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{id}=@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 optionally followed by "@@@var{id}".
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{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Options for filters with several inputs (framesync)
317 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
319 Some filters with several inputs support a common set of options.
320 These options can only be set by name, not with the short notation.
324 The action to take when EOF is encountered on the secondary input; it accepts
325 one of the following values:
329 Repeat the last frame (the default).
333 Pass the main input through.
337 If set to 1, force the output to terminate when the shortest input
338 terminates. Default value is 0.
341 If set to 1, force the filter to extend the last frame of secondary streams
342 until the end of the primary stream. A value of 0 disables this behavior.
346 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
348 @chapter Audio Filters
349 @c man begin AUDIO FILTERS
351 When you configure your FFmpeg build, you can disable any of the
352 existing filters using @code{--disable-filters}.
353 The configure output will show the audio filters included in your
356 Below is a description of the currently available audio filters.
360 A compressor is mainly used to reduce the dynamic range of a signal.
361 Especially modern music is mostly compressed at a high ratio to
362 improve the overall loudness. It's done to get the highest attention
363 of a listener, "fatten" the sound and bring more "power" to the track.
364 If a signal is compressed too much it may sound dull or "dead"
365 afterwards or it may start to "pump" (which could be a powerful effect
366 but can also destroy a track completely).
367 The right compression is the key to reach a professional sound and is
368 the high art of mixing and mastering. Because of its complex settings
369 it may take a long time to get the right feeling for this kind of effect.
371 Compression is done by detecting the volume above a chosen level
372 @code{threshold} and dividing it by the factor set with @code{ratio}.
373 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
374 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
375 the signal would cause distortion of the waveform the reduction can be
376 levelled over the time. This is done by setting "Attack" and "Release".
377 @code{attack} determines how long the signal has to rise above the threshold
378 before any reduction will occur and @code{release} sets the time the signal
379 has to fall below the threshold to reduce the reduction again. Shorter signals
380 than the chosen attack time will be left untouched.
381 The overall reduction of the signal can be made up afterwards with the
382 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
383 raising the makeup to this level results in a signal twice as loud than the
384 source. To gain a softer entry in the compression the @code{knee} flattens the
385 hard edge at the threshold in the range of the chosen decibels.
387 The filter accepts the following options:
391 Set input gain. Default is 1. Range is between 0.015625 and 64.
394 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
395 Default is @code{downward}.
398 If a signal of stream rises above this level it will affect the gain
400 By default it is 0.125. Range is between 0.00097563 and 1.
403 Set a ratio by which the signal is reduced. 1:2 means that if the level
404 rose 4dB above the threshold, it will be only 2dB above after the reduction.
405 Default is 2. Range is between 1 and 20.
408 Amount of milliseconds the signal has to rise above the threshold before gain
409 reduction starts. Default is 20. Range is between 0.01 and 2000.
412 Amount of milliseconds the signal has to fall below the threshold before
413 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
416 Set the amount by how much signal will be amplified after processing.
417 Default is 1. Range is from 1 to 64.
420 Curve the sharp knee around the threshold to enter gain reduction more softly.
421 Default is 2.82843. Range is between 1 and 8.
424 Choose if the @code{average} level between all channels of input stream
425 or the louder(@code{maximum}) channel of input stream affects the
426 reduction. Default is @code{average}.
429 Should the exact signal be taken in case of @code{peak} or an RMS one in case
430 of @code{rms}. Default is @code{rms} which is mostly smoother.
433 How much to use compressed signal in output. Default is 1.
434 Range is between 0 and 1.
438 Simple audio dynamic range compression/expansion filter.
440 The filter accepts the following options:
444 Set contrast. Default is 33. Allowed range is between 0 and 100.
449 Copy the input audio source unchanged to the output. This is mainly useful for
454 Apply cross fade from one input audio stream to another input audio stream.
455 The cross fade is applied for specified duration near the end of first stream.
457 The filter accepts the following options:
461 Specify the number of samples for which the cross fade effect has to last.
462 At the end of the cross fade effect the first input audio will be completely
463 silent. Default is 44100.
466 Specify the duration of the cross fade effect. See
467 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
468 for the accepted syntax.
469 By default the duration is determined by @var{nb_samples}.
470 If set this option is used instead of @var{nb_samples}.
473 Should first stream end overlap with second stream start. Default is enabled.
476 Set curve for cross fade transition for first stream.
479 Set curve for cross fade transition for second stream.
481 For description of available curve types see @ref{afade} filter description.
488 Cross fade from one input to another:
490 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
494 Cross fade from one input to another but without overlapping:
496 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
501 Split audio stream into several bands.
503 This filter splits audio stream into two or more frequency ranges.
504 Summing all streams back will give flat output.
506 The filter accepts the following options:
510 Set split frequencies. Those must be positive and increasing.
513 Set filter order, can be @var{2nd}, @var{4th} or @var{8th}.
514 Default is @var{4th}.
519 Reduce audio bit resolution.
521 This filter is bit crusher with enhanced functionality. A bit crusher
522 is used to audibly reduce number of bits an audio signal is sampled
523 with. This doesn't change the bit depth at all, it just produces the
524 effect. Material reduced in bit depth sounds more harsh and "digital".
525 This filter is able to even round to continuous values instead of discrete
527 Additionally it has a D/C offset which results in different crushing of
528 the lower and the upper half of the signal.
529 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
531 Another feature of this filter is the logarithmic mode.
532 This setting switches from linear distances between bits to logarithmic ones.
533 The result is a much more "natural" sounding crusher which doesn't gate low
534 signals for example. The human ear has a logarithmic perception,
535 so this kind of crushing is much more pleasant.
536 Logarithmic crushing is also able to get anti-aliased.
538 The filter accepts the following options:
554 Can be linear: @code{lin} or logarithmic: @code{log}.
563 Set sample reduction.
566 Enable LFO. By default disabled.
577 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
581 Remove impulsive noise from input audio.
583 Samples detected as impulsive noise are replaced by interpolated samples using
584 autoregressive modelling.
588 Set window size, in milliseconds. Allowed range is from @code{10} to
589 @code{100}. Default value is @code{55} milliseconds.
590 This sets size of window which will be processed at once.
593 Set window overlap, in percentage of window size. Allowed range is from
594 @code{50} to @code{95}. Default value is @code{75} percent.
595 Setting this to a very high value increases impulsive noise removal but makes
596 whole process much slower.
599 Set autoregression order, in percentage of window size. Allowed range is from
600 @code{0} to @code{25}. Default value is @code{2} percent. This option also
601 controls quality of interpolated samples using neighbour good samples.
604 Set threshold value. Allowed range is from @code{1} to @code{100}.
605 Default value is @code{2}.
606 This controls the strength of impulsive noise which is going to be removed.
607 The lower value, the more samples will be detected as impulsive noise.
610 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
611 @code{10}. Default value is @code{2}.
612 If any two samples detected as noise are spaced less than this value then any
613 sample between those two samples will be also detected as noise.
618 It accepts the following values:
621 Select overlap-add method. Even not interpolated samples are slightly
622 changed with this method.
625 Select overlap-save method. Not interpolated samples remain unchanged.
628 Default value is @code{a}.
632 Remove clipped samples from input audio.
634 Samples detected as clipped are replaced by interpolated samples using
635 autoregressive modelling.
639 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
640 Default value is @code{55} milliseconds.
641 This sets size of window which will be processed at once.
644 Set window overlap, in percentage of window size. Allowed range is from @code{50}
645 to @code{95}. Default value is @code{75} percent.
648 Set autoregression order, in percentage of window size. Allowed range is from
649 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
650 quality of interpolated samples using neighbour good samples.
653 Set threshold value. Allowed range is from @code{1} to @code{100}.
654 Default value is @code{10}. Higher values make clip detection less aggressive.
657 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
658 Default value is @code{1000}. Higher values make clip detection less aggressive.
663 It accepts the following values:
666 Select overlap-add method. Even not interpolated samples are slightly changed
670 Select overlap-save method. Not interpolated samples remain unchanged.
673 Default value is @code{a}.
678 Delay one or more audio channels.
680 Samples in delayed channel are filled with silence.
682 The filter accepts the following option:
686 Set list of delays in milliseconds for each channel separated by '|'.
687 Unused delays will be silently ignored. If number of given delays is
688 smaller than number of channels all remaining channels will not be delayed.
689 If you want to delay exact number of samples, append 'S' to number.
690 If you want instead to delay in seconds, append 's' to number.
697 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
698 the second channel (and any other channels that may be present) unchanged.
704 Delay second channel by 500 samples, the third channel by 700 samples and leave
705 the first channel (and any other channels that may be present) unchanged.
711 @section aderivative, aintegral
713 Compute derivative/integral of audio stream.
715 Applying both filters one after another produces original audio.
719 Apply echoing to the input audio.
721 Echoes are reflected sound and can occur naturally amongst mountains
722 (and sometimes large buildings) when talking or shouting; digital echo
723 effects emulate this behaviour and are often used to help fill out the
724 sound of a single instrument or vocal. The time difference between the
725 original signal and the reflection is the @code{delay}, and the
726 loudness of the reflected signal is the @code{decay}.
727 Multiple echoes can have different delays and decays.
729 A description of the accepted parameters follows.
733 Set input gain of reflected signal. Default is @code{0.6}.
736 Set output gain of reflected signal. Default is @code{0.3}.
739 Set list of time intervals in milliseconds between original signal and reflections
740 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
741 Default is @code{1000}.
744 Set list of loudness of reflected signals separated by '|'.
745 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
746 Default is @code{0.5}.
753 Make it sound as if there are twice as many instruments as are actually playing:
755 aecho=0.8:0.88:60:0.4
759 If delay is very short, then it sound like a (metallic) robot playing music:
765 A longer delay will sound like an open air concert in the mountains:
767 aecho=0.8:0.9:1000:0.3
771 Same as above but with one more mountain:
773 aecho=0.8:0.9:1000|1800:0.3|0.25
778 Audio emphasis filter creates or restores material directly taken from LPs or
779 emphased CDs with different filter curves. E.g. to store music on vinyl the
780 signal has to be altered by a filter first to even out the disadvantages of
781 this recording medium.
782 Once the material is played back the inverse filter has to be applied to
783 restore the distortion of the frequency response.
785 The filter accepts the following options:
795 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
796 use @code{production} mode. Default is @code{reproduction} mode.
799 Set filter type. Selects medium. Can be one of the following:
811 select Compact Disc (CD).
817 select 50µs (FM-KF).
819 select 75µs (FM-KF).
825 Modify an audio signal according to the specified expressions.
827 This filter accepts one or more expressions (one for each channel),
828 which are evaluated and used to modify a corresponding audio signal.
830 It accepts the following parameters:
834 Set the '|'-separated expressions list for each separate channel. If
835 the number of input channels is greater than the number of
836 expressions, the last specified expression is used for the remaining
839 @item channel_layout, c
840 Set output channel layout. If not specified, the channel layout is
841 specified by the number of expressions. If set to @samp{same}, it will
842 use by default the same input channel layout.
845 Each expression in @var{exprs} can contain the following constants and functions:
849 channel number of the current expression
852 number of the evaluated sample, starting from 0
858 time of the evaluated sample expressed in seconds
861 @item nb_out_channels
862 input and output number of channels
865 the value of input channel with number @var{CH}
868 Note: this filter is slow. For faster processing you should use a
877 aeval=val(ch)/2:c=same
881 Invert phase of the second channel:
890 Apply fade-in/out effect to input audio.
892 A description of the accepted parameters follows.
896 Specify the effect type, can be either @code{in} for fade-in, or
897 @code{out} for a fade-out effect. Default is @code{in}.
899 @item start_sample, ss
900 Specify the number of the start sample for starting to apply the fade
901 effect. Default is 0.
904 Specify the number of samples for which the fade effect has to last. At
905 the end of the fade-in effect the output audio will have the same
906 volume as the input audio, at the end of the fade-out transition
907 the output audio will be silence. Default is 44100.
910 Specify the start time of the fade effect. Default is 0.
911 The value must be specified as a time duration; see
912 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
913 for the accepted syntax.
914 If set this option is used instead of @var{start_sample}.
917 Specify the duration of the fade effect. See
918 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
919 for the accepted syntax.
920 At the end of the fade-in effect the output audio will have the same
921 volume as the input audio, at the end of the fade-out transition
922 the output audio will be silence.
923 By default the duration is determined by @var{nb_samples}.
924 If set this option is used instead of @var{nb_samples}.
927 Set curve for fade transition.
929 It accepts the following values:
932 select triangular, linear slope (default)
934 select quarter of sine wave
936 select half of sine wave
938 select exponential sine wave
942 select inverted parabola
956 select inverted quarter of sine wave
958 select inverted half of sine wave
960 select double-exponential seat
962 select double-exponential sigmoid
964 select logistic sigmoid
974 Fade in first 15 seconds of audio:
980 Fade out last 25 seconds of a 900 seconds audio:
982 afade=t=out:st=875:d=25
987 Denoise audio samples with FFT.
989 A description of the accepted parameters follows.
993 Set the noise reduction in dB, allowed range is 0.01 to 97.
994 Default value is 12 dB.
997 Set the noise floor in dB, allowed range is -80 to -20.
998 Default value is -50 dB.
1003 It accepts the following values:
1012 Select shellac noise.
1015 Select custom noise, defined in @code{bn} option.
1017 Default value is white noise.
1021 Set custom band noise for every one of 15 bands.
1022 Bands are separated by ' ' or '|'.
1025 Set the residual floor in dB, allowed range is -80 to -20.
1026 Default value is -38 dB.
1029 Enable noise tracking. By default is disabled.
1030 With this enabled, noise floor is automatically adjusted.
1033 Enable residual tracking. By default is disabled.
1036 Set the output mode.
1038 It accepts the following values:
1041 Pass input unchanged.
1044 Pass noise filtered out.
1049 Default value is @var{o}.
1053 @subsection Commands
1055 This filter supports the following commands:
1057 @item sample_noise, sn
1058 Start or stop measuring noise profile.
1059 Syntax for the command is : "start" or "stop" string.
1060 After measuring noise profile is stopped it will be
1061 automatically applied in filtering.
1063 @item noise_reduction, nr
1064 Change noise reduction. Argument is single float number.
1065 Syntax for the command is : "@var{noise_reduction}"
1067 @item noise_floor, nf
1068 Change noise floor. Argument is single float number.
1069 Syntax for the command is : "@var{noise_floor}"
1071 @item output_mode, om
1072 Change output mode operation.
1073 Syntax for the command is : "i", "o" or "n" string.
1077 Apply arbitrary expressions to samples in frequency domain.
1081 Set frequency domain real expression for each separate channel separated
1082 by '|'. Default is "re".
1083 If the number of input channels is greater than the number of
1084 expressions, the last specified expression is used for the remaining
1088 Set frequency domain imaginary expression for each separate channel
1089 separated by '|'. Default is "im".
1091 Each expression in @var{real} and @var{imag} can contain the following
1092 constants and functions:
1099 current frequency bin number
1102 number of available bins
1105 channel number of the current expression
1114 current real part of frequency bin of current channel
1117 current imaginary part of frequency bin of current channel
1120 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1123 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1127 Set window size. Allowed range is from 16 to 131072.
1128 Default is @code{4096}
1131 Set window function. Default is @code{hann}.
1134 Set window overlap. If set to 1, the recommended overlap for selected
1135 window function will be picked. Default is @code{0.75}.
1138 @subsection Examples
1142 Leave almost only low frequencies in audio:
1144 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1151 Apply an arbitrary Frequency Impulse Response filter.
1153 This filter is designed for applying long FIR filters,
1154 up to 60 seconds long.
1156 It can be used as component for digital crossover filters,
1157 room equalization, cross talk cancellation, wavefield synthesis,
1158 auralization, ambiophonics, ambisonics and spatialization.
1160 This filter uses second stream as FIR coefficients.
1161 If second stream holds single channel, it will be used
1162 for all input channels in first stream, otherwise
1163 number of channels in second stream must be same as
1164 number of channels in first stream.
1166 It accepts the following parameters:
1170 Set dry gain. This sets input gain.
1173 Set wet gain. This sets final output gain.
1176 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1179 Enable applying gain measured from power of IR.
1181 Set which approach to use for auto gain measurement.
1185 Do not apply any gain.
1188 select peak gain, very conservative approach. This is default value.
1191 select DC gain, limited application.
1194 select gain to noise approach, this is most popular one.
1198 Set gain to be applied to IR coefficients before filtering.
1199 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1202 Set format of IR stream. Can be @code{mono} or @code{input}.
1203 Default is @code{input}.
1206 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1207 Allowed range is 0.1 to 60 seconds.
1210 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1211 By default it is disabled.
1214 Set for which IR channel to display frequency response. By default is first channel
1215 displayed. This option is used only when @var{response} is enabled.
1218 Set video stream size. This option is used only when @var{response} is enabled.
1221 Set video stream frame rate. This option is used only when @var{response} is enabled.
1224 Set minimal partition size used for convolution. Default is @var{8192}.
1225 Allowed range is from @var{8} to @var{32768}.
1226 Lower values decreases latency at cost of higher CPU usage.
1229 Set maximal partition size used for convolution. Default is @var{8192}.
1230 Allowed range is from @var{8} to @var{32768}.
1231 Lower values may increase CPU usage.
1234 @subsection Examples
1238 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1240 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1247 Set output format constraints for the input audio. The framework will
1248 negotiate the most appropriate format to minimize conversions.
1250 It accepts the following parameters:
1254 A '|'-separated list of requested sample formats.
1257 A '|'-separated list of requested sample rates.
1259 @item channel_layouts
1260 A '|'-separated list of requested channel layouts.
1262 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1263 for the required syntax.
1266 If a parameter is omitted, all values are allowed.
1268 Force the output to either unsigned 8-bit or signed 16-bit stereo
1270 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1275 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1276 processing reduces disturbing noise between useful signals.
1278 Gating is done by detecting the volume below a chosen level @var{threshold}
1279 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1280 floor is set via @var{range}. Because an exact manipulation of the signal
1281 would cause distortion of the waveform the reduction can be levelled over
1282 time. This is done by setting @var{attack} and @var{release}.
1284 @var{attack} determines how long the signal has to fall below the threshold
1285 before any reduction will occur and @var{release} sets the time the signal
1286 has to rise above the threshold to reduce the reduction again.
1287 Shorter signals than the chosen attack time will be left untouched.
1291 Set input level before filtering.
1292 Default is 1. Allowed range is from 0.015625 to 64.
1295 Set the mode of operation. Can be @code{upward} or @code{downward}.
1296 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1297 will be amplified, expanding dynamic range in upward direction.
1298 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1301 Set the level of gain reduction when the signal is below the threshold.
1302 Default is 0.06125. Allowed range is from 0 to 1.
1303 Setting this to 0 disables reduction and then filter behaves like expander.
1306 If a signal rises above this level the gain reduction is released.
1307 Default is 0.125. Allowed range is from 0 to 1.
1310 Set a ratio by which the signal is reduced.
1311 Default is 2. Allowed range is from 1 to 9000.
1314 Amount of milliseconds the signal has to rise above the threshold before gain
1316 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1319 Amount of milliseconds the signal has to fall below the threshold before the
1320 reduction is increased again. Default is 250 milliseconds.
1321 Allowed range is from 0.01 to 9000.
1324 Set amount of amplification of signal after processing.
1325 Default is 1. Allowed range is from 1 to 64.
1328 Curve the sharp knee around the threshold to enter gain reduction more softly.
1329 Default is 2.828427125. Allowed range is from 1 to 8.
1332 Choose if exact signal should be taken for detection or an RMS like one.
1333 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1336 Choose if the average level between all channels or the louder channel affects
1338 Default is @code{average}. Can be @code{average} or @code{maximum}.
1343 Apply an arbitrary Infinite Impulse Response filter.
1345 It accepts the following parameters:
1349 Set numerator/zeros coefficients.
1352 Set denominator/poles coefficients.
1364 Set coefficients format.
1370 Z-plane zeros/poles, cartesian (default)
1372 Z-plane zeros/poles, polar radians
1374 Z-plane zeros/poles, polar degrees
1378 Set kind of processing.
1379 Can be @code{d} - direct or @code{s} - serial cascading. Default is @code{s}.
1382 Set filtering precision.
1386 double-precision floating-point (default)
1388 single-precision floating-point
1396 How much to use filtered signal in output. Default is 1.
1397 Range is between 0 and 1.
1400 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1401 By default it is disabled.
1404 Set for which IR channel to display frequency response. By default is first channel
1405 displayed. This option is used only when @var{response} is enabled.
1408 Set video stream size. This option is used only when @var{response} is enabled.
1411 Coefficients in @code{tf} format are separated by spaces and are in ascending
1414 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1415 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1418 Different coefficients and gains can be provided for every channel, in such case
1419 use '|' to separate coefficients or gains. Last provided coefficients will be
1420 used for all remaining channels.
1422 @subsection Examples
1426 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1428 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1432 Same as above but in @code{zp} format:
1434 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1440 The limiter prevents an input signal from rising over a desired threshold.
1441 This limiter uses lookahead technology to prevent your signal from distorting.
1442 It means that there is a small delay after the signal is processed. Keep in mind
1443 that the delay it produces is the attack time you set.
1445 The filter accepts the following options:
1449 Set input gain. Default is 1.
1452 Set output gain. Default is 1.
1455 Don't let signals above this level pass the limiter. Default is 1.
1458 The limiter will reach its attenuation level in this amount of time in
1459 milliseconds. Default is 5 milliseconds.
1462 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1463 Default is 50 milliseconds.
1466 When gain reduction is always needed ASC takes care of releasing to an
1467 average reduction level rather than reaching a reduction of 0 in the release
1471 Select how much the release time is affected by ASC, 0 means nearly no changes
1472 in release time while 1 produces higher release times.
1475 Auto level output signal. Default is enabled.
1476 This normalizes audio back to 0dB if enabled.
1479 Depending on picked setting it is recommended to upsample input 2x or 4x times
1480 with @ref{aresample} before applying this filter.
1484 Apply a two-pole all-pass filter with central frequency (in Hz)
1485 @var{frequency}, and filter-width @var{width}.
1486 An all-pass filter changes the audio's frequency to phase relationship
1487 without changing its frequency to amplitude relationship.
1489 The filter accepts the following options:
1493 Set frequency in Hz.
1496 Set method to specify band-width of filter.
1511 Specify the band-width of a filter in width_type units.
1514 How much to use filtered signal in output. Default is 1.
1515 Range is between 0 and 1.
1518 Specify which channels to filter, by default all available are filtered.
1521 @subsection Commands
1523 This filter supports the following commands:
1526 Change allpass frequency.
1527 Syntax for the command is : "@var{frequency}"
1530 Change allpass width_type.
1531 Syntax for the command is : "@var{width_type}"
1534 Change allpass width.
1535 Syntax for the command is : "@var{width}"
1539 Syntax for the command is : "@var{mix}"
1546 The filter accepts the following options:
1550 Set the number of loops. Setting this value to -1 will result in infinite loops.
1554 Set maximal number of samples. Default is 0.
1557 Set first sample of loop. Default is 0.
1563 Merge two or more audio streams into a single multi-channel stream.
1565 The filter accepts the following options:
1570 Set the number of inputs. Default is 2.
1574 If the channel layouts of the inputs are disjoint, and therefore compatible,
1575 the channel layout of the output will be set accordingly and the channels
1576 will be reordered as necessary. If the channel layouts of the inputs are not
1577 disjoint, the output will have all the channels of the first input then all
1578 the channels of the second input, in that order, and the channel layout of
1579 the output will be the default value corresponding to the total number of
1582 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1583 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1584 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1585 first input, b1 is the first channel of the second input).
1587 On the other hand, if both input are in stereo, the output channels will be
1588 in the default order: a1, a2, b1, b2, and the channel layout will be
1589 arbitrarily set to 4.0, which may or may not be the expected value.
1591 All inputs must have the same sample rate, and format.
1593 If inputs do not have the same duration, the output will stop with the
1596 @subsection Examples
1600 Merge two mono files into a stereo stream:
1602 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1606 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1608 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
1614 Mixes multiple audio inputs into a single output.
1616 Note that this filter only supports float samples (the @var{amerge}
1617 and @var{pan} audio filters support many formats). If the @var{amix}
1618 input has integer samples then @ref{aresample} will be automatically
1619 inserted to perform the conversion to float samples.
1623 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1625 will mix 3 input audio streams to a single output with the same duration as the
1626 first input and a dropout transition time of 3 seconds.
1628 It accepts the following parameters:
1632 The number of inputs. If unspecified, it defaults to 2.
1635 How to determine the end-of-stream.
1639 The duration of the longest input. (default)
1642 The duration of the shortest input.
1645 The duration of the first input.
1649 @item dropout_transition
1650 The transition time, in seconds, for volume renormalization when an input
1651 stream ends. The default value is 2 seconds.
1654 Specify weight of each input audio stream as sequence.
1655 Each weight is separated by space. By default all inputs have same weight.
1660 Multiply first audio stream with second audio stream and store result
1661 in output audio stream. Multiplication is done by multiplying each
1662 sample from first stream with sample at same position from second stream.
1664 With this element-wise multiplication one can create amplitude fades and
1665 amplitude modulations.
1667 @section anequalizer
1669 High-order parametric multiband equalizer for each channel.
1671 It accepts the following parameters:
1675 This option string is in format:
1676 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1677 Each equalizer band is separated by '|'.
1681 Set channel number to which equalization will be applied.
1682 If input doesn't have that channel the entry is ignored.
1685 Set central frequency for band.
1686 If input doesn't have that frequency the entry is ignored.
1689 Set band width in hertz.
1692 Set band gain in dB.
1695 Set filter type for band, optional, can be:
1699 Butterworth, this is default.
1710 With this option activated frequency response of anequalizer is displayed
1714 Set video stream size. Only useful if curves option is activated.
1717 Set max gain that will be displayed. Only useful if curves option is activated.
1718 Setting this to a reasonable value makes it possible to display gain which is derived from
1719 neighbour bands which are too close to each other and thus produce higher gain
1720 when both are activated.
1723 Set frequency scale used to draw frequency response in video output.
1724 Can be linear or logarithmic. Default is logarithmic.
1727 Set color for each channel curve which is going to be displayed in video stream.
1728 This is list of color names separated by space or by '|'.
1729 Unrecognised or missing colors will be replaced by white color.
1732 @subsection Examples
1736 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1737 for first 2 channels using Chebyshev type 1 filter:
1739 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1743 @subsection Commands
1745 This filter supports the following commands:
1748 Alter existing filter parameters.
1749 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1751 @var{fN} is existing filter number, starting from 0, if no such filter is available
1753 @var{freq} set new frequency parameter.
1754 @var{width} set new width parameter in herz.
1755 @var{gain} set new gain parameter in dB.
1757 Full filter invocation with asendcmd may look like this:
1758 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1763 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1765 Each sample is adjusted by looking for other samples with similar contexts. This
1766 context similarity is defined by comparing their surrounding patches of size
1767 @option{p}. Patches are searched in an area of @option{r} around the sample.
1769 The filter accepts the following options.
1773 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
1776 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
1777 Default value is 2 milliseconds.
1780 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
1781 Default value is 6 milliseconds.
1784 Set the output mode.
1786 It accepts the following values:
1789 Pass input unchanged.
1792 Pass noise filtered out.
1797 Default value is @var{o}.
1801 Set smooth factor. Default value is @var{11}. Allowed range is from @var{1} to @var{15}.
1804 @subsection Commands
1806 This filter supports the following commands:
1809 Change denoise strength. Argument is single float number.
1810 Syntax for the command is : "@var{s}"
1814 Syntax for the command is : "i", "o" or "n" string.
1819 Pass the audio source unchanged to the output.
1823 Pad the end of an audio stream with silence.
1825 This can be used together with @command{ffmpeg} @option{-shortest} to
1826 extend audio streams to the same length as the video stream.
1828 A description of the accepted options follows.
1832 Set silence packet size. Default value is 4096.
1835 Set the number of samples of silence to add to the end. After the
1836 value is reached, the stream is terminated. This option is mutually
1837 exclusive with @option{whole_len}.
1840 Set the minimum total number of samples in the output audio stream. If
1841 the value is longer than the input audio length, silence is added to
1842 the end, until the value is reached. This option is mutually exclusive
1843 with @option{pad_len}.
1846 Specify the duration of samples of silence to add. See
1847 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1848 for the accepted syntax. Used only if set to non-zero value.
1851 Specify the minimum total duration in the output audio stream. See
1852 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1853 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1854 the input audio length, silence is added to the end, until the value is reached.
1855 This option is mutually exclusive with @option{pad_dur}
1858 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1859 nor @option{whole_dur} option is set, the filter will add silence to the end of
1860 the input stream indefinitely.
1862 @subsection Examples
1866 Add 1024 samples of silence to the end of the input:
1872 Make sure the audio output will contain at least 10000 samples, pad
1873 the input with silence if required:
1875 apad=whole_len=10000
1879 Use @command{ffmpeg} to pad the audio input with silence, so that the
1880 video stream will always result the shortest and will be converted
1881 until the end in the output file when using the @option{shortest}
1884 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1889 Add a phasing effect to the input audio.
1891 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1892 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1894 A description of the accepted parameters follows.
1898 Set input gain. Default is 0.4.
1901 Set output gain. Default is 0.74
1904 Set delay in milliseconds. Default is 3.0.
1907 Set decay. Default is 0.4.
1910 Set modulation speed in Hz. Default is 0.5.
1913 Set modulation type. Default is triangular.
1915 It accepts the following values:
1924 Audio pulsator is something between an autopanner and a tremolo.
1925 But it can produce funny stereo effects as well. Pulsator changes the volume
1926 of the left and right channel based on a LFO (low frequency oscillator) with
1927 different waveforms and shifted phases.
1928 This filter have the ability to define an offset between left and right
1929 channel. An offset of 0 means that both LFO shapes match each other.
1930 The left and right channel are altered equally - a conventional tremolo.
1931 An offset of 50% means that the shape of the right channel is exactly shifted
1932 in phase (or moved backwards about half of the frequency) - pulsator acts as
1933 an autopanner. At 1 both curves match again. Every setting in between moves the
1934 phase shift gapless between all stages and produces some "bypassing" sounds with
1935 sine and triangle waveforms. The more you set the offset near 1 (starting from
1936 the 0.5) the faster the signal passes from the left to the right speaker.
1938 The filter accepts the following options:
1942 Set input gain. By default it is 1. Range is [0.015625 - 64].
1945 Set output gain. By default it is 1. Range is [0.015625 - 64].
1948 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1949 sawup or sawdown. Default is sine.
1952 Set modulation. Define how much of original signal is affected by the LFO.
1955 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1958 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1961 Set pulse width. Default is 1. Allowed range is [0 - 2].
1964 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1967 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1971 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1975 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1976 if timing is set to hz.
1982 Resample the input audio to the specified parameters, using the
1983 libswresample library. If none are specified then the filter will
1984 automatically convert between its input and output.
1986 This filter is also able to stretch/squeeze the audio data to make it match
1987 the timestamps or to inject silence / cut out audio to make it match the
1988 timestamps, do a combination of both or do neither.
1990 The filter accepts the syntax
1991 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1992 expresses a sample rate and @var{resampler_options} is a list of
1993 @var{key}=@var{value} pairs, separated by ":". See the
1994 @ref{Resampler Options,,"Resampler Options" section in the
1995 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1996 for the complete list of supported options.
1998 @subsection Examples
2002 Resample the input audio to 44100Hz:
2008 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
2009 samples per second compensation:
2011 aresample=async=1000
2017 Reverse an audio clip.
2019 Warning: This filter requires memory to buffer the entire clip, so trimming
2022 @subsection Examples
2026 Take the first 5 seconds of a clip, and reverse it.
2028 atrim=end=5,areverse
2032 @section asetnsamples
2034 Set the number of samples per each output audio frame.
2036 The last output packet may contain a different number of samples, as
2037 the filter will flush all the remaining samples when the input audio
2040 The filter accepts the following options:
2044 @item nb_out_samples, n
2045 Set the number of frames per each output audio frame. The number is
2046 intended as the number of samples @emph{per each channel}.
2047 Default value is 1024.
2050 If set to 1, the filter will pad the last audio frame with zeroes, so
2051 that the last frame will contain the same number of samples as the
2052 previous ones. Default value is 1.
2055 For example, to set the number of per-frame samples to 1234 and
2056 disable padding for the last frame, use:
2058 asetnsamples=n=1234:p=0
2063 Set the sample rate without altering the PCM data.
2064 This will result in a change of speed and pitch.
2066 The filter accepts the following options:
2069 @item sample_rate, r
2070 Set the output sample rate. Default is 44100 Hz.
2075 Show a line containing various information for each input audio frame.
2076 The input audio is not modified.
2078 The shown line contains a sequence of key/value pairs of the form
2079 @var{key}:@var{value}.
2081 The following values are shown in the output:
2085 The (sequential) number of the input frame, starting from 0.
2088 The presentation timestamp of the input frame, in time base units; the time base
2089 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2092 The presentation timestamp of the input frame in seconds.
2095 position of the frame in the input stream, -1 if this information in
2096 unavailable and/or meaningless (for example in case of synthetic audio)
2105 The sample rate for the audio frame.
2108 The number of samples (per channel) in the frame.
2111 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2112 audio, the data is treated as if all the planes were concatenated.
2114 @item plane_checksums
2115 A list of Adler-32 checksums for each data plane.
2119 Apply audio soft clipping.
2121 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2122 along a smooth curve, rather than the abrupt shape of hard-clipping.
2124 This filter accepts the following options:
2128 Set type of soft-clipping.
2130 It accepts the following values:
2142 Set additional parameter which controls sigmoid function.
2146 Automatic Speech Recognition
2148 This filter uses PocketSphinx for speech recognition. To enable
2149 compilation of this filter, you need to configure FFmpeg with
2150 @code{--enable-pocketsphinx}.
2152 It accepts the following options:
2156 Set sampling rate of input audio. Defaults is @code{16000}.
2157 This need to match speech models, otherwise one will get poor results.
2160 Set dictionary containing acoustic model files.
2163 Set pronunciation dictionary.
2166 Set language model file.
2169 Set language model set.
2172 Set which language model to use.
2175 Set output for log messages.
2178 The filter exports recognized speech as the frame metadata @code{lavfi.asr.text}.
2183 Display time domain statistical information about the audio channels.
2184 Statistics are calculated and displayed for each audio channel and,
2185 where applicable, an overall figure is also given.
2187 It accepts the following option:
2190 Short window length in seconds, used for peak and trough RMS measurement.
2191 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2195 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2196 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2199 Available keys for each channel are:
2241 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2242 this @code{lavfi.astats.Overall.Peak_count}.
2244 For description what each key means read below.
2247 Set number of frame after which stats are going to be recalculated.
2248 Default is disabled.
2250 @item measure_perchannel
2251 Select the entries which need to be measured per channel. The metadata keys can
2252 be used as flags, default is @option{all} which measures everything.
2253 @option{none} disables all per channel measurement.
2255 @item measure_overall
2256 Select the entries which need to be measured overall. The metadata keys can
2257 be used as flags, default is @option{all} which measures everything.
2258 @option{none} disables all overall measurement.
2262 A description of each shown parameter follows:
2266 Mean amplitude displacement from zero.
2269 Minimal sample level.
2272 Maximal sample level.
2274 @item Min difference
2275 Minimal difference between two consecutive samples.
2277 @item Max difference
2278 Maximal difference between two consecutive samples.
2280 @item Mean difference
2281 Mean difference between two consecutive samples.
2282 The average of each difference between two consecutive samples.
2284 @item RMS difference
2285 Root Mean Square difference between two consecutive samples.
2289 Standard peak and RMS level measured in dBFS.
2293 Peak and trough values for RMS level measured over a short window.
2296 Standard ratio of peak to RMS level (note: not in dB).
2299 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2300 (i.e. either @var{Min level} or @var{Max level}).
2303 Number of occasions (not the number of samples) that the signal attained either
2304 @var{Min level} or @var{Max level}.
2307 Overall bit depth of audio. Number of bits used for each sample.
2310 Measured dynamic range of audio in dB.
2312 @item Zero crossings
2313 Number of points where the waveform crosses the zero level axis.
2315 @item Zero crossings rate
2316 Rate of Zero crossings and number of audio samples.
2323 The filter accepts exactly one parameter, the audio tempo. If not
2324 specified then the filter will assume nominal 1.0 tempo. Tempo must
2325 be in the [0.5, 100.0] range.
2327 Note that tempo greater than 2 will skip some samples rather than
2328 blend them in. If for any reason this is a concern it is always
2329 possible to daisy-chain several instances of atempo to achieve the
2330 desired product tempo.
2332 @subsection Examples
2336 Slow down audio to 80% tempo:
2342 To speed up audio to 300% tempo:
2348 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2350 atempo=sqrt(3),atempo=sqrt(3)
2356 Trim the input so that the output contains one continuous subpart of the input.
2358 It accepts the following parameters:
2361 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2362 sample with the timestamp @var{start} will be the first sample in the output.
2365 Specify time of the first audio sample that will be dropped, i.e. the
2366 audio sample immediately preceding the one with the timestamp @var{end} will be
2367 the last sample in the output.
2370 Same as @var{start}, except this option sets the start timestamp in samples
2374 Same as @var{end}, except this option sets the end timestamp in samples instead
2378 The maximum duration of the output in seconds.
2381 The number of the first sample that should be output.
2384 The number of the first sample that should be dropped.
2387 @option{start}, @option{end}, and @option{duration} are expressed as time
2388 duration specifications; see
2389 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2391 Note that the first two sets of the start/end options and the @option{duration}
2392 option look at the frame timestamp, while the _sample options simply count the
2393 samples that pass through the filter. So start/end_pts and start/end_sample will
2394 give different results when the timestamps are wrong, inexact or do not start at
2395 zero. Also note that this filter does not modify the timestamps. If you wish
2396 to have the output timestamps start at zero, insert the asetpts filter after the
2399 If multiple start or end options are set, this filter tries to be greedy and
2400 keep all samples that match at least one of the specified constraints. To keep
2401 only the part that matches all the constraints at once, chain multiple atrim
2404 The defaults are such that all the input is kept. So it is possible to set e.g.
2405 just the end values to keep everything before the specified time.
2410 Drop everything except the second minute of input:
2412 ffmpeg -i INPUT -af atrim=60:120
2416 Keep only the first 1000 samples:
2418 ffmpeg -i INPUT -af atrim=end_sample=1000
2425 Apply a two-pole Butterworth band-pass filter with central
2426 frequency @var{frequency}, and (3dB-point) band-width width.
2427 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2428 instead of the default: constant 0dB peak gain.
2429 The filter roll off at 6dB per octave (20dB per decade).
2431 The filter accepts the following options:
2435 Set the filter's central frequency. Default is @code{3000}.
2438 Constant skirt gain if set to 1. Defaults to 0.
2441 Set method to specify band-width of filter.
2456 Specify the band-width of a filter in width_type units.
2459 How much to use filtered signal in output. Default is 1.
2460 Range is between 0 and 1.
2463 Specify which channels to filter, by default all available are filtered.
2466 @subsection Commands
2468 This filter supports the following commands:
2471 Change bandpass frequency.
2472 Syntax for the command is : "@var{frequency}"
2475 Change bandpass width_type.
2476 Syntax for the command is : "@var{width_type}"
2479 Change bandpass width.
2480 Syntax for the command is : "@var{width}"
2483 Change bandpass mix.
2484 Syntax for the command is : "@var{mix}"
2489 Apply a two-pole Butterworth band-reject filter with central
2490 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2491 The filter roll off at 6dB per octave (20dB per decade).
2493 The filter accepts the following options:
2497 Set the filter's central frequency. Default is @code{3000}.
2500 Set method to specify band-width of filter.
2515 Specify the band-width of a filter in width_type units.
2518 How much to use filtered signal in output. Default is 1.
2519 Range is between 0 and 1.
2522 Specify which channels to filter, by default all available are filtered.
2525 @subsection Commands
2527 This filter supports the following commands:
2530 Change bandreject frequency.
2531 Syntax for the command is : "@var{frequency}"
2534 Change bandreject width_type.
2535 Syntax for the command is : "@var{width_type}"
2538 Change bandreject width.
2539 Syntax for the command is : "@var{width}"
2542 Change bandreject mix.
2543 Syntax for the command is : "@var{mix}"
2546 @section bass, lowshelf
2548 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2549 shelving filter with a response similar to that of a standard
2550 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2552 The filter accepts the following options:
2556 Give the gain at 0 Hz. Its useful range is about -20
2557 (for a large cut) to +20 (for a large boost).
2558 Beware of clipping when using a positive gain.
2561 Set the filter's central frequency and so can be used
2562 to extend or reduce the frequency range to be boosted or cut.
2563 The default value is @code{100} Hz.
2566 Set method to specify band-width of filter.
2581 Determine how steep is the filter's shelf transition.
2584 How much to use filtered signal in output. Default is 1.
2585 Range is between 0 and 1.
2588 Specify which channels to filter, by default all available are filtered.
2591 @subsection Commands
2593 This filter supports the following commands:
2596 Change bass frequency.
2597 Syntax for the command is : "@var{frequency}"
2600 Change bass width_type.
2601 Syntax for the command is : "@var{width_type}"
2605 Syntax for the command is : "@var{width}"
2609 Syntax for the command is : "@var{gain}"
2613 Syntax for the command is : "@var{mix}"
2618 Apply a biquad IIR filter with the given coefficients.
2619 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2620 are the numerator and denominator coefficients respectively.
2621 and @var{channels}, @var{c} specify which channels to filter, by default all
2622 available are filtered.
2624 @subsection Commands
2626 This filter supports the following commands:
2634 Change biquad parameter.
2635 Syntax for the command is : "@var{value}"
2638 How much to use filtered signal in output. Default is 1.
2639 Range is between 0 and 1.
2643 Bauer stereo to binaural transformation, which improves headphone listening of
2644 stereo audio records.
2646 To enable compilation of this filter you need to configure FFmpeg with
2647 @code{--enable-libbs2b}.
2649 It accepts the following parameters:
2653 Pre-defined crossfeed level.
2657 Default level (fcut=700, feed=50).
2660 Chu Moy circuit (fcut=700, feed=60).
2663 Jan Meier circuit (fcut=650, feed=95).
2668 Cut frequency (in Hz).
2677 Remap input channels to new locations.
2679 It accepts the following parameters:
2682 Map channels from input to output. The argument is a '|'-separated list of
2683 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2684 @var{in_channel} form. @var{in_channel} can be either the name of the input
2685 channel (e.g. FL for front left) or its index in the input channel layout.
2686 @var{out_channel} is the name of the output channel or its index in the output
2687 channel layout. If @var{out_channel} is not given then it is implicitly an
2688 index, starting with zero and increasing by one for each mapping.
2690 @item channel_layout
2691 The channel layout of the output stream.
2694 If no mapping is present, the filter will implicitly map input channels to
2695 output channels, preserving indices.
2697 @subsection Examples
2701 For example, assuming a 5.1+downmix input MOV file,
2703 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2705 will create an output WAV file tagged as stereo from the downmix channels of
2709 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2711 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2715 @section channelsplit
2717 Split each channel from an input audio stream into a separate output stream.
2719 It accepts the following parameters:
2721 @item channel_layout
2722 The channel layout of the input stream. The default is "stereo".
2724 A channel layout describing the channels to be extracted as separate output streams
2725 or "all" to extract each input channel as a separate stream. The default is "all".
2727 Choosing channels not present in channel layout in the input will result in an error.
2730 @subsection Examples
2734 For example, assuming a stereo input MP3 file,
2736 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2738 will create an output Matroska file with two audio streams, one containing only
2739 the left channel and the other the right channel.
2742 Split a 5.1 WAV file into per-channel files:
2744 ffmpeg -i in.wav -filter_complex
2745 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2746 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2747 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2752 Extract only LFE from a 5.1 WAV file:
2754 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2755 -map '[LFE]' lfe.wav
2760 Add a chorus effect to the audio.
2762 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2764 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2765 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2766 The modulation depth defines the range the modulated delay is played before or after
2767 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2768 sound tuned around the original one, like in a chorus where some vocals are slightly
2771 It accepts the following parameters:
2774 Set input gain. Default is 0.4.
2777 Set output gain. Default is 0.4.
2780 Set delays. A typical delay is around 40ms to 60ms.
2792 @subsection Examples
2798 chorus=0.7:0.9:55:0.4:0.25:2
2804 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2808 Fuller sounding chorus with three delays:
2810 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
2815 Compress or expand the audio's dynamic range.
2817 It accepts the following parameters:
2823 A list of times in seconds for each channel over which the instantaneous level
2824 of the input signal is averaged to determine its volume. @var{attacks} refers to
2825 increase of volume and @var{decays} refers to decrease of volume. For most
2826 situations, the attack time (response to the audio getting louder) should be
2827 shorter than the decay time, because the human ear is more sensitive to sudden
2828 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2829 a typical value for decay is 0.8 seconds.
2830 If specified number of attacks & decays is lower than number of channels, the last
2831 set attack/decay will be used for all remaining channels.
2834 A list of points for the transfer function, specified in dB relative to the
2835 maximum possible signal amplitude. Each key points list must be defined using
2836 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2837 @code{x0/y0 x1/y1 x2/y2 ....}
2839 The input values must be in strictly increasing order but the transfer function
2840 does not have to be monotonically rising. The point @code{0/0} is assumed but
2841 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2842 function are @code{-70/-70|-60/-20|1/0}.
2845 Set the curve radius in dB for all joints. It defaults to 0.01.
2848 Set the additional gain in dB to be applied at all points on the transfer
2849 function. This allows for easy adjustment of the overall gain.
2853 Set an initial volume, in dB, to be assumed for each channel when filtering
2854 starts. This permits the user to supply a nominal level initially, so that, for
2855 example, a very large gain is not applied to initial signal levels before the
2856 companding has begun to operate. A typical value for audio which is initially
2857 quiet is -90 dB. It defaults to 0.
2860 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2861 delayed before being fed to the volume adjuster. Specifying a delay
2862 approximately equal to the attack/decay times allows the filter to effectively
2863 operate in predictive rather than reactive mode. It defaults to 0.
2867 @subsection Examples
2871 Make music with both quiet and loud passages suitable for listening to in a
2874 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2877 Another example for audio with whisper and explosion parts:
2879 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2883 A noise gate for when the noise is at a lower level than the signal:
2885 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2889 Here is another noise gate, this time for when the noise is at a higher level
2890 than the signal (making it, in some ways, similar to squelch):
2892 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2896 2:1 compression starting at -6dB:
2898 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2902 2:1 compression starting at -9dB:
2904 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2908 2:1 compression starting at -12dB:
2910 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2914 2:1 compression starting at -18dB:
2916 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2920 3:1 compression starting at -15dB:
2922 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2928 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2934 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
2938 Hard limiter at -6dB:
2940 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2944 Hard limiter at -12dB:
2946 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2950 Hard noise gate at -35 dB:
2952 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2958 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2962 @section compensationdelay
2964 Compensation Delay Line is a metric based delay to compensate differing
2965 positions of microphones or speakers.
2967 For example, you have recorded guitar with two microphones placed in
2968 different location. Because the front of sound wave has fixed speed in
2969 normal conditions, the phasing of microphones can vary and depends on
2970 their location and interposition. The best sound mix can be achieved when
2971 these microphones are in phase (synchronized). Note that distance of
2972 ~30 cm between microphones makes one microphone to capture signal in
2973 antiphase to another microphone. That makes the final mix sounding moody.
2974 This filter helps to solve phasing problems by adding different delays
2975 to each microphone track and make them synchronized.
2977 The best result can be reached when you take one track as base and
2978 synchronize other tracks one by one with it.
2979 Remember that synchronization/delay tolerance depends on sample rate, too.
2980 Higher sample rates will give more tolerance.
2982 It accepts the following parameters:
2986 Set millimeters distance. This is compensation distance for fine tuning.
2990 Set cm distance. This is compensation distance for tightening distance setup.
2994 Set meters distance. This is compensation distance for hard distance setup.
2998 Set dry amount. Amount of unprocessed (dry) signal.
3002 Set wet amount. Amount of processed (wet) signal.
3006 Set temperature degree in Celsius. This is the temperature of the environment.
3011 Apply headphone crossfeed filter.
3013 Crossfeed is the process of blending the left and right channels of stereo
3015 It is mainly used to reduce extreme stereo separation of low frequencies.
3017 The intent is to produce more speaker like sound to the listener.
3019 The filter accepts the following options:
3023 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
3024 This sets gain of low shelf filter for side part of stereo image.
3025 Default is -6dB. Max allowed is -30db when strength is set to 1.
3028 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
3029 This sets cut off frequency of low shelf filter. Default is cut off near
3030 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
3033 Set input gain. Default is 0.9.
3036 Set output gain. Default is 1.
3039 @section crystalizer
3040 Simple algorithm to expand audio dynamic range.
3042 The filter accepts the following options:
3046 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
3047 (unchanged sound) to 10.0 (maximum effect).
3050 Enable clipping. By default is enabled.
3054 Apply a DC shift to the audio.
3056 This can be useful to remove a DC offset (caused perhaps by a hardware problem
3057 in the recording chain) from the audio. The effect of a DC offset is reduced
3058 headroom and hence volume. The @ref{astats} filter can be used to determine if
3059 a signal has a DC offset.
3063 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
3067 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
3068 used to prevent clipping.
3073 Apply de-essing to the audio samples.
3077 Set intensity for triggering de-essing. Allowed range is from 0 to 1.
3081 Set amount of ducking on treble part of sound. Allowed range is from 0 to 1.
3085 How much of original frequency content to keep when de-essing. Allowed range is from 0 to 1.
3089 Set the output mode.
3091 It accepts the following values:
3094 Pass input unchanged.
3097 Pass ess filtered out.
3102 Default value is @var{o}.
3108 Measure audio dynamic range.
3110 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3111 is found in transition material. And anything less that 8 have very poor dynamics
3112 and is very compressed.
3114 The filter accepts the following options:
3118 Set window length in seconds used to split audio into segments of equal length.
3119 Default is 3 seconds.
3123 Dynamic Audio Normalizer.
3125 This filter applies a certain amount of gain to the input audio in order
3126 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3127 contrast to more "simple" normalization algorithms, the Dynamic Audio
3128 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3129 This allows for applying extra gain to the "quiet" sections of the audio
3130 while avoiding distortions or clipping the "loud" sections. In other words:
3131 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3132 sections, in the sense that the volume of each section is brought to the
3133 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3134 this goal *without* applying "dynamic range compressing". It will retain 100%
3135 of the dynamic range *within* each section of the audio file.
3139 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3140 Default is 500 milliseconds.
3141 The Dynamic Audio Normalizer processes the input audio in small chunks,
3142 referred to as frames. This is required, because a peak magnitude has no
3143 meaning for just a single sample value. Instead, we need to determine the
3144 peak magnitude for a contiguous sequence of sample values. While a "standard"
3145 normalizer would simply use the peak magnitude of the complete file, the
3146 Dynamic Audio Normalizer determines the peak magnitude individually for each
3147 frame. The length of a frame is specified in milliseconds. By default, the
3148 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3149 been found to give good results with most files.
3150 Note that the exact frame length, in number of samples, will be determined
3151 automatically, based on the sampling rate of the individual input audio file.
3154 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3155 number. Default is 31.
3156 Probably the most important parameter of the Dynamic Audio Normalizer is the
3157 @code{window size} of the Gaussian smoothing filter. The filter's window size
3158 is specified in frames, centered around the current frame. For the sake of
3159 simplicity, this must be an odd number. Consequently, the default value of 31
3160 takes into account the current frame, as well as the 15 preceding frames and
3161 the 15 subsequent frames. Using a larger window results in a stronger
3162 smoothing effect and thus in less gain variation, i.e. slower gain
3163 adaptation. Conversely, using a smaller window results in a weaker smoothing
3164 effect and thus in more gain variation, i.e. faster gain adaptation.
3165 In other words, the more you increase this value, the more the Dynamic Audio
3166 Normalizer will behave like a "traditional" normalization filter. On the
3167 contrary, the more you decrease this value, the more the Dynamic Audio
3168 Normalizer will behave like a dynamic range compressor.
3171 Set the target peak value. This specifies the highest permissible magnitude
3172 level for the normalized audio input. This filter will try to approach the
3173 target peak magnitude as closely as possible, but at the same time it also
3174 makes sure that the normalized signal will never exceed the peak magnitude.
3175 A frame's maximum local gain factor is imposed directly by the target peak
3176 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3177 It is not recommended to go above this value.
3180 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3181 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3182 factor for each input frame, i.e. the maximum gain factor that does not
3183 result in clipping or distortion. The maximum gain factor is determined by
3184 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3185 additionally bounds the frame's maximum gain factor by a predetermined
3186 (global) maximum gain factor. This is done in order to avoid excessive gain
3187 factors in "silent" or almost silent frames. By default, the maximum gain
3188 factor is 10.0, For most inputs the default value should be sufficient and
3189 it usually is not recommended to increase this value. Though, for input
3190 with an extremely low overall volume level, it may be necessary to allow even
3191 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3192 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3193 Instead, a "sigmoid" threshold function will be applied. This way, the
3194 gain factors will smoothly approach the threshold value, but never exceed that
3198 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3199 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3200 This means that the maximum local gain factor for each frame is defined
3201 (only) by the frame's highest magnitude sample. This way, the samples can
3202 be amplified as much as possible without exceeding the maximum signal
3203 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3204 Normalizer can also take into account the frame's root mean square,
3205 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3206 determine the power of a time-varying signal. It is therefore considered
3207 that the RMS is a better approximation of the "perceived loudness" than
3208 just looking at the signal's peak magnitude. Consequently, by adjusting all
3209 frames to a constant RMS value, a uniform "perceived loudness" can be
3210 established. If a target RMS value has been specified, a frame's local gain
3211 factor is defined as the factor that would result in exactly that RMS value.
3212 Note, however, that the maximum local gain factor is still restricted by the
3213 frame's highest magnitude sample, in order to prevent clipping.
3216 Enable channels coupling. By default is enabled.
3217 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3218 amount. This means the same gain factor will be applied to all channels, i.e.
3219 the maximum possible gain factor is determined by the "loudest" channel.
3220 However, in some recordings, it may happen that the volume of the different
3221 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3222 In this case, this option can be used to disable the channel coupling. This way,
3223 the gain factor will be determined independently for each channel, depending
3224 only on the individual channel's highest magnitude sample. This allows for
3225 harmonizing the volume of the different channels.
3228 Enable DC bias correction. By default is disabled.
3229 An audio signal (in the time domain) is a sequence of sample values.
3230 In the Dynamic Audio Normalizer these sample values are represented in the
3231 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3232 audio signal, or "waveform", should be centered around the zero point.
3233 That means if we calculate the mean value of all samples in a file, or in a
3234 single frame, then the result should be 0.0 or at least very close to that
3235 value. If, however, there is a significant deviation of the mean value from
3236 0.0, in either positive or negative direction, this is referred to as a
3237 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3238 Audio Normalizer provides optional DC bias correction.
3239 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3240 the mean value, or "DC correction" offset, of each input frame and subtract
3241 that value from all of the frame's sample values which ensures those samples
3242 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3243 boundaries, the DC correction offset values will be interpolated smoothly
3244 between neighbouring frames.
3246 @item altboundary, b
3247 Enable alternative boundary mode. By default is disabled.
3248 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3249 around each frame. This includes the preceding frames as well as the
3250 subsequent frames. However, for the "boundary" frames, located at the very
3251 beginning and at the very end of the audio file, not all neighbouring
3252 frames are available. In particular, for the first few frames in the audio
3253 file, the preceding frames are not known. And, similarly, for the last few
3254 frames in the audio file, the subsequent frames are not known. Thus, the
3255 question arises which gain factors should be assumed for the missing frames
3256 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3257 to deal with this situation. The default boundary mode assumes a gain factor
3258 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3259 "fade out" at the beginning and at the end of the input, respectively.
3262 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3263 By default, the Dynamic Audio Normalizer does not apply "traditional"
3264 compression. This means that signal peaks will not be pruned and thus the
3265 full dynamic range will be retained within each local neighbourhood. However,
3266 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3267 normalization algorithm with a more "traditional" compression.
3268 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3269 (thresholding) function. If (and only if) the compression feature is enabled,
3270 all input frames will be processed by a soft knee thresholding function prior
3271 to the actual normalization process. Put simply, the thresholding function is
3272 going to prune all samples whose magnitude exceeds a certain threshold value.
3273 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3274 value. Instead, the threshold value will be adjusted for each individual
3276 In general, smaller parameters result in stronger compression, and vice versa.
3277 Values below 3.0 are not recommended, because audible distortion may appear.
3282 Make audio easier to listen to on headphones.
3284 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3285 so that when listened to on headphones the stereo image is moved from
3286 inside your head (standard for headphones) to outside and in front of
3287 the listener (standard for speakers).
3293 Apply a two-pole peaking equalisation (EQ) filter. With this
3294 filter, the signal-level at and around a selected frequency can
3295 be increased or decreased, whilst (unlike bandpass and bandreject
3296 filters) that at all other frequencies is unchanged.
3298 In order to produce complex equalisation curves, this filter can
3299 be given several times, each with a different central frequency.
3301 The filter accepts the following options:
3305 Set the filter's central frequency in Hz.
3308 Set method to specify band-width of filter.
3323 Specify the band-width of a filter in width_type units.
3326 Set the required gain or attenuation in dB.
3327 Beware of clipping when using a positive gain.
3330 How much to use filtered signal in output. Default is 1.
3331 Range is between 0 and 1.
3334 Specify which channels to filter, by default all available are filtered.
3337 @subsection Examples
3340 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3342 equalizer=f=1000:t=h:width=200:g=-10
3346 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3348 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3352 @subsection Commands
3354 This filter supports the following commands:
3357 Change equalizer frequency.
3358 Syntax for the command is : "@var{frequency}"
3361 Change equalizer width_type.
3362 Syntax for the command is : "@var{width_type}"
3365 Change equalizer width.
3366 Syntax for the command is : "@var{width}"
3369 Change equalizer gain.
3370 Syntax for the command is : "@var{gain}"
3373 Change equalizer mix.
3374 Syntax for the command is : "@var{mix}"
3377 @section extrastereo
3379 Linearly increases the difference between left and right channels which
3380 adds some sort of "live" effect to playback.
3382 The filter accepts the following options:
3386 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3387 (average of both channels), with 1.0 sound will be unchanged, with
3388 -1.0 left and right channels will be swapped.
3391 Enable clipping. By default is enabled.
3394 @section firequalizer
3395 Apply FIR Equalization using arbitrary frequency response.
3397 The filter accepts the following option:
3401 Set gain curve equation (in dB). The expression can contain variables:
3404 the evaluated frequency
3408 channel number, set to 0 when multichannels evaluation is disabled
3410 channel id, see libavutil/channel_layout.h, set to the first channel id when
3411 multichannels evaluation is disabled
3415 channel_layout, see libavutil/channel_layout.h
3420 @item gain_interpolate(f)
3421 interpolate gain on frequency f based on gain_entry
3422 @item cubic_interpolate(f)
3423 same as gain_interpolate, but smoother
3425 This option is also available as command. Default is @code{gain_interpolate(f)}.
3428 Set gain entry for gain_interpolate function. The expression can
3432 store gain entry at frequency f with value g
3434 This option is also available as command.
3437 Set filter delay in seconds. Higher value means more accurate.
3438 Default is @code{0.01}.
3441 Set filter accuracy in Hz. Lower value means more accurate.
3442 Default is @code{5}.
3445 Set window function. Acceptable values are:
3448 rectangular window, useful when gain curve is already smooth
3450 hann window (default)
3456 3-terms continuous 1st derivative nuttall window
3458 minimum 3-terms discontinuous nuttall window
3460 4-terms continuous 1st derivative nuttall window
3462 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3464 blackman-harris window
3470 If enabled, use fixed number of audio samples. This improves speed when
3471 filtering with large delay. Default is disabled.
3474 Enable multichannels evaluation on gain. Default is disabled.
3477 Enable zero phase mode by subtracting timestamp to compensate delay.
3478 Default is disabled.
3481 Set scale used by gain. Acceptable values are:
3484 linear frequency, linear gain
3486 linear frequency, logarithmic (in dB) gain (default)
3488 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3490 logarithmic frequency, logarithmic gain
3494 Set file for dumping, suitable for gnuplot.
3497 Set scale for dumpfile. Acceptable values are same with scale option.
3501 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3502 Default is disabled.
3505 Enable minimum phase impulse response. Default is disabled.
3508 @subsection Examples
3513 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3516 lowpass at 1000 Hz with gain_entry:
3518 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3521 custom equalization:
3523 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3526 higher delay with zero phase to compensate delay:
3528 firequalizer=delay=0.1:fixed=on:zero_phase=on
3531 lowpass on left channel, highpass on right channel:
3533 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3534 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3539 Apply a flanging effect to the audio.
3541 The filter accepts the following options:
3545 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3548 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3551 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3555 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3556 Default value is 71.
3559 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3562 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3563 Default value is @var{sinusoidal}.
3566 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3567 Default value is 25.
3570 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3571 Default is @var{linear}.
3575 Apply Haas effect to audio.
3577 Note that this makes most sense to apply on mono signals.
3578 With this filter applied to mono signals it give some directionality and
3579 stretches its stereo image.
3581 The filter accepts the following options:
3585 Set input level. By default is @var{1}, or 0dB
3588 Set output level. By default is @var{1}, or 0dB.
3591 Set gain applied to side part of signal. By default is @var{1}.
3594 Set kind of middle source. Can be one of the following:
3604 Pick middle part signal of stereo image.
3607 Pick side part signal of stereo image.
3611 Change middle phase. By default is disabled.
3614 Set left channel delay. By default is @var{2.05} milliseconds.
3617 Set left channel balance. By default is @var{-1}.
3620 Set left channel gain. By default is @var{1}.
3623 Change left phase. By default is disabled.
3626 Set right channel delay. By defaults is @var{2.12} milliseconds.
3629 Set right channel balance. By default is @var{1}.
3632 Set right channel gain. By default is @var{1}.
3635 Change right phase. By default is enabled.
3640 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3641 embedded HDCD codes is expanded into a 20-bit PCM stream.
3643 The filter supports the Peak Extend and Low-level Gain Adjustment features
3644 of HDCD, and detects the Transient Filter flag.
3647 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3650 When using the filter with wav, note the default encoding for wav is 16-bit,
3651 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3652 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3654 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3655 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3658 The filter accepts the following options:
3661 @item disable_autoconvert
3662 Disable any automatic format conversion or resampling in the filter graph.
3664 @item process_stereo
3665 Process the stereo channels together. If target_gain does not match between
3666 channels, consider it invalid and use the last valid target_gain.
3669 Set the code detect timer period in ms.
3672 Always extend peaks above -3dBFS even if PE isn't signaled.
3675 Replace audio with a solid tone and adjust the amplitude to signal some
3676 specific aspect of the decoding process. The output file can be loaded in
3677 an audio editor alongside the original to aid analysis.
3679 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3686 Gain adjustment level at each sample
3688 Samples where peak extend occurs
3690 Samples where the code detect timer is active
3692 Samples where the target gain does not match between channels
3698 Apply head-related transfer functions (HRTFs) to create virtual
3699 loudspeakers around the user for binaural listening via headphones.
3700 The HRIRs are provided via additional streams, for each channel
3701 one stereo input stream is needed.
3703 The filter accepts the following options:
3707 Set mapping of input streams for convolution.
3708 The argument is a '|'-separated list of channel names in order as they
3709 are given as additional stream inputs for filter.
3710 This also specify number of input streams. Number of input streams
3711 must be not less than number of channels in first stream plus one.
3714 Set gain applied to audio. Value is in dB. Default is 0.
3717 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3718 processing audio in time domain which is slow.
3719 @var{freq} is processing audio in frequency domain which is fast.
3720 Default is @var{freq}.
3723 Set custom gain for LFE channels. Value is in dB. Default is 0.
3726 Set size of frame in number of samples which will be processed at once.
3727 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3730 Set format of hrir stream.
3731 Default value is @var{stereo}. Alternative value is @var{multich}.
3732 If value is set to @var{stereo}, number of additional streams should
3733 be greater or equal to number of input channels in first input stream.
3734 Also each additional stream should have stereo number of channels.
3735 If value is set to @var{multich}, number of additional streams should
3736 be exactly one. Also number of input channels of additional stream
3737 should be equal or greater than twice number of channels of first input
3741 @subsection Examples
3745 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3746 each amovie filter use stereo file with IR coefficients as input.
3747 The files give coefficients for each position of virtual loudspeaker:
3750 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3755 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3756 but now in @var{multich} @var{hrir} format.
3758 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3765 Apply a high-pass filter with 3dB point frequency.
3766 The filter can be either single-pole, or double-pole (the default).
3767 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3769 The filter accepts the following options:
3773 Set frequency in Hz. Default is 3000.
3776 Set number of poles. Default is 2.
3779 Set method to specify band-width of filter.
3794 Specify the band-width of a filter in width_type units.
3795 Applies only to double-pole filter.
3796 The default is 0.707q and gives a Butterworth response.
3799 How much to use filtered signal in output. Default is 1.
3800 Range is between 0 and 1.
3803 Specify which channels to filter, by default all available are filtered.
3806 @subsection Commands
3808 This filter supports the following commands:
3811 Change highpass frequency.
3812 Syntax for the command is : "@var{frequency}"
3815 Change highpass width_type.
3816 Syntax for the command is : "@var{width_type}"
3819 Change highpass width.
3820 Syntax for the command is : "@var{width}"
3823 Change highpass mix.
3824 Syntax for the command is : "@var{mix}"
3829 Join multiple input streams into one multi-channel stream.
3831 It accepts the following parameters:
3835 The number of input streams. It defaults to 2.
3837 @item channel_layout
3838 The desired output channel layout. It defaults to stereo.
3841 Map channels from inputs to output. The argument is a '|'-separated list of
3842 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3843 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3844 can be either the name of the input channel (e.g. FL for front left) or its
3845 index in the specified input stream. @var{out_channel} is the name of the output
3849 The filter will attempt to guess the mappings when they are not specified
3850 explicitly. It does so by first trying to find an unused matching input channel
3851 and if that fails it picks the first unused input channel.
3853 Join 3 inputs (with properly set channel layouts):
3855 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3858 Build a 5.1 output from 6 single-channel streams:
3860 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3861 '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'
3867 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3869 To enable compilation of this filter you need to configure FFmpeg with
3870 @code{--enable-ladspa}.
3874 Specifies the name of LADSPA plugin library to load. If the environment
3875 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3876 each one of the directories specified by the colon separated list in
3877 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3878 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3879 @file{/usr/lib/ladspa/}.
3882 Specifies the plugin within the library. Some libraries contain only
3883 one plugin, but others contain many of them. If this is not set filter
3884 will list all available plugins within the specified library.
3887 Set the '|' separated list of controls which are zero or more floating point
3888 values that determine the behavior of the loaded plugin (for example delay,
3890 Controls need to be defined using the following syntax:
3891 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3892 @var{valuei} is the value set on the @var{i}-th control.
3893 Alternatively they can be also defined using the following syntax:
3894 @var{value0}|@var{value1}|@var{value2}|..., where
3895 @var{valuei} is the value set on the @var{i}-th control.
3896 If @option{controls} is set to @code{help}, all available controls and
3897 their valid ranges are printed.
3899 @item sample_rate, s
3900 Specify the sample rate, default to 44100. Only used if plugin have
3904 Set the number of samples per channel per each output frame, default
3905 is 1024. Only used if plugin have zero inputs.
3908 Set the minimum duration of the sourced audio. See
3909 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3910 for the accepted syntax.
3911 Note that the resulting duration may be greater than the specified duration,
3912 as the generated audio is always cut at the end of a complete frame.
3913 If not specified, or the expressed duration is negative, the audio is
3914 supposed to be generated forever.
3915 Only used if plugin have zero inputs.
3919 @subsection Examples
3923 List all available plugins within amp (LADSPA example plugin) library:
3929 List all available controls and their valid ranges for @code{vcf_notch}
3930 plugin from @code{VCF} library:
3932 ladspa=f=vcf:p=vcf_notch:c=help
3936 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3939 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3943 Add reverberation to the audio using TAP-plugins
3944 (Tom's Audio Processing plugins):
3946 ladspa=file=tap_reverb:tap_reverb
3950 Generate white noise, with 0.2 amplitude:
3952 ladspa=file=cmt:noise_source_white:c=c0=.2
3956 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3957 @code{C* Audio Plugin Suite} (CAPS) library:
3959 ladspa=file=caps:Click:c=c1=20'
3963 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3965 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3969 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3970 @code{SWH Plugins} collection:
3972 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3976 Attenuate low frequencies using Multiband EQ from Steve Harris
3977 @code{SWH Plugins} collection:
3979 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3983 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3986 ladspa=caps:Narrower
3990 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3992 ladspa=caps:White:.2
3996 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3998 ladspa=caps:Fractal:c=c1=1
4002 Dynamic volume normalization using @code{VLevel} plugin:
4004 ladspa=vlevel-ladspa:vlevel_mono
4008 @subsection Commands
4010 This filter supports the following commands:
4013 Modify the @var{N}-th control value.
4015 If the specified value is not valid, it is ignored and prior one is kept.
4020 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
4021 Support for both single pass (livestreams, files) and double pass (files) modes.
4022 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
4023 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
4024 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
4026 The filter accepts the following options:
4030 Set integrated loudness target.
4031 Range is -70.0 - -5.0. Default value is -24.0.
4034 Set loudness range target.
4035 Range is 1.0 - 20.0. Default value is 7.0.
4038 Set maximum true peak.
4039 Range is -9.0 - +0.0. Default value is -2.0.
4041 @item measured_I, measured_i
4042 Measured IL of input file.
4043 Range is -99.0 - +0.0.
4045 @item measured_LRA, measured_lra
4046 Measured LRA of input file.
4047 Range is 0.0 - 99.0.
4049 @item measured_TP, measured_tp
4050 Measured true peak of input file.
4051 Range is -99.0 - +99.0.
4053 @item measured_thresh
4054 Measured threshold of input file.
4055 Range is -99.0 - +0.0.
4058 Set offset gain. Gain is applied before the true-peak limiter.
4059 Range is -99.0 - +99.0. Default is +0.0.
4062 Normalize linearly if possible.
4063 measured_I, measured_LRA, measured_TP, and measured_thresh must also
4064 to be specified in order to use this mode.
4065 Options are true or false. Default is true.
4068 Treat mono input files as "dual-mono". If a mono file is intended for playback
4069 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
4070 If set to @code{true}, this option will compensate for this effect.
4071 Multi-channel input files are not affected by this option.
4072 Options are true or false. Default is false.
4075 Set print format for stats. Options are summary, json, or none.
4076 Default value is none.
4081 Apply a low-pass filter with 3dB point frequency.
4082 The filter can be either single-pole or double-pole (the default).
4083 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
4085 The filter accepts the following options:
4089 Set frequency in Hz. Default is 500.
4092 Set number of poles. Default is 2.
4095 Set method to specify band-width of filter.
4110 Specify the band-width of a filter in width_type units.
4111 Applies only to double-pole filter.
4112 The default is 0.707q and gives a Butterworth response.
4115 How much to use filtered signal in output. Default is 1.
4116 Range is between 0 and 1.
4119 Specify which channels to filter, by default all available are filtered.
4122 @subsection Examples
4125 Lowpass only LFE channel, it LFE is not present it does nothing:
4131 @subsection Commands
4133 This filter supports the following commands:
4136 Change lowpass frequency.
4137 Syntax for the command is : "@var{frequency}"
4140 Change lowpass width_type.
4141 Syntax for the command is : "@var{width_type}"
4144 Change lowpass width.
4145 Syntax for the command is : "@var{width}"
4149 Syntax for the command is : "@var{mix}"
4154 Load a LV2 (LADSPA Version 2) plugin.
4156 To enable compilation of this filter you need to configure FFmpeg with
4157 @code{--enable-lv2}.
4161 Specifies the plugin URI. You may need to escape ':'.
4164 Set the '|' separated list of controls which are zero or more floating point
4165 values that determine the behavior of the loaded plugin (for example delay,
4167 If @option{controls} is set to @code{help}, all available controls and
4168 their valid ranges are printed.
4170 @item sample_rate, s
4171 Specify the sample rate, default to 44100. Only used if plugin have
4175 Set the number of samples per channel per each output frame, default
4176 is 1024. Only used if plugin have zero inputs.
4179 Set the minimum duration of the sourced audio. See
4180 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4181 for the accepted syntax.
4182 Note that the resulting duration may be greater than the specified duration,
4183 as the generated audio is always cut at the end of a complete frame.
4184 If not specified, or the expressed duration is negative, the audio is
4185 supposed to be generated forever.
4186 Only used if plugin have zero inputs.
4189 @subsection Examples
4193 Apply bass enhancer plugin from Calf:
4195 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4199 Apply vinyl plugin from Calf:
4201 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4205 Apply bit crusher plugin from ArtyFX:
4207 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4212 Multiband Compress or expand the audio's dynamic range.
4214 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4215 This is akin to the crossover of a loudspeaker, and results in flat frequency
4216 response when absent compander action.
4218 It accepts the following parameters:
4222 This option syntax is:
4223 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4224 For explanation of each item refer to compand filter documentation.
4230 Mix channels with specific gain levels. The filter accepts the output
4231 channel layout followed by a set of channels definitions.
4233 This filter is also designed to efficiently remap the channels of an audio
4236 The filter accepts parameters of the form:
4237 "@var{l}|@var{outdef}|@var{outdef}|..."
4241 output channel layout or number of channels
4244 output channel specification, of the form:
4245 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4248 output channel to define, either a channel name (FL, FR, etc.) or a channel
4249 number (c0, c1, etc.)
4252 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4255 input channel to use, see out_name for details; it is not possible to mix
4256 named and numbered input channels
4259 If the `=' in a channel specification is replaced by `<', then the gains for
4260 that specification will be renormalized so that the total is 1, thus
4261 avoiding clipping noise.
4263 @subsection Mixing examples
4265 For example, if you want to down-mix from stereo to mono, but with a bigger
4266 factor for the left channel:
4268 pan=1c|c0=0.9*c0+0.1*c1
4271 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4272 7-channels surround:
4274 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4277 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4278 that should be preferred (see "-ac" option) unless you have very specific
4281 @subsection Remapping examples
4283 The channel remapping will be effective if, and only if:
4286 @item gain coefficients are zeroes or ones,
4287 @item only one input per channel output,
4290 If all these conditions are satisfied, the filter will notify the user ("Pure
4291 channel mapping detected"), and use an optimized and lossless method to do the
4294 For example, if you have a 5.1 source and want a stereo audio stream by
4295 dropping the extra channels:
4297 pan="stereo| c0=FL | c1=FR"
4300 Given the same source, you can also switch front left and front right channels
4301 and keep the input channel layout:
4303 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4306 If the input is a stereo audio stream, you can mute the front left channel (and
4307 still keep the stereo channel layout) with:
4312 Still with a stereo audio stream input, you can copy the right channel in both
4313 front left and right:
4315 pan="stereo| c0=FR | c1=FR"
4320 ReplayGain scanner filter. This filter takes an audio stream as an input and
4321 outputs it unchanged.
4322 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4326 Convert the audio sample format, sample rate and channel layout. It is
4327 not meant to be used directly.
4330 Apply time-stretching and pitch-shifting with librubberband.
4332 To enable compilation of this filter, you need to configure FFmpeg with
4333 @code{--enable-librubberband}.
4335 The filter accepts the following options:
4339 Set tempo scale factor.
4342 Set pitch scale factor.
4345 Set transients detector.
4346 Possible values are:
4355 Possible values are:
4364 Possible values are:
4371 Set processing window size.
4372 Possible values are:
4381 Possible values are:
4388 Enable formant preservation when shift pitching.
4389 Possible values are:
4397 Possible values are:
4406 Possible values are:
4413 @section sidechaincompress
4415 This filter acts like normal compressor but has the ability to compress
4416 detected signal using second input signal.
4417 It needs two input streams and returns one output stream.
4418 First input stream will be processed depending on second stream signal.
4419 The filtered signal then can be filtered with other filters in later stages of
4420 processing. See @ref{pan} and @ref{amerge} filter.
4422 The filter accepts the following options:
4426 Set input gain. Default is 1. Range is between 0.015625 and 64.
4429 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
4430 Default is @code{downward}.
4433 If a signal of second stream raises above this level it will affect the gain
4434 reduction of first stream.
4435 By default is 0.125. Range is between 0.00097563 and 1.
4438 Set a ratio about which the signal is reduced. 1:2 means that if the level
4439 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4440 Default is 2. Range is between 1 and 20.
4443 Amount of milliseconds the signal has to rise above the threshold before gain
4444 reduction starts. Default is 20. Range is between 0.01 and 2000.
4447 Amount of milliseconds the signal has to fall below the threshold before
4448 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4451 Set the amount by how much signal will be amplified after processing.
4452 Default is 1. Range is from 1 to 64.
4455 Curve the sharp knee around the threshold to enter gain reduction more softly.
4456 Default is 2.82843. Range is between 1 and 8.
4459 Choose if the @code{average} level between all channels of side-chain stream
4460 or the louder(@code{maximum}) channel of side-chain stream affects the
4461 reduction. Default is @code{average}.
4464 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4465 of @code{rms}. Default is @code{rms} which is mainly smoother.
4468 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4471 How much to use compressed signal in output. Default is 1.
4472 Range is between 0 and 1.
4475 @subsection Examples
4479 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4480 depending on the signal of 2nd input and later compressed signal to be
4481 merged with 2nd input:
4483 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4487 @section sidechaingate
4489 A sidechain gate acts like a normal (wideband) gate but has the ability to
4490 filter the detected signal before sending it to the gain reduction stage.
4491 Normally a gate uses the full range signal to detect a level above the
4493 For example: If you cut all lower frequencies from your sidechain signal
4494 the gate will decrease the volume of your track only if not enough highs
4495 appear. With this technique you are able to reduce the resonation of a
4496 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4498 It needs two input streams and returns one output stream.
4499 First input stream will be processed depending on second stream signal.
4501 The filter accepts the following options:
4505 Set input level before filtering.
4506 Default is 1. Allowed range is from 0.015625 to 64.
4509 Set the mode of operation. Can be @code{upward} or @code{downward}.
4510 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
4511 will be amplified, expanding dynamic range in upward direction.
4512 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
4515 Set the level of gain reduction when the signal is below the threshold.
4516 Default is 0.06125. Allowed range is from 0 to 1.
4517 Setting this to 0 disables reduction and then filter behaves like expander.
4520 If a signal rises above this level the gain reduction is released.
4521 Default is 0.125. Allowed range is from 0 to 1.
4524 Set a ratio about which the signal is reduced.
4525 Default is 2. Allowed range is from 1 to 9000.
4528 Amount of milliseconds the signal has to rise above the threshold before gain
4530 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4533 Amount of milliseconds the signal has to fall below the threshold before the
4534 reduction is increased again. Default is 250 milliseconds.
4535 Allowed range is from 0.01 to 9000.
4538 Set amount of amplification of signal after processing.
4539 Default is 1. Allowed range is from 1 to 64.
4542 Curve the sharp knee around the threshold to enter gain reduction more softly.
4543 Default is 2.828427125. Allowed range is from 1 to 8.
4546 Choose if exact signal should be taken for detection or an RMS like one.
4547 Default is rms. Can be peak or rms.
4550 Choose if the average level between all channels or the louder channel affects
4552 Default is average. Can be average or maximum.
4555 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4558 @section silencedetect
4560 Detect silence in an audio stream.
4562 This filter logs a message when it detects that the input audio volume is less
4563 or equal to a noise tolerance value for a duration greater or equal to the
4564 minimum detected noise duration.
4566 The printed times and duration are expressed in seconds.
4568 The filter accepts the following options:
4572 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4573 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4576 Set silence duration until notification (default is 2 seconds).
4579 Process each channel separately, instead of combined. By default is disabled.
4582 @subsection Examples
4586 Detect 5 seconds of silence with -50dB noise tolerance:
4588 silencedetect=n=-50dB:d=5
4592 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4593 tolerance in @file{silence.mp3}:
4595 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4599 @section silenceremove
4601 Remove silence from the beginning, middle or end of the audio.
4603 The filter accepts the following options:
4607 This value is used to indicate if audio should be trimmed at beginning of
4608 the audio. A value of zero indicates no silence should be trimmed from the
4609 beginning. When specifying a non-zero value, it trims audio up until it
4610 finds non-silence. Normally, when trimming silence from beginning of audio
4611 the @var{start_periods} will be @code{1} but it can be increased to higher
4612 values to trim all audio up to specific count of non-silence periods.
4613 Default value is @code{0}.
4615 @item start_duration
4616 Specify the amount of time that non-silence must be detected before it stops
4617 trimming audio. By increasing the duration, bursts of noises can be treated
4618 as silence and trimmed off. Default value is @code{0}.
4620 @item start_threshold
4621 This indicates what sample value should be treated as silence. For digital
4622 audio, a value of @code{0} may be fine but for audio recorded from analog,
4623 you may wish to increase the value to account for background noise.
4624 Can be specified in dB (in case "dB" is appended to the specified value)
4625 or amplitude ratio. Default value is @code{0}.
4628 Specify max duration of silence at beginning that will be kept after
4629 trimming. Default is 0, which is equal to trimming all samples detected
4633 Specify mode of detection of silence end in start of multi-channel audio.
4634 Can be @var{any} or @var{all}. Default is @var{any}.
4635 With @var{any}, any sample that is detected as non-silence will cause
4636 stopped trimming of silence.
4637 With @var{all}, only if all channels are detected as non-silence will cause
4638 stopped trimming of silence.
4641 Set the count for trimming silence from the end of audio.
4642 To remove silence from the middle of a file, specify a @var{stop_periods}
4643 that is negative. This value is then treated as a positive value and is
4644 used to indicate the effect should restart processing as specified by
4645 @var{start_periods}, making it suitable for removing periods of silence
4646 in the middle of the audio.
4647 Default value is @code{0}.
4650 Specify a duration of silence that must exist before audio is not copied any
4651 more. By specifying a higher duration, silence that is wanted can be left in
4653 Default value is @code{0}.
4655 @item stop_threshold
4656 This is the same as @option{start_threshold} but for trimming silence from
4658 Can be specified in dB (in case "dB" is appended to the specified value)
4659 or amplitude ratio. Default value is @code{0}.
4662 Specify max duration of silence at end that will be kept after
4663 trimming. Default is 0, which is equal to trimming all samples detected
4667 Specify mode of detection of silence start in end of multi-channel audio.
4668 Can be @var{any} or @var{all}. Default is @var{any}.
4669 With @var{any}, any sample that is detected as non-silence will cause
4670 stopped trimming of silence.
4671 With @var{all}, only if all channels are detected as non-silence will cause
4672 stopped trimming of silence.
4675 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4676 and works better with digital silence which is exactly 0.
4677 Default value is @code{rms}.
4680 Set duration in number of seconds used to calculate size of window in number
4681 of samples for detecting silence.
4682 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4685 @subsection Examples
4689 The following example shows how this filter can be used to start a recording
4690 that does not contain the delay at the start which usually occurs between
4691 pressing the record button and the start of the performance:
4693 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4697 Trim all silence encountered from beginning to end where there is more than 1
4698 second of silence in audio:
4700 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4706 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4707 loudspeakers around the user for binaural listening via headphones (audio
4708 formats up to 9 channels supported).
4709 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4710 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4711 Austrian Academy of Sciences.
4713 To enable compilation of this filter you need to configure FFmpeg with
4714 @code{--enable-libmysofa}.
4716 The filter accepts the following options:
4720 Set the SOFA file used for rendering.
4723 Set gain applied to audio. Value is in dB. Default is 0.
4726 Set rotation of virtual loudspeakers in deg. Default is 0.
4729 Set elevation of virtual speakers in deg. Default is 0.
4732 Set distance in meters between loudspeakers and the listener with near-field
4733 HRTFs. Default is 1.
4736 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4737 processing audio in time domain which is slow.
4738 @var{freq} is processing audio in frequency domain which is fast.
4739 Default is @var{freq}.
4742 Set custom positions of virtual loudspeakers. Syntax for this option is:
4743 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4744 Each virtual loudspeaker is described with short channel name following with
4745 azimuth and elevation in degrees.
4746 Each virtual loudspeaker description is separated by '|'.
4747 For example to override front left and front right channel positions use:
4748 'speakers=FL 45 15|FR 345 15'.
4749 Descriptions with unrecognised channel names are ignored.
4752 Set custom gain for LFE channels. Value is in dB. Default is 0.
4755 Set custom frame size in number of samples. Default is 1024.
4756 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4757 is set to @var{freq}.
4760 Should all IRs be normalized upon importing SOFA file.
4761 By default is enabled.
4764 Should nearest IRs be interpolated with neighbor IRs if exact position
4765 does not match. By default is disabled.
4768 Minphase all IRs upon loading of SOFA file. By default is disabled.
4771 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4774 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4777 @subsection Examples
4781 Using ClubFritz6 sofa file:
4783 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4787 Using ClubFritz12 sofa file and bigger radius with small rotation:
4789 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4793 Similar as above but with custom speaker positions for front left, front right, back left and back right
4794 and also with custom gain:
4796 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4800 @section stereotools
4802 This filter has some handy utilities to manage stereo signals, for converting
4803 M/S stereo recordings to L/R signal while having control over the parameters
4804 or spreading the stereo image of master track.
4806 The filter accepts the following options:
4810 Set input level before filtering for both channels. Defaults is 1.
4811 Allowed range is from 0.015625 to 64.
4814 Set output level after filtering for both channels. Defaults is 1.
4815 Allowed range is from 0.015625 to 64.
4818 Set input balance between both channels. Default is 0.
4819 Allowed range is from -1 to 1.
4822 Set output balance between both channels. Default is 0.
4823 Allowed range is from -1 to 1.
4826 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4827 clipping. Disabled by default.
4830 Mute the left channel. Disabled by default.
4833 Mute the right channel. Disabled by default.
4836 Change the phase of the left channel. Disabled by default.
4839 Change the phase of the right channel. Disabled by default.
4842 Set stereo mode. Available values are:
4846 Left/Right to Left/Right, this is default.
4849 Left/Right to Mid/Side.
4852 Mid/Side to Left/Right.
4855 Left/Right to Left/Left.
4858 Left/Right to Right/Right.
4861 Left/Right to Left + Right.
4864 Left/Right to Right/Left.
4867 Mid/Side to Left/Left.
4870 Mid/Side to Right/Right.
4874 Set level of side signal. Default is 1.
4875 Allowed range is from 0.015625 to 64.
4878 Set balance of side signal. Default is 0.
4879 Allowed range is from -1 to 1.
4882 Set level of the middle signal. Default is 1.
4883 Allowed range is from 0.015625 to 64.
4886 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4889 Set stereo base between mono and inversed channels. Default is 0.
4890 Allowed range is from -1 to 1.
4893 Set delay in milliseconds how much to delay left from right channel and
4894 vice versa. Default is 0. Allowed range is from -20 to 20.
4897 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4900 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4902 @item bmode_in, bmode_out
4903 Set balance mode for balance_in/balance_out option.
4905 Can be one of the following:
4909 Classic balance mode. Attenuate one channel at time.
4910 Gain is raised up to 1.
4913 Similar as classic mode above but gain is raised up to 2.
4916 Equal power distribution, from -6dB to +6dB range.
4920 @subsection Examples
4924 Apply karaoke like effect:
4926 stereotools=mlev=0.015625
4930 Convert M/S signal to L/R:
4932 "stereotools=mode=ms>lr"
4936 @section stereowiden
4938 This filter enhance the stereo effect by suppressing signal common to both
4939 channels and by delaying the signal of left into right and vice versa,
4940 thereby widening the stereo effect.
4942 The filter accepts the following options:
4946 Time in milliseconds of the delay of left signal into right and vice versa.
4947 Default is 20 milliseconds.
4950 Amount of gain in delayed signal into right and vice versa. Gives a delay
4951 effect of left signal in right output and vice versa which gives widening
4952 effect. Default is 0.3.
4955 Cross feed of left into right with inverted phase. This helps in suppressing
4956 the mono. If the value is 1 it will cancel all the signal common to both
4957 channels. Default is 0.3.
4960 Set level of input signal of original channel. Default is 0.8.
4963 @section superequalizer
4964 Apply 18 band equalizer.
4966 The filter accepts the following options:
4973 Set 131Hz band gain.
4975 Set 185Hz band gain.
4977 Set 262Hz band gain.
4979 Set 370Hz band gain.
4981 Set 523Hz band gain.
4983 Set 740Hz band gain.
4985 Set 1047Hz band gain.
4987 Set 1480Hz band gain.
4989 Set 2093Hz band gain.
4991 Set 2960Hz band gain.
4993 Set 4186Hz band gain.
4995 Set 5920Hz band gain.
4997 Set 8372Hz band gain.
4999 Set 11840Hz band gain.
5001 Set 16744Hz band gain.
5003 Set 20000Hz band gain.
5007 Apply audio surround upmix filter.
5009 This filter allows to produce multichannel output from audio stream.
5011 The filter accepts the following options:
5015 Set output channel layout. By default, this is @var{5.1}.
5017 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5018 for the required syntax.
5021 Set input channel layout. By default, this is @var{stereo}.
5023 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5024 for the required syntax.
5027 Set input volume level. By default, this is @var{1}.
5030 Set output volume level. By default, this is @var{1}.
5033 Enable LFE channel output if output channel layout has it. By default, this is enabled.
5036 Set LFE low cut off frequency. By default, this is @var{128} Hz.
5039 Set LFE high cut off frequency. By default, this is @var{256} Hz.
5042 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
5043 In @var{add} mode, LFE channel is created from input audio and added to output.
5044 In @var{sub} mode, LFE channel is created from input audio and added to output but
5045 also all non-LFE output channels are subtracted with output LFE channel.
5048 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
5049 Default is @var{90}.
5052 Set front center input volume. By default, this is @var{1}.
5055 Set front center output volume. By default, this is @var{1}.
5058 Set front left input volume. By default, this is @var{1}.
5061 Set front left output volume. By default, this is @var{1}.
5064 Set front right input volume. By default, this is @var{1}.
5067 Set front right output volume. By default, this is @var{1}.
5070 Set side left input volume. By default, this is @var{1}.
5073 Set side left output volume. By default, this is @var{1}.
5076 Set side right input volume. By default, this is @var{1}.
5079 Set side right output volume. By default, this is @var{1}.
5082 Set back left input volume. By default, this is @var{1}.
5085 Set back left output volume. By default, this is @var{1}.
5088 Set back right input volume. By default, this is @var{1}.
5091 Set back right output volume. By default, this is @var{1}.
5094 Set back center input volume. By default, this is @var{1}.
5097 Set back center output volume. By default, this is @var{1}.
5100 Set LFE input volume. By default, this is @var{1}.
5103 Set LFE output volume. By default, this is @var{1}.
5106 Set spread usage of stereo image across X axis for all channels.
5109 Set spread usage of stereo image across Y axis for all channels.
5111 @item fcx, flx, frx, blx, brx, slx, srx, bcx
5112 Set spread usage of stereo image across X axis for each channel.
5114 @item fcy, fly, fry, bly, bry, sly, sry, bcy
5115 Set spread usage of stereo image across Y axis for each channel.
5118 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
5121 Set window function.
5123 It accepts the following values:
5146 Default is @code{hann}.
5149 Set window overlap. If set to 1, the recommended overlap for selected
5150 window function will be picked. Default is @code{0.5}.
5153 @section treble, highshelf
5155 Boost or cut treble (upper) frequencies of the audio using a two-pole
5156 shelving filter with a response similar to that of a standard
5157 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
5159 The filter accepts the following options:
5163 Give the gain at whichever is the lower of ~22 kHz and the
5164 Nyquist frequency. Its useful range is about -20 (for a large cut)
5165 to +20 (for a large boost). Beware of clipping when using a positive gain.
5168 Set the filter's central frequency and so can be used
5169 to extend or reduce the frequency range to be boosted or cut.
5170 The default value is @code{3000} Hz.
5173 Set method to specify band-width of filter.
5188 Determine how steep is the filter's shelf transition.
5191 How much to use filtered signal in output. Default is 1.
5192 Range is between 0 and 1.
5195 Specify which channels to filter, by default all available are filtered.
5198 @subsection Commands
5200 This filter supports the following commands:
5203 Change treble frequency.
5204 Syntax for the command is : "@var{frequency}"
5207 Change treble width_type.
5208 Syntax for the command is : "@var{width_type}"
5211 Change treble width.
5212 Syntax for the command is : "@var{width}"
5216 Syntax for the command is : "@var{gain}"
5220 Syntax for the command is : "@var{mix}"
5225 Sinusoidal amplitude modulation.
5227 The filter accepts the following options:
5231 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
5232 (20 Hz or lower) will result in a tremolo effect.
5233 This filter may also be used as a ring modulator by specifying
5234 a modulation frequency higher than 20 Hz.
5235 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5238 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5239 Default value is 0.5.
5244 Sinusoidal phase modulation.
5246 The filter accepts the following options:
5250 Modulation frequency in Hertz.
5251 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5254 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5255 Default value is 0.5.
5260 Adjust the input audio volume.
5262 It accepts the following parameters:
5266 Set audio volume expression.
5268 Output values are clipped to the maximum value.
5270 The output audio volume is given by the relation:
5272 @var{output_volume} = @var{volume} * @var{input_volume}
5275 The default value for @var{volume} is "1.0".
5278 This parameter represents the mathematical precision.
5280 It determines which input sample formats will be allowed, which affects the
5281 precision of the volume scaling.
5285 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
5287 32-bit floating-point; this limits input sample format to FLT. (default)
5289 64-bit floating-point; this limits input sample format to DBL.
5293 Choose the behaviour on encountering ReplayGain side data in input frames.
5297 Remove ReplayGain side data, ignoring its contents (the default).
5300 Ignore ReplayGain side data, but leave it in the frame.
5303 Prefer the track gain, if present.
5306 Prefer the album gain, if present.
5309 @item replaygain_preamp
5310 Pre-amplification gain in dB to apply to the selected replaygain gain.
5312 Default value for @var{replaygain_preamp} is 0.0.
5315 Set when the volume expression is evaluated.
5317 It accepts the following values:
5320 only evaluate expression once during the filter initialization, or
5321 when the @samp{volume} command is sent
5324 evaluate expression for each incoming frame
5327 Default value is @samp{once}.
5330 The volume expression can contain the following parameters.
5334 frame number (starting at zero)
5337 @item nb_consumed_samples
5338 number of samples consumed by the filter
5340 number of samples in the current frame
5342 original frame position in the file
5348 PTS at start of stream
5350 time at start of stream
5356 last set volume value
5359 Note that when @option{eval} is set to @samp{once} only the
5360 @var{sample_rate} and @var{tb} variables are available, all other
5361 variables will evaluate to NAN.
5363 @subsection Commands
5365 This filter supports the following commands:
5368 Modify the volume expression.
5369 The command accepts the same syntax of the corresponding option.
5371 If the specified expression is not valid, it is kept at its current
5373 @item replaygain_noclip
5374 Prevent clipping by limiting the gain applied.
5376 Default value for @var{replaygain_noclip} is 1.
5380 @subsection Examples
5384 Halve the input audio volume:
5388 volume=volume=-6.0206dB
5391 In all the above example the named key for @option{volume} can be
5392 omitted, for example like in:
5398 Increase input audio power by 6 decibels using fixed-point precision:
5400 volume=volume=6dB:precision=fixed
5404 Fade volume after time 10 with an annihilation period of 5 seconds:
5406 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5410 @section volumedetect
5412 Detect the volume of the input video.
5414 The filter has no parameters. The input is not modified. Statistics about
5415 the volume will be printed in the log when the input stream end is reached.
5417 In particular it will show the mean volume (root mean square), maximum
5418 volume (on a per-sample basis), and the beginning of a histogram of the
5419 registered volume values (from the maximum value to a cumulated 1/1000 of
5422 All volumes are in decibels relative to the maximum PCM value.
5424 @subsection Examples
5426 Here is an excerpt of the output:
5428 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5429 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5430 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5431 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5432 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5433 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5434 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5435 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5436 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5442 The mean square energy is approximately -27 dB, or 10^-2.7.
5444 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5446 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5449 In other words, raising the volume by +4 dB does not cause any clipping,
5450 raising it by +5 dB causes clipping for 6 samples, etc.
5452 @c man end AUDIO FILTERS
5454 @chapter Audio Sources
5455 @c man begin AUDIO SOURCES
5457 Below is a description of the currently available audio sources.
5461 Buffer audio frames, and make them available to the filter chain.
5463 This source is mainly intended for a programmatic use, in particular
5464 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5466 It accepts the following parameters:
5470 The timebase which will be used for timestamps of submitted frames. It must be
5471 either a floating-point number or in @var{numerator}/@var{denominator} form.
5474 The sample rate of the incoming audio buffers.
5477 The sample format of the incoming audio buffers.
5478 Either a sample format name or its corresponding integer representation from
5479 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5481 @item channel_layout
5482 The channel layout of the incoming audio buffers.
5483 Either a channel layout name from channel_layout_map in
5484 @file{libavutil/channel_layout.c} or its corresponding integer representation
5485 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5488 The number of channels of the incoming audio buffers.
5489 If both @var{channels} and @var{channel_layout} are specified, then they
5494 @subsection Examples
5497 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5500 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5501 Since the sample format with name "s16p" corresponds to the number
5502 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5505 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5510 Generate an audio signal specified by an expression.
5512 This source accepts in input one or more expressions (one for each
5513 channel), which are evaluated and used to generate a corresponding
5516 This source accepts the following options:
5520 Set the '|'-separated expressions list for each separate channel. In case the
5521 @option{channel_layout} option is not specified, the selected channel layout
5522 depends on the number of provided expressions. Otherwise the last
5523 specified expression is applied to the remaining output channels.
5525 @item channel_layout, c
5526 Set the channel layout. The number of channels in the specified layout
5527 must be equal to the number of specified expressions.
5530 Set the minimum duration of the sourced audio. See
5531 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5532 for the accepted syntax.
5533 Note that the resulting duration may be greater than the specified
5534 duration, as the generated audio is always cut at the end of a
5537 If not specified, or the expressed duration is negative, the audio is
5538 supposed to be generated forever.
5541 Set the number of samples per channel per each output frame,
5544 @item sample_rate, s
5545 Specify the sample rate, default to 44100.
5548 Each expression in @var{exprs} can contain the following constants:
5552 number of the evaluated sample, starting from 0
5555 time of the evaluated sample expressed in seconds, starting from 0
5562 @subsection Examples
5572 Generate a sin signal with frequency of 440 Hz, set sample rate to
5575 aevalsrc="sin(440*2*PI*t):s=8000"
5579 Generate a two channels signal, specify the channel layout (Front
5580 Center + Back Center) explicitly:
5582 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5586 Generate white noise:
5588 aevalsrc="-2+random(0)"
5592 Generate an amplitude modulated signal:
5594 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5598 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5600 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5607 The null audio source, return unprocessed audio frames. It is mainly useful
5608 as a template and to be employed in analysis / debugging tools, or as
5609 the source for filters which ignore the input data (for example the sox
5612 This source accepts the following options:
5616 @item channel_layout, cl
5618 Specifies the channel layout, and can be either an integer or a string
5619 representing a channel layout. The default value of @var{channel_layout}
5622 Check the channel_layout_map definition in
5623 @file{libavutil/channel_layout.c} for the mapping between strings and
5624 channel layout values.
5626 @item sample_rate, r
5627 Specifies the sample rate, and defaults to 44100.
5630 Set the number of samples per requested frames.
5634 @subsection Examples
5638 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5640 anullsrc=r=48000:cl=4
5644 Do the same operation with a more obvious syntax:
5646 anullsrc=r=48000:cl=mono
5650 All the parameters need to be explicitly defined.
5654 Synthesize a voice utterance using the libflite library.
5656 To enable compilation of this filter you need to configure FFmpeg with
5657 @code{--enable-libflite}.
5659 Note that versions of the flite library prior to 2.0 are not thread-safe.
5661 The filter accepts the following options:
5666 If set to 1, list the names of the available voices and exit
5667 immediately. Default value is 0.
5670 Set the maximum number of samples per frame. Default value is 512.
5673 Set the filename containing the text to speak.
5676 Set the text to speak.
5679 Set the voice to use for the speech synthesis. Default value is
5680 @code{kal}. See also the @var{list_voices} option.
5683 @subsection Examples
5687 Read from file @file{speech.txt}, and synthesize the text using the
5688 standard flite voice:
5690 flite=textfile=speech.txt
5694 Read the specified text selecting the @code{slt} voice:
5696 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5700 Input text to ffmpeg:
5702 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5706 Make @file{ffplay} speak the specified text, using @code{flite} and
5707 the @code{lavfi} device:
5709 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5713 For more information about libflite, check:
5714 @url{http://www.festvox.org/flite/}
5718 Generate a noise audio signal.
5720 The filter accepts the following options:
5723 @item sample_rate, r
5724 Specify the sample rate. Default value is 48000 Hz.
5727 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5731 Specify the duration of the generated audio stream. Not specifying this option
5732 results in noise with an infinite length.
5734 @item color, colour, c
5735 Specify the color of noise. Available noise colors are white, pink, brown,
5736 blue and violet. Default color is white.
5739 Specify a value used to seed the PRNG.
5742 Set the number of samples per each output frame, default is 1024.
5745 @subsection Examples
5750 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5752 anoisesrc=d=60:c=pink:r=44100:a=0.5
5758 Generate odd-tap Hilbert transform FIR coefficients.
5760 The resulting stream can be used with @ref{afir} filter for phase-shifting
5761 the signal by 90 degrees.
5763 This is used in many matrix coding schemes and for analytic signal generation.
5764 The process is often written as a multiplication by i (or j), the imaginary unit.
5766 The filter accepts the following options:
5770 @item sample_rate, s
5771 Set sample rate, default is 44100.
5774 Set length of FIR filter, default is 22051.
5777 Set number of samples per each frame.
5780 Set window function to be used when generating FIR coefficients.
5785 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5787 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5789 The filter accepts the following options:
5792 @item sample_rate, r
5793 Set sample rate, default is 44100.
5796 Set number of samples per each frame. Default is 1024.
5799 Set high-pass frequency. Default is 0.
5802 Set low-pass frequency. Default is 0.
5803 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5804 is higher than 0 then filter will create band-pass filter coefficients,
5805 otherwise band-reject filter coefficients.
5808 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5811 Set Kaiser window beta.
5814 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5817 Enable rounding, by default is disabled.
5820 Set number of taps for high-pass filter.
5823 Set number of taps for low-pass filter.
5828 Generate an audio signal made of a sine wave with amplitude 1/8.
5830 The audio signal is bit-exact.
5832 The filter accepts the following options:
5837 Set the carrier frequency. Default is 440 Hz.
5839 @item beep_factor, b
5840 Enable a periodic beep every second with frequency @var{beep_factor} times
5841 the carrier frequency. Default is 0, meaning the beep is disabled.
5843 @item sample_rate, r
5844 Specify the sample rate, default is 44100.
5847 Specify the duration of the generated audio stream.
5849 @item samples_per_frame
5850 Set the number of samples per output frame.
5852 The expression can contain the following constants:
5856 The (sequential) number of the output audio frame, starting from 0.
5859 The PTS (Presentation TimeStamp) of the output audio frame,
5860 expressed in @var{TB} units.
5863 The PTS of the output audio frame, expressed in seconds.
5866 The timebase of the output audio frames.
5869 Default is @code{1024}.
5872 @subsection Examples
5877 Generate a simple 440 Hz sine wave:
5883 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5887 sine=frequency=220:beep_factor=4:duration=5
5891 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5894 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5898 @c man end AUDIO SOURCES
5900 @chapter Audio Sinks
5901 @c man begin AUDIO SINKS
5903 Below is a description of the currently available audio sinks.
5905 @section abuffersink
5907 Buffer audio frames, and make them available to the end of filter chain.
5909 This sink is mainly intended for programmatic use, in particular
5910 through the interface defined in @file{libavfilter/buffersink.h}
5911 or the options system.
5913 It accepts a pointer to an AVABufferSinkContext structure, which
5914 defines the incoming buffers' formats, to be passed as the opaque
5915 parameter to @code{avfilter_init_filter} for initialization.
5918 Null audio sink; do absolutely nothing with the input audio. It is
5919 mainly useful as a template and for use in analysis / debugging
5922 @c man end AUDIO SINKS
5924 @chapter Video Filters
5925 @c man begin VIDEO FILTERS
5927 When you configure your FFmpeg build, you can disable any of the
5928 existing filters using @code{--disable-filters}.
5929 The configure output will show the video filters included in your
5932 Below is a description of the currently available video filters.
5936 Mark a region of interest in a video frame.
5938 The frame data is passed through unchanged, but metadata is attached
5939 to the frame indicating regions of interest which can affect the
5940 behaviour of later encoding. Multiple regions can be marked by
5941 applying the filter multiple times.
5945 Region distance in pixels from the left edge of the frame.
5947 Region distance in pixels from the top edge of the frame.
5949 Region width in pixels.
5951 Region height in pixels.
5953 The parameters @var{x}, @var{y}, @var{w} and @var{h} are expressions,
5954 and may contain the following variables:
5957 Width of the input frame.
5959 Height of the input frame.
5963 Quantisation offset to apply within the region.
5965 This must be a real value in the range -1 to +1. A value of zero
5966 indicates no quality change. A negative value asks for better quality
5967 (less quantisation), while a positive value asks for worse quality
5968 (greater quantisation).
5970 The range is calibrated so that the extreme values indicate the
5971 largest possible offset - if the rest of the frame is encoded with the
5972 worst possible quality, an offset of -1 indicates that this region
5973 should be encoded with the best possible quality anyway. Intermediate
5974 values are then interpolated in some codec-dependent way.
5976 For example, in 10-bit H.264 the quantisation parameter varies between
5977 -12 and 51. A typical qoffset value of -1/10 therefore indicates that
5978 this region should be encoded with a QP around one-tenth of the full
5979 range better than the rest of the frame. So, if most of the frame
5980 were to be encoded with a QP of around 30, this region would get a QP
5981 of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3).
5982 An extreme value of -1 would indicate that this region should be
5983 encoded with the best possible quality regardless of the treatment of
5984 the rest of the frame - that is, should be encoded at a QP of -12.
5986 If set to true, remove any existing regions of interest marked on the
5987 frame before adding the new one.
5990 @subsection Examples
5994 Mark the centre quarter of the frame as interesting.
5996 addroi=iw/4:ih/4:iw/2:ih/2:-1/10
5999 Mark the 100-pixel-wide region on the left edge of the frame as very
6000 uninteresting (to be encoded at much lower quality than the rest of
6003 addroi=0:0:100:ih:+1/5
6007 @section alphaextract
6009 Extract the alpha component from the input as a grayscale video. This
6010 is especially useful with the @var{alphamerge} filter.
6014 Add or replace the alpha component of the primary input with the
6015 grayscale value of a second input. This is intended for use with
6016 @var{alphaextract} to allow the transmission or storage of frame
6017 sequences that have alpha in a format that doesn't support an alpha
6020 For example, to reconstruct full frames from a normal YUV-encoded video
6021 and a separate video created with @var{alphaextract}, you might use:
6023 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
6026 Since this filter is designed for reconstruction, it operates on frame
6027 sequences without considering timestamps, and terminates when either
6028 input reaches end of stream. This will cause problems if your encoding
6029 pipeline drops frames. If you're trying to apply an image as an
6030 overlay to a video stream, consider the @var{overlay} filter instead.
6034 Amplify differences between current pixel and pixels of adjacent frames in
6035 same pixel location.
6037 This filter accepts the following options:
6041 Set frame radius. Default is 2. Allowed range is from 1 to 63.
6042 For example radius of 3 will instruct filter to calculate average of 7 frames.
6045 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
6048 Set threshold for difference amplification. Any difference greater or equal to
6049 this value will not alter source pixel. Default is 10.
6050 Allowed range is from 0 to 65535.
6053 Set tolerance for difference amplification. Any difference lower to
6054 this value will not alter source pixel. Default is 0.
6055 Allowed range is from 0 to 65535.
6058 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
6059 This option controls maximum possible value that will decrease source pixel value.
6062 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
6063 This option controls maximum possible value that will increase source pixel value.
6066 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
6071 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
6072 and libavformat to work. On the other hand, it is limited to ASS (Advanced
6073 Substation Alpha) subtitles files.
6075 This filter accepts the following option in addition to the common options from
6076 the @ref{subtitles} filter:
6080 Set the shaping engine
6082 Available values are:
6085 The default libass shaping engine, which is the best available.
6087 Fast, font-agnostic shaper that can do only substitutions
6089 Slower shaper using OpenType for substitutions and positioning
6092 The default is @code{auto}.
6096 Apply an Adaptive Temporal Averaging Denoiser to the video input.
6098 The filter accepts the following options:
6102 Set threshold A for 1st plane. Default is 0.02.
6103 Valid range is 0 to 0.3.
6106 Set threshold B for 1st plane. Default is 0.04.
6107 Valid range is 0 to 5.
6110 Set threshold A for 2nd plane. Default is 0.02.
6111 Valid range is 0 to 0.3.
6114 Set threshold B for 2nd plane. Default is 0.04.
6115 Valid range is 0 to 5.
6118 Set threshold A for 3rd plane. Default is 0.02.
6119 Valid range is 0 to 0.3.
6122 Set threshold B for 3rd plane. Default is 0.04.
6123 Valid range is 0 to 5.
6125 Threshold A is designed to react on abrupt changes in the input signal and
6126 threshold B is designed to react on continuous changes in the input signal.
6129 Set number of frames filter will use for averaging. Default is 9. Must be odd
6130 number in range [5, 129].
6133 Set what planes of frame filter will use for averaging. Default is all.
6138 Apply average blur filter.
6140 The filter accepts the following options:
6144 Set horizontal radius size.
6147 Set which planes to filter. By default all planes are filtered.
6150 Set vertical radius size, if zero it will be same as @code{sizeX}.
6151 Default is @code{0}.
6156 Compute the bounding box for the non-black pixels in the input frame
6159 This filter computes the bounding box containing all the pixels with a
6160 luminance value greater than the minimum allowed value.
6161 The parameters describing the bounding box are printed on the filter
6164 The filter accepts the following option:
6168 Set the minimal luminance value. Default is @code{16}.
6171 @section bitplanenoise
6173 Show and measure bit plane noise.
6175 The filter accepts the following options:
6179 Set which plane to analyze. Default is @code{1}.
6182 Filter out noisy pixels from @code{bitplane} set above.
6183 Default is disabled.
6186 @section blackdetect
6188 Detect video intervals that are (almost) completely black. Can be
6189 useful to detect chapter transitions, commercials, or invalid
6190 recordings. Output lines contains the time for the start, end and
6191 duration of the detected black interval expressed in seconds.
6193 In order to display the output lines, you need to set the loglevel at
6194 least to the AV_LOG_INFO value.
6196 The filter accepts the following options:
6199 @item black_min_duration, d
6200 Set the minimum detected black duration expressed in seconds. It must
6201 be a non-negative floating point number.
6203 Default value is 2.0.
6205 @item picture_black_ratio_th, pic_th
6206 Set the threshold for considering a picture "black".
6207 Express the minimum value for the ratio:
6209 @var{nb_black_pixels} / @var{nb_pixels}
6212 for which a picture is considered black.
6213 Default value is 0.98.
6215 @item pixel_black_th, pix_th
6216 Set the threshold for considering a pixel "black".
6218 The threshold expresses the maximum pixel luminance value for which a
6219 pixel is considered "black". The provided value is scaled according to
6220 the following equation:
6222 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
6225 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
6226 the input video format, the range is [0-255] for YUV full-range
6227 formats and [16-235] for YUV non full-range formats.
6229 Default value is 0.10.
6232 The following example sets the maximum pixel threshold to the minimum
6233 value, and detects only black intervals of 2 or more seconds:
6235 blackdetect=d=2:pix_th=0.00
6240 Detect frames that are (almost) completely black. Can be useful to
6241 detect chapter transitions or commercials. Output lines consist of
6242 the frame number of the detected frame, the percentage of blackness,
6243 the position in the file if known or -1 and the timestamp in seconds.
6245 In order to display the output lines, you need to set the loglevel at
6246 least to the AV_LOG_INFO value.
6248 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
6249 The value represents the percentage of pixels in the picture that
6250 are below the threshold value.
6252 It accepts the following parameters:
6257 The percentage of the pixels that have to be below the threshold; it defaults to
6260 @item threshold, thresh
6261 The threshold below which a pixel value is considered black; it defaults to
6266 @section blend, tblend
6268 Blend two video frames into each other.
6270 The @code{blend} filter takes two input streams and outputs one
6271 stream, the first input is the "top" layer and second input is
6272 "bottom" layer. By default, the output terminates when the longest input terminates.
6274 The @code{tblend} (time blend) filter takes two consecutive frames
6275 from one single stream, and outputs the result obtained by blending
6276 the new frame on top of the old frame.
6278 A description of the accepted options follows.
6286 Set blend mode for specific pixel component or all pixel components in case
6287 of @var{all_mode}. Default value is @code{normal}.
6289 Available values for component modes are:
6331 Set blend opacity for specific pixel component or all pixel components in case
6332 of @var{all_opacity}. Only used in combination with pixel component blend modes.
6339 Set blend expression for specific pixel component or all pixel components in case
6340 of @var{all_expr}. Note that related mode options will be ignored if those are set.
6342 The expressions can use the following variables:
6346 The sequential number of the filtered frame, starting from @code{0}.
6350 the coordinates of the current sample
6354 the width and height of currently filtered plane
6358 Width and height scale for the plane being filtered. It is the
6359 ratio between the dimensions of the current plane to the luma plane,
6360 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
6361 the luma plane and @code{0.5,0.5} for the chroma planes.
6364 Time of the current frame, expressed in seconds.
6367 Value of pixel component at current location for first video frame (top layer).
6370 Value of pixel component at current location for second video frame (bottom layer).
6374 The @code{blend} filter also supports the @ref{framesync} options.
6376 @subsection Examples
6380 Apply transition from bottom layer to top layer in first 10 seconds:
6382 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6386 Apply linear horizontal transition from top layer to bottom layer:
6388 blend=all_expr='A*(X/W)+B*(1-X/W)'
6392 Apply 1x1 checkerboard effect:
6394 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6398 Apply uncover left effect:
6400 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6404 Apply uncover down effect:
6406 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6410 Apply uncover up-left effect:
6412 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6416 Split diagonally video and shows top and bottom layer on each side:
6418 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6422 Display differences between the current and the previous frame:
6424 tblend=all_mode=grainextract
6430 Denoise frames using Block-Matching 3D algorithm.
6432 The filter accepts the following options.
6436 Set denoising strength. Default value is 1.
6437 Allowed range is from 0 to 999.9.
6438 The denoising algorithm is very sensitive to sigma, so adjust it
6439 according to the source.
6442 Set local patch size. This sets dimensions in 2D.
6445 Set sliding step for processing blocks. Default value is 4.
6446 Allowed range is from 1 to 64.
6447 Smaller values allows processing more reference blocks and is slower.
6450 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6451 When set to 1, no block matching is done. Larger values allows more blocks
6453 Allowed range is from 1 to 256.
6456 Set radius for search block matching. Default is 9.
6457 Allowed range is from 1 to INT32_MAX.
6460 Set step between two search locations for block matching. Default is 1.
6461 Allowed range is from 1 to 64. Smaller is slower.
6464 Set threshold of mean square error for block matching. Valid range is 0 to
6468 Set thresholding parameter for hard thresholding in 3D transformed domain.
6469 Larger values results in stronger hard-thresholding filtering in frequency
6473 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6474 Default is @code{basic}.
6477 If enabled, filter will use 2nd stream for block matching.
6478 Default is disabled for @code{basic} value of @var{estim} option,
6479 and always enabled if value of @var{estim} is @code{final}.
6482 Set planes to filter. Default is all available except alpha.
6485 @subsection Examples
6489 Basic filtering with bm3d:
6491 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6495 Same as above, but filtering only luma:
6497 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6501 Same as above, but with both estimation modes:
6503 split[a][b],[a]bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
6507 Same as above, but prefilter with @ref{nlmeans} filter instead:
6509 split[a][b],[a]nlmeans=s=3:r=7:p=3[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
6515 Apply a boxblur algorithm to the input video.
6517 It accepts the following parameters:
6521 @item luma_radius, lr
6522 @item luma_power, lp
6523 @item chroma_radius, cr
6524 @item chroma_power, cp
6525 @item alpha_radius, ar
6526 @item alpha_power, ap
6530 A description of the accepted options follows.
6533 @item luma_radius, lr
6534 @item chroma_radius, cr
6535 @item alpha_radius, ar
6536 Set an expression for the box radius in pixels used for blurring the
6537 corresponding input plane.
6539 The radius value must be a non-negative number, and must not be
6540 greater than the value of the expression @code{min(w,h)/2} for the
6541 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6544 Default value for @option{luma_radius} is "2". If not specified,
6545 @option{chroma_radius} and @option{alpha_radius} default to the
6546 corresponding value set for @option{luma_radius}.
6548 The expressions can contain the following constants:
6552 The input width and height in pixels.
6556 The input chroma image width and height in pixels.
6560 The horizontal and vertical chroma subsample values. For example, for the
6561 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6564 @item luma_power, lp
6565 @item chroma_power, cp
6566 @item alpha_power, ap
6567 Specify how many times the boxblur filter is applied to the
6568 corresponding plane.
6570 Default value for @option{luma_power} is 2. If not specified,
6571 @option{chroma_power} and @option{alpha_power} default to the
6572 corresponding value set for @option{luma_power}.
6574 A value of 0 will disable the effect.
6577 @subsection Examples
6581 Apply a boxblur filter with the luma, chroma, and alpha radii
6584 boxblur=luma_radius=2:luma_power=1
6589 Set the luma radius to 2, and alpha and chroma radius to 0:
6591 boxblur=2:1:cr=0:ar=0
6595 Set the luma and chroma radii to a fraction of the video dimension:
6597 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6603 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6604 Deinterlacing Filter").
6606 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6607 interpolation algorithms.
6608 It accepts the following parameters:
6612 The interlacing mode to adopt. It accepts one of the following values:
6616 Output one frame for each frame.
6618 Output one frame for each field.
6621 The default value is @code{send_field}.
6624 The picture field parity assumed for the input interlaced video. It accepts one
6625 of the following values:
6629 Assume the top field is first.
6631 Assume the bottom field is first.
6633 Enable automatic detection of field parity.
6636 The default value is @code{auto}.
6637 If the interlacing is unknown or the decoder does not export this information,
6638 top field first will be assumed.
6641 Specify which frames to deinterlace. Accept one of the following
6646 Deinterlace all frames.
6648 Only deinterlace frames marked as interlaced.
6651 The default value is @code{all}.
6655 Remove all color information for all colors except for certain one.
6657 The filter accepts the following options:
6661 The color which will not be replaced with neutral chroma.
6664 Similarity percentage with the above color.
6665 0.01 matches only the exact key color, while 1.0 matches everything.
6669 0.0 makes pixels either fully gray, or not gray at all.
6670 Higher values result in more preserved color.
6673 Signals that the color passed is already in YUV instead of RGB.
6675 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6676 This can be used to pass exact YUV values as hexadecimal numbers.
6680 YUV colorspace color/chroma keying.
6682 The filter accepts the following options:
6686 The color which will be replaced with transparency.
6689 Similarity percentage with the key color.
6691 0.01 matches only the exact key color, while 1.0 matches everything.
6696 0.0 makes pixels either fully transparent, or not transparent at all.
6698 Higher values result in semi-transparent pixels, with a higher transparency
6699 the more similar the pixels color is to the key color.
6702 Signals that the color passed is already in YUV instead of RGB.
6704 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6705 This can be used to pass exact YUV values as hexadecimal numbers.
6708 @subsection Examples
6712 Make every green pixel in the input image transparent:
6714 ffmpeg -i input.png -vf chromakey=green out.png
6718 Overlay a greenscreen-video on top of a static black background.
6720 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
6724 @section chromashift
6725 Shift chroma pixels horizontally and/or vertically.
6727 The filter accepts the following options:
6730 Set amount to shift chroma-blue horizontally.
6732 Set amount to shift chroma-blue vertically.
6734 Set amount to shift chroma-red horizontally.
6736 Set amount to shift chroma-red vertically.
6738 Set edge mode, can be @var{smear}, default, or @var{warp}.
6743 Display CIE color diagram with pixels overlaid onto it.
6745 The filter accepts the following options:
6760 @item uhdtv, rec2020
6774 Set what gamuts to draw.
6776 See @code{system} option for available values.
6779 Set ciescope size, by default set to 512.
6782 Set intensity used to map input pixel values to CIE diagram.
6785 Set contrast used to draw tongue colors that are out of active color system gamut.
6788 Correct gamma displayed on scope, by default enabled.
6791 Show white point on CIE diagram, by default disabled.
6794 Set input gamma. Used only with XYZ input color space.
6799 Visualize information exported by some codecs.
6801 Some codecs can export information through frames using side-data or other
6802 means. For example, some MPEG based codecs export motion vectors through the
6803 @var{export_mvs} flag in the codec @option{flags2} option.
6805 The filter accepts the following option:
6809 Set motion vectors to visualize.
6811 Available flags for @var{mv} are:
6815 forward predicted MVs of P-frames
6817 forward predicted MVs of B-frames
6819 backward predicted MVs of B-frames
6823 Display quantization parameters using the chroma planes.
6826 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6828 Available flags for @var{mv_type} are:
6832 forward predicted MVs
6834 backward predicted MVs
6837 @item frame_type, ft
6838 Set frame type to visualize motion vectors of.
6840 Available flags for @var{frame_type} are:
6844 intra-coded frames (I-frames)
6846 predicted frames (P-frames)
6848 bi-directionally predicted frames (B-frames)
6852 @subsection Examples
6856 Visualize forward predicted MVs of all frames using @command{ffplay}:
6858 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6862 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6864 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6868 @section colorbalance
6869 Modify intensity of primary colors (red, green and blue) of input frames.
6871 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6872 regions for the red-cyan, green-magenta or blue-yellow balance.
6874 A positive adjustment value shifts the balance towards the primary color, a negative
6875 value towards the complementary color.
6877 The filter accepts the following options:
6883 Adjust red, green and blue shadows (darkest pixels).
6888 Adjust red, green and blue midtones (medium pixels).
6893 Adjust red, green and blue highlights (brightest pixels).
6895 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6898 @subsection Examples
6902 Add red color cast to shadows:
6909 RGB colorspace color keying.
6911 The filter accepts the following options:
6915 The color which will be replaced with transparency.
6918 Similarity percentage with the key color.
6920 0.01 matches only the exact key color, while 1.0 matches everything.
6925 0.0 makes pixels either fully transparent, or not transparent at all.
6927 Higher values result in semi-transparent pixels, with a higher transparency
6928 the more similar the pixels color is to the key color.
6931 @subsection Examples
6935 Make every green pixel in the input image transparent:
6937 ffmpeg -i input.png -vf colorkey=green out.png
6941 Overlay a greenscreen-video on top of a static background image.
6943 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
6948 Remove all color information for all RGB colors except for certain one.
6950 The filter accepts the following options:
6954 The color which will not be replaced with neutral gray.
6957 Similarity percentage with the above color.
6958 0.01 matches only the exact key color, while 1.0 matches everything.
6961 Blend percentage. 0.0 makes pixels fully gray.
6962 Higher values result in more preserved color.
6965 @section colorlevels
6967 Adjust video input frames using levels.
6969 The filter accepts the following options:
6976 Adjust red, green, blue and alpha input black point.
6977 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6983 Adjust red, green, blue and alpha input white point.
6984 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6986 Input levels are used to lighten highlights (bright tones), darken shadows
6987 (dark tones), change the balance of bright and dark tones.
6993 Adjust red, green, blue and alpha output black point.
6994 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
7000 Adjust red, green, blue and alpha output white point.
7001 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
7003 Output levels allows manual selection of a constrained output level range.
7006 @subsection Examples
7010 Make video output darker:
7012 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
7018 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
7022 Make video output lighter:
7024 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
7028 Increase brightness:
7030 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
7034 @section colorchannelmixer
7036 Adjust video input frames by re-mixing color channels.
7038 This filter modifies a color channel by adding the values associated to
7039 the other channels of the same pixels. For example if the value to
7040 modify is red, the output value will be:
7042 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
7045 The filter accepts the following options:
7052 Adjust contribution of input red, green, blue and alpha channels for output red channel.
7053 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
7059 Adjust contribution of input red, green, blue and alpha channels for output green channel.
7060 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
7066 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
7067 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
7073 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
7074 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
7076 Allowed ranges for options are @code{[-2.0, 2.0]}.
7079 @subsection Examples
7083 Convert source to grayscale:
7085 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
7088 Simulate sepia tones:
7090 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
7094 @section colormatrix
7096 Convert color matrix.
7098 The filter accepts the following options:
7103 Specify the source and destination color matrix. Both values must be
7106 The accepted values are:
7134 For example to convert from BT.601 to SMPTE-240M, use the command:
7136 colormatrix=bt601:smpte240m
7141 Convert colorspace, transfer characteristics or color primaries.
7142 Input video needs to have an even size.
7144 The filter accepts the following options:
7149 Specify all color properties at once.
7151 The accepted values are:
7181 Specify output colorspace.
7183 The accepted values are:
7192 BT.470BG or BT.601-6 625
7195 SMPTE-170M or BT.601-6 525
7204 BT.2020 with non-constant luminance
7210 Specify output transfer characteristics.
7212 The accepted values are:
7224 Constant gamma of 2.2
7227 Constant gamma of 2.8
7230 SMPTE-170M, BT.601-6 625 or BT.601-6 525
7248 BT.2020 for 10-bits content
7251 BT.2020 for 12-bits content
7257 Specify output color primaries.
7259 The accepted values are:
7268 BT.470BG or BT.601-6 625
7271 SMPTE-170M or BT.601-6 525
7295 Specify output color range.
7297 The accepted values are:
7300 TV (restricted) range
7303 MPEG (restricted) range
7314 Specify output color format.
7316 The accepted values are:
7319 YUV 4:2:0 planar 8-bits
7322 YUV 4:2:0 planar 10-bits
7325 YUV 4:2:0 planar 12-bits
7328 YUV 4:2:2 planar 8-bits
7331 YUV 4:2:2 planar 10-bits
7334 YUV 4:2:2 planar 12-bits
7337 YUV 4:4:4 planar 8-bits
7340 YUV 4:4:4 planar 10-bits
7343 YUV 4:4:4 planar 12-bits
7348 Do a fast conversion, which skips gamma/primary correction. This will take
7349 significantly less CPU, but will be mathematically incorrect. To get output
7350 compatible with that produced by the colormatrix filter, use fast=1.
7353 Specify dithering mode.
7355 The accepted values are:
7361 Floyd-Steinberg dithering
7365 Whitepoint adaptation mode.
7367 The accepted values are:
7370 Bradford whitepoint adaptation
7373 von Kries whitepoint adaptation
7376 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7380 Override all input properties at once. Same accepted values as @ref{all}.
7383 Override input colorspace. Same accepted values as @ref{space}.
7386 Override input color primaries. Same accepted values as @ref{primaries}.
7389 Override input transfer characteristics. Same accepted values as @ref{trc}.
7392 Override input color range. Same accepted values as @ref{range}.
7396 The filter converts the transfer characteristics, color space and color
7397 primaries to the specified user values. The output value, if not specified,
7398 is set to a default value based on the "all" property. If that property is
7399 also not specified, the filter will log an error. The output color range and
7400 format default to the same value as the input color range and format. The
7401 input transfer characteristics, color space, color primaries and color range
7402 should be set on the input data. If any of these are missing, the filter will
7403 log an error and no conversion will take place.
7405 For example to convert the input to SMPTE-240M, use the command:
7407 colorspace=smpte240m
7410 @section convolution
7412 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7414 The filter accepts the following options:
7421 Set matrix for each plane.
7422 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7423 and from 1 to 49 odd number of signed integers in @var{row} mode.
7429 Set multiplier for calculated value for each plane.
7430 If unset or 0, it will be sum of all matrix elements.
7436 Set bias for each plane. This value is added to the result of the multiplication.
7437 Useful for making the overall image brighter or darker. Default is 0.0.
7443 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7444 Default is @var{square}.
7447 @subsection Examples
7453 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"
7459 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"
7465 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"
7471 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"
7475 Apply laplacian edge detector which includes diagonals:
7477 convolution="1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0"
7483 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"
7489 Apply 2D convolution of video stream in frequency domain using second stream
7492 The filter accepts the following options:
7496 Set which planes to process.
7499 Set which impulse video frames will be processed, can be @var{first}
7500 or @var{all}. Default is @var{all}.
7503 The @code{convolve} filter also supports the @ref{framesync} options.
7507 Copy the input video source unchanged to the output. This is mainly useful for
7512 Video filtering on GPU using Apple's CoreImage API on OSX.
7514 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7515 processed by video hardware. However, software-based OpenGL implementations
7516 exist which means there is no guarantee for hardware processing. It depends on
7519 There are many filters and image generators provided by Apple that come with a
7520 large variety of options. The filter has to be referenced by its name along
7523 The coreimage filter accepts the following options:
7526 List all available filters and generators along with all their respective
7527 options as well as possible minimum and maximum values along with the default
7534 Specify all filters by their respective name and options.
7535 Use @var{list_filters} to determine all valid filter names and options.
7536 Numerical options are specified by a float value and are automatically clamped
7537 to their respective value range. Vector and color options have to be specified
7538 by a list of space separated float values. Character escaping has to be done.
7539 A special option name @code{default} is available to use default options for a
7542 It is required to specify either @code{default} or at least one of the filter options.
7543 All omitted options are used with their default values.
7544 The syntax of the filter string is as follows:
7546 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7550 Specify a rectangle where the output of the filter chain is copied into the
7551 input image. It is given by a list of space separated float values:
7553 output_rect=x\ y\ width\ height
7555 If not given, the output rectangle equals the dimensions of the input image.
7556 The output rectangle is automatically cropped at the borders of the input
7557 image. Negative values are valid for each component.
7559 output_rect=25\ 25\ 100\ 100
7563 Several filters can be chained for successive processing without GPU-HOST
7564 transfers allowing for fast processing of complex filter chains.
7565 Currently, only filters with zero (generators) or exactly one (filters) input
7566 image and one output image are supported. Also, transition filters are not yet
7569 Some filters generate output images with additional padding depending on the
7570 respective filter kernel. The padding is automatically removed to ensure the
7571 filter output has the same size as the input image.
7573 For image generators, the size of the output image is determined by the
7574 previous output image of the filter chain or the input image of the whole
7575 filterchain, respectively. The generators do not use the pixel information of
7576 this image to generate their output. However, the generated output is
7577 blended onto this image, resulting in partial or complete coverage of the
7580 The @ref{coreimagesrc} video source can be used for generating input images
7581 which are directly fed into the filter chain. By using it, providing input
7582 images by another video source or an input video is not required.
7584 @subsection Examples
7589 List all filters available:
7591 coreimage=list_filters=true
7595 Use the CIBoxBlur filter with default options to blur an image:
7597 coreimage=filter=CIBoxBlur@@default
7601 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7602 its center at 100x100 and a radius of 50 pixels:
7604 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7608 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7609 given as complete and escaped command-line for Apple's standard bash shell:
7611 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7617 Crop the input video to given dimensions.
7619 It accepts the following parameters:
7623 The width of the output video. It defaults to @code{iw}.
7624 This expression is evaluated only once during the filter
7625 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7628 The height of the output video. It defaults to @code{ih}.
7629 This expression is evaluated only once during the filter
7630 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7633 The horizontal position, in the input video, of the left edge of the output
7634 video. It defaults to @code{(in_w-out_w)/2}.
7635 This expression is evaluated per-frame.
7638 The vertical position, in the input video, of the top edge of the output video.
7639 It defaults to @code{(in_h-out_h)/2}.
7640 This expression is evaluated per-frame.
7643 If set to 1 will force the output display aspect ratio
7644 to be the same of the input, by changing the output sample aspect
7645 ratio. It defaults to 0.
7648 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7649 width/height/x/y as specified and will not be rounded to nearest smaller value.
7653 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7654 expressions containing the following constants:
7659 The computed values for @var{x} and @var{y}. They are evaluated for
7664 The input width and height.
7668 These are the same as @var{in_w} and @var{in_h}.
7672 The output (cropped) width and height.
7676 These are the same as @var{out_w} and @var{out_h}.
7679 same as @var{iw} / @var{ih}
7682 input sample aspect ratio
7685 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7689 horizontal and vertical chroma subsample values. For example for the
7690 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7693 The number of the input frame, starting from 0.
7696 the position in the file of the input frame, NAN if unknown
7699 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7703 The expression for @var{out_w} may depend on the value of @var{out_h},
7704 and the expression for @var{out_h} may depend on @var{out_w}, but they
7705 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7706 evaluated after @var{out_w} and @var{out_h}.
7708 The @var{x} and @var{y} parameters specify the expressions for the
7709 position of the top-left corner of the output (non-cropped) area. They
7710 are evaluated for each frame. If the evaluated value is not valid, it
7711 is approximated to the nearest valid value.
7713 The expression for @var{x} may depend on @var{y}, and the expression
7714 for @var{y} may depend on @var{x}.
7716 @subsection Examples
7720 Crop area with size 100x100 at position (12,34).
7725 Using named options, the example above becomes:
7727 crop=w=100:h=100:x=12:y=34
7731 Crop the central input area with size 100x100:
7737 Crop the central input area with size 2/3 of the input video:
7739 crop=2/3*in_w:2/3*in_h
7743 Crop the input video central square:
7750 Delimit the rectangle with the top-left corner placed at position
7751 100:100 and the right-bottom corner corresponding to the right-bottom
7752 corner of the input image.
7754 crop=in_w-100:in_h-100:100:100
7758 Crop 10 pixels from the left and right borders, and 20 pixels from
7759 the top and bottom borders
7761 crop=in_w-2*10:in_h-2*20
7765 Keep only the bottom right quarter of the input image:
7767 crop=in_w/2:in_h/2:in_w/2:in_h/2
7771 Crop height for getting Greek harmony:
7773 crop=in_w:1/PHI*in_w
7777 Apply trembling effect:
7779 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)
7783 Apply erratic camera effect depending on timestamp:
7785 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)"
7789 Set x depending on the value of y:
7791 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7795 @subsection Commands
7797 This filter supports the following commands:
7803 Set width/height of the output video and the horizontal/vertical position
7805 The command accepts the same syntax of the corresponding option.
7807 If the specified expression is not valid, it is kept at its current
7813 Auto-detect the crop size.
7815 It calculates the necessary cropping parameters and prints the
7816 recommended parameters via the logging system. The detected dimensions
7817 correspond to the non-black area of the input video.
7819 It accepts the following parameters:
7824 Set higher black value threshold, which can be optionally specified
7825 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7826 value greater to the set value is considered non-black. It defaults to 24.
7827 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7828 on the bitdepth of the pixel format.
7831 The value which the width/height should be divisible by. It defaults to
7832 16. The offset is automatically adjusted to center the video. Use 2 to
7833 get only even dimensions (needed for 4:2:2 video). 16 is best when
7834 encoding to most video codecs.
7836 @item reset_count, reset
7837 Set the counter that determines after how many frames cropdetect will
7838 reset the previously detected largest video area and start over to
7839 detect the current optimal crop area. Default value is 0.
7841 This can be useful when channel logos distort the video area. 0
7842 indicates 'never reset', and returns the largest area encountered during
7849 Delay video filtering until a given wallclock timestamp. The filter first
7850 passes on @option{preroll} amount of frames, then it buffers at most
7851 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7852 it forwards the buffered frames and also any subsequent frames coming in its
7855 The filter can be used synchronize the output of multiple ffmpeg processes for
7856 realtime output devices like decklink. By putting the delay in the filtering
7857 chain and pre-buffering frames the process can pass on data to output almost
7858 immediately after the target wallclock timestamp is reached.
7860 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7866 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7869 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7872 The maximum duration of content to buffer before waiting for the cue expressed
7873 in seconds. Default is 0.
7880 Apply color adjustments using curves.
7882 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7883 component (red, green and blue) has its values defined by @var{N} key points
7884 tied from each other using a smooth curve. The x-axis represents the pixel
7885 values from the input frame, and the y-axis the new pixel values to be set for
7888 By default, a component curve is defined by the two points @var{(0;0)} and
7889 @var{(1;1)}. This creates a straight line where each original pixel value is
7890 "adjusted" to its own value, which means no change to the image.
7892 The filter allows you to redefine these two points and add some more. A new
7893 curve (using a natural cubic spline interpolation) will be define to pass
7894 smoothly through all these new coordinates. The new defined points needs to be
7895 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7896 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7897 the vector spaces, the values will be clipped accordingly.
7899 The filter accepts the following options:
7903 Select one of the available color presets. This option can be used in addition
7904 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7905 options takes priority on the preset values.
7906 Available presets are:
7909 @item color_negative
7912 @item increase_contrast
7914 @item linear_contrast
7915 @item medium_contrast
7917 @item strong_contrast
7920 Default is @code{none}.
7922 Set the master key points. These points will define a second pass mapping. It
7923 is sometimes called a "luminance" or "value" mapping. It can be used with
7924 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7925 post-processing LUT.
7927 Set the key points for the red component.
7929 Set the key points for the green component.
7931 Set the key points for the blue component.
7933 Set the key points for all components (not including master).
7934 Can be used in addition to the other key points component
7935 options. In this case, the unset component(s) will fallback on this
7936 @option{all} setting.
7938 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7940 Save Gnuplot script of the curves in specified file.
7943 To avoid some filtergraph syntax conflicts, each key points list need to be
7944 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7946 @subsection Examples
7950 Increase slightly the middle level of blue:
7952 curves=blue='0/0 0.5/0.58 1/1'
7958 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'
7960 Here we obtain the following coordinates for each components:
7963 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7965 @code{(0;0) (0.50;0.48) (1;1)}
7967 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7971 The previous example can also be achieved with the associated built-in preset:
7973 curves=preset=vintage
7983 Use a Photoshop preset and redefine the points of the green component:
7985 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7989 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7990 and @command{gnuplot}:
7992 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7993 gnuplot -p /tmp/curves.plt
7999 Video data analysis filter.
8001 This filter shows hexadecimal pixel values of part of video.
8003 The filter accepts the following options:
8007 Set output video size.
8010 Set x offset from where to pick pixels.
8013 Set y offset from where to pick pixels.
8016 Set scope mode, can be one of the following:
8019 Draw hexadecimal pixel values with white color on black background.
8022 Draw hexadecimal pixel values with input video pixel color on black
8026 Draw hexadecimal pixel values on color background picked from input video,
8027 the text color is picked in such way so its always visible.
8031 Draw rows and columns numbers on left and top of video.
8034 Set background opacity.
8039 Denoise frames using 2D DCT (frequency domain filtering).
8041 This filter is not designed for real time.
8043 The filter accepts the following options:
8047 Set the noise sigma constant.
8049 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
8050 coefficient (absolute value) below this threshold with be dropped.
8052 If you need a more advanced filtering, see @option{expr}.
8054 Default is @code{0}.
8057 Set number overlapping pixels for each block. Since the filter can be slow, you
8058 may want to reduce this value, at the cost of a less effective filter and the
8059 risk of various artefacts.
8061 If the overlapping value doesn't permit processing the whole input width or
8062 height, a warning will be displayed and according borders won't be denoised.
8064 Default value is @var{blocksize}-1, which is the best possible setting.
8067 Set the coefficient factor expression.
8069 For each coefficient of a DCT block, this expression will be evaluated as a
8070 multiplier value for the coefficient.
8072 If this is option is set, the @option{sigma} option will be ignored.
8074 The absolute value of the coefficient can be accessed through the @var{c}
8078 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
8079 @var{blocksize}, which is the width and height of the processed blocks.
8081 The default value is @var{3} (8x8) and can be raised to @var{4} for a
8082 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
8083 on the speed processing. Also, a larger block size does not necessarily means a
8087 @subsection Examples
8089 Apply a denoise with a @option{sigma} of @code{4.5}:
8094 The same operation can be achieved using the expression system:
8096 dctdnoiz=e='gte(c, 4.5*3)'
8099 Violent denoise using a block size of @code{16x16}:
8106 Remove banding artifacts from input video.
8107 It works by replacing banded pixels with average value of referenced pixels.
8109 The filter accepts the following options:
8116 Set banding detection threshold for each plane. Default is 0.02.
8117 Valid range is 0.00003 to 0.5.
8118 If difference between current pixel and reference pixel is less than threshold,
8119 it will be considered as banded.
8122 Banding detection range in pixels. Default is 16. If positive, random number
8123 in range 0 to set value will be used. If negative, exact absolute value
8125 The range defines square of four pixels around current pixel.
8128 Set direction in radians from which four pixel will be compared. If positive,
8129 random direction from 0 to set direction will be picked. If negative, exact of
8130 absolute value will be picked. For example direction 0, -PI or -2*PI radians
8131 will pick only pixels on same row and -PI/2 will pick only pixels on same
8135 If enabled, current pixel is compared with average value of all four
8136 surrounding pixels. The default is enabled. If disabled current pixel is
8137 compared with all four surrounding pixels. The pixel is considered banded
8138 if only all four differences with surrounding pixels are less than threshold.
8141 If enabled, current pixel is changed if and only if all pixel components are banded,
8142 e.g. banding detection threshold is triggered for all color components.
8143 The default is disabled.
8148 Remove blocking artifacts from input video.
8150 The filter accepts the following options:
8154 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
8155 This controls what kind of deblocking is applied.
8158 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
8164 Set blocking detection thresholds. Allowed range is 0 to 1.
8165 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
8166 Using higher threshold gives more deblocking strength.
8167 Setting @var{alpha} controls threshold detection at exact edge of block.
8168 Remaining options controls threshold detection near the edge. Each one for
8169 below/above or left/right. Setting any of those to @var{0} disables
8173 Set planes to filter. Default is to filter all available planes.
8176 @subsection Examples
8180 Deblock using weak filter and block size of 4 pixels.
8182 deblock=filter=weak:block=4
8186 Deblock using strong filter, block size of 4 pixels and custom thresholds for
8187 deblocking more edges.
8189 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
8193 Similar as above, but filter only first plane.
8195 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
8199 Similar as above, but filter only second and third plane.
8201 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
8208 Drop duplicated frames at regular intervals.
8210 The filter accepts the following options:
8214 Set the number of frames from which one will be dropped. Setting this to
8215 @var{N} means one frame in every batch of @var{N} frames will be dropped.
8216 Default is @code{5}.
8219 Set the threshold for duplicate detection. If the difference metric for a frame
8220 is less than or equal to this value, then it is declared as duplicate. Default
8224 Set scene change threshold. Default is @code{15}.
8228 Set the size of the x and y-axis blocks used during metric calculations.
8229 Larger blocks give better noise suppression, but also give worse detection of
8230 small movements. Must be a power of two. Default is @code{32}.
8233 Mark main input as a pre-processed input and activate clean source input
8234 stream. This allows the input to be pre-processed with various filters to help
8235 the metrics calculation while keeping the frame selection lossless. When set to
8236 @code{1}, the first stream is for the pre-processed input, and the second
8237 stream is the clean source from where the kept frames are chosen. Default is
8241 Set whether or not chroma is considered in the metric calculations. Default is
8247 Apply 2D deconvolution of video stream in frequency domain using second stream
8250 The filter accepts the following options:
8254 Set which planes to process.
8257 Set which impulse video frames will be processed, can be @var{first}
8258 or @var{all}. Default is @var{all}.
8261 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
8262 and height are not same and not power of 2 or if stream prior to convolving
8266 The @code{deconvolve} filter also supports the @ref{framesync} options.
8270 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
8272 It accepts the following options:
8276 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
8277 @var{rainbows} for cross-color reduction.
8280 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
8283 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
8286 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
8289 Set temporal chroma threshold. Lower values increases reduction of cross-color.
8294 Apply deflate effect to the video.
8296 This filter replaces the pixel by the local(3x3) average by taking into account
8297 only values lower than the pixel.
8299 It accepts the following options:
8306 Limit the maximum change for each plane, default is 65535.
8307 If 0, plane will remain unchanged.
8312 Remove temporal frame luminance variations.
8314 It accepts the following options:
8318 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
8321 Set averaging mode to smooth temporal luminance variations.
8323 Available values are:
8348 Do not actually modify frame. Useful when one only wants metadata.
8353 Remove judder produced by partially interlaced telecined content.
8355 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
8356 source was partially telecined content then the output of @code{pullup,dejudder}
8357 will have a variable frame rate. May change the recorded frame rate of the
8358 container. Aside from that change, this filter will not affect constant frame
8361 The option available in this filter is:
8365 Specify the length of the window over which the judder repeats.
8367 Accepts any integer greater than 1. Useful values are:
8371 If the original was telecined from 24 to 30 fps (Film to NTSC).
8374 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8377 If a mixture of the two.
8380 The default is @samp{4}.
8385 Suppress a TV station logo by a simple interpolation of the surrounding
8386 pixels. Just set a rectangle covering the logo and watch it disappear
8387 (and sometimes something even uglier appear - your mileage may vary).
8389 It accepts the following parameters:
8394 Specify the top left corner coordinates of the logo. They must be
8399 Specify the width and height of the logo to clear. They must be
8403 Specify the thickness of the fuzzy edge of the rectangle (added to
8404 @var{w} and @var{h}). The default value is 1. This option is
8405 deprecated, setting higher values should no longer be necessary and
8409 When set to 1, a green rectangle is drawn on the screen to simplify
8410 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8411 The default value is 0.
8413 The rectangle is drawn on the outermost pixels which will be (partly)
8414 replaced with interpolated values. The values of the next pixels
8415 immediately outside this rectangle in each direction will be used to
8416 compute the interpolated pixel values inside the rectangle.
8420 @subsection Examples
8424 Set a rectangle covering the area with top left corner coordinates 0,0
8425 and size 100x77, and a band of size 10:
8427 delogo=x=0:y=0:w=100:h=77:band=10
8434 Remove the rain in the input image/video by applying the derain methods based on
8435 convolutional neural networks. Supported models:
8439 Recurrent Squeeze-and-Excitation Context Aggregation Net (RESCAN).
8440 See @url{http://openaccess.thecvf.com/content_ECCV_2018/papers/Xia_Li_Recurrent_Squeeze-and-Excitation_Context_ECCV_2018_paper.pdf}.
8443 Training scripts as well as scripts for model generation are provided in
8444 the repository at @url{https://github.com/XueweiMeng/derain_filter.git}.
8446 The filter accepts the following options:
8450 Specify which DNN backend to use for model loading and execution. This option accepts
8451 the following values:
8455 Native implementation of DNN loading and execution.
8457 Default value is @samp{native}.
8460 Set path to model file specifying network architecture and its parameters.
8461 Note that different backends use different file formats. TensorFlow backend
8462 can load files for both formats, while native backend can load files for only
8468 Attempt to fix small changes in horizontal and/or vertical shift. This
8469 filter helps remove camera shake from hand-holding a camera, bumping a
8470 tripod, moving on a vehicle, etc.
8472 The filter accepts the following options:
8480 Specify a rectangular area where to limit the search for motion
8482 If desired the search for motion vectors can be limited to a
8483 rectangular area of the frame defined by its top left corner, width
8484 and height. These parameters have the same meaning as the drawbox
8485 filter which can be used to visualise the position of the bounding
8488 This is useful when simultaneous movement of subjects within the frame
8489 might be confused for camera motion by the motion vector search.
8491 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8492 then the full frame is used. This allows later options to be set
8493 without specifying the bounding box for the motion vector search.
8495 Default - search the whole frame.
8499 Specify the maximum extent of movement in x and y directions in the
8500 range 0-64 pixels. Default 16.
8503 Specify how to generate pixels to fill blanks at the edge of the
8504 frame. Available values are:
8507 Fill zeroes at blank locations
8509 Original image at blank locations
8511 Extruded edge value at blank locations
8513 Mirrored edge at blank locations
8515 Default value is @samp{mirror}.
8518 Specify the blocksize to use for motion search. Range 4-128 pixels,
8522 Specify the contrast threshold for blocks. Only blocks with more than
8523 the specified contrast (difference between darkest and lightest
8524 pixels) will be considered. Range 1-255, default 125.
8527 Specify the search strategy. Available values are:
8530 Set exhaustive search
8532 Set less exhaustive search.
8534 Default value is @samp{exhaustive}.
8537 If set then a detailed log of the motion search is written to the
8544 Remove unwanted contamination of foreground colors, caused by reflected color of
8545 greenscreen or bluescreen.
8547 This filter accepts the following options:
8551 Set what type of despill to use.
8554 Set how spillmap will be generated.
8557 Set how much to get rid of still remaining spill.
8560 Controls amount of red in spill area.
8563 Controls amount of green in spill area.
8564 Should be -1 for greenscreen.
8567 Controls amount of blue in spill area.
8568 Should be -1 for bluescreen.
8571 Controls brightness of spill area, preserving colors.
8574 Modify alpha from generated spillmap.
8579 Apply an exact inverse of the telecine operation. It requires a predefined
8580 pattern specified using the pattern option which must be the same as that passed
8581 to the telecine filter.
8583 This filter accepts the following options:
8592 The default value is @code{top}.
8596 A string of numbers representing the pulldown pattern you wish to apply.
8597 The default value is @code{23}.
8600 A number representing position of the first frame with respect to the telecine
8601 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8606 Apply dilation effect to the video.
8608 This filter replaces the pixel by the local(3x3) maximum.
8610 It accepts the following options:
8617 Limit the maximum change for each plane, default is 65535.
8618 If 0, plane will remain unchanged.
8621 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8624 Flags to local 3x3 coordinates maps like this:
8633 Displace pixels as indicated by second and third input stream.
8635 It takes three input streams and outputs one stream, the first input is the
8636 source, and second and third input are displacement maps.
8638 The second input specifies how much to displace pixels along the
8639 x-axis, while the third input specifies how much to displace pixels
8641 If one of displacement map streams terminates, last frame from that
8642 displacement map will be used.
8644 Note that once generated, displacements maps can be reused over and over again.
8646 A description of the accepted options follows.
8650 Set displace behavior for pixels that are out of range.
8652 Available values are:
8655 Missing pixels are replaced by black pixels.
8658 Adjacent pixels will spread out to replace missing pixels.
8661 Out of range pixels are wrapped so they point to pixels of other side.
8664 Out of range pixels will be replaced with mirrored pixels.
8666 Default is @samp{smear}.
8670 @subsection Examples
8674 Add ripple effect to rgb input of video size hd720:
8676 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
8680 Add wave effect to rgb input of video size hd720:
8682 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
8688 Draw a colored box on the input image.
8690 It accepts the following parameters:
8695 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8699 The expressions which specify the width and height of the box; if 0 they are interpreted as
8700 the input width and height. It defaults to 0.
8703 Specify the color of the box to write. For the general syntax of this option,
8704 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8705 value @code{invert} is used, the box edge color is the same as the
8706 video with inverted luma.
8709 The expression which sets the thickness of the box edge.
8710 A value of @code{fill} will create a filled box. Default value is @code{3}.
8712 See below for the list of accepted constants.
8715 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8716 will overwrite the video's color and alpha pixels.
8717 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8720 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8721 following constants:
8725 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8729 horizontal and vertical chroma subsample values. For example for the
8730 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8734 The input width and height.
8737 The input sample aspect ratio.
8741 The x and y offset coordinates where the box is drawn.
8745 The width and height of the drawn box.
8748 The thickness of the drawn box.
8750 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8751 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8755 @subsection Examples
8759 Draw a black box around the edge of the input image:
8765 Draw a box with color red and an opacity of 50%:
8767 drawbox=10:20:200:60:red@@0.5
8770 The previous example can be specified as:
8772 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8776 Fill the box with pink color:
8778 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8782 Draw a 2-pixel red 2.40:1 mask:
8784 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
8790 Draw a grid on the input image.
8792 It accepts the following parameters:
8797 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8801 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8802 input width and height, respectively, minus @code{thickness}, so image gets
8803 framed. Default to 0.
8806 Specify the color of the grid. For the general syntax of this option,
8807 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8808 value @code{invert} is used, the grid color is the same as the
8809 video with inverted luma.
8812 The expression which sets the thickness of the grid line. Default value is @code{1}.
8814 See below for the list of accepted constants.
8817 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8818 will overwrite the video's color and alpha pixels.
8819 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8822 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8823 following constants:
8827 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8831 horizontal and vertical chroma subsample values. For example for the
8832 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8836 The input grid cell width and height.
8839 The input sample aspect ratio.
8843 The x and y coordinates of some point of grid intersection (meant to configure offset).
8847 The width and height of the drawn cell.
8850 The thickness of the drawn cell.
8852 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8853 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8857 @subsection Examples
8861 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8863 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8867 Draw a white 3x3 grid with an opacity of 50%:
8869 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8876 Draw a text string or text from a specified file on top of a video, using the
8877 libfreetype library.
8879 To enable compilation of this filter, you need to configure FFmpeg with
8880 @code{--enable-libfreetype}.
8881 To enable default font fallback and the @var{font} option you need to
8882 configure FFmpeg with @code{--enable-libfontconfig}.
8883 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8884 @code{--enable-libfribidi}.
8888 It accepts the following parameters:
8893 Used to draw a box around text using the background color.
8894 The value must be either 1 (enable) or 0 (disable).
8895 The default value of @var{box} is 0.
8898 Set the width of the border to be drawn around the box using @var{boxcolor}.
8899 The default value of @var{boxborderw} is 0.
8902 The color to be used for drawing box around text. For the syntax of this
8903 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8905 The default value of @var{boxcolor} is "white".
8908 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8909 The default value of @var{line_spacing} is 0.
8912 Set the width of the border to be drawn around the text using @var{bordercolor}.
8913 The default value of @var{borderw} is 0.
8916 Set the color to be used for drawing border around text. For the syntax of this
8917 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8919 The default value of @var{bordercolor} is "black".
8922 Select how the @var{text} is expanded. Can be either @code{none},
8923 @code{strftime} (deprecated) or
8924 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8928 Set a start time for the count. Value is in microseconds. Only applied
8929 in the deprecated strftime expansion mode. To emulate in normal expansion
8930 mode use the @code{pts} function, supplying the start time (in seconds)
8931 as the second argument.
8934 If true, check and fix text coords to avoid clipping.
8937 The color to be used for drawing fonts. For the syntax of this option, check
8938 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8940 The default value of @var{fontcolor} is "black".
8942 @item fontcolor_expr
8943 String which is expanded the same way as @var{text} to obtain dynamic
8944 @var{fontcolor} value. By default this option has empty value and is not
8945 processed. When this option is set, it overrides @var{fontcolor} option.
8948 The font family to be used for drawing text. By default Sans.
8951 The font file to be used for drawing text. The path must be included.
8952 This parameter is mandatory if the fontconfig support is disabled.
8955 Draw the text applying alpha blending. The value can
8956 be a number between 0.0 and 1.0.
8957 The expression accepts the same variables @var{x, y} as well.
8958 The default value is 1.
8959 Please see @var{fontcolor_expr}.
8962 The font size to be used for drawing text.
8963 The default value of @var{fontsize} is 16.
8966 If set to 1, attempt to shape the text (for example, reverse the order of
8967 right-to-left text and join Arabic characters) before drawing it.
8968 Otherwise, just draw the text exactly as given.
8969 By default 1 (if supported).
8972 The flags to be used for loading the fonts.
8974 The flags map the corresponding flags supported by libfreetype, and are
8975 a combination of the following values:
8982 @item vertical_layout
8983 @item force_autohint
8986 @item ignore_global_advance_width
8988 @item ignore_transform
8994 Default value is "default".
8996 For more information consult the documentation for the FT_LOAD_*
9000 The color to be used for drawing a shadow behind the drawn text. For the
9001 syntax of this option, check the @ref{color syntax,,"Color" section in the
9002 ffmpeg-utils manual,ffmpeg-utils}.
9004 The default value of @var{shadowcolor} is "black".
9008 The x and y offsets for the text shadow position with respect to the
9009 position of the text. They can be either positive or negative
9010 values. The default value for both is "0".
9013 The starting frame number for the n/frame_num variable. The default value
9017 The size in number of spaces to use for rendering the tab.
9021 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
9022 format. It can be used with or without text parameter. @var{timecode_rate}
9023 option must be specified.
9025 @item timecode_rate, rate, r
9026 Set the timecode frame rate (timecode only). Value will be rounded to nearest
9027 integer. Minimum value is "1".
9028 Drop-frame timecode is supported for frame rates 30 & 60.
9031 If set to 1, the output of the timecode option will wrap around at 24 hours.
9032 Default is 0 (disabled).
9035 The text string to be drawn. The text must be a sequence of UTF-8
9037 This parameter is mandatory if no file is specified with the parameter
9041 A text file containing text to be drawn. The text must be a sequence
9042 of UTF-8 encoded characters.
9044 This parameter is mandatory if no text string is specified with the
9045 parameter @var{text}.
9047 If both @var{text} and @var{textfile} are specified, an error is thrown.
9050 If set to 1, the @var{textfile} will be reloaded before each frame.
9051 Be sure to update it atomically, or it may be read partially, or even fail.
9055 The expressions which specify the offsets where text will be drawn
9056 within the video frame. They are relative to the top/left border of the
9059 The default value of @var{x} and @var{y} is "0".
9061 See below for the list of accepted constants and functions.
9064 The parameters for @var{x} and @var{y} are expressions containing the
9065 following constants and functions:
9069 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
9073 horizontal and vertical chroma subsample values. For example for the
9074 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
9077 the height of each text line
9085 @item max_glyph_a, ascent
9086 the maximum distance from the baseline to the highest/upper grid
9087 coordinate used to place a glyph outline point, for all the rendered
9089 It is a positive value, due to the grid's orientation with the Y axis
9092 @item max_glyph_d, descent
9093 the maximum distance from the baseline to the lowest grid coordinate
9094 used to place a glyph outline point, for all the rendered glyphs.
9095 This is a negative value, due to the grid's orientation, with the Y axis
9099 maximum glyph height, that is the maximum height for all the glyphs
9100 contained in the rendered text, it is equivalent to @var{ascent} -
9104 maximum glyph width, that is the maximum width for all the glyphs
9105 contained in the rendered text
9108 the number of input frame, starting from 0
9110 @item rand(min, max)
9111 return a random number included between @var{min} and @var{max}
9114 The input sample aspect ratio.
9117 timestamp expressed in seconds, NAN if the input timestamp is unknown
9120 the height of the rendered text
9123 the width of the rendered text
9127 the x and y offset coordinates where the text is drawn.
9129 These parameters allow the @var{x} and @var{y} expressions to refer
9130 to each other, so you can for example specify @code{y=x/dar}.
9133 A one character description of the current frame's picture type.
9136 The current packet's position in the input file or stream
9137 (in bytes, from the start of the input). A value of -1 indicates
9138 this info is not available.
9141 The current packet's duration, in seconds.
9144 The current packet's size (in bytes).
9147 @anchor{drawtext_expansion}
9148 @subsection Text expansion
9150 If @option{expansion} is set to @code{strftime},
9151 the filter recognizes strftime() sequences in the provided text and
9152 expands them accordingly. Check the documentation of strftime(). This
9153 feature is deprecated.
9155 If @option{expansion} is set to @code{none}, the text is printed verbatim.
9157 If @option{expansion} is set to @code{normal} (which is the default),
9158 the following expansion mechanism is used.
9160 The backslash character @samp{\}, followed by any character, always expands to
9161 the second character.
9163 Sequences of the form @code{%@{...@}} are expanded. The text between the
9164 braces is a function name, possibly followed by arguments separated by ':'.
9165 If the arguments contain special characters or delimiters (':' or '@}'),
9166 they should be escaped.
9168 Note that they probably must also be escaped as the value for the
9169 @option{text} option in the filter argument string and as the filter
9170 argument in the filtergraph description, and possibly also for the shell,
9171 that makes up to four levels of escaping; using a text file avoids these
9174 The following functions are available:
9179 The expression evaluation result.
9181 It must take one argument specifying the expression to be evaluated,
9182 which accepts the same constants and functions as the @var{x} and
9183 @var{y} values. Note that not all constants should be used, for
9184 example the text size is not known when evaluating the expression, so
9185 the constants @var{text_w} and @var{text_h} will have an undefined
9188 @item expr_int_format, eif
9189 Evaluate the expression's value and output as formatted integer.
9191 The first argument is the expression to be evaluated, just as for the @var{expr} function.
9192 The second argument specifies the output format. Allowed values are @samp{x},
9193 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
9194 @code{printf} function.
9195 The third parameter is optional and sets the number of positions taken by the output.
9196 It can be used to add padding with zeros from the left.
9199 The time at which the filter is running, expressed in UTC.
9200 It can accept an argument: a strftime() format string.
9203 The time at which the filter is running, expressed in the local time zone.
9204 It can accept an argument: a strftime() format string.
9207 Frame metadata. Takes one or two arguments.
9209 The first argument is mandatory and specifies the metadata key.
9211 The second argument is optional and specifies a default value, used when the
9212 metadata key is not found or empty.
9214 Available metadata can be identified by inspecting entries
9215 starting with TAG included within each frame section
9216 printed by running @code{ffprobe -show_frames}.
9218 String metadata generated in filters leading to
9219 the drawtext filter are also available.
9222 The frame number, starting from 0.
9225 A one character description of the current picture type.
9228 The timestamp of the current frame.
9229 It can take up to three arguments.
9231 The first argument is the format of the timestamp; it defaults to @code{flt}
9232 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
9233 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
9234 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
9235 @code{localtime} stands for the timestamp of the frame formatted as
9236 local time zone time.
9238 The second argument is an offset added to the timestamp.
9240 If the format is set to @code{hms}, a third argument @code{24HH} may be
9241 supplied to present the hour part of the formatted timestamp in 24h format
9244 If the format is set to @code{localtime} or @code{gmtime},
9245 a third argument may be supplied: a strftime() format string.
9246 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
9249 @subsection Commands
9251 This filter supports altering parameters via commands:
9254 Alter existing filter parameters.
9256 Syntax for the argument is the same as for filter invocation, e.g.
9259 fontsize=56:fontcolor=green:text='Hello World'
9262 Full filter invocation with sendcmd would look like this:
9265 sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
9269 If the entire argument can't be parsed or applied as valid values then the filter will
9270 continue with its existing parameters.
9272 @subsection Examples
9276 Draw "Test Text" with font FreeSerif, using the default values for the
9277 optional parameters.
9280 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
9284 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
9285 and y=50 (counting from the top-left corner of the screen), text is
9286 yellow with a red box around it. Both the text and the box have an
9290 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
9291 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
9294 Note that the double quotes are not necessary if spaces are not used
9295 within the parameter list.
9298 Show the text at the center of the video frame:
9300 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
9304 Show the text at a random position, switching to a new position every 30 seconds:
9306 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)"
9310 Show a text line sliding from right to left in the last row of the video
9311 frame. The file @file{LONG_LINE} is assumed to contain a single line
9314 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
9318 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
9320 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
9324 Draw a single green letter "g", at the center of the input video.
9325 The glyph baseline is placed at half screen height.
9327 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
9331 Show text for 1 second every 3 seconds:
9333 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
9337 Use fontconfig to set the font. Note that the colons need to be escaped.
9339 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
9343 Print the date of a real-time encoding (see strftime(3)):
9345 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
9349 Show text fading in and out (appearing/disappearing):
9352 DS=1.0 # display start
9353 DE=10.0 # display end
9354 FID=1.5 # fade in duration
9355 FOD=5 # fade out duration
9356 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 @}"
9360 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
9361 and the @option{fontsize} value are included in the @option{y} offset.
9363 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
9364 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
9369 For more information about libfreetype, check:
9370 @url{http://www.freetype.org/}.
9372 For more information about fontconfig, check:
9373 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
9375 For more information about libfribidi, check:
9376 @url{http://fribidi.org/}.
9380 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
9382 The filter accepts the following options:
9387 Set low and high threshold values used by the Canny thresholding
9390 The high threshold selects the "strong" edge pixels, which are then
9391 connected through 8-connectivity with the "weak" edge pixels selected
9392 by the low threshold.
9394 @var{low} and @var{high} threshold values must be chosen in the range
9395 [0,1], and @var{low} should be lesser or equal to @var{high}.
9397 Default value for @var{low} is @code{20/255}, and default value for @var{high}
9401 Define the drawing mode.
9405 Draw white/gray wires on black background.
9408 Mix the colors to create a paint/cartoon effect.
9411 Apply Canny edge detector on all selected planes.
9413 Default value is @var{wires}.
9416 Select planes for filtering. By default all available planes are filtered.
9419 @subsection Examples
9423 Standard edge detection with custom values for the hysteresis thresholding:
9425 edgedetect=low=0.1:high=0.4
9429 Painting effect without thresholding:
9431 edgedetect=mode=colormix:high=0
9436 Set brightness, contrast, saturation and approximate gamma adjustment.
9438 The filter accepts the following options:
9442 Set the contrast expression. The value must be a float value in range
9443 @code{-2.0} to @code{2.0}. The default value is "1".
9446 Set the brightness expression. The value must be a float value in
9447 range @code{-1.0} to @code{1.0}. The default value is "0".
9450 Set the saturation expression. The value must be a float in
9451 range @code{0.0} to @code{3.0}. The default value is "1".
9454 Set the gamma expression. The value must be a float in range
9455 @code{0.1} to @code{10.0}. The default value is "1".
9458 Set the gamma expression for red. The value must be a float in
9459 range @code{0.1} to @code{10.0}. The default value is "1".
9462 Set the gamma expression for green. The value must be a float in range
9463 @code{0.1} to @code{10.0}. The default value is "1".
9466 Set the gamma expression for blue. The value must be a float in range
9467 @code{0.1} to @code{10.0}. The default value is "1".
9470 Set the gamma weight expression. It can be used to reduce the effect
9471 of a high gamma value on bright image areas, e.g. keep them from
9472 getting overamplified and just plain white. The value must be a float
9473 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
9474 gamma correction all the way down while @code{1.0} leaves it at its
9475 full strength. Default is "1".
9478 Set when the expressions for brightness, contrast, saturation and
9479 gamma expressions are evaluated.
9481 It accepts the following values:
9484 only evaluate expressions once during the filter initialization or
9485 when a command is processed
9488 evaluate expressions for each incoming frame
9491 Default value is @samp{init}.
9494 The expressions accept the following parameters:
9497 frame count of the input frame starting from 0
9500 byte position of the corresponding packet in the input file, NAN if
9504 frame rate of the input video, NAN if the input frame rate is unknown
9507 timestamp expressed in seconds, NAN if the input timestamp is unknown
9510 @subsection Commands
9511 The filter supports the following commands:
9515 Set the contrast expression.
9518 Set the brightness expression.
9521 Set the saturation expression.
9524 Set the gamma expression.
9527 Set the gamma_r expression.
9530 Set gamma_g expression.
9533 Set gamma_b expression.
9536 Set gamma_weight expression.
9538 The command accepts the same syntax of the corresponding option.
9540 If the specified expression is not valid, it is kept at its current
9547 Apply erosion effect to the video.
9549 This filter replaces the pixel by the local(3x3) minimum.
9551 It accepts the following options:
9558 Limit the maximum change for each plane, default is 65535.
9559 If 0, plane will remain unchanged.
9562 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9565 Flags to local 3x3 coordinates maps like this:
9572 @section extractplanes
9574 Extract color channel components from input video stream into
9575 separate grayscale video streams.
9577 The filter accepts the following option:
9581 Set plane(s) to extract.
9583 Available values for planes are:
9594 Choosing planes not available in the input will result in an error.
9595 That means you cannot select @code{r}, @code{g}, @code{b} planes
9596 with @code{y}, @code{u}, @code{v} planes at same time.
9599 @subsection Examples
9603 Extract luma, u and v color channel component from input video frame
9604 into 3 grayscale outputs:
9606 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
9612 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9614 For each input image, the filter will compute the optimal mapping from
9615 the input to the output given the codebook length, that is the number
9616 of distinct output colors.
9618 This filter accepts the following options.
9621 @item codebook_length, l
9622 Set codebook length. The value must be a positive integer, and
9623 represents the number of distinct output colors. Default value is 256.
9626 Set the maximum number of iterations to apply for computing the optimal
9627 mapping. The higher the value the better the result and the higher the
9628 computation time. Default value is 1.
9631 Set a random seed, must be an integer included between 0 and
9632 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9633 will try to use a good random seed on a best effort basis.
9636 Set pal8 output pixel format. This option does not work with codebook
9637 length greater than 256.
9642 Measure graylevel entropy in histogram of color channels of video frames.
9644 It accepts the following parameters:
9648 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9650 @var{diff} mode measures entropy of histogram delta values, absolute differences
9651 between neighbour histogram values.
9656 Apply a fade-in/out effect to the input video.
9658 It accepts the following parameters:
9662 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9664 Default is @code{in}.
9666 @item start_frame, s
9667 Specify the number of the frame to start applying the fade
9668 effect at. Default is 0.
9671 The number of frames that the fade effect lasts. At the end of the
9672 fade-in effect, the output video will have the same intensity as the input video.
9673 At the end of the fade-out transition, the output video will be filled with the
9674 selected @option{color}.
9678 If set to 1, fade only alpha channel, if one exists on the input.
9681 @item start_time, st
9682 Specify the timestamp (in seconds) of the frame to start to apply the fade
9683 effect. If both start_frame and start_time are specified, the fade will start at
9684 whichever comes last. Default is 0.
9687 The number of seconds for which the fade effect has to last. At the end of the
9688 fade-in effect the output video will have the same intensity as the input video,
9689 at the end of the fade-out transition the output video will be filled with the
9690 selected @option{color}.
9691 If both duration and nb_frames are specified, duration is used. Default is 0
9692 (nb_frames is used by default).
9695 Specify the color of the fade. Default is "black".
9698 @subsection Examples
9702 Fade in the first 30 frames of video:
9707 The command above is equivalent to:
9713 Fade out the last 45 frames of a 200-frame video:
9716 fade=type=out:start_frame=155:nb_frames=45
9720 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9722 fade=in:0:25, fade=out:975:25
9726 Make the first 5 frames yellow, then fade in from frame 5-24:
9728 fade=in:5:20:color=yellow
9732 Fade in alpha over first 25 frames of video:
9734 fade=in:0:25:alpha=1
9738 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9740 fade=t=in:st=5.5:d=0.5
9746 Apply arbitrary expressions to samples in frequency domain
9750 Adjust the dc value (gain) of the luma plane of the image. The filter
9751 accepts an integer value in range @code{0} to @code{1000}. The default
9752 value is set to @code{0}.
9755 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9756 filter accepts an integer value in range @code{0} to @code{1000}. The
9757 default value is set to @code{0}.
9760 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9761 filter accepts an integer value in range @code{0} to @code{1000}. The
9762 default value is set to @code{0}.
9765 Set the frequency domain weight expression for the luma plane.
9768 Set the frequency domain weight expression for the 1st chroma plane.
9771 Set the frequency domain weight expression for the 2nd chroma plane.
9774 Set when the expressions are evaluated.
9776 It accepts the following values:
9779 Only evaluate expressions once during the filter initialization.
9782 Evaluate expressions for each incoming frame.
9785 Default value is @samp{init}.
9787 The filter accepts the following variables:
9790 The coordinates of the current sample.
9794 The width and height of the image.
9797 The number of input frame, starting from 0.
9800 @subsection Examples
9806 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9812 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9818 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9824 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9830 Denoise frames using 3D FFT (frequency domain filtering).
9832 The filter accepts the following options:
9836 Set the noise sigma constant. This sets denoising strength.
9837 Default value is 1. Allowed range is from 0 to 30.
9838 Using very high sigma with low overlap may give blocking artifacts.
9841 Set amount of denoising. By default all detected noise is reduced.
9842 Default value is 1. Allowed range is from 0 to 1.
9845 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9846 Actual size of block in pixels is 2 to power of @var{block}, so by default
9847 block size in pixels is 2^4 which is 16.
9850 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9853 Set number of previous frames to use for denoising. By default is set to 0.
9856 Set number of next frames to to use for denoising. By default is set to 0.
9859 Set planes which will be filtered, by default are all available filtered
9865 Extract a single field from an interlaced image using stride
9866 arithmetic to avoid wasting CPU time. The output frames are marked as
9869 The filter accepts the following options:
9873 Specify whether to extract the top (if the value is @code{0} or
9874 @code{top}) or the bottom field (if the value is @code{1} or
9880 Create new frames by copying the top and bottom fields from surrounding frames
9881 supplied as numbers by the hint file.
9885 Set file containing hints: absolute/relative frame numbers.
9887 There must be one line for each frame in a clip. Each line must contain two
9888 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9889 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9890 is current frame number for @code{absolute} mode or out of [-1, 1] range
9891 for @code{relative} mode. First number tells from which frame to pick up top
9892 field and second number tells from which frame to pick up bottom field.
9894 If optionally followed by @code{+} output frame will be marked as interlaced,
9895 else if followed by @code{-} output frame will be marked as progressive, else
9896 it will be marked same as input frame.
9897 If line starts with @code{#} or @code{;} that line is skipped.
9900 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9903 Example of first several lines of @code{hint} file for @code{relative} mode:
9906 1,0 - # second frame, use third's frame top field and second's frame bottom field
9907 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9924 Field matching filter for inverse telecine. It is meant to reconstruct the
9925 progressive frames from a telecined stream. The filter does not drop duplicated
9926 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9927 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9929 The separation of the field matching and the decimation is notably motivated by
9930 the possibility of inserting a de-interlacing filter fallback between the two.
9931 If the source has mixed telecined and real interlaced content,
9932 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9933 But these remaining combed frames will be marked as interlaced, and thus can be
9934 de-interlaced by a later filter such as @ref{yadif} before decimation.
9936 In addition to the various configuration options, @code{fieldmatch} can take an
9937 optional second stream, activated through the @option{ppsrc} option. If
9938 enabled, the frames reconstruction will be based on the fields and frames from
9939 this second stream. This allows the first input to be pre-processed in order to
9940 help the various algorithms of the filter, while keeping the output lossless
9941 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9942 or brightness/contrast adjustments can help.
9944 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9945 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9946 which @code{fieldmatch} is based on. While the semantic and usage are very
9947 close, some behaviour and options names can differ.
9949 The @ref{decimate} filter currently only works for constant frame rate input.
9950 If your input has mixed telecined (30fps) and progressive content with a lower
9951 framerate like 24fps use the following filterchain to produce the necessary cfr
9952 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9954 The filter accepts the following options:
9958 Specify the assumed field order of the input stream. Available values are:
9962 Auto detect parity (use FFmpeg's internal parity value).
9964 Assume bottom field first.
9966 Assume top field first.
9969 Note that it is sometimes recommended not to trust the parity announced by the
9972 Default value is @var{auto}.
9975 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9976 sense that it won't risk creating jerkiness due to duplicate frames when
9977 possible, but if there are bad edits or blended fields it will end up
9978 outputting combed frames when a good match might actually exist. On the other
9979 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9980 but will almost always find a good frame if there is one. The other values are
9981 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9982 jerkiness and creating duplicate frames versus finding good matches in sections
9983 with bad edits, orphaned fields, blended fields, etc.
9985 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9987 Available values are:
9991 2-way matching (p/c)
9993 2-way matching, and trying 3rd match if still combed (p/c + n)
9995 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9997 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9998 still combed (p/c + n + u/b)
10000 3-way matching (p/c/n)
10002 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
10003 detected as combed (p/c/n + u/b)
10006 The parenthesis at the end indicate the matches that would be used for that
10007 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
10010 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
10013 Default value is @var{pc_n}.
10016 Mark the main input stream as a pre-processed input, and enable the secondary
10017 input stream as the clean source to pick the fields from. See the filter
10018 introduction for more details. It is similar to the @option{clip2} feature from
10021 Default value is @code{0} (disabled).
10024 Set the field to match from. It is recommended to set this to the same value as
10025 @option{order} unless you experience matching failures with that setting. In
10026 certain circumstances changing the field that is used to match from can have a
10027 large impact on matching performance. Available values are:
10031 Automatic (same value as @option{order}).
10033 Match from the bottom field.
10035 Match from the top field.
10038 Default value is @var{auto}.
10041 Set whether or not chroma is included during the match comparisons. In most
10042 cases it is recommended to leave this enabled. You should set this to @code{0}
10043 only if your clip has bad chroma problems such as heavy rainbowing or other
10044 artifacts. Setting this to @code{0} could also be used to speed things up at
10045 the cost of some accuracy.
10047 Default value is @code{1}.
10051 These define an exclusion band which excludes the lines between @option{y0} and
10052 @option{y1} from being included in the field matching decision. An exclusion
10053 band can be used to ignore subtitles, a logo, or other things that may
10054 interfere with the matching. @option{y0} sets the starting scan line and
10055 @option{y1} sets the ending line; all lines in between @option{y0} and
10056 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
10057 @option{y0} and @option{y1} to the same value will disable the feature.
10058 @option{y0} and @option{y1} defaults to @code{0}.
10061 Set the scene change detection threshold as a percentage of maximum change on
10062 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
10063 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
10064 @option{scthresh} is @code{[0.0, 100.0]}.
10066 Default value is @code{12.0}.
10069 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
10070 account the combed scores of matches when deciding what match to use as the
10071 final match. Available values are:
10075 No final matching based on combed scores.
10077 Combed scores are only used when a scene change is detected.
10079 Use combed scores all the time.
10082 Default is @var{sc}.
10085 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
10086 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
10087 Available values are:
10091 No forced calculation.
10093 Force p/c/n calculations.
10095 Force p/c/n/u/b calculations.
10098 Default value is @var{none}.
10101 This is the area combing threshold used for combed frame detection. This
10102 essentially controls how "strong" or "visible" combing must be to be detected.
10103 Larger values mean combing must be more visible and smaller values mean combing
10104 can be less visible or strong and still be detected. Valid settings are from
10105 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
10106 be detected as combed). This is basically a pixel difference value. A good
10107 range is @code{[8, 12]}.
10109 Default value is @code{9}.
10112 Sets whether or not chroma is considered in the combed frame decision. Only
10113 disable this if your source has chroma problems (rainbowing, etc.) that are
10114 causing problems for the combed frame detection with chroma enabled. Actually,
10115 using @option{chroma}=@var{0} is usually more reliable, except for the case
10116 where there is chroma only combing in the source.
10118 Default value is @code{0}.
10122 Respectively set the x-axis and y-axis size of the window used during combed
10123 frame detection. This has to do with the size of the area in which
10124 @option{combpel} pixels are required to be detected as combed for a frame to be
10125 declared combed. See the @option{combpel} parameter description for more info.
10126 Possible values are any number that is a power of 2 starting at 4 and going up
10129 Default value is @code{16}.
10132 The number of combed pixels inside any of the @option{blocky} by
10133 @option{blockx} size blocks on the frame for the frame to be detected as
10134 combed. While @option{cthresh} controls how "visible" the combing must be, this
10135 setting controls "how much" combing there must be in any localized area (a
10136 window defined by the @option{blockx} and @option{blocky} settings) on the
10137 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
10138 which point no frames will ever be detected as combed). This setting is known
10139 as @option{MI} in TFM/VFM vocabulary.
10141 Default value is @code{80}.
10144 @anchor{p/c/n/u/b meaning}
10145 @subsection p/c/n/u/b meaning
10147 @subsubsection p/c/n
10149 We assume the following telecined stream:
10152 Top fields: 1 2 2 3 4
10153 Bottom fields: 1 2 3 4 4
10156 The numbers correspond to the progressive frame the fields relate to. Here, the
10157 first two frames are progressive, the 3rd and 4th are combed, and so on.
10159 When @code{fieldmatch} is configured to run a matching from bottom
10160 (@option{field}=@var{bottom}) this is how this input stream get transformed:
10165 B 1 2 3 4 4 <-- matching reference
10174 As a result of the field matching, we can see that some frames get duplicated.
10175 To perform a complete inverse telecine, you need to rely on a decimation filter
10176 after this operation. See for instance the @ref{decimate} filter.
10178 The same operation now matching from top fields (@option{field}=@var{top})
10183 T 1 2 2 3 4 <-- matching reference
10193 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
10194 basically, they refer to the frame and field of the opposite parity:
10197 @item @var{p} matches the field of the opposite parity in the previous frame
10198 @item @var{c} matches the field of the opposite parity in the current frame
10199 @item @var{n} matches the field of the opposite parity in the next frame
10204 The @var{u} and @var{b} matching are a bit special in the sense that they match
10205 from the opposite parity flag. In the following examples, we assume that we are
10206 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
10207 'x' is placed above and below each matched fields.
10209 With bottom matching (@option{field}=@var{bottom}):
10214 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
10215 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
10223 With top matching (@option{field}=@var{top}):
10228 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
10229 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
10237 @subsection Examples
10239 Simple IVTC of a top field first telecined stream:
10241 fieldmatch=order=tff:combmatch=none, decimate
10244 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
10246 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
10249 @section fieldorder
10251 Transform the field order of the input video.
10253 It accepts the following parameters:
10258 The output field order. Valid values are @var{tff} for top field first or @var{bff}
10259 for bottom field first.
10262 The default value is @samp{tff}.
10264 The transformation is done by shifting the picture content up or down
10265 by one line, and filling the remaining line with appropriate picture content.
10266 This method is consistent with most broadcast field order converters.
10268 If the input video is not flagged as being interlaced, or it is already
10269 flagged as being of the required output field order, then this filter does
10270 not alter the incoming video.
10272 It is very useful when converting to or from PAL DV material,
10273 which is bottom field first.
10277 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
10280 @section fifo, afifo
10282 Buffer input images and send them when they are requested.
10284 It is mainly useful when auto-inserted by the libavfilter
10287 It does not take parameters.
10289 @section fillborders
10291 Fill borders of the input video, without changing video stream dimensions.
10292 Sometimes video can have garbage at the four edges and you may not want to
10293 crop video input to keep size multiple of some number.
10295 This filter accepts the following options:
10299 Number of pixels to fill from left border.
10302 Number of pixels to fill from right border.
10305 Number of pixels to fill from top border.
10308 Number of pixels to fill from bottom border.
10313 It accepts the following values:
10316 fill pixels using outermost pixels
10319 fill pixels using mirroring
10322 fill pixels with constant value
10325 Default is @var{smear}.
10328 Set color for pixels in fixed mode. Default is @var{black}.
10333 Find a rectangular object
10335 It accepts the following options:
10339 Filepath of the object image, needs to be in gray8.
10342 Detection threshold, default is 0.5.
10345 Number of mipmaps, default is 3.
10347 @item xmin, ymin, xmax, ymax
10348 Specifies the rectangle in which to search.
10351 @subsection Examples
10355 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
10357 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
10361 @section cover_rect
10363 Cover a rectangular object
10365 It accepts the following options:
10369 Filepath of the optional cover image, needs to be in yuv420.
10374 It accepts the following values:
10377 cover it by the supplied image
10379 cover it by interpolating the surrounding pixels
10382 Default value is @var{blur}.
10385 @subsection Examples
10389 Cover a rectangular object by the supplied image of a given video using @command{ffmpeg}:
10391 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
10397 Flood area with values of same pixel components with another values.
10399 It accepts the following options:
10402 Set pixel x coordinate.
10405 Set pixel y coordinate.
10408 Set source #0 component value.
10411 Set source #1 component value.
10414 Set source #2 component value.
10417 Set source #3 component value.
10420 Set destination #0 component value.
10423 Set destination #1 component value.
10426 Set destination #2 component value.
10429 Set destination #3 component value.
10435 Convert the input video to one of the specified pixel formats.
10436 Libavfilter will try to pick one that is suitable as input to
10439 It accepts the following parameters:
10443 A '|'-separated list of pixel format names, such as
10444 "pix_fmts=yuv420p|monow|rgb24".
10448 @subsection Examples
10452 Convert the input video to the @var{yuv420p} format
10454 format=pix_fmts=yuv420p
10457 Convert the input video to any of the formats in the list
10459 format=pix_fmts=yuv420p|yuv444p|yuv410p
10466 Convert the video to specified constant frame rate by duplicating or dropping
10467 frames as necessary.
10469 It accepts the following parameters:
10473 The desired output frame rate. The default is @code{25}.
10476 Assume the first PTS should be the given value, in seconds. This allows for
10477 padding/trimming at the start of stream. By default, no assumption is made
10478 about the first frame's expected PTS, so no padding or trimming is done.
10479 For example, this could be set to 0 to pad the beginning with duplicates of
10480 the first frame if a video stream starts after the audio stream or to trim any
10481 frames with a negative PTS.
10484 Timestamp (PTS) rounding method.
10486 Possible values are:
10493 round towards -infinity
10495 round towards +infinity
10499 The default is @code{near}.
10502 Action performed when reading the last frame.
10504 Possible values are:
10507 Use same timestamp rounding method as used for other frames.
10509 Pass through last frame if input duration has not been reached yet.
10511 The default is @code{round}.
10515 Alternatively, the options can be specified as a flat string:
10516 @var{fps}[:@var{start_time}[:@var{round}]].
10518 See also the @ref{setpts} filter.
10520 @subsection Examples
10524 A typical usage in order to set the fps to 25:
10530 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10532 fps=fps=film:round=near
10538 Pack two different video streams into a stereoscopic video, setting proper
10539 metadata on supported codecs. The two views should have the same size and
10540 framerate and processing will stop when the shorter video ends. Please note
10541 that you may conveniently adjust view properties with the @ref{scale} and
10544 It accepts the following parameters:
10548 The desired packing format. Supported values are:
10553 The views are next to each other (default).
10556 The views are on top of each other.
10559 The views are packed by line.
10562 The views are packed by column.
10565 The views are temporally interleaved.
10574 # Convert left and right views into a frame-sequential video
10575 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10577 # Convert views into a side-by-side video with the same output resolution as the input
10578 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
10583 Change the frame rate by interpolating new video output frames from the source
10586 This filter is not designed to function correctly with interlaced media. If
10587 you wish to change the frame rate of interlaced media then you are required
10588 to deinterlace before this filter and re-interlace after this filter.
10590 A description of the accepted options follows.
10594 Specify the output frames per second. This option can also be specified
10595 as a value alone. The default is @code{50}.
10598 Specify the start of a range where the output frame will be created as a
10599 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10600 the default is @code{15}.
10603 Specify the end of a range where the output frame will be created as a
10604 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10605 the default is @code{240}.
10608 Specify the level at which a scene change is detected as a value between
10609 0 and 100 to indicate a new scene; a low value reflects a low
10610 probability for the current frame to introduce a new scene, while a higher
10611 value means the current frame is more likely to be one.
10612 The default is @code{8.2}.
10615 Specify flags influencing the filter process.
10617 Available value for @var{flags} is:
10620 @item scene_change_detect, scd
10621 Enable scene change detection using the value of the option @var{scene}.
10622 This flag is enabled by default.
10628 Select one frame every N-th frame.
10630 This filter accepts the following option:
10633 Select frame after every @code{step} frames.
10634 Allowed values are positive integers higher than 0. Default value is @code{1}.
10637 @section freezedetect
10639 Detect frozen video.
10641 This filter logs a message and sets frame metadata when it detects that the
10642 input video has no significant change in content during a specified duration.
10643 Video freeze detection calculates the mean average absolute difference of all
10644 the components of video frames and compares it to a noise floor.
10646 The printed times and duration are expressed in seconds. The
10647 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10648 whose timestamp equals or exceeds the detection duration and it contains the
10649 timestamp of the first frame of the freeze. The
10650 @code{lavfi.freezedetect.freeze_duration} and
10651 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10654 The filter accepts the following options:
10658 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10659 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10663 Set freeze duration until notification (default is 2 seconds).
10669 Apply a frei0r effect to the input video.
10671 To enable the compilation of this filter, you need to install the frei0r
10672 header and configure FFmpeg with @code{--enable-frei0r}.
10674 It accepts the following parameters:
10679 The name of the frei0r effect to load. If the environment variable
10680 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10681 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10682 Otherwise, the standard frei0r paths are searched, in this order:
10683 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10684 @file{/usr/lib/frei0r-1/}.
10686 @item filter_params
10687 A '|'-separated list of parameters to pass to the frei0r effect.
10691 A frei0r effect parameter can be a boolean (its value is either
10692 "y" or "n"), a double, a color (specified as
10693 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10694 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10695 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10696 a position (specified as @var{X}/@var{Y}, where
10697 @var{X} and @var{Y} are floating point numbers) and/or a string.
10699 The number and types of parameters depend on the loaded effect. If an
10700 effect parameter is not specified, the default value is set.
10702 @subsection Examples
10706 Apply the distort0r effect, setting the first two double parameters:
10708 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10712 Apply the colordistance effect, taking a color as the first parameter:
10714 frei0r=colordistance:0.2/0.3/0.4
10715 frei0r=colordistance:violet
10716 frei0r=colordistance:0x112233
10720 Apply the perspective effect, specifying the top left and top right image
10723 frei0r=perspective:0.2/0.2|0.8/0.2
10727 For more information, see
10728 @url{http://frei0r.dyne.org}
10732 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10734 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10735 processing filter, one of them is performed once per block, not per pixel.
10736 This allows for much higher speed.
10738 The filter accepts the following options:
10742 Set quality. This option defines the number of levels for averaging. It accepts
10743 an integer in the range 4-5. Default value is @code{4}.
10746 Force a constant quantization parameter. It accepts an integer in range 0-63.
10747 If not set, the filter will use the QP from the video stream (if available).
10750 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10751 more details but also more artifacts, while higher values make the image smoother
10752 but also blurrier. Default value is @code{0} − PSNR optimal.
10754 @item use_bframe_qp
10755 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10756 option may cause flicker since the B-Frames have often larger QP. Default is
10757 @code{0} (not enabled).
10763 Apply Gaussian blur filter.
10765 The filter accepts the following options:
10769 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10772 Set number of steps for Gaussian approximation. Default is @code{1}.
10775 Set which planes to filter. By default all planes are filtered.
10778 Set vertical sigma, if negative it will be same as @code{sigma}.
10779 Default is @code{-1}.
10784 Apply generic equation to each pixel.
10786 The filter accepts the following options:
10789 @item lum_expr, lum
10790 Set the luminance expression.
10792 Set the chrominance blue expression.
10794 Set the chrominance red expression.
10795 @item alpha_expr, a
10796 Set the alpha expression.
10798 Set the red expression.
10799 @item green_expr, g
10800 Set the green expression.
10802 Set the blue expression.
10805 The colorspace is selected according to the specified options. If one
10806 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10807 options is specified, the filter will automatically select a YCbCr
10808 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10809 @option{blue_expr} options is specified, it will select an RGB
10812 If one of the chrominance expression is not defined, it falls back on the other
10813 one. If no alpha expression is specified it will evaluate to opaque value.
10814 If none of chrominance expressions are specified, they will evaluate
10815 to the luminance expression.
10817 The expressions can use the following variables and functions:
10821 The sequential number of the filtered frame, starting from @code{0}.
10825 The coordinates of the current sample.
10829 The width and height of the image.
10833 Width and height scale depending on the currently filtered plane. It is the
10834 ratio between the corresponding luma plane number of pixels and the current
10835 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10836 @code{0.5,0.5} for chroma planes.
10839 Time of the current frame, expressed in seconds.
10842 Return the value of the pixel at location (@var{x},@var{y}) of the current
10846 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10850 Return the value of the pixel at location (@var{x},@var{y}) of the
10851 blue-difference chroma plane. Return 0 if there is no such plane.
10854 Return the value of the pixel at location (@var{x},@var{y}) of the
10855 red-difference chroma plane. Return 0 if there is no such plane.
10860 Return the value of the pixel at location (@var{x},@var{y}) of the
10861 red/green/blue component. Return 0 if there is no such component.
10864 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10865 plane. Return 0 if there is no such plane.
10868 For functions, if @var{x} and @var{y} are outside the area, the value will be
10869 automatically clipped to the closer edge.
10871 @subsection Examples
10875 Flip the image horizontally:
10881 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10882 wavelength of 100 pixels:
10884 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10888 Generate a fancy enigmatic moving light:
10890 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
10894 Generate a quick emboss effect:
10896 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10900 Modify RGB components depending on pixel position:
10902 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10906 Create a radial gradient that is the same size as the input (also see
10907 the @ref{vignette} filter):
10909 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10915 Fix the banding artifacts that are sometimes introduced into nearly flat
10916 regions by truncation to 8-bit color depth.
10917 Interpolate the gradients that should go where the bands are, and
10920 It is designed for playback only. Do not use it prior to
10921 lossy compression, because compression tends to lose the dither and
10922 bring back the bands.
10924 It accepts the following parameters:
10929 The maximum amount by which the filter will change any one pixel. This is also
10930 the threshold for detecting nearly flat regions. Acceptable values range from
10931 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10935 The neighborhood to fit the gradient to. A larger radius makes for smoother
10936 gradients, but also prevents the filter from modifying the pixels near detailed
10937 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10938 values will be clipped to the valid range.
10942 Alternatively, the options can be specified as a flat string:
10943 @var{strength}[:@var{radius}]
10945 @subsection Examples
10949 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10955 Specify radius, omitting the strength (which will fall-back to the default
10963 @section graphmonitor, agraphmonitor
10964 Show various filtergraph stats.
10966 With this filter one can debug complete filtergraph.
10967 Especially issues with links filling with queued frames.
10969 The filter accepts the following options:
10973 Set video output size. Default is @var{hd720}.
10976 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10979 Set output mode, can be @var{fulll} or @var{compact}.
10980 In @var{compact} mode only filters with some queued frames have displayed stats.
10983 Set flags which enable which stats are shown in video.
10985 Available values for flags are:
10988 Display number of queued frames in each link.
10990 @item frame_count_in
10991 Display number of frames taken from filter.
10993 @item frame_count_out
10994 Display number of frames given out from filter.
10997 Display current filtered frame pts.
11000 Display current filtered frame time.
11003 Display time base for filter link.
11006 Display used format for filter link.
11009 Display video size or number of audio channels in case of audio used by filter link.
11012 Display video frame rate or sample rate in case of audio used by filter link.
11016 Set upper limit for video rate of output stream, Default value is @var{25}.
11017 This guarantee that output video frame rate will not be higher than this value.
11021 A color constancy variation filter which estimates scene illumination via grey edge algorithm
11022 and corrects the scene colors accordingly.
11024 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
11026 The filter accepts the following options:
11030 The order of differentiation to be applied on the scene. Must be chosen in the range
11031 [0,2] and default value is 1.
11034 The Minkowski parameter to be used for calculating the Minkowski distance. Must
11035 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
11036 max value instead of calculating Minkowski distance.
11039 The standard deviation of Gaussian blur to be applied on the scene. Must be
11040 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
11041 can't be equal to 0 if @var{difford} is greater than 0.
11044 @subsection Examples
11050 greyedge=difford=1:minknorm=5:sigma=2
11056 greyedge=difford=1:minknorm=0:sigma=2
11064 Apply a Hald CLUT to a video stream.
11066 First input is the video stream to process, and second one is the Hald CLUT.
11067 The Hald CLUT input can be a simple picture or a complete video stream.
11069 The filter accepts the following options:
11073 Force termination when the shortest input terminates. Default is @code{0}.
11075 Continue applying the last CLUT after the end of the stream. A value of
11076 @code{0} disable the filter after the last frame of the CLUT is reached.
11077 Default is @code{1}.
11080 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
11081 filters share the same internals).
11083 This filter also supports the @ref{framesync} options.
11085 More information about the Hald CLUT can be found on Eskil Steenberg's website
11086 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
11088 @subsection Workflow examples
11090 @subsubsection Hald CLUT video stream
11092 Generate an identity Hald CLUT stream altered with various effects:
11094 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
11097 Note: make sure you use a lossless codec.
11099 Then use it with @code{haldclut} to apply it on some random stream:
11101 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
11104 The Hald CLUT will be applied to the 10 first seconds (duration of
11105 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
11106 to the remaining frames of the @code{mandelbrot} stream.
11108 @subsubsection Hald CLUT with preview
11110 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
11111 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
11112 biggest possible square starting at the top left of the picture. The remaining
11113 padding pixels (bottom or right) will be ignored. This area can be used to add
11114 a preview of the Hald CLUT.
11116 Typically, the following generated Hald CLUT will be supported by the
11117 @code{haldclut} filter:
11120 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
11121 pad=iw+320 [padded_clut];
11122 smptebars=s=320x256, split [a][b];
11123 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
11124 [main][b] overlay=W-320" -frames:v 1 clut.png
11127 It contains the original and a preview of the effect of the CLUT: SMPTE color
11128 bars are displayed on the right-top, and below the same color bars processed by
11131 Then, the effect of this Hald CLUT can be visualized with:
11133 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
11138 Flip the input video horizontally.
11140 For example, to horizontally flip the input video with @command{ffmpeg}:
11142 ffmpeg -i in.avi -vf "hflip" out.avi
11146 This filter applies a global color histogram equalization on a
11149 It can be used to correct video that has a compressed range of pixel
11150 intensities. The filter redistributes the pixel intensities to
11151 equalize their distribution across the intensity range. It may be
11152 viewed as an "automatically adjusting contrast filter". This filter is
11153 useful only for correcting degraded or poorly captured source
11156 The filter accepts the following options:
11160 Determine the amount of equalization to be applied. As the strength
11161 is reduced, the distribution of pixel intensities more-and-more
11162 approaches that of the input frame. The value must be a float number
11163 in the range [0,1] and defaults to 0.200.
11166 Set the maximum intensity that can generated and scale the output
11167 values appropriately. The strength should be set as desired and then
11168 the intensity can be limited if needed to avoid washing-out. The value
11169 must be a float number in the range [0,1] and defaults to 0.210.
11172 Set the antibanding level. If enabled the filter will randomly vary
11173 the luminance of output pixels by a small amount to avoid banding of
11174 the histogram. Possible values are @code{none}, @code{weak} or
11175 @code{strong}. It defaults to @code{none}.
11180 Compute and draw a color distribution histogram for the input video.
11182 The computed histogram is a representation of the color component
11183 distribution in an image.
11185 Standard histogram displays the color components distribution in an image.
11186 Displays color graph for each color component. Shows distribution of
11187 the Y, U, V, A or R, G, B components, depending on input format, in the
11188 current frame. Below each graph a color component scale meter is shown.
11190 The filter accepts the following options:
11194 Set height of level. Default value is @code{200}.
11195 Allowed range is [50, 2048].
11198 Set height of color scale. Default value is @code{12}.
11199 Allowed range is [0, 40].
11203 It accepts the following values:
11206 Per color component graphs are placed below each other.
11209 Per color component graphs are placed side by side.
11212 Presents information identical to that in the @code{parade}, except
11213 that the graphs representing color components are superimposed directly
11216 Default is @code{stack}.
11219 Set mode. Can be either @code{linear}, or @code{logarithmic}.
11220 Default is @code{linear}.
11223 Set what color components to display.
11224 Default is @code{7}.
11227 Set foreground opacity. Default is @code{0.7}.
11230 Set background opacity. Default is @code{0.5}.
11233 @subsection Examples
11238 Calculate and draw histogram:
11240 ffplay -i input -vf histogram
11248 This is a high precision/quality 3d denoise filter. It aims to reduce
11249 image noise, producing smooth images and making still images really
11250 still. It should enhance compressibility.
11252 It accepts the following optional parameters:
11256 A non-negative floating point number which specifies spatial luma strength.
11257 It defaults to 4.0.
11259 @item chroma_spatial
11260 A non-negative floating point number which specifies spatial chroma strength.
11261 It defaults to 3.0*@var{luma_spatial}/4.0.
11264 A floating point number which specifies luma temporal strength. It defaults to
11265 6.0*@var{luma_spatial}/4.0.
11268 A floating point number which specifies chroma temporal strength. It defaults to
11269 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
11272 @anchor{hwdownload}
11273 @section hwdownload
11275 Download hardware frames to system memory.
11277 The input must be in hardware frames, and the output a non-hardware format.
11278 Not all formats will be supported on the output - it may be necessary to insert
11279 an additional @option{format} filter immediately following in the graph to get
11280 the output in a supported format.
11284 Map hardware frames to system memory or to another device.
11286 This filter has several different modes of operation; which one is used depends
11287 on the input and output formats:
11290 Hardware frame input, normal frame output
11292 Map the input frames to system memory and pass them to the output. If the
11293 original hardware frame is later required (for example, after overlaying
11294 something else on part of it), the @option{hwmap} filter can be used again
11295 in the next mode to retrieve it.
11297 Normal frame input, hardware frame output
11299 If the input is actually a software-mapped hardware frame, then unmap it -
11300 that is, return the original hardware frame.
11302 Otherwise, a device must be provided. Create new hardware surfaces on that
11303 device for the output, then map them back to the software format at the input
11304 and give those frames to the preceding filter. This will then act like the
11305 @option{hwupload} filter, but may be able to avoid an additional copy when
11306 the input is already in a compatible format.
11308 Hardware frame input and output
11310 A device must be supplied for the output, either directly or with the
11311 @option{derive_device} option. The input and output devices must be of
11312 different types and compatible - the exact meaning of this is
11313 system-dependent, but typically it means that they must refer to the same
11314 underlying hardware context (for example, refer to the same graphics card).
11316 If the input frames were originally created on the output device, then unmap
11317 to retrieve the original frames.
11319 Otherwise, map the frames to the output device - create new hardware frames
11320 on the output corresponding to the frames on the input.
11323 The following additional parameters are accepted:
11327 Set the frame mapping mode. Some combination of:
11330 The mapped frame should be readable.
11332 The mapped frame should be writeable.
11334 The mapping will always overwrite the entire frame.
11336 This may improve performance in some cases, as the original contents of the
11337 frame need not be loaded.
11339 The mapping must not involve any copying.
11341 Indirect mappings to copies of frames are created in some cases where either
11342 direct mapping is not possible or it would have unexpected properties.
11343 Setting this flag ensures that the mapping is direct and will fail if that is
11346 Defaults to @var{read+write} if not specified.
11348 @item derive_device @var{type}
11349 Rather than using the device supplied at initialisation, instead derive a new
11350 device of type @var{type} from the device the input frames exist on.
11353 In a hardware to hardware mapping, map in reverse - create frames in the sink
11354 and map them back to the source. This may be necessary in some cases where
11355 a mapping in one direction is required but only the opposite direction is
11356 supported by the devices being used.
11358 This option is dangerous - it may break the preceding filter in undefined
11359 ways if there are any additional constraints on that filter's output.
11360 Do not use it without fully understanding the implications of its use.
11366 Upload system memory frames to hardware surfaces.
11368 The device to upload to must be supplied when the filter is initialised. If
11369 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
11372 @anchor{hwupload_cuda}
11373 @section hwupload_cuda
11375 Upload system memory frames to a CUDA device.
11377 It accepts the following optional parameters:
11381 The number of the CUDA device to use
11386 Apply a high-quality magnification filter designed for pixel art. This filter
11387 was originally created by Maxim Stepin.
11389 It accepts the following option:
11393 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
11394 @code{hq3x} and @code{4} for @code{hq4x}.
11395 Default is @code{3}.
11399 Stack input videos horizontally.
11401 All streams must be of same pixel format and of same height.
11403 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
11404 to create same output.
11406 The filter accept the following option:
11410 Set number of input streams. Default is 2.
11413 If set to 1, force the output to terminate when the shortest input
11414 terminates. Default value is 0.
11419 Modify the hue and/or the saturation of the input.
11421 It accepts the following parameters:
11425 Specify the hue angle as a number of degrees. It accepts an expression,
11426 and defaults to "0".
11429 Specify the saturation in the [-10,10] range. It accepts an expression and
11433 Specify the hue angle as a number of radians. It accepts an
11434 expression, and defaults to "0".
11437 Specify the brightness in the [-10,10] range. It accepts an expression and
11441 @option{h} and @option{H} are mutually exclusive, and can't be
11442 specified at the same time.
11444 The @option{b}, @option{h}, @option{H} and @option{s} option values are
11445 expressions containing the following constants:
11449 frame count of the input frame starting from 0
11452 presentation timestamp of the input frame expressed in time base units
11455 frame rate of the input video, NAN if the input frame rate is unknown
11458 timestamp expressed in seconds, NAN if the input timestamp is unknown
11461 time base of the input video
11464 @subsection Examples
11468 Set the hue to 90 degrees and the saturation to 1.0:
11474 Same command but expressing the hue in radians:
11480 Rotate hue and make the saturation swing between 0
11481 and 2 over a period of 1 second:
11483 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11487 Apply a 3 seconds saturation fade-in effect starting at 0:
11489 hue="s=min(t/3\,1)"
11492 The general fade-in expression can be written as:
11494 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11498 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11500 hue="s=max(0\, min(1\, (8-t)/3))"
11503 The general fade-out expression can be written as:
11505 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11510 @subsection Commands
11512 This filter supports the following commands:
11518 Modify the hue and/or the saturation and/or brightness of the input video.
11519 The command accepts the same syntax of the corresponding option.
11521 If the specified expression is not valid, it is kept at its current
11525 @section hysteresis
11527 Grow first stream into second stream by connecting components.
11528 This makes it possible to build more robust edge masks.
11530 This filter accepts the following options:
11534 Set which planes will be processed as bitmap, unprocessed planes will be
11535 copied from first stream.
11536 By default value 0xf, all planes will be processed.
11539 Set threshold which is used in filtering. If pixel component value is higher than
11540 this value filter algorithm for connecting components is activated.
11541 By default value is 0.
11546 Detect video interlacing type.
11548 This filter tries to detect if the input frames are interlaced, progressive,
11549 top or bottom field first. It will also try to detect fields that are
11550 repeated between adjacent frames (a sign of telecine).
11552 Single frame detection considers only immediately adjacent frames when classifying each frame.
11553 Multiple frame detection incorporates the classification history of previous frames.
11555 The filter will log these metadata values:
11558 @item single.current_frame
11559 Detected type of current frame using single-frame detection. One of:
11560 ``tff'' (top field first), ``bff'' (bottom field first),
11561 ``progressive'', or ``undetermined''
11564 Cumulative number of frames detected as top field first using single-frame detection.
11567 Cumulative number of frames detected as top field first using multiple-frame detection.
11570 Cumulative number of frames detected as bottom field first using single-frame detection.
11572 @item multiple.current_frame
11573 Detected type of current frame using multiple-frame detection. One of:
11574 ``tff'' (top field first), ``bff'' (bottom field first),
11575 ``progressive'', or ``undetermined''
11578 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11580 @item single.progressive
11581 Cumulative number of frames detected as progressive using single-frame detection.
11583 @item multiple.progressive
11584 Cumulative number of frames detected as progressive using multiple-frame detection.
11586 @item single.undetermined
11587 Cumulative number of frames that could not be classified using single-frame detection.
11589 @item multiple.undetermined
11590 Cumulative number of frames that could not be classified using multiple-frame detection.
11592 @item repeated.current_frame
11593 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11595 @item repeated.neither
11596 Cumulative number of frames with no repeated field.
11599 Cumulative number of frames with the top field repeated from the previous frame's top field.
11601 @item repeated.bottom
11602 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11605 The filter accepts the following options:
11609 Set interlacing threshold.
11611 Set progressive threshold.
11613 Threshold for repeated field detection.
11615 Number of frames after which a given frame's contribution to the
11616 statistics is halved (i.e., it contributes only 0.5 to its
11617 classification). The default of 0 means that all frames seen are given
11618 full weight of 1.0 forever.
11619 @item analyze_interlaced_flag
11620 When this is not 0 then idet will use the specified number of frames to determine
11621 if the interlaced flag is accurate, it will not count undetermined frames.
11622 If the flag is found to be accurate it will be used without any further
11623 computations, if it is found to be inaccurate it will be cleared without any
11624 further computations. This allows inserting the idet filter as a low computational
11625 method to clean up the interlaced flag
11630 Deinterleave or interleave fields.
11632 This filter allows one to process interlaced images fields without
11633 deinterlacing them. Deinterleaving splits the input frame into 2
11634 fields (so called half pictures). Odd lines are moved to the top
11635 half of the output image, even lines to the bottom half.
11636 You can process (filter) them independently and then re-interleave them.
11638 The filter accepts the following options:
11642 @item chroma_mode, c
11643 @item alpha_mode, a
11644 Available values for @var{luma_mode}, @var{chroma_mode} and
11645 @var{alpha_mode} are:
11651 @item deinterleave, d
11652 Deinterleave fields, placing one above the other.
11654 @item interleave, i
11655 Interleave fields. Reverse the effect of deinterleaving.
11657 Default value is @code{none}.
11659 @item luma_swap, ls
11660 @item chroma_swap, cs
11661 @item alpha_swap, as
11662 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11667 Apply inflate effect to the video.
11669 This filter replaces the pixel by the local(3x3) average by taking into account
11670 only values higher than the pixel.
11672 It accepts the following options:
11679 Limit the maximum change for each plane, default is 65535.
11680 If 0, plane will remain unchanged.
11685 Simple interlacing filter from progressive contents. This interleaves upper (or
11686 lower) lines from odd frames with lower (or upper) lines from even frames,
11687 halving the frame rate and preserving image height.
11690 Original Original New Frame
11691 Frame 'j' Frame 'j+1' (tff)
11692 ========== =========== ==================
11693 Line 0 --------------------> Frame 'j' Line 0
11694 Line 1 Line 1 ----> Frame 'j+1' Line 1
11695 Line 2 ---------------------> Frame 'j' Line 2
11696 Line 3 Line 3 ----> Frame 'j+1' Line 3
11698 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11701 It accepts the following optional parameters:
11705 This determines whether the interlaced frame is taken from the even
11706 (tff - default) or odd (bff) lines of the progressive frame.
11709 Vertical lowpass filter to avoid twitter interlacing and
11710 reduce moire patterns.
11714 Disable vertical lowpass filter
11717 Enable linear filter (default)
11720 Enable complex filter. This will slightly less reduce twitter and moire
11721 but better retain detail and subjective sharpness impression.
11728 Deinterlace input video by applying Donald Graft's adaptive kernel
11729 deinterling. Work on interlaced parts of a video to produce
11730 progressive frames.
11732 The description of the accepted parameters follows.
11736 Set the threshold which affects the filter's tolerance when
11737 determining if a pixel line must be processed. It must be an integer
11738 in the range [0,255] and defaults to 10. A value of 0 will result in
11739 applying the process on every pixels.
11742 Paint pixels exceeding the threshold value to white if set to 1.
11746 Set the fields order. Swap fields if set to 1, leave fields alone if
11750 Enable additional sharpening if set to 1. Default is 0.
11753 Enable twoway sharpening if set to 1. Default is 0.
11756 @subsection Examples
11760 Apply default values:
11762 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11766 Enable additional sharpening:
11772 Paint processed pixels in white:
11780 Slowly update darker pixels.
11782 This filter makes short flashes of light appear longer.
11783 This filter accepts the following options:
11787 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
11790 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
11793 @section lenscorrection
11795 Correct radial lens distortion
11797 This filter can be used to correct for radial distortion as can result from the use
11798 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11799 one can use tools available for example as part of opencv or simply trial-and-error.
11800 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11801 and extract the k1 and k2 coefficients from the resulting matrix.
11803 Note that effectively the same filter is available in the open-source tools Krita and
11804 Digikam from the KDE project.
11806 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11807 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11808 brightness distribution, so you may want to use both filters together in certain
11809 cases, though you will have to take care of ordering, i.e. whether vignetting should
11810 be applied before or after lens correction.
11812 @subsection Options
11814 The filter accepts the following options:
11818 Relative x-coordinate of the focal point of the image, and thereby the center of the
11819 distortion. This value has a range [0,1] and is expressed as fractions of the image
11820 width. Default is 0.5.
11822 Relative y-coordinate of the focal point of the image, and thereby the center of the
11823 distortion. This value has a range [0,1] and is expressed as fractions of the image
11824 height. Default is 0.5.
11826 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11827 no correction. Default is 0.
11829 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11830 0 means no correction. Default is 0.
11833 The formula that generates the correction is:
11835 @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)
11837 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11838 distances from the focal point in the source and target images, respectively.
11842 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11844 The @code{lensfun} filter requires the camera make, camera model, and lens model
11845 to apply the lens correction. The filter will load the lensfun database and
11846 query it to find the corresponding camera and lens entries in the database. As
11847 long as these entries can be found with the given options, the filter can
11848 perform corrections on frames. Note that incomplete strings will result in the
11849 filter choosing the best match with the given options, and the filter will
11850 output the chosen camera and lens models (logged with level "info"). You must
11851 provide the make, camera model, and lens model as they are required.
11853 The filter accepts the following options:
11857 The make of the camera (for example, "Canon"). This option is required.
11860 The model of the camera (for example, "Canon EOS 100D"). This option is
11864 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11865 option is required.
11868 The type of correction to apply. The following values are valid options:
11872 Enables fixing lens vignetting.
11875 Enables fixing lens geometry. This is the default.
11878 Enables fixing chromatic aberrations.
11881 Enables fixing lens vignetting and lens geometry.
11884 Enables fixing lens vignetting and chromatic aberrations.
11887 Enables fixing both lens geometry and chromatic aberrations.
11890 Enables all possible corrections.
11894 The focal length of the image/video (zoom; expected constant for video). For
11895 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11896 range should be chosen when using that lens. Default 18.
11899 The aperture of the image/video (expected constant for video). Note that
11900 aperture is only used for vignetting correction. Default 3.5.
11902 @item focus_distance
11903 The focus distance of the image/video (expected constant for video). Note that
11904 focus distance is only used for vignetting and only slightly affects the
11905 vignetting correction process. If unknown, leave it at the default value (which
11909 The scale factor which is applied after transformation. After correction the
11910 video is no longer necessarily rectangular. This parameter controls how much of
11911 the resulting image is visible. The value 0 means that a value will be chosen
11912 automatically such that there is little or no unmapped area in the output
11913 image. 1.0 means that no additional scaling is done. Lower values may result
11914 in more of the corrected image being visible, while higher values may avoid
11915 unmapped areas in the output.
11917 @item target_geometry
11918 The target geometry of the output image/video. The following values are valid
11922 @item rectilinear (default)
11925 @item equirectangular
11926 @item fisheye_orthographic
11927 @item fisheye_stereographic
11928 @item fisheye_equisolid
11929 @item fisheye_thoby
11932 Apply the reverse of image correction (instead of correcting distortion, apply
11935 @item interpolation
11936 The type of interpolation used when correcting distortion. The following values
11941 @item linear (default)
11946 @subsection Examples
11950 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11951 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11955 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8 -c:v h264 -b:v 8000k output.mov
11959 Apply the same as before, but only for the first 5 seconds of video.
11962 ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8:enable='lte(t\,5)' -c:v h264 -b:v 8000k output.mov
11969 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11970 score between two input videos.
11972 The obtained VMAF score is printed through the logging system.
11974 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11975 After installing the library it can be enabled using:
11976 @code{./configure --enable-libvmaf --enable-version3}.
11977 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11979 The filter has following options:
11983 Set the model path which is to be used for SVM.
11984 Default value: @code{"vmaf_v0.6.1.pkl"}
11987 Set the file path to be used to store logs.
11990 Set the format of the log file (xml or json).
11992 @item enable_transform
11993 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11994 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11995 Default value: @code{false}
11998 Invokes the phone model which will generate VMAF scores higher than in the
11999 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
12002 Enables computing psnr along with vmaf.
12005 Enables computing ssim along with vmaf.
12008 Enables computing ms_ssim along with vmaf.
12011 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
12014 Set number of threads to be used when computing vmaf.
12017 Set interval for frame subsampling used when computing vmaf.
12019 @item enable_conf_interval
12020 Enables confidence interval.
12023 This filter also supports the @ref{framesync} options.
12025 On the below examples the input file @file{main.mpg} being processed is
12026 compared with the reference file @file{ref.mpg}.
12029 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
12032 Example with options:
12034 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
12039 Limits the pixel components values to the specified range [min, max].
12041 The filter accepts the following options:
12045 Lower bound. Defaults to the lowest allowed value for the input.
12048 Upper bound. Defaults to the highest allowed value for the input.
12051 Specify which planes will be processed. Defaults to all available.
12058 The filter accepts the following options:
12062 Set the number of loops. Setting this value to -1 will result in infinite loops.
12066 Set maximal size in number of frames. Default is 0.
12069 Set first frame of loop. Default is 0.
12072 @subsection Examples
12076 Loop single first frame infinitely:
12078 loop=loop=-1:size=1:start=0
12082 Loop single first frame 10 times:
12084 loop=loop=10:size=1:start=0
12088 Loop 10 first frames 5 times:
12090 loop=loop=5:size=10:start=0
12096 Apply a 1D LUT to an input video.
12098 The filter accepts the following options:
12102 Set the 1D LUT file name.
12104 Currently supported formats:
12113 Select interpolation mode.
12115 Available values are:
12119 Use values from the nearest defined point.
12121 Interpolate values using the linear interpolation.
12123 Interpolate values using the cosine interpolation.
12125 Interpolate values using the cubic interpolation.
12127 Interpolate values using the spline interpolation.
12134 Apply a 3D LUT to an input video.
12136 The filter accepts the following options:
12140 Set the 3D LUT file name.
12142 Currently supported formats:
12156 Select interpolation mode.
12158 Available values are:
12162 Use values from the nearest defined point.
12164 Interpolate values using the 8 points defining a cube.
12166 Interpolate values using a tetrahedron.
12172 Turn certain luma values into transparency.
12174 The filter accepts the following options:
12178 Set the luma which will be used as base for transparency.
12179 Default value is @code{0}.
12182 Set the range of luma values to be keyed out.
12183 Default value is @code{0}.
12186 Set the range of softness. Default value is @code{0}.
12187 Use this to control gradual transition from zero to full transparency.
12190 @section lut, lutrgb, lutyuv
12192 Compute a look-up table for binding each pixel component input value
12193 to an output value, and apply it to the input video.
12195 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
12196 to an RGB input video.
12198 These filters accept the following parameters:
12201 set first pixel component expression
12203 set second pixel component expression
12205 set third pixel component expression
12207 set fourth pixel component expression, corresponds to the alpha component
12210 set red component expression
12212 set green component expression
12214 set blue component expression
12216 alpha component expression
12219 set Y/luminance component expression
12221 set U/Cb component expression
12223 set V/Cr component expression
12226 Each of them specifies the expression to use for computing the lookup table for
12227 the corresponding pixel component values.
12229 The exact component associated to each of the @var{c*} options depends on the
12232 The @var{lut} filter requires either YUV or RGB pixel formats in input,
12233 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
12235 The expressions can contain the following constants and functions:
12240 The input width and height.
12243 The input value for the pixel component.
12246 The input value, clipped to the @var{minval}-@var{maxval} range.
12249 The maximum value for the pixel component.
12252 The minimum value for the pixel component.
12255 The negated value for the pixel component value, clipped to the
12256 @var{minval}-@var{maxval} range; it corresponds to the expression
12257 "maxval-clipval+minval".
12260 The computed value in @var{val}, clipped to the
12261 @var{minval}-@var{maxval} range.
12263 @item gammaval(gamma)
12264 The computed gamma correction value of the pixel component value,
12265 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
12267 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
12271 All expressions default to "val".
12273 @subsection Examples
12277 Negate input video:
12279 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
12280 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
12283 The above is the same as:
12285 lutrgb="r=negval:g=negval:b=negval"
12286 lutyuv="y=negval:u=negval:v=negval"
12296 Remove chroma components, turning the video into a graytone image:
12298 lutyuv="u=128:v=128"
12302 Apply a luma burning effect:
12308 Remove green and blue components:
12314 Set a constant alpha channel value on input:
12316 format=rgba,lutrgb=a="maxval-minval/2"
12320 Correct luminance gamma by a factor of 0.5:
12322 lutyuv=y=gammaval(0.5)
12326 Discard least significant bits of luma:
12328 lutyuv=y='bitand(val, 128+64+32)'
12332 Technicolor like effect:
12334 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
12338 @section lut2, tlut2
12340 The @code{lut2} filter takes two input streams and outputs one
12343 The @code{tlut2} (time lut2) filter takes two consecutive frames
12344 from one single stream.
12346 This filter accepts the following parameters:
12349 set first pixel component expression
12351 set second pixel component expression
12353 set third pixel component expression
12355 set fourth pixel component expression, corresponds to the alpha component
12358 set output bit depth, only available for @code{lut2} filter. By default is 0,
12359 which means bit depth is automatically picked from first input format.
12362 Each of them specifies the expression to use for computing the lookup table for
12363 the corresponding pixel component values.
12365 The exact component associated to each of the @var{c*} options depends on the
12368 The expressions can contain the following constants:
12373 The input width and height.
12376 The first input value for the pixel component.
12379 The second input value for the pixel component.
12382 The first input video bit depth.
12385 The second input video bit depth.
12388 All expressions default to "x".
12390 @subsection Examples
12394 Highlight differences between two RGB video streams:
12396 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)'
12400 Highlight differences between two YUV video streams:
12402 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)'
12406 Show max difference between two video streams:
12408 lut2='if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1)))'
12412 @section maskedclamp
12414 Clamp the first input stream with the second input and third input stream.
12416 Returns the value of first stream to be between second input
12417 stream - @code{undershoot} and third input stream + @code{overshoot}.
12419 This filter accepts the following options:
12422 Default value is @code{0}.
12425 Default value is @code{0}.
12428 Set which planes will be processed as bitmap, unprocessed planes will be
12429 copied from first stream.
12430 By default value 0xf, all planes will be processed.
12433 @section maskedmerge
12435 Merge the first input stream with the second input stream using per pixel
12436 weights in the third input stream.
12438 A value of 0 in the third stream pixel component means that pixel component
12439 from first stream is returned unchanged, while maximum value (eg. 255 for
12440 8-bit videos) means that pixel component from second stream is returned
12441 unchanged. Intermediate values define the amount of merging between both
12442 input stream's pixel components.
12444 This filter accepts the following options:
12447 Set which planes will be processed as bitmap, unprocessed planes will be
12448 copied from first stream.
12449 By default value 0xf, all planes will be processed.
12453 Create mask from input video.
12455 For example it is useful to create motion masks after @code{tblend} filter.
12457 This filter accepts the following options:
12461 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
12464 Set high threshold. Any pixel component higher than this value will be set to max value
12465 allowed for current pixel format.
12468 Set planes to filter, by default all available planes are filtered.
12471 Fill all frame pixels with this value.
12474 Set max average pixel value for frame. If sum of all pixel components is higher that this
12475 average, output frame will be completely filled with value set by @var{fill} option.
12476 Typically useful for scene changes when used in combination with @code{tblend} filter.
12481 Apply motion-compensation deinterlacing.
12483 It needs one field per frame as input and must thus be used together
12484 with yadif=1/3 or equivalent.
12486 This filter accepts the following options:
12489 Set the deinterlacing mode.
12491 It accepts one of the following values:
12496 use iterative motion estimation
12498 like @samp{slow}, but use multiple reference frames.
12500 Default value is @samp{fast}.
12503 Set the picture field parity assumed for the input video. It must be
12504 one of the following values:
12508 assume top field first
12510 assume bottom field first
12513 Default value is @samp{bff}.
12516 Set per-block quantization parameter (QP) used by the internal
12519 Higher values should result in a smoother motion vector field but less
12520 optimal individual vectors. Default value is 1.
12523 @section mergeplanes
12525 Merge color channel components from several video streams.
12527 The filter accepts up to 4 input streams, and merge selected input
12528 planes to the output video.
12530 This filter accepts the following options:
12533 Set input to output plane mapping. Default is @code{0}.
12535 The mappings is specified as a bitmap. It should be specified as a
12536 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12537 mapping for the first plane of the output stream. 'A' sets the number of
12538 the input stream to use (from 0 to 3), and 'a' the plane number of the
12539 corresponding input to use (from 0 to 3). The rest of the mappings is
12540 similar, 'Bb' describes the mapping for the output stream second
12541 plane, 'Cc' describes the mapping for the output stream third plane and
12542 'Dd' describes the mapping for the output stream fourth plane.
12545 Set output pixel format. Default is @code{yuva444p}.
12548 @subsection Examples
12552 Merge three gray video streams of same width and height into single video stream:
12554 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12558 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12560 [a0][a1]mergeplanes=0x00010210:yuva444p
12564 Swap Y and A plane in yuva444p stream:
12566 format=yuva444p,mergeplanes=0x03010200:yuva444p
12570 Swap U and V plane in yuv420p stream:
12572 format=yuv420p,mergeplanes=0x000201:yuv420p
12576 Cast a rgb24 clip to yuv444p:
12578 format=rgb24,mergeplanes=0x000102:yuv444p
12584 Estimate and export motion vectors using block matching algorithms.
12585 Motion vectors are stored in frame side data to be used by other filters.
12587 This filter accepts the following options:
12590 Specify the motion estimation method. Accepts one of the following values:
12594 Exhaustive search algorithm.
12596 Three step search algorithm.
12598 Two dimensional logarithmic search algorithm.
12600 New three step search algorithm.
12602 Four step search algorithm.
12604 Diamond search algorithm.
12606 Hexagon-based search algorithm.
12608 Enhanced predictive zonal search algorithm.
12610 Uneven multi-hexagon search algorithm.
12612 Default value is @samp{esa}.
12615 Macroblock size. Default @code{16}.
12618 Search parameter. Default @code{7}.
12621 @section midequalizer
12623 Apply Midway Image Equalization effect using two video streams.
12625 Midway Image Equalization adjusts a pair of images to have the same
12626 histogram, while maintaining their dynamics as much as possible. It's
12627 useful for e.g. matching exposures from a pair of stereo cameras.
12629 This filter has two inputs and one output, which must be of same pixel format, but
12630 may be of different sizes. The output of filter is first input adjusted with
12631 midway histogram of both inputs.
12633 This filter accepts the following option:
12637 Set which planes to process. Default is @code{15}, which is all available planes.
12640 @section minterpolate
12642 Convert the video to specified frame rate using motion interpolation.
12644 This filter accepts the following options:
12647 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}.
12650 Motion interpolation mode. Following values are accepted:
12653 Duplicate previous or next frame for interpolating new ones.
12655 Blend source frames. Interpolated frame is mean of previous and next frames.
12657 Motion compensated interpolation. Following options are effective when this mode is selected:
12661 Motion compensation mode. Following values are accepted:
12664 Overlapped block motion compensation.
12666 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12668 Default mode is @samp{obmc}.
12671 Motion estimation mode. Following values are accepted:
12674 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12676 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12678 Default mode is @samp{bilat}.
12681 The algorithm to be used for motion estimation. Following values are accepted:
12684 Exhaustive search algorithm.
12686 Three step search algorithm.
12688 Two dimensional logarithmic search algorithm.
12690 New three step search algorithm.
12692 Four step search algorithm.
12694 Diamond search algorithm.
12696 Hexagon-based search algorithm.
12698 Enhanced predictive zonal search algorithm.
12700 Uneven multi-hexagon search algorithm.
12702 Default algorithm is @samp{epzs}.
12705 Macroblock size. Default @code{16}.
12708 Motion estimation search parameter. Default @code{32}.
12711 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).
12716 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:
12719 Disable scene change detection.
12721 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12723 Default method is @samp{fdiff}.
12725 @item scd_threshold
12726 Scene change detection threshold. Default is @code{5.0}.
12731 Mix several video input streams into one video stream.
12733 A description of the accepted options follows.
12737 The number of inputs. If unspecified, it defaults to 2.
12740 Specify weight of each input video stream as sequence.
12741 Each weight is separated by space. If number of weights
12742 is smaller than number of @var{frames} last specified
12743 weight will be used for all remaining unset weights.
12746 Specify scale, if it is set it will be multiplied with sum
12747 of each weight multiplied with pixel values to give final destination
12748 pixel value. By default @var{scale} is auto scaled to sum of weights.
12751 Specify how end of stream is determined.
12754 The duration of the longest input. (default)
12757 The duration of the shortest input.
12760 The duration of the first input.
12764 @section mpdecimate
12766 Drop frames that do not differ greatly from the previous frame in
12767 order to reduce frame rate.
12769 The main use of this filter is for very-low-bitrate encoding
12770 (e.g. streaming over dialup modem), but it could in theory be used for
12771 fixing movies that were inverse-telecined incorrectly.
12773 A description of the accepted options follows.
12777 Set the maximum number of consecutive frames which can be dropped (if
12778 positive), or the minimum interval between dropped frames (if
12779 negative). If the value is 0, the frame is dropped disregarding the
12780 number of previous sequentially dropped frames.
12782 Default value is 0.
12787 Set the dropping threshold values.
12789 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12790 represent actual pixel value differences, so a threshold of 64
12791 corresponds to 1 unit of difference for each pixel, or the same spread
12792 out differently over the block.
12794 A frame is a candidate for dropping if no 8x8 blocks differ by more
12795 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12796 meaning the whole image) differ by more than a threshold of @option{lo}.
12798 Default value for @option{hi} is 64*12, default value for @option{lo} is
12799 64*5, and default value for @option{frac} is 0.33.
12805 Negate (invert) the input video.
12807 It accepts the following option:
12812 With value 1, it negates the alpha component, if present. Default value is 0.
12818 Denoise frames using Non-Local Means algorithm.
12820 Each pixel is adjusted by looking for other pixels with similar contexts. This
12821 context similarity is defined by comparing their surrounding patches of size
12822 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12825 Note that the research area defines centers for patches, which means some
12826 patches will be made of pixels outside that research area.
12828 The filter accepts the following options.
12832 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12835 Set patch size. Default is 7. Must be odd number in range [0, 99].
12838 Same as @option{p} but for chroma planes.
12840 The default value is @var{0} and means automatic.
12843 Set research size. Default is 15. Must be odd number in range [0, 99].
12846 Same as @option{r} but for chroma planes.
12848 The default value is @var{0} and means automatic.
12853 Deinterlace video using neural network edge directed interpolation.
12855 This filter accepts the following options:
12859 Mandatory option, without binary file filter can not work.
12860 Currently file can be found here:
12861 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12864 Set which frames to deinterlace, by default it is @code{all}.
12865 Can be @code{all} or @code{interlaced}.
12868 Set mode of operation.
12870 Can be one of the following:
12874 Use frame flags, both fields.
12876 Use frame flags, single field.
12878 Use top field only.
12880 Use bottom field only.
12882 Use both fields, top first.
12884 Use both fields, bottom first.
12888 Set which planes to process, by default filter process all frames.
12891 Set size of local neighborhood around each pixel, used by the predictor neural
12894 Can be one of the following:
12907 Set the number of neurons in predictor neural network.
12908 Can be one of the following:
12919 Controls the number of different neural network predictions that are blended
12920 together to compute the final output value. Can be @code{fast}, default or
12924 Set which set of weights to use in the predictor.
12925 Can be one of the following:
12929 weights trained to minimize absolute error
12931 weights trained to minimize squared error
12935 Controls whether or not the prescreener neural network is used to decide
12936 which pixels should be processed by the predictor neural network and which
12937 can be handled by simple cubic interpolation.
12938 The prescreener is trained to know whether cubic interpolation will be
12939 sufficient for a pixel or whether it should be predicted by the predictor nn.
12940 The computational complexity of the prescreener nn is much less than that of
12941 the predictor nn. Since most pixels can be handled by cubic interpolation,
12942 using the prescreener generally results in much faster processing.
12943 The prescreener is pretty accurate, so the difference between using it and not
12944 using it is almost always unnoticeable.
12946 Can be one of the following:
12954 Default is @code{new}.
12957 Set various debugging flags.
12962 Force libavfilter not to use any of the specified pixel formats for the
12963 input to the next filter.
12965 It accepts the following parameters:
12969 A '|'-separated list of pixel format names, such as
12970 pix_fmts=yuv420p|monow|rgb24".
12974 @subsection Examples
12978 Force libavfilter to use a format different from @var{yuv420p} for the
12979 input to the vflip filter:
12981 noformat=pix_fmts=yuv420p,vflip
12985 Convert the input video to any of the formats not contained in the list:
12987 noformat=yuv420p|yuv444p|yuv410p
12993 Add noise on video input frame.
12995 The filter accepts the following options:
13003 Set noise seed for specific pixel component or all pixel components in case
13004 of @var{all_seed}. Default value is @code{123457}.
13006 @item all_strength, alls
13007 @item c0_strength, c0s
13008 @item c1_strength, c1s
13009 @item c2_strength, c2s
13010 @item c3_strength, c3s
13011 Set noise strength for specific pixel component or all pixel components in case
13012 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
13014 @item all_flags, allf
13015 @item c0_flags, c0f
13016 @item c1_flags, c1f
13017 @item c2_flags, c2f
13018 @item c3_flags, c3f
13019 Set pixel component flags or set flags for all components if @var{all_flags}.
13020 Available values for component flags are:
13023 averaged temporal noise (smoother)
13025 mix random noise with a (semi)regular pattern
13027 temporal noise (noise pattern changes between frames)
13029 uniform noise (gaussian otherwise)
13033 @subsection Examples
13035 Add temporal and uniform noise to input video:
13037 noise=alls=20:allf=t+u
13042 Normalize RGB video (aka histogram stretching, contrast stretching).
13043 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
13045 For each channel of each frame, the filter computes the input range and maps
13046 it linearly to the user-specified output range. The output range defaults
13047 to the full dynamic range from pure black to pure white.
13049 Temporal smoothing can be used on the input range to reduce flickering (rapid
13050 changes in brightness) caused when small dark or bright objects enter or leave
13051 the scene. This is similar to the auto-exposure (automatic gain control) on a
13052 video camera, and, like a video camera, it may cause a period of over- or
13053 under-exposure of the video.
13055 The R,G,B channels can be normalized independently, which may cause some
13056 color shifting, or linked together as a single channel, which prevents
13057 color shifting. Linked normalization preserves hue. Independent normalization
13058 does not, so it can be used to remove some color casts. Independent and linked
13059 normalization can be combined in any ratio.
13061 The normalize filter accepts the following options:
13066 Colors which define the output range. The minimum input value is mapped to
13067 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
13068 The defaults are black and white respectively. Specifying white for
13069 @var{blackpt} and black for @var{whitept} will give color-inverted,
13070 normalized video. Shades of grey can be used to reduce the dynamic range
13071 (contrast). Specifying saturated colors here can create some interesting
13075 The number of previous frames to use for temporal smoothing. The input range
13076 of each channel is smoothed using a rolling average over the current frame
13077 and the @var{smoothing} previous frames. The default is 0 (no temporal
13081 Controls the ratio of independent (color shifting) channel normalization to
13082 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
13083 independent. Defaults to 1.0 (fully independent).
13086 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
13087 expensive no-op. Defaults to 1.0 (full strength).
13091 @subsection Examples
13093 Stretch video contrast to use the full dynamic range, with no temporal
13094 smoothing; may flicker depending on the source content:
13096 normalize=blackpt=black:whitept=white:smoothing=0
13099 As above, but with 50 frames of temporal smoothing; flicker should be
13100 reduced, depending on the source content:
13102 normalize=blackpt=black:whitept=white:smoothing=50
13105 As above, but with hue-preserving linked channel normalization:
13107 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
13110 As above, but with half strength:
13112 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
13115 Map the darkest input color to red, the brightest input color to cyan:
13117 normalize=blackpt=red:whitept=cyan
13122 Pass the video source unchanged to the output.
13125 Optical Character Recognition
13127 This filter uses Tesseract for optical character recognition. To enable
13128 compilation of this filter, you need to configure FFmpeg with
13129 @code{--enable-libtesseract}.
13131 It accepts the following options:
13135 Set datapath to tesseract data. Default is to use whatever was
13136 set at installation.
13139 Set language, default is "eng".
13142 Set character whitelist.
13145 Set character blacklist.
13148 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
13149 The filter exports confidence of recognized words as the frame metadata @code{lavfi.ocr.confidence}.
13153 Apply a video transform using libopencv.
13155 To enable this filter, install the libopencv library and headers and
13156 configure FFmpeg with @code{--enable-libopencv}.
13158 It accepts the following parameters:
13163 The name of the libopencv filter to apply.
13165 @item filter_params
13166 The parameters to pass to the libopencv filter. If not specified, the default
13167 values are assumed.
13171 Refer to the official libopencv documentation for more precise
13173 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
13175 Several libopencv filters are supported; see the following subsections.
13180 Dilate an image by using a specific structuring element.
13181 It corresponds to the libopencv function @code{cvDilate}.
13183 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
13185 @var{struct_el} represents a structuring element, and has the syntax:
13186 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
13188 @var{cols} and @var{rows} represent the number of columns and rows of
13189 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
13190 point, and @var{shape} the shape for the structuring element. @var{shape}
13191 must be "rect", "cross", "ellipse", or "custom".
13193 If the value for @var{shape} is "custom", it must be followed by a
13194 string of the form "=@var{filename}". The file with name
13195 @var{filename} is assumed to represent a binary image, with each
13196 printable character corresponding to a bright pixel. When a custom
13197 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
13198 or columns and rows of the read file are assumed instead.
13200 The default value for @var{struct_el} is "3x3+0x0/rect".
13202 @var{nb_iterations} specifies the number of times the transform is
13203 applied to the image, and defaults to 1.
13207 # Use the default values
13210 # Dilate using a structuring element with a 5x5 cross, iterating two times
13211 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
13213 # Read the shape from the file diamond.shape, iterating two times.
13214 # The file diamond.shape may contain a pattern of characters like this
13220 # The specified columns and rows are ignored
13221 # but the anchor point coordinates are not
13222 ocv=dilate:0x0+2x2/custom=diamond.shape|2
13227 Erode an image by using a specific structuring element.
13228 It corresponds to the libopencv function @code{cvErode}.
13230 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
13231 with the same syntax and semantics as the @ref{dilate} filter.
13235 Smooth the input video.
13237 The filter takes the following parameters:
13238 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
13240 @var{type} is the type of smooth filter to apply, and must be one of
13241 the following values: "blur", "blur_no_scale", "median", "gaussian",
13242 or "bilateral". The default value is "gaussian".
13244 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
13245 depend on the smooth type. @var{param1} and
13246 @var{param2} accept integer positive values or 0. @var{param3} and
13247 @var{param4} accept floating point values.
13249 The default value for @var{param1} is 3. The default value for the
13250 other parameters is 0.
13252 These parameters correspond to the parameters assigned to the
13253 libopencv function @code{cvSmooth}.
13255 @section oscilloscope
13257 2D Video Oscilloscope.
13259 Useful to measure spatial impulse, step responses, chroma delays, etc.
13261 It accepts the following parameters:
13265 Set scope center x position.
13268 Set scope center y position.
13271 Set scope size, relative to frame diagonal.
13274 Set scope tilt/rotation.
13280 Set trace center x position.
13283 Set trace center y position.
13286 Set trace width, relative to width of frame.
13289 Set trace height, relative to height of frame.
13292 Set which components to trace. By default it traces first three components.
13295 Draw trace grid. By default is enabled.
13298 Draw some statistics. By default is enabled.
13301 Draw scope. By default is enabled.
13304 @subsection Examples
13308 Inspect full first row of video frame.
13310 oscilloscope=x=0.5:y=0:s=1
13314 Inspect full last row of video frame.
13316 oscilloscope=x=0.5:y=1:s=1
13320 Inspect full 5th line of video frame of height 1080.
13322 oscilloscope=x=0.5:y=5/1080:s=1
13326 Inspect full last column of video frame.
13328 oscilloscope=x=1:y=0.5:s=1:t=1
13336 Overlay one video on top of another.
13338 It takes two inputs and has one output. The first input is the "main"
13339 video on which the second input is overlaid.
13341 It accepts the following parameters:
13343 A description of the accepted options follows.
13348 Set the expression for the x and y coordinates of the overlaid video
13349 on the main video. Default value is "0" for both expressions. In case
13350 the expression is invalid, it is set to a huge value (meaning that the
13351 overlay will not be displayed within the output visible area).
13354 See @ref{framesync}.
13357 Set when the expressions for @option{x}, and @option{y} are evaluated.
13359 It accepts the following values:
13362 only evaluate expressions once during the filter initialization or
13363 when a command is processed
13366 evaluate expressions for each incoming frame
13369 Default value is @samp{frame}.
13372 See @ref{framesync}.
13375 Set the format for the output video.
13377 It accepts the following values:
13380 force YUV420 output
13383 force YUV422 output
13386 force YUV444 output
13389 force packed RGB output
13392 force planar RGB output
13395 automatically pick format
13398 Default value is @samp{yuv420}.
13401 See @ref{framesync}.
13404 Set format of alpha of the overlaid video, it can be @var{straight} or
13405 @var{premultiplied}. Default is @var{straight}.
13408 The @option{x}, and @option{y} expressions can contain the following
13414 The main input width and height.
13418 The overlay input width and height.
13422 The computed values for @var{x} and @var{y}. They are evaluated for
13427 horizontal and vertical chroma subsample values of the output
13428 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
13432 the number of input frame, starting from 0
13435 the position in the file of the input frame, NAN if unknown
13438 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
13442 This filter also supports the @ref{framesync} options.
13444 Note that the @var{n}, @var{pos}, @var{t} variables are available only
13445 when evaluation is done @emph{per frame}, and will evaluate to NAN
13446 when @option{eval} is set to @samp{init}.
13448 Be aware that frames are taken from each input video in timestamp
13449 order, hence, if their initial timestamps differ, it is a good idea
13450 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
13451 have them begin in the same zero timestamp, as the example for
13452 the @var{movie} filter does.
13454 You can chain together more overlays but you should test the
13455 efficiency of such approach.
13457 @subsection Commands
13459 This filter supports the following commands:
13463 Modify the x and y of the overlay input.
13464 The command accepts the same syntax of the corresponding option.
13466 If the specified expression is not valid, it is kept at its current
13470 @subsection Examples
13474 Draw the overlay at 10 pixels from the bottom right corner of the main
13477 overlay=main_w-overlay_w-10:main_h-overlay_h-10
13480 Using named options the example above becomes:
13482 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
13486 Insert a transparent PNG logo in the bottom left corner of the input,
13487 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
13489 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
13493 Insert 2 different transparent PNG logos (second logo on bottom
13494 right corner) using the @command{ffmpeg} tool:
13496 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
13500 Add a transparent color layer on top of the main video; @code{WxH}
13501 must specify the size of the main input to the overlay filter:
13503 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
13507 Play an original video and a filtered version (here with the deshake
13508 filter) side by side using the @command{ffplay} tool:
13510 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13513 The above command is the same as:
13515 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13519 Make a sliding overlay appearing from the left to the right top part of the
13520 screen starting since time 2:
13522 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13526 Compose output by putting two input videos side to side:
13528 ffmpeg -i left.avi -i right.avi -filter_complex "
13529 nullsrc=size=200x100 [background];
13530 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13531 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13532 [background][left] overlay=shortest=1 [background+left];
13533 [background+left][right] overlay=shortest=1:x=100 [left+right]
13538 Mask 10-20 seconds of a video by applying the delogo filter to a section
13540 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13541 -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]'
13546 Chain several overlays in cascade:
13548 nullsrc=s=200x200 [bg];
13549 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13550 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13551 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13552 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13553 [in3] null, [mid2] overlay=100:100 [out0]
13560 Apply Overcomplete Wavelet denoiser.
13562 The filter accepts the following options:
13568 Larger depth values will denoise lower frequency components more, but
13569 slow down filtering.
13571 Must be an int in the range 8-16, default is @code{8}.
13573 @item luma_strength, ls
13576 Must be a double value in the range 0-1000, default is @code{1.0}.
13578 @item chroma_strength, cs
13579 Set chroma strength.
13581 Must be a double value in the range 0-1000, default is @code{1.0}.
13587 Add paddings to the input image, and place the original input at the
13588 provided @var{x}, @var{y} coordinates.
13590 It accepts the following parameters:
13595 Specify an expression for the size of the output image with the
13596 paddings added. If the value for @var{width} or @var{height} is 0, the
13597 corresponding input size is used for the output.
13599 The @var{width} expression can reference the value set by the
13600 @var{height} expression, and vice versa.
13602 The default value of @var{width} and @var{height} is 0.
13606 Specify the offsets to place the input image at within the padded area,
13607 with respect to the top/left border of the output image.
13609 The @var{x} expression can reference the value set by the @var{y}
13610 expression, and vice versa.
13612 The default value of @var{x} and @var{y} is 0.
13614 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13615 so the input image is centered on the padded area.
13618 Specify the color of the padded area. For the syntax of this option,
13619 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13620 manual,ffmpeg-utils}.
13622 The default value of @var{color} is "black".
13625 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13627 It accepts the following values:
13631 Only evaluate expressions once during the filter initialization or when
13632 a command is processed.
13635 Evaluate expressions for each incoming frame.
13639 Default value is @samp{init}.
13642 Pad to aspect instead to a resolution.
13646 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13647 options are expressions containing the following constants:
13652 The input video width and height.
13656 These are the same as @var{in_w} and @var{in_h}.
13660 The output width and height (the size of the padded area), as
13661 specified by the @var{width} and @var{height} expressions.
13665 These are the same as @var{out_w} and @var{out_h}.
13669 The x and y offsets as specified by the @var{x} and @var{y}
13670 expressions, or NAN if not yet specified.
13673 same as @var{iw} / @var{ih}
13676 input sample aspect ratio
13679 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13683 The horizontal and vertical chroma subsample values. For example for the
13684 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13687 @subsection Examples
13691 Add paddings with the color "violet" to the input video. The output video
13692 size is 640x480, and the top-left corner of the input video is placed at
13695 pad=640:480:0:40:violet
13698 The example above is equivalent to the following command:
13700 pad=width=640:height=480:x=0:y=40:color=violet
13704 Pad the input to get an output with dimensions increased by 3/2,
13705 and put the input video at the center of the padded area:
13707 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13711 Pad the input to get a squared output with size equal to the maximum
13712 value between the input width and height, and put the input video at
13713 the center of the padded area:
13715 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13719 Pad the input to get a final w/h ratio of 16:9:
13721 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13725 In case of anamorphic video, in order to set the output display aspect
13726 correctly, it is necessary to use @var{sar} in the expression,
13727 according to the relation:
13729 (ih * X / ih) * sar = output_dar
13730 X = output_dar / sar
13733 Thus the previous example needs to be modified to:
13735 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13739 Double the output size and put the input video in the bottom-right
13740 corner of the output padded area:
13742 pad="2*iw:2*ih:ow-iw:oh-ih"
13746 @anchor{palettegen}
13747 @section palettegen
13749 Generate one palette for a whole video stream.
13751 It accepts the following options:
13755 Set the maximum number of colors to quantize in the palette.
13756 Note: the palette will still contain 256 colors; the unused palette entries
13759 @item reserve_transparent
13760 Create a palette of 255 colors maximum and reserve the last one for
13761 transparency. Reserving the transparency color is useful for GIF optimization.
13762 If not set, the maximum of colors in the palette will be 256. You probably want
13763 to disable this option for a standalone image.
13766 @item transparency_color
13767 Set the color that will be used as background for transparency.
13770 Set statistics mode.
13772 It accepts the following values:
13775 Compute full frame histograms.
13777 Compute histograms only for the part that differs from previous frame. This
13778 might be relevant to give more importance to the moving part of your input if
13779 the background is static.
13781 Compute new histogram for each frame.
13784 Default value is @var{full}.
13787 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13788 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13789 color quantization of the palette. This information is also visible at
13790 @var{info} logging level.
13792 @subsection Examples
13796 Generate a representative palette of a given video using @command{ffmpeg}:
13798 ffmpeg -i input.mkv -vf palettegen palette.png
13802 @section paletteuse
13804 Use a palette to downsample an input video stream.
13806 The filter takes two inputs: one video stream and a palette. The palette must
13807 be a 256 pixels image.
13809 It accepts the following options:
13813 Select dithering mode. Available algorithms are:
13816 Ordered 8x8 bayer dithering (deterministic)
13818 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13819 Note: this dithering is sometimes considered "wrong" and is included as a
13821 @item floyd_steinberg
13822 Floyd and Steingberg dithering (error diffusion)
13824 Frankie Sierra dithering v2 (error diffusion)
13826 Frankie Sierra dithering v2 "Lite" (error diffusion)
13829 Default is @var{sierra2_4a}.
13832 When @var{bayer} dithering is selected, this option defines the scale of the
13833 pattern (how much the crosshatch pattern is visible). A low value means more
13834 visible pattern for less banding, and higher value means less visible pattern
13835 at the cost of more banding.
13837 The option must be an integer value in the range [0,5]. Default is @var{2}.
13840 If set, define the zone to process
13844 Only the changing rectangle will be reprocessed. This is similar to GIF
13845 cropping/offsetting compression mechanism. This option can be useful for speed
13846 if only a part of the image is changing, and has use cases such as limiting the
13847 scope of the error diffusal @option{dither} to the rectangle that bounds the
13848 moving scene (it leads to more deterministic output if the scene doesn't change
13849 much, and as a result less moving noise and better GIF compression).
13852 Default is @var{none}.
13855 Take new palette for each output frame.
13857 @item alpha_threshold
13858 Sets the alpha threshold for transparency. Alpha values above this threshold
13859 will be treated as completely opaque, and values below this threshold will be
13860 treated as completely transparent.
13862 The option must be an integer value in the range [0,255]. Default is @var{128}.
13865 @subsection Examples
13869 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13870 using @command{ffmpeg}:
13872 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13876 @section perspective
13878 Correct perspective of video not recorded perpendicular to the screen.
13880 A description of the accepted parameters follows.
13891 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13892 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13893 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13894 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13895 then the corners of the source will be sent to the specified coordinates.
13897 The expressions can use the following variables:
13902 the width and height of video frame.
13906 Output frame count.
13909 @item interpolation
13910 Set interpolation for perspective correction.
13912 It accepts the following values:
13918 Default value is @samp{linear}.
13921 Set interpretation of coordinate options.
13923 It accepts the following values:
13927 Send point in the source specified by the given coordinates to
13928 the corners of the destination.
13930 @item 1, destination
13932 Send the corners of the source to the point in the destination specified
13933 by the given coordinates.
13935 Default value is @samp{source}.
13939 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13941 It accepts the following values:
13944 only evaluate expressions once during the filter initialization or
13945 when a command is processed
13948 evaluate expressions for each incoming frame
13951 Default value is @samp{init}.
13956 Delay interlaced video by one field time so that the field order changes.
13958 The intended use is to fix PAL movies that have been captured with the
13959 opposite field order to the film-to-video transfer.
13961 A description of the accepted parameters follows.
13967 It accepts the following values:
13970 Capture field order top-first, transfer bottom-first.
13971 Filter will delay the bottom field.
13974 Capture field order bottom-first, transfer top-first.
13975 Filter will delay the top field.
13978 Capture and transfer with the same field order. This mode only exists
13979 for the documentation of the other options to refer to, but if you
13980 actually select it, the filter will faithfully do nothing.
13983 Capture field order determined automatically by field flags, transfer
13985 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13986 basis using field flags. If no field information is available,
13987 then this works just like @samp{u}.
13990 Capture unknown or varying, transfer opposite.
13991 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13992 analyzing the images and selecting the alternative that produces best
13993 match between the fields.
13996 Capture top-first, transfer unknown or varying.
13997 Filter selects among @samp{t} and @samp{p} using image analysis.
14000 Capture bottom-first, transfer unknown or varying.
14001 Filter selects among @samp{b} and @samp{p} using image analysis.
14004 Capture determined by field flags, transfer unknown or varying.
14005 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
14006 image analysis. If no field information is available, then this works just
14007 like @samp{U}. This is the default mode.
14010 Both capture and transfer unknown or varying.
14011 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
14015 @section pixdesctest
14017 Pixel format descriptor test filter, mainly useful for internal
14018 testing. The output video should be equal to the input video.
14022 format=monow, pixdesctest
14025 can be used to test the monowhite pixel format descriptor definition.
14029 Display sample values of color channels. Mainly useful for checking color
14030 and levels. Minimum supported resolution is 640x480.
14032 The filters accept the following options:
14036 Set scope X position, relative offset on X axis.
14039 Set scope Y position, relative offset on Y axis.
14048 Set window opacity. This window also holds statistics about pixel area.
14051 Set window X position, relative offset on X axis.
14054 Set window Y position, relative offset on Y axis.
14059 Enable the specified chain of postprocessing subfilters using libpostproc. This
14060 library should be automatically selected with a GPL build (@code{--enable-gpl}).
14061 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
14062 Each subfilter and some options have a short and a long name that can be used
14063 interchangeably, i.e. dr/dering are the same.
14065 The filters accept the following options:
14069 Set postprocessing subfilters string.
14072 All subfilters share common options to determine their scope:
14076 Honor the quality commands for this subfilter.
14079 Do chrominance filtering, too (default).
14082 Do luminance filtering only (no chrominance).
14085 Do chrominance filtering only (no luminance).
14088 These options can be appended after the subfilter name, separated by a '|'.
14090 Available subfilters are:
14093 @item hb/hdeblock[|difference[|flatness]]
14094 Horizontal deblocking filter
14097 Difference factor where higher values mean more deblocking (default: @code{32}).
14099 Flatness threshold where lower values mean more deblocking (default: @code{39}).
14102 @item vb/vdeblock[|difference[|flatness]]
14103 Vertical deblocking filter
14106 Difference factor where higher values mean more deblocking (default: @code{32}).
14108 Flatness threshold where lower values mean more deblocking (default: @code{39}).
14111 @item ha/hadeblock[|difference[|flatness]]
14112 Accurate horizontal deblocking filter
14115 Difference factor where higher values mean more deblocking (default: @code{32}).
14117 Flatness threshold where lower values mean more deblocking (default: @code{39}).
14120 @item va/vadeblock[|difference[|flatness]]
14121 Accurate vertical deblocking filter
14124 Difference factor where higher values mean more deblocking (default: @code{32}).
14126 Flatness threshold where lower values mean more deblocking (default: @code{39}).
14130 The horizontal and vertical deblocking filters share the difference and
14131 flatness values so you cannot set different horizontal and vertical
14135 @item h1/x1hdeblock
14136 Experimental horizontal deblocking filter
14138 @item v1/x1vdeblock
14139 Experimental vertical deblocking filter
14144 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
14147 larger -> stronger filtering
14149 larger -> stronger filtering
14151 larger -> stronger filtering
14154 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
14157 Stretch luminance to @code{0-255}.
14160 @item lb/linblenddeint
14161 Linear blend deinterlacing filter that deinterlaces the given block by
14162 filtering all lines with a @code{(1 2 1)} filter.
14164 @item li/linipoldeint
14165 Linear interpolating deinterlacing filter that deinterlaces the given block by
14166 linearly interpolating every second line.
14168 @item ci/cubicipoldeint
14169 Cubic interpolating deinterlacing filter deinterlaces the given block by
14170 cubically interpolating every second line.
14172 @item md/mediandeint
14173 Median deinterlacing filter that deinterlaces the given block by applying a
14174 median filter to every second line.
14176 @item fd/ffmpegdeint
14177 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
14178 second line with a @code{(-1 4 2 4 -1)} filter.
14181 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
14182 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
14184 @item fq/forceQuant[|quantizer]
14185 Overrides the quantizer table from the input with the constant quantizer you
14193 Default pp filter combination (@code{hb|a,vb|a,dr|a})
14196 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
14199 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
14202 @subsection Examples
14206 Apply horizontal and vertical deblocking, deringing and automatic
14207 brightness/contrast:
14213 Apply default filters without brightness/contrast correction:
14219 Apply default filters and temporal denoiser:
14221 pp=default/tmpnoise|1|2|3
14225 Apply deblocking on luminance only, and switch vertical deblocking on or off
14226 automatically depending on available CPU time:
14233 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
14234 similar to spp = 6 with 7 point DCT, where only the center sample is
14237 The filter accepts the following options:
14241 Force a constant quantization parameter. It accepts an integer in range
14242 0 to 63. If not set, the filter will use the QP from the video stream
14246 Set thresholding mode. Available modes are:
14250 Set hard thresholding.
14252 Set soft thresholding (better de-ringing effect, but likely blurrier).
14254 Set medium thresholding (good results, default).
14258 @section premultiply
14259 Apply alpha premultiply effect to input video stream using first plane
14260 of second stream as alpha.
14262 Both streams must have same dimensions and same pixel format.
14264 The filter accepts the following option:
14268 Set which planes will be processed, unprocessed planes will be copied.
14269 By default value 0xf, all planes will be processed.
14272 Do not require 2nd input for processing, instead use alpha plane from input stream.
14276 Apply prewitt operator to input video stream.
14278 The filter accepts the following option:
14282 Set which planes will be processed, unprocessed planes will be copied.
14283 By default value 0xf, all planes will be processed.
14286 Set value which will be multiplied with filtered result.
14289 Set value which will be added to filtered result.
14292 @anchor{program_opencl}
14293 @section program_opencl
14295 Filter video using an OpenCL program.
14300 OpenCL program source file.
14303 Kernel name in program.
14306 Number of inputs to the filter. Defaults to 1.
14309 Size of output frames. Defaults to the same as the first input.
14313 The program source file must contain a kernel function with the given name,
14314 which will be run once for each plane of the output. Each run on a plane
14315 gets enqueued as a separate 2D global NDRange with one work-item for each
14316 pixel to be generated. The global ID offset for each work-item is therefore
14317 the coordinates of a pixel in the destination image.
14319 The kernel function needs to take the following arguments:
14322 Destination image, @var{__write_only image2d_t}.
14324 This image will become the output; the kernel should write all of it.
14326 Frame index, @var{unsigned int}.
14328 This is a counter starting from zero and increasing by one for each frame.
14330 Source images, @var{__read_only image2d_t}.
14332 These are the most recent images on each input. The kernel may read from
14333 them to generate the output, but they can't be written to.
14340 Copy the input to the output (output must be the same size as the input).
14342 __kernel void copy(__write_only image2d_t destination,
14343 unsigned int index,
14344 __read_only image2d_t source)
14346 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
14348 int2 location = (int2)(get_global_id(0), get_global_id(1));
14350 float4 value = read_imagef(source, sampler, location);
14352 write_imagef(destination, location, value);
14357 Apply a simple transformation, rotating the input by an amount increasing
14358 with the index counter. Pixel values are linearly interpolated by the
14359 sampler, and the output need not have the same dimensions as the input.
14361 __kernel void rotate_image(__write_only image2d_t dst,
14362 unsigned int index,
14363 __read_only image2d_t src)
14365 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
14366 CLK_FILTER_LINEAR);
14368 float angle = (float)index / 100.0f;
14370 float2 dst_dim = convert_float2(get_image_dim(dst));
14371 float2 src_dim = convert_float2(get_image_dim(src));
14373 float2 dst_cen = dst_dim / 2.0f;
14374 float2 src_cen = src_dim / 2.0f;
14376 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
14378 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
14380 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
14381 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
14383 src_pos = src_pos * src_dim / dst_dim;
14385 float2 src_loc = src_pos + src_cen;
14387 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
14388 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
14389 write_imagef(dst, dst_loc, 0.5f);
14391 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
14396 Blend two inputs together, with the amount of each input used varying
14397 with the index counter.
14399 __kernel void blend_images(__write_only image2d_t dst,
14400 unsigned int index,
14401 __read_only image2d_t src1,
14402 __read_only image2d_t src2)
14404 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
14405 CLK_FILTER_LINEAR);
14407 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
14409 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
14410 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
14411 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
14413 float4 val1 = read_imagef(src1, sampler, src1_loc);
14414 float4 val2 = read_imagef(src2, sampler, src2_loc);
14416 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
14422 @section pseudocolor
14424 Alter frame colors in video with pseudocolors.
14426 This filter accept the following options:
14430 set pixel first component expression
14433 set pixel second component expression
14436 set pixel third component expression
14439 set pixel fourth component expression, corresponds to the alpha component
14442 set component to use as base for altering colors
14445 Each of them specifies the expression to use for computing the lookup table for
14446 the corresponding pixel component values.
14448 The expressions can contain the following constants and functions:
14453 The input width and height.
14456 The input value for the pixel component.
14458 @item ymin, umin, vmin, amin
14459 The minimum allowed component value.
14461 @item ymax, umax, vmax, amax
14462 The maximum allowed component value.
14465 All expressions default to "val".
14467 @subsection Examples
14471 Change too high luma values to gradient:
14473 pseudocolor="'if(between(val,ymax,amax),lerp(ymin,ymax,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(umax,umin,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(vmin,vmax,(val-ymax)/(amax-ymax)),-1):-1'"
14479 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
14480 Ratio) between two input videos.
14482 This filter takes in input two input videos, the first input is
14483 considered the "main" source and is passed unchanged to the
14484 output. The second input is used as a "reference" video for computing
14487 Both video inputs must have the same resolution and pixel format for
14488 this filter to work correctly. Also it assumes that both inputs
14489 have the same number of frames, which are compared one by one.
14491 The obtained average PSNR is printed through the logging system.
14493 The filter stores the accumulated MSE (mean squared error) of each
14494 frame, and at the end of the processing it is averaged across all frames
14495 equally, and the following formula is applied to obtain the PSNR:
14498 PSNR = 10*log10(MAX^2/MSE)
14501 Where MAX is the average of the maximum values of each component of the
14504 The description of the accepted parameters follows.
14507 @item stats_file, f
14508 If specified the filter will use the named file to save the PSNR of
14509 each individual frame. When filename equals "-" the data is sent to
14512 @item stats_version
14513 Specifies which version of the stats file format to use. Details of
14514 each format are written below.
14515 Default value is 1.
14517 @item stats_add_max
14518 Determines whether the max value is output to the stats log.
14519 Default value is 0.
14520 Requires stats_version >= 2. If this is set and stats_version < 2,
14521 the filter will return an error.
14524 This filter also supports the @ref{framesync} options.
14526 The file printed if @var{stats_file} is selected, contains a sequence of
14527 key/value pairs of the form @var{key}:@var{value} for each compared
14530 If a @var{stats_version} greater than 1 is specified, a header line precedes
14531 the list of per-frame-pair stats, with key value pairs following the frame
14532 format with the following parameters:
14535 @item psnr_log_version
14536 The version of the log file format. Will match @var{stats_version}.
14539 A comma separated list of the per-frame-pair parameters included in
14543 A description of each shown per-frame-pair parameter follows:
14547 sequential number of the input frame, starting from 1
14550 Mean Square Error pixel-by-pixel average difference of the compared
14551 frames, averaged over all the image components.
14553 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14554 Mean Square Error pixel-by-pixel average difference of the compared
14555 frames for the component specified by the suffix.
14557 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14558 Peak Signal to Noise ratio of the compared frames for the component
14559 specified by the suffix.
14561 @item max_avg, max_y, max_u, max_v
14562 Maximum allowed value for each channel, and average over all
14568 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14569 [main][ref] psnr="stats_file=stats.log" [out]
14572 On this example the input file being processed is compared with the
14573 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14574 is stored in @file{stats.log}.
14579 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14580 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14583 The pullup filter is designed to take advantage of future context in making
14584 its decisions. This filter is stateless in the sense that it does not lock
14585 onto a pattern to follow, but it instead looks forward to the following
14586 fields in order to identify matches and rebuild progressive frames.
14588 To produce content with an even framerate, insert the fps filter after
14589 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14590 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14592 The filter accepts the following options:
14599 These options set the amount of "junk" to ignore at the left, right, top, and
14600 bottom of the image, respectively. Left and right are in units of 8 pixels,
14601 while top and bottom are in units of 2 lines.
14602 The default is 8 pixels on each side.
14605 Set the strict breaks. Setting this option to 1 will reduce the chances of
14606 filter generating an occasional mismatched frame, but it may also cause an
14607 excessive number of frames to be dropped during high motion sequences.
14608 Conversely, setting it to -1 will make filter match fields more easily.
14609 This may help processing of video where there is slight blurring between
14610 the fields, but may also cause there to be interlaced frames in the output.
14611 Default value is @code{0}.
14614 Set the metric plane to use. It accepts the following values:
14620 Use chroma blue plane.
14623 Use chroma red plane.
14626 This option may be set to use chroma plane instead of the default luma plane
14627 for doing filter's computations. This may improve accuracy on very clean
14628 source material, but more likely will decrease accuracy, especially if there
14629 is chroma noise (rainbow effect) or any grayscale video.
14630 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14631 load and make pullup usable in realtime on slow machines.
14634 For best results (without duplicated frames in the output file) it is
14635 necessary to change the output frame rate. For example, to inverse
14636 telecine NTSC input:
14638 ffmpeg -i input -vf pullup -r 24000/1001 ...
14643 Change video quantization parameters (QP).
14645 The filter accepts the following option:
14649 Set expression for quantization parameter.
14652 The expression is evaluated through the eval API and can contain, among others,
14653 the following constants:
14657 1 if index is not 129, 0 otherwise.
14660 Sequential index starting from -129 to 128.
14663 @subsection Examples
14667 Some equation like:
14675 Flush video frames from internal cache of frames into a random order.
14676 No frame is discarded.
14677 Inspired by @ref{frei0r} nervous filter.
14681 Set size in number of frames of internal cache, in range from @code{2} to
14682 @code{512}. Default is @code{30}.
14685 Set seed for random number generator, must be an integer included between
14686 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14687 less than @code{0}, the filter will try to use a good random seed on a
14691 @section readeia608
14693 Read closed captioning (EIA-608) information from the top lines of a video frame.
14695 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14696 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14697 with EIA-608 data (starting from 0). A description of each metadata value follows:
14700 @item lavfi.readeia608.X.cc
14701 The two bytes stored as EIA-608 data (printed in hexadecimal).
14703 @item lavfi.readeia608.X.line
14704 The number of the line on which the EIA-608 data was identified and read.
14707 This filter accepts the following options:
14711 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14714 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14717 Set minimal acceptable amplitude change for sync codes detection.
14718 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14721 Set the ratio of width reserved for sync code detection.
14722 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14725 Set the max peaks height difference for sync code detection.
14726 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14729 Set max peaks period difference for sync code detection.
14730 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14733 Set the first two max start code bits differences.
14734 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14737 Set the minimum ratio of bits height compared to 3rd start code bit.
14738 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14741 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14744 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14747 Enable checking the parity bit. In the event of a parity error, the filter will output
14748 @code{0x00} for that character. Default is false.
14751 Lowpass lines prior further proccessing. Default is disabled.
14754 @subsection Examples
14758 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14760 ffprobe -f lavfi -i movie=captioned_video.mov,readeia608 -show_entries frame=pkt_pts_time:frame_tags=lavfi.readeia608.0.cc,lavfi.readeia608.1.cc -of csv
14766 Read vertical interval timecode (VITC) information from the top lines of a
14769 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14770 timecode value, if a valid timecode has been detected. Further metadata key
14771 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14772 timecode data has been found or not.
14774 This filter accepts the following options:
14778 Set the maximum number of lines to scan for VITC data. If the value is set to
14779 @code{-1} the full video frame is scanned. Default is @code{45}.
14782 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14783 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14786 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14787 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14790 @subsection Examples
14794 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14795 draw @code{--:--:--:--} as a placeholder:
14797 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14803 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14805 Destination pixel at position (X, Y) will be picked from source (x, y) position
14806 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14807 value for pixel will be used for destination pixel.
14809 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14810 will have Xmap/Ymap video stream dimensions.
14811 Xmap and Ymap input video streams are 16bit depth, single channel.
14815 Specify pixel format of output from this filter. Can be @code{color} or @code{gray}.
14816 Default is @code{color}.
14819 @section removegrain
14821 The removegrain filter is a spatial denoiser for progressive video.
14825 Set mode for the first plane.
14828 Set mode for the second plane.
14831 Set mode for the third plane.
14834 Set mode for the fourth plane.
14837 Range of mode is from 0 to 24. Description of each mode follows:
14841 Leave input plane unchanged. Default.
14844 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14847 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14850 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14853 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14854 This is equivalent to a median filter.
14857 Line-sensitive clipping giving the minimal change.
14860 Line-sensitive clipping, intermediate.
14863 Line-sensitive clipping, intermediate.
14866 Line-sensitive clipping, intermediate.
14869 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14872 Replaces the target pixel with the closest neighbour.
14875 [1 2 1] horizontal and vertical kernel blur.
14881 Bob mode, interpolates top field from the line where the neighbours
14882 pixels are the closest.
14885 Bob mode, interpolates bottom field from the line where the neighbours
14886 pixels are the closest.
14889 Bob mode, interpolates top field. Same as 13 but with a more complicated
14890 interpolation formula.
14893 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14894 interpolation formula.
14897 Clips the pixel with the minimum and maximum of respectively the maximum and
14898 minimum of each pair of opposite neighbour pixels.
14901 Line-sensitive clipping using opposite neighbours whose greatest distance from
14902 the current pixel is minimal.
14905 Replaces the pixel with the average of its 8 neighbours.
14908 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14911 Clips pixels using the averages of opposite neighbour.
14914 Same as mode 21 but simpler and faster.
14917 Small edge and halo removal, but reputed useless.
14923 @section removelogo
14925 Suppress a TV station logo, using an image file to determine which
14926 pixels comprise the logo. It works by filling in the pixels that
14927 comprise the logo with neighboring pixels.
14929 The filter accepts the following options:
14933 Set the filter bitmap file, which can be any image format supported by
14934 libavformat. The width and height of the image file must match those of the
14935 video stream being processed.
14938 Pixels in the provided bitmap image with a value of zero are not
14939 considered part of the logo, non-zero pixels are considered part of
14940 the logo. If you use white (255) for the logo and black (0) for the
14941 rest, you will be safe. For making the filter bitmap, it is
14942 recommended to take a screen capture of a black frame with the logo
14943 visible, and then using a threshold filter followed by the erode
14944 filter once or twice.
14946 If needed, little splotches can be fixed manually. Remember that if
14947 logo pixels are not covered, the filter quality will be much
14948 reduced. Marking too many pixels as part of the logo does not hurt as
14949 much, but it will increase the amount of blurring needed to cover over
14950 the image and will destroy more information than necessary, and extra
14951 pixels will slow things down on a large logo.
14953 @section repeatfields
14955 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14956 fields based on its value.
14960 Reverse a video clip.
14962 Warning: This filter requires memory to buffer the entire clip, so trimming
14965 @subsection Examples
14969 Take the first 5 seconds of a clip, and reverse it.
14976 Shift R/G/B/A pixels horizontally and/or vertically.
14978 The filter accepts the following options:
14981 Set amount to shift red horizontally.
14983 Set amount to shift red vertically.
14985 Set amount to shift green horizontally.
14987 Set amount to shift green vertically.
14989 Set amount to shift blue horizontally.
14991 Set amount to shift blue vertically.
14993 Set amount to shift alpha horizontally.
14995 Set amount to shift alpha vertically.
14997 Set edge mode, can be @var{smear}, default, or @var{warp}.
15001 Apply roberts cross operator to input video stream.
15003 The filter accepts the following option:
15007 Set which planes will be processed, unprocessed planes will be copied.
15008 By default value 0xf, all planes will be processed.
15011 Set value which will be multiplied with filtered result.
15014 Set value which will be added to filtered result.
15019 Rotate video by an arbitrary angle expressed in radians.
15021 The filter accepts the following options:
15023 A description of the optional parameters follows.
15026 Set an expression for the angle by which to rotate the input video
15027 clockwise, expressed as a number of radians. A negative value will
15028 result in a counter-clockwise rotation. By default it is set to "0".
15030 This expression is evaluated for each frame.
15033 Set the output width expression, default value is "iw".
15034 This expression is evaluated just once during configuration.
15037 Set the output height expression, default value is "ih".
15038 This expression is evaluated just once during configuration.
15041 Enable bilinear interpolation if set to 1, a value of 0 disables
15042 it. Default value is 1.
15045 Set the color used to fill the output area not covered by the rotated
15046 image. For the general syntax of this option, check the
15047 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
15048 If the special value "none" is selected then no
15049 background is printed (useful for example if the background is never shown).
15051 Default value is "black".
15054 The expressions for the angle and the output size can contain the
15055 following constants and functions:
15059 sequential number of the input frame, starting from 0. It is always NAN
15060 before the first frame is filtered.
15063 time in seconds of the input frame, it is set to 0 when the filter is
15064 configured. It is always NAN before the first frame is filtered.
15068 horizontal and vertical chroma subsample values. For example for the
15069 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15073 the input video width and height
15077 the output width and height, that is the size of the padded area as
15078 specified by the @var{width} and @var{height} expressions
15082 the minimal width/height required for completely containing the input
15083 video rotated by @var{a} radians.
15085 These are only available when computing the @option{out_w} and
15086 @option{out_h} expressions.
15089 @subsection Examples
15093 Rotate the input by PI/6 radians clockwise:
15099 Rotate the input by PI/6 radians counter-clockwise:
15105 Rotate the input by 45 degrees clockwise:
15111 Apply a constant rotation with period T, starting from an angle of PI/3:
15113 rotate=PI/3+2*PI*t/T
15117 Make the input video rotation oscillating with a period of T
15118 seconds and an amplitude of A radians:
15120 rotate=A*sin(2*PI/T*t)
15124 Rotate the video, output size is chosen so that the whole rotating
15125 input video is always completely contained in the output:
15127 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
15131 Rotate the video, reduce the output size so that no background is ever
15134 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
15138 @subsection Commands
15140 The filter supports the following commands:
15144 Set the angle expression.
15145 The command accepts the same syntax of the corresponding option.
15147 If the specified expression is not valid, it is kept at its current
15153 Apply Shape Adaptive Blur.
15155 The filter accepts the following options:
15158 @item luma_radius, lr
15159 Set luma blur filter strength, must be a value in range 0.1-4.0, default
15160 value is 1.0. A greater value will result in a more blurred image, and
15161 in slower processing.
15163 @item luma_pre_filter_radius, lpfr
15164 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
15167 @item luma_strength, ls
15168 Set luma maximum difference between pixels to still be considered, must
15169 be a value in the 0.1-100.0 range, default value is 1.0.
15171 @item chroma_radius, cr
15172 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
15173 greater value will result in a more blurred image, and in slower
15176 @item chroma_pre_filter_radius, cpfr
15177 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
15179 @item chroma_strength, cs
15180 Set chroma maximum difference between pixels to still be considered,
15181 must be a value in the -0.9-100.0 range.
15184 Each chroma option value, if not explicitly specified, is set to the
15185 corresponding luma option value.
15190 Scale (resize) the input video, using the libswscale library.
15192 The scale filter forces the output display aspect ratio to be the same
15193 of the input, by changing the output sample aspect ratio.
15195 If the input image format is different from the format requested by
15196 the next filter, the scale filter will convert the input to the
15199 @subsection Options
15200 The filter accepts the following options, or any of the options
15201 supported by the libswscale scaler.
15203 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
15204 the complete list of scaler options.
15209 Set the output video dimension expression. Default value is the input
15212 If the @var{width} or @var{w} value is 0, the input width is used for
15213 the output. If the @var{height} or @var{h} value is 0, the input height
15214 is used for the output.
15216 If one and only one of the values is -n with n >= 1, the scale filter
15217 will use a value that maintains the aspect ratio of the input image,
15218 calculated from the other specified dimension. After that it will,
15219 however, make sure that the calculated dimension is divisible by n and
15220 adjust the value if necessary.
15222 If both values are -n with n >= 1, the behavior will be identical to
15223 both values being set to 0 as previously detailed.
15225 See below for the list of accepted constants for use in the dimension
15229 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
15233 Only evaluate expressions once during the filter initialization or when a command is processed.
15236 Evaluate expressions for each incoming frame.
15240 Default value is @samp{init}.
15244 Set the interlacing mode. It accepts the following values:
15248 Force interlaced aware scaling.
15251 Do not apply interlaced scaling.
15254 Select interlaced aware scaling depending on whether the source frames
15255 are flagged as interlaced or not.
15258 Default value is @samp{0}.
15261 Set libswscale scaling flags. See
15262 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
15263 complete list of values. If not explicitly specified the filter applies
15267 @item param0, param1
15268 Set libswscale input parameters for scaling algorithms that need them. See
15269 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
15270 complete documentation. If not explicitly specified the filter applies
15276 Set the video size. For the syntax of this option, check the
15277 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15279 @item in_color_matrix
15280 @item out_color_matrix
15281 Set in/output YCbCr color space type.
15283 This allows the autodetected value to be overridden as well as allows forcing
15284 a specific value used for the output and encoder.
15286 If not specified, the color space type depends on the pixel format.
15292 Choose automatically.
15295 Format conforming to International Telecommunication Union (ITU)
15296 Recommendation BT.709.
15299 Set color space conforming to the United States Federal Communications
15300 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
15305 Set color space conforming to:
15309 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
15312 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
15315 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
15320 Set color space conforming to SMPTE ST 240:1999.
15323 Set color space conforming to ITU-R BT.2020 non-constant luminance system.
15328 Set in/output YCbCr sample range.
15330 This allows the autodetected value to be overridden as well as allows forcing
15331 a specific value used for the output and encoder. If not specified, the
15332 range depends on the pixel format. Possible values:
15336 Choose automatically.
15339 Set full range (0-255 in case of 8-bit luma).
15341 @item mpeg/limited/tv
15342 Set "MPEG" range (16-235 in case of 8-bit luma).
15345 @item force_original_aspect_ratio
15346 Enable decreasing or increasing output video width or height if necessary to
15347 keep the original aspect ratio. Possible values:
15351 Scale the video as specified and disable this feature.
15354 The output video dimensions will automatically be decreased if needed.
15357 The output video dimensions will automatically be increased if needed.
15361 One useful instance of this option is that when you know a specific device's
15362 maximum allowed resolution, you can use this to limit the output video to
15363 that, while retaining the aspect ratio. For example, device A allows
15364 1280x720 playback, and your video is 1920x800. Using this option (set it to
15365 decrease) and specifying 1280x720 to the command line makes the output
15368 Please note that this is a different thing than specifying -1 for @option{w}
15369 or @option{h}, you still need to specify the output resolution for this option
15374 The values of the @option{w} and @option{h} options are expressions
15375 containing the following constants:
15380 The input width and height
15384 These are the same as @var{in_w} and @var{in_h}.
15388 The output (scaled) width and height
15392 These are the same as @var{out_w} and @var{out_h}
15395 The same as @var{iw} / @var{ih}
15398 input sample aspect ratio
15401 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
15405 horizontal and vertical input chroma subsample values. For example for the
15406 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15410 horizontal and vertical output chroma subsample values. For example for the
15411 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15414 @subsection Examples
15418 Scale the input video to a size of 200x100
15423 This is equivalent to:
15434 Specify a size abbreviation for the output size:
15439 which can also be written as:
15445 Scale the input to 2x:
15447 scale=w=2*iw:h=2*ih
15451 The above is the same as:
15453 scale=2*in_w:2*in_h
15457 Scale the input to 2x with forced interlaced scaling:
15459 scale=2*iw:2*ih:interl=1
15463 Scale the input to half size:
15465 scale=w=iw/2:h=ih/2
15469 Increase the width, and set the height to the same size:
15475 Seek Greek harmony:
15482 Increase the height, and set the width to 3/2 of the height:
15484 scale=w=3/2*oh:h=3/5*ih
15488 Increase the size, making the size a multiple of the chroma
15491 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
15495 Increase the width to a maximum of 500 pixels,
15496 keeping the same aspect ratio as the input:
15498 scale=w='min(500\, iw*3/2):h=-1'
15502 Make pixels square by combining scale and setsar:
15504 scale='trunc(ih*dar):ih',setsar=1/1
15508 Make pixels square by combining scale and setsar,
15509 making sure the resulting resolution is even (required by some codecs):
15511 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
15515 @subsection Commands
15517 This filter supports the following commands:
15521 Set the output video dimension expression.
15522 The command accepts the same syntax of the corresponding option.
15524 If the specified expression is not valid, it is kept at its current
15530 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15531 format conversion on CUDA video frames. Setting the output width and height
15532 works in the same way as for the @var{scale} filter.
15534 The following additional options are accepted:
15537 The pixel format of the output CUDA frames. If set to the string "same" (the
15538 default), the input format will be kept. Note that automatic format negotiation
15539 and conversion is not yet supported for hardware frames
15542 The interpolation algorithm used for resizing. One of the following:
15549 @item cubic2p_bspline
15550 2-parameter cubic (B=1, C=0)
15552 @item cubic2p_catmullrom
15553 2-parameter cubic (B=0, C=1/2)
15555 @item cubic2p_b05c03
15556 2-parameter cubic (B=1/2, C=3/10)
15568 Scale (resize) the input video, based on a reference video.
15570 See the scale filter for available options, scale2ref supports the same but
15571 uses the reference video instead of the main input as basis. scale2ref also
15572 supports the following additional constants for the @option{w} and
15573 @option{h} options:
15578 The main input video's width and height
15581 The same as @var{main_w} / @var{main_h}
15584 The main input video's sample aspect ratio
15586 @item main_dar, mdar
15587 The main input video's display aspect ratio. Calculated from
15588 @code{(main_w / main_h) * main_sar}.
15592 The main input video's horizontal and vertical chroma subsample values.
15593 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15597 @subsection Examples
15601 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15603 'scale2ref[b][a];[a][b]overlay'
15607 Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
15609 [logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
15613 @anchor{selectivecolor}
15614 @section selectivecolor
15616 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15617 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15618 by the "purity" of the color (that is, how saturated it already is).
15620 This filter is similar to the Adobe Photoshop Selective Color tool.
15622 The filter accepts the following options:
15625 @item correction_method
15626 Select color correction method.
15628 Available values are:
15631 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15634 Specified adjustments are relative to the original component value.
15636 Default is @code{absolute}.
15638 Adjustments for red pixels (pixels where the red component is the maximum)
15640 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15642 Adjustments for green pixels (pixels where the green component is the maximum)
15644 Adjustments for cyan pixels (pixels where the red component is the minimum)
15646 Adjustments for blue pixels (pixels where the blue component is the maximum)
15648 Adjustments for magenta pixels (pixels where the green component is the minimum)
15650 Adjustments for white pixels (pixels where all components are greater than 128)
15652 Adjustments for all pixels except pure black and pure white
15654 Adjustments for black pixels (pixels where all components are lesser than 128)
15656 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15659 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15660 4 space separated floating point adjustment values in the [-1,1] range,
15661 respectively to adjust the amount of cyan, magenta, yellow and black for the
15662 pixels of its range.
15664 @subsection Examples
15668 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15669 increase magenta by 27% in blue areas:
15671 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15675 Use a Photoshop selective color preset:
15677 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15681 @anchor{separatefields}
15682 @section separatefields
15684 The @code{separatefields} takes a frame-based video input and splits
15685 each frame into its components fields, producing a new half height clip
15686 with twice the frame rate and twice the frame count.
15688 This filter use field-dominance information in frame to decide which
15689 of each pair of fields to place first in the output.
15690 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15692 @section setdar, setsar
15694 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15697 This is done by changing the specified Sample (aka Pixel) Aspect
15698 Ratio, according to the following equation:
15700 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15703 Keep in mind that the @code{setdar} filter does not modify the pixel
15704 dimensions of the video frame. Also, the display aspect ratio set by
15705 this filter may be changed by later filters in the filterchain,
15706 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15709 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15710 the filter output video.
15712 Note that as a consequence of the application of this filter, the
15713 output display aspect ratio will change according to the equation
15716 Keep in mind that the sample aspect ratio set by the @code{setsar}
15717 filter may be changed by later filters in the filterchain, e.g. if
15718 another "setsar" or a "setdar" filter is applied.
15720 It accepts the following parameters:
15723 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15724 Set the aspect ratio used by the filter.
15726 The parameter can be a floating point number string, an expression, or
15727 a string of the form @var{num}:@var{den}, where @var{num} and
15728 @var{den} are the numerator and denominator of the aspect ratio. If
15729 the parameter is not specified, it is assumed the value "0".
15730 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15734 Set the maximum integer value to use for expressing numerator and
15735 denominator when reducing the expressed aspect ratio to a rational.
15736 Default value is @code{100}.
15740 The parameter @var{sar} is an expression containing
15741 the following constants:
15745 These are approximated values for the mathematical constants e
15746 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15749 The input width and height.
15752 These are the same as @var{w} / @var{h}.
15755 The input sample aspect ratio.
15758 The input display aspect ratio. It is the same as
15759 (@var{w} / @var{h}) * @var{sar}.
15762 Horizontal and vertical chroma subsample values. For example, for the
15763 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15766 @subsection Examples
15771 To change the display aspect ratio to 16:9, specify one of the following:
15778 To change the sample aspect ratio to 10:11, specify:
15784 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15785 1000 in the aspect ratio reduction, use the command:
15787 setdar=ratio=16/9:max=1000
15795 Force field for the output video frame.
15797 The @code{setfield} filter marks the interlace type field for the
15798 output frames. It does not change the input frame, but only sets the
15799 corresponding property, which affects how the frame is treated by
15800 following filters (e.g. @code{fieldorder} or @code{yadif}).
15802 The filter accepts the following options:
15807 Available values are:
15811 Keep the same field property.
15814 Mark the frame as bottom-field-first.
15817 Mark the frame as top-field-first.
15820 Mark the frame as progressive.
15827 Force frame parameter for the output video frame.
15829 The @code{setparams} filter marks interlace and color range for the
15830 output frames. It does not change the input frame, but only sets the
15831 corresponding property, which affects how the frame is treated by
15836 Available values are:
15840 Keep the same field property (default).
15843 Mark the frame as bottom-field-first.
15846 Mark the frame as top-field-first.
15849 Mark the frame as progressive.
15853 Available values are:
15857 Keep the same color range property (default).
15859 @item unspecified, unknown
15860 Mark the frame as unspecified color range.
15862 @item limited, tv, mpeg
15863 Mark the frame as limited range.
15865 @item full, pc, jpeg
15866 Mark the frame as full range.
15869 @item color_primaries
15870 Set the color primaries.
15871 Available values are:
15875 Keep the same color primaries property (default).
15892 Set the color transfer.
15893 Available values are:
15897 Keep the same color trc property (default).
15919 Set the colorspace.
15920 Available values are:
15924 Keep the same colorspace property (default).
15937 @item chroma-derived-nc
15938 @item chroma-derived-c
15945 Show a line containing various information for each input video frame.
15946 The input video is not modified.
15948 This filter supports the following options:
15952 Calculate checksums of each plane. By default enabled.
15955 The shown line contains a sequence of key/value pairs of the form
15956 @var{key}:@var{value}.
15958 The following values are shown in the output:
15962 The (sequential) number of the input frame, starting from 0.
15965 The Presentation TimeStamp of the input frame, expressed as a number of
15966 time base units. The time base unit depends on the filter input pad.
15969 The Presentation TimeStamp of the input frame, expressed as a number of
15973 The position of the frame in the input stream, or -1 if this information is
15974 unavailable and/or meaningless (for example in case of synthetic video).
15977 The pixel format name.
15980 The sample aspect ratio of the input frame, expressed in the form
15981 @var{num}/@var{den}.
15984 The size of the input frame. For the syntax of this option, check the
15985 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15988 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15989 for bottom field first).
15992 This is 1 if the frame is a key frame, 0 otherwise.
15995 The picture type of the input frame ("I" for an I-frame, "P" for a
15996 P-frame, "B" for a B-frame, or "?" for an unknown type).
15997 Also refer to the documentation of the @code{AVPictureType} enum and of
15998 the @code{av_get_picture_type_char} function defined in
15999 @file{libavutil/avutil.h}.
16002 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
16004 @item plane_checksum
16005 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
16006 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
16009 @section showpalette
16011 Displays the 256 colors palette of each frame. This filter is only relevant for
16012 @var{pal8} pixel format frames.
16014 It accepts the following option:
16018 Set the size of the box used to represent one palette color entry. Default is
16019 @code{30} (for a @code{30x30} pixel box).
16022 @section shuffleframes
16024 Reorder and/or duplicate and/or drop video frames.
16026 It accepts the following parameters:
16030 Set the destination indexes of input frames.
16031 This is space or '|' separated list of indexes that maps input frames to output
16032 frames. Number of indexes also sets maximal value that each index may have.
16033 '-1' index have special meaning and that is to drop frame.
16036 The first frame has the index 0. The default is to keep the input unchanged.
16038 @subsection Examples
16042 Swap second and third frame of every three frames of the input:
16044 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
16048 Swap 10th and 1st frame of every ten frames of the input:
16050 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
16054 @section shuffleplanes
16056 Reorder and/or duplicate video planes.
16058 It accepts the following parameters:
16063 The index of the input plane to be used as the first output plane.
16066 The index of the input plane to be used as the second output plane.
16069 The index of the input plane to be used as the third output plane.
16072 The index of the input plane to be used as the fourth output plane.
16076 The first plane has the index 0. The default is to keep the input unchanged.
16078 @subsection Examples
16082 Swap the second and third planes of the input:
16084 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
16088 @anchor{signalstats}
16089 @section signalstats
16090 Evaluate various visual metrics that assist in determining issues associated
16091 with the digitization of analog video media.
16093 By default the filter will log these metadata values:
16097 Display the minimal Y value contained within the input frame. Expressed in
16101 Display the Y value at the 10% percentile within the input frame. Expressed in
16105 Display the average Y value within the input frame. Expressed in range of
16109 Display the Y value at the 90% percentile within the input frame. Expressed in
16113 Display the maximum Y value contained within the input frame. Expressed in
16117 Display the minimal U value contained within the input frame. Expressed in
16121 Display the U value at the 10% percentile within the input frame. Expressed in
16125 Display the average U value within the input frame. Expressed in range of
16129 Display the U value at the 90% percentile within the input frame. Expressed in
16133 Display the maximum U value contained within the input frame. Expressed in
16137 Display the minimal V value contained within the input frame. Expressed in
16141 Display the V value at the 10% percentile within the input frame. Expressed in
16145 Display the average V value within the input frame. Expressed in range of
16149 Display the V value at the 90% percentile within the input frame. Expressed in
16153 Display the maximum V value contained within the input frame. Expressed in
16157 Display the minimal saturation value contained within the input frame.
16158 Expressed in range of [0-~181.02].
16161 Display the saturation value at the 10% percentile within the input frame.
16162 Expressed in range of [0-~181.02].
16165 Display the average saturation value within the input frame. Expressed in range
16169 Display the saturation value at the 90% percentile within the input frame.
16170 Expressed in range of [0-~181.02].
16173 Display the maximum saturation value contained within the input frame.
16174 Expressed in range of [0-~181.02].
16177 Display the median value for hue within the input frame. Expressed in range of
16181 Display the average value for hue within the input frame. Expressed in range of
16185 Display the average of sample value difference between all values of the Y
16186 plane in the current frame and corresponding values of the previous input frame.
16187 Expressed in range of [0-255].
16190 Display the average of sample value difference between all values of the U
16191 plane in the current frame and corresponding values of the previous input frame.
16192 Expressed in range of [0-255].
16195 Display the average of sample value difference between all values of the V
16196 plane in the current frame and corresponding values of the previous input frame.
16197 Expressed in range of [0-255].
16200 Display bit depth of Y plane in current frame.
16201 Expressed in range of [0-16].
16204 Display bit depth of U plane in current frame.
16205 Expressed in range of [0-16].
16208 Display bit depth of V plane in current frame.
16209 Expressed in range of [0-16].
16212 The filter accepts the following options:
16218 @option{stat} specify an additional form of image analysis.
16219 @option{out} output video with the specified type of pixel highlighted.
16221 Both options accept the following values:
16225 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
16226 unlike the neighboring pixels of the same field. Examples of temporal outliers
16227 include the results of video dropouts, head clogs, or tape tracking issues.
16230 Identify @var{vertical line repetition}. Vertical line repetition includes
16231 similar rows of pixels within a frame. In born-digital video vertical line
16232 repetition is common, but this pattern is uncommon in video digitized from an
16233 analog source. When it occurs in video that results from the digitization of an
16234 analog source it can indicate concealment from a dropout compensator.
16237 Identify pixels that fall outside of legal broadcast range.
16241 Set the highlight color for the @option{out} option. The default color is
16245 @subsection Examples
16249 Output data of various video metrics:
16251 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
16255 Output specific data about the minimum and maximum values of the Y plane per frame:
16257 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
16261 Playback video while highlighting pixels that are outside of broadcast range in red.
16263 ffplay example.mov -vf signalstats="out=brng:color=red"
16267 Playback video with signalstats metadata drawn over the frame.
16269 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
16272 The contents of signalstat_drawtext.txt used in the command are:
16275 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
16276 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
16277 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
16278 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
16286 Calculates the MPEG-7 Video Signature. The filter can handle more than one
16287 input. In this case the matching between the inputs can be calculated additionally.
16288 The filter always passes through the first input. The signature of each stream can
16289 be written into a file.
16291 It accepts the following options:
16295 Enable or disable the matching process.
16297 Available values are:
16301 Disable the calculation of a matching (default).
16303 Calculate the matching for the whole video and output whether the whole video
16304 matches or only parts.
16306 Calculate only until a matching is found or the video ends. Should be faster in
16311 Set the number of inputs. The option value must be a non negative integer.
16312 Default value is 1.
16315 Set the path to which the output is written. If there is more than one input,
16316 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
16317 integer), that will be replaced with the input number. If no filename is
16318 specified, no output will be written. This is the default.
16321 Choose the output format.
16323 Available values are:
16327 Use the specified binary representation (default).
16329 Use the specified xml representation.
16333 Set threshold to detect one word as similar. The option value must be an integer
16334 greater than zero. The default value is 9000.
16337 Set threshold to detect all words as similar. The option value must be an integer
16338 greater than zero. The default value is 60000.
16341 Set threshold to detect frames as similar. The option value must be an integer
16342 greater than zero. The default value is 116.
16345 Set the minimum length of a sequence in frames to recognize it as matching
16346 sequence. The option value must be a non negative integer value.
16347 The default value is 0.
16350 Set the minimum relation, that matching frames to all frames must have.
16351 The option value must be a double value between 0 and 1. The default value is 0.5.
16354 @subsection Examples
16358 To calculate the signature of an input video and store it in signature.bin:
16360 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
16364 To detect whether two videos match and store the signatures in XML format in
16365 signature0.xml and signature1.xml:
16367 ffmpeg -i input1.mkv -i input2.mkv -filter_complex "[0:v][1:v] signature=nb_inputs=2:detectmode=full:format=xml:filename=signature%d.xml" -map :v -f null -
16375 Blur the input video without impacting the outlines.
16377 It accepts the following options:
16380 @item luma_radius, lr
16381 Set the luma radius. The option value must be a float number in
16382 the range [0.1,5.0] that specifies the variance of the gaussian filter
16383 used to blur the image (slower if larger). Default value is 1.0.
16385 @item luma_strength, ls
16386 Set the luma strength. The option value must be a float number
16387 in the range [-1.0,1.0] that configures the blurring. A value included
16388 in [0.0,1.0] will blur the image whereas a value included in
16389 [-1.0,0.0] will sharpen the image. Default value is 1.0.
16391 @item luma_threshold, lt
16392 Set the luma threshold used as a coefficient to determine
16393 whether a pixel should be blurred or not. The option value must be an
16394 integer in the range [-30,30]. A value of 0 will filter all the image,
16395 a value included in [0,30] will filter flat areas and a value included
16396 in [-30,0] will filter edges. Default value is 0.
16398 @item chroma_radius, cr
16399 Set the chroma radius. The option value must be a float number in
16400 the range [0.1,5.0] that specifies the variance of the gaussian filter
16401 used to blur the image (slower if larger). Default value is @option{luma_radius}.
16403 @item chroma_strength, cs
16404 Set the chroma strength. The option value must be a float number
16405 in the range [-1.0,1.0] that configures the blurring. A value included
16406 in [0.0,1.0] will blur the image whereas a value included in
16407 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
16409 @item chroma_threshold, ct
16410 Set the chroma threshold used as a coefficient to determine
16411 whether a pixel should be blurred or not. The option value must be an
16412 integer in the range [-30,30]. A value of 0 will filter all the image,
16413 a value included in [0,30] will filter flat areas and a value included
16414 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
16417 If a chroma option is not explicitly set, the corresponding luma value
16422 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
16424 This filter takes in input two input videos, the first input is
16425 considered the "main" source and is passed unchanged to the
16426 output. The second input is used as a "reference" video for computing
16429 Both video inputs must have the same resolution and pixel format for
16430 this filter to work correctly. Also it assumes that both inputs
16431 have the same number of frames, which are compared one by one.
16433 The filter stores the calculated SSIM of each frame.
16435 The description of the accepted parameters follows.
16438 @item stats_file, f
16439 If specified the filter will use the named file to save the SSIM of
16440 each individual frame. When filename equals "-" the data is sent to
16444 The file printed if @var{stats_file} is selected, contains a sequence of
16445 key/value pairs of the form @var{key}:@var{value} for each compared
16448 A description of each shown parameter follows:
16452 sequential number of the input frame, starting from 1
16454 @item Y, U, V, R, G, B
16455 SSIM of the compared frames for the component specified by the suffix.
16458 SSIM of the compared frames for the whole frame.
16461 Same as above but in dB representation.
16464 This filter also supports the @ref{framesync} options.
16468 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16469 [main][ref] ssim="stats_file=stats.log" [out]
16472 On this example the input file being processed is compared with the
16473 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
16474 is stored in @file{stats.log}.
16476 Another example with both psnr and ssim at same time:
16478 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
16483 Convert between different stereoscopic image formats.
16485 The filters accept the following options:
16489 Set stereoscopic image format of input.
16491 Available values for input image formats are:
16494 side by side parallel (left eye left, right eye right)
16497 side by side crosseye (right eye left, left eye right)
16500 side by side parallel with half width resolution
16501 (left eye left, right eye right)
16504 side by side crosseye with half width resolution
16505 (right eye left, left eye right)
16508 above-below (left eye above, right eye below)
16511 above-below (right eye above, left eye below)
16514 above-below with half height resolution
16515 (left eye above, right eye below)
16518 above-below with half height resolution
16519 (right eye above, left eye below)
16522 alternating frames (left eye first, right eye second)
16525 alternating frames (right eye first, left eye second)
16528 interleaved rows (left eye has top row, right eye starts on next row)
16531 interleaved rows (right eye has top row, left eye starts on next row)
16534 interleaved columns, left eye first
16537 interleaved columns, right eye first
16539 Default value is @samp{sbsl}.
16543 Set stereoscopic image format of output.
16547 side by side parallel (left eye left, right eye right)
16550 side by side crosseye (right eye left, left eye right)
16553 side by side parallel with half width resolution
16554 (left eye left, right eye right)
16557 side by side crosseye with half width resolution
16558 (right eye left, left eye right)
16561 above-below (left eye above, right eye below)
16564 above-below (right eye above, left eye below)
16567 above-below with half height resolution
16568 (left eye above, right eye below)
16571 above-below with half height resolution
16572 (right eye above, left eye below)
16575 alternating frames (left eye first, right eye second)
16578 alternating frames (right eye first, left eye second)
16581 interleaved rows (left eye has top row, right eye starts on next row)
16584 interleaved rows (right eye has top row, left eye starts on next row)
16587 anaglyph red/blue gray
16588 (red filter on left eye, blue filter on right eye)
16591 anaglyph red/green gray
16592 (red filter on left eye, green filter on right eye)
16595 anaglyph red/cyan gray
16596 (red filter on left eye, cyan filter on right eye)
16599 anaglyph red/cyan half colored
16600 (red filter on left eye, cyan filter on right eye)
16603 anaglyph red/cyan color
16604 (red filter on left eye, cyan filter on right eye)
16607 anaglyph red/cyan color optimized with the least squares projection of dubois
16608 (red filter on left eye, cyan filter on right eye)
16611 anaglyph green/magenta gray
16612 (green filter on left eye, magenta filter on right eye)
16615 anaglyph green/magenta half colored
16616 (green filter on left eye, magenta filter on right eye)
16619 anaglyph green/magenta colored
16620 (green filter on left eye, magenta filter on right eye)
16623 anaglyph green/magenta color optimized with the least squares projection of dubois
16624 (green filter on left eye, magenta filter on right eye)
16627 anaglyph yellow/blue gray
16628 (yellow filter on left eye, blue filter on right eye)
16631 anaglyph yellow/blue half colored
16632 (yellow filter on left eye, blue filter on right eye)
16635 anaglyph yellow/blue colored
16636 (yellow filter on left eye, blue filter on right eye)
16639 anaglyph yellow/blue color optimized with the least squares projection of dubois
16640 (yellow filter on left eye, blue filter on right eye)
16643 mono output (left eye only)
16646 mono output (right eye only)
16649 checkerboard, left eye first
16652 checkerboard, right eye first
16655 interleaved columns, left eye first
16658 interleaved columns, right eye first
16664 Default value is @samp{arcd}.
16667 @subsection Examples
16671 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16677 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16683 @section streamselect, astreamselect
16684 Select video or audio streams.
16686 The filter accepts the following options:
16690 Set number of inputs. Default is 2.
16693 Set input indexes to remap to outputs.
16696 @subsection Commands
16698 The @code{streamselect} and @code{astreamselect} filter supports the following
16703 Set input indexes to remap to outputs.
16706 @subsection Examples
16710 Select first 5 seconds 1st stream and rest of time 2nd stream:
16712 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16716 Same as above, but for audio:
16718 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16723 Apply sobel operator to input video stream.
16725 The filter accepts the following option:
16729 Set which planes will be processed, unprocessed planes will be copied.
16730 By default value 0xf, all planes will be processed.
16733 Set value which will be multiplied with filtered result.
16736 Set value which will be added to filtered result.
16742 Apply a simple postprocessing filter that compresses and decompresses the image
16743 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16744 and average the results.
16746 The filter accepts the following options:
16750 Set quality. This option defines the number of levels for averaging. It accepts
16751 an integer in the range 0-6. If set to @code{0}, the filter will have no
16752 effect. A value of @code{6} means the higher quality. For each increment of
16753 that value the speed drops by a factor of approximately 2. Default value is
16757 Force a constant quantization parameter. If not set, the filter will use the QP
16758 from the video stream (if available).
16761 Set thresholding mode. Available modes are:
16765 Set hard thresholding (default).
16767 Set soft thresholding (better de-ringing effect, but likely blurrier).
16770 @item use_bframe_qp
16771 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16772 option may cause flicker since the B-Frames have often larger QP. Default is
16773 @code{0} (not enabled).
16778 Scale the input by applying one of the super-resolution methods based on
16779 convolutional neural networks. Supported models:
16783 Super-Resolution Convolutional Neural Network model (SRCNN).
16784 See @url{https://arxiv.org/abs/1501.00092}.
16787 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16788 See @url{https://arxiv.org/abs/1609.05158}.
16791 Training scripts as well as scripts for model file (.pb) saving can be found at
16792 @url{https://github.com/XueweiMeng/sr/tree/sr_dnn_native}. Original repository
16793 is at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16795 Native model files (.model) can be generated from TensorFlow model
16796 files (.pb) by using tools/python/convert.py
16798 The filter accepts the following options:
16802 Specify which DNN backend to use for model loading and execution. This option accepts
16803 the following values:
16807 Native implementation of DNN loading and execution.
16810 TensorFlow backend. To enable this backend you
16811 need to install the TensorFlow for C library (see
16812 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16813 @code{--enable-libtensorflow}
16816 Default value is @samp{native}.
16819 Set path to model file specifying network architecture and its parameters.
16820 Note that different backends use different file formats. TensorFlow backend
16821 can load files for both formats, while native backend can load files for only
16825 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16826 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16827 input upscaled using bicubic upscaling with proper scale factor.
16833 Draw subtitles on top of input video using the libass library.
16835 To enable compilation of this filter you need to configure FFmpeg with
16836 @code{--enable-libass}. This filter also requires a build with libavcodec and
16837 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16838 Alpha) subtitles format.
16840 The filter accepts the following options:
16844 Set the filename of the subtitle file to read. It must be specified.
16846 @item original_size
16847 Specify the size of the original video, the video for which the ASS file
16848 was composed. For the syntax of this option, check the
16849 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16850 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16851 correctly scale the fonts if the aspect ratio has been changed.
16854 Set a directory path containing fonts that can be used by the filter.
16855 These fonts will be used in addition to whatever the font provider uses.
16858 Process alpha channel, by default alpha channel is untouched.
16861 Set subtitles input character encoding. @code{subtitles} filter only. Only
16862 useful if not UTF-8.
16864 @item stream_index, si
16865 Set subtitles stream index. @code{subtitles} filter only.
16868 Override default style or script info parameters of the subtitles. It accepts a
16869 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16872 If the first key is not specified, it is assumed that the first value
16873 specifies the @option{filename}.
16875 For example, to render the file @file{sub.srt} on top of the input
16876 video, use the command:
16881 which is equivalent to:
16883 subtitles=filename=sub.srt
16886 To render the default subtitles stream from file @file{video.mkv}, use:
16888 subtitles=video.mkv
16891 To render the second subtitles stream from that file, use:
16893 subtitles=video.mkv:si=1
16896 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16897 @code{DejaVu Serif}, use:
16899 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16902 @section super2xsai
16904 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16905 Interpolate) pixel art scaling algorithm.
16907 Useful for enlarging pixel art images without reducing sharpness.
16911 Swap two rectangular objects in video.
16913 This filter accepts the following options:
16923 Set 1st rect x coordinate.
16926 Set 1st rect y coordinate.
16929 Set 2nd rect x coordinate.
16932 Set 2nd rect y coordinate.
16934 All expressions are evaluated once for each frame.
16937 The all options are expressions containing the following constants:
16942 The input width and height.
16945 same as @var{w} / @var{h}
16948 input sample aspect ratio
16951 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16954 The number of the input frame, starting from 0.
16957 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16960 the position in the file of the input frame, NAN if unknown
16968 Apply telecine process to the video.
16970 This filter accepts the following options:
16979 The default value is @code{top}.
16983 A string of numbers representing the pulldown pattern you wish to apply.
16984 The default value is @code{23}.
16988 Some typical patterns:
16993 24p: 2332 (preferred)
17000 24p: 222222222223 ("Euro pulldown")
17007 Apply threshold effect to video stream.
17009 This filter needs four video streams to perform thresholding.
17010 First stream is stream we are filtering.
17011 Second stream is holding threshold values, third stream is holding min values,
17012 and last, fourth stream is holding max values.
17014 The filter accepts the following option:
17018 Set which planes will be processed, unprocessed planes will be copied.
17019 By default value 0xf, all planes will be processed.
17022 For example if first stream pixel's component value is less then threshold value
17023 of pixel component from 2nd threshold stream, third stream value will picked,
17024 otherwise fourth stream pixel component value will be picked.
17026 Using color source filter one can perform various types of thresholding:
17028 @subsection Examples
17032 Binary threshold, using gray color as threshold:
17034 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
17038 Inverted binary threshold, using gray color as threshold:
17040 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
17044 Truncate binary threshold, using gray color as threshold:
17046 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
17050 Threshold to zero, using gray color as threshold:
17052 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
17056 Inverted threshold to zero, using gray color as threshold:
17058 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
17063 Select the most representative frame in a given sequence of consecutive frames.
17065 The filter accepts the following options:
17069 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
17070 will pick one of them, and then handle the next batch of @var{n} frames until
17071 the end. Default is @code{100}.
17074 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
17075 value will result in a higher memory usage, so a high value is not recommended.
17077 @subsection Examples
17081 Extract one picture each 50 frames:
17087 Complete example of a thumbnail creation with @command{ffmpeg}:
17089 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
17095 Tile several successive frames together.
17097 The filter accepts the following options:
17102 Set the grid size (i.e. the number of lines and columns). For the syntax of
17103 this option, check the
17104 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17107 Set the maximum number of frames to render in the given area. It must be less
17108 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
17109 the area will be used.
17112 Set the outer border margin in pixels.
17115 Set the inner border thickness (i.e. the number of pixels between frames). For
17116 more advanced padding options (such as having different values for the edges),
17117 refer to the pad video filter.
17120 Specify the color of the unused area. For the syntax of this option, check the
17121 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
17122 The default value of @var{color} is "black".
17125 Set the number of frames to overlap when tiling several successive frames together.
17126 The value must be between @code{0} and @var{nb_frames - 1}.
17129 Set the number of frames to initially be empty before displaying first output frame.
17130 This controls how soon will one get first output frame.
17131 The value must be between @code{0} and @var{nb_frames - 1}.
17134 @subsection Examples
17138 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
17140 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
17142 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
17143 duplicating each output frame to accommodate the originally detected frame
17147 Display @code{5} pictures in an area of @code{3x2} frames,
17148 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
17149 mixed flat and named options:
17151 tile=3x2:nb_frames=5:padding=7:margin=2
17155 @section tinterlace
17157 Perform various types of temporal field interlacing.
17159 Frames are counted starting from 1, so the first input frame is
17162 The filter accepts the following options:
17167 Specify the mode of the interlacing. This option can also be specified
17168 as a value alone. See below for a list of values for this option.
17170 Available values are:
17174 Move odd frames into the upper field, even into the lower field,
17175 generating a double height frame at half frame rate.
17179 Frame 1 Frame 2 Frame 3 Frame 4
17181 11111 22222 33333 44444
17182 11111 22222 33333 44444
17183 11111 22222 33333 44444
17184 11111 22222 33333 44444
17198 Only output odd frames, even frames are dropped, generating a frame with
17199 unchanged height at half frame rate.
17204 Frame 1 Frame 2 Frame 3 Frame 4
17206 11111 22222 33333 44444
17207 11111 22222 33333 44444
17208 11111 22222 33333 44444
17209 11111 22222 33333 44444
17219 Only output even frames, odd frames are dropped, generating a frame with
17220 unchanged height at half frame rate.
17225 Frame 1 Frame 2 Frame 3 Frame 4
17227 11111 22222 33333 44444
17228 11111 22222 33333 44444
17229 11111 22222 33333 44444
17230 11111 22222 33333 44444
17240 Expand each frame to full height, but pad alternate lines with black,
17241 generating a frame with double height at the same input frame rate.
17246 Frame 1 Frame 2 Frame 3 Frame 4
17248 11111 22222 33333 44444
17249 11111 22222 33333 44444
17250 11111 22222 33333 44444
17251 11111 22222 33333 44444
17254 11111 ..... 33333 .....
17255 ..... 22222 ..... 44444
17256 11111 ..... 33333 .....
17257 ..... 22222 ..... 44444
17258 11111 ..... 33333 .....
17259 ..... 22222 ..... 44444
17260 11111 ..... 33333 .....
17261 ..... 22222 ..... 44444
17265 @item interleave_top, 4
17266 Interleave the upper field from odd frames with the lower field from
17267 even frames, generating a frame with unchanged height at half frame rate.
17272 Frame 1 Frame 2 Frame 3 Frame 4
17274 11111<- 22222 33333<- 44444
17275 11111 22222<- 33333 44444<-
17276 11111<- 22222 33333<- 44444
17277 11111 22222<- 33333 44444<-
17287 @item interleave_bottom, 5
17288 Interleave the lower field from odd frames with the upper field from
17289 even frames, generating a frame with unchanged height at half frame rate.
17294 Frame 1 Frame 2 Frame 3 Frame 4
17296 11111 22222<- 33333 44444<-
17297 11111<- 22222 33333<- 44444
17298 11111 22222<- 33333 44444<-
17299 11111<- 22222 33333<- 44444
17309 @item interlacex2, 6
17310 Double frame rate with unchanged height. Frames are inserted each
17311 containing the second temporal field from the previous input frame and
17312 the first temporal field from the next input frame. This mode relies on
17313 the top_field_first flag. Useful for interlaced video displays with no
17314 field synchronisation.
17319 Frame 1 Frame 2 Frame 3 Frame 4
17321 11111 22222 33333 44444
17322 11111 22222 33333 44444
17323 11111 22222 33333 44444
17324 11111 22222 33333 44444
17327 11111 22222 22222 33333 33333 44444 44444
17328 11111 11111 22222 22222 33333 33333 44444
17329 11111 22222 22222 33333 33333 44444 44444
17330 11111 11111 22222 22222 33333 33333 44444
17335 Move odd frames into the upper field, even into the lower field,
17336 generating a double height frame at same frame rate.
17341 Frame 1 Frame 2 Frame 3 Frame 4
17343 11111 22222 33333 44444
17344 11111 22222 33333 44444
17345 11111 22222 33333 44444
17346 11111 22222 33333 44444
17349 11111 33333 33333 55555
17350 22222 22222 44444 44444
17351 11111 33333 33333 55555
17352 22222 22222 44444 44444
17353 11111 33333 33333 55555
17354 22222 22222 44444 44444
17355 11111 33333 33333 55555
17356 22222 22222 44444 44444
17361 Numeric values are deprecated but are accepted for backward
17362 compatibility reasons.
17364 Default mode is @code{merge}.
17367 Specify flags influencing the filter process.
17369 Available value for @var{flags} is:
17372 @item low_pass_filter, vlpf
17373 Enable linear vertical low-pass filtering in the filter.
17374 Vertical low-pass filtering is required when creating an interlaced
17375 destination from a progressive source which contains high-frequency
17376 vertical detail. Filtering will reduce interlace 'twitter' and Moire
17379 @item complex_filter, cvlpf
17380 Enable complex vertical low-pass filtering.
17381 This will slightly less reduce interlace 'twitter' and Moire
17382 patterning but better retain detail and subjective sharpness impression.
17386 Vertical low-pass filtering can only be enabled for @option{mode}
17387 @var{interleave_top} and @var{interleave_bottom}.
17393 Mix successive video frames.
17395 A description of the accepted options follows.
17399 The number of successive frames to mix. If unspecified, it defaults to 3.
17402 Specify weight of each input video frame.
17403 Each weight is separated by space. If number of weights is smaller than
17404 number of @var{frames} last specified weight will be used for all remaining
17408 Specify scale, if it is set it will be multiplied with sum
17409 of each weight multiplied with pixel values to give final destination
17410 pixel value. By default @var{scale} is auto scaled to sum of weights.
17413 @subsection Examples
17417 Average 7 successive frames:
17419 tmix=frames=7:weights="1 1 1 1 1 1 1"
17423 Apply simple temporal convolution:
17425 tmix=frames=3:weights="-1 3 -1"
17429 Similar as above but only showing temporal differences:
17431 tmix=frames=3:weights="-1 2 -1":scale=1
17437 Tone map colors from different dynamic ranges.
17439 This filter expects data in single precision floating point, as it needs to
17440 operate on (and can output) out-of-range values. Another filter, such as
17441 @ref{zscale}, is needed to convert the resulting frame to a usable format.
17443 The tonemapping algorithms implemented only work on linear light, so input
17444 data should be linearized beforehand (and possibly correctly tagged).
17447 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
17450 @subsection Options
17451 The filter accepts the following options.
17455 Set the tone map algorithm to use.
17457 Possible values are:
17460 Do not apply any tone map, only desaturate overbright pixels.
17463 Hard-clip any out-of-range values. Use it for perfect color accuracy for
17464 in-range values, while distorting out-of-range values.
17467 Stretch the entire reference gamut to a linear multiple of the display.
17470 Fit a logarithmic transfer between the tone curves.
17473 Preserve overall image brightness with a simple curve, using nonlinear
17474 contrast, which results in flattening details and degrading color accuracy.
17477 Preserve both dark and bright details better than @var{reinhard}, at the cost
17478 of slightly darkening everything. Use it when detail preservation is more
17479 important than color and brightness accuracy.
17482 Smoothly map out-of-range values, while retaining contrast and colors for
17483 in-range material as much as possible. Use it when color accuracy is more
17484 important than detail preservation.
17490 Tune the tone mapping algorithm.
17492 This affects the following algorithms:
17498 Specifies the scale factor to use while stretching.
17502 Specifies the exponent of the function.
17506 Specify an extra linear coefficient to multiply into the signal before clipping.
17510 Specify the local contrast coefficient at the display peak.
17511 Default to 0.5, which means that in-gamut values will be about half as bright
17518 Specify the transition point from linear to mobius transform. Every value
17519 below this point is guaranteed to be mapped 1:1. The higher the value, the
17520 more accurate the result will be, at the cost of losing bright details.
17521 Default to 0.3, which due to the steep initial slope still preserves in-range
17522 colors fairly accurately.
17526 Apply desaturation for highlights that exceed this level of brightness. The
17527 higher the parameter, the more color information will be preserved. This
17528 setting helps prevent unnaturally blown-out colors for super-highlights, by
17529 (smoothly) turning into white instead. This makes images feel more natural,
17530 at the cost of reducing information about out-of-range colors.
17532 The default of 2.0 is somewhat conservative and will mostly just apply to
17533 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17535 This option works only if the input frame has a supported color tag.
17538 Override signal/nominal/reference peak with this value. Useful when the
17539 embedded peak information in display metadata is not reliable or when tone
17540 mapping from a lower range to a higher range.
17545 Temporarily pad video frames.
17547 The filter accepts the following options:
17551 Specify number of delay frames before input video stream.
17554 Specify number of padding frames after input video stream.
17555 Set to -1 to pad indefinitely.
17558 Set kind of frames added to beginning of stream.
17559 Can be either @var{add} or @var{clone}.
17560 With @var{add} frames of solid-color are added.
17561 With @var{clone} frames are clones of first frame.
17564 Set kind of frames added to end of stream.
17565 Can be either @var{add} or @var{clone}.
17566 With @var{add} frames of solid-color are added.
17567 With @var{clone} frames are clones of last frame.
17569 @item start_duration, stop_duration
17570 Specify the duration of the start/stop delay. See
17571 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17572 for the accepted syntax.
17573 These options override @var{start} and @var{stop}.
17576 Specify the color of the padded area. For the syntax of this option,
17577 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17578 manual,ffmpeg-utils}.
17580 The default value of @var{color} is "black".
17586 Transpose rows with columns in the input video and optionally flip it.
17588 It accepts the following parameters:
17593 Specify the transposition direction.
17595 Can assume the following values:
17597 @item 0, 4, cclock_flip
17598 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17606 Rotate by 90 degrees clockwise, that is:
17614 Rotate by 90 degrees counterclockwise, that is:
17621 @item 3, 7, clock_flip
17622 Rotate by 90 degrees clockwise and vertically flip, that is:
17630 For values between 4-7, the transposition is only done if the input
17631 video geometry is portrait and not landscape. These values are
17632 deprecated, the @code{passthrough} option should be used instead.
17634 Numerical values are deprecated, and should be dropped in favor of
17635 symbolic constants.
17638 Do not apply the transposition if the input geometry matches the one
17639 specified by the specified value. It accepts the following values:
17642 Always apply transposition.
17644 Preserve portrait geometry (when @var{height} >= @var{width}).
17646 Preserve landscape geometry (when @var{width} >= @var{height}).
17649 Default value is @code{none}.
17652 For example to rotate by 90 degrees clockwise and preserve portrait
17655 transpose=dir=1:passthrough=portrait
17658 The command above can also be specified as:
17660 transpose=1:portrait
17663 @section transpose_npp
17665 Transpose rows with columns in the input video and optionally flip it.
17666 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17668 It accepts the following parameters:
17673 Specify the transposition direction.
17675 Can assume the following values:
17678 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17681 Rotate by 90 degrees clockwise.
17684 Rotate by 90 degrees counterclockwise.
17687 Rotate by 90 degrees clockwise and vertically flip.
17691 Do not apply the transposition if the input geometry matches the one
17692 specified by the specified value. It accepts the following values:
17695 Always apply transposition. (default)
17697 Preserve portrait geometry (when @var{height} >= @var{width}).
17699 Preserve landscape geometry (when @var{width} >= @var{height}).
17705 Trim the input so that the output contains one continuous subpart of the input.
17707 It accepts the following parameters:
17710 Specify the time of the start of the kept section, i.e. the frame with the
17711 timestamp @var{start} will be the first frame in the output.
17714 Specify the time of the first frame that will be dropped, i.e. the frame
17715 immediately preceding the one with the timestamp @var{end} will be the last
17716 frame in the output.
17719 This is the same as @var{start}, except this option sets the start timestamp
17720 in timebase units instead of seconds.
17723 This is the same as @var{end}, except this option sets the end timestamp
17724 in timebase units instead of seconds.
17727 The maximum duration of the output in seconds.
17730 The number of the first frame that should be passed to the output.
17733 The number of the first frame that should be dropped.
17736 @option{start}, @option{end}, and @option{duration} are expressed as time
17737 duration specifications; see
17738 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17739 for the accepted syntax.
17741 Note that the first two sets of the start/end options and the @option{duration}
17742 option look at the frame timestamp, while the _frame variants simply count the
17743 frames that pass through the filter. Also note that this filter does not modify
17744 the timestamps. If you wish for the output timestamps to start at zero, insert a
17745 setpts filter after the trim filter.
17747 If multiple start or end options are set, this filter tries to be greedy and
17748 keep all the frames that match at least one of the specified constraints. To keep
17749 only the part that matches all the constraints at once, chain multiple trim
17752 The defaults are such that all the input is kept. So it is possible to set e.g.
17753 just the end values to keep everything before the specified time.
17758 Drop everything except the second minute of input:
17760 ffmpeg -i INPUT -vf trim=60:120
17764 Keep only the first second:
17766 ffmpeg -i INPUT -vf trim=duration=1
17771 @section unpremultiply
17772 Apply alpha unpremultiply effect to input video stream using first plane
17773 of second stream as alpha.
17775 Both streams must have same dimensions and same pixel format.
17777 The filter accepts the following option:
17781 Set which planes will be processed, unprocessed planes will be copied.
17782 By default value 0xf, all planes will be processed.
17784 If the format has 1 or 2 components, then luma is bit 0.
17785 If the format has 3 or 4 components:
17786 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17787 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17788 If present, the alpha channel is always the last bit.
17791 Do not require 2nd input for processing, instead use alpha plane from input stream.
17797 Sharpen or blur the input video.
17799 It accepts the following parameters:
17802 @item luma_msize_x, lx
17803 Set the luma matrix horizontal size. It must be an odd integer between
17804 3 and 23. The default value is 5.
17806 @item luma_msize_y, ly
17807 Set the luma matrix vertical size. It must be an odd integer between 3
17808 and 23. The default value is 5.
17810 @item luma_amount, la
17811 Set the luma effect strength. It must be a floating point number, reasonable
17812 values lay between -1.5 and 1.5.
17814 Negative values will blur the input video, while positive values will
17815 sharpen it, a value of zero will disable the effect.
17817 Default value is 1.0.
17819 @item chroma_msize_x, cx
17820 Set the chroma matrix horizontal size. It must be an odd integer
17821 between 3 and 23. The default value is 5.
17823 @item chroma_msize_y, cy
17824 Set the chroma matrix vertical size. It must be an odd integer
17825 between 3 and 23. The default value is 5.
17827 @item chroma_amount, ca
17828 Set the chroma effect strength. It must be a floating point number, reasonable
17829 values lay between -1.5 and 1.5.
17831 Negative values will blur the input video, while positive values will
17832 sharpen it, a value of zero will disable the effect.
17834 Default value is 0.0.
17838 All parameters are optional and default to the equivalent of the
17839 string '5:5:1.0:5:5:0.0'.
17841 @subsection Examples
17845 Apply strong luma sharpen effect:
17847 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17851 Apply a strong blur of both luma and chroma parameters:
17853 unsharp=7:7:-2:7:7:-2
17859 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17860 the image at several (or - in the case of @option{quality} level @code{8} - all)
17861 shifts and average the results.
17863 The way this differs from the behavior of spp is that uspp actually encodes &
17864 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17865 DCT similar to MJPEG.
17867 The filter accepts the following options:
17871 Set quality. This option defines the number of levels for averaging. It accepts
17872 an integer in the range 0-8. If set to @code{0}, the filter will have no
17873 effect. A value of @code{8} means the higher quality. For each increment of
17874 that value the speed drops by a factor of approximately 2. Default value is
17878 Force a constant quantization parameter. If not set, the filter will use the QP
17879 from the video stream (if available).
17882 @section vaguedenoiser
17884 Apply a wavelet based denoiser.
17886 It transforms each frame from the video input into the wavelet domain,
17887 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17888 the obtained coefficients. It does an inverse wavelet transform after.
17889 Due to wavelet properties, it should give a nice smoothed result, and
17890 reduced noise, without blurring picture features.
17892 This filter accepts the following options:
17896 The filtering strength. The higher, the more filtered the video will be.
17897 Hard thresholding can use a higher threshold than soft thresholding
17898 before the video looks overfiltered. Default value is 2.
17901 The filtering method the filter will use.
17903 It accepts the following values:
17906 All values under the threshold will be zeroed.
17909 All values under the threshold will be zeroed. All values above will be
17910 reduced by the threshold.
17913 Scales or nullifies coefficients - intermediary between (more) soft and
17914 (less) hard thresholding.
17917 Default is garrote.
17920 Number of times, the wavelet will decompose the picture. Picture can't
17921 be decomposed beyond a particular point (typically, 8 for a 640x480
17922 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17925 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17928 A list of the planes to process. By default all planes are processed.
17931 @section vectorscope
17933 Display 2 color component values in the two dimensional graph (which is called
17936 This filter accepts the following options:
17940 Set vectorscope mode.
17942 It accepts the following values:
17945 Gray values are displayed on graph, higher brightness means more pixels have
17946 same component color value on location in graph. This is the default mode.
17949 Gray values are displayed on graph. Surrounding pixels values which are not
17950 present in video frame are drawn in gradient of 2 color components which are
17951 set by option @code{x} and @code{y}. The 3rd color component is static.
17954 Actual color components values present in video frame are displayed on graph.
17957 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17958 on graph increases value of another color component, which is luminance by
17959 default values of @code{x} and @code{y}.
17962 Actual colors present in video frame are displayed on graph. If two different
17963 colors map to same position on graph then color with higher value of component
17964 not present in graph is picked.
17967 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17968 component picked from radial gradient.
17972 Set which color component will be represented on X-axis. Default is @code{1}.
17975 Set which color component will be represented on Y-axis. Default is @code{2}.
17978 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17979 of color component which represents frequency of (X, Y) location in graph.
17984 No envelope, this is default.
17987 Instant envelope, even darkest single pixel will be clearly highlighted.
17990 Hold maximum and minimum values presented in graph over time. This way you
17991 can still spot out of range values without constantly looking at vectorscope.
17994 Peak and instant envelope combined together.
17998 Set what kind of graticule to draw.
18006 Set graticule opacity.
18009 Set graticule flags.
18013 Draw graticule for white point.
18016 Draw graticule for black point.
18019 Draw color points short names.
18023 Set background opacity.
18025 @item lthreshold, l
18026 Set low threshold for color component not represented on X or Y axis.
18027 Values lower than this value will be ignored. Default is 0.
18028 Note this value is multiplied with actual max possible value one pixel component
18029 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
18032 @item hthreshold, h
18033 Set high threshold for color component not represented on X or Y axis.
18034 Values higher than this value will be ignored. Default is 1.
18035 Note this value is multiplied with actual max possible value one pixel component
18036 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
18037 is 0.9 * 255 = 230.
18039 @item colorspace, c
18040 Set what kind of colorspace to use when drawing graticule.
18049 @anchor{vidstabdetect}
18050 @section vidstabdetect
18052 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
18053 @ref{vidstabtransform} for pass 2.
18055 This filter generates a file with relative translation and rotation
18056 transform information about subsequent frames, which is then used by
18057 the @ref{vidstabtransform} filter.
18059 To enable compilation of this filter you need to configure FFmpeg with
18060 @code{--enable-libvidstab}.
18062 This filter accepts the following options:
18066 Set the path to the file used to write the transforms information.
18067 Default value is @file{transforms.trf}.
18070 Set how shaky the video is and how quick the camera is. It accepts an
18071 integer in the range 1-10, a value of 1 means little shakiness, a
18072 value of 10 means strong shakiness. Default value is 5.
18075 Set the accuracy of the detection process. It must be a value in the
18076 range 1-15. A value of 1 means low accuracy, a value of 15 means high
18077 accuracy. Default value is 15.
18080 Set stepsize of the search process. The region around minimum is
18081 scanned with 1 pixel resolution. Default value is 6.
18084 Set minimum contrast. Below this value a local measurement field is
18085 discarded. Must be a floating point value in the range 0-1. Default
18089 Set reference frame number for tripod mode.
18091 If enabled, the motion of the frames is compared to a reference frame
18092 in the filtered stream, identified by the specified number. The idea
18093 is to compensate all movements in a more-or-less static scene and keep
18094 the camera view absolutely still.
18096 If set to 0, it is disabled. The frames are counted starting from 1.
18099 Show fields and transforms in the resulting frames. It accepts an
18100 integer in the range 0-2. Default value is 0, which disables any
18104 @subsection Examples
18108 Use default values:
18114 Analyze strongly shaky movie and put the results in file
18115 @file{mytransforms.trf}:
18117 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
18121 Visualize the result of internal transformations in the resulting
18124 vidstabdetect=show=1
18128 Analyze a video with medium shakiness using @command{ffmpeg}:
18130 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
18134 @anchor{vidstabtransform}
18135 @section vidstabtransform
18137 Video stabilization/deshaking: pass 2 of 2,
18138 see @ref{vidstabdetect} for pass 1.
18140 Read a file with transform information for each frame and
18141 apply/compensate them. Together with the @ref{vidstabdetect}
18142 filter this can be used to deshake videos. See also
18143 @url{http://public.hronopik.de/vid.stab}. It is important to also use
18144 the @ref{unsharp} filter, see below.
18146 To enable compilation of this filter you need to configure FFmpeg with
18147 @code{--enable-libvidstab}.
18149 @subsection Options
18153 Set path to the file used to read the transforms. Default value is
18154 @file{transforms.trf}.
18157 Set the number of frames (value*2 + 1) used for lowpass filtering the
18158 camera movements. Default value is 10.
18160 For example a number of 10 means that 21 frames are used (10 in the
18161 past and 10 in the future) to smoothen the motion in the video. A
18162 larger value leads to a smoother video, but limits the acceleration of
18163 the camera (pan/tilt movements). 0 is a special case where a static
18164 camera is simulated.
18167 Set the camera path optimization algorithm.
18169 Accepted values are:
18172 gaussian kernel low-pass filter on camera motion (default)
18174 averaging on transformations
18178 Set maximal number of pixels to translate frames. Default value is -1,
18182 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
18183 value is -1, meaning no limit.
18186 Specify how to deal with borders that may be visible due to movement
18189 Available values are:
18192 keep image information from previous frame (default)
18194 fill the border black
18198 Invert transforms if set to 1. Default value is 0.
18201 Consider transforms as relative to previous frame if set to 1,
18202 absolute if set to 0. Default value is 0.
18205 Set percentage to zoom. A positive value will result in a zoom-in
18206 effect, a negative value in a zoom-out effect. Default value is 0 (no
18210 Set optimal zooming to avoid borders.
18212 Accepted values are:
18217 optimal static zoom value is determined (only very strong movements
18218 will lead to visible borders) (default)
18220 optimal adaptive zoom value is determined (no borders will be
18221 visible), see @option{zoomspeed}
18224 Note that the value given at zoom is added to the one calculated here.
18227 Set percent to zoom maximally each frame (enabled when
18228 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
18232 Specify type of interpolation.
18234 Available values are:
18239 linear only horizontal
18241 linear in both directions (default)
18243 cubic in both directions (slow)
18247 Enable virtual tripod mode if set to 1, which is equivalent to
18248 @code{relative=0:smoothing=0}. Default value is 0.
18250 Use also @code{tripod} option of @ref{vidstabdetect}.
18253 Increase log verbosity if set to 1. Also the detected global motions
18254 are written to the temporary file @file{global_motions.trf}. Default
18258 @subsection Examples
18262 Use @command{ffmpeg} for a typical stabilization with default values:
18264 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
18267 Note the use of the @ref{unsharp} filter which is always recommended.
18270 Zoom in a bit more and load transform data from a given file:
18272 vidstabtransform=zoom=5:input="mytransforms.trf"
18276 Smoothen the video even more:
18278 vidstabtransform=smoothing=30
18284 Flip the input video vertically.
18286 For example, to vertically flip a video with @command{ffmpeg}:
18288 ffmpeg -i in.avi -vf "vflip" out.avi
18293 Detect variable frame rate video.
18295 This filter tries to detect if the input is variable or constant frame rate.
18297 At end it will output number of frames detected as having variable delta pts,
18298 and ones with constant delta pts.
18299 If there was frames with variable delta, than it will also show min and max delta
18304 Boost or alter saturation.
18306 The filter accepts the following options:
18309 Set strength of boost if positive value or strength of alter if negative value.
18310 Default is 0. Allowed range is from -2 to 2.
18313 Set the red balance. Default is 1. Allowed range is from -10 to 10.
18316 Set the green balance. Default is 1. Allowed range is from -10 to 10.
18319 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
18322 Set the red luma coefficient.
18325 Set the green luma coefficient.
18328 Set the blue luma coefficient.
18331 If @code{intensity} is negative and this is set to 1, colors will change,
18332 otherwise colors will be less saturated, more towards gray.
18338 Make or reverse a natural vignetting effect.
18340 The filter accepts the following options:
18344 Set lens angle expression as a number of radians.
18346 The value is clipped in the @code{[0,PI/2]} range.
18348 Default value: @code{"PI/5"}
18352 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
18356 Set forward/backward mode.
18358 Available modes are:
18361 The larger the distance from the central point, the darker the image becomes.
18364 The larger the distance from the central point, the brighter the image becomes.
18365 This can be used to reverse a vignette effect, though there is no automatic
18366 detection to extract the lens @option{angle} and other settings (yet). It can
18367 also be used to create a burning effect.
18370 Default value is @samp{forward}.
18373 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
18375 It accepts the following values:
18378 Evaluate expressions only once during the filter initialization.
18381 Evaluate expressions for each incoming frame. This is way slower than the
18382 @samp{init} mode since it requires all the scalers to be re-computed, but it
18383 allows advanced dynamic expressions.
18386 Default value is @samp{init}.
18389 Set dithering to reduce the circular banding effects. Default is @code{1}
18393 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
18394 Setting this value to the SAR of the input will make a rectangular vignetting
18395 following the dimensions of the video.
18397 Default is @code{1/1}.
18400 @subsection Expressions
18402 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
18403 following parameters.
18408 input width and height
18411 the number of input frame, starting from 0
18414 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
18415 @var{TB} units, NAN if undefined
18418 frame rate of the input video, NAN if the input frame rate is unknown
18421 the PTS (Presentation TimeStamp) of the filtered video frame,
18422 expressed in seconds, NAN if undefined
18425 time base of the input video
18429 @subsection Examples
18433 Apply simple strong vignetting effect:
18439 Make a flickering vignetting:
18441 vignette='PI/4+random(1)*PI/50':eval=frame
18446 @section vmafmotion
18448 Obtain the average vmaf motion score of a video.
18449 It is one of the component filters of VMAF.
18451 The obtained average motion score is printed through the logging system.
18453 In the below example the input file @file{ref.mpg} is being processed and score
18457 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
18461 Stack input videos vertically.
18463 All streams must be of same pixel format and of same width.
18465 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
18466 to create same output.
18468 The filter accept the following option:
18472 Set number of input streams. Default is 2.
18475 If set to 1, force the output to terminate when the shortest input
18476 terminates. Default value is 0.
18481 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
18482 Deinterlacing Filter").
18484 Based on the process described by Martin Weston for BBC R&D, and
18485 implemented based on the de-interlace algorithm written by Jim
18486 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
18487 uses filter coefficients calculated by BBC R&D.
18489 This filter use field-dominance information in frame to decide which
18490 of each pair of fields to place first in the output.
18491 If it gets it wrong use @ref{setfield} filter before @code{w3fdif} filter.
18493 There are two sets of filter coefficients, so called "simple":
18494 and "complex". Which set of filter coefficients is used can
18495 be set by passing an optional parameter:
18499 Set the interlacing filter coefficients. Accepts one of the following values:
18503 Simple filter coefficient set.
18505 More-complex filter coefficient set.
18507 Default value is @samp{complex}.
18510 Specify which frames to deinterlace. Accept one of the following values:
18514 Deinterlace all frames,
18516 Only deinterlace frames marked as interlaced.
18519 Default value is @samp{all}.
18523 Video waveform monitor.
18525 The waveform monitor plots color component intensity. By default luminance
18526 only. Each column of the waveform corresponds to a column of pixels in the
18529 It accepts the following options:
18533 Can be either @code{row}, or @code{column}. Default is @code{column}.
18534 In row mode, the graph on the left side represents color component value 0 and
18535 the right side represents value = 255. In column mode, the top side represents
18536 color component value = 0 and bottom side represents value = 255.
18539 Set intensity. Smaller values are useful to find out how many values of the same
18540 luminance are distributed across input rows/columns.
18541 Default value is @code{0.04}. Allowed range is [0, 1].
18544 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18545 In mirrored mode, higher values will be represented on the left
18546 side for @code{row} mode and at the top for @code{column} mode. Default is
18547 @code{1} (mirrored).
18551 It accepts the following values:
18554 Presents information identical to that in the @code{parade}, except
18555 that the graphs representing color components are superimposed directly
18558 This display mode makes it easier to spot relative differences or similarities
18559 in overlapping areas of the color components that are supposed to be identical,
18560 such as neutral whites, grays, or blacks.
18563 Display separate graph for the color components side by side in
18564 @code{row} mode or one below the other in @code{column} mode.
18567 Display separate graph for the color components side by side in
18568 @code{column} mode or one below the other in @code{row} mode.
18570 Using this display mode makes it easy to spot color casts in the highlights
18571 and shadows of an image, by comparing the contours of the top and the bottom
18572 graphs of each waveform. Since whites, grays, and blacks are characterized
18573 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18574 should display three waveforms of roughly equal width/height. If not, the
18575 correction is easy to perform by making level adjustments the three waveforms.
18577 Default is @code{stack}.
18579 @item components, c
18580 Set which color components to display. Default is 1, which means only luminance
18581 or red color component if input is in RGB colorspace. If is set for example to
18582 7 it will display all 3 (if) available color components.
18587 No envelope, this is default.
18590 Instant envelope, minimum and maximum values presented in graph will be easily
18591 visible even with small @code{step} value.
18594 Hold minimum and maximum values presented in graph across time. This way you
18595 can still spot out of range values without constantly looking at waveforms.
18598 Peak and instant envelope combined together.
18604 No filtering, this is default.
18607 Luma and chroma combined together.
18610 Similar as above, but shows difference between blue and red chroma.
18613 Similar as above, but use different colors.
18616 Displays only chroma.
18619 Displays actual color value on waveform.
18622 Similar as above, but with luma showing frequency of chroma values.
18626 Set which graticule to display.
18630 Do not display graticule.
18633 Display green graticule showing legal broadcast ranges.
18636 Display orange graticule showing legal broadcast ranges.
18640 Set graticule opacity.
18643 Set graticule flags.
18647 Draw numbers above lines. By default enabled.
18650 Draw dots instead of lines.
18654 Set scale used for displaying graticule.
18661 Default is digital.
18664 Set background opacity.
18667 @section weave, doubleweave
18669 The @code{weave} takes a field-based video input and join
18670 each two sequential fields into single frame, producing a new double
18671 height clip with half the frame rate and half the frame count.
18673 The @code{doubleweave} works same as @code{weave} but without
18674 halving frame rate and frame count.
18676 It accepts the following option:
18680 Set first field. Available values are:
18684 Set the frame as top-field-first.
18687 Set the frame as bottom-field-first.
18691 @subsection Examples
18695 Interlace video using @ref{select} and @ref{separatefields} filter:
18697 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18702 Apply the xBR high-quality magnification filter which is designed for pixel
18703 art. It follows a set of edge-detection rules, see
18704 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18706 It accepts the following option:
18710 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18711 @code{3xBR} and @code{4} for @code{4xBR}.
18712 Default is @code{3}.
18716 Pick median pixels from several input videos.
18718 The filter accept the following options:
18722 Set number of inputs.
18723 Default is 3. Allowed range is from 3 to 255.
18724 If number of inputs is even number, than result will be mean value between two median values.
18727 Set which planes to filter. Default value is @code{15}, by which all planes are processed.
18731 Stack video inputs into custom layout.
18733 All streams must be of same pixel format.
18735 The filter accept the following option:
18739 Set number of input streams. Default is 2.
18742 Specify layout of inputs.
18743 This option requires the desired layout configuration to be explicitly set by the user.
18744 This sets position of each video input in output. Each input
18745 is separated by '|'.
18746 The first number represents the column, and the second number represents the row.
18747 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18748 where X is video input from which to take width or height.
18749 Multiple values can be used when separated by '+'. In such
18750 case values are summed together.
18752 For 2 inputs, a default layout of @code{0_0|w0_0} is set. In all other cases,
18753 a layout must be set by the user.
18756 If set to 1, force the output to terminate when the shortest input
18757 terminates. Default value is 0.
18760 @subsection Examples
18764 Display 4 inputs into 2x2 grid,
18765 note that if inputs are of different sizes unused gaps might appear,
18766 as not all of output video is used.
18768 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18772 Display 4 inputs into 1x4 grid,
18773 note that if inputs are of different sizes unused gaps might appear,
18774 as not all of output video is used.
18776 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18780 Display 9 inputs into 3x3 grid,
18781 note that if inputs are of different sizes unused gaps might appear,
18782 as not all of output video is used.
18784 xstack=inputs=9:layout=w3_0|w3_h0+h2|w3_h0|0_h4|0_0|w3+w1_0|0_h1+h2|w3+w1_h0|w3+w1_h1+h2
18791 Deinterlace the input video ("yadif" means "yet another deinterlacing
18794 It accepts the following parameters:
18800 The interlacing mode to adopt. It accepts one of the following values:
18803 @item 0, send_frame
18804 Output one frame for each frame.
18805 @item 1, send_field
18806 Output one frame for each field.
18807 @item 2, send_frame_nospatial
18808 Like @code{send_frame}, but it skips the spatial interlacing check.
18809 @item 3, send_field_nospatial
18810 Like @code{send_field}, but it skips the spatial interlacing check.
18813 The default value is @code{send_frame}.
18816 The picture field parity assumed for the input interlaced video. It accepts one
18817 of the following values:
18821 Assume the top field is first.
18823 Assume the bottom field is first.
18825 Enable automatic detection of field parity.
18828 The default value is @code{auto}.
18829 If the interlacing is unknown or the decoder does not export this information,
18830 top field first will be assumed.
18833 Specify which frames to deinterlace. Accept one of the following
18838 Deinterlace all frames.
18839 @item 1, interlaced
18840 Only deinterlace frames marked as interlaced.
18843 The default value is @code{all}.
18846 @section yadif_cuda
18848 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18849 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18852 It accepts the following parameters:
18858 The interlacing mode to adopt. It accepts one of the following values:
18861 @item 0, send_frame
18862 Output one frame for each frame.
18863 @item 1, send_field
18864 Output one frame for each field.
18865 @item 2, send_frame_nospatial
18866 Like @code{send_frame}, but it skips the spatial interlacing check.
18867 @item 3, send_field_nospatial
18868 Like @code{send_field}, but it skips the spatial interlacing check.
18871 The default value is @code{send_frame}.
18874 The picture field parity assumed for the input interlaced video. It accepts one
18875 of the following values:
18879 Assume the top field is first.
18881 Assume the bottom field is first.
18883 Enable automatic detection of field parity.
18886 The default value is @code{auto}.
18887 If the interlacing is unknown or the decoder does not export this information,
18888 top field first will be assumed.
18891 Specify which frames to deinterlace. Accept one of the following
18896 Deinterlace all frames.
18897 @item 1, interlaced
18898 Only deinterlace frames marked as interlaced.
18901 The default value is @code{all}.
18906 Apply Zoom & Pan effect.
18908 This filter accepts the following options:
18912 Set the zoom expression. Range is 1-10. Default is 1.
18916 Set the x and y expression. Default is 0.
18919 Set the duration expression in number of frames.
18920 This sets for how many number of frames effect will last for
18921 single input image.
18924 Set the output image size, default is 'hd720'.
18927 Set the output frame rate, default is '25'.
18930 Each expression can contain the following constants:
18949 Output frame count.
18953 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18954 for current input frame.
18958 'x' and 'y' of last output frame of previous input frame or 0 when there was
18959 not yet such frame (first input frame).
18962 Last calculated zoom from 'z' expression for current input frame.
18965 Last calculated zoom of last output frame of previous input frame.
18968 Number of output frames for current input frame. Calculated from 'd' expression
18969 for each input frame.
18972 number of output frames created for previous input frame
18975 Rational number: input width / input height
18978 sample aspect ratio
18981 display aspect ratio
18985 @subsection Examples
18989 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18991 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
18995 Zoom-in up to 1.5 and pan always at center of picture:
18997 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
19001 Same as above but without pausing:
19003 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
19009 Scale (resize) the input video, using the z.lib library:
19010 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
19011 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
19013 The zscale filter forces the output display aspect ratio to be the same
19014 as the input, by changing the output sample aspect ratio.
19016 If the input image format is different from the format requested by
19017 the next filter, the zscale filter will convert the input to the
19020 @subsection Options
19021 The filter accepts the following options.
19026 Set the output video dimension expression. Default value is the input
19029 If the @var{width} or @var{w} value is 0, the input width is used for
19030 the output. If the @var{height} or @var{h} value is 0, the input height
19031 is used for the output.
19033 If one and only one of the values is -n with n >= 1, the zscale filter
19034 will use a value that maintains the aspect ratio of the input image,
19035 calculated from the other specified dimension. After that it will,
19036 however, make sure that the calculated dimension is divisible by n and
19037 adjust the value if necessary.
19039 If both values are -n with n >= 1, the behavior will be identical to
19040 both values being set to 0 as previously detailed.
19042 See below for the list of accepted constants for use in the dimension
19046 Set the video size. For the syntax of this option, check the
19047 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19050 Set the dither type.
19052 Possible values are:
19057 @item error_diffusion
19063 Set the resize filter type.
19065 Possible values are:
19075 Default is bilinear.
19078 Set the color range.
19080 Possible values are:
19087 Default is same as input.
19090 Set the color primaries.
19092 Possible values are:
19102 Default is same as input.
19105 Set the transfer characteristics.
19107 Possible values are:
19121 Default is same as input.
19124 Set the colorspace matrix.
19126 Possible value are:
19137 Default is same as input.
19140 Set the input color range.
19142 Possible values are:
19149 Default is same as input.
19151 @item primariesin, pin
19152 Set the input color primaries.
19154 Possible values are:
19164 Default is same as input.
19166 @item transferin, tin
19167 Set the input transfer characteristics.
19169 Possible values are:
19180 Default is same as input.
19182 @item matrixin, min
19183 Set the input colorspace matrix.
19185 Possible value are:
19197 Set the output chroma location.
19199 Possible values are:
19210 @item chromalin, cin
19211 Set the input chroma location.
19213 Possible values are:
19225 Set the nominal peak luminance.
19228 The values of the @option{w} and @option{h} options are expressions
19229 containing the following constants:
19234 The input width and height
19238 These are the same as @var{in_w} and @var{in_h}.
19242 The output (scaled) width and height
19246 These are the same as @var{out_w} and @var{out_h}
19249 The same as @var{iw} / @var{ih}
19252 input sample aspect ratio
19255 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
19259 horizontal and vertical input chroma subsample values. For example for the
19260 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
19264 horizontal and vertical output chroma subsample values. For example for the
19265 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
19271 @c man end VIDEO FILTERS
19273 @chapter OpenCL Video Filters
19274 @c man begin OPENCL VIDEO FILTERS
19276 Below is a description of the currently available OpenCL video filters.
19278 To enable compilation of these filters you need to configure FFmpeg with
19279 @code{--enable-opencl}.
19281 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
19284 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
19285 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
19286 given device parameters.
19288 @item -filter_hw_device @var{name}
19289 Pass the hardware device called @var{name} to all filters in any filter graph.
19293 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
19297 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
19299 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
19303 Since OpenCL filters are not able to access frame data in normal memory, all frame data needs to be uploaded(@ref{hwupload}) to hardware surfaces connected to the appropriate device before being used and then downloaded(@ref{hwdownload}) back to normal memory. Note that @ref{hwupload} will upload to a surface with the same layout as the software frame, so it may be necessary to add a @ref{format} filter immediately before to get the input into the right format and @ref{hwdownload} does not support all formats on the output - it may be necessary to insert an additional @ref{format} filter immediately following in the graph to get the output in a supported format.
19305 @section avgblur_opencl
19307 Apply average blur filter.
19309 The filter accepts the following options:
19313 Set horizontal radius size.
19314 Range is @code{[1, 1024]} and default value is @code{1}.
19317 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19320 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
19323 @subsection Example
19327 Apply average blur filter with horizontal and vertical size of 3, setting each pixel of the output to the average value of the 7x7 region centered on it in the input. For pixels on the edges of the image, the region does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
19329 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
19333 @section boxblur_opencl
19335 Apply a boxblur algorithm to the input video.
19337 It accepts the following parameters:
19341 @item luma_radius, lr
19342 @item luma_power, lp
19343 @item chroma_radius, cr
19344 @item chroma_power, cp
19345 @item alpha_radius, ar
19346 @item alpha_power, ap
19350 A description of the accepted options follows.
19353 @item luma_radius, lr
19354 @item chroma_radius, cr
19355 @item alpha_radius, ar
19356 Set an expression for the box radius in pixels used for blurring the
19357 corresponding input plane.
19359 The radius value must be a non-negative number, and must not be
19360 greater than the value of the expression @code{min(w,h)/2} for the
19361 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
19364 Default value for @option{luma_radius} is "2". If not specified,
19365 @option{chroma_radius} and @option{alpha_radius} default to the
19366 corresponding value set for @option{luma_radius}.
19368 The expressions can contain the following constants:
19372 The input width and height in pixels.
19376 The input chroma image width and height in pixels.
19380 The horizontal and vertical chroma subsample values. For example, for the
19381 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
19384 @item luma_power, lp
19385 @item chroma_power, cp
19386 @item alpha_power, ap
19387 Specify how many times the boxblur filter is applied to the
19388 corresponding plane.
19390 Default value for @option{luma_power} is 2. If not specified,
19391 @option{chroma_power} and @option{alpha_power} default to the
19392 corresponding value set for @option{luma_power}.
19394 A value of 0 will disable the effect.
19397 @subsection Examples
19399 Apply boxblur filter, setting each pixel of the output to the average value of box-radiuses @var{luma_radius}, @var{chroma_radius}, @var{alpha_radius} for each plane respectively. The filter will apply @var{luma_power}, @var{chroma_power}, @var{alpha_power} times onto the corresponding plane. For pixels on the edges of the image, the radius does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
19403 Apply a boxblur filter with the luma, chroma, and alpha radius
19404 set to 2 and luma, chroma, and alpha power set to 3. The filter will run 3 times with box-radius set to 2 for every plane of the image.
19406 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
19407 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
19411 Apply a boxblur filter with luma radius set to 2, luma_power to 1, chroma_radius to 4, chroma_power to 5, alpha_radius to 3 and alpha_power to 7.
19413 For the luma plane, a 2x2 box radius will be run once.
19415 For the chroma plane, a 4x4 box radius will be run 5 times.
19417 For the alpha plane, a 3x3 box radius will be run 7 times.
19419 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
19423 @section convolution_opencl
19425 Apply convolution of 3x3, 5x5, 7x7 matrix.
19427 The filter accepts the following options:
19434 Set matrix for each plane.
19435 Matrix is sequence of 9, 25 or 49 signed numbers.
19436 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
19442 Set multiplier for calculated value for each plane.
19443 If unset or 0, it will be sum of all matrix elements.
19444 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
19450 Set bias for each plane. This value is added to the result of the multiplication.
19451 Useful for making the overall image brighter or darker.
19452 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
19456 @subsection Examples
19462 -i INPUT -vf "hwupload, convolution_opencl=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, hwdownload" OUTPUT
19468 -i INPUT -vf "hwupload, convolution_opencl=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, hwdownload" OUTPUT
19472 Apply edge enhance:
19474 -i INPUT -vf "hwupload, convolution_opencl=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, hwdownload" OUTPUT
19480 -i INPUT -vf "hwupload, convolution_opencl=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, hwdownload" OUTPUT
19484 Apply laplacian edge detector which includes diagonals:
19486 -i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0, hwdownload" OUTPUT
19492 -i INPUT -vf "hwupload, convolution_opencl=-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, hwdownload" OUTPUT
19496 @section dilation_opencl
19498 Apply dilation effect to the video.
19500 This filter replaces the pixel by the local(3x3) maximum.
19502 It accepts the following options:
19509 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19510 If @code{0}, plane will remain unchanged.
19513 Flag which specifies the pixel to refer to.
19514 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19516 Flags to local 3x3 coordinates region centered on @code{x}:
19525 @subsection Example
19529 Apply dilation filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local maximum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local maximum is more then threshold of the corresponding plane, output pixel will be set to input pixel + threshold of corresponding plane.
19531 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19535 @section erosion_opencl
19537 Apply erosion effect to the video.
19539 This filter replaces the pixel by the local(3x3) minimum.
19541 It accepts the following options:
19548 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19549 If @code{0}, plane will remain unchanged.
19552 Flag which specifies the pixel to refer to.
19553 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19555 Flags to local 3x3 coordinates region centered on @code{x}:
19564 @subsection Example
19568 Apply erosion filter with threshold0 set to 30, threshold1 set 40, threshold2 set to 50 and coordinates set to 231, setting each pixel of the output to the local minimum between pixels: 1, 2, 3, 6, 7, 8 of the 3x3 region centered on it in the input. If the difference between input pixel and local minimum is more then threshold of the corresponding plane, output pixel will be set to input pixel - threshold of corresponding plane.
19570 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19574 @section colorkey_opencl
19575 RGB colorspace color keying.
19577 The filter accepts the following options:
19581 The color which will be replaced with transparency.
19584 Similarity percentage with the key color.
19586 0.01 matches only the exact key color, while 1.0 matches everything.
19591 0.0 makes pixels either fully transparent, or not transparent at all.
19593 Higher values result in semi-transparent pixels, with a higher transparency
19594 the more similar the pixels color is to the key color.
19597 @subsection Examples
19601 Make every semi-green pixel in the input transparent with some slight blending:
19603 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
19607 @section nlmeans_opencl
19609 Non-local Means denoise filter through OpenCL, this filter accepts same options as @ref{nlmeans}.
19611 @section overlay_opencl
19613 Overlay one video on top of another.
19615 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19616 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19618 The filter accepts the following options:
19623 Set the x coordinate of the overlaid video on the main video.
19624 Default value is @code{0}.
19627 Set the x coordinate of the overlaid video on the main video.
19628 Default value is @code{0}.
19632 @subsection Examples
19636 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19638 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19641 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19643 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19648 @section prewitt_opencl
19650 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19652 The filter accepts the following option:
19656 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19659 Set value which will be multiplied with filtered result.
19660 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19663 Set value which will be added to filtered result.
19664 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19667 @subsection Example
19671 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19673 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19677 @section roberts_opencl
19678 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19680 The filter accepts the following option:
19684 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19687 Set value which will be multiplied with filtered result.
19688 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19691 Set value which will be added to filtered result.
19692 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19695 @subsection Example
19699 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19701 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19705 @section sobel_opencl
19707 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19709 The filter accepts the following option:
19713 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19716 Set value which will be multiplied with filtered result.
19717 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19720 Set value which will be added to filtered result.
19721 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19724 @subsection Example
19728 Apply sobel operator with scale set to 2 and delta set to 10
19730 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19734 @section tonemap_opencl
19736 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19738 It accepts the following parameters:
19742 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19745 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19748 Apply desaturation for highlights that exceed this level of brightness. The
19749 higher the parameter, the more color information will be preserved. This
19750 setting helps prevent unnaturally blown-out colors for super-highlights, by
19751 (smoothly) turning into white instead. This makes images feel more natural,
19752 at the cost of reducing information about out-of-range colors.
19754 The default value is 0.5, and the algorithm here is a little different from
19755 the cpu version tonemap currently. A setting of 0.0 disables this option.
19758 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19759 is used to detect whether the scene has changed or not. If the distance between
19760 the current frame average brightness and the current running average exceeds
19761 a threshold value, we would re-calculate scene average and peak brightness.
19762 The default value is 0.2.
19765 Specify the output pixel format.
19767 Currently supported formats are:
19774 Set the output color range.
19776 Possible values are:
19782 Default is same as input.
19785 Set the output color primaries.
19787 Possible values are:
19793 Default is same as input.
19796 Set the output transfer characteristics.
19798 Possible values are:
19807 Set the output colorspace matrix.
19809 Possible value are:
19815 Default is same as input.
19819 @subsection Example
19823 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19825 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19829 @section unsharp_opencl
19831 Sharpen or blur the input video.
19833 It accepts the following parameters:
19836 @item luma_msize_x, lx
19837 Set the luma matrix horizontal size.
19838 Range is @code{[1, 23]} and default value is @code{5}.
19840 @item luma_msize_y, ly
19841 Set the luma matrix vertical size.
19842 Range is @code{[1, 23]} and default value is @code{5}.
19844 @item luma_amount, la
19845 Set the luma effect strength.
19846 Range is @code{[-10, 10]} and default value is @code{1.0}.
19848 Negative values will blur the input video, while positive values will
19849 sharpen it, a value of zero will disable the effect.
19851 @item chroma_msize_x, cx
19852 Set the chroma matrix horizontal size.
19853 Range is @code{[1, 23]} and default value is @code{5}.
19855 @item chroma_msize_y, cy
19856 Set the chroma matrix vertical size.
19857 Range is @code{[1, 23]} and default value is @code{5}.
19859 @item chroma_amount, ca
19860 Set the chroma effect strength.
19861 Range is @code{[-10, 10]} and default value is @code{0.0}.
19863 Negative values will blur the input video, while positive values will
19864 sharpen it, a value of zero will disable the effect.
19868 All parameters are optional and default to the equivalent of the
19869 string '5:5:1.0:5:5:0.0'.
19871 @subsection Examples
19875 Apply strong luma sharpen effect:
19877 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19881 Apply a strong blur of both luma and chroma parameters:
19883 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19887 @c man end OPENCL VIDEO FILTERS
19889 @chapter Video Sources
19890 @c man begin VIDEO SOURCES
19892 Below is a description of the currently available video sources.
19896 Buffer video frames, and make them available to the filter chain.
19898 This source is mainly intended for a programmatic use, in particular
19899 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19901 It accepts the following parameters:
19906 Specify the size (width and height) of the buffered video frames. For the
19907 syntax of this option, check the
19908 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19911 The input video width.
19914 The input video height.
19917 A string representing the pixel format of the buffered video frames.
19918 It may be a number corresponding to a pixel format, or a pixel format
19922 Specify the timebase assumed by the timestamps of the buffered frames.
19925 Specify the frame rate expected for the video stream.
19927 @item pixel_aspect, sar
19928 The sample (pixel) aspect ratio of the input video.
19931 Specify the optional parameters to be used for the scale filter which
19932 is automatically inserted when an input change is detected in the
19933 input size or format.
19935 @item hw_frames_ctx
19936 When using a hardware pixel format, this should be a reference to an
19937 AVHWFramesContext describing input frames.
19942 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19945 will instruct the source to accept video frames with size 320x240 and
19946 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19947 square pixels (1:1 sample aspect ratio).
19948 Since the pixel format with name "yuv410p" corresponds to the number 6
19949 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19950 this example corresponds to:
19952 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19955 Alternatively, the options can be specified as a flat string, but this
19956 syntax is deprecated:
19958 @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}]
19962 Create a pattern generated by an elementary cellular automaton.
19964 The initial state of the cellular automaton can be defined through the
19965 @option{filename} and @option{pattern} options. If such options are
19966 not specified an initial state is created randomly.
19968 At each new frame a new row in the video is filled with the result of
19969 the cellular automaton next generation. The behavior when the whole
19970 frame is filled is defined by the @option{scroll} option.
19972 This source accepts the following options:
19976 Read the initial cellular automaton state, i.e. the starting row, from
19977 the specified file.
19978 In the file, each non-whitespace character is considered an alive
19979 cell, a newline will terminate the row, and further characters in the
19980 file will be ignored.
19983 Read the initial cellular automaton state, i.e. the starting row, from
19984 the specified string.
19986 Each non-whitespace character in the string is considered an alive
19987 cell, a newline will terminate the row, and further characters in the
19988 string will be ignored.
19991 Set the video rate, that is the number of frames generated per second.
19994 @item random_fill_ratio, ratio
19995 Set the random fill ratio for the initial cellular automaton row. It
19996 is a floating point number value ranging from 0 to 1, defaults to
19999 This option is ignored when a file or a pattern is specified.
20001 @item random_seed, seed
20002 Set the seed for filling randomly the initial row, must be an integer
20003 included between 0 and UINT32_MAX. If not specified, or if explicitly
20004 set to -1, the filter will try to use a good random seed on a best
20008 Set the cellular automaton rule, it is a number ranging from 0 to 255.
20009 Default value is 110.
20012 Set the size of the output video. For the syntax of this option, check the
20013 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20015 If @option{filename} or @option{pattern} is specified, the size is set
20016 by default to the width of the specified initial state row, and the
20017 height is set to @var{width} * PHI.
20019 If @option{size} is set, it must contain the width of the specified
20020 pattern string, and the specified pattern will be centered in the
20023 If a filename or a pattern string is not specified, the size value
20024 defaults to "320x518" (used for a randomly generated initial state).
20027 If set to 1, scroll the output upward when all the rows in the output
20028 have been already filled. If set to 0, the new generated row will be
20029 written over the top row just after the bottom row is filled.
20032 @item start_full, full
20033 If set to 1, completely fill the output with generated rows before
20034 outputting the first frame.
20035 This is the default behavior, for disabling set the value to 0.
20038 If set to 1, stitch the left and right row edges together.
20039 This is the default behavior, for disabling set the value to 0.
20042 @subsection Examples
20046 Read the initial state from @file{pattern}, and specify an output of
20049 cellauto=f=pattern:s=200x400
20053 Generate a random initial row with a width of 200 cells, with a fill
20056 cellauto=ratio=2/3:s=200x200
20060 Create a pattern generated by rule 18 starting by a single alive cell
20061 centered on an initial row with width 100:
20063 cellauto=p=@@:s=100x400:full=0:rule=18
20067 Specify a more elaborated initial pattern:
20069 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
20074 @anchor{coreimagesrc}
20075 @section coreimagesrc
20076 Video source generated on GPU using Apple's CoreImage API on OSX.
20078 This video source is a specialized version of the @ref{coreimage} video filter.
20079 Use a core image generator at the beginning of the applied filterchain to
20080 generate the content.
20082 The coreimagesrc video source accepts the following options:
20084 @item list_generators
20085 List all available generators along with all their respective options as well as
20086 possible minimum and maximum values along with the default values.
20088 list_generators=true
20092 Specify the size of the sourced video. For the syntax of this option, check the
20093 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20094 The default value is @code{320x240}.
20097 Specify the frame rate of the sourced video, as the number of frames
20098 generated per second. It has to be a string in the format
20099 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
20100 number or a valid video frame rate abbreviation. The default value is
20104 Set the sample aspect ratio of the sourced video.
20107 Set the duration of the sourced video. See
20108 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20109 for the accepted syntax.
20111 If not specified, or the expressed duration is negative, the video is
20112 supposed to be generated forever.
20115 Additionally, all options of the @ref{coreimage} video filter are accepted.
20116 A complete filterchain can be used for further processing of the
20117 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
20118 and examples for details.
20120 @subsection Examples
20125 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
20126 given as complete and escaped command-line for Apple's standard bash shell:
20128 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
20130 This example is equivalent to the QRCode example of @ref{coreimage} without the
20131 need for a nullsrc video source.
20135 @section mandelbrot
20137 Generate a Mandelbrot set fractal, and progressively zoom towards the
20138 point specified with @var{start_x} and @var{start_y}.
20140 This source accepts the following options:
20145 Set the terminal pts value. Default value is 400.
20148 Set the terminal scale value.
20149 Must be a floating point value. Default value is 0.3.
20152 Set the inner coloring mode, that is the algorithm used to draw the
20153 Mandelbrot fractal internal region.
20155 It shall assume one of the following values:
20160 Show time until convergence.
20162 Set color based on point closest to the origin of the iterations.
20167 Default value is @var{mincol}.
20170 Set the bailout value. Default value is 10.0.
20173 Set the maximum of iterations performed by the rendering
20174 algorithm. Default value is 7189.
20177 Set outer coloring mode.
20178 It shall assume one of following values:
20180 @item iteration_count
20181 Set iteration count mode.
20182 @item normalized_iteration_count
20183 set normalized iteration count mode.
20185 Default value is @var{normalized_iteration_count}.
20188 Set frame rate, expressed as number of frames per second. Default
20192 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
20193 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
20196 Set the initial scale value. Default value is 3.0.
20199 Set the initial x position. Must be a floating point value between
20200 -100 and 100. Default value is -0.743643887037158704752191506114774.
20203 Set the initial y position. Must be a floating point value between
20204 -100 and 100. Default value is -0.131825904205311970493132056385139.
20209 Generate various test patterns, as generated by the MPlayer test filter.
20211 The size of the generated video is fixed, and is 256x256.
20212 This source is useful in particular for testing encoding features.
20214 This source accepts the following options:
20219 Specify the frame rate of the sourced video, as the number of frames
20220 generated per second. It has to be a string in the format
20221 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
20222 number or a valid video frame rate abbreviation. The default value is
20226 Set the duration of the sourced video. See
20227 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20228 for the accepted syntax.
20230 If not specified, or the expressed duration is negative, the video is
20231 supposed to be generated forever.
20235 Set the number or the name of the test to perform. Supported tests are:
20251 Default value is "all", which will cycle through the list of all tests.
20256 mptestsrc=t=dc_luma
20259 will generate a "dc_luma" test pattern.
20261 @section frei0r_src
20263 Provide a frei0r source.
20265 To enable compilation of this filter you need to install the frei0r
20266 header and configure FFmpeg with @code{--enable-frei0r}.
20268 This source accepts the following parameters:
20273 The size of the video to generate. For the syntax of this option, check the
20274 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20277 The framerate of the generated video. It may be a string of the form
20278 @var{num}/@var{den} or a frame rate abbreviation.
20281 The name to the frei0r source to load. For more information regarding frei0r and
20282 how to set the parameters, read the @ref{frei0r} section in the video filters
20285 @item filter_params
20286 A '|'-separated list of parameters to pass to the frei0r source.
20290 For example, to generate a frei0r partik0l source with size 200x200
20291 and frame rate 10 which is overlaid on the overlay filter main input:
20293 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
20298 Generate a life pattern.
20300 This source is based on a generalization of John Conway's life game.
20302 The sourced input represents a life grid, each pixel represents a cell
20303 which can be in one of two possible states, alive or dead. Every cell
20304 interacts with its eight neighbours, which are the cells that are
20305 horizontally, vertically, or diagonally adjacent.
20307 At each interaction the grid evolves according to the adopted rule,
20308 which specifies the number of neighbor alive cells which will make a
20309 cell stay alive or born. The @option{rule} option allows one to specify
20312 This source accepts the following options:
20316 Set the file from which to read the initial grid state. In the file,
20317 each non-whitespace character is considered an alive cell, and newline
20318 is used to delimit the end of each row.
20320 If this option is not specified, the initial grid is generated
20324 Set the video rate, that is the number of frames generated per second.
20327 @item random_fill_ratio, ratio
20328 Set the random fill ratio for the initial random grid. It is a
20329 floating point number value ranging from 0 to 1, defaults to 1/PHI.
20330 It is ignored when a file is specified.
20332 @item random_seed, seed
20333 Set the seed for filling the initial random grid, must be an integer
20334 included between 0 and UINT32_MAX. If not specified, or if explicitly
20335 set to -1, the filter will try to use a good random seed on a best
20341 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
20342 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
20343 @var{NS} specifies the number of alive neighbor cells which make a
20344 live cell stay alive, and @var{NB} the number of alive neighbor cells
20345 which make a dead cell to become alive (i.e. to "born").
20346 "s" and "b" can be used in place of "S" and "B", respectively.
20348 Alternatively a rule can be specified by an 18-bits integer. The 9
20349 high order bits are used to encode the next cell state if it is alive
20350 for each number of neighbor alive cells, the low order bits specify
20351 the rule for "borning" new cells. Higher order bits encode for an
20352 higher number of neighbor cells.
20353 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
20354 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
20356 Default value is "S23/B3", which is the original Conway's game of life
20357 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
20358 cells, and will born a new cell if there are three alive cells around
20362 Set the size of the output video. For the syntax of this option, check the
20363 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20365 If @option{filename} is specified, the size is set by default to the
20366 same size of the input file. If @option{size} is set, it must contain
20367 the size specified in the input file, and the initial grid defined in
20368 that file is centered in the larger resulting area.
20370 If a filename is not specified, the size value defaults to "320x240"
20371 (used for a randomly generated initial grid).
20374 If set to 1, stitch the left and right grid edges together, and the
20375 top and bottom edges also. Defaults to 1.
20378 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
20379 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
20380 value from 0 to 255.
20383 Set the color of living (or new born) cells.
20386 Set the color of dead cells. If @option{mold} is set, this is the first color
20387 used to represent a dead cell.
20390 Set mold color, for definitely dead and moldy cells.
20392 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
20393 ffmpeg-utils manual,ffmpeg-utils}.
20396 @subsection Examples
20400 Read a grid from @file{pattern}, and center it on a grid of size
20403 life=f=pattern:s=300x300
20407 Generate a random grid of size 200x200, with a fill ratio of 2/3:
20409 life=ratio=2/3:s=200x200
20413 Specify a custom rule for evolving a randomly generated grid:
20419 Full example with slow death effect (mold) using @command{ffplay}:
20421 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
20428 @anchor{haldclutsrc}
20431 @anchor{pal100bars}
20432 @anchor{rgbtestsrc}
20434 @anchor{smptehdbars}
20437 @anchor{yuvtestsrc}
20438 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
20440 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
20442 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
20444 The @code{color} source provides an uniformly colored input.
20446 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
20447 @ref{haldclut} filter.
20449 The @code{nullsrc} source returns unprocessed video frames. It is
20450 mainly useful to be employed in analysis / debugging tools, or as the
20451 source for filters which ignore the input data.
20453 The @code{pal75bars} source generates a color bars pattern, based on
20454 EBU PAL recommendations with 75% color levels.
20456 The @code{pal100bars} source generates a color bars pattern, based on
20457 EBU PAL recommendations with 100% color levels.
20459 The @code{rgbtestsrc} source generates an RGB test pattern useful for
20460 detecting RGB vs BGR issues. You should see a red, green and blue
20461 stripe from top to bottom.
20463 The @code{smptebars} source generates a color bars pattern, based on
20464 the SMPTE Engineering Guideline EG 1-1990.
20466 The @code{smptehdbars} source generates a color bars pattern, based on
20467 the SMPTE RP 219-2002.
20469 The @code{testsrc} source generates a test video pattern, showing a
20470 color pattern, a scrolling gradient and a timestamp. This is mainly
20471 intended for testing purposes.
20473 The @code{testsrc2} source is similar to testsrc, but supports more
20474 pixel formats instead of just @code{rgb24}. This allows using it as an
20475 input for other tests without requiring a format conversion.
20477 The @code{yuvtestsrc} source generates an YUV test pattern. You should
20478 see a y, cb and cr stripe from top to bottom.
20480 The sources accept the following parameters:
20485 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
20486 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
20487 pixels to be used as identity matrix for 3D lookup tables. Each component is
20488 coded on a @code{1/(N*N)} scale.
20491 Specify the color of the source, only available in the @code{color}
20492 source. For the syntax of this option, check the
20493 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
20496 Specify the size of the sourced video. For the syntax of this option, check the
20497 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20498 The default value is @code{320x240}.
20500 This option is not available with the @code{allrgb}, @code{allyuv}, and
20501 @code{haldclutsrc} filters.
20504 Specify the frame rate of the sourced video, as the number of frames
20505 generated per second. It has to be a string in the format
20506 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
20507 number or a valid video frame rate abbreviation. The default value is
20511 Set the duration of the sourced video. See
20512 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20513 for the accepted syntax.
20515 If not specified, or the expressed duration is negative, the video is
20516 supposed to be generated forever.
20519 Set the sample aspect ratio of the sourced video.
20522 Specify the alpha (opacity) of the background, only available in the
20523 @code{testsrc2} source. The value must be between 0 (fully transparent) and
20524 255 (fully opaque, the default).
20527 Set the number of decimals to show in the timestamp, only available in the
20528 @code{testsrc} source.
20530 The displayed timestamp value will correspond to the original
20531 timestamp value multiplied by the power of 10 of the specified
20532 value. Default value is 0.
20535 @subsection Examples
20539 Generate a video with a duration of 5.3 seconds, with size
20540 176x144 and a frame rate of 10 frames per second:
20542 testsrc=duration=5.3:size=qcif:rate=10
20546 The following graph description will generate a red source
20547 with an opacity of 0.2, with size "qcif" and a frame rate of 10
20550 color=c=red@@0.2:s=qcif:r=10
20554 If the input content is to be ignored, @code{nullsrc} can be used. The
20555 following command generates noise in the luminance plane by employing
20556 the @code{geq} filter:
20558 nullsrc=s=256x256, geq=random(1)*255:128:128
20562 @subsection Commands
20564 The @code{color} source supports the following commands:
20568 Set the color of the created image. Accepts the same syntax of the
20569 corresponding @option{color} option.
20574 Generate video using an OpenCL program.
20579 OpenCL program source file.
20582 Kernel name in program.
20585 Size of frames to generate. This must be set.
20588 Pixel format to use for the generated frames. This must be set.
20591 Number of frames generated every second. Default value is '25'.
20595 For details of how the program loading works, see the @ref{program_opencl}
20602 Generate a colour ramp by setting pixel values from the position of the pixel
20603 in the output image. (Note that this will work with all pixel formats, but
20604 the generated output will not be the same.)
20606 __kernel void ramp(__write_only image2d_t dst,
20607 unsigned int index)
20609 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20612 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20614 write_imagef(dst, loc, val);
20619 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20621 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20622 unsigned int index)
20624 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20626 float4 value = 0.0f;
20627 int x = loc.x + index;
20628 int y = loc.y + index;
20629 while (x > 0 || y > 0) {
20630 if (x % 3 == 1 && y % 3 == 1) {
20638 write_imagef(dst, loc, value);
20644 @c man end VIDEO SOURCES
20646 @chapter Video Sinks
20647 @c man begin VIDEO SINKS
20649 Below is a description of the currently available video sinks.
20651 @section buffersink
20653 Buffer video frames, and make them available to the end of the filter
20656 This sink is mainly intended for programmatic use, in particular
20657 through the interface defined in @file{libavfilter/buffersink.h}
20658 or the options system.
20660 It accepts a pointer to an AVBufferSinkContext structure, which
20661 defines the incoming buffers' formats, to be passed as the opaque
20662 parameter to @code{avfilter_init_filter} for initialization.
20666 Null video sink: do absolutely nothing with the input video. It is
20667 mainly useful as a template and for use in analysis / debugging
20670 @c man end VIDEO SINKS
20672 @chapter Multimedia Filters
20673 @c man begin MULTIMEDIA FILTERS
20675 Below is a description of the currently available multimedia filters.
20679 Convert input audio to a video output, displaying the audio bit scope.
20681 The filter accepts the following options:
20685 Set frame rate, expressed as number of frames per second. Default
20689 Specify the video size for the output. For the syntax of this option, check the
20690 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20691 Default value is @code{1024x256}.
20694 Specify list of colors separated by space or by '|' which will be used to
20695 draw channels. Unrecognized or missing colors will be replaced
20699 @section ahistogram
20701 Convert input audio to a video output, displaying the volume histogram.
20703 The filter accepts the following options:
20707 Specify how histogram is calculated.
20709 It accepts the following values:
20712 Use single histogram for all channels.
20714 Use separate histogram for each channel.
20716 Default is @code{single}.
20719 Set frame rate, expressed as number of frames per second. Default
20723 Specify the video size for the output. For the syntax of this option, check the
20724 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20725 Default value is @code{hd720}.
20730 It accepts the following values:
20741 reverse logarithmic
20743 Default is @code{log}.
20746 Set amplitude scale.
20748 It accepts the following values:
20755 Default is @code{log}.
20758 Set how much frames to accumulate in histogram.
20759 Default is 1. Setting this to -1 accumulates all frames.
20762 Set histogram ratio of window height.
20765 Set sonogram sliding.
20767 It accepts the following values:
20770 replace old rows with new ones.
20772 scroll from top to bottom.
20774 Default is @code{replace}.
20777 @section aphasemeter
20779 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20780 representing mean phase of current audio frame. A video output can also be produced and is
20781 enabled by default. The audio is passed through as first output.
20783 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20784 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20785 and @code{1} means channels are in phase.
20787 The filter accepts the following options, all related to its video output:
20791 Set the output frame rate. Default value is @code{25}.
20794 Set the video size for the output. For the syntax of this option, check the
20795 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20796 Default value is @code{800x400}.
20801 Specify the red, green, blue contrast. Default values are @code{2},
20802 @code{7} and @code{1}.
20803 Allowed range is @code{[0, 255]}.
20806 Set color which will be used for drawing median phase. If color is
20807 @code{none} which is default, no median phase value will be drawn.
20810 Enable video output. Default is enabled.
20813 @section avectorscope
20815 Convert input audio to a video output, representing the audio vector
20818 The filter is used to measure the difference between channels of stereo
20819 audio stream. A monoaural signal, consisting of identical left and right
20820 signal, results in straight vertical line. Any stereo separation is visible
20821 as a deviation from this line, creating a Lissajous figure.
20822 If the straight (or deviation from it) but horizontal line appears this
20823 indicates that the left and right channels are out of phase.
20825 The filter accepts the following options:
20829 Set the vectorscope mode.
20831 Available values are:
20834 Lissajous rotated by 45 degrees.
20837 Same as above but not rotated.
20840 Shape resembling half of circle.
20843 Default value is @samp{lissajous}.
20846 Set the video size for the output. For the syntax of this option, check the
20847 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20848 Default value is @code{400x400}.
20851 Set the output frame rate. Default value is @code{25}.
20857 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20858 @code{160}, @code{80} and @code{255}.
20859 Allowed range is @code{[0, 255]}.
20865 Specify the red, green, blue and alpha fade. Default values are @code{15},
20866 @code{10}, @code{5} and @code{5}.
20867 Allowed range is @code{[0, 255]}.
20870 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20871 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20874 Set the vectorscope drawing mode.
20876 Available values are:
20879 Draw dot for each sample.
20882 Draw line between previous and current sample.
20885 Default value is @samp{dot}.
20888 Specify amplitude scale of audio samples.
20890 Available values are:
20906 Swap left channel axis with right channel axis.
20916 Mirror only x axis.
20919 Mirror only y axis.
20927 @subsection Examples
20931 Complete example using @command{ffplay}:
20933 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20934 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20938 @section bench, abench
20940 Benchmark part of a filtergraph.
20942 The filter accepts the following options:
20946 Start or stop a timer.
20948 Available values are:
20951 Get the current time, set it as frame metadata (using the key
20952 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20955 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20956 the input frame metadata to get the time difference. Time difference, average,
20957 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20958 @code{min}) are then printed. The timestamps are expressed in seconds.
20962 @subsection Examples
20966 Benchmark @ref{selectivecolor} filter:
20968 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20974 Concatenate audio and video streams, joining them together one after the
20977 The filter works on segments of synchronized video and audio streams. All
20978 segments must have the same number of streams of each type, and that will
20979 also be the number of streams at output.
20981 The filter accepts the following options:
20986 Set the number of segments. Default is 2.
20989 Set the number of output video streams, that is also the number of video
20990 streams in each segment. Default is 1.
20993 Set the number of output audio streams, that is also the number of audio
20994 streams in each segment. Default is 0.
20997 Activate unsafe mode: do not fail if segments have a different format.
21001 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
21002 @var{a} audio outputs.
21004 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
21005 segment, in the same order as the outputs, then the inputs for the second
21008 Related streams do not always have exactly the same duration, for various
21009 reasons including codec frame size or sloppy authoring. For that reason,
21010 related synchronized streams (e.g. a video and its audio track) should be
21011 concatenated at once. The concat filter will use the duration of the longest
21012 stream in each segment (except the last one), and if necessary pad shorter
21013 audio streams with silence.
21015 For this filter to work correctly, all segments must start at timestamp 0.
21017 All corresponding streams must have the same parameters in all segments; the
21018 filtering system will automatically select a common pixel format for video
21019 streams, and a common sample format, sample rate and channel layout for
21020 audio streams, but other settings, such as resolution, must be converted
21021 explicitly by the user.
21023 Different frame rates are acceptable but will result in variable frame rate
21024 at output; be sure to configure the output file to handle it.
21026 @subsection Examples
21030 Concatenate an opening, an episode and an ending, all in bilingual version
21031 (video in stream 0, audio in streams 1 and 2):
21033 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
21034 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
21035 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
21036 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
21040 Concatenate two parts, handling audio and video separately, using the
21041 (a)movie sources, and adjusting the resolution:
21043 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
21044 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
21045 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
21047 Note that a desync will happen at the stitch if the audio and video streams
21048 do not have exactly the same duration in the first file.
21052 @subsection Commands
21054 This filter supports the following commands:
21057 Close the current segment and step to the next one
21060 @section drawgraph, adrawgraph
21062 Draw a graph using input video or audio metadata.
21064 It accepts the following parameters:
21068 Set 1st frame metadata key from which metadata values will be used to draw a graph.
21071 Set 1st foreground color expression.
21074 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
21077 Set 2nd foreground color expression.
21080 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
21083 Set 3rd foreground color expression.
21086 Set 4th frame metadata key from which metadata values will be used to draw a graph.
21089 Set 4th foreground color expression.
21092 Set minimal value of metadata value.
21095 Set maximal value of metadata value.
21098 Set graph background color. Default is white.
21103 Available values for mode is:
21110 Default is @code{line}.
21115 Available values for slide is:
21118 Draw new frame when right border is reached.
21121 Replace old columns with new ones.
21124 Scroll from right to left.
21127 Scroll from left to right.
21130 Draw single picture.
21133 Default is @code{frame}.
21136 Set size of graph video. For the syntax of this option, check the
21137 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21138 The default value is @code{900x256}.
21140 The foreground color expressions can use the following variables:
21143 Minimal value of metadata value.
21146 Maximal value of metadata value.
21149 Current metadata key value.
21152 The color is defined as 0xAABBGGRR.
21155 Example using metadata from @ref{signalstats} filter:
21157 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
21160 Example using metadata from @ref{ebur128} filter:
21162 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
21168 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
21169 level. By default, it logs a message at a frequency of 10Hz with the
21170 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
21171 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
21173 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
21174 sample format is double-precision floating point. The input stream will be converted to
21175 this specification, if needed. Users may need to insert aformat and/or aresample filters
21176 after this filter to obtain the original parameters.
21178 The filter also has a video output (see the @var{video} option) with a real
21179 time graph to observe the loudness evolution. The graphic contains the logged
21180 message mentioned above, so it is not printed anymore when this option is set,
21181 unless the verbose logging is set. The main graphing area contains the
21182 short-term loudness (3 seconds of analysis), and the gauge on the right is for
21183 the momentary loudness (400 milliseconds), but can optionally be configured
21184 to instead display short-term loudness (see @var{gauge}).
21186 The green area marks a +/- 1LU target range around the target loudness
21187 (-23LUFS by default, unless modified through @var{target}).
21189 More information about the Loudness Recommendation EBU R128 on
21190 @url{http://tech.ebu.ch/loudness}.
21192 The filter accepts the following options:
21197 Activate the video output. The audio stream is passed unchanged whether this
21198 option is set or no. The video stream will be the first output stream if
21199 activated. Default is @code{0}.
21202 Set the video size. This option is for video only. For the syntax of this
21204 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21205 Default and minimum resolution is @code{640x480}.
21208 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
21209 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
21210 other integer value between this range is allowed.
21213 Set metadata injection. If set to @code{1}, the audio input will be segmented
21214 into 100ms output frames, each of them containing various loudness information
21215 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
21217 Default is @code{0}.
21220 Force the frame logging level.
21222 Available values are:
21225 information logging level
21227 verbose logging level
21230 By default, the logging level is set to @var{info}. If the @option{video} or
21231 the @option{metadata} options are set, it switches to @var{verbose}.
21236 Available modes can be cumulated (the option is a @code{flag} type). Possible
21240 Disable any peak mode (default).
21242 Enable sample-peak mode.
21244 Simple peak mode looking for the higher sample value. It logs a message
21245 for sample-peak (identified by @code{SPK}).
21247 Enable true-peak mode.
21249 If enabled, the peak lookup is done on an over-sampled version of the input
21250 stream for better peak accuracy. It logs a message for true-peak.
21251 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
21252 This mode requires a build with @code{libswresample}.
21256 Treat mono input files as "dual mono". If a mono file is intended for playback
21257 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
21258 If set to @code{true}, this option will compensate for this effect.
21259 Multi-channel input files are not affected by this option.
21262 Set a specific pan law to be used for the measurement of dual mono files.
21263 This parameter is optional, and has a default value of -3.01dB.
21266 Set a specific target level (in LUFS) used as relative zero in the visualization.
21267 This parameter is optional and has a default value of -23LUFS as specified
21268 by EBU R128. However, material published online may prefer a level of -16LUFS
21269 (e.g. for use with podcasts or video platforms).
21272 Set the value displayed by the gauge. Valid values are @code{momentary} and s
21273 @code{shortterm}. By default the momentary value will be used, but in certain
21274 scenarios it may be more useful to observe the short term value instead (e.g.
21278 Sets the display scale for the loudness. Valid parameters are @code{absolute}
21279 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
21280 video output, not the summary or continuous log output.
21283 @subsection Examples
21287 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
21289 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
21293 Run an analysis with @command{ffmpeg}:
21295 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
21299 @section interleave, ainterleave
21301 Temporally interleave frames from several inputs.
21303 @code{interleave} works with video inputs, @code{ainterleave} with audio.
21305 These filters read frames from several inputs and send the oldest
21306 queued frame to the output.
21308 Input streams must have well defined, monotonically increasing frame
21311 In order to submit one frame to output, these filters need to enqueue
21312 at least one frame for each input, so they cannot work in case one
21313 input is not yet terminated and will not receive incoming frames.
21315 For example consider the case when one input is a @code{select} filter
21316 which always drops input frames. The @code{interleave} filter will keep
21317 reading from that input, but it will never be able to send new frames
21318 to output until the input sends an end-of-stream signal.
21320 Also, depending on inputs synchronization, the filters will drop
21321 frames in case one input receives more frames than the other ones, and
21322 the queue is already filled.
21324 These filters accept the following options:
21328 Set the number of different inputs, it is 2 by default.
21331 @subsection Examples
21335 Interleave frames belonging to different streams using @command{ffmpeg}:
21337 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
21341 Add flickering blur effect:
21343 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
21347 @section metadata, ametadata
21349 Manipulate frame metadata.
21351 This filter accepts the following options:
21355 Set mode of operation of the filter.
21357 Can be one of the following:
21361 If both @code{value} and @code{key} is set, select frames
21362 which have such metadata. If only @code{key} is set, select
21363 every frame that has such key in metadata.
21366 Add new metadata @code{key} and @code{value}. If key is already available
21370 Modify value of already present key.
21373 If @code{value} is set, delete only keys that have such value.
21374 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
21378 Print key and its value if metadata was found. If @code{key} is not set print all
21379 metadata values available in frame.
21383 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
21386 Set metadata value which will be used. This option is mandatory for
21387 @code{modify} and @code{add} mode.
21390 Which function to use when comparing metadata value and @code{value}.
21392 Can be one of following:
21396 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
21399 Values are interpreted as strings, returns true if metadata value starts with
21400 the @code{value} option string.
21403 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
21406 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
21409 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
21412 Values are interpreted as floats, returns true if expression from option @code{expr}
21417 Set expression which is used when @code{function} is set to @code{expr}.
21418 The expression is evaluated through the eval API and can contain the following
21423 Float representation of @code{value} from metadata key.
21426 Float representation of @code{value} as supplied by user in @code{value} option.
21430 If specified in @code{print} mode, output is written to the named file. Instead of
21431 plain filename any writable url can be specified. Filename ``-'' is a shorthand
21432 for standard output. If @code{file} option is not set, output is written to the log
21433 with AV_LOG_INFO loglevel.
21437 @subsection Examples
21441 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
21444 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
21447 Print silencedetect output to file @file{metadata.txt}.
21449 silencedetect,ametadata=mode=print:file=metadata.txt
21452 Direct all metadata to a pipe with file descriptor 4.
21454 metadata=mode=print:file='pipe\:4'
21458 @section perms, aperms
21460 Set read/write permissions for the output frames.
21462 These filters are mainly aimed at developers to test direct path in the
21463 following filter in the filtergraph.
21465 The filters accept the following options:
21469 Select the permissions mode.
21471 It accepts the following values:
21474 Do nothing. This is the default.
21476 Set all the output frames read-only.
21478 Set all the output frames directly writable.
21480 Make the frame read-only if writable, and writable if read-only.
21482 Set each output frame read-only or writable randomly.
21486 Set the seed for the @var{random} mode, must be an integer included between
21487 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
21488 @code{-1}, the filter will try to use a good random seed on a best effort
21492 Note: in case of auto-inserted filter between the permission filter and the
21493 following one, the permission might not be received as expected in that
21494 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
21495 perms/aperms filter can avoid this problem.
21497 @section realtime, arealtime
21499 Slow down filtering to match real time approximately.
21501 These filters will pause the filtering for a variable amount of time to
21502 match the output rate with the input timestamps.
21503 They are similar to the @option{re} option to @code{ffmpeg}.
21505 They accept the following options:
21509 Time limit for the pauses. Any pause longer than that will be considered
21510 a timestamp discontinuity and reset the timer. Default is 2 seconds.
21512 Speed factor for processing. The value must be a float larger than zero.
21513 Values larger than 1.0 will result in faster than realtime processing,
21514 smaller will slow processing down. The @var{limit} is automatically adapted
21515 accordingly. Default is 1.0.
21517 A processing speed faster than what is possible without these filters cannot
21522 @section select, aselect
21524 Select frames to pass in output.
21526 This filter accepts the following options:
21531 Set expression, which is evaluated for each input frame.
21533 If the expression is evaluated to zero, the frame is discarded.
21535 If the evaluation result is negative or NaN, the frame is sent to the
21536 first output; otherwise it is sent to the output with index
21537 @code{ceil(val)-1}, assuming that the input index starts from 0.
21539 For example a value of @code{1.2} corresponds to the output with index
21540 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
21543 Set the number of outputs. The output to which to send the selected
21544 frame is based on the result of the evaluation. Default value is 1.
21547 The expression can contain the following constants:
21551 The (sequential) number of the filtered frame, starting from 0.
21554 The (sequential) number of the selected frame, starting from 0.
21556 @item prev_selected_n
21557 The sequential number of the last selected frame. It's NAN if undefined.
21560 The timebase of the input timestamps.
21563 The PTS (Presentation TimeStamp) of the filtered video frame,
21564 expressed in @var{TB} units. It's NAN if undefined.
21567 The PTS of the filtered video frame,
21568 expressed in seconds. It's NAN if undefined.
21571 The PTS of the previously filtered video frame. It's NAN if undefined.
21573 @item prev_selected_pts
21574 The PTS of the last previously filtered video frame. It's NAN if undefined.
21576 @item prev_selected_t
21577 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
21580 The PTS of the first video frame in the video. It's NAN if undefined.
21583 The time of the first video frame in the video. It's NAN if undefined.
21585 @item pict_type @emph{(video only)}
21586 The type of the filtered frame. It can assume one of the following
21598 @item interlace_type @emph{(video only)}
21599 The frame interlace type. It can assume one of the following values:
21602 The frame is progressive (not interlaced).
21604 The frame is top-field-first.
21606 The frame is bottom-field-first.
21609 @item consumed_sample_n @emph{(audio only)}
21610 the number of selected samples before the current frame
21612 @item samples_n @emph{(audio only)}
21613 the number of samples in the current frame
21615 @item sample_rate @emph{(audio only)}
21616 the input sample rate
21619 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21622 the position in the file of the filtered frame, -1 if the information
21623 is not available (e.g. for synthetic video)
21625 @item scene @emph{(video only)}
21626 value between 0 and 1 to indicate a new scene; a low value reflects a low
21627 probability for the current frame to introduce a new scene, while a higher
21628 value means the current frame is more likely to be one (see the example below)
21630 @item concatdec_select
21631 The concat demuxer can select only part of a concat input file by setting an
21632 inpoint and an outpoint, but the output packets may not be entirely contained
21633 in the selected interval. By using this variable, it is possible to skip frames
21634 generated by the concat demuxer which are not exactly contained in the selected
21637 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21638 and the @var{lavf.concat.duration} packet metadata values which are also
21639 present in the decoded frames.
21641 The @var{concatdec_select} variable is -1 if the frame pts is at least
21642 start_time and either the duration metadata is missing or the frame pts is less
21643 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21646 That basically means that an input frame is selected if its pts is within the
21647 interval set by the concat demuxer.
21651 The default value of the select expression is "1".
21653 @subsection Examples
21657 Select all frames in input:
21662 The example above is the same as:
21674 Select only I-frames:
21676 select='eq(pict_type\,I)'
21680 Select one frame every 100:
21682 select='not(mod(n\,100))'
21686 Select only frames contained in the 10-20 time interval:
21688 select=between(t\,10\,20)
21692 Select only I-frames contained in the 10-20 time interval:
21694 select=between(t\,10\,20)*eq(pict_type\,I)
21698 Select frames with a minimum distance of 10 seconds:
21700 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21704 Use aselect to select only audio frames with samples number > 100:
21706 aselect='gt(samples_n\,100)'
21710 Create a mosaic of the first scenes:
21712 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21715 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21719 Send even and odd frames to separate outputs, and compose them:
21721 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21725 Select useful frames from an ffconcat file which is using inpoints and
21726 outpoints but where the source files are not intra frame only.
21728 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21732 @section sendcmd, asendcmd
21734 Send commands to filters in the filtergraph.
21736 These filters read commands to be sent to other filters in the
21739 @code{sendcmd} must be inserted between two video filters,
21740 @code{asendcmd} must be inserted between two audio filters, but apart
21741 from that they act the same way.
21743 The specification of commands can be provided in the filter arguments
21744 with the @var{commands} option, or in a file specified by the
21745 @var{filename} option.
21747 These filters accept the following options:
21750 Set the commands to be read and sent to the other filters.
21752 Set the filename of the commands to be read and sent to the other
21756 @subsection Commands syntax
21758 A commands description consists of a sequence of interval
21759 specifications, comprising a list of commands to be executed when a
21760 particular event related to that interval occurs. The occurring event
21761 is typically the current frame time entering or leaving a given time
21764 An interval is specified by the following syntax:
21766 @var{START}[-@var{END}] @var{COMMANDS};
21769 The time interval is specified by the @var{START} and @var{END} times.
21770 @var{END} is optional and defaults to the maximum time.
21772 The current frame time is considered within the specified interval if
21773 it is included in the interval [@var{START}, @var{END}), that is when
21774 the time is greater or equal to @var{START} and is lesser than
21777 @var{COMMANDS} consists of a sequence of one or more command
21778 specifications, separated by ",", relating to that interval. The
21779 syntax of a command specification is given by:
21781 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21784 @var{FLAGS} is optional and specifies the type of events relating to
21785 the time interval which enable sending the specified command, and must
21786 be a non-null sequence of identifier flags separated by "+" or "|" and
21787 enclosed between "[" and "]".
21789 The following flags are recognized:
21792 The command is sent when the current frame timestamp enters the
21793 specified interval. In other words, the command is sent when the
21794 previous frame timestamp was not in the given interval, and the
21798 The command is sent when the current frame timestamp leaves the
21799 specified interval. In other words, the command is sent when the
21800 previous frame timestamp was in the given interval, and the
21804 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21807 @var{TARGET} specifies the target of the command, usually the name of
21808 the filter class or a specific filter instance name.
21810 @var{COMMAND} specifies the name of the command for the target filter.
21812 @var{ARG} is optional and specifies the optional list of argument for
21813 the given @var{COMMAND}.
21815 Between one interval specification and another, whitespaces, or
21816 sequences of characters starting with @code{#} until the end of line,
21817 are ignored and can be used to annotate comments.
21819 A simplified BNF description of the commands specification syntax
21822 @var{COMMAND_FLAG} ::= "enter" | "leave"
21823 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21824 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21825 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21826 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21827 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21830 @subsection Examples
21834 Specify audio tempo change at second 4:
21836 asendcmd=c='4.0 atempo tempo 1.5',atempo
21840 Target a specific filter instance:
21842 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21846 Specify a list of drawtext and hue commands in a file.
21848 # show text in the interval 5-10
21849 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21850 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21852 # desaturate the image in the interval 15-20
21853 15.0-20.0 [enter] hue s 0,
21854 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21856 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21858 # apply an exponential saturation fade-out effect, starting from time 25
21859 25 [enter] hue s exp(25-t)
21862 A filtergraph allowing to read and process the above command list
21863 stored in a file @file{test.cmd}, can be specified with:
21865 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21870 @section setpts, asetpts
21872 Change the PTS (presentation timestamp) of the input frames.
21874 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21876 This filter accepts the following options:
21881 The expression which is evaluated for each frame to construct its timestamp.
21885 The expression is evaluated through the eval API and can contain the following
21889 @item FRAME_RATE, FR
21890 frame rate, only defined for constant frame-rate video
21893 The presentation timestamp in input
21896 The count of the input frame for video or the number of consumed samples,
21897 not including the current frame for audio, starting from 0.
21899 @item NB_CONSUMED_SAMPLES
21900 The number of consumed samples, not including the current frame (only
21903 @item NB_SAMPLES, S
21904 The number of samples in the current frame (only audio)
21906 @item SAMPLE_RATE, SR
21907 The audio sample rate.
21910 The PTS of the first frame.
21913 the time in seconds of the first frame
21916 State whether the current frame is interlaced.
21919 the time in seconds of the current frame
21922 original position in the file of the frame, or undefined if undefined
21923 for the current frame
21926 The previous input PTS.
21929 previous input time in seconds
21932 The previous output PTS.
21935 previous output time in seconds
21938 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21942 The wallclock (RTC) time at the start of the movie in microseconds.
21945 The timebase of the input timestamps.
21949 @subsection Examples
21953 Start counting PTS from zero
21955 setpts=PTS-STARTPTS
21959 Apply fast motion effect:
21965 Apply slow motion effect:
21971 Set fixed rate of 25 frames per second:
21977 Set fixed rate 25 fps with some jitter:
21979 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21983 Apply an offset of 10 seconds to the input PTS:
21989 Generate timestamps from a "live source" and rebase onto the current timebase:
21991 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21995 Generate timestamps by counting samples:
22004 Force color range for the output video frame.
22006 The @code{setrange} filter marks the color range property for the
22007 output frames. It does not change the input frame, but only sets the
22008 corresponding property, which affects how the frame is treated by
22011 The filter accepts the following options:
22016 Available values are:
22020 Keep the same color range property.
22022 @item unspecified, unknown
22023 Set the color range as unspecified.
22025 @item limited, tv, mpeg
22026 Set the color range as limited.
22028 @item full, pc, jpeg
22029 Set the color range as full.
22033 @section settb, asettb
22035 Set the timebase to use for the output frames timestamps.
22036 It is mainly useful for testing timebase configuration.
22038 It accepts the following parameters:
22043 The expression which is evaluated into the output timebase.
22047 The value for @option{tb} is an arithmetic expression representing a
22048 rational. The expression can contain the constants "AVTB" (the default
22049 timebase), "intb" (the input timebase) and "sr" (the sample rate,
22050 audio only). Default value is "intb".
22052 @subsection Examples
22056 Set the timebase to 1/25:
22062 Set the timebase to 1/10:
22068 Set the timebase to 1001/1000:
22074 Set the timebase to 2*intb:
22080 Set the default timebase value:
22087 Convert input audio to a video output representing frequency spectrum
22088 logarithmically using Brown-Puckette constant Q transform algorithm with
22089 direct frequency domain coefficient calculation (but the transform itself
22090 is not really constant Q, instead the Q factor is actually variable/clamped),
22091 with musical tone scale, from E0 to D#10.
22093 The filter accepts the following options:
22097 Specify the video size for the output. It must be even. For the syntax of this option,
22098 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22099 Default value is @code{1920x1080}.
22102 Set the output frame rate. Default value is @code{25}.
22105 Set the bargraph height. It must be even. Default value is @code{-1} which
22106 computes the bargraph height automatically.
22109 Set the axis height. It must be even. Default value is @code{-1} which computes
22110 the axis height automatically.
22113 Set the sonogram height. It must be even. Default value is @code{-1} which
22114 computes the sonogram height automatically.
22117 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
22118 instead. Default value is @code{1}.
22120 @item sono_v, volume
22121 Specify the sonogram volume expression. It can contain variables:
22124 the @var{bar_v} evaluated expression
22125 @item frequency, freq, f
22126 the frequency where it is evaluated
22127 @item timeclamp, tc
22128 the value of @var{timeclamp} option
22132 @item a_weighting(f)
22133 A-weighting of equal loudness
22134 @item b_weighting(f)
22135 B-weighting of equal loudness
22136 @item c_weighting(f)
22137 C-weighting of equal loudness.
22139 Default value is @code{16}.
22141 @item bar_v, volume2
22142 Specify the bargraph volume expression. It can contain variables:
22145 the @var{sono_v} evaluated expression
22146 @item frequency, freq, f
22147 the frequency where it is evaluated
22148 @item timeclamp, tc
22149 the value of @var{timeclamp} option
22153 @item a_weighting(f)
22154 A-weighting of equal loudness
22155 @item b_weighting(f)
22156 B-weighting of equal loudness
22157 @item c_weighting(f)
22158 C-weighting of equal loudness.
22160 Default value is @code{sono_v}.
22162 @item sono_g, gamma
22163 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
22164 higher gamma makes the spectrum having more range. Default value is @code{3}.
22165 Acceptable range is @code{[1, 7]}.
22167 @item bar_g, gamma2
22168 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
22172 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
22173 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
22175 @item timeclamp, tc
22176 Specify the transform timeclamp. At low frequency, there is trade-off between
22177 accuracy in time domain and frequency domain. If timeclamp is lower,
22178 event in time domain is represented more accurately (such as fast bass drum),
22179 otherwise event in frequency domain is represented more accurately
22180 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
22183 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
22184 limits future samples by applying asymmetric windowing in time domain, useful
22185 when low latency is required. Accepted range is @code{[0, 1]}.
22188 Specify the transform base frequency. Default value is @code{20.01523126408007475},
22189 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
22192 Specify the transform end frequency. Default value is @code{20495.59681441799654},
22193 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
22196 This option is deprecated and ignored.
22199 Specify the transform length in time domain. Use this option to control accuracy
22200 trade-off between time domain and frequency domain at every frequency sample.
22201 It can contain variables:
22203 @item frequency, freq, f
22204 the frequency where it is evaluated
22205 @item timeclamp, tc
22206 the value of @var{timeclamp} option.
22208 Default value is @code{384*tc/(384+tc*f)}.
22211 Specify the transform count for every video frame. Default value is @code{6}.
22212 Acceptable range is @code{[1, 30]}.
22215 Specify the transform count for every single pixel. Default value is @code{0},
22216 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
22219 Specify font file for use with freetype to draw the axis. If not specified,
22220 use embedded font. Note that drawing with font file or embedded font is not
22221 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
22225 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
22226 The : in the pattern may be replaced by | to avoid unnecessary escaping.
22229 Specify font color expression. This is arithmetic expression that should return
22230 integer value 0xRRGGBB. It can contain variables:
22232 @item frequency, freq, f
22233 the frequency where it is evaluated
22234 @item timeclamp, tc
22235 the value of @var{timeclamp} option
22240 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
22241 @item r(x), g(x), b(x)
22242 red, green, and blue value of intensity x.
22244 Default value is @code{st(0, (midi(f)-59.5)/12);
22245 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
22246 r(1-ld(1)) + b(ld(1))}.
22249 Specify image file to draw the axis. This option override @var{fontfile} and
22250 @var{fontcolor} option.
22253 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
22254 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
22255 Default value is @code{1}.
22258 Set colorspace. The accepted values are:
22261 Unspecified (default)
22270 BT.470BG or BT.601-6 625
22273 SMPTE-170M or BT.601-6 525
22279 BT.2020 with non-constant luminance
22284 Set spectrogram color scheme. This is list of floating point values with format
22285 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
22286 The default is @code{1|0.5|0|0|0.5|1}.
22290 @subsection Examples
22294 Playing audio while showing the spectrum:
22296 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
22300 Same as above, but with frame rate 30 fps:
22302 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
22306 Playing at 1280x720:
22308 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
22312 Disable sonogram display:
22318 A1 and its harmonics: A1, A2, (near)E3, A3:
22320 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),
22321 asplit[a][out1]; [a] showcqt [out0]'
22325 Same as above, but with more accuracy in frequency domain:
22327 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),
22328 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
22334 bar_v=10:sono_v=bar_v*a_weighting(f)
22338 Custom gamma, now spectrum is linear to the amplitude.
22344 Custom tlength equation:
22346 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)))'
22350 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
22352 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
22356 Custom font using fontconfig:
22358 font='Courier New,Monospace,mono|bold'
22362 Custom frequency range with custom axis using image file:
22364 axisfile=myaxis.png:basefreq=40:endfreq=10000
22370 Convert input audio to video output representing the audio power spectrum.
22371 Audio amplitude is on Y-axis while frequency is on X-axis.
22373 The filter accepts the following options:
22377 Specify size of video. For the syntax of this option, check the
22378 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22379 Default is @code{1024x512}.
22383 This set how each frequency bin will be represented.
22385 It accepts the following values:
22391 Default is @code{bar}.
22394 Set amplitude scale.
22396 It accepts the following values:
22410 Default is @code{log}.
22413 Set frequency scale.
22415 It accepts the following values:
22424 Reverse logarithmic scale.
22426 Default is @code{lin}.
22429 Set window size. Allowed range is from 16 to 65536.
22431 Default is @code{2048}
22434 Set windowing function.
22436 It accepts the following values:
22459 Default is @code{hanning}.
22462 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22463 which means optimal overlap for selected window function will be picked.
22466 Set time averaging. Setting this to 0 will display current maximal peaks.
22467 Default is @code{1}, which means time averaging is disabled.
22470 Specify list of colors separated by space or by '|' which will be used to
22471 draw channel frequencies. Unrecognized or missing colors will be replaced
22475 Set channel display mode.
22477 It accepts the following values:
22482 Default is @code{combined}.
22485 Set minimum amplitude used in @code{log} amplitude scaler.
22489 @section showspatial
22491 Convert stereo input audio to a video output, representing the spatial relationship
22492 between two channels.
22494 The filter accepts the following options:
22498 Specify the video size for the output. For the syntax of this option, check the
22499 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22500 Default value is @code{512x512}.
22503 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
22506 Set window function.
22508 It accepts the following values:
22533 Default value is @code{hann}.
22536 Set ratio of overlap window. Default value is @code{0.5}.
22537 When value is @code{1} overlap is set to recommended size for specific
22538 window function currently used.
22541 @anchor{showspectrum}
22542 @section showspectrum
22544 Convert input audio to a video output, representing the audio frequency
22547 The filter accepts the following options:
22551 Specify the video size for the output. For the syntax of this option, check the
22552 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22553 Default value is @code{640x512}.
22556 Specify how the spectrum should slide along the window.
22558 It accepts the following values:
22561 the samples start again on the left when they reach the right
22563 the samples scroll from right to left
22565 frames are only produced when the samples reach the right
22567 the samples scroll from left to right
22570 Default value is @code{replace}.
22573 Specify display mode.
22575 It accepts the following values:
22578 all channels are displayed in the same row
22580 all channels are displayed in separate rows
22583 Default value is @samp{combined}.
22586 Specify display color mode.
22588 It accepts the following values:
22591 each channel is displayed in a separate color
22593 each channel is displayed using the same color scheme
22595 each channel is displayed using the rainbow color scheme
22597 each channel is displayed using the moreland color scheme
22599 each channel is displayed using the nebulae color scheme
22601 each channel is displayed using the fire color scheme
22603 each channel is displayed using the fiery color scheme
22605 each channel is displayed using the fruit color scheme
22607 each channel is displayed using the cool color scheme
22609 each channel is displayed using the magma color scheme
22611 each channel is displayed using the green color scheme
22613 each channel is displayed using the viridis color scheme
22615 each channel is displayed using the plasma color scheme
22617 each channel is displayed using the cividis color scheme
22619 each channel is displayed using the terrain color scheme
22622 Default value is @samp{channel}.
22625 Specify scale used for calculating intensity color values.
22627 It accepts the following values:
22632 square root, default
22643 Default value is @samp{sqrt}.
22646 Specify frequency scale.
22648 It accepts the following values:
22656 Default value is @samp{lin}.
22659 Set saturation modifier for displayed colors. Negative values provide
22660 alternative color scheme. @code{0} is no saturation at all.
22661 Saturation must be in [-10.0, 10.0] range.
22662 Default value is @code{1}.
22665 Set window function.
22667 It accepts the following values:
22692 Default value is @code{hann}.
22695 Set orientation of time vs frequency axis. Can be @code{vertical} or
22696 @code{horizontal}. Default is @code{vertical}.
22699 Set ratio of overlap window. Default value is @code{0}.
22700 When value is @code{1} overlap is set to recommended size for specific
22701 window function currently used.
22704 Set scale gain for calculating intensity color values.
22705 Default value is @code{1}.
22708 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22711 Set color rotation, must be in [-1.0, 1.0] range.
22712 Default value is @code{0}.
22715 Set start frequency from which to display spectrogram. Default is @code{0}.
22718 Set stop frequency to which to display spectrogram. Default is @code{0}.
22721 Set upper frame rate limit. Default is @code{auto}, unlimited.
22724 Draw time and frequency axes and legends. Default is disabled.
22727 The usage is very similar to the showwaves filter; see the examples in that
22730 @subsection Examples
22734 Large window with logarithmic color scaling:
22736 showspectrum=s=1280x480:scale=log
22740 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22742 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22743 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22747 @section showspectrumpic
22749 Convert input audio to a single video frame, representing the audio frequency
22752 The filter accepts the following options:
22756 Specify the video size for the output. For the syntax of this option, check the
22757 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22758 Default value is @code{4096x2048}.
22761 Specify display mode.
22763 It accepts the following values:
22766 all channels are displayed in the same row
22768 all channels are displayed in separate rows
22770 Default value is @samp{combined}.
22773 Specify display color mode.
22775 It accepts the following values:
22778 each channel is displayed in a separate color
22780 each channel is displayed using the same color scheme
22782 each channel is displayed using the rainbow color scheme
22784 each channel is displayed using the moreland color scheme
22786 each channel is displayed using the nebulae color scheme
22788 each channel is displayed using the fire color scheme
22790 each channel is displayed using the fiery color scheme
22792 each channel is displayed using the fruit color scheme
22794 each channel is displayed using the cool color scheme
22796 each channel is displayed using the magma color scheme
22798 each channel is displayed using the green color scheme
22800 each channel is displayed using the viridis color scheme
22802 each channel is displayed using the plasma color scheme
22804 each channel is displayed using the cividis color scheme
22806 each channel is displayed using the terrain color scheme
22808 Default value is @samp{intensity}.
22811 Specify scale used for calculating intensity color values.
22813 It accepts the following values:
22818 square root, default
22828 Default value is @samp{log}.
22831 Specify frequency scale.
22833 It accepts the following values:
22841 Default value is @samp{lin}.
22844 Set saturation modifier for displayed colors. Negative values provide
22845 alternative color scheme. @code{0} is no saturation at all.
22846 Saturation must be in [-10.0, 10.0] range.
22847 Default value is @code{1}.
22850 Set window function.
22852 It accepts the following values:
22876 Default value is @code{hann}.
22879 Set orientation of time vs frequency axis. Can be @code{vertical} or
22880 @code{horizontal}. Default is @code{vertical}.
22883 Set scale gain for calculating intensity color values.
22884 Default value is @code{1}.
22887 Draw time and frequency axes and legends. Default is enabled.
22890 Set color rotation, must be in [-1.0, 1.0] range.
22891 Default value is @code{0}.
22894 Set start frequency from which to display spectrogram. Default is @code{0}.
22897 Set stop frequency to which to display spectrogram. Default is @code{0}.
22900 @subsection Examples
22904 Extract an audio spectrogram of a whole audio track
22905 in a 1024x1024 picture using @command{ffmpeg}:
22907 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22911 @section showvolume
22913 Convert input audio volume to a video output.
22915 The filter accepts the following options:
22922 Set border width, allowed range is [0, 5]. Default is 1.
22925 Set channel width, allowed range is [80, 8192]. Default is 400.
22928 Set channel height, allowed range is [1, 900]. Default is 20.
22931 Set fade, allowed range is [0, 1]. Default is 0.95.
22934 Set volume color expression.
22936 The expression can use the following variables:
22940 Current max volume of channel in dB.
22946 Current channel number, starting from 0.
22950 If set, displays channel names. Default is enabled.
22953 If set, displays volume values. Default is enabled.
22956 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22957 default is @code{h}.
22960 Set step size, allowed range is [0, 5]. Default is 0, which means
22964 Set background opacity, allowed range is [0, 1]. Default is 0.
22967 Set metering mode, can be peak: @code{p} or rms: @code{r},
22968 default is @code{p}.
22971 Set display scale, can be linear: @code{lin} or log: @code{log},
22972 default is @code{lin}.
22976 If set to > 0., display a line for the max level
22977 in the previous seconds.
22978 default is disabled: @code{0.}
22981 The color of the max line. Use when @code{dm} option is set to > 0.
22982 default is: @code{orange}
22987 Convert input audio to a video output, representing the samples waves.
22989 The filter accepts the following options:
22993 Specify the video size for the output. For the syntax of this option, check the
22994 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22995 Default value is @code{600x240}.
23000 Available values are:
23003 Draw a point for each sample.
23006 Draw a vertical line for each sample.
23009 Draw a point for each sample and a line between them.
23012 Draw a centered vertical line for each sample.
23015 Default value is @code{point}.
23018 Set the number of samples which are printed on the same column. A
23019 larger value will decrease the frame rate. Must be a positive
23020 integer. This option can be set only if the value for @var{rate}
23021 is not explicitly specified.
23024 Set the (approximate) output frame rate. This is done by setting the
23025 option @var{n}. Default value is "25".
23027 @item split_channels
23028 Set if channels should be drawn separately or overlap. Default value is 0.
23031 Set colors separated by '|' which are going to be used for drawing of each channel.
23034 Set amplitude scale.
23036 Available values are:
23054 Set the draw mode. This is mostly useful to set for high @var{n}.
23056 Available values are:
23059 Scale pixel values for each drawn sample.
23062 Draw every sample directly.
23065 Default value is @code{scale}.
23068 @subsection Examples
23072 Output the input file audio and the corresponding video representation
23075 amovie=a.mp3,asplit[out0],showwaves[out1]
23079 Create a synthetic signal and show it with showwaves, forcing a
23080 frame rate of 30 frames per second:
23082 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
23086 @section showwavespic
23088 Convert input audio to a single video frame, representing the samples waves.
23090 The filter accepts the following options:
23094 Specify the video size for the output. For the syntax of this option, check the
23095 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
23096 Default value is @code{600x240}.
23098 @item split_channels
23099 Set if channels should be drawn separately or overlap. Default value is 0.
23102 Set colors separated by '|' which are going to be used for drawing of each channel.
23105 Set amplitude scale.
23107 Available values are:
23127 Available values are:
23130 Scale pixel values for each drawn sample.
23133 Draw every sample directly.
23136 Default value is @code{scale}.
23139 @subsection Examples
23143 Extract a channel split representation of the wave form of a whole audio track
23144 in a 1024x800 picture using @command{ffmpeg}:
23146 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
23150 @section sidedata, asidedata
23152 Delete frame side data, or select frames based on it.
23154 This filter accepts the following options:
23158 Set mode of operation of the filter.
23160 Can be one of the following:
23164 Select every frame with side data of @code{type}.
23167 Delete side data of @code{type}. If @code{type} is not set, delete all side
23173 Set side data type used with all modes. Must be set for @code{select} mode. For
23174 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
23175 in @file{libavutil/frame.h}. For example, to choose
23176 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
23180 @section spectrumsynth
23182 Sythesize audio from 2 input video spectrums, first input stream represents
23183 magnitude across time and second represents phase across time.
23184 The filter will transform from frequency domain as displayed in videos back
23185 to time domain as presented in audio output.
23187 This filter is primarily created for reversing processed @ref{showspectrum}
23188 filter outputs, but can synthesize sound from other spectrograms too.
23189 But in such case results are going to be poor if the phase data is not
23190 available, because in such cases phase data need to be recreated, usually
23191 it's just recreated from random noise.
23192 For best results use gray only output (@code{channel} color mode in
23193 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
23194 @code{lin} scale for phase video. To produce phase, for 2nd video, use
23195 @code{data} option. Inputs videos should generally use @code{fullframe}
23196 slide mode as that saves resources needed for decoding video.
23198 The filter accepts the following options:
23202 Specify sample rate of output audio, the sample rate of audio from which
23203 spectrum was generated may differ.
23206 Set number of channels represented in input video spectrums.
23209 Set scale which was used when generating magnitude input spectrum.
23210 Can be @code{lin} or @code{log}. Default is @code{log}.
23213 Set slide which was used when generating inputs spectrums.
23214 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
23215 Default is @code{fullframe}.
23218 Set window function used for resynthesis.
23221 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
23222 which means optimal overlap for selected window function will be picked.
23225 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
23226 Default is @code{vertical}.
23229 @subsection Examples
23233 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
23234 then resynthesize videos back to audio with spectrumsynth:
23236 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
23237 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
23238 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
23242 @section split, asplit
23244 Split input into several identical outputs.
23246 @code{asplit} works with audio input, @code{split} with video.
23248 The filter accepts a single parameter which specifies the number of outputs. If
23249 unspecified, it defaults to 2.
23251 @subsection Examples
23255 Create two separate outputs from the same input:
23257 [in] split [out0][out1]
23261 To create 3 or more outputs, you need to specify the number of
23264 [in] asplit=3 [out0][out1][out2]
23268 Create two separate outputs from the same input, one cropped and
23271 [in] split [splitout1][splitout2];
23272 [splitout1] crop=100:100:0:0 [cropout];
23273 [splitout2] pad=200:200:100:100 [padout];
23277 Create 5 copies of the input audio with @command{ffmpeg}:
23279 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
23285 Receive commands sent through a libzmq client, and forward them to
23286 filters in the filtergraph.
23288 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
23289 must be inserted between two video filters, @code{azmq} between two
23290 audio filters. Both are capable to send messages to any filter type.
23292 To enable these filters you need to install the libzmq library and
23293 headers and configure FFmpeg with @code{--enable-libzmq}.
23295 For more information about libzmq see:
23296 @url{http://www.zeromq.org/}
23298 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
23299 receives messages sent through a network interface defined by the
23300 @option{bind_address} (or the abbreviation "@option{b}") option.
23301 Default value of this option is @file{tcp://localhost:5555}. You may
23302 want to alter this value to your needs, but do not forget to escape any
23303 ':' signs (see @ref{filtergraph escaping}).
23305 The received message must be in the form:
23307 @var{TARGET} @var{COMMAND} [@var{ARG}]
23310 @var{TARGET} specifies the target of the command, usually the name of
23311 the filter class or a specific filter instance name. The default
23312 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
23313 but you can override this by using the @samp{filter_name@@id} syntax
23314 (see @ref{Filtergraph syntax}).
23316 @var{COMMAND} specifies the name of the command for the target filter.
23318 @var{ARG} is optional and specifies the optional argument list for the
23319 given @var{COMMAND}.
23321 Upon reception, the message is processed and the corresponding command
23322 is injected into the filtergraph. Depending on the result, the filter
23323 will send a reply to the client, adopting the format:
23325 @var{ERROR_CODE} @var{ERROR_REASON}
23329 @var{MESSAGE} is optional.
23331 @subsection Examples
23333 Look at @file{tools/zmqsend} for an example of a zmq client which can
23334 be used to send commands processed by these filters.
23336 Consider the following filtergraph generated by @command{ffplay}.
23337 In this example the last overlay filter has an instance name. All other
23338 filters will have default instance names.
23341 ffplay -dumpgraph 1 -f lavfi "
23342 color=s=100x100:c=red [l];
23343 color=s=100x100:c=blue [r];
23344 nullsrc=s=200x100, zmq [bg];
23345 [bg][l] overlay [bg+l];
23346 [bg+l][r] overlay@@my=x=100 "
23349 To change the color of the left side of the video, the following
23350 command can be used:
23352 echo Parsed_color_0 c yellow | tools/zmqsend
23355 To change the right side:
23357 echo Parsed_color_1 c pink | tools/zmqsend
23360 To change the position of the right side:
23362 echo overlay@@my x 150 | tools/zmqsend
23366 @c man end MULTIMEDIA FILTERS
23368 @chapter Multimedia Sources
23369 @c man begin MULTIMEDIA SOURCES
23371 Below is a description of the currently available multimedia sources.
23375 This is the same as @ref{movie} source, except it selects an audio
23381 Read audio and/or video stream(s) from a movie container.
23383 It accepts the following parameters:
23387 The name of the resource to read (not necessarily a file; it can also be a
23388 device or a stream accessed through some protocol).
23390 @item format_name, f
23391 Specifies the format assumed for the movie to read, and can be either
23392 the name of a container or an input device. If not specified, the
23393 format is guessed from @var{movie_name} or by probing.
23395 @item seek_point, sp
23396 Specifies the seek point in seconds. The frames will be output
23397 starting from this seek point. The parameter is evaluated with
23398 @code{av_strtod}, so the numerical value may be suffixed by an IS
23399 postfix. The default value is "0".
23402 Specifies the streams to read. Several streams can be specified,
23403 separated by "+". The source will then have as many outputs, in the
23404 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
23405 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
23406 respectively the default (best suited) video and audio stream. Default
23407 is "dv", or "da" if the filter is called as "amovie".
23409 @item stream_index, si
23410 Specifies the index of the video stream to read. If the value is -1,
23411 the most suitable video stream will be automatically selected. The default
23412 value is "-1". Deprecated. If the filter is called "amovie", it will select
23413 audio instead of video.
23416 Specifies how many times to read the stream in sequence.
23417 If the value is 0, the stream will be looped infinitely.
23418 Default value is "1".
23420 Note that when the movie is looped the source timestamps are not
23421 changed, so it will generate non monotonically increasing timestamps.
23423 @item discontinuity
23424 Specifies the time difference between frames above which the point is
23425 considered a timestamp discontinuity which is removed by adjusting the later
23429 It allows overlaying a second video on top of the main input of
23430 a filtergraph, as shown in this graph:
23432 input -----------> deltapts0 --> overlay --> output
23435 movie --> scale--> deltapts1 -------+
23437 @subsection Examples
23441 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
23442 on top of the input labelled "in":
23444 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
23445 [in] setpts=PTS-STARTPTS [main];
23446 [main][over] overlay=16:16 [out]
23450 Read from a video4linux2 device, and overlay it on top of the input
23453 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
23454 [in] setpts=PTS-STARTPTS [main];
23455 [main][over] overlay=16:16 [out]
23459 Read the first video stream and the audio stream with id 0x81 from
23460 dvd.vob; the video is connected to the pad named "video" and the audio is
23461 connected to the pad named "audio":
23463 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
23467 @subsection Commands
23469 Both movie and amovie support the following commands:
23472 Perform seek using "av_seek_frame".
23473 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
23476 @var{stream_index}: If stream_index is -1, a default
23477 stream is selected, and @var{timestamp} is automatically converted
23478 from AV_TIME_BASE units to the stream specific time_base.
23480 @var{timestamp}: Timestamp in AVStream.time_base units
23481 or, if no stream is specified, in AV_TIME_BASE units.
23483 @var{flags}: Flags which select direction and seeking mode.
23487 Get movie duration in AV_TIME_BASE units.
23491 @c man end MULTIMEDIA SOURCES