The @code{color} source provides an uniformly colored input.
15329 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
15330 @ref{haldclut} filter.
15332 The @code{nullsrc} source returns unprocessed video frames. It is
15333 mainly useful to be employed in analysis / debugging tools, or as the
15334 source for filters which ignore the input data.
15336 The @code{rgbtestsrc} source generates an RGB test pattern useful for
15337 detecting RGB vs BGR issues. You should see a red, green and blue
15338 stripe from top to bottom.
15340 The @code{smptebars} source generates a color bars pattern, based on
15341 the SMPTE Engineering Guideline EG 1-1990.
15343 The @code{smptehdbars} source generates a color bars pattern, based on
15344 the SMPTE RP 219-2002.
15346 The @code{testsrc} source generates a test video pattern, showing a
15347 color pattern, a scrolling gradient and a timestamp. This is mainly
15348 intended for testing purposes.
15350 The @code{testsrc2} source is similar to testsrc, but supports more
15351 pixel formats instead of just @code{rgb24}. This allows using it as an
15352 input for other tests without requiring a format conversion.
15354 The @code{yuvtestsrc} source generates an YUV test pattern. You should
15355 see a y, cb and cr stripe from top to bottom.
15357 The sources accept the following parameters:
15362 Specify the color of the source, only available in the @code{color}
15363 source. For the syntax of this option, check the "Color" section in the
15364 ffmpeg-utils manual.
15367 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
15368 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
15369 pixels to be used as identity matrix for 3D lookup tables. Each component is
15370 coded on a @code{1/(N*N)} scale.
15373 Specify the size of the sourced video. For the syntax of this option, check the
15374 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15375 The default value is @code{320x240}.
15377 This option is not available with the @code{haldclutsrc} filter.
15380 Specify the frame rate of the sourced video, as the number of frames
15381 generated per second. It has to be a string in the format
15382 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15383 number or a valid video frame rate abbreviation. The default value is
15387 Set the sample aspect ratio of the sourced video.
15390 Set the duration of the sourced video. See
15391 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15392 for the accepted syntax.
15394 If not specified, or the expressed duration is negative, the video is
15395 supposed to be generated forever.
15398 Set the number of decimals to show in the timestamp, only available in the
15399 @code{testsrc} source.
15401 The displayed timestamp value will correspond to the original
15402 timestamp value multiplied by the power of 10 of the specified
15403 value. Default value is 0.
15406 For example the following:
15408 testsrc=duration=5.3:size=qcif:rate=10
15411 will generate a video with a duration of 5.3 seconds, with size
15412 176x144 and a frame rate of 10 frames per second.
15414 The following graph description will generate a red source
15415 with an opacity of 0.2, with size "qcif" and a frame rate of 10
15418 color=c=red@@0.2:s=qcif:r=10
15421 If the input content is to be ignored, @code{nullsrc} can be used. The
15422 following command generates noise in the luminance plane by employing
15423 the @code{geq} filter:
15425 nullsrc=s=256x256, geq=random(1)*255:128:128
15428 @subsection Commands
15430 The @code{color} source supports the following commands:
15434 Set the color of the created image. Accepts the same syntax of the
15435 corresponding @option{color} option.
15438 @c man end VIDEO SOURCES
15440 @chapter Video Sinks
15441 @c man begin VIDEO SINKS
15443 Below is a description of the currently available video sinks.
15445 @section buffersink
15447 Buffer video frames, and make them available to the end of the filter
15450 This sink is mainly intended for programmatic use, in particular
15451 through the interface defined in @file{libavfilter/buffersink.h}
15452 or the options system.
15454 It accepts a pointer to an AVBufferSinkContext structure, which
15455 defines the incoming buffers' formats, to be passed as the opaque
15456 parameter to @code{avfilter_init_filter} for initialization.
15460 Null video sink: do absolutely nothing with the input video. It is
15461 mainly useful as a template and for use in analysis / debugging
15464 @c man end VIDEO SINKS
15466 @chapter Multimedia Filters
15467 @c man begin MULTIMEDIA FILTERS
15469 Below is a description of the currently available multimedia filters.
15471 @section ahistogram
15473 Convert input audio to a video output, displaying the volume histogram.
15475 The filter accepts the following options:
15479 Specify how histogram is calculated.
15481 It accepts the following values:
15484 Use single histogram for all channels.
15486 Use separate histogram for each channel.
15488 Default is @code{single}.
15491 Set frame rate, expressed as number of frames per second. Default
15495 Specify the video size for the output. For the syntax of this option, check the
15496 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15497 Default value is @code{hd720}.
15502 It accepts the following values:
15513 reverse logarithmic
15515 Default is @code{log}.
15518 Set amplitude scale.
15520 It accepts the following values:
15527 Default is @code{log}.
15530 Set how much frames to accumulate in histogram.
15531 Defauls is 1. Setting this to -1 accumulates all frames.
15534 Set histogram ratio of window height.
15537 Set sonogram sliding.
15539 It accepts the following values:
15542 replace old rows with new ones.
15544 scroll from top to bottom.
15546 Default is @code{replace}.
15549 @section aphasemeter
15551 Convert input audio to a video output, displaying the audio phase.
15553 The filter accepts the following options:
15557 Set the output frame rate. Default value is @code{25}.
15560 Set the video size for the output. For the syntax of this option, check the
15561 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15562 Default value is @code{800x400}.
15567 Specify the red, green, blue contrast. Default values are @code{2},
15568 @code{7} and @code{1}.
15569 Allowed range is @code{[0, 255]}.
15572 Set color which will be used for drawing median phase. If color is
15573 @code{none} which is default, no median phase value will be drawn.
15576 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
15577 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
15578 The @code{-1} means left and right channels are completely out of phase and
15579 @code{1} means channels are in phase.
15581 @section avectorscope
15583 Convert input audio to a video output, representing the audio vector
15586 The filter is used to measure the difference between channels of stereo
15587 audio stream. A monoaural signal, consisting of identical left and right
15588 signal, results in straight vertical line. Any stereo separation is visible
15589 as a deviation from this line, creating a Lissajous figure.
15590 If the straight (or deviation from it) but horizontal line appears this
15591 indicates that the left and right channels are out of phase.
15593 The filter accepts the following options:
15597 Set the vectorscope mode.
15599 Available values are:
15602 Lissajous rotated by 45 degrees.
15605 Same as above but not rotated.
15608 Shape resembling half of circle.
15611 Default value is @samp{lissajous}.
15614 Set the video size for the output. For the syntax of this option, check the
15615 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15616 Default value is @code{400x400}.
15619 Set the output frame rate. Default value is @code{25}.
15625 Specify the red, green, blue and alpha contrast. Default values are @code{40},
15626 @code{160}, @code{80} and @code{255}.
15627 Allowed range is @code{[0, 255]}.
15633 Specify the red, green, blue and alpha fade. Default values are @code{15},
15634 @code{10}, @code{5} and @code{5}.
15635 Allowed range is @code{[0, 255]}.
15638 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
15641 Set the vectorscope drawing mode.
15643 Available values are:
15646 Draw dot for each sample.
15649 Draw line between previous and current sample.
15652 Default value is @samp{dot}.
15655 Specify amplitude scale of audio samples.
15657 Available values are:
15674 @subsection Examples
15678 Complete example using @command{ffplay}:
15680 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
15681 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
15685 @section bench, abench
15687 Benchmark part of a filtergraph.
15689 The filter accepts the following options:
15693 Start or stop a timer.
15695 Available values are:
15698 Get the current time, set it as frame metadata (using the key
15699 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
15702 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
15703 the input frame metadata to get the time difference. Time difference, average,
15704 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
15705 @code{min}) are then printed. The timestamps are expressed in seconds.
15709 @subsection Examples
15713 Benchmark @ref{selectivecolor} filter:
15715 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
15721 Concatenate audio and video streams, joining them together one after the
15724 The filter works on segments of synchronized video and audio streams. All
15725 segments must have the same number of streams of each type, and that will
15726 also be the number of streams at output.
15728 The filter accepts the following options:
15733 Set the number of segments. Default is 2.
15736 Set the number of output video streams, that is also the number of video
15737 streams in each segment. Default is 1.
15740 Set the number of output audio streams, that is also the number of audio
15741 streams in each segment. Default is 0.
15744 Activate unsafe mode: do not fail if segments have a different format.
15748 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
15749 @var{a} audio outputs.
15751 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
15752 segment, in the same order as the outputs, then the inputs for the second
15755 Related streams do not always have exactly the same duration, for various
15756 reasons including codec frame size or sloppy authoring. For that reason,
15757 related synchronized streams (e.g. a video and its audio track) should be
15758 concatenated at once. The concat filter will use the duration of the longest
15759 stream in each segment (except the last one), and if necessary pad shorter
15760 audio streams with silence.
15762 For this filter to work correctly, all segments must start at timestamp 0.
15764 All corresponding streams must have the same parameters in all segments; the
15765 filtering system will automatically select a common pixel format for video
15766 streams, and a common sample format, sample rate and channel layout for
15767 audio streams, but other settings, such as resolution, must be converted
15768 explicitly by the user.
15770 Different frame rates are acceptable but will result in variable frame rate
15771 at output; be sure to configure the output file to handle it.
15773 @subsection Examples
15777 Concatenate an opening, an episode and an ending, all in bilingual version
15778 (video in stream 0, audio in streams 1 and 2):
15780 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
15781 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
15782 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
15783 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
15787 Concatenate two parts, handling audio and video separately, using the
15788 (a)movie sources, and adjusting the resolution:
15790 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
15791 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
15792 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
15794 Note that a desync will happen at the stitch if the audio and video streams
15795 do not have exactly the same duration in the first file.
15799 @section drawgraph, adrawgraph
15801 Draw a graph using input video or audio metadata.
15803 It accepts the following parameters:
15807 Set 1st frame metadata key from which metadata values will be used to draw a graph.
15810 Set 1st foreground color expression.
15813 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
15816 Set 2nd foreground color expression.
15819 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
15822 Set 3rd foreground color expression.
15825 Set 4th frame metadata key from which metadata values will be used to draw a graph.
15828 Set 4th foreground color expression.
15831 Set minimal value of metadata value.
15834 Set maximal value of metadata value.
15837 Set graph background color. Default is white.
15842 Available values for mode is:
15849 Default is @code{line}.
15854 Available values for slide is:
15857 Draw new frame when right border is reached.
15860 Replace old columns with new ones.
15863 Scroll from right to left.
15866 Scroll from left to right.
15869 Draw single picture.
15872 Default is @code{frame}.
15875 Set size of graph video. For the syntax of this option, check the
15876 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15877 The default value is @code{900x256}.
15879 The foreground color expressions can use the following variables:
15882 Minimal value of metadata value.
15885 Maximal value of metadata value.
15888 Current metadata key value.
15891 The color is defined as 0xAABBGGRR.
15894 Example using metadata from @ref{signalstats} filter:
15896 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
15899 Example using metadata from @ref{ebur128} filter:
15901 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
15907 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
15908 it unchanged. By default, it logs a message at a frequency of 10Hz with the
15909 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
15910 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
15912 The filter also has a video output (see the @var{video} option) with a real
15913 time graph to observe the loudness evolution. The graphic contains the logged
15914 message mentioned above, so it is not printed anymore when this option is set,
15915 unless the verbose logging is set. The main graphing area contains the
15916 short-term loudness (3 seconds of analysis), and the gauge on the right is for
15917 the momentary loudness (400 milliseconds).
15919 More information about the Loudness Recommendation EBU R128 on
15920 @url{http://tech.ebu.ch/loudness}.
15922 The filter accepts the following options:
15927 Activate the video output. The audio stream is passed unchanged whether this
15928 option is set or no. The video stream will be the first output stream if
15929 activated. Default is @code{0}.
15932 Set the video size. This option is for video only. For the syntax of this
15934 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15935 Default and minimum resolution is @code{640x480}.
15938 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
15939 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
15940 other integer value between this range is allowed.
15943 Set metadata injection. If set to @code{1}, the audio input will be segmented
15944 into 100ms output frames, each of them containing various loudness information
15945 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
15947 Default is @code{0}.
15950 Force the frame logging level.
15952 Available values are:
15955 information logging level
15957 verbose logging level
15960 By default, the logging level is set to @var{info}. If the @option{video} or
15961 the @option{metadata} options are set, it switches to @var{verbose}.
15966 Available modes can be cumulated (the option is a @code{flag} type). Possible
15970 Disable any peak mode (default).
15972 Enable sample-peak mode.
15974 Simple peak mode looking for the higher sample value. It logs a message
15975 for sample-peak (identified by @code{SPK}).
15977 Enable true-peak mode.
15979 If enabled, the peak lookup is done on an over-sampled version of the input
15980 stream for better peak accuracy. It logs a message for true-peak.
15981 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
15982 This mode requires a build with @code{libswresample}.
15986 Treat mono input files as "dual mono". If a mono file is intended for playback
15987 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
15988 If set to @code{true}, this option will compensate for this effect.
15989 Multi-channel input files are not affected by this option.
15992 Set a specific pan law to be used for the measurement of dual mono files.
15993 This parameter is optional, and has a default value of -3.01dB.
15996 @subsection Examples
16000 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
16002 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
16006 Run an analysis with @command{ffmpeg}:
16008 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
16012 @section interleave, ainterleave
16014 Temporally interleave frames from several inputs.
16016 @code{interleave} works with video inputs, @code{ainterleave} with audio.
16018 These filters read frames from several inputs and send the oldest
16019 queued frame to the output.
16021 Input streams must have well defined, monotonically increasing frame
16024 In order to submit one frame to output, these filters need to enqueue
16025 at least one frame for each input, so they cannot work in case one
16026 input is not yet terminated and will not receive incoming frames.
16028 For example consider the case when one input is a @code{select} filter
16029 which always drops input frames. The @code{interleave} filter will keep
16030 reading from that input, but it will never be able to send new frames
16031 to output until the input sends an end-of-stream signal.
16033 Also, depending on inputs synchronization, the filters will drop
16034 frames in case one input receives more frames than the other ones, and
16035 the queue is already filled.
16037 These filters accept the following options:
16041 Set the number of different inputs, it is 2 by default.
16044 @subsection Examples
16048 Interleave frames belonging to different streams using @command{ffmpeg}:
16050 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
16054 Add flickering blur effect:
16056 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
16060 @section metadata, ametadata
16062 Manipulate frame metadata.
16064 This filter accepts the following options:
16068 Set mode of operation of the filter.
16070 Can be one of the following:
16074 If both @code{value} and @code{key} is set, select frames
16075 which have such metadata. If only @code{key} is set, select
16076 every frame that has such key in metadata.
16079 Add new metadata @code{key} and @code{value}. If key is already available
16083 Modify value of already present key.
16086 If @code{value} is set, delete only keys that have such value.
16087 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
16091 Print key and its value if metadata was found. If @code{key} is not set print all
16092 metadata values available in frame.
16096 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
16099 Set metadata value which will be used. This option is mandatory for
16100 @code{modify} and @code{add} mode.
16103 Which function to use when comparing metadata value and @code{value}.
16105 Can be one of following:
16109 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
16112 Values are interpreted as strings, returns true if metadata value starts with
16113 the @code{value} option string.
16116 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
16119 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
16122 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
16125 Values are interpreted as floats, returns true if expression from option @code{expr}
16130 Set expression which is used when @code{function} is set to @code{expr}.
16131 The expression is evaluated through the eval API and can contain the following
16136 Float representation of @code{value} from metadata key.
16139 Float representation of @code{value} as supplied by user in @code{value} option.
16142 If specified in @code{print} mode, output is written to the named file. Instead of
16143 plain filename any writable url can be specified. Filename ``-'' is a shorthand
16144 for standard output. If @code{file} option is not set, output is written to the log
16145 with AV_LOG_INFO loglevel.
16150 @subsection Examples
16154 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
16157 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
16160 Print silencedetect output to file @file{metadata.txt}.
16162 silencedetect,ametadata=mode=print:file=metadata.txt
16165 Direct all metadata to a pipe with file descriptor 4.
16167 metadata=mode=print:file='pipe\:4'
16171 @section perms, aperms
16173 Set read/write permissions for the output frames.
16175 These filters are mainly aimed at developers to test direct path in the
16176 following filter in the filtergraph.
16178 The filters accept the following options:
16182 Select the permissions mode.
16184 It accepts the following values:
16187 Do nothing. This is the default.
16189 Set all the output frames read-only.
16191 Set all the output frames directly writable.
16193 Make the frame read-only if writable, and writable if read-only.
16195 Set each output frame read-only or writable randomly.
16199 Set the seed for the @var{random} mode, must be an integer included between
16200 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16201 @code{-1}, the filter will try to use a good random seed on a best effort
16205 Note: in case of auto-inserted filter between the permission filter and the
16206 following one, the permission might not be received as expected in that
16207 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
16208 perms/aperms filter can avoid this problem.
16210 @section realtime, arealtime
16212 Slow down filtering to match real time approximatively.
16214 These filters will pause the filtering for a variable amount of time to
16215 match the output rate with the input timestamps.
16216 They are similar to the @option{re} option to @code{ffmpeg}.
16218 They accept the following options:
16222 Time limit for the pauses. Any pause longer than that will be considered
16223 a timestamp discontinuity and reset the timer. Default is 2 seconds.
16227 @section select, aselect
16229 Select frames to pass in output.
16231 This filter accepts the following options:
16236 Set expression, which is evaluated for each input frame.
16238 If the expression is evaluated to zero, the frame is discarded.
16240 If the evaluation result is negative or NaN, the frame is sent to the
16241 first output; otherwise it is sent to the output with index
16242 @code{ceil(val)-1}, assuming that the input index starts from 0.
16244 For example a value of @code{1.2} corresponds to the output with index
16245 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
16248 Set the number of outputs. The output to which to send the selected
16249 frame is based on the result of the evaluation. Default value is 1.
16252 The expression can contain the following constants:
16256 The (sequential) number of the filtered frame, starting from 0.
16259 The (sequential) number of the selected frame, starting from 0.
16261 @item prev_selected_n
16262 The sequential number of the last selected frame. It's NAN if undefined.
16265 The timebase of the input timestamps.
16268 The PTS (Presentation TimeStamp) of the filtered video frame,
16269 expressed in @var{TB} units. It's NAN if undefined.
16272 The PTS of the filtered video frame,
16273 expressed in seconds. It's NAN if undefined.
16276 The PTS of the previously filtered video frame. It's NAN if undefined.
16278 @item prev_selected_pts
16279 The PTS of the last previously filtered video frame. It's NAN if undefined.
16281 @item prev_selected_t
16282 The PTS of the last previously selected video frame. It's NAN if undefined.
16285 The PTS of the first video frame in the video. It's NAN if undefined.
16288 The time of the first video frame in the video. It's NAN if undefined.
16290 @item pict_type @emph{(video only)}
16291 The type of the filtered frame. It can assume one of the following
16303 @item interlace_type @emph{(video only)}
16304 The frame interlace type. It can assume one of the following values:
16307 The frame is progressive (not interlaced).
16309 The frame is top-field-first.
16311 The frame is bottom-field-first.
16314 @item consumed_sample_n @emph{(audio only)}
16315 the number of selected samples before the current frame
16317 @item samples_n @emph{(audio only)}
16318 the number of samples in the current frame
16320 @item sample_rate @emph{(audio only)}
16321 the input sample rate
16324 This is 1 if the filtered frame is a key-frame, 0 otherwise.
16327 the position in the file of the filtered frame, -1 if the information
16328 is not available (e.g. for synthetic video)
16330 @item scene @emph{(video only)}
16331 value between 0 and 1 to indicate a new scene; a low value reflects a low
16332 probability for the current frame to introduce a new scene, while a higher
16333 value means the current frame is more likely to be one (see the example below)
16335 @item concatdec_select
16336 The concat demuxer can select only part of a concat input file by setting an
16337 inpoint and an outpoint, but the output packets may not be entirely contained
16338 in the selected interval. By using this variable, it is possible to skip frames
16339 generated by the concat demuxer which are not exactly contained in the selected
16342 This works by comparing the frame pts against the @var{lavf.concat.start_time}
16343 and the @var{lavf.concat.duration} packet metadata values which are also
16344 present in the decoded frames.
16346 The @var{concatdec_select} variable is -1 if the frame pts is at least
16347 start_time and either the duration metadata is missing or the frame pts is less
16348 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
16351 That basically means that an input frame is selected if its pts is within the
16352 interval set by the concat demuxer.
16356 The default value of the select expression is "1".
16358 @subsection Examples
16362 Select all frames in input:
16367 The example above is the same as:
16379 Select only I-frames:
16381 select='eq(pict_type\,I)'
16385 Select one frame every 100:
16387 select='not(mod(n\,100))'
16391 Select only frames contained in the 10-20 time interval:
16393 select=between(t\,10\,20)
16397 Select only I-frames contained in the 10-20 time interval:
16399 select=between(t\,10\,20)*eq(pict_type\,I)
16403 Select frames with a minimum distance of 10 seconds:
16405 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
16409 Use aselect to select only audio frames with samples number > 100:
16411 aselect='gt(samples_n\,100)'
16415 Create a mosaic of the first scenes:
16417 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
16420 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
16424 Send even and odd frames to separate outputs, and compose them:
16426 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
16430 Select useful frames from an ffconcat file which is using inpoints and
16431 outpoints but where the source files are not intra frame only.
16433 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
16437 @section sendcmd, asendcmd
16439 Send commands to filters in the filtergraph.
16441 These filters read commands to be sent to other filters in the
16444 @code{sendcmd} must be inserted between two video filters,
16445 @code{asendcmd} must be inserted between two audio filters, but apart
16446 from that they act the same way.
16448 The specification of commands can be provided in the filter arguments
16449 with the @var{commands} option, or in a file specified by the
16450 @var{filename} option.
16452 These filters accept the following options:
16455 Set the commands to be read and sent to the other filters.
16457 Set the filename of the commands to be read and sent to the other
16461 @subsection Commands syntax
16463 A commands description consists of a sequence of interval
16464 specifications, comprising a list of commands to be executed when a
16465 particular event related to that interval occurs. The occurring event
16466 is typically the current frame time entering or leaving a given time
16469 An interval is specified by the following syntax:
16471 @var{START}[-@var{END}] @var{COMMANDS};
16474 The time interval is specified by the @var{START} and @var{END} times.
16475 @var{END} is optional and defaults to the maximum time.
16477 The current frame time is considered within the specified interval if
16478 it is included in the interval [@var{START}, @var{END}), that is when
16479 the time is greater or equal to @var{START} and is lesser than
16482 @var{COMMANDS} consists of a sequence of one or more command
16483 specifications, separated by ",", relating to that interval. The
16484 syntax of a command specification is given by:
16486 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
16489 @var{FLAGS} is optional and specifies the type of events relating to
16490 the time interval which enable sending the specified command, and must
16491 be a non-null sequence of identifier flags separated by "+" or "|" and
16492 enclosed between "[" and "]".
16494 The following flags are recognized:
16497 The command is sent when the current frame timestamp enters the
16498 specified interval. In other words, the command is sent when the
16499 previous frame timestamp was not in the given interval, and the
16503 The command is sent when the current frame timestamp leaves the
16504 specified interval. In other words, the command is sent when the
16505 previous frame timestamp was in the given interval, and the
16509 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
16512 @var{TARGET} specifies the target of the command, usually the name of
16513 the filter class or a specific filter instance name.
16515 @var{COMMAND} specifies the name of the command for the target filter.
16517 @var{ARG} is optional and specifies the optional list of argument for
16518 the given @var{COMMAND}.
16520 Between one interval specification and another, whitespaces, or
16521 sequences of characters starting with @code{#} until the end of line,
16522 are ignored and can be used to annotate comments.
16524 A simplified BNF description of the commands specification syntax
16527 @var{COMMAND_FLAG} ::= "enter" | "leave"
16528 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
16529 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
16530 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
16531 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
16532 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
16535 @subsection Examples
16539 Specify audio tempo change at second 4:
16541 asendcmd=c='4.0 atempo tempo 1.5',atempo
16545 Specify a list of drawtext and hue commands in a file.
16547 # show text in the interval 5-10
16548 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
16549 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
16551 # desaturate the image in the interval 15-20
16552 15.0-20.0 [enter] hue s 0,
16553 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
16555 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
16557 # apply an exponential saturation fade-out effect, starting from time 25
16558 25 [enter] hue s exp(25-t)
16561 A filtergraph allowing to read and process the above command list
16562 stored in a file @file{test.cmd}, can be specified with:
16564 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
16569 @section setpts, asetpts
16571 Change the PTS (presentation timestamp) of the input frames.
16573 @code{setpts} works on video frames, @code{asetpts} on audio frames.
16575 This filter accepts the following options:
16580 The expression which is evaluated for each frame to construct its timestamp.
16584 The expression is evaluated through the eval API and can contain the following
16589 frame rate, only defined for constant frame-rate video
16592 The presentation timestamp in input
16595 The count of the input frame for video or the number of consumed samples,
16596 not including the current frame for audio, starting from 0.
16598 @item NB_CONSUMED_SAMPLES
16599 The number of consumed samples, not including the current frame (only
16602 @item NB_SAMPLES, S
16603 The number of samples in the current frame (only audio)
16605 @item SAMPLE_RATE, SR
16606 The audio sample rate.
16609 The PTS of the first frame.
16612 the time in seconds of the first frame
16615 State whether the current frame is interlaced.
16618 the time in seconds of the current frame
16621 original position in the file of the frame, or undefined if undefined
16622 for the current frame
16625 The previous input PTS.
16628 previous input time in seconds
16631 The previous output PTS.
16634 previous output time in seconds
16637 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
16641 The wallclock (RTC) time at the start of the movie in microseconds.
16644 The timebase of the input timestamps.
16648 @subsection Examples
16652 Start counting PTS from zero
16654 setpts=PTS-STARTPTS
16658 Apply fast motion effect:
16664 Apply slow motion effect:
16670 Set fixed rate of 25 frames per second:
16676 Set fixed rate 25 fps with some jitter:
16678 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
16682 Apply an offset of 10 seconds to the input PTS:
16688 Generate timestamps from a "live source" and rebase onto the current timebase:
16690 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
16694 Generate timestamps by counting samples:
16701 @section settb, asettb
16703 Set the timebase to use for the output frames timestamps.
16704 It is mainly useful for testing timebase configuration.
16706 It accepts the following parameters:
16711 The expression which is evaluated into the output timebase.
16715 The value for @option{tb} is an arithmetic expression representing a
16716 rational. The expression can contain the constants "AVTB" (the default
16717 timebase), "intb" (the input timebase) and "sr" (the sample rate,
16718 audio only). Default value is "intb".
16720 @subsection Examples
16724 Set the timebase to 1/25:
16730 Set the timebase to 1/10:
16736 Set the timebase to 1001/1000:
16742 Set the timebase to 2*intb:
16748 Set the default timebase value:
16755 Convert input audio to a video output representing frequency spectrum
16756 logarithmically using Brown-Puckette constant Q transform algorithm with
16757 direct frequency domain coefficient calculation (but the transform itself
16758 is not really constant Q, instead the Q factor is actually variable/clamped),
16759 with musical tone scale, from E0 to D#10.
16761 The filter accepts the following options:
16765 Specify the video size for the output. It must be even. For the syntax of this option,
16766 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16767 Default value is @code{1920x1080}.
16770 Set the output frame rate. Default value is @code{25}.
16773 Set the bargraph height. It must be even. Default value is @code{-1} which
16774 computes the bargraph height automatically.
16777 Set the axis height. It must be even. Default value is @code{-1} which computes
16778 the axis height automatically.
16781 Set the sonogram height. It must be even. Default value is @code{-1} which
16782 computes the sonogram height automatically.
16785 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
16786 instead. Default value is @code{1}.
16788 @item sono_v, volume
16789 Specify the sonogram volume expression. It can contain variables:
16792 the @var{bar_v} evaluated expression
16793 @item frequency, freq, f
16794 the frequency where it is evaluated
16795 @item timeclamp, tc
16796 the value of @var{timeclamp} option
16800 @item a_weighting(f)
16801 A-weighting of equal loudness
16802 @item b_weighting(f)
16803 B-weighting of equal loudness
16804 @item c_weighting(f)
16805 C-weighting of equal loudness.
16807 Default value is @code{16}.
16809 @item bar_v, volume2
16810 Specify the bargraph volume expression. It can contain variables:
16813 the @var{sono_v} evaluated expression
16814 @item frequency, freq, f
16815 the frequency where it is evaluated
16816 @item timeclamp, tc
16817 the value of @var{timeclamp} option
16821 @item a_weighting(f)
16822 A-weighting of equal loudness
16823 @item b_weighting(f)
16824 B-weighting of equal loudness
16825 @item c_weighting(f)
16826 C-weighting of equal loudness.
16828 Default value is @code{sono_v}.
16830 @item sono_g, gamma
16831 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
16832 higher gamma makes the spectrum having more range. Default value is @code{3}.
16833 Acceptable range is @code{[1, 7]}.
16835 @item bar_g, gamma2
16836 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
16840 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
16841 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
16843 @item timeclamp, tc
16844 Specify the transform timeclamp. At low frequency, there is trade-off between
16845 accuracy in time domain and frequency domain. If timeclamp is lower,
16846 event in time domain is represented more accurately (such as fast bass drum),
16847 otherwise event in frequency domain is represented more accurately
16848 (such as bass guitar). Acceptable range is @code{[0.1, 1]}. Default value is @code{0.17}.
16851 Specify the transform base frequency. Default value is @code{20.01523126408007475},
16852 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
16855 Specify the transform end frequency. Default value is @code{20495.59681441799654},
16856 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
16859 This option is deprecated and ignored.
16862 Specify the transform length in time domain. Use this option to control accuracy
16863 trade-off between time domain and frequency domain at every frequency sample.
16864 It can contain variables:
16866 @item frequency, freq, f
16867 the frequency where it is evaluated
16868 @item timeclamp, tc
16869 the value of @var{timeclamp} option.
16871 Default value is @code{384*tc/(384+tc*f)}.
16874 Specify the transform count for every video frame. Default value is @code{6}.
16875 Acceptable range is @code{[1, 30]}.
16878 Specify the transform count for every single pixel. Default value is @code{0},
16879 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
16882 Specify font file for use with freetype to draw the axis. If not specified,
16883 use embedded font. Note that drawing with font file or embedded font is not
16884 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
16888 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
16889 The : in the pattern may be replaced by | to avoid unnecessary escaping.
16892 Specify font color expression. This is arithmetic expression that should return
16893 integer value 0xRRGGBB. It can contain variables:
16895 @item frequency, freq, f
16896 the frequency where it is evaluated
16897 @item timeclamp, tc
16898 the value of @var{timeclamp} option
16903 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
16904 @item r(x), g(x), b(x)
16905 red, green, and blue value of intensity x.
16907 Default value is @code{st(0, (midi(f)-59.5)/12);
16908 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
16909 r(1-ld(1)) + b(ld(1))}.
16912 Specify image file to draw the axis. This option override @var{fontfile} and
16913 @var{fontcolor} option.
16916 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
16917 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
16918 Default value is @code{1}.
16921 Set colorspace. The accepted values are:
16924 Unspecified (default)
16933 BT.470BG or BT.601-6 625
16936 SMPTE-170M or BT.601-6 525
16942 BT.2020 with non-constant luminance
16947 Set spectrogram color scheme. This is list of floating point values with format
16948 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
16949 The default is @code{1|0.5|0|0|0.5|1}.
16953 @subsection Examples
16957 Playing audio while showing the spectrum:
16959 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
16963 Same as above, but with frame rate 30 fps:
16965 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
16969 Playing at 1280x720:
16971 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
16975 Disable sonogram display:
16981 A1 and its harmonics: A1, A2, (near)E3, A3:
16983 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),
16984 asplit[a][out1]; [a] showcqt [out0]'
16988 Same as above, but with more accuracy in frequency domain:
16990 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),
16991 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
16997 bar_v=10:sono_v=bar_v*a_weighting(f)
17001 Custom gamma, now spectrum is linear to the amplitude.
17007 Custom tlength equation:
17009 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)))'
17013 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
17015 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
17019 Custom font using fontconfig:
17021 font='Courier New,Monospace,mono|bold'
17025 Custom frequency range with custom axis using image file:
17027 axisfile=myaxis.png:basefreq=40:endfreq=10000
17033 Convert input audio to video output representing the audio power spectrum.
17034 Audio amplitude is on Y-axis while frequency is on X-axis.
17036 The filter accepts the following options:
17040 Specify size of video. For the syntax of this option, check the
17041 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17042 Default is @code{1024x512}.
17046 This set how each frequency bin will be represented.
17048 It accepts the following values:
17054 Default is @code{bar}.
17057 Set amplitude scale.
17059 It accepts the following values:
17073 Default is @code{log}.
17076 Set frequency scale.
17078 It accepts the following values:
17087 Reverse logarithmic scale.
17089 Default is @code{lin}.
17094 It accepts the following values:
17110 Default is @code{w2048}
17113 Set windowing function.
17115 It accepts the following values:
17137 Default is @code{hanning}.
17140 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17141 which means optimal overlap for selected window function will be picked.
17144 Set time averaging. Setting this to 0 will display current maximal peaks.
17145 Default is @code{1}, which means time averaging is disabled.
17148 Specify list of colors separated by space or by '|' which will be used to
17149 draw channel frequencies. Unrecognized or missing colors will be replaced
17153 Set channel display mode.
17155 It accepts the following values:
17160 Default is @code{combined}.
17163 Set minimum amplitude used in @code{log} amplitude scaler.
17167 @anchor{showspectrum}
17168 @section showspectrum
17170 Convert input audio to a video output, representing the audio frequency
17173 The filter accepts the following options:
17177 Specify the video size for the output. For the syntax of this option, check the
17178 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17179 Default value is @code{640x512}.
17182 Specify how the spectrum should slide along the window.
17184 It accepts the following values:
17187 the samples start again on the left when they reach the right
17189 the samples scroll from right to left
17191 frames are only produced when the samples reach the right
17193 the samples scroll from left to right
17196 Default value is @code{replace}.
17199 Specify display mode.
17201 It accepts the following values:
17204 all channels are displayed in the same row
17206 all channels are displayed in separate rows
17209 Default value is @samp{combined}.
17212 Specify display color mode.
17214 It accepts the following values:
17217 each channel is displayed in a separate color
17219 each channel is displayed using the same color scheme
17221 each channel is displayed using the rainbow color scheme
17223 each channel is displayed using the moreland color scheme
17225 each channel is displayed using the nebulae color scheme
17227 each channel is displayed using the fire color scheme
17229 each channel is displayed using the fiery color scheme
17231 each channel is displayed using the fruit color scheme
17233 each channel is displayed using the cool color scheme
17236 Default value is @samp{channel}.
17239 Specify scale used for calculating intensity color values.
17241 It accepts the following values:
17246 square root, default
17257 Default value is @samp{sqrt}.
17260 Set saturation modifier for displayed colors. Negative values provide
17261 alternative color scheme. @code{0} is no saturation at all.
17262 Saturation must be in [-10.0, 10.0] range.
17263 Default value is @code{1}.
17266 Set window function.
17268 It accepts the following values:
17292 Default value is @code{hann}.
17295 Set orientation of time vs frequency axis. Can be @code{vertical} or
17296 @code{horizontal}. Default is @code{vertical}.
17299 Set ratio of overlap window. Default value is @code{0}.
17300 When value is @code{1} overlap is set to recommended size for specific
17301 window function currently used.
17304 Set scale gain for calculating intensity color values.
17305 Default value is @code{1}.
17308 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
17311 Set color rotation, must be in [-1.0, 1.0] range.
17312 Default value is @code{0}.
17315 The usage is very similar to the showwaves filter; see the examples in that
17318 @subsection Examples
17322 Large window with logarithmic color scaling:
17324 showspectrum=s=1280x480:scale=log
17328 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
17330 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17331 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
17335 @section showspectrumpic
17337 Convert input audio to a single video frame, representing the audio frequency
17340 The filter accepts the following options:
17344 Specify the video size for the output. For the syntax of this option, check the
17345 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17346 Default value is @code{4096x2048}.
17349 Specify display mode.
17351 It accepts the following values:
17354 all channels are displayed in the same row
17356 all channels are displayed in separate rows
17358 Default value is @samp{combined}.
17361 Specify display color mode.
17363 It accepts the following values:
17366 each channel is displayed in a separate color
17368 each channel is displayed using the same color scheme
17370 each channel is displayed using the rainbow color scheme
17372 each channel is displayed using the moreland color scheme
17374 each channel is displayed using the nebulae color scheme
17376 each channel is displayed using the fire color scheme
17378 each channel is displayed using the fiery color scheme
17380 each channel is displayed using the fruit color scheme
17382 each channel is displayed using the cool color scheme
17384 Default value is @samp{intensity}.
17387 Specify scale used for calculating intensity color values.
17389 It accepts the following values:
17394 square root, default
17404 Default value is @samp{log}.
17407 Set saturation modifier for displayed colors. Negative values provide
17408 alternative color scheme. @code{0} is no saturation at all.
17409 Saturation must be in [-10.0, 10.0] range.
17410 Default value is @code{1}.
17413 Set window function.
17415 It accepts the following values:
17438 Default value is @code{hann}.
17441 Set orientation of time vs frequency axis. Can be @code{vertical} or
17442 @code{horizontal}. Default is @code{vertical}.
17445 Set scale gain for calculating intensity color values.
17446 Default value is @code{1}.
17449 Draw time and frequency axes and legends. Default is enabled.
17452 Set color rotation, must be in [-1.0, 1.0] range.
17453 Default value is @code{0}.
17456 @subsection Examples
17460 Extract an audio spectrogram of a whole audio track
17461 in a 1024x1024 picture using @command{ffmpeg}:
17463 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
17467 @section showvolume
17469 Convert input audio volume to a video output.
17471 The filter accepts the following options:
17478 Set border width, allowed range is [0, 5]. Default is 1.
17481 Set channel width, allowed range is [80, 8192]. Default is 400.
17484 Set channel height, allowed range is [1, 900]. Default is 20.
17487 Set fade, allowed range is [0.001, 1]. Default is 0.95.
17490 Set volume color expression.
17492 The expression can use the following variables:
17496 Current max volume of channel in dB.
17502 Current channel number, starting from 0.
17506 If set, displays channel names. Default is enabled.
17509 If set, displays volume values. Default is enabled.
17512 Set orientation, can be @code{horizontal} or @code{vertical},
17513 default is @code{horizontal}.
17516 Set step size, allowed range s [0, 5]. Default is 0, which means
17522 Convert input audio to a video output, representing the samples waves.
17524 The filter accepts the following options:
17528 Specify the video size for the output. For the syntax of this option, check the
17529 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17530 Default value is @code{600x240}.
17535 Available values are:
17538 Draw a point for each sample.
17541 Draw a vertical line for each sample.
17544 Draw a point for each sample and a line between them.
17547 Draw a centered vertical line for each sample.
17550 Default value is @code{point}.
17553 Set the number of samples which are printed on the same column. A
17554 larger value will decrease the frame rate. Must be a positive
17555 integer. This option can be set only if the value for @var{rate}
17556 is not explicitly specified.
17559 Set the (approximate) output frame rate. This is done by setting the
17560 option @var{n}. Default value is "25".
17562 @item split_channels
17563 Set if channels should be drawn separately or overlap. Default value is 0.
17566 Set colors separated by '|' which are going to be used for drawing of each channel.
17569 Set amplitude scale.
17571 Available values are:
17589 @subsection Examples
17593 Output the input file audio and the corresponding video representation
17596 amovie=a.mp3,asplit[out0],showwaves[out1]
17600 Create a synthetic signal and show it with showwaves, forcing a
17601 frame rate of 30 frames per second:
17603 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
17607 @section showwavespic
17609 Convert input audio to a single video frame, representing the samples waves.
17611 The filter accepts the following options:
17615 Specify the video size for the output. For the syntax of this option, check the
17616 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17617 Default value is @code{600x240}.
17619 @item split_channels
17620 Set if channels should be drawn separately or overlap. Default value is 0.
17623 Set colors separated by '|' which are going to be used for drawing of each channel.
17626 Set amplitude scale. Can be linear @code{lin} or logarithmic @code{log}.
17630 @subsection Examples
17634 Extract a channel split representation of the wave form of a whole audio track
17635 in a 1024x800 picture using @command{ffmpeg}:
17637 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
17641 @section sidedata, asidedata
17643 Delete frame side data, or select frames based on it.
17645 This filter accepts the following options:
17649 Set mode of operation of the filter.
17651 Can be one of the following:
17655 Select every frame with side data of @code{type}.
17658 Delete side data of @code{type}. If @code{type} is not set, delete all side
17664 Set side data type used with all modes. Must be set for @code{select} mode. For
17665 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
17666 in @file{libavutil/frame.h}. For example, to choose
17667 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
17671 @section spectrumsynth
17673 Sythesize audio from 2 input video spectrums, first input stream represents
17674 magnitude across time and second represents phase across time.
17675 The filter will transform from frequency domain as displayed in videos back
17676 to time domain as presented in audio output.
17678 This filter is primarily created for reversing processed @ref{showspectrum}
17679 filter outputs, but can synthesize sound from other spectrograms too.
17680 But in such case results are going to be poor if the phase data is not
17681 available, because in such cases phase data need to be recreated, usually
17682 its just recreated from random noise.
17683 For best results use gray only output (@code{channel} color mode in
17684 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
17685 @code{lin} scale for phase video. To produce phase, for 2nd video, use
17686 @code{data} option. Inputs videos should generally use @code{fullframe}
17687 slide mode as that saves resources needed for decoding video.
17689 The filter accepts the following options:
17693 Specify sample rate of output audio, the sample rate of audio from which
17694 spectrum was generated may differ.
17697 Set number of channels represented in input video spectrums.
17700 Set scale which was used when generating magnitude input spectrum.
17701 Can be @code{lin} or @code{log}. Default is @code{log}.
17704 Set slide which was used when generating inputs spectrums.
17705 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
17706 Default is @code{fullframe}.
17709 Set window function used for resynthesis.
17712 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17713 which means optimal overlap for selected window function will be picked.
17716 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
17717 Default is @code{vertical}.
17720 @subsection Examples
17724 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
17725 then resynthesize videos back to audio with spectrumsynth:
17727 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
17728 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
17729 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
17733 @section split, asplit
17735 Split input into several identical outputs.
17737 @code{asplit} works with audio input, @code{split} with video.
17739 The filter accepts a single parameter which specifies the number of outputs. If
17740 unspecified, it defaults to 2.
17742 @subsection Examples
17746 Create two separate outputs from the same input:
17748 [in] split [out0][out1]
17752 To create 3 or more outputs, you need to specify the number of
17755 [in] asplit=3 [out0][out1][out2]
17759 Create two separate outputs from the same input, one cropped and
17762 [in] split [splitout1][splitout2];
17763 [splitout1] crop=100:100:0:0 [cropout];
17764 [splitout2] pad=200:200:100:100 [padout];
17768 Create 5 copies of the input audio with @command{ffmpeg}:
17770 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
17776 Receive commands sent through a libzmq client, and forward them to
17777 filters in the filtergraph.
17779 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
17780 must be inserted between two video filters, @code{azmq} between two
17783 To enable these filters you need to install the libzmq library and
17784 headers and configure FFmpeg with @code{--enable-libzmq}.
17786 For more information about libzmq see:
17787 @url{http://www.zeromq.org/}
17789 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
17790 receives messages sent through a network interface defined by the
17791 @option{bind_address} option.
17793 The received message must be in the form:
17795 @var{TARGET} @var{COMMAND} [@var{ARG}]
17798 @var{TARGET} specifies the target of the command, usually the name of
17799 the filter class or a specific filter instance name.
17801 @var{COMMAND} specifies the name of the command for the target filter.
17803 @var{ARG} is optional and specifies the optional argument list for the
17804 given @var{COMMAND}.
17806 Upon reception, the message is processed and the corresponding command
17807 is injected into the filtergraph. Depending on the result, the filter
17808 will send a reply to the client, adopting the format:
17810 @var{ERROR_CODE} @var{ERROR_REASON}
17814 @var{MESSAGE} is optional.
17816 @subsection Examples
17818 Look at @file{tools/zmqsend} for an example of a zmq client which can
17819 be used to send commands processed by these filters.
17821 Consider the following filtergraph generated by @command{ffplay}
17823 ffplay -dumpgraph 1 -f lavfi "
17824 color=s=100x100:c=red [l];
17825 color=s=100x100:c=blue [r];
17826 nullsrc=s=200x100, zmq [bg];
17827 [bg][l] overlay [bg+l];
17828 [bg+l][r] overlay=x=100 "
17831 To change the color of the left side of the video, the following
17832 command can be used:
17834 echo Parsed_color_0 c yellow | tools/zmqsend
17837 To change the right side:
17839 echo Parsed_color_1 c pink | tools/zmqsend
17842 @c man end MULTIMEDIA FILTERS
17844 @chapter Multimedia Sources
17845 @c man begin MULTIMEDIA SOURCES
17847 Below is a description of the currently available multimedia sources.
17851 This is the same as @ref{movie} source, except it selects an audio
17857 Read audio and/or video stream(s) from a movie container.
17859 It accepts the following parameters:
17863 The name of the resource to read (not necessarily a file; it can also be a
17864 device or a stream accessed through some protocol).
17866 @item format_name, f
17867 Specifies the format assumed for the movie to read, and can be either
17868 the name of a container or an input device. If not specified, the
17869 format is guessed from @var{movie_name} or by probing.
17871 @item seek_point, sp
17872 Specifies the seek point in seconds. The frames will be output
17873 starting from this seek point. The parameter is evaluated with
17874 @code{av_strtod}, so the numerical value may be suffixed by an IS
17875 postfix. The default value is "0".
17878 Specifies the streams to read. Several streams can be specified,
17879 separated by "+". The source will then have as many outputs, in the
17880 same order. The syntax is explained in the ``Stream specifiers''
17881 section in the ffmpeg manual. Two special names, "dv" and "da" specify
17882 respectively the default (best suited) video and audio stream. Default
17883 is "dv", or "da" if the filter is called as "amovie".
17885 @item stream_index, si
17886 Specifies the index of the video stream to read. If the value is -1,
17887 the most suitable video stream will be automatically selected. The default
17888 value is "-1". Deprecated. If the filter is called "amovie", it will select
17889 audio instead of video.
17892 Specifies how many times to read the stream in sequence.
17893 If the value is less than 1, the stream will be read again and again.
17894 Default value is "1".
17896 Note that when the movie is looped the source timestamps are not
17897 changed, so it will generate non monotonically increasing timestamps.
17899 @item discontinuity
17900 Specifies the time difference between frames above which the point is
17901 considered a timestamp discontinuity which is removed by adjusting the later
17905 It allows overlaying a second video on top of the main input of
17906 a filtergraph, as shown in this graph:
17908 input -----------> deltapts0 --> overlay --> output
17911 movie --> scale--> deltapts1 -------+
17913 @subsection Examples
17917 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
17918 on top of the input labelled "in":
17920 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
17921 [in] setpts=PTS-STARTPTS [main];
17922 [main][over] overlay=16:16 [out]
17926 Read from a video4linux2 device, and overlay it on top of the input
17929 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
17930 [in] setpts=PTS-STARTPTS [main];
17931 [main][over] overlay=16:16 [out]
17935 Read the first video stream and the audio stream with id 0x81 from
17936 dvd.vob; the video is connected to the pad named "video" and the audio is
17937 connected to the pad named "audio":
17939 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
17943 @subsection Commands
17945 Both movie and amovie support the following commands:
17948 Perform seek using "av_seek_frame".
17949 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
17952 @var{stream_index}: If stream_index is -1, a default
17953 stream is selected, and @var{timestamp} is automatically converted
17954 from AV_TIME_BASE units to the stream specific time_base.
17956 @var{timestamp}: Timestamp in AVStream.time_base units
17957 or, if no stream is specified, in AV_TIME_BASE units.
17959 @var{flags}: Flags which select direction and seeking mode.
17963 Get movie duration in AV_TIME_BASE units.
17967 @c man end MULTIMEDIA SOURCES