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})
1129 It accepts the following values:
1145 Default is @code{w4096}
1148 Set window function. Default is @code{hann}.
1151 Set window overlap. If set to 1, the recommended overlap for selected
1152 window function will be picked. Default is @code{0.75}.
1155 @subsection Examples
1159 Leave almost only low frequencies in audio:
1161 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1168 Apply an arbitrary Frequency Impulse Response filter.
1170 This filter is designed for applying long FIR filters,
1171 up to 60 seconds long.
1173 It can be used as component for digital crossover filters,
1174 room equalization, cross talk cancellation, wavefield synthesis,
1175 auralization, ambiophonics, ambisonics and spatialization.
1177 This filter uses second stream as FIR coefficients.
1178 If second stream holds single channel, it will be used
1179 for all input channels in first stream, otherwise
1180 number of channels in second stream must be same as
1181 number of channels in first stream.
1183 It accepts the following parameters:
1187 Set dry gain. This sets input gain.
1190 Set wet gain. This sets final output gain.
1193 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1196 Enable applying gain measured from power of IR.
1198 Set which approach to use for auto gain measurement.
1202 Do not apply any gain.
1205 select peak gain, very conservative approach. This is default value.
1208 select DC gain, limited application.
1211 select gain to noise approach, this is most popular one.
1215 Set gain to be applied to IR coefficients before filtering.
1216 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1219 Set format of IR stream. Can be @code{mono} or @code{input}.
1220 Default is @code{input}.
1223 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1224 Allowed range is 0.1 to 60 seconds.
1227 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1228 By default it is disabled.
1231 Set for which IR channel to display frequency response. By default is first channel
1232 displayed. This option is used only when @var{response} is enabled.
1235 Set video stream size. This option is used only when @var{response} is enabled.
1238 Set video stream frame rate. This option is used only when @var{response} is enabled.
1241 Set minimal partition size used for convolution. Default is @var{8192}.
1242 Allowed range is from @var{8} to @var{32768}.
1243 Lower values decreases latency at cost of higher CPU usage.
1246 Set maximal partition size used for convolution. Default is @var{8192}.
1247 Allowed range is from @var{8} to @var{32768}.
1248 Lower values may increase CPU usage.
1251 @subsection Examples
1255 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1257 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1264 Set output format constraints for the input audio. The framework will
1265 negotiate the most appropriate format to minimize conversions.
1267 It accepts the following parameters:
1271 A '|'-separated list of requested sample formats.
1274 A '|'-separated list of requested sample rates.
1276 @item channel_layouts
1277 A '|'-separated list of requested channel layouts.
1279 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1280 for the required syntax.
1283 If a parameter is omitted, all values are allowed.
1285 Force the output to either unsigned 8-bit or signed 16-bit stereo
1287 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1292 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1293 processing reduces disturbing noise between useful signals.
1295 Gating is done by detecting the volume below a chosen level @var{threshold}
1296 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1297 floor is set via @var{range}. Because an exact manipulation of the signal
1298 would cause distortion of the waveform the reduction can be levelled over
1299 time. This is done by setting @var{attack} and @var{release}.
1301 @var{attack} determines how long the signal has to fall below the threshold
1302 before any reduction will occur and @var{release} sets the time the signal
1303 has to rise above the threshold to reduce the reduction again.
1304 Shorter signals than the chosen attack time will be left untouched.
1308 Set input level before filtering.
1309 Default is 1. Allowed range is from 0.015625 to 64.
1312 Set the mode of operation. Can be @code{upward} or @code{downward}.
1313 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1314 will be amplified, expanding dynamic range in upward direction.
1315 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1318 Set the level of gain reduction when the signal is below the threshold.
1319 Default is 0.06125. Allowed range is from 0 to 1.
1320 Setting this to 0 disables reduction and then filter behaves like expander.
1323 If a signal rises above this level the gain reduction is released.
1324 Default is 0.125. Allowed range is from 0 to 1.
1327 Set a ratio by which the signal is reduced.
1328 Default is 2. Allowed range is from 1 to 9000.
1331 Amount of milliseconds the signal has to rise above the threshold before gain
1333 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1336 Amount of milliseconds the signal has to fall below the threshold before the
1337 reduction is increased again. Default is 250 milliseconds.
1338 Allowed range is from 0.01 to 9000.
1341 Set amount of amplification of signal after processing.
1342 Default is 1. Allowed range is from 1 to 64.
1345 Curve the sharp knee around the threshold to enter gain reduction more softly.
1346 Default is 2.828427125. Allowed range is from 1 to 8.
1349 Choose if exact signal should be taken for detection or an RMS like one.
1350 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1353 Choose if the average level between all channels or the louder channel affects
1355 Default is @code{average}. Can be @code{average} or @code{maximum}.
1360 Apply an arbitrary Infinite Impulse Response filter.
1362 It accepts the following parameters:
1366 Set numerator/zeros coefficients.
1369 Set denominator/poles coefficients.
1381 Set coefficients format.
1387 Z-plane zeros/poles, cartesian (default)
1389 Z-plane zeros/poles, polar radians
1391 Z-plane zeros/poles, polar degrees
1395 Set kind of processing.
1396 Can be @code{d} - direct or @code{s} - serial cascading. Default is @code{s}.
1399 Set filtering precision.
1403 double-precision floating-point (default)
1405 single-precision floating-point
1413 Show IR frequency response, magnitude and phase in additional video stream.
1414 By default it is disabled.
1417 Set for which IR channel to display frequency response. By default is first channel
1418 displayed. This option is used only when @var{response} is enabled.
1421 Set video stream size. This option is used only when @var{response} is enabled.
1424 Coefficients in @code{tf} format are separated by spaces and are in ascending
1427 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1428 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1431 Different coefficients and gains can be provided for every channel, in such case
1432 use '|' to separate coefficients or gains. Last provided coefficients will be
1433 used for all remaining channels.
1435 @subsection Examples
1439 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1441 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
1445 Same as above but in @code{zp} format:
1447 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
1453 The limiter prevents an input signal from rising over a desired threshold.
1454 This limiter uses lookahead technology to prevent your signal from distorting.
1455 It means that there is a small delay after the signal is processed. Keep in mind
1456 that the delay it produces is the attack time you set.
1458 The filter accepts the following options:
1462 Set input gain. Default is 1.
1465 Set output gain. Default is 1.
1468 Don't let signals above this level pass the limiter. Default is 1.
1471 The limiter will reach its attenuation level in this amount of time in
1472 milliseconds. Default is 5 milliseconds.
1475 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1476 Default is 50 milliseconds.
1479 When gain reduction is always needed ASC takes care of releasing to an
1480 average reduction level rather than reaching a reduction of 0 in the release
1484 Select how much the release time is affected by ASC, 0 means nearly no changes
1485 in release time while 1 produces higher release times.
1488 Auto level output signal. Default is enabled.
1489 This normalizes audio back to 0dB if enabled.
1492 Depending on picked setting it is recommended to upsample input 2x or 4x times
1493 with @ref{aresample} before applying this filter.
1497 Apply a two-pole all-pass filter with central frequency (in Hz)
1498 @var{frequency}, and filter-width @var{width}.
1499 An all-pass filter changes the audio's frequency to phase relationship
1500 without changing its frequency to amplitude relationship.
1502 The filter accepts the following options:
1506 Set frequency in Hz.
1509 Set method to specify band-width of filter.
1524 Specify the band-width of a filter in width_type units.
1527 Specify which channels to filter, by default all available are filtered.
1530 @subsection Commands
1532 This filter supports the following commands:
1535 Change allpass frequency.
1536 Syntax for the command is : "@var{frequency}"
1539 Change allpass width_type.
1540 Syntax for the command is : "@var{width_type}"
1543 Change allpass width.
1544 Syntax for the command is : "@var{width}"
1551 The filter accepts the following options:
1555 Set the number of loops. Setting this value to -1 will result in infinite loops.
1559 Set maximal number of samples. Default is 0.
1562 Set first sample of loop. Default is 0.
1568 Merge two or more audio streams into a single multi-channel stream.
1570 The filter accepts the following options:
1575 Set the number of inputs. Default is 2.
1579 If the channel layouts of the inputs are disjoint, and therefore compatible,
1580 the channel layout of the output will be set accordingly and the channels
1581 will be reordered as necessary. If the channel layouts of the inputs are not
1582 disjoint, the output will have all the channels of the first input then all
1583 the channels of the second input, in that order, and the channel layout of
1584 the output will be the default value corresponding to the total number of
1587 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1588 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1589 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1590 first input, b1 is the first channel of the second input).
1592 On the other hand, if both input are in stereo, the output channels will be
1593 in the default order: a1, a2, b1, b2, and the channel layout will be
1594 arbitrarily set to 4.0, which may or may not be the expected value.
1596 All inputs must have the same sample rate, and format.
1598 If inputs do not have the same duration, the output will stop with the
1601 @subsection Examples
1605 Merge two mono files into a stereo stream:
1607 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1611 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1613 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
1619 Mixes multiple audio inputs into a single output.
1621 Note that this filter only supports float samples (the @var{amerge}
1622 and @var{pan} audio filters support many formats). If the @var{amix}
1623 input has integer samples then @ref{aresample} will be automatically
1624 inserted to perform the conversion to float samples.
1628 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1630 will mix 3 input audio streams to a single output with the same duration as the
1631 first input and a dropout transition time of 3 seconds.
1633 It accepts the following parameters:
1637 The number of inputs. If unspecified, it defaults to 2.
1640 How to determine the end-of-stream.
1644 The duration of the longest input. (default)
1647 The duration of the shortest input.
1650 The duration of the first input.
1654 @item dropout_transition
1655 The transition time, in seconds, for volume renormalization when an input
1656 stream ends. The default value is 2 seconds.
1659 Specify weight of each input audio stream as sequence.
1660 Each weight is separated by space. By default all inputs have same weight.
1665 Multiply first audio stream with second audio stream and store result
1666 in output audio stream. Multiplication is done by multiplying each
1667 sample from first stream with sample at same position from second stream.
1669 With this element-wise multiplication one can create amplitude fades and
1670 amplitude modulations.
1672 @section anequalizer
1674 High-order parametric multiband equalizer for each channel.
1676 It accepts the following parameters:
1680 This option string is in format:
1681 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1682 Each equalizer band is separated by '|'.
1686 Set channel number to which equalization will be applied.
1687 If input doesn't have that channel the entry is ignored.
1690 Set central frequency for band.
1691 If input doesn't have that frequency the entry is ignored.
1694 Set band width in hertz.
1697 Set band gain in dB.
1700 Set filter type for band, optional, can be:
1704 Butterworth, this is default.
1715 With this option activated frequency response of anequalizer is displayed
1719 Set video stream size. Only useful if curves option is activated.
1722 Set max gain that will be displayed. Only useful if curves option is activated.
1723 Setting this to a reasonable value makes it possible to display gain which is derived from
1724 neighbour bands which are too close to each other and thus produce higher gain
1725 when both are activated.
1728 Set frequency scale used to draw frequency response in video output.
1729 Can be linear or logarithmic. Default is logarithmic.
1732 Set color for each channel curve which is going to be displayed in video stream.
1733 This is list of color names separated by space or by '|'.
1734 Unrecognised or missing colors will be replaced by white color.
1737 @subsection Examples
1741 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1742 for first 2 channels using Chebyshev type 1 filter:
1744 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1748 @subsection Commands
1750 This filter supports the following commands:
1753 Alter existing filter parameters.
1754 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1756 @var{fN} is existing filter number, starting from 0, if no such filter is available
1758 @var{freq} set new frequency parameter.
1759 @var{width} set new width parameter in herz.
1760 @var{gain} set new gain parameter in dB.
1762 Full filter invocation with asendcmd may look like this:
1763 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1768 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1770 Each sample is adjusted by looking for other samples with similar contexts. This
1771 context similarity is defined by comparing their surrounding patches of size
1772 @option{p}. Patches are searched in an area of @option{r} around the sample.
1774 The filter accepts the following options.
1778 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
1781 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
1782 Default value is 2 milliseconds.
1785 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
1786 Default value is 6 milliseconds.
1789 Set the output mode.
1791 It accepts the following values:
1794 Pass input unchanged.
1797 Pass noise filtered out.
1802 Default value is @var{o}.
1808 Pass the audio source unchanged to the output.
1812 Pad the end of an audio stream with silence.
1814 This can be used together with @command{ffmpeg} @option{-shortest} to
1815 extend audio streams to the same length as the video stream.
1817 A description of the accepted options follows.
1821 Set silence packet size. Default value is 4096.
1824 Set the number of samples of silence to add to the end. After the
1825 value is reached, the stream is terminated. This option is mutually
1826 exclusive with @option{whole_len}.
1829 Set the minimum total number of samples in the output audio stream. If
1830 the value is longer than the input audio length, silence is added to
1831 the end, until the value is reached. This option is mutually exclusive
1832 with @option{pad_len}.
1835 Specify the duration of samples of silence to add. See
1836 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1837 for the accepted syntax. Used only if set to non-zero value.
1840 Specify the minimum total duration in the output audio stream. See
1841 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1842 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1843 the input audio length, silence is added to the end, until the value is reached.
1844 This option is mutually exclusive with @option{pad_dur}
1847 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1848 nor @option{whole_dur} option is set, the filter will add silence to the end of
1849 the input stream indefinitely.
1851 @subsection Examples
1855 Add 1024 samples of silence to the end of the input:
1861 Make sure the audio output will contain at least 10000 samples, pad
1862 the input with silence if required:
1864 apad=whole_len=10000
1868 Use @command{ffmpeg} to pad the audio input with silence, so that the
1869 video stream will always result the shortest and will be converted
1870 until the end in the output file when using the @option{shortest}
1873 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1878 Add a phasing effect to the input audio.
1880 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1881 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1883 A description of the accepted parameters follows.
1887 Set input gain. Default is 0.4.
1890 Set output gain. Default is 0.74
1893 Set delay in milliseconds. Default is 3.0.
1896 Set decay. Default is 0.4.
1899 Set modulation speed in Hz. Default is 0.5.
1902 Set modulation type. Default is triangular.
1904 It accepts the following values:
1913 Audio pulsator is something between an autopanner and a tremolo.
1914 But it can produce funny stereo effects as well. Pulsator changes the volume
1915 of the left and right channel based on a LFO (low frequency oscillator) with
1916 different waveforms and shifted phases.
1917 This filter have the ability to define an offset between left and right
1918 channel. An offset of 0 means that both LFO shapes match each other.
1919 The left and right channel are altered equally - a conventional tremolo.
1920 An offset of 50% means that the shape of the right channel is exactly shifted
1921 in phase (or moved backwards about half of the frequency) - pulsator acts as
1922 an autopanner. At 1 both curves match again. Every setting in between moves the
1923 phase shift gapless between all stages and produces some "bypassing" sounds with
1924 sine and triangle waveforms. The more you set the offset near 1 (starting from
1925 the 0.5) the faster the signal passes from the left to the right speaker.
1927 The filter accepts the following options:
1931 Set input gain. By default it is 1. Range is [0.015625 - 64].
1934 Set output gain. By default it is 1. Range is [0.015625 - 64].
1937 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1938 sawup or sawdown. Default is sine.
1941 Set modulation. Define how much of original signal is affected by the LFO.
1944 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1947 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1950 Set pulse width. Default is 1. Allowed range is [0 - 2].
1953 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1956 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1960 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1964 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1965 if timing is set to hz.
1971 Resample the input audio to the specified parameters, using the
1972 libswresample library. If none are specified then the filter will
1973 automatically convert between its input and output.
1975 This filter is also able to stretch/squeeze the audio data to make it match
1976 the timestamps or to inject silence / cut out audio to make it match the
1977 timestamps, do a combination of both or do neither.
1979 The filter accepts the syntax
1980 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1981 expresses a sample rate and @var{resampler_options} is a list of
1982 @var{key}=@var{value} pairs, separated by ":". See the
1983 @ref{Resampler Options,,"Resampler Options" section in the
1984 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1985 for the complete list of supported options.
1987 @subsection Examples
1991 Resample the input audio to 44100Hz:
1997 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1998 samples per second compensation:
2000 aresample=async=1000
2006 Reverse an audio clip.
2008 Warning: This filter requires memory to buffer the entire clip, so trimming
2011 @subsection Examples
2015 Take the first 5 seconds of a clip, and reverse it.
2017 atrim=end=5,areverse
2021 @section asetnsamples
2023 Set the number of samples per each output audio frame.
2025 The last output packet may contain a different number of samples, as
2026 the filter will flush all the remaining samples when the input audio
2029 The filter accepts the following options:
2033 @item nb_out_samples, n
2034 Set the number of frames per each output audio frame. The number is
2035 intended as the number of samples @emph{per each channel}.
2036 Default value is 1024.
2039 If set to 1, the filter will pad the last audio frame with zeroes, so
2040 that the last frame will contain the same number of samples as the
2041 previous ones. Default value is 1.
2044 For example, to set the number of per-frame samples to 1234 and
2045 disable padding for the last frame, use:
2047 asetnsamples=n=1234:p=0
2052 Set the sample rate without altering the PCM data.
2053 This will result in a change of speed and pitch.
2055 The filter accepts the following options:
2058 @item sample_rate, r
2059 Set the output sample rate. Default is 44100 Hz.
2064 Show a line containing various information for each input audio frame.
2065 The input audio is not modified.
2067 The shown line contains a sequence of key/value pairs of the form
2068 @var{key}:@var{value}.
2070 The following values are shown in the output:
2074 The (sequential) number of the input frame, starting from 0.
2077 The presentation timestamp of the input frame, in time base units; the time base
2078 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2081 The presentation timestamp of the input frame in seconds.
2084 position of the frame in the input stream, -1 if this information in
2085 unavailable and/or meaningless (for example in case of synthetic audio)
2094 The sample rate for the audio frame.
2097 The number of samples (per channel) in the frame.
2100 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2101 audio, the data is treated as if all the planes were concatenated.
2103 @item plane_checksums
2104 A list of Adler-32 checksums for each data plane.
2110 Display time domain statistical information about the audio channels.
2111 Statistics are calculated and displayed for each audio channel and,
2112 where applicable, an overall figure is also given.
2114 It accepts the following option:
2117 Short window length in seconds, used for peak and trough RMS measurement.
2118 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2122 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2123 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2126 Available keys for each channel are:
2162 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2163 this @code{lavfi.astats.Overall.Peak_count}.
2165 For description what each key means read below.
2168 Set number of frame after which stats are going to be recalculated.
2169 Default is disabled.
2171 @item measure_perchannel
2172 Select the entries which need to be measured per channel. The metadata keys can
2173 be used as flags, default is @option{all} which measures everything.
2174 @option{none} disables all per channel measurement.
2176 @item measure_overall
2177 Select the entries which need to be measured overall. The metadata keys can
2178 be used as flags, default is @option{all} which measures everything.
2179 @option{none} disables all overall measurement.
2183 A description of each shown parameter follows:
2187 Mean amplitude displacement from zero.
2190 Minimal sample level.
2193 Maximal sample level.
2195 @item Min difference
2196 Minimal difference between two consecutive samples.
2198 @item Max difference
2199 Maximal difference between two consecutive samples.
2201 @item Mean difference
2202 Mean difference between two consecutive samples.
2203 The average of each difference between two consecutive samples.
2205 @item RMS difference
2206 Root Mean Square difference between two consecutive samples.
2210 Standard peak and RMS level measured in dBFS.
2214 Peak and trough values for RMS level measured over a short window.
2217 Standard ratio of peak to RMS level (note: not in dB).
2220 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2221 (i.e. either @var{Min level} or @var{Max level}).
2224 Number of occasions (not the number of samples) that the signal attained either
2225 @var{Min level} or @var{Max level}.
2228 Overall bit depth of audio. Number of bits used for each sample.
2231 Measured dynamic range of audio in dB.
2233 @item Zero crossings
2234 Number of points where the waveform crosses the zero level axis.
2236 @item Zero crossings rate
2237 Rate of Zero crossings and number of audio samples.
2244 The filter accepts exactly one parameter, the audio tempo. If not
2245 specified then the filter will assume nominal 1.0 tempo. Tempo must
2246 be in the [0.5, 100.0] range.
2248 Note that tempo greater than 2 will skip some samples rather than
2249 blend them in. If for any reason this is a concern it is always
2250 possible to daisy-chain several instances of atempo to achieve the
2251 desired product tempo.
2253 @subsection Examples
2257 Slow down audio to 80% tempo:
2263 To speed up audio to 300% tempo:
2269 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2271 atempo=sqrt(3),atempo=sqrt(3)
2277 Trim the input so that the output contains one continuous subpart of the input.
2279 It accepts the following parameters:
2282 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2283 sample with the timestamp @var{start} will be the first sample in the output.
2286 Specify time of the first audio sample that will be dropped, i.e. the
2287 audio sample immediately preceding the one with the timestamp @var{end} will be
2288 the last sample in the output.
2291 Same as @var{start}, except this option sets the start timestamp in samples
2295 Same as @var{end}, except this option sets the end timestamp in samples instead
2299 The maximum duration of the output in seconds.
2302 The number of the first sample that should be output.
2305 The number of the first sample that should be dropped.
2308 @option{start}, @option{end}, and @option{duration} are expressed as time
2309 duration specifications; see
2310 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2312 Note that the first two sets of the start/end options and the @option{duration}
2313 option look at the frame timestamp, while the _sample options simply count the
2314 samples that pass through the filter. So start/end_pts and start/end_sample will
2315 give different results when the timestamps are wrong, inexact or do not start at
2316 zero. Also note that this filter does not modify the timestamps. If you wish
2317 to have the output timestamps start at zero, insert the asetpts filter after the
2320 If multiple start or end options are set, this filter tries to be greedy and
2321 keep all samples that match at least one of the specified constraints. To keep
2322 only the part that matches all the constraints at once, chain multiple atrim
2325 The defaults are such that all the input is kept. So it is possible to set e.g.
2326 just the end values to keep everything before the specified time.
2331 Drop everything except the second minute of input:
2333 ffmpeg -i INPUT -af atrim=60:120
2337 Keep only the first 1000 samples:
2339 ffmpeg -i INPUT -af atrim=end_sample=1000
2346 Apply a two-pole Butterworth band-pass filter with central
2347 frequency @var{frequency}, and (3dB-point) band-width width.
2348 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2349 instead of the default: constant 0dB peak gain.
2350 The filter roll off at 6dB per octave (20dB per decade).
2352 The filter accepts the following options:
2356 Set the filter's central frequency. Default is @code{3000}.
2359 Constant skirt gain if set to 1. Defaults to 0.
2362 Set method to specify band-width of filter.
2377 Specify the band-width of a filter in width_type units.
2380 Specify which channels to filter, by default all available are filtered.
2383 @subsection Commands
2385 This filter supports the following commands:
2388 Change bandpass frequency.
2389 Syntax for the command is : "@var{frequency}"
2392 Change bandpass width_type.
2393 Syntax for the command is : "@var{width_type}"
2396 Change bandpass width.
2397 Syntax for the command is : "@var{width}"
2402 Apply a two-pole Butterworth band-reject filter with central
2403 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2404 The filter roll off at 6dB per octave (20dB per decade).
2406 The filter accepts the following options:
2410 Set the filter's central frequency. Default is @code{3000}.
2413 Set method to specify band-width of filter.
2428 Specify the band-width of a filter in width_type units.
2431 Specify which channels to filter, by default all available are filtered.
2434 @subsection Commands
2436 This filter supports the following commands:
2439 Change bandreject frequency.
2440 Syntax for the command is : "@var{frequency}"
2443 Change bandreject width_type.
2444 Syntax for the command is : "@var{width_type}"
2447 Change bandreject width.
2448 Syntax for the command is : "@var{width}"
2451 @section bass, lowshelf
2453 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2454 shelving filter with a response similar to that of a standard
2455 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2457 The filter accepts the following options:
2461 Give the gain at 0 Hz. Its useful range is about -20
2462 (for a large cut) to +20 (for a large boost).
2463 Beware of clipping when using a positive gain.
2466 Set the filter's central frequency and so can be used
2467 to extend or reduce the frequency range to be boosted or cut.
2468 The default value is @code{100} Hz.
2471 Set method to specify band-width of filter.
2486 Determine how steep is the filter's shelf transition.
2489 Specify which channels to filter, by default all available are filtered.
2492 @subsection Commands
2494 This filter supports the following commands:
2497 Change bass frequency.
2498 Syntax for the command is : "@var{frequency}"
2501 Change bass width_type.
2502 Syntax for the command is : "@var{width_type}"
2506 Syntax for the command is : "@var{width}"
2510 Syntax for the command is : "@var{gain}"
2515 Apply a biquad IIR filter with the given coefficients.
2516 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2517 are the numerator and denominator coefficients respectively.
2518 and @var{channels}, @var{c} specify which channels to filter, by default all
2519 available are filtered.
2521 @subsection Commands
2523 This filter supports the following commands:
2531 Change biquad parameter.
2532 Syntax for the command is : "@var{value}"
2536 Bauer stereo to binaural transformation, which improves headphone listening of
2537 stereo audio records.
2539 To enable compilation of this filter you need to configure FFmpeg with
2540 @code{--enable-libbs2b}.
2542 It accepts the following parameters:
2546 Pre-defined crossfeed level.
2550 Default level (fcut=700, feed=50).
2553 Chu Moy circuit (fcut=700, feed=60).
2556 Jan Meier circuit (fcut=650, feed=95).
2561 Cut frequency (in Hz).
2570 Remap input channels to new locations.
2572 It accepts the following parameters:
2575 Map channels from input to output. The argument is a '|'-separated list of
2576 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2577 @var{in_channel} form. @var{in_channel} can be either the name of the input
2578 channel (e.g. FL for front left) or its index in the input channel layout.
2579 @var{out_channel} is the name of the output channel or its index in the output
2580 channel layout. If @var{out_channel} is not given then it is implicitly an
2581 index, starting with zero and increasing by one for each mapping.
2583 @item channel_layout
2584 The channel layout of the output stream.
2587 If no mapping is present, the filter will implicitly map input channels to
2588 output channels, preserving indices.
2590 @subsection Examples
2594 For example, assuming a 5.1+downmix input MOV file,
2596 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2598 will create an output WAV file tagged as stereo from the downmix channels of
2602 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2604 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2608 @section channelsplit
2610 Split each channel from an input audio stream into a separate output stream.
2612 It accepts the following parameters:
2614 @item channel_layout
2615 The channel layout of the input stream. The default is "stereo".
2617 A channel layout describing the channels to be extracted as separate output streams
2618 or "all" to extract each input channel as a separate stream. The default is "all".
2620 Choosing channels not present in channel layout in the input will result in an error.
2623 @subsection Examples
2627 For example, assuming a stereo input MP3 file,
2629 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2631 will create an output Matroska file with two audio streams, one containing only
2632 the left channel and the other the right channel.
2635 Split a 5.1 WAV file into per-channel files:
2637 ffmpeg -i in.wav -filter_complex
2638 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2639 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2640 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2645 Extract only LFE from a 5.1 WAV file:
2647 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2648 -map '[LFE]' lfe.wav
2653 Add a chorus effect to the audio.
2655 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2657 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2658 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2659 The modulation depth defines the range the modulated delay is played before or after
2660 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2661 sound tuned around the original one, like in a chorus where some vocals are slightly
2664 It accepts the following parameters:
2667 Set input gain. Default is 0.4.
2670 Set output gain. Default is 0.4.
2673 Set delays. A typical delay is around 40ms to 60ms.
2685 @subsection Examples
2691 chorus=0.7:0.9:55:0.4:0.25:2
2697 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2701 Fuller sounding chorus with three delays:
2703 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
2708 Compress or expand the audio's dynamic range.
2710 It accepts the following parameters:
2716 A list of times in seconds for each channel over which the instantaneous level
2717 of the input signal is averaged to determine its volume. @var{attacks} refers to
2718 increase of volume and @var{decays} refers to decrease of volume. For most
2719 situations, the attack time (response to the audio getting louder) should be
2720 shorter than the decay time, because the human ear is more sensitive to sudden
2721 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2722 a typical value for decay is 0.8 seconds.
2723 If specified number of attacks & decays is lower than number of channels, the last
2724 set attack/decay will be used for all remaining channels.
2727 A list of points for the transfer function, specified in dB relative to the
2728 maximum possible signal amplitude. Each key points list must be defined using
2729 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2730 @code{x0/y0 x1/y1 x2/y2 ....}
2732 The input values must be in strictly increasing order but the transfer function
2733 does not have to be monotonically rising. The point @code{0/0} is assumed but
2734 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2735 function are @code{-70/-70|-60/-20|1/0}.
2738 Set the curve radius in dB for all joints. It defaults to 0.01.
2741 Set the additional gain in dB to be applied at all points on the transfer
2742 function. This allows for easy adjustment of the overall gain.
2746 Set an initial volume, in dB, to be assumed for each channel when filtering
2747 starts. This permits the user to supply a nominal level initially, so that, for
2748 example, a very large gain is not applied to initial signal levels before the
2749 companding has begun to operate. A typical value for audio which is initially
2750 quiet is -90 dB. It defaults to 0.
2753 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2754 delayed before being fed to the volume adjuster. Specifying a delay
2755 approximately equal to the attack/decay times allows the filter to effectively
2756 operate in predictive rather than reactive mode. It defaults to 0.
2760 @subsection Examples
2764 Make music with both quiet and loud passages suitable for listening to in a
2767 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2770 Another example for audio with whisper and explosion parts:
2772 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2776 A noise gate for when the noise is at a lower level than the signal:
2778 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2782 Here is another noise gate, this time for when the noise is at a higher level
2783 than the signal (making it, in some ways, similar to squelch):
2785 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2789 2:1 compression starting at -6dB:
2791 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2795 2:1 compression starting at -9dB:
2797 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2801 2:1 compression starting at -12dB:
2803 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2807 2:1 compression starting at -18dB:
2809 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2813 3:1 compression starting at -15dB:
2815 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2821 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2827 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
2831 Hard limiter at -6dB:
2833 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2837 Hard limiter at -12dB:
2839 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2843 Hard noise gate at -35 dB:
2845 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2851 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2855 @section compensationdelay
2857 Compensation Delay Line is a metric based delay to compensate differing
2858 positions of microphones or speakers.
2860 For example, you have recorded guitar with two microphones placed in
2861 different location. Because the front of sound wave has fixed speed in
2862 normal conditions, the phasing of microphones can vary and depends on
2863 their location and interposition. The best sound mix can be achieved when
2864 these microphones are in phase (synchronized). Note that distance of
2865 ~30 cm between microphones makes one microphone to capture signal in
2866 antiphase to another microphone. That makes the final mix sounding moody.
2867 This filter helps to solve phasing problems by adding different delays
2868 to each microphone track and make them synchronized.
2870 The best result can be reached when you take one track as base and
2871 synchronize other tracks one by one with it.
2872 Remember that synchronization/delay tolerance depends on sample rate, too.
2873 Higher sample rates will give more tolerance.
2875 It accepts the following parameters:
2879 Set millimeters distance. This is compensation distance for fine tuning.
2883 Set cm distance. This is compensation distance for tightening distance setup.
2887 Set meters distance. This is compensation distance for hard distance setup.
2891 Set dry amount. Amount of unprocessed (dry) signal.
2895 Set wet amount. Amount of processed (wet) signal.
2899 Set temperature degree in Celsius. This is the temperature of the environment.
2904 Apply headphone crossfeed filter.
2906 Crossfeed is the process of blending the left and right channels of stereo
2908 It is mainly used to reduce extreme stereo separation of low frequencies.
2910 The intent is to produce more speaker like sound to the listener.
2912 The filter accepts the following options:
2916 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2917 This sets gain of low shelf filter for side part of stereo image.
2918 Default is -6dB. Max allowed is -30db when strength is set to 1.
2921 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2922 This sets cut off frequency of low shelf filter. Default is cut off near
2923 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2926 Set input gain. Default is 0.9.
2929 Set output gain. Default is 1.
2932 @section crystalizer
2933 Simple algorithm to expand audio dynamic range.
2935 The filter accepts the following options:
2939 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2940 (unchanged sound) to 10.0 (maximum effect).
2943 Enable clipping. By default is enabled.
2947 Apply a DC shift to the audio.
2949 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2950 in the recording chain) from the audio. The effect of a DC offset is reduced
2951 headroom and hence volume. The @ref{astats} filter can be used to determine if
2952 a signal has a DC offset.
2956 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2960 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2961 used to prevent clipping.
2965 Measure audio dynamic range.
2967 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2968 is found in transition material. And anything less that 8 have very poor dynamics
2969 and is very compressed.
2971 The filter accepts the following options:
2975 Set window length in seconds used to split audio into segments of equal length.
2976 Default is 3 seconds.
2980 Dynamic Audio Normalizer.
2982 This filter applies a certain amount of gain to the input audio in order
2983 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2984 contrast to more "simple" normalization algorithms, the Dynamic Audio
2985 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2986 This allows for applying extra gain to the "quiet" sections of the audio
2987 while avoiding distortions or clipping the "loud" sections. In other words:
2988 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2989 sections, in the sense that the volume of each section is brought to the
2990 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2991 this goal *without* applying "dynamic range compressing". It will retain 100%
2992 of the dynamic range *within* each section of the audio file.
2996 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2997 Default is 500 milliseconds.
2998 The Dynamic Audio Normalizer processes the input audio in small chunks,
2999 referred to as frames. This is required, because a peak magnitude has no
3000 meaning for just a single sample value. Instead, we need to determine the
3001 peak magnitude for a contiguous sequence of sample values. While a "standard"
3002 normalizer would simply use the peak magnitude of the complete file, the
3003 Dynamic Audio Normalizer determines the peak magnitude individually for each
3004 frame. The length of a frame is specified in milliseconds. By default, the
3005 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3006 been found to give good results with most files.
3007 Note that the exact frame length, in number of samples, will be determined
3008 automatically, based on the sampling rate of the individual input audio file.
3011 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3012 number. Default is 31.
3013 Probably the most important parameter of the Dynamic Audio Normalizer is the
3014 @code{window size} of the Gaussian smoothing filter. The filter's window size
3015 is specified in frames, centered around the current frame. For the sake of
3016 simplicity, this must be an odd number. Consequently, the default value of 31
3017 takes into account the current frame, as well as the 15 preceding frames and
3018 the 15 subsequent frames. Using a larger window results in a stronger
3019 smoothing effect and thus in less gain variation, i.e. slower gain
3020 adaptation. Conversely, using a smaller window results in a weaker smoothing
3021 effect and thus in more gain variation, i.e. faster gain adaptation.
3022 In other words, the more you increase this value, the more the Dynamic Audio
3023 Normalizer will behave like a "traditional" normalization filter. On the
3024 contrary, the more you decrease this value, the more the Dynamic Audio
3025 Normalizer will behave like a dynamic range compressor.
3028 Set the target peak value. This specifies the highest permissible magnitude
3029 level for the normalized audio input. This filter will try to approach the
3030 target peak magnitude as closely as possible, but at the same time it also
3031 makes sure that the normalized signal will never exceed the peak magnitude.
3032 A frame's maximum local gain factor is imposed directly by the target peak
3033 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3034 It is not recommended to go above this value.
3037 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3038 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3039 factor for each input frame, i.e. the maximum gain factor that does not
3040 result in clipping or distortion. The maximum gain factor is determined by
3041 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3042 additionally bounds the frame's maximum gain factor by a predetermined
3043 (global) maximum gain factor. This is done in order to avoid excessive gain
3044 factors in "silent" or almost silent frames. By default, the maximum gain
3045 factor is 10.0, For most inputs the default value should be sufficient and
3046 it usually is not recommended to increase this value. Though, for input
3047 with an extremely low overall volume level, it may be necessary to allow even
3048 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3049 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3050 Instead, a "sigmoid" threshold function will be applied. This way, the
3051 gain factors will smoothly approach the threshold value, but never exceed that
3055 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3056 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3057 This means that the maximum local gain factor for each frame is defined
3058 (only) by the frame's highest magnitude sample. This way, the samples can
3059 be amplified as much as possible without exceeding the maximum signal
3060 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3061 Normalizer can also take into account the frame's root mean square,
3062 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3063 determine the power of a time-varying signal. It is therefore considered
3064 that the RMS is a better approximation of the "perceived loudness" than
3065 just looking at the signal's peak magnitude. Consequently, by adjusting all
3066 frames to a constant RMS value, a uniform "perceived loudness" can be
3067 established. If a target RMS value has been specified, a frame's local gain
3068 factor is defined as the factor that would result in exactly that RMS value.
3069 Note, however, that the maximum local gain factor is still restricted by the
3070 frame's highest magnitude sample, in order to prevent clipping.
3073 Enable channels coupling. By default is enabled.
3074 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3075 amount. This means the same gain factor will be applied to all channels, i.e.
3076 the maximum possible gain factor is determined by the "loudest" channel.
3077 However, in some recordings, it may happen that the volume of the different
3078 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3079 In this case, this option can be used to disable the channel coupling. This way,
3080 the gain factor will be determined independently for each channel, depending
3081 only on the individual channel's highest magnitude sample. This allows for
3082 harmonizing the volume of the different channels.
3085 Enable DC bias correction. By default is disabled.
3086 An audio signal (in the time domain) is a sequence of sample values.
3087 In the Dynamic Audio Normalizer these sample values are represented in the
3088 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3089 audio signal, or "waveform", should be centered around the zero point.
3090 That means if we calculate the mean value of all samples in a file, or in a
3091 single frame, then the result should be 0.0 or at least very close to that
3092 value. If, however, there is a significant deviation of the mean value from
3093 0.0, in either positive or negative direction, this is referred to as a
3094 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3095 Audio Normalizer provides optional DC bias correction.
3096 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3097 the mean value, or "DC correction" offset, of each input frame and subtract
3098 that value from all of the frame's sample values which ensures those samples
3099 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3100 boundaries, the DC correction offset values will be interpolated smoothly
3101 between neighbouring frames.
3104 Enable alternative boundary mode. By default is disabled.
3105 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3106 around each frame. This includes the preceding frames as well as the
3107 subsequent frames. However, for the "boundary" frames, located at the very
3108 beginning and at the very end of the audio file, not all neighbouring
3109 frames are available. In particular, for the first few frames in the audio
3110 file, the preceding frames are not known. And, similarly, for the last few
3111 frames in the audio file, the subsequent frames are not known. Thus, the
3112 question arises which gain factors should be assumed for the missing frames
3113 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3114 to deal with this situation. The default boundary mode assumes a gain factor
3115 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3116 "fade out" at the beginning and at the end of the input, respectively.
3119 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3120 By default, the Dynamic Audio Normalizer does not apply "traditional"
3121 compression. This means that signal peaks will not be pruned and thus the
3122 full dynamic range will be retained within each local neighbourhood. However,
3123 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3124 normalization algorithm with a more "traditional" compression.
3125 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3126 (thresholding) function. If (and only if) the compression feature is enabled,
3127 all input frames will be processed by a soft knee thresholding function prior
3128 to the actual normalization process. Put simply, the thresholding function is
3129 going to prune all samples whose magnitude exceeds a certain threshold value.
3130 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3131 value. Instead, the threshold value will be adjusted for each individual
3133 In general, smaller parameters result in stronger compression, and vice versa.
3134 Values below 3.0 are not recommended, because audible distortion may appear.
3139 Make audio easier to listen to on headphones.
3141 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3142 so that when listened to on headphones the stereo image is moved from
3143 inside your head (standard for headphones) to outside and in front of
3144 the listener (standard for speakers).
3150 Apply a two-pole peaking equalisation (EQ) filter. With this
3151 filter, the signal-level at and around a selected frequency can
3152 be increased or decreased, whilst (unlike bandpass and bandreject
3153 filters) that at all other frequencies is unchanged.
3155 In order to produce complex equalisation curves, this filter can
3156 be given several times, each with a different central frequency.
3158 The filter accepts the following options:
3162 Set the filter's central frequency in Hz.
3165 Set method to specify band-width of filter.
3180 Specify the band-width of a filter in width_type units.
3183 Set the required gain or attenuation in dB.
3184 Beware of clipping when using a positive gain.
3187 Specify which channels to filter, by default all available are filtered.
3190 @subsection Examples
3193 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3195 equalizer=f=1000:t=h:width=200:g=-10
3199 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3201 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3205 @subsection Commands
3207 This filter supports the following commands:
3210 Change equalizer frequency.
3211 Syntax for the command is : "@var{frequency}"
3214 Change equalizer width_type.
3215 Syntax for the command is : "@var{width_type}"
3218 Change equalizer width.
3219 Syntax for the command is : "@var{width}"
3222 Change equalizer gain.
3223 Syntax for the command is : "@var{gain}"
3226 @section extrastereo
3228 Linearly increases the difference between left and right channels which
3229 adds some sort of "live" effect to playback.
3231 The filter accepts the following options:
3235 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3236 (average of both channels), with 1.0 sound will be unchanged, with
3237 -1.0 left and right channels will be swapped.
3240 Enable clipping. By default is enabled.
3243 @section firequalizer
3244 Apply FIR Equalization using arbitrary frequency response.
3246 The filter accepts the following option:
3250 Set gain curve equation (in dB). The expression can contain variables:
3253 the evaluated frequency
3257 channel number, set to 0 when multichannels evaluation is disabled
3259 channel id, see libavutil/channel_layout.h, set to the first channel id when
3260 multichannels evaluation is disabled
3264 channel_layout, see libavutil/channel_layout.h
3269 @item gain_interpolate(f)
3270 interpolate gain on frequency f based on gain_entry
3271 @item cubic_interpolate(f)
3272 same as gain_interpolate, but smoother
3274 This option is also available as command. Default is @code{gain_interpolate(f)}.
3277 Set gain entry for gain_interpolate function. The expression can
3281 store gain entry at frequency f with value g
3283 This option is also available as command.
3286 Set filter delay in seconds. Higher value means more accurate.
3287 Default is @code{0.01}.
3290 Set filter accuracy in Hz. Lower value means more accurate.
3291 Default is @code{5}.
3294 Set window function. Acceptable values are:
3297 rectangular window, useful when gain curve is already smooth
3299 hann window (default)
3305 3-terms continuous 1st derivative nuttall window
3307 minimum 3-terms discontinuous nuttall window
3309 4-terms continuous 1st derivative nuttall window
3311 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3313 blackman-harris window
3319 If enabled, use fixed number of audio samples. This improves speed when
3320 filtering with large delay. Default is disabled.
3323 Enable multichannels evaluation on gain. Default is disabled.
3326 Enable zero phase mode by subtracting timestamp to compensate delay.
3327 Default is disabled.
3330 Set scale used by gain. Acceptable values are:
3333 linear frequency, linear gain
3335 linear frequency, logarithmic (in dB) gain (default)
3337 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3339 logarithmic frequency, logarithmic gain
3343 Set file for dumping, suitable for gnuplot.
3346 Set scale for dumpfile. Acceptable values are same with scale option.
3350 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3351 Default is disabled.
3354 Enable minimum phase impulse response. Default is disabled.
3357 @subsection Examples
3362 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3365 lowpass at 1000 Hz with gain_entry:
3367 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3370 custom equalization:
3372 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3375 higher delay with zero phase to compensate delay:
3377 firequalizer=delay=0.1:fixed=on:zero_phase=on
3380 lowpass on left channel, highpass on right channel:
3382 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3383 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3388 Apply a flanging effect to the audio.
3390 The filter accepts the following options:
3394 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3397 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3400 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3404 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3405 Default value is 71.
3408 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3411 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3412 Default value is @var{sinusoidal}.
3415 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3416 Default value is 25.
3419 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3420 Default is @var{linear}.
3424 Apply Haas effect to audio.
3426 Note that this makes most sense to apply on mono signals.
3427 With this filter applied to mono signals it give some directionality and
3428 stretches its stereo image.
3430 The filter accepts the following options:
3434 Set input level. By default is @var{1}, or 0dB
3437 Set output level. By default is @var{1}, or 0dB.
3440 Set gain applied to side part of signal. By default is @var{1}.
3443 Set kind of middle source. Can be one of the following:
3453 Pick middle part signal of stereo image.
3456 Pick side part signal of stereo image.
3460 Change middle phase. By default is disabled.
3463 Set left channel delay. By default is @var{2.05} milliseconds.
3466 Set left channel balance. By default is @var{-1}.
3469 Set left channel gain. By default is @var{1}.
3472 Change left phase. By default is disabled.
3475 Set right channel delay. By defaults is @var{2.12} milliseconds.
3478 Set right channel balance. By default is @var{1}.
3481 Set right channel gain. By default is @var{1}.
3484 Change right phase. By default is enabled.
3489 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3490 embedded HDCD codes is expanded into a 20-bit PCM stream.
3492 The filter supports the Peak Extend and Low-level Gain Adjustment features
3493 of HDCD, and detects the Transient Filter flag.
3496 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3499 When using the filter with wav, note the default encoding for wav is 16-bit,
3500 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3501 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3503 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3504 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3507 The filter accepts the following options:
3510 @item disable_autoconvert
3511 Disable any automatic format conversion or resampling in the filter graph.
3513 @item process_stereo
3514 Process the stereo channels together. If target_gain does not match between
3515 channels, consider it invalid and use the last valid target_gain.
3518 Set the code detect timer period in ms.
3521 Always extend peaks above -3dBFS even if PE isn't signaled.
3524 Replace audio with a solid tone and adjust the amplitude to signal some
3525 specific aspect of the decoding process. The output file can be loaded in
3526 an audio editor alongside the original to aid analysis.
3528 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3535 Gain adjustment level at each sample
3537 Samples where peak extend occurs
3539 Samples where the code detect timer is active
3541 Samples where the target gain does not match between channels
3547 Apply head-related transfer functions (HRTFs) to create virtual
3548 loudspeakers around the user for binaural listening via headphones.
3549 The HRIRs are provided via additional streams, for each channel
3550 one stereo input stream is needed.
3552 The filter accepts the following options:
3556 Set mapping of input streams for convolution.
3557 The argument is a '|'-separated list of channel names in order as they
3558 are given as additional stream inputs for filter.
3559 This also specify number of input streams. Number of input streams
3560 must be not less than number of channels in first stream plus one.
3563 Set gain applied to audio. Value is in dB. Default is 0.
3566 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3567 processing audio in time domain which is slow.
3568 @var{freq} is processing audio in frequency domain which is fast.
3569 Default is @var{freq}.
3572 Set custom gain for LFE channels. Value is in dB. Default is 0.
3575 Set size of frame in number of samples which will be processed at once.
3576 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3579 Set format of hrir stream.
3580 Default value is @var{stereo}. Alternative value is @var{multich}.
3581 If value is set to @var{stereo}, number of additional streams should
3582 be greater or equal to number of input channels in first input stream.
3583 Also each additional stream should have stereo number of channels.
3584 If value is set to @var{multich}, number of additional streams should
3585 be exactly one. Also number of input channels of additional stream
3586 should be equal or greater than twice number of channels of first input
3590 @subsection Examples
3594 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3595 each amovie filter use stereo file with IR coefficients as input.
3596 The files give coefficients for each position of virtual loudspeaker:
3599 -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"
3604 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3605 but now in @var{multich} @var{hrir} format.
3607 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"
3614 Apply a high-pass filter with 3dB point frequency.
3615 The filter can be either single-pole, or double-pole (the default).
3616 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3618 The filter accepts the following options:
3622 Set frequency in Hz. Default is 3000.
3625 Set number of poles. Default is 2.
3628 Set method to specify band-width of filter.
3643 Specify the band-width of a filter in width_type units.
3644 Applies only to double-pole filter.
3645 The default is 0.707q and gives a Butterworth response.
3648 Specify which channels to filter, by default all available are filtered.
3651 @subsection Commands
3653 This filter supports the following commands:
3656 Change highpass frequency.
3657 Syntax for the command is : "@var{frequency}"
3660 Change highpass width_type.
3661 Syntax for the command is : "@var{width_type}"
3664 Change highpass width.
3665 Syntax for the command is : "@var{width}"
3670 Join multiple input streams into one multi-channel stream.
3672 It accepts the following parameters:
3676 The number of input streams. It defaults to 2.
3678 @item channel_layout
3679 The desired output channel layout. It defaults to stereo.
3682 Map channels from inputs to output. The argument is a '|'-separated list of
3683 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3684 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3685 can be either the name of the input channel (e.g. FL for front left) or its
3686 index in the specified input stream. @var{out_channel} is the name of the output
3690 The filter will attempt to guess the mappings when they are not specified
3691 explicitly. It does so by first trying to find an unused matching input channel
3692 and if that fails it picks the first unused input channel.
3694 Join 3 inputs (with properly set channel layouts):
3696 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3699 Build a 5.1 output from 6 single-channel streams:
3701 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3702 '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'
3708 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3710 To enable compilation of this filter you need to configure FFmpeg with
3711 @code{--enable-ladspa}.
3715 Specifies the name of LADSPA plugin library to load. If the environment
3716 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3717 each one of the directories specified by the colon separated list in
3718 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3719 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3720 @file{/usr/lib/ladspa/}.
3723 Specifies the plugin within the library. Some libraries contain only
3724 one plugin, but others contain many of them. If this is not set filter
3725 will list all available plugins within the specified library.
3728 Set the '|' separated list of controls which are zero or more floating point
3729 values that determine the behavior of the loaded plugin (for example delay,
3731 Controls need to be defined using the following syntax:
3732 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3733 @var{valuei} is the value set on the @var{i}-th control.
3734 Alternatively they can be also defined using the following syntax:
3735 @var{value0}|@var{value1}|@var{value2}|..., where
3736 @var{valuei} is the value set on the @var{i}-th control.
3737 If @option{controls} is set to @code{help}, all available controls and
3738 their valid ranges are printed.
3740 @item sample_rate, s
3741 Specify the sample rate, default to 44100. Only used if plugin have
3745 Set the number of samples per channel per each output frame, default
3746 is 1024. Only used if plugin have zero inputs.
3749 Set the minimum duration of the sourced audio. See
3750 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3751 for the accepted syntax.
3752 Note that the resulting duration may be greater than the specified duration,
3753 as the generated audio is always cut at the end of a complete frame.
3754 If not specified, or the expressed duration is negative, the audio is
3755 supposed to be generated forever.
3756 Only used if plugin have zero inputs.
3760 @subsection Examples
3764 List all available plugins within amp (LADSPA example plugin) library:
3770 List all available controls and their valid ranges for @code{vcf_notch}
3771 plugin from @code{VCF} library:
3773 ladspa=f=vcf:p=vcf_notch:c=help
3777 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3780 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3784 Add reverberation to the audio using TAP-plugins
3785 (Tom's Audio Processing plugins):
3787 ladspa=file=tap_reverb:tap_reverb
3791 Generate white noise, with 0.2 amplitude:
3793 ladspa=file=cmt:noise_source_white:c=c0=.2
3797 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3798 @code{C* Audio Plugin Suite} (CAPS) library:
3800 ladspa=file=caps:Click:c=c1=20'
3804 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3806 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3810 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3811 @code{SWH Plugins} collection:
3813 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3817 Attenuate low frequencies using Multiband EQ from Steve Harris
3818 @code{SWH Plugins} collection:
3820 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3824 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3827 ladspa=caps:Narrower
3831 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3833 ladspa=caps:White:.2
3837 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3839 ladspa=caps:Fractal:c=c1=1
3843 Dynamic volume normalization using @code{VLevel} plugin:
3845 ladspa=vlevel-ladspa:vlevel_mono
3849 @subsection Commands
3851 This filter supports the following commands:
3854 Modify the @var{N}-th control value.
3856 If the specified value is not valid, it is ignored and prior one is kept.
3861 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3862 Support for both single pass (livestreams, files) and double pass (files) modes.
3863 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3864 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3865 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3867 The filter accepts the following options:
3871 Set integrated loudness target.
3872 Range is -70.0 - -5.0. Default value is -24.0.
3875 Set loudness range target.
3876 Range is 1.0 - 20.0. Default value is 7.0.
3879 Set maximum true peak.
3880 Range is -9.0 - +0.0. Default value is -2.0.
3882 @item measured_I, measured_i
3883 Measured IL of input file.
3884 Range is -99.0 - +0.0.
3886 @item measured_LRA, measured_lra
3887 Measured LRA of input file.
3888 Range is 0.0 - 99.0.
3890 @item measured_TP, measured_tp
3891 Measured true peak of input file.
3892 Range is -99.0 - +99.0.
3894 @item measured_thresh
3895 Measured threshold of input file.
3896 Range is -99.0 - +0.0.
3899 Set offset gain. Gain is applied before the true-peak limiter.
3900 Range is -99.0 - +99.0. Default is +0.0.
3903 Normalize linearly if possible.
3904 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3905 to be specified in order to use this mode.
3906 Options are true or false. Default is true.
3909 Treat mono input files as "dual-mono". If a mono file is intended for playback
3910 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3911 If set to @code{true}, this option will compensate for this effect.
3912 Multi-channel input files are not affected by this option.
3913 Options are true or false. Default is false.
3916 Set print format for stats. Options are summary, json, or none.
3917 Default value is none.
3922 Apply a low-pass filter with 3dB point frequency.
3923 The filter can be either single-pole or double-pole (the default).
3924 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3926 The filter accepts the following options:
3930 Set frequency in Hz. Default is 500.
3933 Set number of poles. Default is 2.
3936 Set method to specify band-width of filter.
3951 Specify the band-width of a filter in width_type units.
3952 Applies only to double-pole filter.
3953 The default is 0.707q and gives a Butterworth response.
3956 Specify which channels to filter, by default all available are filtered.
3959 @subsection Examples
3962 Lowpass only LFE channel, it LFE is not present it does nothing:
3968 @subsection Commands
3970 This filter supports the following commands:
3973 Change lowpass frequency.
3974 Syntax for the command is : "@var{frequency}"
3977 Change lowpass width_type.
3978 Syntax for the command is : "@var{width_type}"
3981 Change lowpass width.
3982 Syntax for the command is : "@var{width}"
3987 Load a LV2 (LADSPA Version 2) plugin.
3989 To enable compilation of this filter you need to configure FFmpeg with
3990 @code{--enable-lv2}.
3994 Specifies the plugin URI. You may need to escape ':'.
3997 Set the '|' separated list of controls which are zero or more floating point
3998 values that determine the behavior of the loaded plugin (for example delay,
4000 If @option{controls} is set to @code{help}, all available controls and
4001 their valid ranges are printed.
4003 @item sample_rate, s
4004 Specify the sample rate, default to 44100. Only used if plugin have
4008 Set the number of samples per channel per each output frame, default
4009 is 1024. Only used if plugin have zero inputs.
4012 Set the minimum duration of the sourced audio. See
4013 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4014 for the accepted syntax.
4015 Note that the resulting duration may be greater than the specified duration,
4016 as the generated audio is always cut at the end of a complete frame.
4017 If not specified, or the expressed duration is negative, the audio is
4018 supposed to be generated forever.
4019 Only used if plugin have zero inputs.
4022 @subsection Examples
4026 Apply bass enhancer plugin from Calf:
4028 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4032 Apply vinyl plugin from Calf:
4034 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4038 Apply bit crusher plugin from ArtyFX:
4040 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4045 Multiband Compress or expand the audio's dynamic range.
4047 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4048 This is akin to the crossover of a loudspeaker, and results in flat frequency
4049 response when absent compander action.
4051 It accepts the following parameters:
4055 This option syntax is:
4056 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4057 For explanation of each item refer to compand filter documentation.
4063 Mix channels with specific gain levels. The filter accepts the output
4064 channel layout followed by a set of channels definitions.
4066 This filter is also designed to efficiently remap the channels of an audio
4069 The filter accepts parameters of the form:
4070 "@var{l}|@var{outdef}|@var{outdef}|..."
4074 output channel layout or number of channels
4077 output channel specification, of the form:
4078 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4081 output channel to define, either a channel name (FL, FR, etc.) or a channel
4082 number (c0, c1, etc.)
4085 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4088 input channel to use, see out_name for details; it is not possible to mix
4089 named and numbered input channels
4092 If the `=' in a channel specification is replaced by `<', then the gains for
4093 that specification will be renormalized so that the total is 1, thus
4094 avoiding clipping noise.
4096 @subsection Mixing examples
4098 For example, if you want to down-mix from stereo to mono, but with a bigger
4099 factor for the left channel:
4101 pan=1c|c0=0.9*c0+0.1*c1
4104 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4105 7-channels surround:
4107 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4110 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4111 that should be preferred (see "-ac" option) unless you have very specific
4114 @subsection Remapping examples
4116 The channel remapping will be effective if, and only if:
4119 @item gain coefficients are zeroes or ones,
4120 @item only one input per channel output,
4123 If all these conditions are satisfied, the filter will notify the user ("Pure
4124 channel mapping detected"), and use an optimized and lossless method to do the
4127 For example, if you have a 5.1 source and want a stereo audio stream by
4128 dropping the extra channels:
4130 pan="stereo| c0=FL | c1=FR"
4133 Given the same source, you can also switch front left and front right channels
4134 and keep the input channel layout:
4136 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4139 If the input is a stereo audio stream, you can mute the front left channel (and
4140 still keep the stereo channel layout) with:
4145 Still with a stereo audio stream input, you can copy the right channel in both
4146 front left and right:
4148 pan="stereo| c0=FR | c1=FR"
4153 ReplayGain scanner filter. This filter takes an audio stream as an input and
4154 outputs it unchanged.
4155 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4159 Convert the audio sample format, sample rate and channel layout. It is
4160 not meant to be used directly.
4163 Apply time-stretching and pitch-shifting with librubberband.
4165 To enable compilation of this filter, you need to configure FFmpeg with
4166 @code{--enable-librubberband}.
4168 The filter accepts the following options:
4172 Set tempo scale factor.
4175 Set pitch scale factor.
4178 Set transients detector.
4179 Possible values are:
4188 Possible values are:
4197 Possible values are:
4204 Set processing window size.
4205 Possible values are:
4214 Possible values are:
4221 Enable formant preservation when shift pitching.
4222 Possible values are:
4230 Possible values are:
4239 Possible values are:
4246 @section sidechaincompress
4248 This filter acts like normal compressor but has the ability to compress
4249 detected signal using second input signal.
4250 It needs two input streams and returns one output stream.
4251 First input stream will be processed depending on second stream signal.
4252 The filtered signal then can be filtered with other filters in later stages of
4253 processing. See @ref{pan} and @ref{amerge} filter.
4255 The filter accepts the following options:
4259 Set input gain. Default is 1. Range is between 0.015625 and 64.
4262 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
4263 Default is @code{downward}.
4266 If a signal of second stream raises above this level it will affect the gain
4267 reduction of first stream.
4268 By default is 0.125. Range is between 0.00097563 and 1.
4271 Set a ratio about which the signal is reduced. 1:2 means that if the level
4272 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4273 Default is 2. Range is between 1 and 20.
4276 Amount of milliseconds the signal has to rise above the threshold before gain
4277 reduction starts. Default is 20. Range is between 0.01 and 2000.
4280 Amount of milliseconds the signal has to fall below the threshold before
4281 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4284 Set the amount by how much signal will be amplified after processing.
4285 Default is 1. Range is from 1 to 64.
4288 Curve the sharp knee around the threshold to enter gain reduction more softly.
4289 Default is 2.82843. Range is between 1 and 8.
4292 Choose if the @code{average} level between all channels of side-chain stream
4293 or the louder(@code{maximum}) channel of side-chain stream affects the
4294 reduction. Default is @code{average}.
4297 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4298 of @code{rms}. Default is @code{rms} which is mainly smoother.
4301 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4304 How much to use compressed signal in output. Default is 1.
4305 Range is between 0 and 1.
4308 @subsection Examples
4312 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4313 depending on the signal of 2nd input and later compressed signal to be
4314 merged with 2nd input:
4316 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4320 @section sidechaingate
4322 A sidechain gate acts like a normal (wideband) gate but has the ability to
4323 filter the detected signal before sending it to the gain reduction stage.
4324 Normally a gate uses the full range signal to detect a level above the
4326 For example: If you cut all lower frequencies from your sidechain signal
4327 the gate will decrease the volume of your track only if not enough highs
4328 appear. With this technique you are able to reduce the resonation of a
4329 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4331 It needs two input streams and returns one output stream.
4332 First input stream will be processed depending on second stream signal.
4334 The filter accepts the following options:
4338 Set input level before filtering.
4339 Default is 1. Allowed range is from 0.015625 to 64.
4342 Set the mode of operation. Can be @code{upward} or @code{downward}.
4343 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
4344 will be amplified, expanding dynamic range in upward direction.
4345 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
4348 Set the level of gain reduction when the signal is below the threshold.
4349 Default is 0.06125. Allowed range is from 0 to 1.
4350 Setting this to 0 disables reduction and then filter behaves like expander.
4353 If a signal rises above this level the gain reduction is released.
4354 Default is 0.125. Allowed range is from 0 to 1.
4357 Set a ratio about which the signal is reduced.
4358 Default is 2. Allowed range is from 1 to 9000.
4361 Amount of milliseconds the signal has to rise above the threshold before gain
4363 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4366 Amount of milliseconds the signal has to fall below the threshold before the
4367 reduction is increased again. Default is 250 milliseconds.
4368 Allowed range is from 0.01 to 9000.
4371 Set amount of amplification of signal after processing.
4372 Default is 1. Allowed range is from 1 to 64.
4375 Curve the sharp knee around the threshold to enter gain reduction more softly.
4376 Default is 2.828427125. Allowed range is from 1 to 8.
4379 Choose if exact signal should be taken for detection or an RMS like one.
4380 Default is rms. Can be peak or rms.
4383 Choose if the average level between all channels or the louder channel affects
4385 Default is average. Can be average or maximum.
4388 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4391 @section silencedetect
4393 Detect silence in an audio stream.
4395 This filter logs a message when it detects that the input audio volume is less
4396 or equal to a noise tolerance value for a duration greater or equal to the
4397 minimum detected noise duration.
4399 The printed times and duration are expressed in seconds.
4401 The filter accepts the following options:
4405 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4406 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4409 Set silence duration until notification (default is 2 seconds).
4412 Process each channel separately, instead of combined. By default is disabled.
4415 @subsection Examples
4419 Detect 5 seconds of silence with -50dB noise tolerance:
4421 silencedetect=n=-50dB:d=5
4425 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4426 tolerance in @file{silence.mp3}:
4428 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4432 @section silenceremove
4434 Remove silence from the beginning, middle or end of the audio.
4436 The filter accepts the following options:
4440 This value is used to indicate if audio should be trimmed at beginning of
4441 the audio. A value of zero indicates no silence should be trimmed from the
4442 beginning. When specifying a non-zero value, it trims audio up until it
4443 finds non-silence. Normally, when trimming silence from beginning of audio
4444 the @var{start_periods} will be @code{1} but it can be increased to higher
4445 values to trim all audio up to specific count of non-silence periods.
4446 Default value is @code{0}.
4448 @item start_duration
4449 Specify the amount of time that non-silence must be detected before it stops
4450 trimming audio. By increasing the duration, bursts of noises can be treated
4451 as silence and trimmed off. Default value is @code{0}.
4453 @item start_threshold
4454 This indicates what sample value should be treated as silence. For digital
4455 audio, a value of @code{0} may be fine but for audio recorded from analog,
4456 you may wish to increase the value to account for background noise.
4457 Can be specified in dB (in case "dB" is appended to the specified value)
4458 or amplitude ratio. Default value is @code{0}.
4461 Specify max duration of silence at beginning that will be kept after
4462 trimming. Default is 0, which is equal to trimming all samples detected
4466 Specify mode of detection of silence end in start of multi-channel audio.
4467 Can be @var{any} or @var{all}. Default is @var{any}.
4468 With @var{any}, any sample that is detected as non-silence will cause
4469 stopped trimming of silence.
4470 With @var{all}, only if all channels are detected as non-silence will cause
4471 stopped trimming of silence.
4474 Set the count for trimming silence from the end of audio.
4475 To remove silence from the middle of a file, specify a @var{stop_periods}
4476 that is negative. This value is then treated as a positive value and is
4477 used to indicate the effect should restart processing as specified by
4478 @var{start_periods}, making it suitable for removing periods of silence
4479 in the middle of the audio.
4480 Default value is @code{0}.
4483 Specify a duration of silence that must exist before audio is not copied any
4484 more. By specifying a higher duration, silence that is wanted can be left in
4486 Default value is @code{0}.
4488 @item stop_threshold
4489 This is the same as @option{start_threshold} but for trimming silence from
4491 Can be specified in dB (in case "dB" is appended to the specified value)
4492 or amplitude ratio. Default value is @code{0}.
4495 Specify max duration of silence at end that will be kept after
4496 trimming. Default is 0, which is equal to trimming all samples detected
4500 Specify mode of detection of silence start in end of multi-channel audio.
4501 Can be @var{any} or @var{all}. Default is @var{any}.
4502 With @var{any}, any sample that is detected as non-silence will cause
4503 stopped trimming of silence.
4504 With @var{all}, only if all channels are detected as non-silence will cause
4505 stopped trimming of silence.
4508 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4509 and works better with digital silence which is exactly 0.
4510 Default value is @code{rms}.
4513 Set duration in number of seconds used to calculate size of window in number
4514 of samples for detecting silence.
4515 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4518 @subsection Examples
4522 The following example shows how this filter can be used to start a recording
4523 that does not contain the delay at the start which usually occurs between
4524 pressing the record button and the start of the performance:
4526 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4530 Trim all silence encountered from beginning to end where there is more than 1
4531 second of silence in audio:
4533 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4539 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4540 loudspeakers around the user for binaural listening via headphones (audio
4541 formats up to 9 channels supported).
4542 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4543 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4544 Austrian Academy of Sciences.
4546 To enable compilation of this filter you need to configure FFmpeg with
4547 @code{--enable-libmysofa}.
4549 The filter accepts the following options:
4553 Set the SOFA file used for rendering.
4556 Set gain applied to audio. Value is in dB. Default is 0.
4559 Set rotation of virtual loudspeakers in deg. Default is 0.
4562 Set elevation of virtual speakers in deg. Default is 0.
4565 Set distance in meters between loudspeakers and the listener with near-field
4566 HRTFs. Default is 1.
4569 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4570 processing audio in time domain which is slow.
4571 @var{freq} is processing audio in frequency domain which is fast.
4572 Default is @var{freq}.
4575 Set custom positions of virtual loudspeakers. Syntax for this option is:
4576 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4577 Each virtual loudspeaker is described with short channel name following with
4578 azimuth and elevation in degrees.
4579 Each virtual loudspeaker description is separated by '|'.
4580 For example to override front left and front right channel positions use:
4581 'speakers=FL 45 15|FR 345 15'.
4582 Descriptions with unrecognised channel names are ignored.
4585 Set custom gain for LFE channels. Value is in dB. Default is 0.
4588 Set custom frame size in number of samples. Default is 1024.
4589 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4590 is set to @var{freq}.
4593 Should all IRs be normalized upon importing SOFA file.
4594 By default is enabled.
4597 Should nearest IRs be interpolated with neighbor IRs if exact position
4598 does not match. By default is disabled.
4601 Minphase all IRs upon loading of SOFA file. By default is disabled.
4604 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4607 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4610 @subsection Examples
4614 Using ClubFritz6 sofa file:
4616 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4620 Using ClubFritz12 sofa file and bigger radius with small rotation:
4622 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4626 Similar as above but with custom speaker positions for front left, front right, back left and back right
4627 and also with custom gain:
4629 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4633 @section stereotools
4635 This filter has some handy utilities to manage stereo signals, for converting
4636 M/S stereo recordings to L/R signal while having control over the parameters
4637 or spreading the stereo image of master track.
4639 The filter accepts the following options:
4643 Set input level before filtering for both channels. Defaults is 1.
4644 Allowed range is from 0.015625 to 64.
4647 Set output level after filtering for both channels. Defaults is 1.
4648 Allowed range is from 0.015625 to 64.
4651 Set input balance between both channels. Default is 0.
4652 Allowed range is from -1 to 1.
4655 Set output balance between both channels. Default is 0.
4656 Allowed range is from -1 to 1.
4659 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4660 clipping. Disabled by default.
4663 Mute the left channel. Disabled by default.
4666 Mute the right channel. Disabled by default.
4669 Change the phase of the left channel. Disabled by default.
4672 Change the phase of the right channel. Disabled by default.
4675 Set stereo mode. Available values are:
4679 Left/Right to Left/Right, this is default.
4682 Left/Right to Mid/Side.
4685 Mid/Side to Left/Right.
4688 Left/Right to Left/Left.
4691 Left/Right to Right/Right.
4694 Left/Right to Left + Right.
4697 Left/Right to Right/Left.
4700 Mid/Side to Left/Left.
4703 Mid/Side to Right/Right.
4707 Set level of side signal. Default is 1.
4708 Allowed range is from 0.015625 to 64.
4711 Set balance of side signal. Default is 0.
4712 Allowed range is from -1 to 1.
4715 Set level of the middle signal. Default is 1.
4716 Allowed range is from 0.015625 to 64.
4719 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4722 Set stereo base between mono and inversed channels. Default is 0.
4723 Allowed range is from -1 to 1.
4726 Set delay in milliseconds how much to delay left from right channel and
4727 vice versa. Default is 0. Allowed range is from -20 to 20.
4730 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4733 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4735 @item bmode_in, bmode_out
4736 Set balance mode for balance_in/balance_out option.
4738 Can be one of the following:
4742 Classic balance mode. Attenuate one channel at time.
4743 Gain is raised up to 1.
4746 Similar as classic mode above but gain is raised up to 2.
4749 Equal power distribution, from -6dB to +6dB range.
4753 @subsection Examples
4757 Apply karaoke like effect:
4759 stereotools=mlev=0.015625
4763 Convert M/S signal to L/R:
4765 "stereotools=mode=ms>lr"
4769 @section stereowiden
4771 This filter enhance the stereo effect by suppressing signal common to both
4772 channels and by delaying the signal of left into right and vice versa,
4773 thereby widening the stereo effect.
4775 The filter accepts the following options:
4779 Time in milliseconds of the delay of left signal into right and vice versa.
4780 Default is 20 milliseconds.
4783 Amount of gain in delayed signal into right and vice versa. Gives a delay
4784 effect of left signal in right output and vice versa which gives widening
4785 effect. Default is 0.3.
4788 Cross feed of left into right with inverted phase. This helps in suppressing
4789 the mono. If the value is 1 it will cancel all the signal common to both
4790 channels. Default is 0.3.
4793 Set level of input signal of original channel. Default is 0.8.
4796 @section superequalizer
4797 Apply 18 band equalizer.
4799 The filter accepts the following options:
4806 Set 131Hz band gain.
4808 Set 185Hz band gain.
4810 Set 262Hz band gain.
4812 Set 370Hz band gain.
4814 Set 523Hz band gain.
4816 Set 740Hz band gain.
4818 Set 1047Hz band gain.
4820 Set 1480Hz band gain.
4822 Set 2093Hz band gain.
4824 Set 2960Hz band gain.
4826 Set 4186Hz band gain.
4828 Set 5920Hz band gain.
4830 Set 8372Hz band gain.
4832 Set 11840Hz band gain.
4834 Set 16744Hz band gain.
4836 Set 20000Hz band gain.
4840 Apply audio surround upmix filter.
4842 This filter allows to produce multichannel output from audio stream.
4844 The filter accepts the following options:
4848 Set output channel layout. By default, this is @var{5.1}.
4850 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4851 for the required syntax.
4854 Set input channel layout. By default, this is @var{stereo}.
4856 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4857 for the required syntax.
4860 Set input volume level. By default, this is @var{1}.
4863 Set output volume level. By default, this is @var{1}.
4866 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4869 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4872 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4875 Set front center input volume. By default, this is @var{1}.
4878 Set front center output volume. By default, this is @var{1}.
4881 Set LFE input volume. By default, this is @var{1}.
4884 Set LFE output volume. By default, this is @var{1}.
4887 Set window function, default is @code{hann}.
4890 Set window overlap. If set to 1, the recommended overlap for selected
4891 window function will be picked. Default is @code{0.5}.
4894 @section treble, highshelf
4896 Boost or cut treble (upper) frequencies of the audio using a two-pole
4897 shelving filter with a response similar to that of a standard
4898 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4900 The filter accepts the following options:
4904 Give the gain at whichever is the lower of ~22 kHz and the
4905 Nyquist frequency. Its useful range is about -20 (for a large cut)
4906 to +20 (for a large boost). Beware of clipping when using a positive gain.
4909 Set the filter's central frequency and so can be used
4910 to extend or reduce the frequency range to be boosted or cut.
4911 The default value is @code{3000} Hz.
4914 Set method to specify band-width of filter.
4929 Determine how steep is the filter's shelf transition.
4932 Specify which channels to filter, by default all available are filtered.
4935 @subsection Commands
4937 This filter supports the following commands:
4940 Change treble frequency.
4941 Syntax for the command is : "@var{frequency}"
4944 Change treble width_type.
4945 Syntax for the command is : "@var{width_type}"
4948 Change treble width.
4949 Syntax for the command is : "@var{width}"
4953 Syntax for the command is : "@var{gain}"
4958 Sinusoidal amplitude modulation.
4960 The filter accepts the following options:
4964 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4965 (20 Hz or lower) will result in a tremolo effect.
4966 This filter may also be used as a ring modulator by specifying
4967 a modulation frequency higher than 20 Hz.
4968 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4971 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4972 Default value is 0.5.
4977 Sinusoidal phase modulation.
4979 The filter accepts the following options:
4983 Modulation frequency in Hertz.
4984 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4987 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4988 Default value is 0.5.
4993 Adjust the input audio volume.
4995 It accepts the following parameters:
4999 Set audio volume expression.
5001 Output values are clipped to the maximum value.
5003 The output audio volume is given by the relation:
5005 @var{output_volume} = @var{volume} * @var{input_volume}
5008 The default value for @var{volume} is "1.0".
5011 This parameter represents the mathematical precision.
5013 It determines which input sample formats will be allowed, which affects the
5014 precision of the volume scaling.
5018 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
5020 32-bit floating-point; this limits input sample format to FLT. (default)
5022 64-bit floating-point; this limits input sample format to DBL.
5026 Choose the behaviour on encountering ReplayGain side data in input frames.
5030 Remove ReplayGain side data, ignoring its contents (the default).
5033 Ignore ReplayGain side data, but leave it in the frame.
5036 Prefer the track gain, if present.
5039 Prefer the album gain, if present.
5042 @item replaygain_preamp
5043 Pre-amplification gain in dB to apply to the selected replaygain gain.
5045 Default value for @var{replaygain_preamp} is 0.0.
5048 Set when the volume expression is evaluated.
5050 It accepts the following values:
5053 only evaluate expression once during the filter initialization, or
5054 when the @samp{volume} command is sent
5057 evaluate expression for each incoming frame
5060 Default value is @samp{once}.
5063 The volume expression can contain the following parameters.
5067 frame number (starting at zero)
5070 @item nb_consumed_samples
5071 number of samples consumed by the filter
5073 number of samples in the current frame
5075 original frame position in the file
5081 PTS at start of stream
5083 time at start of stream
5089 last set volume value
5092 Note that when @option{eval} is set to @samp{once} only the
5093 @var{sample_rate} and @var{tb} variables are available, all other
5094 variables will evaluate to NAN.
5096 @subsection Commands
5098 This filter supports the following commands:
5101 Modify the volume expression.
5102 The command accepts the same syntax of the corresponding option.
5104 If the specified expression is not valid, it is kept at its current
5106 @item replaygain_noclip
5107 Prevent clipping by limiting the gain applied.
5109 Default value for @var{replaygain_noclip} is 1.
5113 @subsection Examples
5117 Halve the input audio volume:
5121 volume=volume=-6.0206dB
5124 In all the above example the named key for @option{volume} can be
5125 omitted, for example like in:
5131 Increase input audio power by 6 decibels using fixed-point precision:
5133 volume=volume=6dB:precision=fixed
5137 Fade volume after time 10 with an annihilation period of 5 seconds:
5139 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5143 @section volumedetect
5145 Detect the volume of the input video.
5147 The filter has no parameters. The input is not modified. Statistics about
5148 the volume will be printed in the log when the input stream end is reached.
5150 In particular it will show the mean volume (root mean square), maximum
5151 volume (on a per-sample basis), and the beginning of a histogram of the
5152 registered volume values (from the maximum value to a cumulated 1/1000 of
5155 All volumes are in decibels relative to the maximum PCM value.
5157 @subsection Examples
5159 Here is an excerpt of the output:
5161 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5162 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5163 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5164 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5165 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5166 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5167 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5168 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5169 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5175 The mean square energy is approximately -27 dB, or 10^-2.7.
5177 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5179 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5182 In other words, raising the volume by +4 dB does not cause any clipping,
5183 raising it by +5 dB causes clipping for 6 samples, etc.
5185 @c man end AUDIO FILTERS
5187 @chapter Audio Sources
5188 @c man begin AUDIO SOURCES
5190 Below is a description of the currently available audio sources.
5194 Buffer audio frames, and make them available to the filter chain.
5196 This source is mainly intended for a programmatic use, in particular
5197 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5199 It accepts the following parameters:
5203 The timebase which will be used for timestamps of submitted frames. It must be
5204 either a floating-point number or in @var{numerator}/@var{denominator} form.
5207 The sample rate of the incoming audio buffers.
5210 The sample format of the incoming audio buffers.
5211 Either a sample format name or its corresponding integer representation from
5212 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5214 @item channel_layout
5215 The channel layout of the incoming audio buffers.
5216 Either a channel layout name from channel_layout_map in
5217 @file{libavutil/channel_layout.c} or its corresponding integer representation
5218 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5221 The number of channels of the incoming audio buffers.
5222 If both @var{channels} and @var{channel_layout} are specified, then they
5227 @subsection Examples
5230 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5233 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5234 Since the sample format with name "s16p" corresponds to the number
5235 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5238 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5243 Generate an audio signal specified by an expression.
5245 This source accepts in input one or more expressions (one for each
5246 channel), which are evaluated and used to generate a corresponding
5249 This source accepts the following options:
5253 Set the '|'-separated expressions list for each separate channel. In case the
5254 @option{channel_layout} option is not specified, the selected channel layout
5255 depends on the number of provided expressions. Otherwise the last
5256 specified expression is applied to the remaining output channels.
5258 @item channel_layout, c
5259 Set the channel layout. The number of channels in the specified layout
5260 must be equal to the number of specified expressions.
5263 Set the minimum duration of the sourced audio. See
5264 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5265 for the accepted syntax.
5266 Note that the resulting duration may be greater than the specified
5267 duration, as the generated audio is always cut at the end of a
5270 If not specified, or the expressed duration is negative, the audio is
5271 supposed to be generated forever.
5274 Set the number of samples per channel per each output frame,
5277 @item sample_rate, s
5278 Specify the sample rate, default to 44100.
5281 Each expression in @var{exprs} can contain the following constants:
5285 number of the evaluated sample, starting from 0
5288 time of the evaluated sample expressed in seconds, starting from 0
5295 @subsection Examples
5305 Generate a sin signal with frequency of 440 Hz, set sample rate to
5308 aevalsrc="sin(440*2*PI*t):s=8000"
5312 Generate a two channels signal, specify the channel layout (Front
5313 Center + Back Center) explicitly:
5315 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5319 Generate white noise:
5321 aevalsrc="-2+random(0)"
5325 Generate an amplitude modulated signal:
5327 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5331 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5333 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5340 The null audio source, return unprocessed audio frames. It is mainly useful
5341 as a template and to be employed in analysis / debugging tools, or as
5342 the source for filters which ignore the input data (for example the sox
5345 This source accepts the following options:
5349 @item channel_layout, cl
5351 Specifies the channel layout, and can be either an integer or a string
5352 representing a channel layout. The default value of @var{channel_layout}
5355 Check the channel_layout_map definition in
5356 @file{libavutil/channel_layout.c} for the mapping between strings and
5357 channel layout values.
5359 @item sample_rate, r
5360 Specifies the sample rate, and defaults to 44100.
5363 Set the number of samples per requested frames.
5367 @subsection Examples
5371 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5373 anullsrc=r=48000:cl=4
5377 Do the same operation with a more obvious syntax:
5379 anullsrc=r=48000:cl=mono
5383 All the parameters need to be explicitly defined.
5387 Synthesize a voice utterance using the libflite library.
5389 To enable compilation of this filter you need to configure FFmpeg with
5390 @code{--enable-libflite}.
5392 Note that versions of the flite library prior to 2.0 are not thread-safe.
5394 The filter accepts the following options:
5399 If set to 1, list the names of the available voices and exit
5400 immediately. Default value is 0.
5403 Set the maximum number of samples per frame. Default value is 512.
5406 Set the filename containing the text to speak.
5409 Set the text to speak.
5412 Set the voice to use for the speech synthesis. Default value is
5413 @code{kal}. See also the @var{list_voices} option.
5416 @subsection Examples
5420 Read from file @file{speech.txt}, and synthesize the text using the
5421 standard flite voice:
5423 flite=textfile=speech.txt
5427 Read the specified text selecting the @code{slt} voice:
5429 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5433 Input text to ffmpeg:
5435 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5439 Make @file{ffplay} speak the specified text, using @code{flite} and
5440 the @code{lavfi} device:
5442 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5446 For more information about libflite, check:
5447 @url{http://www.festvox.org/flite/}
5451 Generate a noise audio signal.
5453 The filter accepts the following options:
5456 @item sample_rate, r
5457 Specify the sample rate. Default value is 48000 Hz.
5460 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5464 Specify the duration of the generated audio stream. Not specifying this option
5465 results in noise with an infinite length.
5467 @item color, colour, c
5468 Specify the color of noise. Available noise colors are white, pink, brown,
5469 blue and violet. Default color is white.
5472 Specify a value used to seed the PRNG.
5475 Set the number of samples per each output frame, default is 1024.
5478 @subsection Examples
5483 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5485 anoisesrc=d=60:c=pink:r=44100:a=0.5
5491 Generate odd-tap Hilbert transform FIR coefficients.
5493 The resulting stream can be used with @ref{afir} filter for phase-shifting
5494 the signal by 90 degrees.
5496 This is used in many matrix coding schemes and for analytic signal generation.
5497 The process is often written as a multiplication by i (or j), the imaginary unit.
5499 The filter accepts the following options:
5503 @item sample_rate, s
5504 Set sample rate, default is 44100.
5507 Set length of FIR filter, default is 22051.
5510 Set number of samples per each frame.
5513 Set window function to be used when generating FIR coefficients.
5518 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5520 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5522 The filter accepts the following options:
5525 @item sample_rate, r
5526 Set sample rate, default is 44100.
5529 Set number of samples per each frame. Default is 1024.
5532 Set high-pass frequency. Default is 0.
5535 Set low-pass frequency. Default is 0.
5536 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5537 is higher than 0 then filter will create band-pass filter coefficients,
5538 otherwise band-reject filter coefficients.
5541 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5544 Set Kaiser window beta.
5547 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5550 Enable rounding, by default is disabled.
5553 Set number of taps for high-pass filter.
5556 Set number of taps for low-pass filter.
5561 Generate an audio signal made of a sine wave with amplitude 1/8.
5563 The audio signal is bit-exact.
5565 The filter accepts the following options:
5570 Set the carrier frequency. Default is 440 Hz.
5572 @item beep_factor, b
5573 Enable a periodic beep every second with frequency @var{beep_factor} times
5574 the carrier frequency. Default is 0, meaning the beep is disabled.
5576 @item sample_rate, r
5577 Specify the sample rate, default is 44100.
5580 Specify the duration of the generated audio stream.
5582 @item samples_per_frame
5583 Set the number of samples per output frame.
5585 The expression can contain the following constants:
5589 The (sequential) number of the output audio frame, starting from 0.
5592 The PTS (Presentation TimeStamp) of the output audio frame,
5593 expressed in @var{TB} units.
5596 The PTS of the output audio frame, expressed in seconds.
5599 The timebase of the output audio frames.
5602 Default is @code{1024}.
5605 @subsection Examples
5610 Generate a simple 440 Hz sine wave:
5616 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5620 sine=frequency=220:beep_factor=4:duration=5
5624 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5627 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5631 @c man end AUDIO SOURCES
5633 @chapter Audio Sinks
5634 @c man begin AUDIO SINKS
5636 Below is a description of the currently available audio sinks.
5638 @section abuffersink
5640 Buffer audio frames, and make them available to the end of filter chain.
5642 This sink is mainly intended for programmatic use, in particular
5643 through the interface defined in @file{libavfilter/buffersink.h}
5644 or the options system.
5646 It accepts a pointer to an AVABufferSinkContext structure, which
5647 defines the incoming buffers' formats, to be passed as the opaque
5648 parameter to @code{avfilter_init_filter} for initialization.
5651 Null audio sink; do absolutely nothing with the input audio. It is
5652 mainly useful as a template and for use in analysis / debugging
5655 @c man end AUDIO SINKS
5657 @chapter Video Filters
5658 @c man begin VIDEO FILTERS
5660 When you configure your FFmpeg build, you can disable any of the
5661 existing filters using @code{--disable-filters}.
5662 The configure output will show the video filters included in your
5665 Below is a description of the currently available video filters.
5667 @section alphaextract
5669 Extract the alpha component from the input as a grayscale video. This
5670 is especially useful with the @var{alphamerge} filter.
5674 Add or replace the alpha component of the primary input with the
5675 grayscale value of a second input. This is intended for use with
5676 @var{alphaextract} to allow the transmission or storage of frame
5677 sequences that have alpha in a format that doesn't support an alpha
5680 For example, to reconstruct full frames from a normal YUV-encoded video
5681 and a separate video created with @var{alphaextract}, you might use:
5683 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5686 Since this filter is designed for reconstruction, it operates on frame
5687 sequences without considering timestamps, and terminates when either
5688 input reaches end of stream. This will cause problems if your encoding
5689 pipeline drops frames. If you're trying to apply an image as an
5690 overlay to a video stream, consider the @var{overlay} filter instead.
5694 Amplify differences between current pixel and pixels of adjacent frames in
5695 same pixel location.
5697 This filter accepts the following options:
5701 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5702 For example radius of 3 will instruct filter to calculate average of 7 frames.
5705 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5708 Set threshold for difference amplification. Any difference greater or equal to
5709 this value will not alter source pixel. Default is 10.
5710 Allowed range is from 0 to 65535.
5713 Set tolerance for difference amplification. Any difference lower to
5714 this value will not alter source pixel. Default is 0.
5715 Allowed range is from 0 to 65535.
5718 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5719 This option controls maximum possible value that will decrease source pixel value.
5722 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5723 This option controls maximum possible value that will increase source pixel value.
5726 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5731 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5732 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5733 Substation Alpha) subtitles files.
5735 This filter accepts the following option in addition to the common options from
5736 the @ref{subtitles} filter:
5740 Set the shaping engine
5742 Available values are:
5745 The default libass shaping engine, which is the best available.
5747 Fast, font-agnostic shaper that can do only substitutions
5749 Slower shaper using OpenType for substitutions and positioning
5752 The default is @code{auto}.
5756 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5758 The filter accepts the following options:
5762 Set threshold A for 1st plane. Default is 0.02.
5763 Valid range is 0 to 0.3.
5766 Set threshold B for 1st plane. Default is 0.04.
5767 Valid range is 0 to 5.
5770 Set threshold A for 2nd plane. Default is 0.02.
5771 Valid range is 0 to 0.3.
5774 Set threshold B for 2nd plane. Default is 0.04.
5775 Valid range is 0 to 5.
5778 Set threshold A for 3rd plane. Default is 0.02.
5779 Valid range is 0 to 0.3.
5782 Set threshold B for 3rd plane. Default is 0.04.
5783 Valid range is 0 to 5.
5785 Threshold A is designed to react on abrupt changes in the input signal and
5786 threshold B is designed to react on continuous changes in the input signal.
5789 Set number of frames filter will use for averaging. Default is 9. Must be odd
5790 number in range [5, 129].
5793 Set what planes of frame filter will use for averaging. Default is all.
5798 Apply average blur filter.
5800 The filter accepts the following options:
5804 Set horizontal radius size.
5807 Set which planes to filter. By default all planes are filtered.
5810 Set vertical radius size, if zero it will be same as @code{sizeX}.
5811 Default is @code{0}.
5816 Compute the bounding box for the non-black pixels in the input frame
5819 This filter computes the bounding box containing all the pixels with a
5820 luminance value greater than the minimum allowed value.
5821 The parameters describing the bounding box are printed on the filter
5824 The filter accepts the following option:
5828 Set the minimal luminance value. Default is @code{16}.
5831 @section bitplanenoise
5833 Show and measure bit plane noise.
5835 The filter accepts the following options:
5839 Set which plane to analyze. Default is @code{1}.
5842 Filter out noisy pixels from @code{bitplane} set above.
5843 Default is disabled.
5846 @section blackdetect
5848 Detect video intervals that are (almost) completely black. Can be
5849 useful to detect chapter transitions, commercials, or invalid
5850 recordings. Output lines contains the time for the start, end and
5851 duration of the detected black interval expressed in seconds.
5853 In order to display the output lines, you need to set the loglevel at
5854 least to the AV_LOG_INFO value.
5856 The filter accepts the following options:
5859 @item black_min_duration, d
5860 Set the minimum detected black duration expressed in seconds. It must
5861 be a non-negative floating point number.
5863 Default value is 2.0.
5865 @item picture_black_ratio_th, pic_th
5866 Set the threshold for considering a picture "black".
5867 Express the minimum value for the ratio:
5869 @var{nb_black_pixels} / @var{nb_pixels}
5872 for which a picture is considered black.
5873 Default value is 0.98.
5875 @item pixel_black_th, pix_th
5876 Set the threshold for considering a pixel "black".
5878 The threshold expresses the maximum pixel luminance value for which a
5879 pixel is considered "black". The provided value is scaled according to
5880 the following equation:
5882 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5885 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5886 the input video format, the range is [0-255] for YUV full-range
5887 formats and [16-235] for YUV non full-range formats.
5889 Default value is 0.10.
5892 The following example sets the maximum pixel threshold to the minimum
5893 value, and detects only black intervals of 2 or more seconds:
5895 blackdetect=d=2:pix_th=0.00
5900 Detect frames that are (almost) completely black. Can be useful to
5901 detect chapter transitions or commercials. Output lines consist of
5902 the frame number of the detected frame, the percentage of blackness,
5903 the position in the file if known or -1 and the timestamp in seconds.
5905 In order to display the output lines, you need to set the loglevel at
5906 least to the AV_LOG_INFO value.
5908 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5909 The value represents the percentage of pixels in the picture that
5910 are below the threshold value.
5912 It accepts the following parameters:
5917 The percentage of the pixels that have to be below the threshold; it defaults to
5920 @item threshold, thresh
5921 The threshold below which a pixel value is considered black; it defaults to
5926 @section blend, tblend
5928 Blend two video frames into each other.
5930 The @code{blend} filter takes two input streams and outputs one
5931 stream, the first input is the "top" layer and second input is
5932 "bottom" layer. By default, the output terminates when the longest input terminates.
5934 The @code{tblend} (time blend) filter takes two consecutive frames
5935 from one single stream, and outputs the result obtained by blending
5936 the new frame on top of the old frame.
5938 A description of the accepted options follows.
5946 Set blend mode for specific pixel component or all pixel components in case
5947 of @var{all_mode}. Default value is @code{normal}.
5949 Available values for component modes are:
5991 Set blend opacity for specific pixel component or all pixel components in case
5992 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5999 Set blend expression for specific pixel component or all pixel components in case
6000 of @var{all_expr}. Note that related mode options will be ignored if those are set.
6002 The expressions can use the following variables:
6006 The sequential number of the filtered frame, starting from @code{0}.
6010 the coordinates of the current sample
6014 the width and height of currently filtered plane
6018 Width and height scale for the plane being filtered. It is the
6019 ratio between the dimensions of the current plane to the luma plane,
6020 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
6021 the luma plane and @code{0.5,0.5} for the chroma planes.
6024 Time of the current frame, expressed in seconds.
6027 Value of pixel component at current location for first video frame (top layer).
6030 Value of pixel component at current location for second video frame (bottom layer).
6034 The @code{blend} filter also supports the @ref{framesync} options.
6036 @subsection Examples
6040 Apply transition from bottom layer to top layer in first 10 seconds:
6042 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6046 Apply linear horizontal transition from top layer to bottom layer:
6048 blend=all_expr='A*(X/W)+B*(1-X/W)'
6052 Apply 1x1 checkerboard effect:
6054 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6058 Apply uncover left effect:
6060 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6064 Apply uncover down effect:
6066 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6070 Apply uncover up-left effect:
6072 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6076 Split diagonally video and shows top and bottom layer on each side:
6078 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6082 Display differences between the current and the previous frame:
6084 tblend=all_mode=grainextract
6090 Denoise frames using Block-Matching 3D algorithm.
6092 The filter accepts the following options.
6096 Set denoising strength. Default value is 1.
6097 Allowed range is from 0 to 999.9.
6098 The denoising algorithm is very sensitive to sigma, so adjust it
6099 according to the source.
6102 Set local patch size. This sets dimensions in 2D.
6105 Set sliding step for processing blocks. Default value is 4.
6106 Allowed range is from 1 to 64.
6107 Smaller values allows processing more reference blocks and is slower.
6110 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6111 When set to 1, no block matching is done. Larger values allows more blocks
6113 Allowed range is from 1 to 256.
6116 Set radius for search block matching. Default is 9.
6117 Allowed range is from 1 to INT32_MAX.
6120 Set step between two search locations for block matching. Default is 1.
6121 Allowed range is from 1 to 64. Smaller is slower.
6124 Set threshold of mean square error for block matching. Valid range is 0 to
6128 Set thresholding parameter for hard thresholding in 3D transformed domain.
6129 Larger values results in stronger hard-thresholding filtering in frequency
6133 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6134 Default is @code{basic}.
6137 If enabled, filter will use 2nd stream for block matching.
6138 Default is disabled for @code{basic} value of @var{estim} option,
6139 and always enabled if value of @var{estim} is @code{final}.
6142 Set planes to filter. Default is all available except alpha.
6145 @subsection Examples
6149 Basic filtering with bm3d:
6151 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6155 Same as above, but filtering only luma:
6157 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6161 Same as above, but with both estimation modes:
6163 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
6167 Same as above, but prefilter with @ref{nlmeans} filter instead:
6169 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
6175 Apply a boxblur algorithm to the input video.
6177 It accepts the following parameters:
6181 @item luma_radius, lr
6182 @item luma_power, lp
6183 @item chroma_radius, cr
6184 @item chroma_power, cp
6185 @item alpha_radius, ar
6186 @item alpha_power, ap
6190 A description of the accepted options follows.
6193 @item luma_radius, lr
6194 @item chroma_radius, cr
6195 @item alpha_radius, ar
6196 Set an expression for the box radius in pixels used for blurring the
6197 corresponding input plane.
6199 The radius value must be a non-negative number, and must not be
6200 greater than the value of the expression @code{min(w,h)/2} for the
6201 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6204 Default value for @option{luma_radius} is "2". If not specified,
6205 @option{chroma_radius} and @option{alpha_radius} default to the
6206 corresponding value set for @option{luma_radius}.
6208 The expressions can contain the following constants:
6212 The input width and height in pixels.
6216 The input chroma image width and height in pixels.
6220 The horizontal and vertical chroma subsample values. For example, for the
6221 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6224 @item luma_power, lp
6225 @item chroma_power, cp
6226 @item alpha_power, ap
6227 Specify how many times the boxblur filter is applied to the
6228 corresponding plane.
6230 Default value for @option{luma_power} is 2. If not specified,
6231 @option{chroma_power} and @option{alpha_power} default to the
6232 corresponding value set for @option{luma_power}.
6234 A value of 0 will disable the effect.
6237 @subsection Examples
6241 Apply a boxblur filter with the luma, chroma, and alpha radii
6244 boxblur=luma_radius=2:luma_power=1
6249 Set the luma radius to 2, and alpha and chroma radius to 0:
6251 boxblur=2:1:cr=0:ar=0
6255 Set the luma and chroma radii to a fraction of the video dimension:
6257 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6263 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6264 Deinterlacing Filter").
6266 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6267 interpolation algorithms.
6268 It accepts the following parameters:
6272 The interlacing mode to adopt. It accepts one of the following values:
6276 Output one frame for each frame.
6278 Output one frame for each field.
6281 The default value is @code{send_field}.
6284 The picture field parity assumed for the input interlaced video. It accepts one
6285 of the following values:
6289 Assume the top field is first.
6291 Assume the bottom field is first.
6293 Enable automatic detection of field parity.
6296 The default value is @code{auto}.
6297 If the interlacing is unknown or the decoder does not export this information,
6298 top field first will be assumed.
6301 Specify which frames to deinterlace. Accept one of the following
6306 Deinterlace all frames.
6308 Only deinterlace frames marked as interlaced.
6311 The default value is @code{all}.
6315 Remove all color information for all colors except for certain one.
6317 The filter accepts the following options:
6321 The color which will not be replaced with neutral chroma.
6324 Similarity percentage with the above color.
6325 0.01 matches only the exact key color, while 1.0 matches everything.
6328 Signals that the color passed is already in YUV instead of RGB.
6330 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6331 This can be used to pass exact YUV values as hexadecimal numbers.
6335 YUV colorspace color/chroma keying.
6337 The filter accepts the following options:
6341 The color which will be replaced with transparency.
6344 Similarity percentage with the key color.
6346 0.01 matches only the exact key color, while 1.0 matches everything.
6351 0.0 makes pixels either fully transparent, or not transparent at all.
6353 Higher values result in semi-transparent pixels, with a higher transparency
6354 the more similar the pixels color is to the key color.
6357 Signals that the color passed is already in YUV instead of RGB.
6359 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6360 This can be used to pass exact YUV values as hexadecimal numbers.
6363 @subsection Examples
6367 Make every green pixel in the input image transparent:
6369 ffmpeg -i input.png -vf chromakey=green out.png
6373 Overlay a greenscreen-video on top of a static black background.
6375 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
6379 @section chromashift
6380 Shift chroma pixels horizontally and/or vertically.
6382 The filter accepts the following options:
6385 Set amount to shift chroma-blue horizontally.
6387 Set amount to shift chroma-blue vertically.
6389 Set amount to shift chroma-red horizontally.
6391 Set amount to shift chroma-red vertically.
6393 Set edge mode, can be @var{smear}, default, or @var{warp}.
6398 Display CIE color diagram with pixels overlaid onto it.
6400 The filter accepts the following options:
6415 @item uhdtv, rec2020
6428 Set what gamuts to draw.
6430 See @code{system} option for available values.
6433 Set ciescope size, by default set to 512.
6436 Set intensity used to map input pixel values to CIE diagram.
6439 Set contrast used to draw tongue colors that are out of active color system gamut.
6442 Correct gamma displayed on scope, by default enabled.
6445 Show white point on CIE diagram, by default disabled.
6448 Set input gamma. Used only with XYZ input color space.
6453 Visualize information exported by some codecs.
6455 Some codecs can export information through frames using side-data or other
6456 means. For example, some MPEG based codecs export motion vectors through the
6457 @var{export_mvs} flag in the codec @option{flags2} option.
6459 The filter accepts the following option:
6463 Set motion vectors to visualize.
6465 Available flags for @var{mv} are:
6469 forward predicted MVs of P-frames
6471 forward predicted MVs of B-frames
6473 backward predicted MVs of B-frames
6477 Display quantization parameters using the chroma planes.
6480 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6482 Available flags for @var{mv_type} are:
6486 forward predicted MVs
6488 backward predicted MVs
6491 @item frame_type, ft
6492 Set frame type to visualize motion vectors of.
6494 Available flags for @var{frame_type} are:
6498 intra-coded frames (I-frames)
6500 predicted frames (P-frames)
6502 bi-directionally predicted frames (B-frames)
6506 @subsection Examples
6510 Visualize forward predicted MVs of all frames using @command{ffplay}:
6512 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6516 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6518 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6522 @section colorbalance
6523 Modify intensity of primary colors (red, green and blue) of input frames.
6525 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6526 regions for the red-cyan, green-magenta or blue-yellow balance.
6528 A positive adjustment value shifts the balance towards the primary color, a negative
6529 value towards the complementary color.
6531 The filter accepts the following options:
6537 Adjust red, green and blue shadows (darkest pixels).
6542 Adjust red, green and blue midtones (medium pixels).
6547 Adjust red, green and blue highlights (brightest pixels).
6549 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6552 @subsection Examples
6556 Add red color cast to shadows:
6563 RGB colorspace color keying.
6565 The filter accepts the following options:
6569 The color which will be replaced with transparency.
6572 Similarity percentage with the key color.
6574 0.01 matches only the exact key color, while 1.0 matches everything.
6579 0.0 makes pixels either fully transparent, or not transparent at all.
6581 Higher values result in semi-transparent pixels, with a higher transparency
6582 the more similar the pixels color is to the key color.
6585 @subsection Examples
6589 Make every green pixel in the input image transparent:
6591 ffmpeg -i input.png -vf colorkey=green out.png
6595 Overlay a greenscreen-video on top of a static background image.
6597 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
6601 @section colorlevels
6603 Adjust video input frames using levels.
6605 The filter accepts the following options:
6612 Adjust red, green, blue and alpha input black point.
6613 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6619 Adjust red, green, blue and alpha input white point.
6620 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6622 Input levels are used to lighten highlights (bright tones), darken shadows
6623 (dark tones), change the balance of bright and dark tones.
6629 Adjust red, green, blue and alpha output black point.
6630 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6636 Adjust red, green, blue and alpha output white point.
6637 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6639 Output levels allows manual selection of a constrained output level range.
6642 @subsection Examples
6646 Make video output darker:
6648 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6654 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6658 Make video output lighter:
6660 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6664 Increase brightness:
6666 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6670 @section colorchannelmixer
6672 Adjust video input frames by re-mixing color channels.
6674 This filter modifies a color channel by adding the values associated to
6675 the other channels of the same pixels. For example if the value to
6676 modify is red, the output value will be:
6678 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6681 The filter accepts the following options:
6688 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6689 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6695 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6696 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6702 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6703 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6709 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6710 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6712 Allowed ranges for options are @code{[-2.0, 2.0]}.
6715 @subsection Examples
6719 Convert source to grayscale:
6721 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6724 Simulate sepia tones:
6726 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6730 @section colormatrix
6732 Convert color matrix.
6734 The filter accepts the following options:
6739 Specify the source and destination color matrix. Both values must be
6742 The accepted values are:
6770 For example to convert from BT.601 to SMPTE-240M, use the command:
6772 colormatrix=bt601:smpte240m
6777 Convert colorspace, transfer characteristics or color primaries.
6778 Input video needs to have an even size.
6780 The filter accepts the following options:
6785 Specify all color properties at once.
6787 The accepted values are:
6817 Specify output colorspace.
6819 The accepted values are:
6828 BT.470BG or BT.601-6 625
6831 SMPTE-170M or BT.601-6 525
6840 BT.2020 with non-constant luminance
6846 Specify output transfer characteristics.
6848 The accepted values are:
6860 Constant gamma of 2.2
6863 Constant gamma of 2.8
6866 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6884 BT.2020 for 10-bits content
6887 BT.2020 for 12-bits content
6893 Specify output color primaries.
6895 The accepted values are:
6904 BT.470BG or BT.601-6 625
6907 SMPTE-170M or BT.601-6 525
6931 Specify output color range.
6933 The accepted values are:
6936 TV (restricted) range
6939 MPEG (restricted) range
6950 Specify output color format.
6952 The accepted values are:
6955 YUV 4:2:0 planar 8-bits
6958 YUV 4:2:0 planar 10-bits
6961 YUV 4:2:0 planar 12-bits
6964 YUV 4:2:2 planar 8-bits
6967 YUV 4:2:2 planar 10-bits
6970 YUV 4:2:2 planar 12-bits
6973 YUV 4:4:4 planar 8-bits
6976 YUV 4:4:4 planar 10-bits
6979 YUV 4:4:4 planar 12-bits
6984 Do a fast conversion, which skips gamma/primary correction. This will take
6985 significantly less CPU, but will be mathematically incorrect. To get output
6986 compatible with that produced by the colormatrix filter, use fast=1.
6989 Specify dithering mode.
6991 The accepted values are:
6997 Floyd-Steinberg dithering
7001 Whitepoint adaptation mode.
7003 The accepted values are:
7006 Bradford whitepoint adaptation
7009 von Kries whitepoint adaptation
7012 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7016 Override all input properties at once. Same accepted values as @ref{all}.
7019 Override input colorspace. Same accepted values as @ref{space}.
7022 Override input color primaries. Same accepted values as @ref{primaries}.
7025 Override input transfer characteristics. Same accepted values as @ref{trc}.
7028 Override input color range. Same accepted values as @ref{range}.
7032 The filter converts the transfer characteristics, color space and color
7033 primaries to the specified user values. The output value, if not specified,
7034 is set to a default value based on the "all" property. If that property is
7035 also not specified, the filter will log an error. The output color range and
7036 format default to the same value as the input color range and format. The
7037 input transfer characteristics, color space, color primaries and color range
7038 should be set on the input data. If any of these are missing, the filter will
7039 log an error and no conversion will take place.
7041 For example to convert the input to SMPTE-240M, use the command:
7043 colorspace=smpte240m
7046 @section convolution
7048 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7050 The filter accepts the following options:
7057 Set matrix for each plane.
7058 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7059 and from 1 to 49 odd number of signed integers in @var{row} mode.
7065 Set multiplier for calculated value for each plane.
7066 If unset or 0, it will be sum of all matrix elements.
7072 Set bias for each plane. This value is added to the result of the multiplication.
7073 Useful for making the overall image brighter or darker. Default is 0.0.
7079 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7080 Default is @var{square}.
7083 @subsection Examples
7089 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"
7095 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"
7101 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"
7107 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"
7111 Apply laplacian edge detector which includes diagonals:
7113 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"
7119 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"
7125 Apply 2D convolution of video stream in frequency domain using second stream
7128 The filter accepts the following options:
7132 Set which planes to process.
7135 Set which impulse video frames will be processed, can be @var{first}
7136 or @var{all}. Default is @var{all}.
7139 The @code{convolve} filter also supports the @ref{framesync} options.
7143 Copy the input video source unchanged to the output. This is mainly useful for
7148 Video filtering on GPU using Apple's CoreImage API on OSX.
7150 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7151 processed by video hardware. However, software-based OpenGL implementations
7152 exist which means there is no guarantee for hardware processing. It depends on
7155 There are many filters and image generators provided by Apple that come with a
7156 large variety of options. The filter has to be referenced by its name along
7159 The coreimage filter accepts the following options:
7162 List all available filters and generators along with all their respective
7163 options as well as possible minimum and maximum values along with the default
7170 Specify all filters by their respective name and options.
7171 Use @var{list_filters} to determine all valid filter names and options.
7172 Numerical options are specified by a float value and are automatically clamped
7173 to their respective value range. Vector and color options have to be specified
7174 by a list of space separated float values. Character escaping has to be done.
7175 A special option name @code{default} is available to use default options for a
7178 It is required to specify either @code{default} or at least one of the filter options.
7179 All omitted options are used with their default values.
7180 The syntax of the filter string is as follows:
7182 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7186 Specify a rectangle where the output of the filter chain is copied into the
7187 input image. It is given by a list of space separated float values:
7189 output_rect=x\ y\ width\ height
7191 If not given, the output rectangle equals the dimensions of the input image.
7192 The output rectangle is automatically cropped at the borders of the input
7193 image. Negative values are valid for each component.
7195 output_rect=25\ 25\ 100\ 100
7199 Several filters can be chained for successive processing without GPU-HOST
7200 transfers allowing for fast processing of complex filter chains.
7201 Currently, only filters with zero (generators) or exactly one (filters) input
7202 image and one output image are supported. Also, transition filters are not yet
7205 Some filters generate output images with additional padding depending on the
7206 respective filter kernel. The padding is automatically removed to ensure the
7207 filter output has the same size as the input image.
7209 For image generators, the size of the output image is determined by the
7210 previous output image of the filter chain or the input image of the whole
7211 filterchain, respectively. The generators do not use the pixel information of
7212 this image to generate their output. However, the generated output is
7213 blended onto this image, resulting in partial or complete coverage of the
7216 The @ref{coreimagesrc} video source can be used for generating input images
7217 which are directly fed into the filter chain. By using it, providing input
7218 images by another video source or an input video is not required.
7220 @subsection Examples
7225 List all filters available:
7227 coreimage=list_filters=true
7231 Use the CIBoxBlur filter with default options to blur an image:
7233 coreimage=filter=CIBoxBlur@@default
7237 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7238 its center at 100x100 and a radius of 50 pixels:
7240 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7244 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7245 given as complete and escaped command-line for Apple's standard bash shell:
7247 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7253 Crop the input video to given dimensions.
7255 It accepts the following parameters:
7259 The width of the output video. It defaults to @code{iw}.
7260 This expression is evaluated only once during the filter
7261 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7264 The height of the output video. It defaults to @code{ih}.
7265 This expression is evaluated only once during the filter
7266 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7269 The horizontal position, in the input video, of the left edge of the output
7270 video. It defaults to @code{(in_w-out_w)/2}.
7271 This expression is evaluated per-frame.
7274 The vertical position, in the input video, of the top edge of the output video.
7275 It defaults to @code{(in_h-out_h)/2}.
7276 This expression is evaluated per-frame.
7279 If set to 1 will force the output display aspect ratio
7280 to be the same of the input, by changing the output sample aspect
7281 ratio. It defaults to 0.
7284 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7285 width/height/x/y as specified and will not be rounded to nearest smaller value.
7289 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7290 expressions containing the following constants:
7295 The computed values for @var{x} and @var{y}. They are evaluated for
7300 The input width and height.
7304 These are the same as @var{in_w} and @var{in_h}.
7308 The output (cropped) width and height.
7312 These are the same as @var{out_w} and @var{out_h}.
7315 same as @var{iw} / @var{ih}
7318 input sample aspect ratio
7321 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7325 horizontal and vertical chroma subsample values. For example for the
7326 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7329 The number of the input frame, starting from 0.
7332 the position in the file of the input frame, NAN if unknown
7335 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7339 The expression for @var{out_w} may depend on the value of @var{out_h},
7340 and the expression for @var{out_h} may depend on @var{out_w}, but they
7341 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7342 evaluated after @var{out_w} and @var{out_h}.
7344 The @var{x} and @var{y} parameters specify the expressions for the
7345 position of the top-left corner of the output (non-cropped) area. They
7346 are evaluated for each frame. If the evaluated value is not valid, it
7347 is approximated to the nearest valid value.
7349 The expression for @var{x} may depend on @var{y}, and the expression
7350 for @var{y} may depend on @var{x}.
7352 @subsection Examples
7356 Crop area with size 100x100 at position (12,34).
7361 Using named options, the example above becomes:
7363 crop=w=100:h=100:x=12:y=34
7367 Crop the central input area with size 100x100:
7373 Crop the central input area with size 2/3 of the input video:
7375 crop=2/3*in_w:2/3*in_h
7379 Crop the input video central square:
7386 Delimit the rectangle with the top-left corner placed at position
7387 100:100 and the right-bottom corner corresponding to the right-bottom
7388 corner of the input image.
7390 crop=in_w-100:in_h-100:100:100
7394 Crop 10 pixels from the left and right borders, and 20 pixels from
7395 the top and bottom borders
7397 crop=in_w-2*10:in_h-2*20
7401 Keep only the bottom right quarter of the input image:
7403 crop=in_w/2:in_h/2:in_w/2:in_h/2
7407 Crop height for getting Greek harmony:
7409 crop=in_w:1/PHI*in_w
7413 Apply trembling effect:
7415 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)
7419 Apply erratic camera effect depending on timestamp:
7421 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)"
7425 Set x depending on the value of y:
7427 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7431 @subsection Commands
7433 This filter supports the following commands:
7439 Set width/height of the output video and the horizontal/vertical position
7441 The command accepts the same syntax of the corresponding option.
7443 If the specified expression is not valid, it is kept at its current
7449 Auto-detect the crop size.
7451 It calculates the necessary cropping parameters and prints the
7452 recommended parameters via the logging system. The detected dimensions
7453 correspond to the non-black area of the input video.
7455 It accepts the following parameters:
7460 Set higher black value threshold, which can be optionally specified
7461 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7462 value greater to the set value is considered non-black. It defaults to 24.
7463 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7464 on the bitdepth of the pixel format.
7467 The value which the width/height should be divisible by. It defaults to
7468 16. The offset is automatically adjusted to center the video. Use 2 to
7469 get only even dimensions (needed for 4:2:2 video). 16 is best when
7470 encoding to most video codecs.
7472 @item reset_count, reset
7473 Set the counter that determines after how many frames cropdetect will
7474 reset the previously detected largest video area and start over to
7475 detect the current optimal crop area. Default value is 0.
7477 This can be useful when channel logos distort the video area. 0
7478 indicates 'never reset', and returns the largest area encountered during
7485 Delay video filtering until a given wallclock timestamp. The filter first
7486 passes on @option{preroll} amount of frames, then it buffers at most
7487 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7488 it forwards the buffered frames and also any subsequent frames coming in its
7491 The filter can be used synchronize the output of multiple ffmpeg processes for
7492 realtime output devices like decklink. By putting the delay in the filtering
7493 chain and pre-buffering frames the process can pass on data to output almost
7494 immediately after the target wallclock timestamp is reached.
7496 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7502 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7505 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7508 The maximum duration of content to buffer before waiting for the cue expressed
7509 in seconds. Default is 0.
7516 Apply color adjustments using curves.
7518 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7519 component (red, green and blue) has its values defined by @var{N} key points
7520 tied from each other using a smooth curve. The x-axis represents the pixel
7521 values from the input frame, and the y-axis the new pixel values to be set for
7524 By default, a component curve is defined by the two points @var{(0;0)} and
7525 @var{(1;1)}. This creates a straight line where each original pixel value is
7526 "adjusted" to its own value, which means no change to the image.
7528 The filter allows you to redefine these two points and add some more. A new
7529 curve (using a natural cubic spline interpolation) will be define to pass
7530 smoothly through all these new coordinates. The new defined points needs to be
7531 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7532 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7533 the vector spaces, the values will be clipped accordingly.
7535 The filter accepts the following options:
7539 Select one of the available color presets. This option can be used in addition
7540 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7541 options takes priority on the preset values.
7542 Available presets are:
7545 @item color_negative
7548 @item increase_contrast
7550 @item linear_contrast
7551 @item medium_contrast
7553 @item strong_contrast
7556 Default is @code{none}.
7558 Set the master key points. These points will define a second pass mapping. It
7559 is sometimes called a "luminance" or "value" mapping. It can be used with
7560 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7561 post-processing LUT.
7563 Set the key points for the red component.
7565 Set the key points for the green component.
7567 Set the key points for the blue component.
7569 Set the key points for all components (not including master).
7570 Can be used in addition to the other key points component
7571 options. In this case, the unset component(s) will fallback on this
7572 @option{all} setting.
7574 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7576 Save Gnuplot script of the curves in specified file.
7579 To avoid some filtergraph syntax conflicts, each key points list need to be
7580 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7582 @subsection Examples
7586 Increase slightly the middle level of blue:
7588 curves=blue='0/0 0.5/0.58 1/1'
7594 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'
7596 Here we obtain the following coordinates for each components:
7599 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7601 @code{(0;0) (0.50;0.48) (1;1)}
7603 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7607 The previous example can also be achieved with the associated built-in preset:
7609 curves=preset=vintage
7619 Use a Photoshop preset and redefine the points of the green component:
7621 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7625 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7626 and @command{gnuplot}:
7628 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7629 gnuplot -p /tmp/curves.plt
7635 Video data analysis filter.
7637 This filter shows hexadecimal pixel values of part of video.
7639 The filter accepts the following options:
7643 Set output video size.
7646 Set x offset from where to pick pixels.
7649 Set y offset from where to pick pixels.
7652 Set scope mode, can be one of the following:
7655 Draw hexadecimal pixel values with white color on black background.
7658 Draw hexadecimal pixel values with input video pixel color on black
7662 Draw hexadecimal pixel values on color background picked from input video,
7663 the text color is picked in such way so its always visible.
7667 Draw rows and columns numbers on left and top of video.
7670 Set background opacity.
7675 Denoise frames using 2D DCT (frequency domain filtering).
7677 This filter is not designed for real time.
7679 The filter accepts the following options:
7683 Set the noise sigma constant.
7685 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7686 coefficient (absolute value) below this threshold with be dropped.
7688 If you need a more advanced filtering, see @option{expr}.
7690 Default is @code{0}.
7693 Set number overlapping pixels for each block. Since the filter can be slow, you
7694 may want to reduce this value, at the cost of a less effective filter and the
7695 risk of various artefacts.
7697 If the overlapping value doesn't permit processing the whole input width or
7698 height, a warning will be displayed and according borders won't be denoised.
7700 Default value is @var{blocksize}-1, which is the best possible setting.
7703 Set the coefficient factor expression.
7705 For each coefficient of a DCT block, this expression will be evaluated as a
7706 multiplier value for the coefficient.
7708 If this is option is set, the @option{sigma} option will be ignored.
7710 The absolute value of the coefficient can be accessed through the @var{c}
7714 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7715 @var{blocksize}, which is the width and height of the processed blocks.
7717 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7718 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7719 on the speed processing. Also, a larger block size does not necessarily means a
7723 @subsection Examples
7725 Apply a denoise with a @option{sigma} of @code{4.5}:
7730 The same operation can be achieved using the expression system:
7732 dctdnoiz=e='gte(c, 4.5*3)'
7735 Violent denoise using a block size of @code{16x16}:
7742 Remove banding artifacts from input video.
7743 It works by replacing banded pixels with average value of referenced pixels.
7745 The filter accepts the following options:
7752 Set banding detection threshold for each plane. Default is 0.02.
7753 Valid range is 0.00003 to 0.5.
7754 If difference between current pixel and reference pixel is less than threshold,
7755 it will be considered as banded.
7758 Banding detection range in pixels. Default is 16. If positive, random number
7759 in range 0 to set value will be used. If negative, exact absolute value
7761 The range defines square of four pixels around current pixel.
7764 Set direction in radians from which four pixel will be compared. If positive,
7765 random direction from 0 to set direction will be picked. If negative, exact of
7766 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7767 will pick only pixels on same row and -PI/2 will pick only pixels on same
7771 If enabled, current pixel is compared with average value of all four
7772 surrounding pixels. The default is enabled. If disabled current pixel is
7773 compared with all four surrounding pixels. The pixel is considered banded
7774 if only all four differences with surrounding pixels are less than threshold.
7777 If enabled, current pixel is changed if and only if all pixel components are banded,
7778 e.g. banding detection threshold is triggered for all color components.
7779 The default is disabled.
7784 Remove blocking artifacts from input video.
7786 The filter accepts the following options:
7790 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7791 This controls what kind of deblocking is applied.
7794 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7800 Set blocking detection thresholds. Allowed range is 0 to 1.
7801 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7802 Using higher threshold gives more deblocking strength.
7803 Setting @var{alpha} controls threshold detection at exact edge of block.
7804 Remaining options controls threshold detection near the edge. Each one for
7805 below/above or left/right. Setting any of those to @var{0} disables
7809 Set planes to filter. Default is to filter all available planes.
7812 @subsection Examples
7816 Deblock using weak filter and block size of 4 pixels.
7818 deblock=filter=weak:block=4
7822 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7823 deblocking more edges.
7825 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7829 Similar as above, but filter only first plane.
7831 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7835 Similar as above, but filter only second and third plane.
7837 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7844 Drop duplicated frames at regular intervals.
7846 The filter accepts the following options:
7850 Set the number of frames from which one will be dropped. Setting this to
7851 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7852 Default is @code{5}.
7855 Set the threshold for duplicate detection. If the difference metric for a frame
7856 is less than or equal to this value, then it is declared as duplicate. Default
7860 Set scene change threshold. Default is @code{15}.
7864 Set the size of the x and y-axis blocks used during metric calculations.
7865 Larger blocks give better noise suppression, but also give worse detection of
7866 small movements. Must be a power of two. Default is @code{32}.
7869 Mark main input as a pre-processed input and activate clean source input
7870 stream. This allows the input to be pre-processed with various filters to help
7871 the metrics calculation while keeping the frame selection lossless. When set to
7872 @code{1}, the first stream is for the pre-processed input, and the second
7873 stream is the clean source from where the kept frames are chosen. Default is
7877 Set whether or not chroma is considered in the metric calculations. Default is
7883 Apply 2D deconvolution of video stream in frequency domain using second stream
7886 The filter accepts the following options:
7890 Set which planes to process.
7893 Set which impulse video frames will be processed, can be @var{first}
7894 or @var{all}. Default is @var{all}.
7897 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7898 and height are not same and not power of 2 or if stream prior to convolving
7902 The @code{deconvolve} filter also supports the @ref{framesync} options.
7906 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
7908 It accepts the following options:
7912 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
7913 @var{rainbows} for cross-color reduction.
7916 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
7919 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
7922 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
7925 Set temporal chroma threshold. Lower values increases reduction of cross-color.
7930 Apply deflate effect to the video.
7932 This filter replaces the pixel by the local(3x3) average by taking into account
7933 only values lower than the pixel.
7935 It accepts the following options:
7942 Limit the maximum change for each plane, default is 65535.
7943 If 0, plane will remain unchanged.
7948 Remove temporal frame luminance variations.
7950 It accepts the following options:
7954 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7957 Set averaging mode to smooth temporal luminance variations.
7959 Available values are:
7984 Do not actually modify frame. Useful when one only wants metadata.
7989 Remove judder produced by partially interlaced telecined content.
7991 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7992 source was partially telecined content then the output of @code{pullup,dejudder}
7993 will have a variable frame rate. May change the recorded frame rate of the
7994 container. Aside from that change, this filter will not affect constant frame
7997 The option available in this filter is:
8001 Specify the length of the window over which the judder repeats.
8003 Accepts any integer greater than 1. Useful values are:
8007 If the original was telecined from 24 to 30 fps (Film to NTSC).
8010 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8013 If a mixture of the two.
8016 The default is @samp{4}.
8021 Suppress a TV station logo by a simple interpolation of the surrounding
8022 pixels. Just set a rectangle covering the logo and watch it disappear
8023 (and sometimes something even uglier appear - your mileage may vary).
8025 It accepts the following parameters:
8030 Specify the top left corner coordinates of the logo. They must be
8035 Specify the width and height of the logo to clear. They must be
8039 Specify the thickness of the fuzzy edge of the rectangle (added to
8040 @var{w} and @var{h}). The default value is 1. This option is
8041 deprecated, setting higher values should no longer be necessary and
8045 When set to 1, a green rectangle is drawn on the screen to simplify
8046 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8047 The default value is 0.
8049 The rectangle is drawn on the outermost pixels which will be (partly)
8050 replaced with interpolated values. The values of the next pixels
8051 immediately outside this rectangle in each direction will be used to
8052 compute the interpolated pixel values inside the rectangle.
8056 @subsection Examples
8060 Set a rectangle covering the area with top left corner coordinates 0,0
8061 and size 100x77, and a band of size 10:
8063 delogo=x=0:y=0:w=100:h=77:band=10
8070 Attempt to fix small changes in horizontal and/or vertical shift. This
8071 filter helps remove camera shake from hand-holding a camera, bumping a
8072 tripod, moving on a vehicle, etc.
8074 The filter accepts the following options:
8082 Specify a rectangular area where to limit the search for motion
8084 If desired the search for motion vectors can be limited to a
8085 rectangular area of the frame defined by its top left corner, width
8086 and height. These parameters have the same meaning as the drawbox
8087 filter which can be used to visualise the position of the bounding
8090 This is useful when simultaneous movement of subjects within the frame
8091 might be confused for camera motion by the motion vector search.
8093 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8094 then the full frame is used. This allows later options to be set
8095 without specifying the bounding box for the motion vector search.
8097 Default - search the whole frame.
8101 Specify the maximum extent of movement in x and y directions in the
8102 range 0-64 pixels. Default 16.
8105 Specify how to generate pixels to fill blanks at the edge of the
8106 frame. Available values are:
8109 Fill zeroes at blank locations
8111 Original image at blank locations
8113 Extruded edge value at blank locations
8115 Mirrored edge at blank locations
8117 Default value is @samp{mirror}.
8120 Specify the blocksize to use for motion search. Range 4-128 pixels,
8124 Specify the contrast threshold for blocks. Only blocks with more than
8125 the specified contrast (difference between darkest and lightest
8126 pixels) will be considered. Range 1-255, default 125.
8129 Specify the search strategy. Available values are:
8132 Set exhaustive search
8134 Set less exhaustive search.
8136 Default value is @samp{exhaustive}.
8139 If set then a detailed log of the motion search is written to the
8146 Remove unwanted contamination of foreground colors, caused by reflected color of
8147 greenscreen or bluescreen.
8149 This filter accepts the following options:
8153 Set what type of despill to use.
8156 Set how spillmap will be generated.
8159 Set how much to get rid of still remaining spill.
8162 Controls amount of red in spill area.
8165 Controls amount of green in spill area.
8166 Should be -1 for greenscreen.
8169 Controls amount of blue in spill area.
8170 Should be -1 for bluescreen.
8173 Controls brightness of spill area, preserving colors.
8176 Modify alpha from generated spillmap.
8181 Apply an exact inverse of the telecine operation. It requires a predefined
8182 pattern specified using the pattern option which must be the same as that passed
8183 to the telecine filter.
8185 This filter accepts the following options:
8194 The default value is @code{top}.
8198 A string of numbers representing the pulldown pattern you wish to apply.
8199 The default value is @code{23}.
8202 A number representing position of the first frame with respect to the telecine
8203 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8208 Apply dilation effect to the video.
8210 This filter replaces the pixel by the local(3x3) maximum.
8212 It accepts the following options:
8219 Limit the maximum change for each plane, default is 65535.
8220 If 0, plane will remain unchanged.
8223 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8226 Flags to local 3x3 coordinates maps like this:
8235 Displace pixels as indicated by second and third input stream.
8237 It takes three input streams and outputs one stream, the first input is the
8238 source, and second and third input are displacement maps.
8240 The second input specifies how much to displace pixels along the
8241 x-axis, while the third input specifies how much to displace pixels
8243 If one of displacement map streams terminates, last frame from that
8244 displacement map will be used.
8246 Note that once generated, displacements maps can be reused over and over again.
8248 A description of the accepted options follows.
8252 Set displace behavior for pixels that are out of range.
8254 Available values are:
8257 Missing pixels are replaced by black pixels.
8260 Adjacent pixels will spread out to replace missing pixels.
8263 Out of range pixels are wrapped so they point to pixels of other side.
8266 Out of range pixels will be replaced with mirrored pixels.
8268 Default is @samp{smear}.
8272 @subsection Examples
8276 Add ripple effect to rgb input of video size hd720:
8278 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
8282 Add wave effect to rgb input of video size hd720:
8284 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
8290 Draw a colored box on the input image.
8292 It accepts the following parameters:
8297 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8301 The expressions which specify the width and height of the box; if 0 they are interpreted as
8302 the input width and height. It defaults to 0.
8305 Specify the color of the box to write. For the general syntax of this option,
8306 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8307 value @code{invert} is used, the box edge color is the same as the
8308 video with inverted luma.
8311 The expression which sets the thickness of the box edge.
8312 A value of @code{fill} will create a filled box. Default value is @code{3}.
8314 See below for the list of accepted constants.
8317 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8318 will overwrite the video's color and alpha pixels.
8319 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8322 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8323 following constants:
8327 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8331 horizontal and vertical chroma subsample values. For example for the
8332 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8336 The input width and height.
8339 The input sample aspect ratio.
8343 The x and y offset coordinates where the box is drawn.
8347 The width and height of the drawn box.
8350 The thickness of the drawn box.
8352 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8353 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8357 @subsection Examples
8361 Draw a black box around the edge of the input image:
8367 Draw a box with color red and an opacity of 50%:
8369 drawbox=10:20:200:60:red@@0.5
8372 The previous example can be specified as:
8374 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8378 Fill the box with pink color:
8380 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8384 Draw a 2-pixel red 2.40:1 mask:
8386 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
8392 Draw a grid on the input image.
8394 It accepts the following parameters:
8399 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8403 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8404 input width and height, respectively, minus @code{thickness}, so image gets
8405 framed. Default to 0.
8408 Specify the color of the grid. For the general syntax of this option,
8409 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8410 value @code{invert} is used, the grid color is the same as the
8411 video with inverted luma.
8414 The expression which sets the thickness of the grid line. Default value is @code{1}.
8416 See below for the list of accepted constants.
8419 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8420 will overwrite the video's color and alpha pixels.
8421 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8424 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8425 following constants:
8429 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8433 horizontal and vertical chroma subsample values. For example for the
8434 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8438 The input grid cell width and height.
8441 The input sample aspect ratio.
8445 The x and y coordinates of some point of grid intersection (meant to configure offset).
8449 The width and height of the drawn cell.
8452 The thickness of the drawn cell.
8454 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8455 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8459 @subsection Examples
8463 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8465 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8469 Draw a white 3x3 grid with an opacity of 50%:
8471 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8478 Draw a text string or text from a specified file on top of a video, using the
8479 libfreetype library.
8481 To enable compilation of this filter, you need to configure FFmpeg with
8482 @code{--enable-libfreetype}.
8483 To enable default font fallback and the @var{font} option you need to
8484 configure FFmpeg with @code{--enable-libfontconfig}.
8485 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8486 @code{--enable-libfribidi}.
8490 It accepts the following parameters:
8495 Used to draw a box around text using the background color.
8496 The value must be either 1 (enable) or 0 (disable).
8497 The default value of @var{box} is 0.
8500 Set the width of the border to be drawn around the box using @var{boxcolor}.
8501 The default value of @var{boxborderw} is 0.
8504 The color to be used for drawing box around text. For the syntax of this
8505 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8507 The default value of @var{boxcolor} is "white".
8510 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8511 The default value of @var{line_spacing} is 0.
8514 Set the width of the border to be drawn around the text using @var{bordercolor}.
8515 The default value of @var{borderw} is 0.
8518 Set the color to be used for drawing border around text. For the syntax of this
8519 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8521 The default value of @var{bordercolor} is "black".
8524 Select how the @var{text} is expanded. Can be either @code{none},
8525 @code{strftime} (deprecated) or
8526 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8530 Set a start time for the count. Value is in microseconds. Only applied
8531 in the deprecated strftime expansion mode. To emulate in normal expansion
8532 mode use the @code{pts} function, supplying the start time (in seconds)
8533 as the second argument.
8536 If true, check and fix text coords to avoid clipping.
8539 The color to be used for drawing fonts. For the syntax of this option, check
8540 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8542 The default value of @var{fontcolor} is "black".
8544 @item fontcolor_expr
8545 String which is expanded the same way as @var{text} to obtain dynamic
8546 @var{fontcolor} value. By default this option has empty value and is not
8547 processed. When this option is set, it overrides @var{fontcolor} option.
8550 The font family to be used for drawing text. By default Sans.
8553 The font file to be used for drawing text. The path must be included.
8554 This parameter is mandatory if the fontconfig support is disabled.
8557 Draw the text applying alpha blending. The value can
8558 be a number between 0.0 and 1.0.
8559 The expression accepts the same variables @var{x, y} as well.
8560 The default value is 1.
8561 Please see @var{fontcolor_expr}.
8564 The font size to be used for drawing text.
8565 The default value of @var{fontsize} is 16.
8568 If set to 1, attempt to shape the text (for example, reverse the order of
8569 right-to-left text and join Arabic characters) before drawing it.
8570 Otherwise, just draw the text exactly as given.
8571 By default 1 (if supported).
8574 The flags to be used for loading the fonts.
8576 The flags map the corresponding flags supported by libfreetype, and are
8577 a combination of the following values:
8584 @item vertical_layout
8585 @item force_autohint
8588 @item ignore_global_advance_width
8590 @item ignore_transform
8596 Default value is "default".
8598 For more information consult the documentation for the FT_LOAD_*
8602 The color to be used for drawing a shadow behind the drawn text. For the
8603 syntax of this option, check the @ref{color syntax,,"Color" section in the
8604 ffmpeg-utils manual,ffmpeg-utils}.
8606 The default value of @var{shadowcolor} is "black".
8610 The x and y offsets for the text shadow position with respect to the
8611 position of the text. They can be either positive or negative
8612 values. The default value for both is "0".
8615 The starting frame number for the n/frame_num variable. The default value
8619 The size in number of spaces to use for rendering the tab.
8623 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8624 format. It can be used with or without text parameter. @var{timecode_rate}
8625 option must be specified.
8627 @item timecode_rate, rate, r
8628 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8629 integer. Minimum value is "1".
8630 Drop-frame timecode is supported for frame rates 30 & 60.
8633 If set to 1, the output of the timecode option will wrap around at 24 hours.
8634 Default is 0 (disabled).
8637 The text string to be drawn. The text must be a sequence of UTF-8
8639 This parameter is mandatory if no file is specified with the parameter
8643 A text file containing text to be drawn. The text must be a sequence
8644 of UTF-8 encoded characters.
8646 This parameter is mandatory if no text string is specified with the
8647 parameter @var{text}.
8649 If both @var{text} and @var{textfile} are specified, an error is thrown.
8652 If set to 1, the @var{textfile} will be reloaded before each frame.
8653 Be sure to update it atomically, or it may be read partially, or even fail.
8657 The expressions which specify the offsets where text will be drawn
8658 within the video frame. They are relative to the top/left border of the
8661 The default value of @var{x} and @var{y} is "0".
8663 See below for the list of accepted constants and functions.
8666 The parameters for @var{x} and @var{y} are expressions containing the
8667 following constants and functions:
8671 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8675 horizontal and vertical chroma subsample values. For example for the
8676 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8679 the height of each text line
8687 @item max_glyph_a, ascent
8688 the maximum distance from the baseline to the highest/upper grid
8689 coordinate used to place a glyph outline point, for all the rendered
8691 It is a positive value, due to the grid's orientation with the Y axis
8694 @item max_glyph_d, descent
8695 the maximum distance from the baseline to the lowest grid coordinate
8696 used to place a glyph outline point, for all the rendered glyphs.
8697 This is a negative value, due to the grid's orientation, with the Y axis
8701 maximum glyph height, that is the maximum height for all the glyphs
8702 contained in the rendered text, it is equivalent to @var{ascent} -
8706 maximum glyph width, that is the maximum width for all the glyphs
8707 contained in the rendered text
8710 the number of input frame, starting from 0
8712 @item rand(min, max)
8713 return a random number included between @var{min} and @var{max}
8716 The input sample aspect ratio.
8719 timestamp expressed in seconds, NAN if the input timestamp is unknown
8722 the height of the rendered text
8725 the width of the rendered text
8729 the x and y offset coordinates where the text is drawn.
8731 These parameters allow the @var{x} and @var{y} expressions to refer
8732 each other, so you can for example specify @code{y=x/dar}.
8735 @anchor{drawtext_expansion}
8736 @subsection Text expansion
8738 If @option{expansion} is set to @code{strftime},
8739 the filter recognizes strftime() sequences in the provided text and
8740 expands them accordingly. Check the documentation of strftime(). This
8741 feature is deprecated.
8743 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8745 If @option{expansion} is set to @code{normal} (which is the default),
8746 the following expansion mechanism is used.
8748 The backslash character @samp{\}, followed by any character, always expands to
8749 the second character.
8751 Sequences of the form @code{%@{...@}} are expanded. The text between the
8752 braces is a function name, possibly followed by arguments separated by ':'.
8753 If the arguments contain special characters or delimiters (':' or '@}'),
8754 they should be escaped.
8756 Note that they probably must also be escaped as the value for the
8757 @option{text} option in the filter argument string and as the filter
8758 argument in the filtergraph description, and possibly also for the shell,
8759 that makes up to four levels of escaping; using a text file avoids these
8762 The following functions are available:
8767 The expression evaluation result.
8769 It must take one argument specifying the expression to be evaluated,
8770 which accepts the same constants and functions as the @var{x} and
8771 @var{y} values. Note that not all constants should be used, for
8772 example the text size is not known when evaluating the expression, so
8773 the constants @var{text_w} and @var{text_h} will have an undefined
8776 @item expr_int_format, eif
8777 Evaluate the expression's value and output as formatted integer.
8779 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8780 The second argument specifies the output format. Allowed values are @samp{x},
8781 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8782 @code{printf} function.
8783 The third parameter is optional and sets the number of positions taken by the output.
8784 It can be used to add padding with zeros from the left.
8787 The time at which the filter is running, expressed in UTC.
8788 It can accept an argument: a strftime() format string.
8791 The time at which the filter is running, expressed in the local time zone.
8792 It can accept an argument: a strftime() format string.
8795 Frame metadata. Takes one or two arguments.
8797 The first argument is mandatory and specifies the metadata key.
8799 The second argument is optional and specifies a default value, used when the
8800 metadata key is not found or empty.
8803 The frame number, starting from 0.
8806 A 1 character description of the current picture type.
8809 The timestamp of the current frame.
8810 It can take up to three arguments.
8812 The first argument is the format of the timestamp; it defaults to @code{flt}
8813 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8814 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8815 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8816 @code{localtime} stands for the timestamp of the frame formatted as
8817 local time zone time.
8819 The second argument is an offset added to the timestamp.
8821 If the format is set to @code{hms}, a third argument @code{24HH} may be
8822 supplied to present the hour part of the formatted timestamp in 24h format
8825 If the format is set to @code{localtime} or @code{gmtime},
8826 a third argument may be supplied: a strftime() format string.
8827 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8830 @subsection Examples
8834 Draw "Test Text" with font FreeSerif, using the default values for the
8835 optional parameters.
8838 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8842 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8843 and y=50 (counting from the top-left corner of the screen), text is
8844 yellow with a red box around it. Both the text and the box have an
8848 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8849 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8852 Note that the double quotes are not necessary if spaces are not used
8853 within the parameter list.
8856 Show the text at the center of the video frame:
8858 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8862 Show the text at a random position, switching to a new position every 30 seconds:
8864 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)"
8868 Show a text line sliding from right to left in the last row of the video
8869 frame. The file @file{LONG_LINE} is assumed to contain a single line
8872 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8876 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8878 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8882 Draw a single green letter "g", at the center of the input video.
8883 The glyph baseline is placed at half screen height.
8885 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8889 Show text for 1 second every 3 seconds:
8891 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8895 Use fontconfig to set the font. Note that the colons need to be escaped.
8897 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8901 Print the date of a real-time encoding (see strftime(3)):
8903 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8907 Show text fading in and out (appearing/disappearing):
8910 DS=1.0 # display start
8911 DE=10.0 # display end
8912 FID=1.5 # fade in duration
8913 FOD=5 # fade out duration
8914 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 @}"
8918 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8919 and the @option{fontsize} value are included in the @option{y} offset.
8921 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8922 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8927 For more information about libfreetype, check:
8928 @url{http://www.freetype.org/}.
8930 For more information about fontconfig, check:
8931 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8933 For more information about libfribidi, check:
8934 @url{http://fribidi.org/}.
8938 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8940 The filter accepts the following options:
8945 Set low and high threshold values used by the Canny thresholding
8948 The high threshold selects the "strong" edge pixels, which are then
8949 connected through 8-connectivity with the "weak" edge pixels selected
8950 by the low threshold.
8952 @var{low} and @var{high} threshold values must be chosen in the range
8953 [0,1], and @var{low} should be lesser or equal to @var{high}.
8955 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8959 Define the drawing mode.
8963 Draw white/gray wires on black background.
8966 Mix the colors to create a paint/cartoon effect.
8969 Apply Canny edge detector on all selected planes.
8971 Default value is @var{wires}.
8974 Select planes for filtering. By default all available planes are filtered.
8977 @subsection Examples
8981 Standard edge detection with custom values for the hysteresis thresholding:
8983 edgedetect=low=0.1:high=0.4
8987 Painting effect without thresholding:
8989 edgedetect=mode=colormix:high=0
8994 Set brightness, contrast, saturation and approximate gamma adjustment.
8996 The filter accepts the following options:
9000 Set the contrast expression. The value must be a float value in range
9001 @code{-2.0} to @code{2.0}. The default value is "1".
9004 Set the brightness expression. The value must be a float value in
9005 range @code{-1.0} to @code{1.0}. The default value is "0".
9008 Set the saturation expression. The value must be a float in
9009 range @code{0.0} to @code{3.0}. The default value is "1".
9012 Set the gamma expression. The value must be a float in range
9013 @code{0.1} to @code{10.0}. The default value is "1".
9016 Set the gamma expression for red. The value must be a float in
9017 range @code{0.1} to @code{10.0}. The default value is "1".
9020 Set the gamma expression for green. The value must be a float in range
9021 @code{0.1} to @code{10.0}. The default value is "1".
9024 Set the gamma expression for blue. The value must be a float in range
9025 @code{0.1} to @code{10.0}. The default value is "1".
9028 Set the gamma weight expression. It can be used to reduce the effect
9029 of a high gamma value on bright image areas, e.g. keep them from
9030 getting overamplified and just plain white. The value must be a float
9031 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
9032 gamma correction all the way down while @code{1.0} leaves it at its
9033 full strength. Default is "1".
9036 Set when the expressions for brightness, contrast, saturation and
9037 gamma expressions are evaluated.
9039 It accepts the following values:
9042 only evaluate expressions once during the filter initialization or
9043 when a command is processed
9046 evaluate expressions for each incoming frame
9049 Default value is @samp{init}.
9052 The expressions accept the following parameters:
9055 frame count of the input frame starting from 0
9058 byte position of the corresponding packet in the input file, NAN if
9062 frame rate of the input video, NAN if the input frame rate is unknown
9065 timestamp expressed in seconds, NAN if the input timestamp is unknown
9068 @subsection Commands
9069 The filter supports the following commands:
9073 Set the contrast expression.
9076 Set the brightness expression.
9079 Set the saturation expression.
9082 Set the gamma expression.
9085 Set the gamma_r expression.
9088 Set gamma_g expression.
9091 Set gamma_b expression.
9094 Set gamma_weight expression.
9096 The command accepts the same syntax of the corresponding option.
9098 If the specified expression is not valid, it is kept at its current
9105 Apply erosion effect to the video.
9107 This filter replaces the pixel by the local(3x3) minimum.
9109 It accepts the following options:
9116 Limit the maximum change for each plane, default is 65535.
9117 If 0, plane will remain unchanged.
9120 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9123 Flags to local 3x3 coordinates maps like this:
9130 @section extractplanes
9132 Extract color channel components from input video stream into
9133 separate grayscale video streams.
9135 The filter accepts the following option:
9139 Set plane(s) to extract.
9141 Available values for planes are:
9152 Choosing planes not available in the input will result in an error.
9153 That means you cannot select @code{r}, @code{g}, @code{b} planes
9154 with @code{y}, @code{u}, @code{v} planes at same time.
9157 @subsection Examples
9161 Extract luma, u and v color channel component from input video frame
9162 into 3 grayscale outputs:
9164 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
9170 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9172 For each input image, the filter will compute the optimal mapping from
9173 the input to the output given the codebook length, that is the number
9174 of distinct output colors.
9176 This filter accepts the following options.
9179 @item codebook_length, l
9180 Set codebook length. The value must be a positive integer, and
9181 represents the number of distinct output colors. Default value is 256.
9184 Set the maximum number of iterations to apply for computing the optimal
9185 mapping. The higher the value the better the result and the higher the
9186 computation time. Default value is 1.
9189 Set a random seed, must be an integer included between 0 and
9190 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9191 will try to use a good random seed on a best effort basis.
9194 Set pal8 output pixel format. This option does not work with codebook
9195 length greater than 256.
9200 Measure graylevel entropy in histogram of color channels of video frames.
9202 It accepts the following parameters:
9206 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9208 @var{diff} mode measures entropy of histogram delta values, absolute differences
9209 between neighbour histogram values.
9214 Apply a fade-in/out effect to the input video.
9216 It accepts the following parameters:
9220 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9222 Default is @code{in}.
9224 @item start_frame, s
9225 Specify the number of the frame to start applying the fade
9226 effect at. Default is 0.
9229 The number of frames that the fade effect lasts. At the end of the
9230 fade-in effect, the output video will have the same intensity as the input video.
9231 At the end of the fade-out transition, the output video will be filled with the
9232 selected @option{color}.
9236 If set to 1, fade only alpha channel, if one exists on the input.
9239 @item start_time, st
9240 Specify the timestamp (in seconds) of the frame to start to apply the fade
9241 effect. If both start_frame and start_time are specified, the fade will start at
9242 whichever comes last. Default is 0.
9245 The number of seconds for which the fade effect has to last. At the end of the
9246 fade-in effect the output video will have the same intensity as the input video,
9247 at the end of the fade-out transition the output video will be filled with the
9248 selected @option{color}.
9249 If both duration and nb_frames are specified, duration is used. Default is 0
9250 (nb_frames is used by default).
9253 Specify the color of the fade. Default is "black".
9256 @subsection Examples
9260 Fade in the first 30 frames of video:
9265 The command above is equivalent to:
9271 Fade out the last 45 frames of a 200-frame video:
9274 fade=type=out:start_frame=155:nb_frames=45
9278 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9280 fade=in:0:25, fade=out:975:25
9284 Make the first 5 frames yellow, then fade in from frame 5-24:
9286 fade=in:5:20:color=yellow
9290 Fade in alpha over first 25 frames of video:
9292 fade=in:0:25:alpha=1
9296 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9298 fade=t=in:st=5.5:d=0.5
9304 Apply arbitrary expressions to samples in frequency domain
9308 Adjust the dc value (gain) of the luma plane of the image. The filter
9309 accepts an integer value in range @code{0} to @code{1000}. The default
9310 value is set to @code{0}.
9313 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9314 filter accepts an integer value in range @code{0} to @code{1000}. The
9315 default value is set to @code{0}.
9318 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9319 filter accepts an integer value in range @code{0} to @code{1000}. The
9320 default value is set to @code{0}.
9323 Set the frequency domain weight expression for the luma plane.
9326 Set the frequency domain weight expression for the 1st chroma plane.
9329 Set the frequency domain weight expression for the 2nd chroma plane.
9332 Set when the expressions are evaluated.
9334 It accepts the following values:
9337 Only evaluate expressions once during the filter initialization.
9340 Evaluate expressions for each incoming frame.
9343 Default value is @samp{init}.
9345 The filter accepts the following variables:
9348 The coordinates of the current sample.
9352 The width and height of the image.
9355 The number of input frame, starting from 0.
9358 @subsection Examples
9364 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9370 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9376 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9382 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9388 Denoise frames using 3D FFT (frequency domain filtering).
9390 The filter accepts the following options:
9394 Set the noise sigma constant. This sets denoising strength.
9395 Default value is 1. Allowed range is from 0 to 30.
9396 Using very high sigma with low overlap may give blocking artifacts.
9399 Set amount of denoising. By default all detected noise is reduced.
9400 Default value is 1. Allowed range is from 0 to 1.
9403 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9404 Actual size of block in pixels is 2 to power of @var{block}, so by default
9405 block size in pixels is 2^4 which is 16.
9408 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9411 Set number of previous frames to use for denoising. By default is set to 0.
9414 Set number of next frames to to use for denoising. By default is set to 0.
9417 Set planes which will be filtered, by default are all available filtered
9423 Extract a single field from an interlaced image using stride
9424 arithmetic to avoid wasting CPU time. The output frames are marked as
9427 The filter accepts the following options:
9431 Specify whether to extract the top (if the value is @code{0} or
9432 @code{top}) or the bottom field (if the value is @code{1} or
9438 Create new frames by copying the top and bottom fields from surrounding frames
9439 supplied as numbers by the hint file.
9443 Set file containing hints: absolute/relative frame numbers.
9445 There must be one line for each frame in a clip. Each line must contain two
9446 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9447 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9448 is current frame number for @code{absolute} mode or out of [-1, 1] range
9449 for @code{relative} mode. First number tells from which frame to pick up top
9450 field and second number tells from which frame to pick up bottom field.
9452 If optionally followed by @code{+} output frame will be marked as interlaced,
9453 else if followed by @code{-} output frame will be marked as progressive, else
9454 it will be marked same as input frame.
9455 If line starts with @code{#} or @code{;} that line is skipped.
9458 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9461 Example of first several lines of @code{hint} file for @code{relative} mode:
9464 1,0 - # second frame, use third's frame top field and second's frame bottom field
9465 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9482 Field matching filter for inverse telecine. It is meant to reconstruct the
9483 progressive frames from a telecined stream. The filter does not drop duplicated
9484 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9485 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9487 The separation of the field matching and the decimation is notably motivated by
9488 the possibility of inserting a de-interlacing filter fallback between the two.
9489 If the source has mixed telecined and real interlaced content,
9490 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9491 But these remaining combed frames will be marked as interlaced, and thus can be
9492 de-interlaced by a later filter such as @ref{yadif} before decimation.
9494 In addition to the various configuration options, @code{fieldmatch} can take an
9495 optional second stream, activated through the @option{ppsrc} option. If
9496 enabled, the frames reconstruction will be based on the fields and frames from
9497 this second stream. This allows the first input to be pre-processed in order to
9498 help the various algorithms of the filter, while keeping the output lossless
9499 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9500 or brightness/contrast adjustments can help.
9502 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9503 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9504 which @code{fieldmatch} is based on. While the semantic and usage are very
9505 close, some behaviour and options names can differ.
9507 The @ref{decimate} filter currently only works for constant frame rate input.
9508 If your input has mixed telecined (30fps) and progressive content with a lower
9509 framerate like 24fps use the following filterchain to produce the necessary cfr
9510 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9512 The filter accepts the following options:
9516 Specify the assumed field order of the input stream. Available values are:
9520 Auto detect parity (use FFmpeg's internal parity value).
9522 Assume bottom field first.
9524 Assume top field first.
9527 Note that it is sometimes recommended not to trust the parity announced by the
9530 Default value is @var{auto}.
9533 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9534 sense that it won't risk creating jerkiness due to duplicate frames when
9535 possible, but if there are bad edits or blended fields it will end up
9536 outputting combed frames when a good match might actually exist. On the other
9537 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9538 but will almost always find a good frame if there is one. The other values are
9539 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9540 jerkiness and creating duplicate frames versus finding good matches in sections
9541 with bad edits, orphaned fields, blended fields, etc.
9543 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9545 Available values are:
9549 2-way matching (p/c)
9551 2-way matching, and trying 3rd match if still combed (p/c + n)
9553 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9555 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9556 still combed (p/c + n + u/b)
9558 3-way matching (p/c/n)
9560 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9561 detected as combed (p/c/n + u/b)
9564 The parenthesis at the end indicate the matches that would be used for that
9565 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9568 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9571 Default value is @var{pc_n}.
9574 Mark the main input stream as a pre-processed input, and enable the secondary
9575 input stream as the clean source to pick the fields from. See the filter
9576 introduction for more details. It is similar to the @option{clip2} feature from
9579 Default value is @code{0} (disabled).
9582 Set the field to match from. It is recommended to set this to the same value as
9583 @option{order} unless you experience matching failures with that setting. In
9584 certain circumstances changing the field that is used to match from can have a
9585 large impact on matching performance. Available values are:
9589 Automatic (same value as @option{order}).
9591 Match from the bottom field.
9593 Match from the top field.
9596 Default value is @var{auto}.
9599 Set whether or not chroma is included during the match comparisons. In most
9600 cases it is recommended to leave this enabled. You should set this to @code{0}
9601 only if your clip has bad chroma problems such as heavy rainbowing or other
9602 artifacts. Setting this to @code{0} could also be used to speed things up at
9603 the cost of some accuracy.
9605 Default value is @code{1}.
9609 These define an exclusion band which excludes the lines between @option{y0} and
9610 @option{y1} from being included in the field matching decision. An exclusion
9611 band can be used to ignore subtitles, a logo, or other things that may
9612 interfere with the matching. @option{y0} sets the starting scan line and
9613 @option{y1} sets the ending line; all lines in between @option{y0} and
9614 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9615 @option{y0} and @option{y1} to the same value will disable the feature.
9616 @option{y0} and @option{y1} defaults to @code{0}.
9619 Set the scene change detection threshold as a percentage of maximum change on
9620 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9621 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9622 @option{scthresh} is @code{[0.0, 100.0]}.
9624 Default value is @code{12.0}.
9627 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9628 account the combed scores of matches when deciding what match to use as the
9629 final match. Available values are:
9633 No final matching based on combed scores.
9635 Combed scores are only used when a scene change is detected.
9637 Use combed scores all the time.
9640 Default is @var{sc}.
9643 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9644 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9645 Available values are:
9649 No forced calculation.
9651 Force p/c/n calculations.
9653 Force p/c/n/u/b calculations.
9656 Default value is @var{none}.
9659 This is the area combing threshold used for combed frame detection. This
9660 essentially controls how "strong" or "visible" combing must be to be detected.
9661 Larger values mean combing must be more visible and smaller values mean combing
9662 can be less visible or strong and still be detected. Valid settings are from
9663 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9664 be detected as combed). This is basically a pixel difference value. A good
9665 range is @code{[8, 12]}.
9667 Default value is @code{9}.
9670 Sets whether or not chroma is considered in the combed frame decision. Only
9671 disable this if your source has chroma problems (rainbowing, etc.) that are
9672 causing problems for the combed frame detection with chroma enabled. Actually,
9673 using @option{chroma}=@var{0} is usually more reliable, except for the case
9674 where there is chroma only combing in the source.
9676 Default value is @code{0}.
9680 Respectively set the x-axis and y-axis size of the window used during combed
9681 frame detection. This has to do with the size of the area in which
9682 @option{combpel} pixels are required to be detected as combed for a frame to be
9683 declared combed. See the @option{combpel} parameter description for more info.
9684 Possible values are any number that is a power of 2 starting at 4 and going up
9687 Default value is @code{16}.
9690 The number of combed pixels inside any of the @option{blocky} by
9691 @option{blockx} size blocks on the frame for the frame to be detected as
9692 combed. While @option{cthresh} controls how "visible" the combing must be, this
9693 setting controls "how much" combing there must be in any localized area (a
9694 window defined by the @option{blockx} and @option{blocky} settings) on the
9695 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9696 which point no frames will ever be detected as combed). This setting is known
9697 as @option{MI} in TFM/VFM vocabulary.
9699 Default value is @code{80}.
9702 @anchor{p/c/n/u/b meaning}
9703 @subsection p/c/n/u/b meaning
9705 @subsubsection p/c/n
9707 We assume the following telecined stream:
9710 Top fields: 1 2 2 3 4
9711 Bottom fields: 1 2 3 4 4
9714 The numbers correspond to the progressive frame the fields relate to. Here, the
9715 first two frames are progressive, the 3rd and 4th are combed, and so on.
9717 When @code{fieldmatch} is configured to run a matching from bottom
9718 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9723 B 1 2 3 4 4 <-- matching reference
9732 As a result of the field matching, we can see that some frames get duplicated.
9733 To perform a complete inverse telecine, you need to rely on a decimation filter
9734 after this operation. See for instance the @ref{decimate} filter.
9736 The same operation now matching from top fields (@option{field}=@var{top})
9741 T 1 2 2 3 4 <-- matching reference
9751 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9752 basically, they refer to the frame and field of the opposite parity:
9755 @item @var{p} matches the field of the opposite parity in the previous frame
9756 @item @var{c} matches the field of the opposite parity in the current frame
9757 @item @var{n} matches the field of the opposite parity in the next frame
9762 The @var{u} and @var{b} matching are a bit special in the sense that they match
9763 from the opposite parity flag. In the following examples, we assume that we are
9764 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9765 'x' is placed above and below each matched fields.
9767 With bottom matching (@option{field}=@var{bottom}):
9772 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9773 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9781 With top matching (@option{field}=@var{top}):
9786 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9787 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9795 @subsection Examples
9797 Simple IVTC of a top field first telecined stream:
9799 fieldmatch=order=tff:combmatch=none, decimate
9802 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9804 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9809 Transform the field order of the input video.
9811 It accepts the following parameters:
9816 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9817 for bottom field first.
9820 The default value is @samp{tff}.
9822 The transformation is done by shifting the picture content up or down
9823 by one line, and filling the remaining line with appropriate picture content.
9824 This method is consistent with most broadcast field order converters.
9826 If the input video is not flagged as being interlaced, or it is already
9827 flagged as being of the required output field order, then this filter does
9828 not alter the incoming video.
9830 It is very useful when converting to or from PAL DV material,
9831 which is bottom field first.
9835 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9838 @section fifo, afifo
9840 Buffer input images and send them when they are requested.
9842 It is mainly useful when auto-inserted by the libavfilter
9845 It does not take parameters.
9847 @section fillborders
9849 Fill borders of the input video, without changing video stream dimensions.
9850 Sometimes video can have garbage at the four edges and you may not want to
9851 crop video input to keep size multiple of some number.
9853 This filter accepts the following options:
9857 Number of pixels to fill from left border.
9860 Number of pixels to fill from right border.
9863 Number of pixels to fill from top border.
9866 Number of pixels to fill from bottom border.
9871 It accepts the following values:
9874 fill pixels using outermost pixels
9877 fill pixels using mirroring
9880 fill pixels with constant value
9883 Default is @var{smear}.
9886 Set color for pixels in fixed mode. Default is @var{black}.
9891 Find a rectangular object
9893 It accepts the following options:
9897 Filepath of the object image, needs to be in gray8.
9900 Detection threshold, default is 0.5.
9903 Number of mipmaps, default is 3.
9905 @item xmin, ymin, xmax, ymax
9906 Specifies the rectangle in which to search.
9909 @subsection Examples
9913 Generate a representative palette of a given video using @command{ffmpeg}:
9915 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9921 Cover a rectangular object
9923 It accepts the following options:
9927 Filepath of the optional cover image, needs to be in yuv420.
9932 It accepts the following values:
9935 cover it by the supplied image
9937 cover it by interpolating the surrounding pixels
9940 Default value is @var{blur}.
9943 @subsection Examples
9947 Generate a representative palette of a given video using @command{ffmpeg}:
9949 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9955 Flood area with values of same pixel components with another values.
9957 It accepts the following options:
9960 Set pixel x coordinate.
9963 Set pixel y coordinate.
9966 Set source #0 component value.
9969 Set source #1 component value.
9972 Set source #2 component value.
9975 Set source #3 component value.
9978 Set destination #0 component value.
9981 Set destination #1 component value.
9984 Set destination #2 component value.
9987 Set destination #3 component value.
9993 Convert the input video to one of the specified pixel formats.
9994 Libavfilter will try to pick one that is suitable as input to
9997 It accepts the following parameters:
10001 A '|'-separated list of pixel format names, such as
10002 "pix_fmts=yuv420p|monow|rgb24".
10006 @subsection Examples
10010 Convert the input video to the @var{yuv420p} format
10012 format=pix_fmts=yuv420p
10015 Convert the input video to any of the formats in the list
10017 format=pix_fmts=yuv420p|yuv444p|yuv410p
10024 Convert the video to specified constant frame rate by duplicating or dropping
10025 frames as necessary.
10027 It accepts the following parameters:
10031 The desired output frame rate. The default is @code{25}.
10034 Assume the first PTS should be the given value, in seconds. This allows for
10035 padding/trimming at the start of stream. By default, no assumption is made
10036 about the first frame's expected PTS, so no padding or trimming is done.
10037 For example, this could be set to 0 to pad the beginning with duplicates of
10038 the first frame if a video stream starts after the audio stream or to trim any
10039 frames with a negative PTS.
10042 Timestamp (PTS) rounding method.
10044 Possible values are:
10051 round towards -infinity
10053 round towards +infinity
10057 The default is @code{near}.
10060 Action performed when reading the last frame.
10062 Possible values are:
10065 Use same timestamp rounding method as used for other frames.
10067 Pass through last frame if input duration has not been reached yet.
10069 The default is @code{round}.
10073 Alternatively, the options can be specified as a flat string:
10074 @var{fps}[:@var{start_time}[:@var{round}]].
10076 See also the @ref{setpts} filter.
10078 @subsection Examples
10082 A typical usage in order to set the fps to 25:
10088 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10090 fps=fps=film:round=near
10096 Pack two different video streams into a stereoscopic video, setting proper
10097 metadata on supported codecs. The two views should have the same size and
10098 framerate and processing will stop when the shorter video ends. Please note
10099 that you may conveniently adjust view properties with the @ref{scale} and
10102 It accepts the following parameters:
10106 The desired packing format. Supported values are:
10111 The views are next to each other (default).
10114 The views are on top of each other.
10117 The views are packed by line.
10120 The views are packed by column.
10123 The views are temporally interleaved.
10132 # Convert left and right views into a frame-sequential video
10133 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10135 # Convert views into a side-by-side video with the same output resolution as the input
10136 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
10141 Change the frame rate by interpolating new video output frames from the source
10144 This filter is not designed to function correctly with interlaced media. If
10145 you wish to change the frame rate of interlaced media then you are required
10146 to deinterlace before this filter and re-interlace after this filter.
10148 A description of the accepted options follows.
10152 Specify the output frames per second. This option can also be specified
10153 as a value alone. The default is @code{50}.
10156 Specify the start of a range where the output frame will be created as a
10157 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10158 the default is @code{15}.
10161 Specify the end of a range where the output frame will be created as a
10162 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10163 the default is @code{240}.
10166 Specify the level at which a scene change is detected as a value between
10167 0 and 100 to indicate a new scene; a low value reflects a low
10168 probability for the current frame to introduce a new scene, while a higher
10169 value means the current frame is more likely to be one.
10170 The default is @code{8.2}.
10173 Specify flags influencing the filter process.
10175 Available value for @var{flags} is:
10178 @item scene_change_detect, scd
10179 Enable scene change detection using the value of the option @var{scene}.
10180 This flag is enabled by default.
10186 Select one frame every N-th frame.
10188 This filter accepts the following option:
10191 Select frame after every @code{step} frames.
10192 Allowed values are positive integers higher than 0. Default value is @code{1}.
10195 @section freezedetect
10197 Detect frozen video.
10199 This filter logs a message and sets frame metadata when it detects that the
10200 input video has no significant change in content during a specified duration.
10201 Video freeze detection calculates the mean average absolute difference of all
10202 the components of video frames and compares it to a noise floor.
10204 The printed times and duration are expressed in seconds. The
10205 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10206 whose timestamp equals or exceeds the detection duration and it contains the
10207 timestamp of the first frame of the freeze. The
10208 @code{lavfi.freezedetect.freeze_duration} and
10209 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10212 The filter accepts the following options:
10216 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10217 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10221 Set freeze duration until notification (default is 2 seconds).
10227 Apply a frei0r effect to the input video.
10229 To enable the compilation of this filter, you need to install the frei0r
10230 header and configure FFmpeg with @code{--enable-frei0r}.
10232 It accepts the following parameters:
10237 The name of the frei0r effect to load. If the environment variable
10238 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10239 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10240 Otherwise, the standard frei0r paths are searched, in this order:
10241 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10242 @file{/usr/lib/frei0r-1/}.
10244 @item filter_params
10245 A '|'-separated list of parameters to pass to the frei0r effect.
10249 A frei0r effect parameter can be a boolean (its value is either
10250 "y" or "n"), a double, a color (specified as
10251 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10252 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10253 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10254 a position (specified as @var{X}/@var{Y}, where
10255 @var{X} and @var{Y} are floating point numbers) and/or a string.
10257 The number and types of parameters depend on the loaded effect. If an
10258 effect parameter is not specified, the default value is set.
10260 @subsection Examples
10264 Apply the distort0r effect, setting the first two double parameters:
10266 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10270 Apply the colordistance effect, taking a color as the first parameter:
10272 frei0r=colordistance:0.2/0.3/0.4
10273 frei0r=colordistance:violet
10274 frei0r=colordistance:0x112233
10278 Apply the perspective effect, specifying the top left and top right image
10281 frei0r=perspective:0.2/0.2|0.8/0.2
10285 For more information, see
10286 @url{http://frei0r.dyne.org}
10290 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10292 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10293 processing filter, one of them is performed once per block, not per pixel.
10294 This allows for much higher speed.
10296 The filter accepts the following options:
10300 Set quality. This option defines the number of levels for averaging. It accepts
10301 an integer in the range 4-5. Default value is @code{4}.
10304 Force a constant quantization parameter. It accepts an integer in range 0-63.
10305 If not set, the filter will use the QP from the video stream (if available).
10308 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10309 more details but also more artifacts, while higher values make the image smoother
10310 but also blurrier. Default value is @code{0} − PSNR optimal.
10312 @item use_bframe_qp
10313 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10314 option may cause flicker since the B-Frames have often larger QP. Default is
10315 @code{0} (not enabled).
10321 Apply Gaussian blur filter.
10323 The filter accepts the following options:
10327 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10330 Set number of steps for Gaussian approximation. Default is @code{1}.
10333 Set which planes to filter. By default all planes are filtered.
10336 Set vertical sigma, if negative it will be same as @code{sigma}.
10337 Default is @code{-1}.
10342 Apply generic equation to each pixel.
10344 The filter accepts the following options:
10347 @item lum_expr, lum
10348 Set the luminance expression.
10350 Set the chrominance blue expression.
10352 Set the chrominance red expression.
10353 @item alpha_expr, a
10354 Set the alpha expression.
10356 Set the red expression.
10357 @item green_expr, g
10358 Set the green expression.
10360 Set the blue expression.
10363 The colorspace is selected according to the specified options. If one
10364 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10365 options is specified, the filter will automatically select a YCbCr
10366 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10367 @option{blue_expr} options is specified, it will select an RGB
10370 If one of the chrominance expression is not defined, it falls back on the other
10371 one. If no alpha expression is specified it will evaluate to opaque value.
10372 If none of chrominance expressions are specified, they will evaluate
10373 to the luminance expression.
10375 The expressions can use the following variables and functions:
10379 The sequential number of the filtered frame, starting from @code{0}.
10383 The coordinates of the current sample.
10387 The width and height of the image.
10391 Width and height scale depending on the currently filtered plane. It is the
10392 ratio between the corresponding luma plane number of pixels and the current
10393 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10394 @code{0.5,0.5} for chroma planes.
10397 Time of the current frame, expressed in seconds.
10400 Return the value of the pixel at location (@var{x},@var{y}) of the current
10404 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10408 Return the value of the pixel at location (@var{x},@var{y}) of the
10409 blue-difference chroma plane. Return 0 if there is no such plane.
10412 Return the value of the pixel at location (@var{x},@var{y}) of the
10413 red-difference chroma plane. Return 0 if there is no such plane.
10418 Return the value of the pixel at location (@var{x},@var{y}) of the
10419 red/green/blue component. Return 0 if there is no such component.
10422 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10423 plane. Return 0 if there is no such plane.
10426 For functions, if @var{x} and @var{y} are outside the area, the value will be
10427 automatically clipped to the closer edge.
10429 @subsection Examples
10433 Flip the image horizontally:
10439 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10440 wavelength of 100 pixels:
10442 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10446 Generate a fancy enigmatic moving light:
10448 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
10452 Generate a quick emboss effect:
10454 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10458 Modify RGB components depending on pixel position:
10460 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10464 Create a radial gradient that is the same size as the input (also see
10465 the @ref{vignette} filter):
10467 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10473 Fix the banding artifacts that are sometimes introduced into nearly flat
10474 regions by truncation to 8-bit color depth.
10475 Interpolate the gradients that should go where the bands are, and
10478 It is designed for playback only. Do not use it prior to
10479 lossy compression, because compression tends to lose the dither and
10480 bring back the bands.
10482 It accepts the following parameters:
10487 The maximum amount by which the filter will change any one pixel. This is also
10488 the threshold for detecting nearly flat regions. Acceptable values range from
10489 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10493 The neighborhood to fit the gradient to. A larger radius makes for smoother
10494 gradients, but also prevents the filter from modifying the pixels near detailed
10495 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10496 values will be clipped to the valid range.
10500 Alternatively, the options can be specified as a flat string:
10501 @var{strength}[:@var{radius}]
10503 @subsection Examples
10507 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10513 Specify radius, omitting the strength (which will fall-back to the default
10521 @section graphmonitor, agraphmonitor
10522 Show various filtergraph stats.
10524 With this filter one can debug complete filtergraph.
10525 Especially issues with links filling with queued frames.
10527 The filter accepts the following options:
10531 Set video output size. Default is @var{hd720}.
10534 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10537 Set output mode, can be @var{fulll} or @var{compact}.
10538 In @var{compact} mode only filters with some queued frames have displayed stats.
10541 Set flags which enable which stats are shown in video.
10543 Available values for flags are:
10546 Display number of queued frames in each link.
10548 @item frame_count_in
10549 Display number of frames taken from filter.
10551 @item frame_count_out
10552 Display number of frames given out from filter.
10555 Display current filtered frame pts.
10558 Display current filtered frame time.
10561 Display time base for filter link.
10564 Display used format for filter link.
10567 Display video size or number of audio channels in case of audio used by filter link.
10570 Display video frame rate or sample rate in case of audio used by filter link.
10574 Set upper limit for video rate of output stream, Default value is @var{25}.
10575 This guarantee that output video frame rate will not be higher than this value.
10579 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10580 and corrects the scene colors accordingly.
10582 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10584 The filter accepts the following options:
10588 The order of differentiation to be applied on the scene. Must be chosen in the range
10589 [0,2] and default value is 1.
10592 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10593 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10594 max value instead of calculating Minkowski distance.
10597 The standard deviation of Gaussian blur to be applied on the scene. Must be
10598 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10599 can't be equal to 0 if @var{difford} is greater than 0.
10602 @subsection Examples
10608 greyedge=difford=1:minknorm=5:sigma=2
10614 greyedge=difford=1:minknorm=0:sigma=2
10622 Apply a Hald CLUT to a video stream.
10624 First input is the video stream to process, and second one is the Hald CLUT.
10625 The Hald CLUT input can be a simple picture or a complete video stream.
10627 The filter accepts the following options:
10631 Force termination when the shortest input terminates. Default is @code{0}.
10633 Continue applying the last CLUT after the end of the stream. A value of
10634 @code{0} disable the filter after the last frame of the CLUT is reached.
10635 Default is @code{1}.
10638 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10639 filters share the same internals).
10641 More information about the Hald CLUT can be found on Eskil Steenberg's website
10642 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10644 @subsection Workflow examples
10646 @subsubsection Hald CLUT video stream
10648 Generate an identity Hald CLUT stream altered with various effects:
10650 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
10653 Note: make sure you use a lossless codec.
10655 Then use it with @code{haldclut} to apply it on some random stream:
10657 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10660 The Hald CLUT will be applied to the 10 first seconds (duration of
10661 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10662 to the remaining frames of the @code{mandelbrot} stream.
10664 @subsubsection Hald CLUT with preview
10666 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10667 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10668 biggest possible square starting at the top left of the picture. The remaining
10669 padding pixels (bottom or right) will be ignored. This area can be used to add
10670 a preview of the Hald CLUT.
10672 Typically, the following generated Hald CLUT will be supported by the
10673 @code{haldclut} filter:
10676 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10677 pad=iw+320 [padded_clut];
10678 smptebars=s=320x256, split [a][b];
10679 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10680 [main][b] overlay=W-320" -frames:v 1 clut.png
10683 It contains the original and a preview of the effect of the CLUT: SMPTE color
10684 bars are displayed on the right-top, and below the same color bars processed by
10687 Then, the effect of this Hald CLUT can be visualized with:
10689 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10694 Flip the input video horizontally.
10696 For example, to horizontally flip the input video with @command{ffmpeg}:
10698 ffmpeg -i in.avi -vf "hflip" out.avi
10702 This filter applies a global color histogram equalization on a
10705 It can be used to correct video that has a compressed range of pixel
10706 intensities. The filter redistributes the pixel intensities to
10707 equalize their distribution across the intensity range. It may be
10708 viewed as an "automatically adjusting contrast filter". This filter is
10709 useful only for correcting degraded or poorly captured source
10712 The filter accepts the following options:
10716 Determine the amount of equalization to be applied. As the strength
10717 is reduced, the distribution of pixel intensities more-and-more
10718 approaches that of the input frame. The value must be a float number
10719 in the range [0,1] and defaults to 0.200.
10722 Set the maximum intensity that can generated and scale the output
10723 values appropriately. The strength should be set as desired and then
10724 the intensity can be limited if needed to avoid washing-out. The value
10725 must be a float number in the range [0,1] and defaults to 0.210.
10728 Set the antibanding level. If enabled the filter will randomly vary
10729 the luminance of output pixels by a small amount to avoid banding of
10730 the histogram. Possible values are @code{none}, @code{weak} or
10731 @code{strong}. It defaults to @code{none}.
10736 Compute and draw a color distribution histogram for the input video.
10738 The computed histogram is a representation of the color component
10739 distribution in an image.
10741 Standard histogram displays the color components distribution in an image.
10742 Displays color graph for each color component. Shows distribution of
10743 the Y, U, V, A or R, G, B components, depending on input format, in the
10744 current frame. Below each graph a color component scale meter is shown.
10746 The filter accepts the following options:
10750 Set height of level. Default value is @code{200}.
10751 Allowed range is [50, 2048].
10754 Set height of color scale. Default value is @code{12}.
10755 Allowed range is [0, 40].
10759 It accepts the following values:
10762 Per color component graphs are placed below each other.
10765 Per color component graphs are placed side by side.
10768 Presents information identical to that in the @code{parade}, except
10769 that the graphs representing color components are superimposed directly
10772 Default is @code{stack}.
10775 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10776 Default is @code{linear}.
10779 Set what color components to display.
10780 Default is @code{7}.
10783 Set foreground opacity. Default is @code{0.7}.
10786 Set background opacity. Default is @code{0.5}.
10789 @subsection Examples
10794 Calculate and draw histogram:
10796 ffplay -i input -vf histogram
10804 This is a high precision/quality 3d denoise filter. It aims to reduce
10805 image noise, producing smooth images and making still images really
10806 still. It should enhance compressibility.
10808 It accepts the following optional parameters:
10812 A non-negative floating point number which specifies spatial luma strength.
10813 It defaults to 4.0.
10815 @item chroma_spatial
10816 A non-negative floating point number which specifies spatial chroma strength.
10817 It defaults to 3.0*@var{luma_spatial}/4.0.
10820 A floating point number which specifies luma temporal strength. It defaults to
10821 6.0*@var{luma_spatial}/4.0.
10824 A floating point number which specifies chroma temporal strength. It defaults to
10825 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10828 @anchor{hwdownload}
10829 @section hwdownload
10831 Download hardware frames to system memory.
10833 The input must be in hardware frames, and the output a non-hardware format.
10834 Not all formats will be supported on the output - it may be necessary to insert
10835 an additional @option{format} filter immediately following in the graph to get
10836 the output in a supported format.
10840 Map hardware frames to system memory or to another device.
10842 This filter has several different modes of operation; which one is used depends
10843 on the input and output formats:
10846 Hardware frame input, normal frame output
10848 Map the input frames to system memory and pass them to the output. If the
10849 original hardware frame is later required (for example, after overlaying
10850 something else on part of it), the @option{hwmap} filter can be used again
10851 in the next mode to retrieve it.
10853 Normal frame input, hardware frame output
10855 If the input is actually a software-mapped hardware frame, then unmap it -
10856 that is, return the original hardware frame.
10858 Otherwise, a device must be provided. Create new hardware surfaces on that
10859 device for the output, then map them back to the software format at the input
10860 and give those frames to the preceding filter. This will then act like the
10861 @option{hwupload} filter, but may be able to avoid an additional copy when
10862 the input is already in a compatible format.
10864 Hardware frame input and output
10866 A device must be supplied for the output, either directly or with the
10867 @option{derive_device} option. The input and output devices must be of
10868 different types and compatible - the exact meaning of this is
10869 system-dependent, but typically it means that they must refer to the same
10870 underlying hardware context (for example, refer to the same graphics card).
10872 If the input frames were originally created on the output device, then unmap
10873 to retrieve the original frames.
10875 Otherwise, map the frames to the output device - create new hardware frames
10876 on the output corresponding to the frames on the input.
10879 The following additional parameters are accepted:
10883 Set the frame mapping mode. Some combination of:
10886 The mapped frame should be readable.
10888 The mapped frame should be writeable.
10890 The mapping will always overwrite the entire frame.
10892 This may improve performance in some cases, as the original contents of the
10893 frame need not be loaded.
10895 The mapping must not involve any copying.
10897 Indirect mappings to copies of frames are created in some cases where either
10898 direct mapping is not possible or it would have unexpected properties.
10899 Setting this flag ensures that the mapping is direct and will fail if that is
10902 Defaults to @var{read+write} if not specified.
10904 @item derive_device @var{type}
10905 Rather than using the device supplied at initialisation, instead derive a new
10906 device of type @var{type} from the device the input frames exist on.
10909 In a hardware to hardware mapping, map in reverse - create frames in the sink
10910 and map them back to the source. This may be necessary in some cases where
10911 a mapping in one direction is required but only the opposite direction is
10912 supported by the devices being used.
10914 This option is dangerous - it may break the preceding filter in undefined
10915 ways if there are any additional constraints on that filter's output.
10916 Do not use it without fully understanding the implications of its use.
10922 Upload system memory frames to hardware surfaces.
10924 The device to upload to must be supplied when the filter is initialised. If
10925 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10928 @anchor{hwupload_cuda}
10929 @section hwupload_cuda
10931 Upload system memory frames to a CUDA device.
10933 It accepts the following optional parameters:
10937 The number of the CUDA device to use
10942 Apply a high-quality magnification filter designed for pixel art. This filter
10943 was originally created by Maxim Stepin.
10945 It accepts the following option:
10949 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10950 @code{hq3x} and @code{4} for @code{hq4x}.
10951 Default is @code{3}.
10955 Stack input videos horizontally.
10957 All streams must be of same pixel format and of same height.
10959 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10960 to create same output.
10962 The filter accept the following option:
10966 Set number of input streams. Default is 2.
10969 If set to 1, force the output to terminate when the shortest input
10970 terminates. Default value is 0.
10975 Modify the hue and/or the saturation of the input.
10977 It accepts the following parameters:
10981 Specify the hue angle as a number of degrees. It accepts an expression,
10982 and defaults to "0".
10985 Specify the saturation in the [-10,10] range. It accepts an expression and
10989 Specify the hue angle as a number of radians. It accepts an
10990 expression, and defaults to "0".
10993 Specify the brightness in the [-10,10] range. It accepts an expression and
10997 @option{h} and @option{H} are mutually exclusive, and can't be
10998 specified at the same time.
11000 The @option{b}, @option{h}, @option{H} and @option{s} option values are
11001 expressions containing the following constants:
11005 frame count of the input frame starting from 0
11008 presentation timestamp of the input frame expressed in time base units
11011 frame rate of the input video, NAN if the input frame rate is unknown
11014 timestamp expressed in seconds, NAN if the input timestamp is unknown
11017 time base of the input video
11020 @subsection Examples
11024 Set the hue to 90 degrees and the saturation to 1.0:
11030 Same command but expressing the hue in radians:
11036 Rotate hue and make the saturation swing between 0
11037 and 2 over a period of 1 second:
11039 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11043 Apply a 3 seconds saturation fade-in effect starting at 0:
11045 hue="s=min(t/3\,1)"
11048 The general fade-in expression can be written as:
11050 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11054 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11056 hue="s=max(0\, min(1\, (8-t)/3))"
11059 The general fade-out expression can be written as:
11061 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11066 @subsection Commands
11068 This filter supports the following commands:
11074 Modify the hue and/or the saturation and/or brightness of the input video.
11075 The command accepts the same syntax of the corresponding option.
11077 If the specified expression is not valid, it is kept at its current
11081 @section hysteresis
11083 Grow first stream into second stream by connecting components.
11084 This makes it possible to build more robust edge masks.
11086 This filter accepts the following options:
11090 Set which planes will be processed as bitmap, unprocessed planes will be
11091 copied from first stream.
11092 By default value 0xf, all planes will be processed.
11095 Set threshold which is used in filtering. If pixel component value is higher than
11096 this value filter algorithm for connecting components is activated.
11097 By default value is 0.
11102 Detect video interlacing type.
11104 This filter tries to detect if the input frames are interlaced, progressive,
11105 top or bottom field first. It will also try to detect fields that are
11106 repeated between adjacent frames (a sign of telecine).
11108 Single frame detection considers only immediately adjacent frames when classifying each frame.
11109 Multiple frame detection incorporates the classification history of previous frames.
11111 The filter will log these metadata values:
11114 @item single.current_frame
11115 Detected type of current frame using single-frame detection. One of:
11116 ``tff'' (top field first), ``bff'' (bottom field first),
11117 ``progressive'', or ``undetermined''
11120 Cumulative number of frames detected as top field first using single-frame detection.
11123 Cumulative number of frames detected as top field first using multiple-frame detection.
11126 Cumulative number of frames detected as bottom field first using single-frame detection.
11128 @item multiple.current_frame
11129 Detected type of current frame using multiple-frame detection. One of:
11130 ``tff'' (top field first), ``bff'' (bottom field first),
11131 ``progressive'', or ``undetermined''
11134 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11136 @item single.progressive
11137 Cumulative number of frames detected as progressive using single-frame detection.
11139 @item multiple.progressive
11140 Cumulative number of frames detected as progressive using multiple-frame detection.
11142 @item single.undetermined
11143 Cumulative number of frames that could not be classified using single-frame detection.
11145 @item multiple.undetermined
11146 Cumulative number of frames that could not be classified using multiple-frame detection.
11148 @item repeated.current_frame
11149 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11151 @item repeated.neither
11152 Cumulative number of frames with no repeated field.
11155 Cumulative number of frames with the top field repeated from the previous frame's top field.
11157 @item repeated.bottom
11158 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11161 The filter accepts the following options:
11165 Set interlacing threshold.
11167 Set progressive threshold.
11169 Threshold for repeated field detection.
11171 Number of frames after which a given frame's contribution to the
11172 statistics is halved (i.e., it contributes only 0.5 to its
11173 classification). The default of 0 means that all frames seen are given
11174 full weight of 1.0 forever.
11175 @item analyze_interlaced_flag
11176 When this is not 0 then idet will use the specified number of frames to determine
11177 if the interlaced flag is accurate, it will not count undetermined frames.
11178 If the flag is found to be accurate it will be used without any further
11179 computations, if it is found to be inaccurate it will be cleared without any
11180 further computations. This allows inserting the idet filter as a low computational
11181 method to clean up the interlaced flag
11186 Deinterleave or interleave fields.
11188 This filter allows one to process interlaced images fields without
11189 deinterlacing them. Deinterleaving splits the input frame into 2
11190 fields (so called half pictures). Odd lines are moved to the top
11191 half of the output image, even lines to the bottom half.
11192 You can process (filter) them independently and then re-interleave them.
11194 The filter accepts the following options:
11198 @item chroma_mode, c
11199 @item alpha_mode, a
11200 Available values for @var{luma_mode}, @var{chroma_mode} and
11201 @var{alpha_mode} are:
11207 @item deinterleave, d
11208 Deinterleave fields, placing one above the other.
11210 @item interleave, i
11211 Interleave fields. Reverse the effect of deinterleaving.
11213 Default value is @code{none}.
11215 @item luma_swap, ls
11216 @item chroma_swap, cs
11217 @item alpha_swap, as
11218 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11223 Apply inflate effect to the video.
11225 This filter replaces the pixel by the local(3x3) average by taking into account
11226 only values higher than the pixel.
11228 It accepts the following options:
11235 Limit the maximum change for each plane, default is 65535.
11236 If 0, plane will remain unchanged.
11241 Simple interlacing filter from progressive contents. This interleaves upper (or
11242 lower) lines from odd frames with lower (or upper) lines from even frames,
11243 halving the frame rate and preserving image height.
11246 Original Original New Frame
11247 Frame 'j' Frame 'j+1' (tff)
11248 ========== =========== ==================
11249 Line 0 --------------------> Frame 'j' Line 0
11250 Line 1 Line 1 ----> Frame 'j+1' Line 1
11251 Line 2 ---------------------> Frame 'j' Line 2
11252 Line 3 Line 3 ----> Frame 'j+1' Line 3
11254 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11257 It accepts the following optional parameters:
11261 This determines whether the interlaced frame is taken from the even
11262 (tff - default) or odd (bff) lines of the progressive frame.
11265 Vertical lowpass filter to avoid twitter interlacing and
11266 reduce moire patterns.
11270 Disable vertical lowpass filter
11273 Enable linear filter (default)
11276 Enable complex filter. This will slightly less reduce twitter and moire
11277 but better retain detail and subjective sharpness impression.
11284 Deinterlace input video by applying Donald Graft's adaptive kernel
11285 deinterling. Work on interlaced parts of a video to produce
11286 progressive frames.
11288 The description of the accepted parameters follows.
11292 Set the threshold which affects the filter's tolerance when
11293 determining if a pixel line must be processed. It must be an integer
11294 in the range [0,255] and defaults to 10. A value of 0 will result in
11295 applying the process on every pixels.
11298 Paint pixels exceeding the threshold value to white if set to 1.
11302 Set the fields order. Swap fields if set to 1, leave fields alone if
11306 Enable additional sharpening if set to 1. Default is 0.
11309 Enable twoway sharpening if set to 1. Default is 0.
11312 @subsection Examples
11316 Apply default values:
11318 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11322 Enable additional sharpening:
11328 Paint processed pixels in white:
11334 @section lenscorrection
11336 Correct radial lens distortion
11338 This filter can be used to correct for radial distortion as can result from the use
11339 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11340 one can use tools available for example as part of opencv or simply trial-and-error.
11341 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11342 and extract the k1 and k2 coefficients from the resulting matrix.
11344 Note that effectively the same filter is available in the open-source tools Krita and
11345 Digikam from the KDE project.
11347 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11348 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11349 brightness distribution, so you may want to use both filters together in certain
11350 cases, though you will have to take care of ordering, i.e. whether vignetting should
11351 be applied before or after lens correction.
11353 @subsection Options
11355 The filter accepts the following options:
11359 Relative x-coordinate of the focal point of the image, and thereby the center of the
11360 distortion. This value has a range [0,1] and is expressed as fractions of the image
11361 width. Default is 0.5.
11363 Relative y-coordinate of the focal point of the image, and thereby the center of the
11364 distortion. This value has a range [0,1] and is expressed as fractions of the image
11365 height. Default is 0.5.
11367 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11368 no correction. Default is 0.
11370 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11371 0 means no correction. Default is 0.
11374 The formula that generates the correction is:
11376 @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)
11378 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11379 distances from the focal point in the source and target images, respectively.
11383 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11385 The @code{lensfun} filter requires the camera make, camera model, and lens model
11386 to apply the lens correction. The filter will load the lensfun database and
11387 query it to find the corresponding camera and lens entries in the database. As
11388 long as these entries can be found with the given options, the filter can
11389 perform corrections on frames. Note that incomplete strings will result in the
11390 filter choosing the best match with the given options, and the filter will
11391 output the chosen camera and lens models (logged with level "info"). You must
11392 provide the make, camera model, and lens model as they are required.
11394 The filter accepts the following options:
11398 The make of the camera (for example, "Canon"). This option is required.
11401 The model of the camera (for example, "Canon EOS 100D"). This option is
11405 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11406 option is required.
11409 The type of correction to apply. The following values are valid options:
11413 Enables fixing lens vignetting.
11416 Enables fixing lens geometry. This is the default.
11419 Enables fixing chromatic aberrations.
11422 Enables fixing lens vignetting and lens geometry.
11425 Enables fixing lens vignetting and chromatic aberrations.
11428 Enables fixing both lens geometry and chromatic aberrations.
11431 Enables all possible corrections.
11435 The focal length of the image/video (zoom; expected constant for video). For
11436 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11437 range should be chosen when using that lens. Default 18.
11440 The aperture of the image/video (expected constant for video). Note that
11441 aperture is only used for vignetting correction. Default 3.5.
11443 @item focus_distance
11444 The focus distance of the image/video (expected constant for video). Note that
11445 focus distance is only used for vignetting and only slightly affects the
11446 vignetting correction process. If unknown, leave it at the default value (which
11450 The scale factor which is applied after transformation. After correction the
11451 video is no longer necessarily rectangular. This parameter controls how much of
11452 the resulting image is visible. The value 0 means that a value will be chosen
11453 automatically such that there is little or no unmapped area in the output
11454 image. 1.0 means that no additional scaling is done. Lower values may result
11455 in more of the corrected image being visible, while higher values may avoid
11456 unmapped areas in the output.
11458 @item target_geometry
11459 The target geometry of the output image/video. The following values are valid
11463 @item rectilinear (default)
11466 @item equirectangular
11467 @item fisheye_orthographic
11468 @item fisheye_stereographic
11469 @item fisheye_equisolid
11470 @item fisheye_thoby
11473 Apply the reverse of image correction (instead of correcting distortion, apply
11476 @item interpolation
11477 The type of interpolation used when correcting distortion. The following values
11482 @item linear (default)
11487 @subsection Examples
11491 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11492 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11496 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
11500 Apply the same as before, but only for the first 5 seconds of video.
11503 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
11510 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11511 score between two input videos.
11513 The obtained VMAF score is printed through the logging system.
11515 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11516 After installing the library it can be enabled using:
11517 @code{./configure --enable-libvmaf --enable-version3}.
11518 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11520 The filter has following options:
11524 Set the model path which is to be used for SVM.
11525 Default value: @code{"vmaf_v0.6.1.pkl"}
11528 Set the file path to be used to store logs.
11531 Set the format of the log file (xml or json).
11533 @item enable_transform
11534 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11535 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11536 Default value: @code{false}
11539 Invokes the phone model which will generate VMAF scores higher than in the
11540 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11543 Enables computing psnr along with vmaf.
11546 Enables computing ssim along with vmaf.
11549 Enables computing ms_ssim along with vmaf.
11552 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11555 Set number of threads to be used when computing vmaf.
11558 Set interval for frame subsampling used when computing vmaf.
11560 @item enable_conf_interval
11561 Enables confidence interval.
11564 This filter also supports the @ref{framesync} options.
11566 On the below examples the input file @file{main.mpg} being processed is
11567 compared with the reference file @file{ref.mpg}.
11570 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11573 Example with options:
11575 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11580 Limits the pixel components values to the specified range [min, max].
11582 The filter accepts the following options:
11586 Lower bound. Defaults to the lowest allowed value for the input.
11589 Upper bound. Defaults to the highest allowed value for the input.
11592 Specify which planes will be processed. Defaults to all available.
11599 The filter accepts the following options:
11603 Set the number of loops. Setting this value to -1 will result in infinite loops.
11607 Set maximal size in number of frames. Default is 0.
11610 Set first frame of loop. Default is 0.
11613 @subsection Examples
11617 Loop single first frame infinitely:
11619 loop=loop=-1:size=1:start=0
11623 Loop single first frame 10 times:
11625 loop=loop=10:size=1:start=0
11629 Loop 10 first frames 5 times:
11631 loop=loop=5:size=10:start=0
11637 Apply a 1D LUT to an input video.
11639 The filter accepts the following options:
11643 Set the 1D LUT file name.
11645 Currently supported formats:
11652 Select interpolation mode.
11654 Available values are:
11658 Use values from the nearest defined point.
11660 Interpolate values using the linear interpolation.
11662 Interpolate values using the cosine interpolation.
11664 Interpolate values using the cubic interpolation.
11666 Interpolate values using the spline interpolation.
11673 Apply a 3D LUT to an input video.
11675 The filter accepts the following options:
11679 Set the 3D LUT file name.
11681 Currently supported formats:
11693 Select interpolation mode.
11695 Available values are:
11699 Use values from the nearest defined point.
11701 Interpolate values using the 8 points defining a cube.
11703 Interpolate values using a tetrahedron.
11707 This filter also supports the @ref{framesync} options.
11711 Turn certain luma values into transparency.
11713 The filter accepts the following options:
11717 Set the luma which will be used as base for transparency.
11718 Default value is @code{0}.
11721 Set the range of luma values to be keyed out.
11722 Default value is @code{0}.
11725 Set the range of softness. Default value is @code{0}.
11726 Use this to control gradual transition from zero to full transparency.
11729 @section lut, lutrgb, lutyuv
11731 Compute a look-up table for binding each pixel component input value
11732 to an output value, and apply it to the input video.
11734 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11735 to an RGB input video.
11737 These filters accept the following parameters:
11740 set first pixel component expression
11742 set second pixel component expression
11744 set third pixel component expression
11746 set fourth pixel component expression, corresponds to the alpha component
11749 set red component expression
11751 set green component expression
11753 set blue component expression
11755 alpha component expression
11758 set Y/luminance component expression
11760 set U/Cb component expression
11762 set V/Cr component expression
11765 Each of them specifies the expression to use for computing the lookup table for
11766 the corresponding pixel component values.
11768 The exact component associated to each of the @var{c*} options depends on the
11771 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11772 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11774 The expressions can contain the following constants and functions:
11779 The input width and height.
11782 The input value for the pixel component.
11785 The input value, clipped to the @var{minval}-@var{maxval} range.
11788 The maximum value for the pixel component.
11791 The minimum value for the pixel component.
11794 The negated value for the pixel component value, clipped to the
11795 @var{minval}-@var{maxval} range; it corresponds to the expression
11796 "maxval-clipval+minval".
11799 The computed value in @var{val}, clipped to the
11800 @var{minval}-@var{maxval} range.
11802 @item gammaval(gamma)
11803 The computed gamma correction value of the pixel component value,
11804 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11806 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11810 All expressions default to "val".
11812 @subsection Examples
11816 Negate input video:
11818 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11819 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11822 The above is the same as:
11824 lutrgb="r=negval:g=negval:b=negval"
11825 lutyuv="y=negval:u=negval:v=negval"
11835 Remove chroma components, turning the video into a graytone image:
11837 lutyuv="u=128:v=128"
11841 Apply a luma burning effect:
11847 Remove green and blue components:
11853 Set a constant alpha channel value on input:
11855 format=rgba,lutrgb=a="maxval-minval/2"
11859 Correct luminance gamma by a factor of 0.5:
11861 lutyuv=y=gammaval(0.5)
11865 Discard least significant bits of luma:
11867 lutyuv=y='bitand(val, 128+64+32)'
11871 Technicolor like effect:
11873 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11877 @section lut2, tlut2
11879 The @code{lut2} filter takes two input streams and outputs one
11882 The @code{tlut2} (time lut2) filter takes two consecutive frames
11883 from one single stream.
11885 This filter accepts the following parameters:
11888 set first pixel component expression
11890 set second pixel component expression
11892 set third pixel component expression
11894 set fourth pixel component expression, corresponds to the alpha component
11897 set output bit depth, only available for @code{lut2} filter. By default is 0,
11898 which means bit depth is automatically picked from first input format.
11901 Each of them specifies the expression to use for computing the lookup table for
11902 the corresponding pixel component values.
11904 The exact component associated to each of the @var{c*} options depends on the
11907 The expressions can contain the following constants:
11912 The input width and height.
11915 The first input value for the pixel component.
11918 The second input value for the pixel component.
11921 The first input video bit depth.
11924 The second input video bit depth.
11927 All expressions default to "x".
11929 @subsection Examples
11933 Highlight differences between two RGB video streams:
11935 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)'
11939 Highlight differences between two YUV video streams:
11941 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)'
11945 Show max difference between two video streams:
11947 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)))'
11951 @section maskedclamp
11953 Clamp the first input stream with the second input and third input stream.
11955 Returns the value of first stream to be between second input
11956 stream - @code{undershoot} and third input stream + @code{overshoot}.
11958 This filter accepts the following options:
11961 Default value is @code{0}.
11964 Default value is @code{0}.
11967 Set which planes will be processed as bitmap, unprocessed planes will be
11968 copied from first stream.
11969 By default value 0xf, all planes will be processed.
11972 @section maskedmerge
11974 Merge the first input stream with the second input stream using per pixel
11975 weights in the third input stream.
11977 A value of 0 in the third stream pixel component means that pixel component
11978 from first stream is returned unchanged, while maximum value (eg. 255 for
11979 8-bit videos) means that pixel component from second stream is returned
11980 unchanged. Intermediate values define the amount of merging between both
11981 input stream's pixel components.
11983 This filter accepts the following options:
11986 Set which planes will be processed as bitmap, unprocessed planes will be
11987 copied from first stream.
11988 By default value 0xf, all planes will be processed.
11992 Create mask from input video.
11994 For example it is useful to create motion masks after @code{tblend} filter.
11996 This filter accepts the following options:
12000 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
12003 Set high threshold. Any pixel component higher than this value will be set to max value
12004 allowed for current pixel format.
12007 Set planes to filter, by default all available planes are filtered.
12010 Fill all frame pixels with this value.
12013 Set max average pixel value for frame. If sum of all pixel components is higher that this
12014 average, output frame will be completely filled with value set by @var{fill} option.
12015 Typically useful for scene changes when used in combination with @code{tblend} filter.
12020 Apply motion-compensation deinterlacing.
12022 It needs one field per frame as input and must thus be used together
12023 with yadif=1/3 or equivalent.
12025 This filter accepts the following options:
12028 Set the deinterlacing mode.
12030 It accepts one of the following values:
12035 use iterative motion estimation
12037 like @samp{slow}, but use multiple reference frames.
12039 Default value is @samp{fast}.
12042 Set the picture field parity assumed for the input video. It must be
12043 one of the following values:
12047 assume top field first
12049 assume bottom field first
12052 Default value is @samp{bff}.
12055 Set per-block quantization parameter (QP) used by the internal
12058 Higher values should result in a smoother motion vector field but less
12059 optimal individual vectors. Default value is 1.
12062 @section mergeplanes
12064 Merge color channel components from several video streams.
12066 The filter accepts up to 4 input streams, and merge selected input
12067 planes to the output video.
12069 This filter accepts the following options:
12072 Set input to output plane mapping. Default is @code{0}.
12074 The mappings is specified as a bitmap. It should be specified as a
12075 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12076 mapping for the first plane of the output stream. 'A' sets the number of
12077 the input stream to use (from 0 to 3), and 'a' the plane number of the
12078 corresponding input to use (from 0 to 3). The rest of the mappings is
12079 similar, 'Bb' describes the mapping for the output stream second
12080 plane, 'Cc' describes the mapping for the output stream third plane and
12081 'Dd' describes the mapping for the output stream fourth plane.
12084 Set output pixel format. Default is @code{yuva444p}.
12087 @subsection Examples
12091 Merge three gray video streams of same width and height into single video stream:
12093 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12097 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12099 [a0][a1]mergeplanes=0x00010210:yuva444p
12103 Swap Y and A plane in yuva444p stream:
12105 format=yuva444p,mergeplanes=0x03010200:yuva444p
12109 Swap U and V plane in yuv420p stream:
12111 format=yuv420p,mergeplanes=0x000201:yuv420p
12115 Cast a rgb24 clip to yuv444p:
12117 format=rgb24,mergeplanes=0x000102:yuv444p
12123 Estimate and export motion vectors using block matching algorithms.
12124 Motion vectors are stored in frame side data to be used by other filters.
12126 This filter accepts the following options:
12129 Specify the motion estimation method. Accepts one of the following values:
12133 Exhaustive search algorithm.
12135 Three step search algorithm.
12137 Two dimensional logarithmic search algorithm.
12139 New three step search algorithm.
12141 Four step search algorithm.
12143 Diamond search algorithm.
12145 Hexagon-based search algorithm.
12147 Enhanced predictive zonal search algorithm.
12149 Uneven multi-hexagon search algorithm.
12151 Default value is @samp{esa}.
12154 Macroblock size. Default @code{16}.
12157 Search parameter. Default @code{7}.
12160 @section midequalizer
12162 Apply Midway Image Equalization effect using two video streams.
12164 Midway Image Equalization adjusts a pair of images to have the same
12165 histogram, while maintaining their dynamics as much as possible. It's
12166 useful for e.g. matching exposures from a pair of stereo cameras.
12168 This filter has two inputs and one output, which must be of same pixel format, but
12169 may be of different sizes. The output of filter is first input adjusted with
12170 midway histogram of both inputs.
12172 This filter accepts the following option:
12176 Set which planes to process. Default is @code{15}, which is all available planes.
12179 @section minterpolate
12181 Convert the video to specified frame rate using motion interpolation.
12183 This filter accepts the following options:
12186 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}.
12189 Motion interpolation mode. Following values are accepted:
12192 Duplicate previous or next frame for interpolating new ones.
12194 Blend source frames. Interpolated frame is mean of previous and next frames.
12196 Motion compensated interpolation. Following options are effective when this mode is selected:
12200 Motion compensation mode. Following values are accepted:
12203 Overlapped block motion compensation.
12205 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12207 Default mode is @samp{obmc}.
12210 Motion estimation mode. Following values are accepted:
12213 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12215 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12217 Default mode is @samp{bilat}.
12220 The algorithm to be used for motion estimation. Following values are accepted:
12223 Exhaustive search algorithm.
12225 Three step search algorithm.
12227 Two dimensional logarithmic search algorithm.
12229 New three step search algorithm.
12231 Four step search algorithm.
12233 Diamond search algorithm.
12235 Hexagon-based search algorithm.
12237 Enhanced predictive zonal search algorithm.
12239 Uneven multi-hexagon search algorithm.
12241 Default algorithm is @samp{epzs}.
12244 Macroblock size. Default @code{16}.
12247 Motion estimation search parameter. Default @code{32}.
12250 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).
12255 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:
12258 Disable scene change detection.
12260 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12262 Default method is @samp{fdiff}.
12264 @item scd_threshold
12265 Scene change detection threshold. Default is @code{5.0}.
12270 Mix several video input streams into one video stream.
12272 A description of the accepted options follows.
12276 The number of inputs. If unspecified, it defaults to 2.
12279 Specify weight of each input video stream as sequence.
12280 Each weight is separated by space. If number of weights
12281 is smaller than number of @var{frames} last specified
12282 weight will be used for all remaining unset weights.
12285 Specify scale, if it is set it will be multiplied with sum
12286 of each weight multiplied with pixel values to give final destination
12287 pixel value. By default @var{scale} is auto scaled to sum of weights.
12290 Specify how end of stream is determined.
12293 The duration of the longest input. (default)
12296 The duration of the shortest input.
12299 The duration of the first input.
12303 @section mpdecimate
12305 Drop frames that do not differ greatly from the previous frame in
12306 order to reduce frame rate.
12308 The main use of this filter is for very-low-bitrate encoding
12309 (e.g. streaming over dialup modem), but it could in theory be used for
12310 fixing movies that were inverse-telecined incorrectly.
12312 A description of the accepted options follows.
12316 Set the maximum number of consecutive frames which can be dropped (if
12317 positive), or the minimum interval between dropped frames (if
12318 negative). If the value is 0, the frame is dropped disregarding the
12319 number of previous sequentially dropped frames.
12321 Default value is 0.
12326 Set the dropping threshold values.
12328 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12329 represent actual pixel value differences, so a threshold of 64
12330 corresponds to 1 unit of difference for each pixel, or the same spread
12331 out differently over the block.
12333 A frame is a candidate for dropping if no 8x8 blocks differ by more
12334 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12335 meaning the whole image) differ by more than a threshold of @option{lo}.
12337 Default value for @option{hi} is 64*12, default value for @option{lo} is
12338 64*5, and default value for @option{frac} is 0.33.
12344 Negate (invert) the input video.
12346 It accepts the following option:
12351 With value 1, it negates the alpha component, if present. Default value is 0.
12357 Denoise frames using Non-Local Means algorithm.
12359 Each pixel is adjusted by looking for other pixels with similar contexts. This
12360 context similarity is defined by comparing their surrounding patches of size
12361 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12364 Note that the research area defines centers for patches, which means some
12365 patches will be made of pixels outside that research area.
12367 The filter accepts the following options.
12371 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12374 Set patch size. Default is 7. Must be odd number in range [0, 99].
12377 Same as @option{p} but for chroma planes.
12379 The default value is @var{0} and means automatic.
12382 Set research size. Default is 15. Must be odd number in range [0, 99].
12385 Same as @option{r} but for chroma planes.
12387 The default value is @var{0} and means automatic.
12392 Deinterlace video using neural network edge directed interpolation.
12394 This filter accepts the following options:
12398 Mandatory option, without binary file filter can not work.
12399 Currently file can be found here:
12400 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12403 Set which frames to deinterlace, by default it is @code{all}.
12404 Can be @code{all} or @code{interlaced}.
12407 Set mode of operation.
12409 Can be one of the following:
12413 Use frame flags, both fields.
12415 Use frame flags, single field.
12417 Use top field only.
12419 Use bottom field only.
12421 Use both fields, top first.
12423 Use both fields, bottom first.
12427 Set which planes to process, by default filter process all frames.
12430 Set size of local neighborhood around each pixel, used by the predictor neural
12433 Can be one of the following:
12446 Set the number of neurons in predictor neural network.
12447 Can be one of the following:
12458 Controls the number of different neural network predictions that are blended
12459 together to compute the final output value. Can be @code{fast}, default or
12463 Set which set of weights to use in the predictor.
12464 Can be one of the following:
12468 weights trained to minimize absolute error
12470 weights trained to minimize squared error
12474 Controls whether or not the prescreener neural network is used to decide
12475 which pixels should be processed by the predictor neural network and which
12476 can be handled by simple cubic interpolation.
12477 The prescreener is trained to know whether cubic interpolation will be
12478 sufficient for a pixel or whether it should be predicted by the predictor nn.
12479 The computational complexity of the prescreener nn is much less than that of
12480 the predictor nn. Since most pixels can be handled by cubic interpolation,
12481 using the prescreener generally results in much faster processing.
12482 The prescreener is pretty accurate, so the difference between using it and not
12483 using it is almost always unnoticeable.
12485 Can be one of the following:
12493 Default is @code{new}.
12496 Set various debugging flags.
12501 Force libavfilter not to use any of the specified pixel formats for the
12502 input to the next filter.
12504 It accepts the following parameters:
12508 A '|'-separated list of pixel format names, such as
12509 pix_fmts=yuv420p|monow|rgb24".
12513 @subsection Examples
12517 Force libavfilter to use a format different from @var{yuv420p} for the
12518 input to the vflip filter:
12520 noformat=pix_fmts=yuv420p,vflip
12524 Convert the input video to any of the formats not contained in the list:
12526 noformat=yuv420p|yuv444p|yuv410p
12532 Add noise on video input frame.
12534 The filter accepts the following options:
12542 Set noise seed for specific pixel component or all pixel components in case
12543 of @var{all_seed}. Default value is @code{123457}.
12545 @item all_strength, alls
12546 @item c0_strength, c0s
12547 @item c1_strength, c1s
12548 @item c2_strength, c2s
12549 @item c3_strength, c3s
12550 Set noise strength for specific pixel component or all pixel components in case
12551 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12553 @item all_flags, allf
12554 @item c0_flags, c0f
12555 @item c1_flags, c1f
12556 @item c2_flags, c2f
12557 @item c3_flags, c3f
12558 Set pixel component flags or set flags for all components if @var{all_flags}.
12559 Available values for component flags are:
12562 averaged temporal noise (smoother)
12564 mix random noise with a (semi)regular pattern
12566 temporal noise (noise pattern changes between frames)
12568 uniform noise (gaussian otherwise)
12572 @subsection Examples
12574 Add temporal and uniform noise to input video:
12576 noise=alls=20:allf=t+u
12581 Normalize RGB video (aka histogram stretching, contrast stretching).
12582 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12584 For each channel of each frame, the filter computes the input range and maps
12585 it linearly to the user-specified output range. The output range defaults
12586 to the full dynamic range from pure black to pure white.
12588 Temporal smoothing can be used on the input range to reduce flickering (rapid
12589 changes in brightness) caused when small dark or bright objects enter or leave
12590 the scene. This is similar to the auto-exposure (automatic gain control) on a
12591 video camera, and, like a video camera, it may cause a period of over- or
12592 under-exposure of the video.
12594 The R,G,B channels can be normalized independently, which may cause some
12595 color shifting, or linked together as a single channel, which prevents
12596 color shifting. Linked normalization preserves hue. Independent normalization
12597 does not, so it can be used to remove some color casts. Independent and linked
12598 normalization can be combined in any ratio.
12600 The normalize filter accepts the following options:
12605 Colors which define the output range. The minimum input value is mapped to
12606 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12607 The defaults are black and white respectively. Specifying white for
12608 @var{blackpt} and black for @var{whitept} will give color-inverted,
12609 normalized video. Shades of grey can be used to reduce the dynamic range
12610 (contrast). Specifying saturated colors here can create some interesting
12614 The number of previous frames to use for temporal smoothing. The input range
12615 of each channel is smoothed using a rolling average over the current frame
12616 and the @var{smoothing} previous frames. The default is 0 (no temporal
12620 Controls the ratio of independent (color shifting) channel normalization to
12621 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12622 independent. Defaults to 1.0 (fully independent).
12625 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12626 expensive no-op. Defaults to 1.0 (full strength).
12630 @subsection Examples
12632 Stretch video contrast to use the full dynamic range, with no temporal
12633 smoothing; may flicker depending on the source content:
12635 normalize=blackpt=black:whitept=white:smoothing=0
12638 As above, but with 50 frames of temporal smoothing; flicker should be
12639 reduced, depending on the source content:
12641 normalize=blackpt=black:whitept=white:smoothing=50
12644 As above, but with hue-preserving linked channel normalization:
12646 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12649 As above, but with half strength:
12651 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12654 Map the darkest input color to red, the brightest input color to cyan:
12656 normalize=blackpt=red:whitept=cyan
12661 Pass the video source unchanged to the output.
12664 Optical Character Recognition
12666 This filter uses Tesseract for optical character recognition. To enable
12667 compilation of this filter, you need to configure FFmpeg with
12668 @code{--enable-libtesseract}.
12670 It accepts the following options:
12674 Set datapath to tesseract data. Default is to use whatever was
12675 set at installation.
12678 Set language, default is "eng".
12681 Set character whitelist.
12684 Set character blacklist.
12687 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12691 Apply a video transform using libopencv.
12693 To enable this filter, install the libopencv library and headers and
12694 configure FFmpeg with @code{--enable-libopencv}.
12696 It accepts the following parameters:
12701 The name of the libopencv filter to apply.
12703 @item filter_params
12704 The parameters to pass to the libopencv filter. If not specified, the default
12705 values are assumed.
12709 Refer to the official libopencv documentation for more precise
12711 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12713 Several libopencv filters are supported; see the following subsections.
12718 Dilate an image by using a specific structuring element.
12719 It corresponds to the libopencv function @code{cvDilate}.
12721 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12723 @var{struct_el} represents a structuring element, and has the syntax:
12724 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12726 @var{cols} and @var{rows} represent the number of columns and rows of
12727 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12728 point, and @var{shape} the shape for the structuring element. @var{shape}
12729 must be "rect", "cross", "ellipse", or "custom".
12731 If the value for @var{shape} is "custom", it must be followed by a
12732 string of the form "=@var{filename}". The file with name
12733 @var{filename} is assumed to represent a binary image, with each
12734 printable character corresponding to a bright pixel. When a custom
12735 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12736 or columns and rows of the read file are assumed instead.
12738 The default value for @var{struct_el} is "3x3+0x0/rect".
12740 @var{nb_iterations} specifies the number of times the transform is
12741 applied to the image, and defaults to 1.
12745 # Use the default values
12748 # Dilate using a structuring element with a 5x5 cross, iterating two times
12749 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12751 # Read the shape from the file diamond.shape, iterating two times.
12752 # The file diamond.shape may contain a pattern of characters like this
12758 # The specified columns and rows are ignored
12759 # but the anchor point coordinates are not
12760 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12765 Erode an image by using a specific structuring element.
12766 It corresponds to the libopencv function @code{cvErode}.
12768 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12769 with the same syntax and semantics as the @ref{dilate} filter.
12773 Smooth the input video.
12775 The filter takes the following parameters:
12776 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12778 @var{type} is the type of smooth filter to apply, and must be one of
12779 the following values: "blur", "blur_no_scale", "median", "gaussian",
12780 or "bilateral". The default value is "gaussian".
12782 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12783 depend on the smooth type. @var{param1} and
12784 @var{param2} accept integer positive values or 0. @var{param3} and
12785 @var{param4} accept floating point values.
12787 The default value for @var{param1} is 3. The default value for the
12788 other parameters is 0.
12790 These parameters correspond to the parameters assigned to the
12791 libopencv function @code{cvSmooth}.
12793 @section oscilloscope
12795 2D Video Oscilloscope.
12797 Useful to measure spatial impulse, step responses, chroma delays, etc.
12799 It accepts the following parameters:
12803 Set scope center x position.
12806 Set scope center y position.
12809 Set scope size, relative to frame diagonal.
12812 Set scope tilt/rotation.
12818 Set trace center x position.
12821 Set trace center y position.
12824 Set trace width, relative to width of frame.
12827 Set trace height, relative to height of frame.
12830 Set which components to trace. By default it traces first three components.
12833 Draw trace grid. By default is enabled.
12836 Draw some statistics. By default is enabled.
12839 Draw scope. By default is enabled.
12842 @subsection Examples
12846 Inspect full first row of video frame.
12848 oscilloscope=x=0.5:y=0:s=1
12852 Inspect full last row of video frame.
12854 oscilloscope=x=0.5:y=1:s=1
12858 Inspect full 5th line of video frame of height 1080.
12860 oscilloscope=x=0.5:y=5/1080:s=1
12864 Inspect full last column of video frame.
12866 oscilloscope=x=1:y=0.5:s=1:t=1
12874 Overlay one video on top of another.
12876 It takes two inputs and has one output. The first input is the "main"
12877 video on which the second input is overlaid.
12879 It accepts the following parameters:
12881 A description of the accepted options follows.
12886 Set the expression for the x and y coordinates of the overlaid video
12887 on the main video. Default value is "0" for both expressions. In case
12888 the expression is invalid, it is set to a huge value (meaning that the
12889 overlay will not be displayed within the output visible area).
12892 See @ref{framesync}.
12895 Set when the expressions for @option{x}, and @option{y} are evaluated.
12897 It accepts the following values:
12900 only evaluate expressions once during the filter initialization or
12901 when a command is processed
12904 evaluate expressions for each incoming frame
12907 Default value is @samp{frame}.
12910 See @ref{framesync}.
12913 Set the format for the output video.
12915 It accepts the following values:
12918 force YUV420 output
12921 force YUV422 output
12924 force YUV444 output
12927 force packed RGB output
12930 force planar RGB output
12933 automatically pick format
12936 Default value is @samp{yuv420}.
12939 See @ref{framesync}.
12942 Set format of alpha of the overlaid video, it can be @var{straight} or
12943 @var{premultiplied}. Default is @var{straight}.
12946 The @option{x}, and @option{y} expressions can contain the following
12952 The main input width and height.
12956 The overlay input width and height.
12960 The computed values for @var{x} and @var{y}. They are evaluated for
12965 horizontal and vertical chroma subsample values of the output
12966 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
12970 the number of input frame, starting from 0
12973 the position in the file of the input frame, NAN if unknown
12976 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
12980 This filter also supports the @ref{framesync} options.
12982 Note that the @var{n}, @var{pos}, @var{t} variables are available only
12983 when evaluation is done @emph{per frame}, and will evaluate to NAN
12984 when @option{eval} is set to @samp{init}.
12986 Be aware that frames are taken from each input video in timestamp
12987 order, hence, if their initial timestamps differ, it is a good idea
12988 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
12989 have them begin in the same zero timestamp, as the example for
12990 the @var{movie} filter does.
12992 You can chain together more overlays but you should test the
12993 efficiency of such approach.
12995 @subsection Commands
12997 This filter supports the following commands:
13001 Modify the x and y of the overlay input.
13002 The command accepts the same syntax of the corresponding option.
13004 If the specified expression is not valid, it is kept at its current
13008 @subsection Examples
13012 Draw the overlay at 10 pixels from the bottom right corner of the main
13015 overlay=main_w-overlay_w-10:main_h-overlay_h-10
13018 Using named options the example above becomes:
13020 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
13024 Insert a transparent PNG logo in the bottom left corner of the input,
13025 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
13027 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
13031 Insert 2 different transparent PNG logos (second logo on bottom
13032 right corner) using the @command{ffmpeg} tool:
13034 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
13038 Add a transparent color layer on top of the main video; @code{WxH}
13039 must specify the size of the main input to the overlay filter:
13041 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
13045 Play an original video and a filtered version (here with the deshake
13046 filter) side by side using the @command{ffplay} tool:
13048 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13051 The above command is the same as:
13053 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13057 Make a sliding overlay appearing from the left to the right top part of the
13058 screen starting since time 2:
13060 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13064 Compose output by putting two input videos side to side:
13066 ffmpeg -i left.avi -i right.avi -filter_complex "
13067 nullsrc=size=200x100 [background];
13068 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13069 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13070 [background][left] overlay=shortest=1 [background+left];
13071 [background+left][right] overlay=shortest=1:x=100 [left+right]
13076 Mask 10-20 seconds of a video by applying the delogo filter to a section
13078 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13079 -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]'
13084 Chain several overlays in cascade:
13086 nullsrc=s=200x200 [bg];
13087 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13088 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13089 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13090 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13091 [in3] null, [mid2] overlay=100:100 [out0]
13098 Apply Overcomplete Wavelet denoiser.
13100 The filter accepts the following options:
13106 Larger depth values will denoise lower frequency components more, but
13107 slow down filtering.
13109 Must be an int in the range 8-16, default is @code{8}.
13111 @item luma_strength, ls
13114 Must be a double value in the range 0-1000, default is @code{1.0}.
13116 @item chroma_strength, cs
13117 Set chroma strength.
13119 Must be a double value in the range 0-1000, default is @code{1.0}.
13125 Add paddings to the input image, and place the original input at the
13126 provided @var{x}, @var{y} coordinates.
13128 It accepts the following parameters:
13133 Specify an expression for the size of the output image with the
13134 paddings added. If the value for @var{width} or @var{height} is 0, the
13135 corresponding input size is used for the output.
13137 The @var{width} expression can reference the value set by the
13138 @var{height} expression, and vice versa.
13140 The default value of @var{width} and @var{height} is 0.
13144 Specify the offsets to place the input image at within the padded area,
13145 with respect to the top/left border of the output image.
13147 The @var{x} expression can reference the value set by the @var{y}
13148 expression, and vice versa.
13150 The default value of @var{x} and @var{y} is 0.
13152 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13153 so the input image is centered on the padded area.
13156 Specify the color of the padded area. For the syntax of this option,
13157 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13158 manual,ffmpeg-utils}.
13160 The default value of @var{color} is "black".
13163 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13165 It accepts the following values:
13169 Only evaluate expressions once during the filter initialization or when
13170 a command is processed.
13173 Evaluate expressions for each incoming frame.
13177 Default value is @samp{init}.
13180 Pad to aspect instead to a resolution.
13184 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13185 options are expressions containing the following constants:
13190 The input video width and height.
13194 These are the same as @var{in_w} and @var{in_h}.
13198 The output width and height (the size of the padded area), as
13199 specified by the @var{width} and @var{height} expressions.
13203 These are the same as @var{out_w} and @var{out_h}.
13207 The x and y offsets as specified by the @var{x} and @var{y}
13208 expressions, or NAN if not yet specified.
13211 same as @var{iw} / @var{ih}
13214 input sample aspect ratio
13217 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13221 The horizontal and vertical chroma subsample values. For example for the
13222 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13225 @subsection Examples
13229 Add paddings with the color "violet" to the input video. The output video
13230 size is 640x480, and the top-left corner of the input video is placed at
13233 pad=640:480:0:40:violet
13236 The example above is equivalent to the following command:
13238 pad=width=640:height=480:x=0:y=40:color=violet
13242 Pad the input to get an output with dimensions increased by 3/2,
13243 and put the input video at the center of the padded area:
13245 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13249 Pad the input to get a squared output with size equal to the maximum
13250 value between the input width and height, and put the input video at
13251 the center of the padded area:
13253 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13257 Pad the input to get a final w/h ratio of 16:9:
13259 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13263 In case of anamorphic video, in order to set the output display aspect
13264 correctly, it is necessary to use @var{sar} in the expression,
13265 according to the relation:
13267 (ih * X / ih) * sar = output_dar
13268 X = output_dar / sar
13271 Thus the previous example needs to be modified to:
13273 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13277 Double the output size and put the input video in the bottom-right
13278 corner of the output padded area:
13280 pad="2*iw:2*ih:ow-iw:oh-ih"
13284 @anchor{palettegen}
13285 @section palettegen
13287 Generate one palette for a whole video stream.
13289 It accepts the following options:
13293 Set the maximum number of colors to quantize in the palette.
13294 Note: the palette will still contain 256 colors; the unused palette entries
13297 @item reserve_transparent
13298 Create a palette of 255 colors maximum and reserve the last one for
13299 transparency. Reserving the transparency color is useful for GIF optimization.
13300 If not set, the maximum of colors in the palette will be 256. You probably want
13301 to disable this option for a standalone image.
13304 @item transparency_color
13305 Set the color that will be used as background for transparency.
13308 Set statistics mode.
13310 It accepts the following values:
13313 Compute full frame histograms.
13315 Compute histograms only for the part that differs from previous frame. This
13316 might be relevant to give more importance to the moving part of your input if
13317 the background is static.
13319 Compute new histogram for each frame.
13322 Default value is @var{full}.
13325 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13326 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13327 color quantization of the palette. This information is also visible at
13328 @var{info} logging level.
13330 @subsection Examples
13334 Generate a representative palette of a given video using @command{ffmpeg}:
13336 ffmpeg -i input.mkv -vf palettegen palette.png
13340 @section paletteuse
13342 Use a palette to downsample an input video stream.
13344 The filter takes two inputs: one video stream and a palette. The palette must
13345 be a 256 pixels image.
13347 It accepts the following options:
13351 Select dithering mode. Available algorithms are:
13354 Ordered 8x8 bayer dithering (deterministic)
13356 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13357 Note: this dithering is sometimes considered "wrong" and is included as a
13359 @item floyd_steinberg
13360 Floyd and Steingberg dithering (error diffusion)
13362 Frankie Sierra dithering v2 (error diffusion)
13364 Frankie Sierra dithering v2 "Lite" (error diffusion)
13367 Default is @var{sierra2_4a}.
13370 When @var{bayer} dithering is selected, this option defines the scale of the
13371 pattern (how much the crosshatch pattern is visible). A low value means more
13372 visible pattern for less banding, and higher value means less visible pattern
13373 at the cost of more banding.
13375 The option must be an integer value in the range [0,5]. Default is @var{2}.
13378 If set, define the zone to process
13382 Only the changing rectangle will be reprocessed. This is similar to GIF
13383 cropping/offsetting compression mechanism. This option can be useful for speed
13384 if only a part of the image is changing, and has use cases such as limiting the
13385 scope of the error diffusal @option{dither} to the rectangle that bounds the
13386 moving scene (it leads to more deterministic output if the scene doesn't change
13387 much, and as a result less moving noise and better GIF compression).
13390 Default is @var{none}.
13393 Take new palette for each output frame.
13395 @item alpha_threshold
13396 Sets the alpha threshold for transparency. Alpha values above this threshold
13397 will be treated as completely opaque, and values below this threshold will be
13398 treated as completely transparent.
13400 The option must be an integer value in the range [0,255]. Default is @var{128}.
13403 @subsection Examples
13407 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13408 using @command{ffmpeg}:
13410 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13414 @section perspective
13416 Correct perspective of video not recorded perpendicular to the screen.
13418 A description of the accepted parameters follows.
13429 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13430 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13431 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13432 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13433 then the corners of the source will be sent to the specified coordinates.
13435 The expressions can use the following variables:
13440 the width and height of video frame.
13444 Output frame count.
13447 @item interpolation
13448 Set interpolation for perspective correction.
13450 It accepts the following values:
13456 Default value is @samp{linear}.
13459 Set interpretation of coordinate options.
13461 It accepts the following values:
13465 Send point in the source specified by the given coordinates to
13466 the corners of the destination.
13468 @item 1, destination
13470 Send the corners of the source to the point in the destination specified
13471 by the given coordinates.
13473 Default value is @samp{source}.
13477 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13479 It accepts the following values:
13482 only evaluate expressions once during the filter initialization or
13483 when a command is processed
13486 evaluate expressions for each incoming frame
13489 Default value is @samp{init}.
13494 Delay interlaced video by one field time so that the field order changes.
13496 The intended use is to fix PAL movies that have been captured with the
13497 opposite field order to the film-to-video transfer.
13499 A description of the accepted parameters follows.
13505 It accepts the following values:
13508 Capture field order top-first, transfer bottom-first.
13509 Filter will delay the bottom field.
13512 Capture field order bottom-first, transfer top-first.
13513 Filter will delay the top field.
13516 Capture and transfer with the same field order. This mode only exists
13517 for the documentation of the other options to refer to, but if you
13518 actually select it, the filter will faithfully do nothing.
13521 Capture field order determined automatically by field flags, transfer
13523 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13524 basis using field flags. If no field information is available,
13525 then this works just like @samp{u}.
13528 Capture unknown or varying, transfer opposite.
13529 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13530 analyzing the images and selecting the alternative that produces best
13531 match between the fields.
13534 Capture top-first, transfer unknown or varying.
13535 Filter selects among @samp{t} and @samp{p} using image analysis.
13538 Capture bottom-first, transfer unknown or varying.
13539 Filter selects among @samp{b} and @samp{p} using image analysis.
13542 Capture determined by field flags, transfer unknown or varying.
13543 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13544 image analysis. If no field information is available, then this works just
13545 like @samp{U}. This is the default mode.
13548 Both capture and transfer unknown or varying.
13549 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13553 @section pixdesctest
13555 Pixel format descriptor test filter, mainly useful for internal
13556 testing. The output video should be equal to the input video.
13560 format=monow, pixdesctest
13563 can be used to test the monowhite pixel format descriptor definition.
13567 Display sample values of color channels. Mainly useful for checking color
13568 and levels. Minimum supported resolution is 640x480.
13570 The filters accept the following options:
13574 Set scope X position, relative offset on X axis.
13577 Set scope Y position, relative offset on Y axis.
13586 Set window opacity. This window also holds statistics about pixel area.
13589 Set window X position, relative offset on X axis.
13592 Set window Y position, relative offset on Y axis.
13597 Enable the specified chain of postprocessing subfilters using libpostproc. This
13598 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13599 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13600 Each subfilter and some options have a short and a long name that can be used
13601 interchangeably, i.e. dr/dering are the same.
13603 The filters accept the following options:
13607 Set postprocessing subfilters string.
13610 All subfilters share common options to determine their scope:
13614 Honor the quality commands for this subfilter.
13617 Do chrominance filtering, too (default).
13620 Do luminance filtering only (no chrominance).
13623 Do chrominance filtering only (no luminance).
13626 These options can be appended after the subfilter name, separated by a '|'.
13628 Available subfilters are:
13631 @item hb/hdeblock[|difference[|flatness]]
13632 Horizontal deblocking filter
13635 Difference factor where higher values mean more deblocking (default: @code{32}).
13637 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13640 @item vb/vdeblock[|difference[|flatness]]
13641 Vertical deblocking filter
13644 Difference factor where higher values mean more deblocking (default: @code{32}).
13646 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13649 @item ha/hadeblock[|difference[|flatness]]
13650 Accurate horizontal deblocking filter
13653 Difference factor where higher values mean more deblocking (default: @code{32}).
13655 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13658 @item va/vadeblock[|difference[|flatness]]
13659 Accurate vertical deblocking filter
13662 Difference factor where higher values mean more deblocking (default: @code{32}).
13664 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13668 The horizontal and vertical deblocking filters share the difference and
13669 flatness values so you cannot set different horizontal and vertical
13673 @item h1/x1hdeblock
13674 Experimental horizontal deblocking filter
13676 @item v1/x1vdeblock
13677 Experimental vertical deblocking filter
13682 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13685 larger -> stronger filtering
13687 larger -> stronger filtering
13689 larger -> stronger filtering
13692 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13695 Stretch luminance to @code{0-255}.
13698 @item lb/linblenddeint
13699 Linear blend deinterlacing filter that deinterlaces the given block by
13700 filtering all lines with a @code{(1 2 1)} filter.
13702 @item li/linipoldeint
13703 Linear interpolating deinterlacing filter that deinterlaces the given block by
13704 linearly interpolating every second line.
13706 @item ci/cubicipoldeint
13707 Cubic interpolating deinterlacing filter deinterlaces the given block by
13708 cubically interpolating every second line.
13710 @item md/mediandeint
13711 Median deinterlacing filter that deinterlaces the given block by applying a
13712 median filter to every second line.
13714 @item fd/ffmpegdeint
13715 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13716 second line with a @code{(-1 4 2 4 -1)} filter.
13719 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13720 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13722 @item fq/forceQuant[|quantizer]
13723 Overrides the quantizer table from the input with the constant quantizer you
13731 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13734 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13737 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13740 @subsection Examples
13744 Apply horizontal and vertical deblocking, deringing and automatic
13745 brightness/contrast:
13751 Apply default filters without brightness/contrast correction:
13757 Apply default filters and temporal denoiser:
13759 pp=default/tmpnoise|1|2|3
13763 Apply deblocking on luminance only, and switch vertical deblocking on or off
13764 automatically depending on available CPU time:
13771 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13772 similar to spp = 6 with 7 point DCT, where only the center sample is
13775 The filter accepts the following options:
13779 Force a constant quantization parameter. It accepts an integer in range
13780 0 to 63. If not set, the filter will use the QP from the video stream
13784 Set thresholding mode. Available modes are:
13788 Set hard thresholding.
13790 Set soft thresholding (better de-ringing effect, but likely blurrier).
13792 Set medium thresholding (good results, default).
13796 @section premultiply
13797 Apply alpha premultiply effect to input video stream using first plane
13798 of second stream as alpha.
13800 Both streams must have same dimensions and same pixel format.
13802 The filter accepts the following option:
13806 Set which planes will be processed, unprocessed planes will be copied.
13807 By default value 0xf, all planes will be processed.
13810 Do not require 2nd input for processing, instead use alpha plane from input stream.
13814 Apply prewitt operator to input video stream.
13816 The filter accepts the following option:
13820 Set which planes will be processed, unprocessed planes will be copied.
13821 By default value 0xf, all planes will be processed.
13824 Set value which will be multiplied with filtered result.
13827 Set value which will be added to filtered result.
13830 @anchor{program_opencl}
13831 @section program_opencl
13833 Filter video using an OpenCL program.
13838 OpenCL program source file.
13841 Kernel name in program.
13844 Number of inputs to the filter. Defaults to 1.
13847 Size of output frames. Defaults to the same as the first input.
13851 The program source file must contain a kernel function with the given name,
13852 which will be run once for each plane of the output. Each run on a plane
13853 gets enqueued as a separate 2D global NDRange with one work-item for each
13854 pixel to be generated. The global ID offset for each work-item is therefore
13855 the coordinates of a pixel in the destination image.
13857 The kernel function needs to take the following arguments:
13860 Destination image, @var{__write_only image2d_t}.
13862 This image will become the output; the kernel should write all of it.
13864 Frame index, @var{unsigned int}.
13866 This is a counter starting from zero and increasing by one for each frame.
13868 Source images, @var{__read_only image2d_t}.
13870 These are the most recent images on each input. The kernel may read from
13871 them to generate the output, but they can't be written to.
13878 Copy the input to the output (output must be the same size as the input).
13880 __kernel void copy(__write_only image2d_t destination,
13881 unsigned int index,
13882 __read_only image2d_t source)
13884 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13886 int2 location = (int2)(get_global_id(0), get_global_id(1));
13888 float4 value = read_imagef(source, sampler, location);
13890 write_imagef(destination, location, value);
13895 Apply a simple transformation, rotating the input by an amount increasing
13896 with the index counter. Pixel values are linearly interpolated by the
13897 sampler, and the output need not have the same dimensions as the input.
13899 __kernel void rotate_image(__write_only image2d_t dst,
13900 unsigned int index,
13901 __read_only image2d_t src)
13903 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13904 CLK_FILTER_LINEAR);
13906 float angle = (float)index / 100.0f;
13908 float2 dst_dim = convert_float2(get_image_dim(dst));
13909 float2 src_dim = convert_float2(get_image_dim(src));
13911 float2 dst_cen = dst_dim / 2.0f;
13912 float2 src_cen = src_dim / 2.0f;
13914 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13916 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13918 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13919 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13921 src_pos = src_pos * src_dim / dst_dim;
13923 float2 src_loc = src_pos + src_cen;
13925 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13926 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13927 write_imagef(dst, dst_loc, 0.5f);
13929 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13934 Blend two inputs together, with the amount of each input used varying
13935 with the index counter.
13937 __kernel void blend_images(__write_only image2d_t dst,
13938 unsigned int index,
13939 __read_only image2d_t src1,
13940 __read_only image2d_t src2)
13942 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13943 CLK_FILTER_LINEAR);
13945 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
13947 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13948 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
13949 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
13951 float4 val1 = read_imagef(src1, sampler, src1_loc);
13952 float4 val2 = read_imagef(src2, sampler, src2_loc);
13954 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
13960 @section pseudocolor
13962 Alter frame colors in video with pseudocolors.
13964 This filter accept the following options:
13968 set pixel first component expression
13971 set pixel second component expression
13974 set pixel third component expression
13977 set pixel fourth component expression, corresponds to the alpha component
13980 set component to use as base for altering colors
13983 Each of them specifies the expression to use for computing the lookup table for
13984 the corresponding pixel component values.
13986 The expressions can contain the following constants and functions:
13991 The input width and height.
13994 The input value for the pixel component.
13996 @item ymin, umin, vmin, amin
13997 The minimum allowed component value.
13999 @item ymax, umax, vmax, amax
14000 The maximum allowed component value.
14003 All expressions default to "val".
14005 @subsection Examples
14009 Change too high luma values to gradient:
14011 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'"
14017 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
14018 Ratio) between two input videos.
14020 This filter takes in input two input videos, the first input is
14021 considered the "main" source and is passed unchanged to the
14022 output. The second input is used as a "reference" video for computing
14025 Both video inputs must have the same resolution and pixel format for
14026 this filter to work correctly. Also it assumes that both inputs
14027 have the same number of frames, which are compared one by one.
14029 The obtained average PSNR is printed through the logging system.
14031 The filter stores the accumulated MSE (mean squared error) of each
14032 frame, and at the end of the processing it is averaged across all frames
14033 equally, and the following formula is applied to obtain the PSNR:
14036 PSNR = 10*log10(MAX^2/MSE)
14039 Where MAX is the average of the maximum values of each component of the
14042 The description of the accepted parameters follows.
14045 @item stats_file, f
14046 If specified the filter will use the named file to save the PSNR of
14047 each individual frame. When filename equals "-" the data is sent to
14050 @item stats_version
14051 Specifies which version of the stats file format to use. Details of
14052 each format are written below.
14053 Default value is 1.
14055 @item stats_add_max
14056 Determines whether the max value is output to the stats log.
14057 Default value is 0.
14058 Requires stats_version >= 2. If this is set and stats_version < 2,
14059 the filter will return an error.
14062 This filter also supports the @ref{framesync} options.
14064 The file printed if @var{stats_file} is selected, contains a sequence of
14065 key/value pairs of the form @var{key}:@var{value} for each compared
14068 If a @var{stats_version} greater than 1 is specified, a header line precedes
14069 the list of per-frame-pair stats, with key value pairs following the frame
14070 format with the following parameters:
14073 @item psnr_log_version
14074 The version of the log file format. Will match @var{stats_version}.
14077 A comma separated list of the per-frame-pair parameters included in
14081 A description of each shown per-frame-pair parameter follows:
14085 sequential number of the input frame, starting from 1
14088 Mean Square Error pixel-by-pixel average difference of the compared
14089 frames, averaged over all the image components.
14091 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14092 Mean Square Error pixel-by-pixel average difference of the compared
14093 frames for the component specified by the suffix.
14095 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14096 Peak Signal to Noise ratio of the compared frames for the component
14097 specified by the suffix.
14099 @item max_avg, max_y, max_u, max_v
14100 Maximum allowed value for each channel, and average over all
14106 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14107 [main][ref] psnr="stats_file=stats.log" [out]
14110 On this example the input file being processed is compared with the
14111 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14112 is stored in @file{stats.log}.
14117 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14118 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14121 The pullup filter is designed to take advantage of future context in making
14122 its decisions. This filter is stateless in the sense that it does not lock
14123 onto a pattern to follow, but it instead looks forward to the following
14124 fields in order to identify matches and rebuild progressive frames.
14126 To produce content with an even framerate, insert the fps filter after
14127 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14128 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14130 The filter accepts the following options:
14137 These options set the amount of "junk" to ignore at the left, right, top, and
14138 bottom of the image, respectively. Left and right are in units of 8 pixels,
14139 while top and bottom are in units of 2 lines.
14140 The default is 8 pixels on each side.
14143 Set the strict breaks. Setting this option to 1 will reduce the chances of
14144 filter generating an occasional mismatched frame, but it may also cause an
14145 excessive number of frames to be dropped during high motion sequences.
14146 Conversely, setting it to -1 will make filter match fields more easily.
14147 This may help processing of video where there is slight blurring between
14148 the fields, but may also cause there to be interlaced frames in the output.
14149 Default value is @code{0}.
14152 Set the metric plane to use. It accepts the following values:
14158 Use chroma blue plane.
14161 Use chroma red plane.
14164 This option may be set to use chroma plane instead of the default luma plane
14165 for doing filter's computations. This may improve accuracy on very clean
14166 source material, but more likely will decrease accuracy, especially if there
14167 is chroma noise (rainbow effect) or any grayscale video.
14168 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14169 load and make pullup usable in realtime on slow machines.
14172 For best results (without duplicated frames in the output file) it is
14173 necessary to change the output frame rate. For example, to inverse
14174 telecine NTSC input:
14176 ffmpeg -i input -vf pullup -r 24000/1001 ...
14181 Change video quantization parameters (QP).
14183 The filter accepts the following option:
14187 Set expression for quantization parameter.
14190 The expression is evaluated through the eval API and can contain, among others,
14191 the following constants:
14195 1 if index is not 129, 0 otherwise.
14198 Sequential index starting from -129 to 128.
14201 @subsection Examples
14205 Some equation like:
14213 Flush video frames from internal cache of frames into a random order.
14214 No frame is discarded.
14215 Inspired by @ref{frei0r} nervous filter.
14219 Set size in number of frames of internal cache, in range from @code{2} to
14220 @code{512}. Default is @code{30}.
14223 Set seed for random number generator, must be an integer included between
14224 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14225 less than @code{0}, the filter will try to use a good random seed on a
14229 @section readeia608
14231 Read closed captioning (EIA-608) information from the top lines of a video frame.
14233 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14234 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14235 with EIA-608 data (starting from 0). A description of each metadata value follows:
14238 @item lavfi.readeia608.X.cc
14239 The two bytes stored as EIA-608 data (printed in hexadecimal).
14241 @item lavfi.readeia608.X.line
14242 The number of the line on which the EIA-608 data was identified and read.
14245 This filter accepts the following options:
14249 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14252 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14255 Set minimal acceptable amplitude change for sync codes detection.
14256 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14259 Set the ratio of width reserved for sync code detection.
14260 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14263 Set the max peaks height difference for sync code detection.
14264 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14267 Set max peaks period difference for sync code detection.
14268 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14271 Set the first two max start code bits differences.
14272 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14275 Set the minimum ratio of bits height compared to 3rd start code bit.
14276 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14279 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14282 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14285 Enable checking the parity bit. In the event of a parity error, the filter will output
14286 @code{0x00} for that character. Default is false.
14289 @subsection Examples
14293 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14295 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
14301 Read vertical interval timecode (VITC) information from the top lines of a
14304 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14305 timecode value, if a valid timecode has been detected. Further metadata key
14306 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14307 timecode data has been found or not.
14309 This filter accepts the following options:
14313 Set the maximum number of lines to scan for VITC data. If the value is set to
14314 @code{-1} the full video frame is scanned. Default is @code{45}.
14317 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14318 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14321 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14322 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14325 @subsection Examples
14329 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14330 draw @code{--:--:--:--} as a placeholder:
14332 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14338 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14340 Destination pixel at position (X, Y) will be picked from source (x, y) position
14341 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14342 value for pixel will be used for destination pixel.
14344 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14345 will have Xmap/Ymap video stream dimensions.
14346 Xmap and Ymap input video streams are 16bit depth, single channel.
14348 @section removegrain
14350 The removegrain filter is a spatial denoiser for progressive video.
14354 Set mode for the first plane.
14357 Set mode for the second plane.
14360 Set mode for the third plane.
14363 Set mode for the fourth plane.
14366 Range of mode is from 0 to 24. Description of each mode follows:
14370 Leave input plane unchanged. Default.
14373 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14376 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14379 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14382 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14383 This is equivalent to a median filter.
14386 Line-sensitive clipping giving the minimal change.
14389 Line-sensitive clipping, intermediate.
14392 Line-sensitive clipping, intermediate.
14395 Line-sensitive clipping, intermediate.
14398 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14401 Replaces the target pixel with the closest neighbour.
14404 [1 2 1] horizontal and vertical kernel blur.
14410 Bob mode, interpolates top field from the line where the neighbours
14411 pixels are the closest.
14414 Bob mode, interpolates bottom field from the line where the neighbours
14415 pixels are the closest.
14418 Bob mode, interpolates top field. Same as 13 but with a more complicated
14419 interpolation formula.
14422 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14423 interpolation formula.
14426 Clips the pixel with the minimum and maximum of respectively the maximum and
14427 minimum of each pair of opposite neighbour pixels.
14430 Line-sensitive clipping using opposite neighbours whose greatest distance from
14431 the current pixel is minimal.
14434 Replaces the pixel with the average of its 8 neighbours.
14437 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14440 Clips pixels using the averages of opposite neighbour.
14443 Same as mode 21 but simpler and faster.
14446 Small edge and halo removal, but reputed useless.
14452 @section removelogo
14454 Suppress a TV station logo, using an image file to determine which
14455 pixels comprise the logo. It works by filling in the pixels that
14456 comprise the logo with neighboring pixels.
14458 The filter accepts the following options:
14462 Set the filter bitmap file, which can be any image format supported by
14463 libavformat. The width and height of the image file must match those of the
14464 video stream being processed.
14467 Pixels in the provided bitmap image with a value of zero are not
14468 considered part of the logo, non-zero pixels are considered part of
14469 the logo. If you use white (255) for the logo and black (0) for the
14470 rest, you will be safe. For making the filter bitmap, it is
14471 recommended to take a screen capture of a black frame with the logo
14472 visible, and then using a threshold filter followed by the erode
14473 filter once or twice.
14475 If needed, little splotches can be fixed manually. Remember that if
14476 logo pixels are not covered, the filter quality will be much
14477 reduced. Marking too many pixels as part of the logo does not hurt as
14478 much, but it will increase the amount of blurring needed to cover over
14479 the image and will destroy more information than necessary, and extra
14480 pixels will slow things down on a large logo.
14482 @section repeatfields
14484 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14485 fields based on its value.
14489 Reverse a video clip.
14491 Warning: This filter requires memory to buffer the entire clip, so trimming
14494 @subsection Examples
14498 Take the first 5 seconds of a clip, and reverse it.
14505 Shift R/G/B/A pixels horizontally and/or vertically.
14507 The filter accepts the following options:
14510 Set amount to shift red horizontally.
14512 Set amount to shift red vertically.
14514 Set amount to shift green horizontally.
14516 Set amount to shift green vertically.
14518 Set amount to shift blue horizontally.
14520 Set amount to shift blue vertically.
14522 Set amount to shift alpha horizontally.
14524 Set amount to shift alpha vertically.
14526 Set edge mode, can be @var{smear}, default, or @var{warp}.
14530 Apply roberts cross operator to input video stream.
14532 The filter accepts the following option:
14536 Set which planes will be processed, unprocessed planes will be copied.
14537 By default value 0xf, all planes will be processed.
14540 Set value which will be multiplied with filtered result.
14543 Set value which will be added to filtered result.
14548 Rotate video by an arbitrary angle expressed in radians.
14550 The filter accepts the following options:
14552 A description of the optional parameters follows.
14555 Set an expression for the angle by which to rotate the input video
14556 clockwise, expressed as a number of radians. A negative value will
14557 result in a counter-clockwise rotation. By default it is set to "0".
14559 This expression is evaluated for each frame.
14562 Set the output width expression, default value is "iw".
14563 This expression is evaluated just once during configuration.
14566 Set the output height expression, default value is "ih".
14567 This expression is evaluated just once during configuration.
14570 Enable bilinear interpolation if set to 1, a value of 0 disables
14571 it. Default value is 1.
14574 Set the color used to fill the output area not covered by the rotated
14575 image. For the general syntax of this option, check the
14576 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14577 If the special value "none" is selected then no
14578 background is printed (useful for example if the background is never shown).
14580 Default value is "black".
14583 The expressions for the angle and the output size can contain the
14584 following constants and functions:
14588 sequential number of the input frame, starting from 0. It is always NAN
14589 before the first frame is filtered.
14592 time in seconds of the input frame, it is set to 0 when the filter is
14593 configured. It is always NAN before the first frame is filtered.
14597 horizontal and vertical chroma subsample values. For example for the
14598 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14602 the input video width and height
14606 the output width and height, that is the size of the padded area as
14607 specified by the @var{width} and @var{height} expressions
14611 the minimal width/height required for completely containing the input
14612 video rotated by @var{a} radians.
14614 These are only available when computing the @option{out_w} and
14615 @option{out_h} expressions.
14618 @subsection Examples
14622 Rotate the input by PI/6 radians clockwise:
14628 Rotate the input by PI/6 radians counter-clockwise:
14634 Rotate the input by 45 degrees clockwise:
14640 Apply a constant rotation with period T, starting from an angle of PI/3:
14642 rotate=PI/3+2*PI*t/T
14646 Make the input video rotation oscillating with a period of T
14647 seconds and an amplitude of A radians:
14649 rotate=A*sin(2*PI/T*t)
14653 Rotate the video, output size is chosen so that the whole rotating
14654 input video is always completely contained in the output:
14656 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14660 Rotate the video, reduce the output size so that no background is ever
14663 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14667 @subsection Commands
14669 The filter supports the following commands:
14673 Set the angle expression.
14674 The command accepts the same syntax of the corresponding option.
14676 If the specified expression is not valid, it is kept at its current
14682 Apply Shape Adaptive Blur.
14684 The filter accepts the following options:
14687 @item luma_radius, lr
14688 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14689 value is 1.0. A greater value will result in a more blurred image, and
14690 in slower processing.
14692 @item luma_pre_filter_radius, lpfr
14693 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14696 @item luma_strength, ls
14697 Set luma maximum difference between pixels to still be considered, must
14698 be a value in the 0.1-100.0 range, default value is 1.0.
14700 @item chroma_radius, cr
14701 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14702 greater value will result in a more blurred image, and in slower
14705 @item chroma_pre_filter_radius, cpfr
14706 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14708 @item chroma_strength, cs
14709 Set chroma maximum difference between pixels to still be considered,
14710 must be a value in the -0.9-100.0 range.
14713 Each chroma option value, if not explicitly specified, is set to the
14714 corresponding luma option value.
14719 Scale (resize) the input video, using the libswscale library.
14721 The scale filter forces the output display aspect ratio to be the same
14722 of the input, by changing the output sample aspect ratio.
14724 If the input image format is different from the format requested by
14725 the next filter, the scale filter will convert the input to the
14728 @subsection Options
14729 The filter accepts the following options, or any of the options
14730 supported by the libswscale scaler.
14732 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14733 the complete list of scaler options.
14738 Set the output video dimension expression. Default value is the input
14741 If the @var{width} or @var{w} value is 0, the input width is used for
14742 the output. If the @var{height} or @var{h} value is 0, the input height
14743 is used for the output.
14745 If one and only one of the values is -n with n >= 1, the scale filter
14746 will use a value that maintains the aspect ratio of the input image,
14747 calculated from the other specified dimension. After that it will,
14748 however, make sure that the calculated dimension is divisible by n and
14749 adjust the value if necessary.
14751 If both values are -n with n >= 1, the behavior will be identical to
14752 both values being set to 0 as previously detailed.
14754 See below for the list of accepted constants for use in the dimension
14758 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14762 Only evaluate expressions once during the filter initialization or when a command is processed.
14765 Evaluate expressions for each incoming frame.
14769 Default value is @samp{init}.
14773 Set the interlacing mode. It accepts the following values:
14777 Force interlaced aware scaling.
14780 Do not apply interlaced scaling.
14783 Select interlaced aware scaling depending on whether the source frames
14784 are flagged as interlaced or not.
14787 Default value is @samp{0}.
14790 Set libswscale scaling flags. See
14791 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14792 complete list of values. If not explicitly specified the filter applies
14796 @item param0, param1
14797 Set libswscale input parameters for scaling algorithms that need them. See
14798 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14799 complete documentation. If not explicitly specified the filter applies
14805 Set the video size. For the syntax of this option, check the
14806 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14808 @item in_color_matrix
14809 @item out_color_matrix
14810 Set in/output YCbCr color space type.
14812 This allows the autodetected value to be overridden as well as allows forcing
14813 a specific value used for the output and encoder.
14815 If not specified, the color space type depends on the pixel format.
14821 Choose automatically.
14824 Format conforming to International Telecommunication Union (ITU)
14825 Recommendation BT.709.
14828 Set color space conforming to the United States Federal Communications
14829 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14832 Set color space conforming to:
14836 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14839 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14842 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14847 Set color space conforming to SMPTE ST 240:1999.
14852 Set in/output YCbCr sample range.
14854 This allows the autodetected value to be overridden as well as allows forcing
14855 a specific value used for the output and encoder. If not specified, the
14856 range depends on the pixel format. Possible values:
14860 Choose automatically.
14863 Set full range (0-255 in case of 8-bit luma).
14865 @item mpeg/limited/tv
14866 Set "MPEG" range (16-235 in case of 8-bit luma).
14869 @item force_original_aspect_ratio
14870 Enable decreasing or increasing output video width or height if necessary to
14871 keep the original aspect ratio. Possible values:
14875 Scale the video as specified and disable this feature.
14878 The output video dimensions will automatically be decreased if needed.
14881 The output video dimensions will automatically be increased if needed.
14885 One useful instance of this option is that when you know a specific device's
14886 maximum allowed resolution, you can use this to limit the output video to
14887 that, while retaining the aspect ratio. For example, device A allows
14888 1280x720 playback, and your video is 1920x800. Using this option (set it to
14889 decrease) and specifying 1280x720 to the command line makes the output
14892 Please note that this is a different thing than specifying -1 for @option{w}
14893 or @option{h}, you still need to specify the output resolution for this option
14898 The values of the @option{w} and @option{h} options are expressions
14899 containing the following constants:
14904 The input width and height
14908 These are the same as @var{in_w} and @var{in_h}.
14912 The output (scaled) width and height
14916 These are the same as @var{out_w} and @var{out_h}
14919 The same as @var{iw} / @var{ih}
14922 input sample aspect ratio
14925 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14929 horizontal and vertical input chroma subsample values. For example for the
14930 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14934 horizontal and vertical output chroma subsample values. For example for the
14935 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14938 @subsection Examples
14942 Scale the input video to a size of 200x100
14947 This is equivalent to:
14958 Specify a size abbreviation for the output size:
14963 which can also be written as:
14969 Scale the input to 2x:
14971 scale=w=2*iw:h=2*ih
14975 The above is the same as:
14977 scale=2*in_w:2*in_h
14981 Scale the input to 2x with forced interlaced scaling:
14983 scale=2*iw:2*ih:interl=1
14987 Scale the input to half size:
14989 scale=w=iw/2:h=ih/2
14993 Increase the width, and set the height to the same size:
14999 Seek Greek harmony:
15006 Increase the height, and set the width to 3/2 of the height:
15008 scale=w=3/2*oh:h=3/5*ih
15012 Increase the size, making the size a multiple of the chroma
15015 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
15019 Increase the width to a maximum of 500 pixels,
15020 keeping the same aspect ratio as the input:
15022 scale=w='min(500\, iw*3/2):h=-1'
15026 Make pixels square by combining scale and setsar:
15028 scale='trunc(ih*dar):ih',setsar=1/1
15032 Make pixels square by combining scale and setsar,
15033 making sure the resulting resolution is even (required by some codecs):
15035 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
15039 @subsection Commands
15041 This filter supports the following commands:
15045 Set the output video dimension expression.
15046 The command accepts the same syntax of the corresponding option.
15048 If the specified expression is not valid, it is kept at its current
15054 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15055 format conversion on CUDA video frames. Setting the output width and height
15056 works in the same way as for the @var{scale} filter.
15058 The following additional options are accepted:
15061 The pixel format of the output CUDA frames. If set to the string "same" (the
15062 default), the input format will be kept. Note that automatic format negotiation
15063 and conversion is not yet supported for hardware frames
15066 The interpolation algorithm used for resizing. One of the following:
15073 @item cubic2p_bspline
15074 2-parameter cubic (B=1, C=0)
15076 @item cubic2p_catmullrom
15077 2-parameter cubic (B=0, C=1/2)
15079 @item cubic2p_b05c03
15080 2-parameter cubic (B=1/2, C=3/10)
15092 Scale (resize) the input video, based on a reference video.
15094 See the scale filter for available options, scale2ref supports the same but
15095 uses the reference video instead of the main input as basis. scale2ref also
15096 supports the following additional constants for the @option{w} and
15097 @option{h} options:
15102 The main input video's width and height
15105 The same as @var{main_w} / @var{main_h}
15108 The main input video's sample aspect ratio
15110 @item main_dar, mdar
15111 The main input video's display aspect ratio. Calculated from
15112 @code{(main_w / main_h) * main_sar}.
15116 The main input video's horizontal and vertical chroma subsample values.
15117 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15121 @subsection Examples
15125 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15127 'scale2ref[b][a];[a][b]overlay'
15131 @anchor{selectivecolor}
15132 @section selectivecolor
15134 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15135 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15136 by the "purity" of the color (that is, how saturated it already is).
15138 This filter is similar to the Adobe Photoshop Selective Color tool.
15140 The filter accepts the following options:
15143 @item correction_method
15144 Select color correction method.
15146 Available values are:
15149 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15152 Specified adjustments are relative to the original component value.
15154 Default is @code{absolute}.
15156 Adjustments for red pixels (pixels where the red component is the maximum)
15158 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15160 Adjustments for green pixels (pixels where the green component is the maximum)
15162 Adjustments for cyan pixels (pixels where the red component is the minimum)
15164 Adjustments for blue pixels (pixels where the blue component is the maximum)
15166 Adjustments for magenta pixels (pixels where the green component is the minimum)
15168 Adjustments for white pixels (pixels where all components are greater than 128)
15170 Adjustments for all pixels except pure black and pure white
15172 Adjustments for black pixels (pixels where all components are lesser than 128)
15174 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15177 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15178 4 space separated floating point adjustment values in the [-1,1] range,
15179 respectively to adjust the amount of cyan, magenta, yellow and black for the
15180 pixels of its range.
15182 @subsection Examples
15186 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15187 increase magenta by 27% in blue areas:
15189 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15193 Use a Photoshop selective color preset:
15195 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15199 @anchor{separatefields}
15200 @section separatefields
15202 The @code{separatefields} takes a frame-based video input and splits
15203 each frame into its components fields, producing a new half height clip
15204 with twice the frame rate and twice the frame count.
15206 This filter use field-dominance information in frame to decide which
15207 of each pair of fields to place first in the output.
15208 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15210 @section setdar, setsar
15212 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15215 This is done by changing the specified Sample (aka Pixel) Aspect
15216 Ratio, according to the following equation:
15218 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15221 Keep in mind that the @code{setdar} filter does not modify the pixel
15222 dimensions of the video frame. Also, the display aspect ratio set by
15223 this filter may be changed by later filters in the filterchain,
15224 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15227 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15228 the filter output video.
15230 Note that as a consequence of the application of this filter, the
15231 output display aspect ratio will change according to the equation
15234 Keep in mind that the sample aspect ratio set by the @code{setsar}
15235 filter may be changed by later filters in the filterchain, e.g. if
15236 another "setsar" or a "setdar" filter is applied.
15238 It accepts the following parameters:
15241 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15242 Set the aspect ratio used by the filter.
15244 The parameter can be a floating point number string, an expression, or
15245 a string of the form @var{num}:@var{den}, where @var{num} and
15246 @var{den} are the numerator and denominator of the aspect ratio. If
15247 the parameter is not specified, it is assumed the value "0".
15248 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15252 Set the maximum integer value to use for expressing numerator and
15253 denominator when reducing the expressed aspect ratio to a rational.
15254 Default value is @code{100}.
15258 The parameter @var{sar} is an expression containing
15259 the following constants:
15263 These are approximated values for the mathematical constants e
15264 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15267 The input width and height.
15270 These are the same as @var{w} / @var{h}.
15273 The input sample aspect ratio.
15276 The input display aspect ratio. It is the same as
15277 (@var{w} / @var{h}) * @var{sar}.
15280 Horizontal and vertical chroma subsample values. For example, for the
15281 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15284 @subsection Examples
15289 To change the display aspect ratio to 16:9, specify one of the following:
15296 To change the sample aspect ratio to 10:11, specify:
15302 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15303 1000 in the aspect ratio reduction, use the command:
15305 setdar=ratio=16/9:max=1000
15313 Force field for the output video frame.
15315 The @code{setfield} filter marks the interlace type field for the
15316 output frames. It does not change the input frame, but only sets the
15317 corresponding property, which affects how the frame is treated by
15318 following filters (e.g. @code{fieldorder} or @code{yadif}).
15320 The filter accepts the following options:
15325 Available values are:
15329 Keep the same field property.
15332 Mark the frame as bottom-field-first.
15335 Mark the frame as top-field-first.
15338 Mark the frame as progressive.
15345 Force frame parameter for the output video frame.
15347 The @code{setparams} filter marks interlace and color range for the
15348 output frames. It does not change the input frame, but only sets the
15349 corresponding property, which affects how the frame is treated by
15354 Available values are:
15358 Keep the same field property (default).
15361 Mark the frame as bottom-field-first.
15364 Mark the frame as top-field-first.
15367 Mark the frame as progressive.
15371 Available values are:
15375 Keep the same color range property (default).
15377 @item unspecified, unknown
15378 Mark the frame as unspecified color range.
15380 @item limited, tv, mpeg
15381 Mark the frame as limited range.
15383 @item full, pc, jpeg
15384 Mark the frame as full range.
15387 @item color_primaries
15388 Set the color primaries.
15389 Available values are:
15393 Keep the same color primaries property (default).
15410 Set the color transfer.
15411 Available values are:
15415 Keep the same color trc property (default).
15437 Set the colorspace.
15438 Available values are:
15442 Keep the same colorspace property (default).
15455 @item chroma-derived-nc
15456 @item chroma-derived-c
15463 Show a line containing various information for each input video frame.
15464 The input video is not modified.
15466 This filter supports the following options:
15470 Calculate checksums of each plane. By default enabled.
15473 The shown line contains a sequence of key/value pairs of the form
15474 @var{key}:@var{value}.
15476 The following values are shown in the output:
15480 The (sequential) number of the input frame, starting from 0.
15483 The Presentation TimeStamp of the input frame, expressed as a number of
15484 time base units. The time base unit depends on the filter input pad.
15487 The Presentation TimeStamp of the input frame, expressed as a number of
15491 The position of the frame in the input stream, or -1 if this information is
15492 unavailable and/or meaningless (for example in case of synthetic video).
15495 The pixel format name.
15498 The sample aspect ratio of the input frame, expressed in the form
15499 @var{num}/@var{den}.
15502 The size of the input frame. For the syntax of this option, check the
15503 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15506 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15507 for bottom field first).
15510 This is 1 if the frame is a key frame, 0 otherwise.
15513 The picture type of the input frame ("I" for an I-frame, "P" for a
15514 P-frame, "B" for a B-frame, or "?" for an unknown type).
15515 Also refer to the documentation of the @code{AVPictureType} enum and of
15516 the @code{av_get_picture_type_char} function defined in
15517 @file{libavutil/avutil.h}.
15520 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15522 @item plane_checksum
15523 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15524 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15527 @section showpalette
15529 Displays the 256 colors palette of each frame. This filter is only relevant for
15530 @var{pal8} pixel format frames.
15532 It accepts the following option:
15536 Set the size of the box used to represent one palette color entry. Default is
15537 @code{30} (for a @code{30x30} pixel box).
15540 @section shuffleframes
15542 Reorder and/or duplicate and/or drop video frames.
15544 It accepts the following parameters:
15548 Set the destination indexes of input frames.
15549 This is space or '|' separated list of indexes that maps input frames to output
15550 frames. Number of indexes also sets maximal value that each index may have.
15551 '-1' index have special meaning and that is to drop frame.
15554 The first frame has the index 0. The default is to keep the input unchanged.
15556 @subsection Examples
15560 Swap second and third frame of every three frames of the input:
15562 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15566 Swap 10th and 1st frame of every ten frames of the input:
15568 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15572 @section shuffleplanes
15574 Reorder and/or duplicate video planes.
15576 It accepts the following parameters:
15581 The index of the input plane to be used as the first output plane.
15584 The index of the input plane to be used as the second output plane.
15587 The index of the input plane to be used as the third output plane.
15590 The index of the input plane to be used as the fourth output plane.
15594 The first plane has the index 0. The default is to keep the input unchanged.
15596 @subsection Examples
15600 Swap the second and third planes of the input:
15602 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15606 @anchor{signalstats}
15607 @section signalstats
15608 Evaluate various visual metrics that assist in determining issues associated
15609 with the digitization of analog video media.
15611 By default the filter will log these metadata values:
15615 Display the minimal Y value contained within the input frame. Expressed in
15619 Display the Y value at the 10% percentile within the input frame. Expressed in
15623 Display the average Y value within the input frame. Expressed in range of
15627 Display the Y value at the 90% percentile within the input frame. Expressed in
15631 Display the maximum Y value contained within the input frame. Expressed in
15635 Display the minimal U value contained within the input frame. Expressed in
15639 Display the U value at the 10% percentile within the input frame. Expressed in
15643 Display the average U value within the input frame. Expressed in range of
15647 Display the U value at the 90% percentile within the input frame. Expressed in
15651 Display the maximum U value contained within the input frame. Expressed in
15655 Display the minimal V value contained within the input frame. Expressed in
15659 Display the V value at the 10% percentile within the input frame. Expressed in
15663 Display the average V value within the input frame. Expressed in range of
15667 Display the V value at the 90% percentile within the input frame. Expressed in
15671 Display the maximum V value contained within the input frame. Expressed in
15675 Display the minimal saturation value contained within the input frame.
15676 Expressed in range of [0-~181.02].
15679 Display the saturation value at the 10% percentile within the input frame.
15680 Expressed in range of [0-~181.02].
15683 Display the average saturation value within the input frame. Expressed in range
15687 Display the saturation value at the 90% percentile within the input frame.
15688 Expressed in range of [0-~181.02].
15691 Display the maximum saturation value contained within the input frame.
15692 Expressed in range of [0-~181.02].
15695 Display the median value for hue within the input frame. Expressed in range of
15699 Display the average value for hue within the input frame. Expressed in range of
15703 Display the average of sample value difference between all values of the Y
15704 plane in the current frame and corresponding values of the previous input frame.
15705 Expressed in range of [0-255].
15708 Display the average of sample value difference between all values of the U
15709 plane in the current frame and corresponding values of the previous input frame.
15710 Expressed in range of [0-255].
15713 Display the average of sample value difference between all values of the V
15714 plane in the current frame and corresponding values of the previous input frame.
15715 Expressed in range of [0-255].
15718 Display bit depth of Y plane in current frame.
15719 Expressed in range of [0-16].
15722 Display bit depth of U plane in current frame.
15723 Expressed in range of [0-16].
15726 Display bit depth of V plane in current frame.
15727 Expressed in range of [0-16].
15730 The filter accepts the following options:
15736 @option{stat} specify an additional form of image analysis.
15737 @option{out} output video with the specified type of pixel highlighted.
15739 Both options accept the following values:
15743 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15744 unlike the neighboring pixels of the same field. Examples of temporal outliers
15745 include the results of video dropouts, head clogs, or tape tracking issues.
15748 Identify @var{vertical line repetition}. Vertical line repetition includes
15749 similar rows of pixels within a frame. In born-digital video vertical line
15750 repetition is common, but this pattern is uncommon in video digitized from an
15751 analog source. When it occurs in video that results from the digitization of an
15752 analog source it can indicate concealment from a dropout compensator.
15755 Identify pixels that fall outside of legal broadcast range.
15759 Set the highlight color for the @option{out} option. The default color is
15763 @subsection Examples
15767 Output data of various video metrics:
15769 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15773 Output specific data about the minimum and maximum values of the Y plane per frame:
15775 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15779 Playback video while highlighting pixels that are outside of broadcast range in red.
15781 ffplay example.mov -vf signalstats="out=brng:color=red"
15785 Playback video with signalstats metadata drawn over the frame.
15787 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15790 The contents of signalstat_drawtext.txt used in the command are:
15793 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15794 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15795 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15796 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15804 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15805 input. In this case the matching between the inputs can be calculated additionally.
15806 The filter always passes through the first input. The signature of each stream can
15807 be written into a file.
15809 It accepts the following options:
15813 Enable or disable the matching process.
15815 Available values are:
15819 Disable the calculation of a matching (default).
15821 Calculate the matching for the whole video and output whether the whole video
15822 matches or only parts.
15824 Calculate only until a matching is found or the video ends. Should be faster in
15829 Set the number of inputs. The option value must be a non negative integer.
15830 Default value is 1.
15833 Set the path to which the output is written. If there is more than one input,
15834 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15835 integer), that will be replaced with the input number. If no filename is
15836 specified, no output will be written. This is the default.
15839 Choose the output format.
15841 Available values are:
15845 Use the specified binary representation (default).
15847 Use the specified xml representation.
15851 Set threshold to detect one word as similar. The option value must be an integer
15852 greater than zero. The default value is 9000.
15855 Set threshold to detect all words as similar. The option value must be an integer
15856 greater than zero. The default value is 60000.
15859 Set threshold to detect frames as similar. The option value must be an integer
15860 greater than zero. The default value is 116.
15863 Set the minimum length of a sequence in frames to recognize it as matching
15864 sequence. The option value must be a non negative integer value.
15865 The default value is 0.
15868 Set the minimum relation, that matching frames to all frames must have.
15869 The option value must be a double value between 0 and 1. The default value is 0.5.
15872 @subsection Examples
15876 To calculate the signature of an input video and store it in signature.bin:
15878 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15882 To detect whether two videos match and store the signatures in XML format in
15883 signature0.xml and signature1.xml:
15885 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 -
15893 Blur the input video without impacting the outlines.
15895 It accepts the following options:
15898 @item luma_radius, lr
15899 Set the luma radius. The option value must be a float number in
15900 the range [0.1,5.0] that specifies the variance of the gaussian filter
15901 used to blur the image (slower if larger). Default value is 1.0.
15903 @item luma_strength, ls
15904 Set the luma strength. The option value must be a float number
15905 in the range [-1.0,1.0] that configures the blurring. A value included
15906 in [0.0,1.0] will blur the image whereas a value included in
15907 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15909 @item luma_threshold, lt
15910 Set the luma threshold used as a coefficient to determine
15911 whether a pixel should be blurred or not. The option value must be an
15912 integer in the range [-30,30]. A value of 0 will filter all the image,
15913 a value included in [0,30] will filter flat areas and a value included
15914 in [-30,0] will filter edges. Default value is 0.
15916 @item chroma_radius, cr
15917 Set the chroma radius. The option value must be a float number in
15918 the range [0.1,5.0] that specifies the variance of the gaussian filter
15919 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15921 @item chroma_strength, cs
15922 Set the chroma strength. The option value must be a float number
15923 in the range [-1.0,1.0] that configures the blurring. A value included
15924 in [0.0,1.0] will blur the image whereas a value included in
15925 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15927 @item chroma_threshold, ct
15928 Set the chroma threshold used as a coefficient to determine
15929 whether a pixel should be blurred or not. The option value must be an
15930 integer in the range [-30,30]. A value of 0 will filter all the image,
15931 a value included in [0,30] will filter flat areas and a value included
15932 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15935 If a chroma option is not explicitly set, the corresponding luma value
15940 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
15942 This filter takes in input two input videos, the first input is
15943 considered the "main" source and is passed unchanged to the
15944 output. The second input is used as a "reference" video for computing
15947 Both video inputs must have the same resolution and pixel format for
15948 this filter to work correctly. Also it assumes that both inputs
15949 have the same number of frames, which are compared one by one.
15951 The filter stores the calculated SSIM of each frame.
15953 The description of the accepted parameters follows.
15956 @item stats_file, f
15957 If specified the filter will use the named file to save the SSIM of
15958 each individual frame. When filename equals "-" the data is sent to
15962 The file printed if @var{stats_file} is selected, contains a sequence of
15963 key/value pairs of the form @var{key}:@var{value} for each compared
15966 A description of each shown parameter follows:
15970 sequential number of the input frame, starting from 1
15972 @item Y, U, V, R, G, B
15973 SSIM of the compared frames for the component specified by the suffix.
15976 SSIM of the compared frames for the whole frame.
15979 Same as above but in dB representation.
15982 This filter also supports the @ref{framesync} options.
15986 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15987 [main][ref] ssim="stats_file=stats.log" [out]
15990 On this example the input file being processed is compared with the
15991 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
15992 is stored in @file{stats.log}.
15994 Another example with both psnr and ssim at same time:
15996 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
16001 Convert between different stereoscopic image formats.
16003 The filters accept the following options:
16007 Set stereoscopic image format of input.
16009 Available values for input image formats are:
16012 side by side parallel (left eye left, right eye right)
16015 side by side crosseye (right eye left, left eye right)
16018 side by side parallel with half width resolution
16019 (left eye left, right eye right)
16022 side by side crosseye with half width resolution
16023 (right eye left, left eye right)
16026 above-below (left eye above, right eye below)
16029 above-below (right eye above, left eye below)
16032 above-below with half height resolution
16033 (left eye above, right eye below)
16036 above-below with half height resolution
16037 (right eye above, left eye below)
16040 alternating frames (left eye first, right eye second)
16043 alternating frames (right eye first, left eye second)
16046 interleaved rows (left eye has top row, right eye starts on next row)
16049 interleaved rows (right eye has top row, left eye starts on next row)
16052 interleaved columns, left eye first
16055 interleaved columns, right eye first
16057 Default value is @samp{sbsl}.
16061 Set stereoscopic image format of output.
16065 side by side parallel (left eye left, right eye right)
16068 side by side crosseye (right eye left, left eye right)
16071 side by side parallel with half width resolution
16072 (left eye left, right eye right)
16075 side by side crosseye with half width resolution
16076 (right eye left, left eye right)
16079 above-below (left eye above, right eye below)
16082 above-below (right eye above, left eye below)
16085 above-below with half height resolution
16086 (left eye above, right eye below)
16089 above-below with half height resolution
16090 (right eye above, left eye below)
16093 alternating frames (left eye first, right eye second)
16096 alternating frames (right eye first, left eye second)
16099 interleaved rows (left eye has top row, right eye starts on next row)
16102 interleaved rows (right eye has top row, left eye starts on next row)
16105 anaglyph red/blue gray
16106 (red filter on left eye, blue filter on right eye)
16109 anaglyph red/green gray
16110 (red filter on left eye, green filter on right eye)
16113 anaglyph red/cyan gray
16114 (red filter on left eye, cyan filter on right eye)
16117 anaglyph red/cyan half colored
16118 (red filter on left eye, cyan filter on right eye)
16121 anaglyph red/cyan color
16122 (red filter on left eye, cyan filter on right eye)
16125 anaglyph red/cyan color optimized with the least squares projection of dubois
16126 (red filter on left eye, cyan filter on right eye)
16129 anaglyph green/magenta gray
16130 (green filter on left eye, magenta filter on right eye)
16133 anaglyph green/magenta half colored
16134 (green filter on left eye, magenta filter on right eye)
16137 anaglyph green/magenta colored
16138 (green filter on left eye, magenta filter on right eye)
16141 anaglyph green/magenta color optimized with the least squares projection of dubois
16142 (green filter on left eye, magenta filter on right eye)
16145 anaglyph yellow/blue gray
16146 (yellow filter on left eye, blue filter on right eye)
16149 anaglyph yellow/blue half colored
16150 (yellow filter on left eye, blue filter on right eye)
16153 anaglyph yellow/blue colored
16154 (yellow filter on left eye, blue filter on right eye)
16157 anaglyph yellow/blue color optimized with the least squares projection of dubois
16158 (yellow filter on left eye, blue filter on right eye)
16161 mono output (left eye only)
16164 mono output (right eye only)
16167 checkerboard, left eye first
16170 checkerboard, right eye first
16173 interleaved columns, left eye first
16176 interleaved columns, right eye first
16182 Default value is @samp{arcd}.
16185 @subsection Examples
16189 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16195 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16201 @section streamselect, astreamselect
16202 Select video or audio streams.
16204 The filter accepts the following options:
16208 Set number of inputs. Default is 2.
16211 Set input indexes to remap to outputs.
16214 @subsection Commands
16216 The @code{streamselect} and @code{astreamselect} filter supports the following
16221 Set input indexes to remap to outputs.
16224 @subsection Examples
16228 Select first 5 seconds 1st stream and rest of time 2nd stream:
16230 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16234 Same as above, but for audio:
16236 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16241 Apply sobel operator to input video stream.
16243 The filter accepts the following option:
16247 Set which planes will be processed, unprocessed planes will be copied.
16248 By default value 0xf, all planes will be processed.
16251 Set value which will be multiplied with filtered result.
16254 Set value which will be added to filtered result.
16260 Apply a simple postprocessing filter that compresses and decompresses the image
16261 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16262 and average the results.
16264 The filter accepts the following options:
16268 Set quality. This option defines the number of levels for averaging. It accepts
16269 an integer in the range 0-6. If set to @code{0}, the filter will have no
16270 effect. A value of @code{6} means the higher quality. For each increment of
16271 that value the speed drops by a factor of approximately 2. Default value is
16275 Force a constant quantization parameter. If not set, the filter will use the QP
16276 from the video stream (if available).
16279 Set thresholding mode. Available modes are:
16283 Set hard thresholding (default).
16285 Set soft thresholding (better de-ringing effect, but likely blurrier).
16288 @item use_bframe_qp
16289 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16290 option may cause flicker since the B-Frames have often larger QP. Default is
16291 @code{0} (not enabled).
16296 Scale the input by applying one of the super-resolution methods based on
16297 convolutional neural networks. Supported models:
16301 Super-Resolution Convolutional Neural Network model (SRCNN).
16302 See @url{https://arxiv.org/abs/1501.00092}.
16305 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16306 See @url{https://arxiv.org/abs/1609.05158}.
16309 Training scripts as well as scripts for model generation are provided in
16310 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16312 The filter accepts the following options:
16316 Specify which DNN backend to use for model loading and execution. This option accepts
16317 the following values:
16321 Native implementation of DNN loading and execution.
16324 TensorFlow backend. To enable this backend you
16325 need to install the TensorFlow for C library (see
16326 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16327 @code{--enable-libtensorflow}
16330 Default value is @samp{native}.
16333 Set path to model file specifying network architecture and its parameters.
16334 Note that different backends use different file formats. TensorFlow backend
16335 can load files for both formats, while native backend can load files for only
16339 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16340 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16341 input upscaled using bicubic upscaling with proper scale factor.
16347 Draw subtitles on top of input video using the libass library.
16349 To enable compilation of this filter you need to configure FFmpeg with
16350 @code{--enable-libass}. This filter also requires a build with libavcodec and
16351 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16352 Alpha) subtitles format.
16354 The filter accepts the following options:
16358 Set the filename of the subtitle file to read. It must be specified.
16360 @item original_size
16361 Specify the size of the original video, the video for which the ASS file
16362 was composed. For the syntax of this option, check the
16363 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16364 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16365 correctly scale the fonts if the aspect ratio has been changed.
16368 Set a directory path containing fonts that can be used by the filter.
16369 These fonts will be used in addition to whatever the font provider uses.
16372 Process alpha channel, by default alpha channel is untouched.
16375 Set subtitles input character encoding. @code{subtitles} filter only. Only
16376 useful if not UTF-8.
16378 @item stream_index, si
16379 Set subtitles stream index. @code{subtitles} filter only.
16382 Override default style or script info parameters of the subtitles. It accepts a
16383 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16386 If the first key is not specified, it is assumed that the first value
16387 specifies the @option{filename}.
16389 For example, to render the file @file{sub.srt} on top of the input
16390 video, use the command:
16395 which is equivalent to:
16397 subtitles=filename=sub.srt
16400 To render the default subtitles stream from file @file{video.mkv}, use:
16402 subtitles=video.mkv
16405 To render the second subtitles stream from that file, use:
16407 subtitles=video.mkv:si=1
16410 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16411 @code{DejaVu Serif}, use:
16413 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16416 @section super2xsai
16418 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16419 Interpolate) pixel art scaling algorithm.
16421 Useful for enlarging pixel art images without reducing sharpness.
16425 Swap two rectangular objects in video.
16427 This filter accepts the following options:
16437 Set 1st rect x coordinate.
16440 Set 1st rect y coordinate.
16443 Set 2nd rect x coordinate.
16446 Set 2nd rect y coordinate.
16448 All expressions are evaluated once for each frame.
16451 The all options are expressions containing the following constants:
16456 The input width and height.
16459 same as @var{w} / @var{h}
16462 input sample aspect ratio
16465 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16468 The number of the input frame, starting from 0.
16471 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16474 the position in the file of the input frame, NAN if unknown
16482 Apply telecine process to the video.
16484 This filter accepts the following options:
16493 The default value is @code{top}.
16497 A string of numbers representing the pulldown pattern you wish to apply.
16498 The default value is @code{23}.
16502 Some typical patterns:
16507 24p: 2332 (preferred)
16514 24p: 222222222223 ("Euro pulldown")
16521 Apply threshold effect to video stream.
16523 This filter needs four video streams to perform thresholding.
16524 First stream is stream we are filtering.
16525 Second stream is holding threshold values, third stream is holding min values,
16526 and last, fourth stream is holding max values.
16528 The filter accepts the following option:
16532 Set which planes will be processed, unprocessed planes will be copied.
16533 By default value 0xf, all planes will be processed.
16536 For example if first stream pixel's component value is less then threshold value
16537 of pixel component from 2nd threshold stream, third stream value will picked,
16538 otherwise fourth stream pixel component value will be picked.
16540 Using color source filter one can perform various types of thresholding:
16542 @subsection Examples
16546 Binary threshold, using gray color as threshold:
16548 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16552 Inverted binary threshold, using gray color as threshold:
16554 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16558 Truncate binary threshold, using gray color as threshold:
16560 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16564 Threshold to zero, using gray color as threshold:
16566 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16570 Inverted threshold to zero, using gray color as threshold:
16572 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16577 Select the most representative frame in a given sequence of consecutive frames.
16579 The filter accepts the following options:
16583 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16584 will pick one of them, and then handle the next batch of @var{n} frames until
16585 the end. Default is @code{100}.
16588 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16589 value will result in a higher memory usage, so a high value is not recommended.
16591 @subsection Examples
16595 Extract one picture each 50 frames:
16601 Complete example of a thumbnail creation with @command{ffmpeg}:
16603 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16609 Tile several successive frames together.
16611 The filter accepts the following options:
16616 Set the grid size (i.e. the number of lines and columns). For the syntax of
16617 this option, check the
16618 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16621 Set the maximum number of frames to render in the given area. It must be less
16622 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16623 the area will be used.
16626 Set the outer border margin in pixels.
16629 Set the inner border thickness (i.e. the number of pixels between frames). For
16630 more advanced padding options (such as having different values for the edges),
16631 refer to the pad video filter.
16634 Specify the color of the unused area. For the syntax of this option, check the
16635 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16636 The default value of @var{color} is "black".
16639 Set the number of frames to overlap when tiling several successive frames together.
16640 The value must be between @code{0} and @var{nb_frames - 1}.
16643 Set the number of frames to initially be empty before displaying first output frame.
16644 This controls how soon will one get first output frame.
16645 The value must be between @code{0} and @var{nb_frames - 1}.
16648 @subsection Examples
16652 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16654 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16656 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16657 duplicating each output frame to accommodate the originally detected frame
16661 Display @code{5} pictures in an area of @code{3x2} frames,
16662 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16663 mixed flat and named options:
16665 tile=3x2:nb_frames=5:padding=7:margin=2
16669 @section tinterlace
16671 Perform various types of temporal field interlacing.
16673 Frames are counted starting from 1, so the first input frame is
16676 The filter accepts the following options:
16681 Specify the mode of the interlacing. This option can also be specified
16682 as a value alone. See below for a list of values for this option.
16684 Available values are:
16688 Move odd frames into the upper field, even into the lower field,
16689 generating a double height frame at half frame rate.
16693 Frame 1 Frame 2 Frame 3 Frame 4
16695 11111 22222 33333 44444
16696 11111 22222 33333 44444
16697 11111 22222 33333 44444
16698 11111 22222 33333 44444
16712 Only output odd frames, even frames are dropped, generating a frame with
16713 unchanged height at half frame rate.
16718 Frame 1 Frame 2 Frame 3 Frame 4
16720 11111 22222 33333 44444
16721 11111 22222 33333 44444
16722 11111 22222 33333 44444
16723 11111 22222 33333 44444
16733 Only output even frames, odd frames are dropped, generating a frame with
16734 unchanged height at half frame rate.
16739 Frame 1 Frame 2 Frame 3 Frame 4
16741 11111 22222 33333 44444
16742 11111 22222 33333 44444
16743 11111 22222 33333 44444
16744 11111 22222 33333 44444
16754 Expand each frame to full height, but pad alternate lines with black,
16755 generating a frame with double height at the same input frame rate.
16760 Frame 1 Frame 2 Frame 3 Frame 4
16762 11111 22222 33333 44444
16763 11111 22222 33333 44444
16764 11111 22222 33333 44444
16765 11111 22222 33333 44444
16768 11111 ..... 33333 .....
16769 ..... 22222 ..... 44444
16770 11111 ..... 33333 .....
16771 ..... 22222 ..... 44444
16772 11111 ..... 33333 .....
16773 ..... 22222 ..... 44444
16774 11111 ..... 33333 .....
16775 ..... 22222 ..... 44444
16779 @item interleave_top, 4
16780 Interleave the upper field from odd frames with the lower field from
16781 even frames, generating a frame with unchanged height at half frame rate.
16786 Frame 1 Frame 2 Frame 3 Frame 4
16788 11111<- 22222 33333<- 44444
16789 11111 22222<- 33333 44444<-
16790 11111<- 22222 33333<- 44444
16791 11111 22222<- 33333 44444<-
16801 @item interleave_bottom, 5
16802 Interleave the lower field from odd frames with the upper field from
16803 even frames, generating a frame with unchanged height at half frame rate.
16808 Frame 1 Frame 2 Frame 3 Frame 4
16810 11111 22222<- 33333 44444<-
16811 11111<- 22222 33333<- 44444
16812 11111 22222<- 33333 44444<-
16813 11111<- 22222 33333<- 44444
16823 @item interlacex2, 6
16824 Double frame rate with unchanged height. Frames are inserted each
16825 containing the second temporal field from the previous input frame and
16826 the first temporal field from the next input frame. This mode relies on
16827 the top_field_first flag. Useful for interlaced video displays with no
16828 field synchronisation.
16833 Frame 1 Frame 2 Frame 3 Frame 4
16835 11111 22222 33333 44444
16836 11111 22222 33333 44444
16837 11111 22222 33333 44444
16838 11111 22222 33333 44444
16841 11111 22222 22222 33333 33333 44444 44444
16842 11111 11111 22222 22222 33333 33333 44444
16843 11111 22222 22222 33333 33333 44444 44444
16844 11111 11111 22222 22222 33333 33333 44444
16849 Move odd frames into the upper field, even into the lower field,
16850 generating a double height frame at same frame rate.
16855 Frame 1 Frame 2 Frame 3 Frame 4
16857 11111 22222 33333 44444
16858 11111 22222 33333 44444
16859 11111 22222 33333 44444
16860 11111 22222 33333 44444
16863 11111 33333 33333 55555
16864 22222 22222 44444 44444
16865 11111 33333 33333 55555
16866 22222 22222 44444 44444
16867 11111 33333 33333 55555
16868 22222 22222 44444 44444
16869 11111 33333 33333 55555
16870 22222 22222 44444 44444
16875 Numeric values are deprecated but are accepted for backward
16876 compatibility reasons.
16878 Default mode is @code{merge}.
16881 Specify flags influencing the filter process.
16883 Available value for @var{flags} is:
16886 @item low_pass_filter, vlfp
16887 Enable linear vertical low-pass filtering in the filter.
16888 Vertical low-pass filtering is required when creating an interlaced
16889 destination from a progressive source which contains high-frequency
16890 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16893 @item complex_filter, cvlfp
16894 Enable complex vertical low-pass filtering.
16895 This will slightly less reduce interlace 'twitter' and Moire
16896 patterning but better retain detail and subjective sharpness impression.
16900 Vertical low-pass filtering can only be enabled for @option{mode}
16901 @var{interleave_top} and @var{interleave_bottom}.
16907 Mix successive video frames.
16909 A description of the accepted options follows.
16913 The number of successive frames to mix. If unspecified, it defaults to 3.
16916 Specify weight of each input video frame.
16917 Each weight is separated by space. If number of weights is smaller than
16918 number of @var{frames} last specified weight will be used for all remaining
16922 Specify scale, if it is set it will be multiplied with sum
16923 of each weight multiplied with pixel values to give final destination
16924 pixel value. By default @var{scale} is auto scaled to sum of weights.
16927 @subsection Examples
16931 Average 7 successive frames:
16933 tmix=frames=7:weights="1 1 1 1 1 1 1"
16937 Apply simple temporal convolution:
16939 tmix=frames=3:weights="-1 3 -1"
16943 Similar as above but only showing temporal differences:
16945 tmix=frames=3:weights="-1 2 -1":scale=1
16951 Tone map colors from different dynamic ranges.
16953 This filter expects data in single precision floating point, as it needs to
16954 operate on (and can output) out-of-range values. Another filter, such as
16955 @ref{zscale}, is needed to convert the resulting frame to a usable format.
16957 The tonemapping algorithms implemented only work on linear light, so input
16958 data should be linearized beforehand (and possibly correctly tagged).
16961 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
16964 @subsection Options
16965 The filter accepts the following options.
16969 Set the tone map algorithm to use.
16971 Possible values are:
16974 Do not apply any tone map, only desaturate overbright pixels.
16977 Hard-clip any out-of-range values. Use it for perfect color accuracy for
16978 in-range values, while distorting out-of-range values.
16981 Stretch the entire reference gamut to a linear multiple of the display.
16984 Fit a logarithmic transfer between the tone curves.
16987 Preserve overall image brightness with a simple curve, using nonlinear
16988 contrast, which results in flattening details and degrading color accuracy.
16991 Preserve both dark and bright details better than @var{reinhard}, at the cost
16992 of slightly darkening everything. Use it when detail preservation is more
16993 important than color and brightness accuracy.
16996 Smoothly map out-of-range values, while retaining contrast and colors for
16997 in-range material as much as possible. Use it when color accuracy is more
16998 important than detail preservation.
17004 Tune the tone mapping algorithm.
17006 This affects the following algorithms:
17012 Specifies the scale factor to use while stretching.
17016 Specifies the exponent of the function.
17020 Specify an extra linear coefficient to multiply into the signal before clipping.
17024 Specify the local contrast coefficient at the display peak.
17025 Default to 0.5, which means that in-gamut values will be about half as bright
17032 Specify the transition point from linear to mobius transform. Every value
17033 below this point is guaranteed to be mapped 1:1. The higher the value, the
17034 more accurate the result will be, at the cost of losing bright details.
17035 Default to 0.3, which due to the steep initial slope still preserves in-range
17036 colors fairly accurately.
17040 Apply desaturation for highlights that exceed this level of brightness. The
17041 higher the parameter, the more color information will be preserved. This
17042 setting helps prevent unnaturally blown-out colors for super-highlights, by
17043 (smoothly) turning into white instead. This makes images feel more natural,
17044 at the cost of reducing information about out-of-range colors.
17046 The default of 2.0 is somewhat conservative and will mostly just apply to
17047 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17049 This option works only if the input frame has a supported color tag.
17052 Override signal/nominal/reference peak with this value. Useful when the
17053 embedded peak information in display metadata is not reliable or when tone
17054 mapping from a lower range to a higher range.
17059 Temporarily pad video frames.
17061 The filter accepts the following options:
17065 Specify number of delay frames before input video stream.
17068 Specify number of padding frames after input video stream.
17069 Set to -1 to pad indefinitely.
17072 Set kind of frames added to beginning of stream.
17073 Can be either @var{add} or @var{clone}.
17074 With @var{add} frames of solid-color are added.
17075 With @var{clone} frames are clones of first frame.
17078 Set kind of frames added to end of stream.
17079 Can be either @var{add} or @var{clone}.
17080 With @var{add} frames of solid-color are added.
17081 With @var{clone} frames are clones of last frame.
17083 @item start_duration, stop_duration
17084 Specify the duration of the start/stop delay. See
17085 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17086 for the accepted syntax.
17087 These options override @var{start} and @var{stop}.
17090 Specify the color of the padded area. For the syntax of this option,
17091 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17092 manual,ffmpeg-utils}.
17094 The default value of @var{color} is "black".
17100 Transpose rows with columns in the input video and optionally flip it.
17102 It accepts the following parameters:
17107 Specify the transposition direction.
17109 Can assume the following values:
17111 @item 0, 4, cclock_flip
17112 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17120 Rotate by 90 degrees clockwise, that is:
17128 Rotate by 90 degrees counterclockwise, that is:
17135 @item 3, 7, clock_flip
17136 Rotate by 90 degrees clockwise and vertically flip, that is:
17144 For values between 4-7, the transposition is only done if the input
17145 video geometry is portrait and not landscape. These values are
17146 deprecated, the @code{passthrough} option should be used instead.
17148 Numerical values are deprecated, and should be dropped in favor of
17149 symbolic constants.
17152 Do not apply the transposition if the input geometry matches the one
17153 specified by the specified value. It accepts the following values:
17156 Always apply transposition.
17158 Preserve portrait geometry (when @var{height} >= @var{width}).
17160 Preserve landscape geometry (when @var{width} >= @var{height}).
17163 Default value is @code{none}.
17166 For example to rotate by 90 degrees clockwise and preserve portrait
17169 transpose=dir=1:passthrough=portrait
17172 The command above can also be specified as:
17174 transpose=1:portrait
17177 @section transpose_npp
17179 Transpose rows with columns in the input video and optionally flip it.
17180 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17182 It accepts the following parameters:
17187 Specify the transposition direction.
17189 Can assume the following values:
17192 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17195 Rotate by 90 degrees clockwise.
17198 Rotate by 90 degrees counterclockwise.
17201 Rotate by 90 degrees clockwise and vertically flip.
17205 Do not apply the transposition if the input geometry matches the one
17206 specified by the specified value. It accepts the following values:
17209 Always apply transposition. (default)
17211 Preserve portrait geometry (when @var{height} >= @var{width}).
17213 Preserve landscape geometry (when @var{width} >= @var{height}).
17219 Trim the input so that the output contains one continuous subpart of the input.
17221 It accepts the following parameters:
17224 Specify the time of the start of the kept section, i.e. the frame with the
17225 timestamp @var{start} will be the first frame in the output.
17228 Specify the time of the first frame that will be dropped, i.e. the frame
17229 immediately preceding the one with the timestamp @var{end} will be the last
17230 frame in the output.
17233 This is the same as @var{start}, except this option sets the start timestamp
17234 in timebase units instead of seconds.
17237 This is the same as @var{end}, except this option sets the end timestamp
17238 in timebase units instead of seconds.
17241 The maximum duration of the output in seconds.
17244 The number of the first frame that should be passed to the output.
17247 The number of the first frame that should be dropped.
17250 @option{start}, @option{end}, and @option{duration} are expressed as time
17251 duration specifications; see
17252 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17253 for the accepted syntax.
17255 Note that the first two sets of the start/end options and the @option{duration}
17256 option look at the frame timestamp, while the _frame variants simply count the
17257 frames that pass through the filter. Also note that this filter does not modify
17258 the timestamps. If you wish for the output timestamps to start at zero, insert a
17259 setpts filter after the trim filter.
17261 If multiple start or end options are set, this filter tries to be greedy and
17262 keep all the frames that match at least one of the specified constraints. To keep
17263 only the part that matches all the constraints at once, chain multiple trim
17266 The defaults are such that all the input is kept. So it is possible to set e.g.
17267 just the end values to keep everything before the specified time.
17272 Drop everything except the second minute of input:
17274 ffmpeg -i INPUT -vf trim=60:120
17278 Keep only the first second:
17280 ffmpeg -i INPUT -vf trim=duration=1
17285 @section unpremultiply
17286 Apply alpha unpremultiply effect to input video stream using first plane
17287 of second stream as alpha.
17289 Both streams must have same dimensions and same pixel format.
17291 The filter accepts the following option:
17295 Set which planes will be processed, unprocessed planes will be copied.
17296 By default value 0xf, all planes will be processed.
17298 If the format has 1 or 2 components, then luma is bit 0.
17299 If the format has 3 or 4 components:
17300 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17301 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17302 If present, the alpha channel is always the last bit.
17305 Do not require 2nd input for processing, instead use alpha plane from input stream.
17311 Sharpen or blur the input video.
17313 It accepts the following parameters:
17316 @item luma_msize_x, lx
17317 Set the luma matrix horizontal size. It must be an odd integer between
17318 3 and 23. The default value is 5.
17320 @item luma_msize_y, ly
17321 Set the luma matrix vertical size. It must be an odd integer between 3
17322 and 23. The default value is 5.
17324 @item luma_amount, la
17325 Set the luma effect strength. It must be a floating point number, reasonable
17326 values lay between -1.5 and 1.5.
17328 Negative values will blur the input video, while positive values will
17329 sharpen it, a value of zero will disable the effect.
17331 Default value is 1.0.
17333 @item chroma_msize_x, cx
17334 Set the chroma matrix horizontal size. It must be an odd integer
17335 between 3 and 23. The default value is 5.
17337 @item chroma_msize_y, cy
17338 Set the chroma matrix vertical size. It must be an odd integer
17339 between 3 and 23. The default value is 5.
17341 @item chroma_amount, ca
17342 Set the chroma effect strength. It must be a floating point number, reasonable
17343 values lay between -1.5 and 1.5.
17345 Negative values will blur the input video, while positive values will
17346 sharpen it, a value of zero will disable the effect.
17348 Default value is 0.0.
17352 All parameters are optional and default to the equivalent of the
17353 string '5:5:1.0:5:5:0.0'.
17355 @subsection Examples
17359 Apply strong luma sharpen effect:
17361 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17365 Apply a strong blur of both luma and chroma parameters:
17367 unsharp=7:7:-2:7:7:-2
17373 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17374 the image at several (or - in the case of @option{quality} level @code{8} - all)
17375 shifts and average the results.
17377 The way this differs from the behavior of spp is that uspp actually encodes &
17378 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17379 DCT similar to MJPEG.
17381 The filter accepts the following options:
17385 Set quality. This option defines the number of levels for averaging. It accepts
17386 an integer in the range 0-8. If set to @code{0}, the filter will have no
17387 effect. A value of @code{8} means the higher quality. For each increment of
17388 that value the speed drops by a factor of approximately 2. Default value is
17392 Force a constant quantization parameter. If not set, the filter will use the QP
17393 from the video stream (if available).
17396 @section vaguedenoiser
17398 Apply a wavelet based denoiser.
17400 It transforms each frame from the video input into the wavelet domain,
17401 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17402 the obtained coefficients. It does an inverse wavelet transform after.
17403 Due to wavelet properties, it should give a nice smoothed result, and
17404 reduced noise, without blurring picture features.
17406 This filter accepts the following options:
17410 The filtering strength. The higher, the more filtered the video will be.
17411 Hard thresholding can use a higher threshold than soft thresholding
17412 before the video looks overfiltered. Default value is 2.
17415 The filtering method the filter will use.
17417 It accepts the following values:
17420 All values under the threshold will be zeroed.
17423 All values under the threshold will be zeroed. All values above will be
17424 reduced by the threshold.
17427 Scales or nullifies coefficients - intermediary between (more) soft and
17428 (less) hard thresholding.
17431 Default is garrote.
17434 Number of times, the wavelet will decompose the picture. Picture can't
17435 be decomposed beyond a particular point (typically, 8 for a 640x480
17436 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17439 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17442 A list of the planes to process. By default all planes are processed.
17445 @section vectorscope
17447 Display 2 color component values in the two dimensional graph (which is called
17450 This filter accepts the following options:
17454 Set vectorscope mode.
17456 It accepts the following values:
17459 Gray values are displayed on graph, higher brightness means more pixels have
17460 same component color value on location in graph. This is the default mode.
17463 Gray values are displayed on graph. Surrounding pixels values which are not
17464 present in video frame are drawn in gradient of 2 color components which are
17465 set by option @code{x} and @code{y}. The 3rd color component is static.
17468 Actual color components values present in video frame are displayed on graph.
17471 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17472 on graph increases value of another color component, which is luminance by
17473 default values of @code{x} and @code{y}.
17476 Actual colors present in video frame are displayed on graph. If two different
17477 colors map to same position on graph then color with higher value of component
17478 not present in graph is picked.
17481 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17482 component picked from radial gradient.
17486 Set which color component will be represented on X-axis. Default is @code{1}.
17489 Set which color component will be represented on Y-axis. Default is @code{2}.
17492 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17493 of color component which represents frequency of (X, Y) location in graph.
17498 No envelope, this is default.
17501 Instant envelope, even darkest single pixel will be clearly highlighted.
17504 Hold maximum and minimum values presented in graph over time. This way you
17505 can still spot out of range values without constantly looking at vectorscope.
17508 Peak and instant envelope combined together.
17512 Set what kind of graticule to draw.
17520 Set graticule opacity.
17523 Set graticule flags.
17527 Draw graticule for white point.
17530 Draw graticule for black point.
17533 Draw color points short names.
17537 Set background opacity.
17539 @item lthreshold, l
17540 Set low threshold for color component not represented on X or Y axis.
17541 Values lower than this value will be ignored. Default is 0.
17542 Note this value is multiplied with actual max possible value one pixel component
17543 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17546 @item hthreshold, h
17547 Set high threshold for color component not represented on X or Y axis.
17548 Values higher than this value will be ignored. Default is 1.
17549 Note this value is multiplied with actual max possible value one pixel component
17550 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17551 is 0.9 * 255 = 230.
17553 @item colorspace, c
17554 Set what kind of colorspace to use when drawing graticule.
17563 @anchor{vidstabdetect}
17564 @section vidstabdetect
17566 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17567 @ref{vidstabtransform} for pass 2.
17569 This filter generates a file with relative translation and rotation
17570 transform information about subsequent frames, which is then used by
17571 the @ref{vidstabtransform} filter.
17573 To enable compilation of this filter you need to configure FFmpeg with
17574 @code{--enable-libvidstab}.
17576 This filter accepts the following options:
17580 Set the path to the file used to write the transforms information.
17581 Default value is @file{transforms.trf}.
17584 Set how shaky the video is and how quick the camera is. It accepts an
17585 integer in the range 1-10, a value of 1 means little shakiness, a
17586 value of 10 means strong shakiness. Default value is 5.
17589 Set the accuracy of the detection process. It must be a value in the
17590 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17591 accuracy. Default value is 15.
17594 Set stepsize of the search process. The region around minimum is
17595 scanned with 1 pixel resolution. Default value is 6.
17598 Set minimum contrast. Below this value a local measurement field is
17599 discarded. Must be a floating point value in the range 0-1. Default
17603 Set reference frame number for tripod mode.
17605 If enabled, the motion of the frames is compared to a reference frame
17606 in the filtered stream, identified by the specified number. The idea
17607 is to compensate all movements in a more-or-less static scene and keep
17608 the camera view absolutely still.
17610 If set to 0, it is disabled. The frames are counted starting from 1.
17613 Show fields and transforms in the resulting frames. It accepts an
17614 integer in the range 0-2. Default value is 0, which disables any
17618 @subsection Examples
17622 Use default values:
17628 Analyze strongly shaky movie and put the results in file
17629 @file{mytransforms.trf}:
17631 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17635 Visualize the result of internal transformations in the resulting
17638 vidstabdetect=show=1
17642 Analyze a video with medium shakiness using @command{ffmpeg}:
17644 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17648 @anchor{vidstabtransform}
17649 @section vidstabtransform
17651 Video stabilization/deshaking: pass 2 of 2,
17652 see @ref{vidstabdetect} for pass 1.
17654 Read a file with transform information for each frame and
17655 apply/compensate them. Together with the @ref{vidstabdetect}
17656 filter this can be used to deshake videos. See also
17657 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17658 the @ref{unsharp} filter, see below.
17660 To enable compilation of this filter you need to configure FFmpeg with
17661 @code{--enable-libvidstab}.
17663 @subsection Options
17667 Set path to the file used to read the transforms. Default value is
17668 @file{transforms.trf}.
17671 Set the number of frames (value*2 + 1) used for lowpass filtering the
17672 camera movements. Default value is 10.
17674 For example a number of 10 means that 21 frames are used (10 in the
17675 past and 10 in the future) to smoothen the motion in the video. A
17676 larger value leads to a smoother video, but limits the acceleration of
17677 the camera (pan/tilt movements). 0 is a special case where a static
17678 camera is simulated.
17681 Set the camera path optimization algorithm.
17683 Accepted values are:
17686 gaussian kernel low-pass filter on camera motion (default)
17688 averaging on transformations
17692 Set maximal number of pixels to translate frames. Default value is -1,
17696 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17697 value is -1, meaning no limit.
17700 Specify how to deal with borders that may be visible due to movement
17703 Available values are:
17706 keep image information from previous frame (default)
17708 fill the border black
17712 Invert transforms if set to 1. Default value is 0.
17715 Consider transforms as relative to previous frame if set to 1,
17716 absolute if set to 0. Default value is 0.
17719 Set percentage to zoom. A positive value will result in a zoom-in
17720 effect, a negative value in a zoom-out effect. Default value is 0 (no
17724 Set optimal zooming to avoid borders.
17726 Accepted values are:
17731 optimal static zoom value is determined (only very strong movements
17732 will lead to visible borders) (default)
17734 optimal adaptive zoom value is determined (no borders will be
17735 visible), see @option{zoomspeed}
17738 Note that the value given at zoom is added to the one calculated here.
17741 Set percent to zoom maximally each frame (enabled when
17742 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17746 Specify type of interpolation.
17748 Available values are:
17753 linear only horizontal
17755 linear in both directions (default)
17757 cubic in both directions (slow)
17761 Enable virtual tripod mode if set to 1, which is equivalent to
17762 @code{relative=0:smoothing=0}. Default value is 0.
17764 Use also @code{tripod} option of @ref{vidstabdetect}.
17767 Increase log verbosity if set to 1. Also the detected global motions
17768 are written to the temporary file @file{global_motions.trf}. Default
17772 @subsection Examples
17776 Use @command{ffmpeg} for a typical stabilization with default values:
17778 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17781 Note the use of the @ref{unsharp} filter which is always recommended.
17784 Zoom in a bit more and load transform data from a given file:
17786 vidstabtransform=zoom=5:input="mytransforms.trf"
17790 Smoothen the video even more:
17792 vidstabtransform=smoothing=30
17798 Flip the input video vertically.
17800 For example, to vertically flip a video with @command{ffmpeg}:
17802 ffmpeg -i in.avi -vf "vflip" out.avi
17807 Detect variable frame rate video.
17809 This filter tries to detect if the input is variable or constant frame rate.
17811 At end it will output number of frames detected as having variable delta pts,
17812 and ones with constant delta pts.
17813 If there was frames with variable delta, than it will also show min and max delta
17818 Boost or alter saturation.
17820 The filter accepts the following options:
17823 Set strength of boost if positive value or strength of alter if negative value.
17824 Default is 0. Allowed range is from -2 to 2.
17827 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17830 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17833 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17836 Set the red luma coefficient.
17839 Set the green luma coefficient.
17842 Set the blue luma coefficient.
17848 Make or reverse a natural vignetting effect.
17850 The filter accepts the following options:
17854 Set lens angle expression as a number of radians.
17856 The value is clipped in the @code{[0,PI/2]} range.
17858 Default value: @code{"PI/5"}
17862 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17866 Set forward/backward mode.
17868 Available modes are:
17871 The larger the distance from the central point, the darker the image becomes.
17874 The larger the distance from the central point, the brighter the image becomes.
17875 This can be used to reverse a vignette effect, though there is no automatic
17876 detection to extract the lens @option{angle} and other settings (yet). It can
17877 also be used to create a burning effect.
17880 Default value is @samp{forward}.
17883 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17885 It accepts the following values:
17888 Evaluate expressions only once during the filter initialization.
17891 Evaluate expressions for each incoming frame. This is way slower than the
17892 @samp{init} mode since it requires all the scalers to be re-computed, but it
17893 allows advanced dynamic expressions.
17896 Default value is @samp{init}.
17899 Set dithering to reduce the circular banding effects. Default is @code{1}
17903 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17904 Setting this value to the SAR of the input will make a rectangular vignetting
17905 following the dimensions of the video.
17907 Default is @code{1/1}.
17910 @subsection Expressions
17912 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17913 following parameters.
17918 input width and height
17921 the number of input frame, starting from 0
17924 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17925 @var{TB} units, NAN if undefined
17928 frame rate of the input video, NAN if the input frame rate is unknown
17931 the PTS (Presentation TimeStamp) of the filtered video frame,
17932 expressed in seconds, NAN if undefined
17935 time base of the input video
17939 @subsection Examples
17943 Apply simple strong vignetting effect:
17949 Make a flickering vignetting:
17951 vignette='PI/4+random(1)*PI/50':eval=frame
17956 @section vmafmotion
17958 Obtain the average vmaf motion score of a video.
17959 It is one of the component filters of VMAF.
17961 The obtained average motion score is printed through the logging system.
17963 In the below example the input file @file{ref.mpg} is being processed and score
17967 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
17971 Stack input videos vertically.
17973 All streams must be of same pixel format and of same width.
17975 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
17976 to create same output.
17978 The filter accept the following option:
17982 Set number of input streams. Default is 2.
17985 If set to 1, force the output to terminate when the shortest input
17986 terminates. Default value is 0.
17991 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
17992 Deinterlacing Filter").
17994 Based on the process described by Martin Weston for BBC R&D, and
17995 implemented based on the de-interlace algorithm written by Jim
17996 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
17997 uses filter coefficients calculated by BBC R&D.
17999 There are two sets of filter coefficients, so called "simple":
18000 and "complex". Which set of filter coefficients is used can
18001 be set by passing an optional parameter:
18005 Set the interlacing filter coefficients. Accepts one of the following values:
18009 Simple filter coefficient set.
18011 More-complex filter coefficient set.
18013 Default value is @samp{complex}.
18016 Specify which frames to deinterlace. Accept one of the following values:
18020 Deinterlace all frames,
18022 Only deinterlace frames marked as interlaced.
18025 Default value is @samp{all}.
18029 Video waveform monitor.
18031 The waveform monitor plots color component intensity. By default luminance
18032 only. Each column of the waveform corresponds to a column of pixels in the
18035 It accepts the following options:
18039 Can be either @code{row}, or @code{column}. Default is @code{column}.
18040 In row mode, the graph on the left side represents color component value 0 and
18041 the right side represents value = 255. In column mode, the top side represents
18042 color component value = 0 and bottom side represents value = 255.
18045 Set intensity. Smaller values are useful to find out how many values of the same
18046 luminance are distributed across input rows/columns.
18047 Default value is @code{0.04}. Allowed range is [0, 1].
18050 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18051 In mirrored mode, higher values will be represented on the left
18052 side for @code{row} mode and at the top for @code{column} mode. Default is
18053 @code{1} (mirrored).
18057 It accepts the following values:
18060 Presents information identical to that in the @code{parade}, except
18061 that the graphs representing color components are superimposed directly
18064 This display mode makes it easier to spot relative differences or similarities
18065 in overlapping areas of the color components that are supposed to be identical,
18066 such as neutral whites, grays, or blacks.
18069 Display separate graph for the color components side by side in
18070 @code{row} mode or one below the other in @code{column} mode.
18073 Display separate graph for the color components side by side in
18074 @code{column} mode or one below the other in @code{row} mode.
18076 Using this display mode makes it easy to spot color casts in the highlights
18077 and shadows of an image, by comparing the contours of the top and the bottom
18078 graphs of each waveform. Since whites, grays, and blacks are characterized
18079 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18080 should display three waveforms of roughly equal width/height. If not, the
18081 correction is easy to perform by making level adjustments the three waveforms.
18083 Default is @code{stack}.
18085 @item components, c
18086 Set which color components to display. Default is 1, which means only luminance
18087 or red color component if input is in RGB colorspace. If is set for example to
18088 7 it will display all 3 (if) available color components.
18093 No envelope, this is default.
18096 Instant envelope, minimum and maximum values presented in graph will be easily
18097 visible even with small @code{step} value.
18100 Hold minimum and maximum values presented in graph across time. This way you
18101 can still spot out of range values without constantly looking at waveforms.
18104 Peak and instant envelope combined together.
18110 No filtering, this is default.
18113 Luma and chroma combined together.
18116 Similar as above, but shows difference between blue and red chroma.
18119 Similar as above, but use different colors.
18122 Displays only chroma.
18125 Displays actual color value on waveform.
18128 Similar as above, but with luma showing frequency of chroma values.
18132 Set which graticule to display.
18136 Do not display graticule.
18139 Display green graticule showing legal broadcast ranges.
18142 Display orange graticule showing legal broadcast ranges.
18146 Set graticule opacity.
18149 Set graticule flags.
18153 Draw numbers above lines. By default enabled.
18156 Draw dots instead of lines.
18160 Set scale used for displaying graticule.
18167 Default is digital.
18170 Set background opacity.
18173 @section weave, doubleweave
18175 The @code{weave} takes a field-based video input and join
18176 each two sequential fields into single frame, producing a new double
18177 height clip with half the frame rate and half the frame count.
18179 The @code{doubleweave} works same as @code{weave} but without
18180 halving frame rate and frame count.
18182 It accepts the following option:
18186 Set first field. Available values are:
18190 Set the frame as top-field-first.
18193 Set the frame as bottom-field-first.
18197 @subsection Examples
18201 Interlace video using @ref{select} and @ref{separatefields} filter:
18203 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18208 Apply the xBR high-quality magnification filter which is designed for pixel
18209 art. It follows a set of edge-detection rules, see
18210 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18212 It accepts the following option:
18216 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18217 @code{3xBR} and @code{4} for @code{4xBR}.
18218 Default is @code{3}.
18222 Stack video inputs into custom layout.
18224 All streams must be of same pixel format.
18226 The filter accept the following option:
18230 Set number of input streams. Default is 2.
18233 Specify layout of inputs.
18234 This option requires the desired layout configuration to be explicitly set by the user.
18235 This sets position of each video input in output. Each input
18236 is separated by '|'.
18237 The first number represents the column, and the second number represents the row.
18238 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18239 where X is video input from which to take width or height.
18240 Multiple values can be used when separated by '+'. In such
18241 case values are summed together.
18244 If set to 1, force the output to terminate when the shortest input
18245 terminates. Default value is 0.
18248 @subsection Examples
18252 Display 4 inputs into 2x2 grid,
18253 note that if inputs are of different sizes unused gaps might appear,
18254 as not all of output video is used.
18256 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18260 Display 4 inputs into 1x4 grid,
18261 note that if inputs are of different sizes unused gaps might appear,
18262 as not all of output video is used.
18264 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18268 Display 9 inputs into 3x3 grid,
18269 note that if inputs are of different sizes unused gaps might appear,
18270 as not all of output video is used.
18272 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
18279 Deinterlace the input video ("yadif" means "yet another deinterlacing
18282 It accepts the following parameters:
18288 The interlacing mode to adopt. It accepts one of the following values:
18291 @item 0, send_frame
18292 Output one frame for each frame.
18293 @item 1, send_field
18294 Output one frame for each field.
18295 @item 2, send_frame_nospatial
18296 Like @code{send_frame}, but it skips the spatial interlacing check.
18297 @item 3, send_field_nospatial
18298 Like @code{send_field}, but it skips the spatial interlacing check.
18301 The default value is @code{send_frame}.
18304 The picture field parity assumed for the input interlaced video. It accepts one
18305 of the following values:
18309 Assume the top field is first.
18311 Assume the bottom field is first.
18313 Enable automatic detection of field parity.
18316 The default value is @code{auto}.
18317 If the interlacing is unknown or the decoder does not export this information,
18318 top field first will be assumed.
18321 Specify which frames to deinterlace. Accept one of the following
18326 Deinterlace all frames.
18327 @item 1, interlaced
18328 Only deinterlace frames marked as interlaced.
18331 The default value is @code{all}.
18334 @section yadif_cuda
18336 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18337 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18340 It accepts the following parameters:
18346 The interlacing mode to adopt. It accepts one of the following values:
18349 @item 0, send_frame
18350 Output one frame for each frame.
18351 @item 1, send_field
18352 Output one frame for each field.
18353 @item 2, send_frame_nospatial
18354 Like @code{send_frame}, but it skips the spatial interlacing check.
18355 @item 3, send_field_nospatial
18356 Like @code{send_field}, but it skips the spatial interlacing check.
18359 The default value is @code{send_frame}.
18362 The picture field parity assumed for the input interlaced video. It accepts one
18363 of the following values:
18367 Assume the top field is first.
18369 Assume the bottom field is first.
18371 Enable automatic detection of field parity.
18374 The default value is @code{auto}.
18375 If the interlacing is unknown or the decoder does not export this information,
18376 top field first will be assumed.
18379 Specify which frames to deinterlace. Accept one of the following
18384 Deinterlace all frames.
18385 @item 1, interlaced
18386 Only deinterlace frames marked as interlaced.
18389 The default value is @code{all}.
18394 Apply Zoom & Pan effect.
18396 This filter accepts the following options:
18400 Set the zoom expression. Range is 1-10. Default is 1.
18404 Set the x and y expression. Default is 0.
18407 Set the duration expression in number of frames.
18408 This sets for how many number of frames effect will last for
18409 single input image.
18412 Set the output image size, default is 'hd720'.
18415 Set the output frame rate, default is '25'.
18418 Each expression can contain the following constants:
18437 Output frame count.
18441 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18442 for current input frame.
18446 'x' and 'y' of last output frame of previous input frame or 0 when there was
18447 not yet such frame (first input frame).
18450 Last calculated zoom from 'z' expression for current input frame.
18453 Last calculated zoom of last output frame of previous input frame.
18456 Number of output frames for current input frame. Calculated from 'd' expression
18457 for each input frame.
18460 number of output frames created for previous input frame
18463 Rational number: input width / input height
18466 sample aspect ratio
18469 display aspect ratio
18473 @subsection Examples
18477 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18479 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
18483 Zoom-in up to 1.5 and pan always at center of picture:
18485 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18489 Same as above but without pausing:
18491 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18497 Scale (resize) the input video, using the z.lib library:
18498 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18499 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18501 The zscale filter forces the output display aspect ratio to be the same
18502 as the input, by changing the output sample aspect ratio.
18504 If the input image format is different from the format requested by
18505 the next filter, the zscale filter will convert the input to the
18508 @subsection Options
18509 The filter accepts the following options.
18514 Set the output video dimension expression. Default value is the input
18517 If the @var{width} or @var{w} value is 0, the input width is used for
18518 the output. If the @var{height} or @var{h} value is 0, the input height
18519 is used for the output.
18521 If one and only one of the values is -n with n >= 1, the zscale filter
18522 will use a value that maintains the aspect ratio of the input image,
18523 calculated from the other specified dimension. After that it will,
18524 however, make sure that the calculated dimension is divisible by n and
18525 adjust the value if necessary.
18527 If both values are -n with n >= 1, the behavior will be identical to
18528 both values being set to 0 as previously detailed.
18530 See below for the list of accepted constants for use in the dimension
18534 Set the video size. For the syntax of this option, check the
18535 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18538 Set the dither type.
18540 Possible values are:
18545 @item error_diffusion
18551 Set the resize filter type.
18553 Possible values are:
18563 Default is bilinear.
18566 Set the color range.
18568 Possible values are:
18575 Default is same as input.
18578 Set the color primaries.
18580 Possible values are:
18590 Default is same as input.
18593 Set the transfer characteristics.
18595 Possible values are:
18609 Default is same as input.
18612 Set the colorspace matrix.
18614 Possible value are:
18625 Default is same as input.
18628 Set the input color range.
18630 Possible values are:
18637 Default is same as input.
18639 @item primariesin, pin
18640 Set the input color primaries.
18642 Possible values are:
18652 Default is same as input.
18654 @item transferin, tin
18655 Set the input transfer characteristics.
18657 Possible values are:
18668 Default is same as input.
18670 @item matrixin, min
18671 Set the input colorspace matrix.
18673 Possible value are:
18685 Set the output chroma location.
18687 Possible values are:
18698 @item chromalin, cin
18699 Set the input chroma location.
18701 Possible values are:
18713 Set the nominal peak luminance.
18716 The values of the @option{w} and @option{h} options are expressions
18717 containing the following constants:
18722 The input width and height
18726 These are the same as @var{in_w} and @var{in_h}.
18730 The output (scaled) width and height
18734 These are the same as @var{out_w} and @var{out_h}
18737 The same as @var{iw} / @var{ih}
18740 input sample aspect ratio
18743 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18747 horizontal and vertical input chroma subsample values. For example for the
18748 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18752 horizontal and vertical output chroma subsample values. For example for the
18753 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18759 @c man end VIDEO FILTERS
18761 @chapter OpenCL Video Filters
18762 @c man begin OPENCL VIDEO FILTERS
18764 Below is a description of the currently available OpenCL video filters.
18766 To enable compilation of these filters you need to configure FFmpeg with
18767 @code{--enable-opencl}.
18769 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18772 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18773 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18774 given device parameters.
18776 @item -filter_hw_device @var{name}
18777 Pass the hardware device called @var{name} to all filters in any filter graph.
18781 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18785 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18787 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18791 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.
18793 @section avgblur_opencl
18795 Apply average blur filter.
18797 The filter accepts the following options:
18801 Set horizontal radius size.
18802 Range is @code{[1, 1024]} and default value is @code{1}.
18805 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18808 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18811 @subsection Example
18815 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.
18817 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18821 @section boxblur_opencl
18823 Apply a boxblur algorithm to the input video.
18825 It accepts the following parameters:
18829 @item luma_radius, lr
18830 @item luma_power, lp
18831 @item chroma_radius, cr
18832 @item chroma_power, cp
18833 @item alpha_radius, ar
18834 @item alpha_power, ap
18838 A description of the accepted options follows.
18841 @item luma_radius, lr
18842 @item chroma_radius, cr
18843 @item alpha_radius, ar
18844 Set an expression for the box radius in pixels used for blurring the
18845 corresponding input plane.
18847 The radius value must be a non-negative number, and must not be
18848 greater than the value of the expression @code{min(w,h)/2} for the
18849 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18852 Default value for @option{luma_radius} is "2". If not specified,
18853 @option{chroma_radius} and @option{alpha_radius} default to the
18854 corresponding value set for @option{luma_radius}.
18856 The expressions can contain the following constants:
18860 The input width and height in pixels.
18864 The input chroma image width and height in pixels.
18868 The horizontal and vertical chroma subsample values. For example, for the
18869 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
18872 @item luma_power, lp
18873 @item chroma_power, cp
18874 @item alpha_power, ap
18875 Specify how many times the boxblur filter is applied to the
18876 corresponding plane.
18878 Default value for @option{luma_power} is 2. If not specified,
18879 @option{chroma_power} and @option{alpha_power} default to the
18880 corresponding value set for @option{luma_power}.
18882 A value of 0 will disable the effect.
18885 @subsection Examples
18887 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.
18891 Apply a boxblur filter with the luma, chroma, and alpha radius
18892 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.
18894 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
18895 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
18899 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.
18901 For the luma plane, a 2x2 box radius will be run once.
18903 For the chroma plane, a 4x4 box radius will be run 5 times.
18905 For the alpha plane, a 3x3 box radius will be run 7 times.
18907 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
18911 @section convolution_opencl
18913 Apply convolution of 3x3, 5x5, 7x7 matrix.
18915 The filter accepts the following options:
18922 Set matrix for each plane.
18923 Matrix is sequence of 9, 25 or 49 signed numbers.
18924 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
18930 Set multiplier for calculated value for each plane.
18931 If unset or 0, it will be sum of all matrix elements.
18932 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
18938 Set bias for each plane. This value is added to the result of the multiplication.
18939 Useful for making the overall image brighter or darker.
18940 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
18944 @subsection Examples
18950 -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
18956 -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
18960 Apply edge enhance:
18962 -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
18968 -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
18972 Apply laplacian edge detector which includes diagonals:
18974 -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
18980 -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
18984 @section dilation_opencl
18986 Apply dilation effect to the video.
18988 This filter replaces the pixel by the local(3x3) maximum.
18990 It accepts the following options:
18997 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18998 If @code{0}, plane will remain unchanged.
19001 Flag which specifies the pixel to refer to.
19002 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19004 Flags to local 3x3 coordinates region centered on @code{x}:
19013 @subsection Example
19017 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.
19019 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19023 @section erosion_opencl
19025 Apply erosion effect to the video.
19027 This filter replaces the pixel by the local(3x3) minimum.
19029 It accepts the following options:
19036 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19037 If @code{0}, plane will remain unchanged.
19040 Flag which specifies the pixel to refer to.
19041 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19043 Flags to local 3x3 coordinates region centered on @code{x}:
19052 @subsection Example
19056 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.
19058 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19062 @section overlay_opencl
19064 Overlay one video on top of another.
19066 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19067 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19069 The filter accepts the following options:
19074 Set the x coordinate of the overlaid video on the main video.
19075 Default value is @code{0}.
19078 Set the x coordinate of the overlaid video on the main video.
19079 Default value is @code{0}.
19083 @subsection Examples
19087 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19089 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19092 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19094 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19099 @section prewitt_opencl
19101 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19103 The filter accepts the following option:
19107 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19110 Set value which will be multiplied with filtered result.
19111 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19114 Set value which will be added to filtered result.
19115 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19118 @subsection Example
19122 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19124 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19128 @section roberts_opencl
19129 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19131 The filter accepts the following option:
19135 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19138 Set value which will be multiplied with filtered result.
19139 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19142 Set value which will be added to filtered result.
19143 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19146 @subsection Example
19150 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19152 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19156 @section sobel_opencl
19158 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19160 The filter accepts the following option:
19164 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19167 Set value which will be multiplied with filtered result.
19168 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19171 Set value which will be added to filtered result.
19172 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19175 @subsection Example
19179 Apply sobel operator with scale set to 2 and delta set to 10
19181 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19185 @section tonemap_opencl
19187 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19189 It accepts the following parameters:
19193 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19196 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19199 Apply desaturation for highlights that exceed this level of brightness. The
19200 higher the parameter, the more color information will be preserved. This
19201 setting helps prevent unnaturally blown-out colors for super-highlights, by
19202 (smoothly) turning into white instead. This makes images feel more natural,
19203 at the cost of reducing information about out-of-range colors.
19205 The default value is 0.5, and the algorithm here is a little different from
19206 the cpu version tonemap currently. A setting of 0.0 disables this option.
19209 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19210 is used to detect whether the scene has changed or not. If the distance between
19211 the current frame average brightness and the current running average exceeds
19212 a threshold value, we would re-calculate scene average and peak brightness.
19213 The default value is 0.2.
19216 Specify the output pixel format.
19218 Currently supported formats are:
19225 Set the output color range.
19227 Possible values are:
19233 Default is same as input.
19236 Set the output color primaries.
19238 Possible values are:
19244 Default is same as input.
19247 Set the output transfer characteristics.
19249 Possible values are:
19258 Set the output colorspace matrix.
19260 Possible value are:
19266 Default is same as input.
19270 @subsection Example
19274 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19276 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19280 @section unsharp_opencl
19282 Sharpen or blur the input video.
19284 It accepts the following parameters:
19287 @item luma_msize_x, lx
19288 Set the luma matrix horizontal size.
19289 Range is @code{[1, 23]} and default value is @code{5}.
19291 @item luma_msize_y, ly
19292 Set the luma matrix vertical size.
19293 Range is @code{[1, 23]} and default value is @code{5}.
19295 @item luma_amount, la
19296 Set the luma effect strength.
19297 Range is @code{[-10, 10]} and default value is @code{1.0}.
19299 Negative values will blur the input video, while positive values will
19300 sharpen it, a value of zero will disable the effect.
19302 @item chroma_msize_x, cx
19303 Set the chroma matrix horizontal size.
19304 Range is @code{[1, 23]} and default value is @code{5}.
19306 @item chroma_msize_y, cy
19307 Set the chroma matrix vertical size.
19308 Range is @code{[1, 23]} and default value is @code{5}.
19310 @item chroma_amount, ca
19311 Set the chroma effect strength.
19312 Range is @code{[-10, 10]} and default value is @code{0.0}.
19314 Negative values will blur the input video, while positive values will
19315 sharpen it, a value of zero will disable the effect.
19319 All parameters are optional and default to the equivalent of the
19320 string '5:5:1.0:5:5:0.0'.
19322 @subsection Examples
19326 Apply strong luma sharpen effect:
19328 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19332 Apply a strong blur of both luma and chroma parameters:
19334 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19338 @c man end OPENCL VIDEO FILTERS
19340 @chapter Video Sources
19341 @c man begin VIDEO SOURCES
19343 Below is a description of the currently available video sources.
19347 Buffer video frames, and make them available to the filter chain.
19349 This source is mainly intended for a programmatic use, in particular
19350 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19352 It accepts the following parameters:
19357 Specify the size (width and height) of the buffered video frames. For the
19358 syntax of this option, check the
19359 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19362 The input video width.
19365 The input video height.
19368 A string representing the pixel format of the buffered video frames.
19369 It may be a number corresponding to a pixel format, or a pixel format
19373 Specify the timebase assumed by the timestamps of the buffered frames.
19376 Specify the frame rate expected for the video stream.
19378 @item pixel_aspect, sar
19379 The sample (pixel) aspect ratio of the input video.
19382 Specify the optional parameters to be used for the scale filter which
19383 is automatically inserted when an input change is detected in the
19384 input size or format.
19386 @item hw_frames_ctx
19387 When using a hardware pixel format, this should be a reference to an
19388 AVHWFramesContext describing input frames.
19393 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19396 will instruct the source to accept video frames with size 320x240 and
19397 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19398 square pixels (1:1 sample aspect ratio).
19399 Since the pixel format with name "yuv410p" corresponds to the number 6
19400 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19401 this example corresponds to:
19403 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19406 Alternatively, the options can be specified as a flat string, but this
19407 syntax is deprecated:
19409 @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}]
19413 Create a pattern generated by an elementary cellular automaton.
19415 The initial state of the cellular automaton can be defined through the
19416 @option{filename} and @option{pattern} options. If such options are
19417 not specified an initial state is created randomly.
19419 At each new frame a new row in the video is filled with the result of
19420 the cellular automaton next generation. The behavior when the whole
19421 frame is filled is defined by the @option{scroll} option.
19423 This source accepts the following options:
19427 Read the initial cellular automaton state, i.e. the starting row, from
19428 the specified file.
19429 In the file, each non-whitespace character is considered an alive
19430 cell, a newline will terminate the row, and further characters in the
19431 file will be ignored.
19434 Read the initial cellular automaton state, i.e. the starting row, from
19435 the specified string.
19437 Each non-whitespace character in the string is considered an alive
19438 cell, a newline will terminate the row, and further characters in the
19439 string will be ignored.
19442 Set the video rate, that is the number of frames generated per second.
19445 @item random_fill_ratio, ratio
19446 Set the random fill ratio for the initial cellular automaton row. It
19447 is a floating point number value ranging from 0 to 1, defaults to
19450 This option is ignored when a file or a pattern is specified.
19452 @item random_seed, seed
19453 Set the seed for filling randomly the initial row, must be an integer
19454 included between 0 and UINT32_MAX. If not specified, or if explicitly
19455 set to -1, the filter will try to use a good random seed on a best
19459 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19460 Default value is 110.
19463 Set the size of the output video. For the syntax of this option, check the
19464 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19466 If @option{filename} or @option{pattern} is specified, the size is set
19467 by default to the width of the specified initial state row, and the
19468 height is set to @var{width} * PHI.
19470 If @option{size} is set, it must contain the width of the specified
19471 pattern string, and the specified pattern will be centered in the
19474 If a filename or a pattern string is not specified, the size value
19475 defaults to "320x518" (used for a randomly generated initial state).
19478 If set to 1, scroll the output upward when all the rows in the output
19479 have been already filled. If set to 0, the new generated row will be
19480 written over the top row just after the bottom row is filled.
19483 @item start_full, full
19484 If set to 1, completely fill the output with generated rows before
19485 outputting the first frame.
19486 This is the default behavior, for disabling set the value to 0.
19489 If set to 1, stitch the left and right row edges together.
19490 This is the default behavior, for disabling set the value to 0.
19493 @subsection Examples
19497 Read the initial state from @file{pattern}, and specify an output of
19500 cellauto=f=pattern:s=200x400
19504 Generate a random initial row with a width of 200 cells, with a fill
19507 cellauto=ratio=2/3:s=200x200
19511 Create a pattern generated by rule 18 starting by a single alive cell
19512 centered on an initial row with width 100:
19514 cellauto=p=@@:s=100x400:full=0:rule=18
19518 Specify a more elaborated initial pattern:
19520 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19525 @anchor{coreimagesrc}
19526 @section coreimagesrc
19527 Video source generated on GPU using Apple's CoreImage API on OSX.
19529 This video source is a specialized version of the @ref{coreimage} video filter.
19530 Use a core image generator at the beginning of the applied filterchain to
19531 generate the content.
19533 The coreimagesrc video source accepts the following options:
19535 @item list_generators
19536 List all available generators along with all their respective options as well as
19537 possible minimum and maximum values along with the default values.
19539 list_generators=true
19543 Specify the size of the sourced video. For the syntax of this option, check the
19544 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19545 The default value is @code{320x240}.
19548 Specify the frame rate of the sourced video, as the number of frames
19549 generated per second. It has to be a string in the format
19550 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19551 number or a valid video frame rate abbreviation. The default value is
19555 Set the sample aspect ratio of the sourced video.
19558 Set the duration of the sourced video. See
19559 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19560 for the accepted syntax.
19562 If not specified, or the expressed duration is negative, the video is
19563 supposed to be generated forever.
19566 Additionally, all options of the @ref{coreimage} video filter are accepted.
19567 A complete filterchain can be used for further processing of the
19568 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19569 and examples for details.
19571 @subsection Examples
19576 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19577 given as complete and escaped command-line for Apple's standard bash shell:
19579 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19581 This example is equivalent to the QRCode example of @ref{coreimage} without the
19582 need for a nullsrc video source.
19586 @section mandelbrot
19588 Generate a Mandelbrot set fractal, and progressively zoom towards the
19589 point specified with @var{start_x} and @var{start_y}.
19591 This source accepts the following options:
19596 Set the terminal pts value. Default value is 400.
19599 Set the terminal scale value.
19600 Must be a floating point value. Default value is 0.3.
19603 Set the inner coloring mode, that is the algorithm used to draw the
19604 Mandelbrot fractal internal region.
19606 It shall assume one of the following values:
19611 Show time until convergence.
19613 Set color based on point closest to the origin of the iterations.
19618 Default value is @var{mincol}.
19621 Set the bailout value. Default value is 10.0.
19624 Set the maximum of iterations performed by the rendering
19625 algorithm. Default value is 7189.
19628 Set outer coloring mode.
19629 It shall assume one of following values:
19631 @item iteration_count
19632 Set iteration count mode.
19633 @item normalized_iteration_count
19634 set normalized iteration count mode.
19636 Default value is @var{normalized_iteration_count}.
19639 Set frame rate, expressed as number of frames per second. Default
19643 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19644 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19647 Set the initial scale value. Default value is 3.0.
19650 Set the initial x position. Must be a floating point value between
19651 -100 and 100. Default value is -0.743643887037158704752191506114774.
19654 Set the initial y position. Must be a floating point value between
19655 -100 and 100. Default value is -0.131825904205311970493132056385139.
19660 Generate various test patterns, as generated by the MPlayer test filter.
19662 The size of the generated video is fixed, and is 256x256.
19663 This source is useful in particular for testing encoding features.
19665 This source accepts the following options:
19670 Specify the frame rate of the sourced video, as the number of frames
19671 generated per second. It has to be a string in the format
19672 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19673 number or a valid video frame rate abbreviation. The default value is
19677 Set the duration of the sourced video. See
19678 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19679 for the accepted syntax.
19681 If not specified, or the expressed duration is negative, the video is
19682 supposed to be generated forever.
19686 Set the number or the name of the test to perform. Supported tests are:
19702 Default value is "all", which will cycle through the list of all tests.
19707 mptestsrc=t=dc_luma
19710 will generate a "dc_luma" test pattern.
19712 @section frei0r_src
19714 Provide a frei0r source.
19716 To enable compilation of this filter you need to install the frei0r
19717 header and configure FFmpeg with @code{--enable-frei0r}.
19719 This source accepts the following parameters:
19724 The size of the video to generate. For the syntax of this option, check the
19725 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19728 The framerate of the generated video. It may be a string of the form
19729 @var{num}/@var{den} or a frame rate abbreviation.
19732 The name to the frei0r source to load. For more information regarding frei0r and
19733 how to set the parameters, read the @ref{frei0r} section in the video filters
19736 @item filter_params
19737 A '|'-separated list of parameters to pass to the frei0r source.
19741 For example, to generate a frei0r partik0l source with size 200x200
19742 and frame rate 10 which is overlaid on the overlay filter main input:
19744 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19749 Generate a life pattern.
19751 This source is based on a generalization of John Conway's life game.
19753 The sourced input represents a life grid, each pixel represents a cell
19754 which can be in one of two possible states, alive or dead. Every cell
19755 interacts with its eight neighbours, which are the cells that are
19756 horizontally, vertically, or diagonally adjacent.
19758 At each interaction the grid evolves according to the adopted rule,
19759 which specifies the number of neighbor alive cells which will make a
19760 cell stay alive or born. The @option{rule} option allows one to specify
19763 This source accepts the following options:
19767 Set the file from which to read the initial grid state. In the file,
19768 each non-whitespace character is considered an alive cell, and newline
19769 is used to delimit the end of each row.
19771 If this option is not specified, the initial grid is generated
19775 Set the video rate, that is the number of frames generated per second.
19778 @item random_fill_ratio, ratio
19779 Set the random fill ratio for the initial random grid. It is a
19780 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19781 It is ignored when a file is specified.
19783 @item random_seed, seed
19784 Set the seed for filling the initial random grid, must be an integer
19785 included between 0 and UINT32_MAX. If not specified, or if explicitly
19786 set to -1, the filter will try to use a good random seed on a best
19792 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19793 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19794 @var{NS} specifies the number of alive neighbor cells which make a
19795 live cell stay alive, and @var{NB} the number of alive neighbor cells
19796 which make a dead cell to become alive (i.e. to "born").
19797 "s" and "b" can be used in place of "S" and "B", respectively.
19799 Alternatively a rule can be specified by an 18-bits integer. The 9
19800 high order bits are used to encode the next cell state if it is alive
19801 for each number of neighbor alive cells, the low order bits specify
19802 the rule for "borning" new cells. Higher order bits encode for an
19803 higher number of neighbor cells.
19804 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19805 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19807 Default value is "S23/B3", which is the original Conway's game of life
19808 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19809 cells, and will born a new cell if there are three alive cells around
19813 Set the size of the output video. For the syntax of this option, check the
19814 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19816 If @option{filename} is specified, the size is set by default to the
19817 same size of the input file. If @option{size} is set, it must contain
19818 the size specified in the input file, and the initial grid defined in
19819 that file is centered in the larger resulting area.
19821 If a filename is not specified, the size value defaults to "320x240"
19822 (used for a randomly generated initial grid).
19825 If set to 1, stitch the left and right grid edges together, and the
19826 top and bottom edges also. Defaults to 1.
19829 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
19830 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
19831 value from 0 to 255.
19834 Set the color of living (or new born) cells.
19837 Set the color of dead cells. If @option{mold} is set, this is the first color
19838 used to represent a dead cell.
19841 Set mold color, for definitely dead and moldy cells.
19843 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
19844 ffmpeg-utils manual,ffmpeg-utils}.
19847 @subsection Examples
19851 Read a grid from @file{pattern}, and center it on a grid of size
19854 life=f=pattern:s=300x300
19858 Generate a random grid of size 200x200, with a fill ratio of 2/3:
19860 life=ratio=2/3:s=200x200
19864 Specify a custom rule for evolving a randomly generated grid:
19870 Full example with slow death effect (mold) using @command{ffplay}:
19872 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
19879 @anchor{haldclutsrc}
19882 @anchor{pal100bars}
19883 @anchor{rgbtestsrc}
19885 @anchor{smptehdbars}
19888 @anchor{yuvtestsrc}
19889 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
19891 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
19893 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
19895 The @code{color} source provides an uniformly colored input.
19897 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
19898 @ref{haldclut} filter.
19900 The @code{nullsrc} source returns unprocessed video frames. It is
19901 mainly useful to be employed in analysis / debugging tools, or as the
19902 source for filters which ignore the input data.
19904 The @code{pal75bars} source generates a color bars pattern, based on
19905 EBU PAL recommendations with 75% color levels.
19907 The @code{pal100bars} source generates a color bars pattern, based on
19908 EBU PAL recommendations with 100% color levels.
19910 The @code{rgbtestsrc} source generates an RGB test pattern useful for
19911 detecting RGB vs BGR issues. You should see a red, green and blue
19912 stripe from top to bottom.
19914 The @code{smptebars} source generates a color bars pattern, based on
19915 the SMPTE Engineering Guideline EG 1-1990.
19917 The @code{smptehdbars} source generates a color bars pattern, based on
19918 the SMPTE RP 219-2002.
19920 The @code{testsrc} source generates a test video pattern, showing a
19921 color pattern, a scrolling gradient and a timestamp. This is mainly
19922 intended for testing purposes.
19924 The @code{testsrc2} source is similar to testsrc, but supports more
19925 pixel formats instead of just @code{rgb24}. This allows using it as an
19926 input for other tests without requiring a format conversion.
19928 The @code{yuvtestsrc} source generates an YUV test pattern. You should
19929 see a y, cb and cr stripe from top to bottom.
19931 The sources accept the following parameters:
19936 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
19937 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
19938 pixels to be used as identity matrix for 3D lookup tables. Each component is
19939 coded on a @code{1/(N*N)} scale.
19942 Specify the color of the source, only available in the @code{color}
19943 source. For the syntax of this option, check the
19944 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19947 Specify the size of the sourced video. For the syntax of this option, check the
19948 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19949 The default value is @code{320x240}.
19951 This option is not available with the @code{allrgb}, @code{allyuv}, and
19952 @code{haldclutsrc} filters.
19955 Specify the frame rate of the sourced video, as the number of frames
19956 generated per second. It has to be a string in the format
19957 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19958 number or a valid video frame rate abbreviation. The default value is
19962 Set the duration of the sourced video. See
19963 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19964 for the accepted syntax.
19966 If not specified, or the expressed duration is negative, the video is
19967 supposed to be generated forever.
19970 Set the sample aspect ratio of the sourced video.
19973 Specify the alpha (opacity) of the background, only available in the
19974 @code{testsrc2} source. The value must be between 0 (fully transparent) and
19975 255 (fully opaque, the default).
19978 Set the number of decimals to show in the timestamp, only available in the
19979 @code{testsrc} source.
19981 The displayed timestamp value will correspond to the original
19982 timestamp value multiplied by the power of 10 of the specified
19983 value. Default value is 0.
19986 @subsection Examples
19990 Generate a video with a duration of 5.3 seconds, with size
19991 176x144 and a frame rate of 10 frames per second:
19993 testsrc=duration=5.3:size=qcif:rate=10
19997 The following graph description will generate a red source
19998 with an opacity of 0.2, with size "qcif" and a frame rate of 10
20001 color=c=red@@0.2:s=qcif:r=10
20005 If the input content is to be ignored, @code{nullsrc} can be used. The
20006 following command generates noise in the luminance plane by employing
20007 the @code{geq} filter:
20009 nullsrc=s=256x256, geq=random(1)*255:128:128
20013 @subsection Commands
20015 The @code{color} source supports the following commands:
20019 Set the color of the created image. Accepts the same syntax of the
20020 corresponding @option{color} option.
20025 Generate video using an OpenCL program.
20030 OpenCL program source file.
20033 Kernel name in program.
20036 Size of frames to generate. This must be set.
20039 Pixel format to use for the generated frames. This must be set.
20042 Number of frames generated every second. Default value is '25'.
20046 For details of how the program loading works, see the @ref{program_opencl}
20053 Generate a colour ramp by setting pixel values from the position of the pixel
20054 in the output image. (Note that this will work with all pixel formats, but
20055 the generated output will not be the same.)
20057 __kernel void ramp(__write_only image2d_t dst,
20058 unsigned int index)
20060 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20063 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20065 write_imagef(dst, loc, val);
20070 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20072 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20073 unsigned int index)
20075 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20077 float4 value = 0.0f;
20078 int x = loc.x + index;
20079 int y = loc.y + index;
20080 while (x > 0 || y > 0) {
20081 if (x % 3 == 1 && y % 3 == 1) {
20089 write_imagef(dst, loc, value);
20095 @c man end VIDEO SOURCES
20097 @chapter Video Sinks
20098 @c man begin VIDEO SINKS
20100 Below is a description of the currently available video sinks.
20102 @section buffersink
20104 Buffer video frames, and make them available to the end of the filter
20107 This sink is mainly intended for programmatic use, in particular
20108 through the interface defined in @file{libavfilter/buffersink.h}
20109 or the options system.
20111 It accepts a pointer to an AVBufferSinkContext structure, which
20112 defines the incoming buffers' formats, to be passed as the opaque
20113 parameter to @code{avfilter_init_filter} for initialization.
20117 Null video sink: do absolutely nothing with the input video. It is
20118 mainly useful as a template and for use in analysis / debugging
20121 @c man end VIDEO SINKS
20123 @chapter Multimedia Filters
20124 @c man begin MULTIMEDIA FILTERS
20126 Below is a description of the currently available multimedia filters.
20130 Convert input audio to a video output, displaying the audio bit scope.
20132 The filter accepts the following options:
20136 Set frame rate, expressed as number of frames per second. Default
20140 Specify the video size for the output. For the syntax of this option, check the
20141 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20142 Default value is @code{1024x256}.
20145 Specify list of colors separated by space or by '|' which will be used to
20146 draw channels. Unrecognized or missing colors will be replaced
20150 @section ahistogram
20152 Convert input audio to a video output, displaying the volume histogram.
20154 The filter accepts the following options:
20158 Specify how histogram is calculated.
20160 It accepts the following values:
20163 Use single histogram for all channels.
20165 Use separate histogram for each channel.
20167 Default is @code{single}.
20170 Set frame rate, expressed as number of frames per second. Default
20174 Specify the video size for the output. For the syntax of this option, check the
20175 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20176 Default value is @code{hd720}.
20181 It accepts the following values:
20192 reverse logarithmic
20194 Default is @code{log}.
20197 Set amplitude scale.
20199 It accepts the following values:
20206 Default is @code{log}.
20209 Set how much frames to accumulate in histogram.
20210 Default is 1. Setting this to -1 accumulates all frames.
20213 Set histogram ratio of window height.
20216 Set sonogram sliding.
20218 It accepts the following values:
20221 replace old rows with new ones.
20223 scroll from top to bottom.
20225 Default is @code{replace}.
20228 @section aphasemeter
20230 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20231 representing mean phase of current audio frame. A video output can also be produced and is
20232 enabled by default. The audio is passed through as first output.
20234 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20235 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20236 and @code{1} means channels are in phase.
20238 The filter accepts the following options, all related to its video output:
20242 Set the output frame rate. Default value is @code{25}.
20245 Set the video size for the output. For the syntax of this option, check the
20246 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20247 Default value is @code{800x400}.
20252 Specify the red, green, blue contrast. Default values are @code{2},
20253 @code{7} and @code{1}.
20254 Allowed range is @code{[0, 255]}.
20257 Set color which will be used for drawing median phase. If color is
20258 @code{none} which is default, no median phase value will be drawn.
20261 Enable video output. Default is enabled.
20264 @section avectorscope
20266 Convert input audio to a video output, representing the audio vector
20269 The filter is used to measure the difference between channels of stereo
20270 audio stream. A monoaural signal, consisting of identical left and right
20271 signal, results in straight vertical line. Any stereo separation is visible
20272 as a deviation from this line, creating a Lissajous figure.
20273 If the straight (or deviation from it) but horizontal line appears this
20274 indicates that the left and right channels are out of phase.
20276 The filter accepts the following options:
20280 Set the vectorscope mode.
20282 Available values are:
20285 Lissajous rotated by 45 degrees.
20288 Same as above but not rotated.
20291 Shape resembling half of circle.
20294 Default value is @samp{lissajous}.
20297 Set the video size for the output. For the syntax of this option, check the
20298 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20299 Default value is @code{400x400}.
20302 Set the output frame rate. Default value is @code{25}.
20308 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20309 @code{160}, @code{80} and @code{255}.
20310 Allowed range is @code{[0, 255]}.
20316 Specify the red, green, blue and alpha fade. Default values are @code{15},
20317 @code{10}, @code{5} and @code{5}.
20318 Allowed range is @code{[0, 255]}.
20321 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20322 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20325 Set the vectorscope drawing mode.
20327 Available values are:
20330 Draw dot for each sample.
20333 Draw line between previous and current sample.
20336 Default value is @samp{dot}.
20339 Specify amplitude scale of audio samples.
20341 Available values are:
20357 Swap left channel axis with right channel axis.
20367 Mirror only x axis.
20370 Mirror only y axis.
20378 @subsection Examples
20382 Complete example using @command{ffplay}:
20384 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20385 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20389 @section bench, abench
20391 Benchmark part of a filtergraph.
20393 The filter accepts the following options:
20397 Start or stop a timer.
20399 Available values are:
20402 Get the current time, set it as frame metadata (using the key
20403 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20406 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20407 the input frame metadata to get the time difference. Time difference, average,
20408 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20409 @code{min}) are then printed. The timestamps are expressed in seconds.
20413 @subsection Examples
20417 Benchmark @ref{selectivecolor} filter:
20419 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20425 Concatenate audio and video streams, joining them together one after the
20428 The filter works on segments of synchronized video and audio streams. All
20429 segments must have the same number of streams of each type, and that will
20430 also be the number of streams at output.
20432 The filter accepts the following options:
20437 Set the number of segments. Default is 2.
20440 Set the number of output video streams, that is also the number of video
20441 streams in each segment. Default is 1.
20444 Set the number of output audio streams, that is also the number of audio
20445 streams in each segment. Default is 0.
20448 Activate unsafe mode: do not fail if segments have a different format.
20452 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20453 @var{a} audio outputs.
20455 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20456 segment, in the same order as the outputs, then the inputs for the second
20459 Related streams do not always have exactly the same duration, for various
20460 reasons including codec frame size or sloppy authoring. For that reason,
20461 related synchronized streams (e.g. a video and its audio track) should be
20462 concatenated at once. The concat filter will use the duration of the longest
20463 stream in each segment (except the last one), and if necessary pad shorter
20464 audio streams with silence.
20466 For this filter to work correctly, all segments must start at timestamp 0.
20468 All corresponding streams must have the same parameters in all segments; the
20469 filtering system will automatically select a common pixel format for video
20470 streams, and a common sample format, sample rate and channel layout for
20471 audio streams, but other settings, such as resolution, must be converted
20472 explicitly by the user.
20474 Different frame rates are acceptable but will result in variable frame rate
20475 at output; be sure to configure the output file to handle it.
20477 @subsection Examples
20481 Concatenate an opening, an episode and an ending, all in bilingual version
20482 (video in stream 0, audio in streams 1 and 2):
20484 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20485 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20486 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20487 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20491 Concatenate two parts, handling audio and video separately, using the
20492 (a)movie sources, and adjusting the resolution:
20494 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20495 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20496 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20498 Note that a desync will happen at the stitch if the audio and video streams
20499 do not have exactly the same duration in the first file.
20503 @subsection Commands
20505 This filter supports the following commands:
20508 Close the current segment and step to the next one
20511 @section drawgraph, adrawgraph
20513 Draw a graph using input video or audio metadata.
20515 It accepts the following parameters:
20519 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20522 Set 1st foreground color expression.
20525 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20528 Set 2nd foreground color expression.
20531 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20534 Set 3rd foreground color expression.
20537 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20540 Set 4th foreground color expression.
20543 Set minimal value of metadata value.
20546 Set maximal value of metadata value.
20549 Set graph background color. Default is white.
20554 Available values for mode is:
20561 Default is @code{line}.
20566 Available values for slide is:
20569 Draw new frame when right border is reached.
20572 Replace old columns with new ones.
20575 Scroll from right to left.
20578 Scroll from left to right.
20581 Draw single picture.
20584 Default is @code{frame}.
20587 Set size of graph video. For the syntax of this option, check the
20588 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20589 The default value is @code{900x256}.
20591 The foreground color expressions can use the following variables:
20594 Minimal value of metadata value.
20597 Maximal value of metadata value.
20600 Current metadata key value.
20603 The color is defined as 0xAABBGGRR.
20606 Example using metadata from @ref{signalstats} filter:
20608 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20611 Example using metadata from @ref{ebur128} filter:
20613 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20619 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
20620 level. By default, it logs a message at a frequency of 10Hz with the
20621 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20622 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20624 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
20625 sample format is double-precision floating point. The input stream will be converted to
20626 this specification, if needed. Users may need to insert aformat and/or aresample filters
20627 after this filter to obtain the original parameters.
20629 The filter also has a video output (see the @var{video} option) with a real
20630 time graph to observe the loudness evolution. The graphic contains the logged
20631 message mentioned above, so it is not printed anymore when this option is set,
20632 unless the verbose logging is set. The main graphing area contains the
20633 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20634 the momentary loudness (400 milliseconds), but can optionally be configured
20635 to instead display short-term loudness (see @var{gauge}).
20637 The green area marks a +/- 1LU target range around the target loudness
20638 (-23LUFS by default, unless modified through @var{target}).
20640 More information about the Loudness Recommendation EBU R128 on
20641 @url{http://tech.ebu.ch/loudness}.
20643 The filter accepts the following options:
20648 Activate the video output. The audio stream is passed unchanged whether this
20649 option is set or no. The video stream will be the first output stream if
20650 activated. Default is @code{0}.
20653 Set the video size. This option is for video only. For the syntax of this
20655 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20656 Default and minimum resolution is @code{640x480}.
20659 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20660 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20661 other integer value between this range is allowed.
20664 Set metadata injection. If set to @code{1}, the audio input will be segmented
20665 into 100ms output frames, each of them containing various loudness information
20666 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20668 Default is @code{0}.
20671 Force the frame logging level.
20673 Available values are:
20676 information logging level
20678 verbose logging level
20681 By default, the logging level is set to @var{info}. If the @option{video} or
20682 the @option{metadata} options are set, it switches to @var{verbose}.
20687 Available modes can be cumulated (the option is a @code{flag} type). Possible
20691 Disable any peak mode (default).
20693 Enable sample-peak mode.
20695 Simple peak mode looking for the higher sample value. It logs a message
20696 for sample-peak (identified by @code{SPK}).
20698 Enable true-peak mode.
20700 If enabled, the peak lookup is done on an over-sampled version of the input
20701 stream for better peak accuracy. It logs a message for true-peak.
20702 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20703 This mode requires a build with @code{libswresample}.
20707 Treat mono input files as "dual mono". If a mono file is intended for playback
20708 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20709 If set to @code{true}, this option will compensate for this effect.
20710 Multi-channel input files are not affected by this option.
20713 Set a specific pan law to be used for the measurement of dual mono files.
20714 This parameter is optional, and has a default value of -3.01dB.
20717 Set a specific target level (in LUFS) used as relative zero in the visualization.
20718 This parameter is optional and has a default value of -23LUFS as specified
20719 by EBU R128. However, material published online may prefer a level of -16LUFS
20720 (e.g. for use with podcasts or video platforms).
20723 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20724 @code{shortterm}. By default the momentary value will be used, but in certain
20725 scenarios it may be more useful to observe the short term value instead (e.g.
20729 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20730 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20731 video output, not the summary or continuous log output.
20734 @subsection Examples
20738 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20740 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20744 Run an analysis with @command{ffmpeg}:
20746 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20750 @section interleave, ainterleave
20752 Temporally interleave frames from several inputs.
20754 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20756 These filters read frames from several inputs and send the oldest
20757 queued frame to the output.
20759 Input streams must have well defined, monotonically increasing frame
20762 In order to submit one frame to output, these filters need to enqueue
20763 at least one frame for each input, so they cannot work in case one
20764 input is not yet terminated and will not receive incoming frames.
20766 For example consider the case when one input is a @code{select} filter
20767 which always drops input frames. The @code{interleave} filter will keep
20768 reading from that input, but it will never be able to send new frames
20769 to output until the input sends an end-of-stream signal.
20771 Also, depending on inputs synchronization, the filters will drop
20772 frames in case one input receives more frames than the other ones, and
20773 the queue is already filled.
20775 These filters accept the following options:
20779 Set the number of different inputs, it is 2 by default.
20782 @subsection Examples
20786 Interleave frames belonging to different streams using @command{ffmpeg}:
20788 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20792 Add flickering blur effect:
20794 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20798 @section metadata, ametadata
20800 Manipulate frame metadata.
20802 This filter accepts the following options:
20806 Set mode of operation of the filter.
20808 Can be one of the following:
20812 If both @code{value} and @code{key} is set, select frames
20813 which have such metadata. If only @code{key} is set, select
20814 every frame that has such key in metadata.
20817 Add new metadata @code{key} and @code{value}. If key is already available
20821 Modify value of already present key.
20824 If @code{value} is set, delete only keys that have such value.
20825 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
20829 Print key and its value if metadata was found. If @code{key} is not set print all
20830 metadata values available in frame.
20834 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
20837 Set metadata value which will be used. This option is mandatory for
20838 @code{modify} and @code{add} mode.
20841 Which function to use when comparing metadata value and @code{value}.
20843 Can be one of following:
20847 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
20850 Values are interpreted as strings, returns true if metadata value starts with
20851 the @code{value} option string.
20854 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
20857 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
20860 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
20863 Values are interpreted as floats, returns true if expression from option @code{expr}
20868 Set expression which is used when @code{function} is set to @code{expr}.
20869 The expression is evaluated through the eval API and can contain the following
20874 Float representation of @code{value} from metadata key.
20877 Float representation of @code{value} as supplied by user in @code{value} option.
20881 If specified in @code{print} mode, output is written to the named file. Instead of
20882 plain filename any writable url can be specified. Filename ``-'' is a shorthand
20883 for standard output. If @code{file} option is not set, output is written to the log
20884 with AV_LOG_INFO loglevel.
20888 @subsection Examples
20892 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
20895 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
20898 Print silencedetect output to file @file{metadata.txt}.
20900 silencedetect,ametadata=mode=print:file=metadata.txt
20903 Direct all metadata to a pipe with file descriptor 4.
20905 metadata=mode=print:file='pipe\:4'
20909 @section perms, aperms
20911 Set read/write permissions for the output frames.
20913 These filters are mainly aimed at developers to test direct path in the
20914 following filter in the filtergraph.
20916 The filters accept the following options:
20920 Select the permissions mode.
20922 It accepts the following values:
20925 Do nothing. This is the default.
20927 Set all the output frames read-only.
20929 Set all the output frames directly writable.
20931 Make the frame read-only if writable, and writable if read-only.
20933 Set each output frame read-only or writable randomly.
20937 Set the seed for the @var{random} mode, must be an integer included between
20938 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
20939 @code{-1}, the filter will try to use a good random seed on a best effort
20943 Note: in case of auto-inserted filter between the permission filter and the
20944 following one, the permission might not be received as expected in that
20945 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
20946 perms/aperms filter can avoid this problem.
20948 @section realtime, arealtime
20950 Slow down filtering to match real time approximately.
20952 These filters will pause the filtering for a variable amount of time to
20953 match the output rate with the input timestamps.
20954 They are similar to the @option{re} option to @code{ffmpeg}.
20956 They accept the following options:
20960 Time limit for the pauses. Any pause longer than that will be considered
20961 a timestamp discontinuity and reset the timer. Default is 2 seconds.
20965 @section select, aselect
20967 Select frames to pass in output.
20969 This filter accepts the following options:
20974 Set expression, which is evaluated for each input frame.
20976 If the expression is evaluated to zero, the frame is discarded.
20978 If the evaluation result is negative or NaN, the frame is sent to the
20979 first output; otherwise it is sent to the output with index
20980 @code{ceil(val)-1}, assuming that the input index starts from 0.
20982 For example a value of @code{1.2} corresponds to the output with index
20983 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
20986 Set the number of outputs. The output to which to send the selected
20987 frame is based on the result of the evaluation. Default value is 1.
20990 The expression can contain the following constants:
20994 The (sequential) number of the filtered frame, starting from 0.
20997 The (sequential) number of the selected frame, starting from 0.
20999 @item prev_selected_n
21000 The sequential number of the last selected frame. It's NAN if undefined.
21003 The timebase of the input timestamps.
21006 The PTS (Presentation TimeStamp) of the filtered video frame,
21007 expressed in @var{TB} units. It's NAN if undefined.
21010 The PTS of the filtered video frame,
21011 expressed in seconds. It's NAN if undefined.
21014 The PTS of the previously filtered video frame. It's NAN if undefined.
21016 @item prev_selected_pts
21017 The PTS of the last previously filtered video frame. It's NAN if undefined.
21019 @item prev_selected_t
21020 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
21023 The PTS of the first video frame in the video. It's NAN if undefined.
21026 The time of the first video frame in the video. It's NAN if undefined.
21028 @item pict_type @emph{(video only)}
21029 The type of the filtered frame. It can assume one of the following
21041 @item interlace_type @emph{(video only)}
21042 The frame interlace type. It can assume one of the following values:
21045 The frame is progressive (not interlaced).
21047 The frame is top-field-first.
21049 The frame is bottom-field-first.
21052 @item consumed_sample_n @emph{(audio only)}
21053 the number of selected samples before the current frame
21055 @item samples_n @emph{(audio only)}
21056 the number of samples in the current frame
21058 @item sample_rate @emph{(audio only)}
21059 the input sample rate
21062 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21065 the position in the file of the filtered frame, -1 if the information
21066 is not available (e.g. for synthetic video)
21068 @item scene @emph{(video only)}
21069 value between 0 and 1 to indicate a new scene; a low value reflects a low
21070 probability for the current frame to introduce a new scene, while a higher
21071 value means the current frame is more likely to be one (see the example below)
21073 @item concatdec_select
21074 The concat demuxer can select only part of a concat input file by setting an
21075 inpoint and an outpoint, but the output packets may not be entirely contained
21076 in the selected interval. By using this variable, it is possible to skip frames
21077 generated by the concat demuxer which are not exactly contained in the selected
21080 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21081 and the @var{lavf.concat.duration} packet metadata values which are also
21082 present in the decoded frames.
21084 The @var{concatdec_select} variable is -1 if the frame pts is at least
21085 start_time and either the duration metadata is missing or the frame pts is less
21086 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21089 That basically means that an input frame is selected if its pts is within the
21090 interval set by the concat demuxer.
21094 The default value of the select expression is "1".
21096 @subsection Examples
21100 Select all frames in input:
21105 The example above is the same as:
21117 Select only I-frames:
21119 select='eq(pict_type\,I)'
21123 Select one frame every 100:
21125 select='not(mod(n\,100))'
21129 Select only frames contained in the 10-20 time interval:
21131 select=between(t\,10\,20)
21135 Select only I-frames contained in the 10-20 time interval:
21137 select=between(t\,10\,20)*eq(pict_type\,I)
21141 Select frames with a minimum distance of 10 seconds:
21143 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21147 Use aselect to select only audio frames with samples number > 100:
21149 aselect='gt(samples_n\,100)'
21153 Create a mosaic of the first scenes:
21155 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21158 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21162 Send even and odd frames to separate outputs, and compose them:
21164 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21168 Select useful frames from an ffconcat file which is using inpoints and
21169 outpoints but where the source files are not intra frame only.
21171 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21175 @section sendcmd, asendcmd
21177 Send commands to filters in the filtergraph.
21179 These filters read commands to be sent to other filters in the
21182 @code{sendcmd} must be inserted between two video filters,
21183 @code{asendcmd} must be inserted between two audio filters, but apart
21184 from that they act the same way.
21186 The specification of commands can be provided in the filter arguments
21187 with the @var{commands} option, or in a file specified by the
21188 @var{filename} option.
21190 These filters accept the following options:
21193 Set the commands to be read and sent to the other filters.
21195 Set the filename of the commands to be read and sent to the other
21199 @subsection Commands syntax
21201 A commands description consists of a sequence of interval
21202 specifications, comprising a list of commands to be executed when a
21203 particular event related to that interval occurs. The occurring event
21204 is typically the current frame time entering or leaving a given time
21207 An interval is specified by the following syntax:
21209 @var{START}[-@var{END}] @var{COMMANDS};
21212 The time interval is specified by the @var{START} and @var{END} times.
21213 @var{END} is optional and defaults to the maximum time.
21215 The current frame time is considered within the specified interval if
21216 it is included in the interval [@var{START}, @var{END}), that is when
21217 the time is greater or equal to @var{START} and is lesser than
21220 @var{COMMANDS} consists of a sequence of one or more command
21221 specifications, separated by ",", relating to that interval. The
21222 syntax of a command specification is given by:
21224 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21227 @var{FLAGS} is optional and specifies the type of events relating to
21228 the time interval which enable sending the specified command, and must
21229 be a non-null sequence of identifier flags separated by "+" or "|" and
21230 enclosed between "[" and "]".
21232 The following flags are recognized:
21235 The command is sent when the current frame timestamp enters the
21236 specified interval. In other words, the command is sent when the
21237 previous frame timestamp was not in the given interval, and the
21241 The command is sent when the current frame timestamp leaves the
21242 specified interval. In other words, the command is sent when the
21243 previous frame timestamp was in the given interval, and the
21247 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21250 @var{TARGET} specifies the target of the command, usually the name of
21251 the filter class or a specific filter instance name.
21253 @var{COMMAND} specifies the name of the command for the target filter.
21255 @var{ARG} is optional and specifies the optional list of argument for
21256 the given @var{COMMAND}.
21258 Between one interval specification and another, whitespaces, or
21259 sequences of characters starting with @code{#} until the end of line,
21260 are ignored and can be used to annotate comments.
21262 A simplified BNF description of the commands specification syntax
21265 @var{COMMAND_FLAG} ::= "enter" | "leave"
21266 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21267 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21268 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21269 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21270 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21273 @subsection Examples
21277 Specify audio tempo change at second 4:
21279 asendcmd=c='4.0 atempo tempo 1.5',atempo
21283 Target a specific filter instance:
21285 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21289 Specify a list of drawtext and hue commands in a file.
21291 # show text in the interval 5-10
21292 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21293 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21295 # desaturate the image in the interval 15-20
21296 15.0-20.0 [enter] hue s 0,
21297 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21299 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21301 # apply an exponential saturation fade-out effect, starting from time 25
21302 25 [enter] hue s exp(25-t)
21305 A filtergraph allowing to read and process the above command list
21306 stored in a file @file{test.cmd}, can be specified with:
21308 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21313 @section setpts, asetpts
21315 Change the PTS (presentation timestamp) of the input frames.
21317 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21319 This filter accepts the following options:
21324 The expression which is evaluated for each frame to construct its timestamp.
21328 The expression is evaluated through the eval API and can contain the following
21332 @item FRAME_RATE, FR
21333 frame rate, only defined for constant frame-rate video
21336 The presentation timestamp in input
21339 The count of the input frame for video or the number of consumed samples,
21340 not including the current frame for audio, starting from 0.
21342 @item NB_CONSUMED_SAMPLES
21343 The number of consumed samples, not including the current frame (only
21346 @item NB_SAMPLES, S
21347 The number of samples in the current frame (only audio)
21349 @item SAMPLE_RATE, SR
21350 The audio sample rate.
21353 The PTS of the first frame.
21356 the time in seconds of the first frame
21359 State whether the current frame is interlaced.
21362 the time in seconds of the current frame
21365 original position in the file of the frame, or undefined if undefined
21366 for the current frame
21369 The previous input PTS.
21372 previous input time in seconds
21375 The previous output PTS.
21378 previous output time in seconds
21381 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21385 The wallclock (RTC) time at the start of the movie in microseconds.
21388 The timebase of the input timestamps.
21392 @subsection Examples
21396 Start counting PTS from zero
21398 setpts=PTS-STARTPTS
21402 Apply fast motion effect:
21408 Apply slow motion effect:
21414 Set fixed rate of 25 frames per second:
21420 Set fixed rate 25 fps with some jitter:
21422 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21426 Apply an offset of 10 seconds to the input PTS:
21432 Generate timestamps from a "live source" and rebase onto the current timebase:
21434 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21438 Generate timestamps by counting samples:
21447 Force color range for the output video frame.
21449 The @code{setrange} filter marks the color range property for the
21450 output frames. It does not change the input frame, but only sets the
21451 corresponding property, which affects how the frame is treated by
21454 The filter accepts the following options:
21459 Available values are:
21463 Keep the same color range property.
21465 @item unspecified, unknown
21466 Set the color range as unspecified.
21468 @item limited, tv, mpeg
21469 Set the color range as limited.
21471 @item full, pc, jpeg
21472 Set the color range as full.
21476 @section settb, asettb
21478 Set the timebase to use for the output frames timestamps.
21479 It is mainly useful for testing timebase configuration.
21481 It accepts the following parameters:
21486 The expression which is evaluated into the output timebase.
21490 The value for @option{tb} is an arithmetic expression representing a
21491 rational. The expression can contain the constants "AVTB" (the default
21492 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21493 audio only). Default value is "intb".
21495 @subsection Examples
21499 Set the timebase to 1/25:
21505 Set the timebase to 1/10:
21511 Set the timebase to 1001/1000:
21517 Set the timebase to 2*intb:
21523 Set the default timebase value:
21530 Convert input audio to a video output representing frequency spectrum
21531 logarithmically using Brown-Puckette constant Q transform algorithm with
21532 direct frequency domain coefficient calculation (but the transform itself
21533 is not really constant Q, instead the Q factor is actually variable/clamped),
21534 with musical tone scale, from E0 to D#10.
21536 The filter accepts the following options:
21540 Specify the video size for the output. It must be even. For the syntax of this option,
21541 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21542 Default value is @code{1920x1080}.
21545 Set the output frame rate. Default value is @code{25}.
21548 Set the bargraph height. It must be even. Default value is @code{-1} which
21549 computes the bargraph height automatically.
21552 Set the axis height. It must be even. Default value is @code{-1} which computes
21553 the axis height automatically.
21556 Set the sonogram height. It must be even. Default value is @code{-1} which
21557 computes the sonogram height automatically.
21560 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21561 instead. Default value is @code{1}.
21563 @item sono_v, volume
21564 Specify the sonogram volume expression. It can contain variables:
21567 the @var{bar_v} evaluated expression
21568 @item frequency, freq, f
21569 the frequency where it is evaluated
21570 @item timeclamp, tc
21571 the value of @var{timeclamp} option
21575 @item a_weighting(f)
21576 A-weighting of equal loudness
21577 @item b_weighting(f)
21578 B-weighting of equal loudness
21579 @item c_weighting(f)
21580 C-weighting of equal loudness.
21582 Default value is @code{16}.
21584 @item bar_v, volume2
21585 Specify the bargraph volume expression. It can contain variables:
21588 the @var{sono_v} evaluated expression
21589 @item frequency, freq, f
21590 the frequency where it is evaluated
21591 @item timeclamp, tc
21592 the value of @var{timeclamp} option
21596 @item a_weighting(f)
21597 A-weighting of equal loudness
21598 @item b_weighting(f)
21599 B-weighting of equal loudness
21600 @item c_weighting(f)
21601 C-weighting of equal loudness.
21603 Default value is @code{sono_v}.
21605 @item sono_g, gamma
21606 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21607 higher gamma makes the spectrum having more range. Default value is @code{3}.
21608 Acceptable range is @code{[1, 7]}.
21610 @item bar_g, gamma2
21611 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21615 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21616 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21618 @item timeclamp, tc
21619 Specify the transform timeclamp. At low frequency, there is trade-off between
21620 accuracy in time domain and frequency domain. If timeclamp is lower,
21621 event in time domain is represented more accurately (such as fast bass drum),
21622 otherwise event in frequency domain is represented more accurately
21623 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21626 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21627 limits future samples by applying asymmetric windowing in time domain, useful
21628 when low latency is required. Accepted range is @code{[0, 1]}.
21631 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21632 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21635 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21636 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21639 This option is deprecated and ignored.
21642 Specify the transform length in time domain. Use this option to control accuracy
21643 trade-off between time domain and frequency domain at every frequency sample.
21644 It can contain variables:
21646 @item frequency, freq, f
21647 the frequency where it is evaluated
21648 @item timeclamp, tc
21649 the value of @var{timeclamp} option.
21651 Default value is @code{384*tc/(384+tc*f)}.
21654 Specify the transform count for every video frame. Default value is @code{6}.
21655 Acceptable range is @code{[1, 30]}.
21658 Specify the transform count for every single pixel. Default value is @code{0},
21659 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21662 Specify font file for use with freetype to draw the axis. If not specified,
21663 use embedded font. Note that drawing with font file or embedded font is not
21664 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21668 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21669 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21672 Specify font color expression. This is arithmetic expression that should return
21673 integer value 0xRRGGBB. It can contain variables:
21675 @item frequency, freq, f
21676 the frequency where it is evaluated
21677 @item timeclamp, tc
21678 the value of @var{timeclamp} option
21683 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21684 @item r(x), g(x), b(x)
21685 red, green, and blue value of intensity x.
21687 Default value is @code{st(0, (midi(f)-59.5)/12);
21688 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21689 r(1-ld(1)) + b(ld(1))}.
21692 Specify image file to draw the axis. This option override @var{fontfile} and
21693 @var{fontcolor} option.
21696 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21697 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21698 Default value is @code{1}.
21701 Set colorspace. The accepted values are:
21704 Unspecified (default)
21713 BT.470BG or BT.601-6 625
21716 SMPTE-170M or BT.601-6 525
21722 BT.2020 with non-constant luminance
21727 Set spectrogram color scheme. This is list of floating point values with format
21728 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21729 The default is @code{1|0.5|0|0|0.5|1}.
21733 @subsection Examples
21737 Playing audio while showing the spectrum:
21739 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21743 Same as above, but with frame rate 30 fps:
21745 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21749 Playing at 1280x720:
21751 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21755 Disable sonogram display:
21761 A1 and its harmonics: A1, A2, (near)E3, A3:
21763 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),
21764 asplit[a][out1]; [a] showcqt [out0]'
21768 Same as above, but with more accuracy in frequency domain:
21770 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),
21771 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21777 bar_v=10:sono_v=bar_v*a_weighting(f)
21781 Custom gamma, now spectrum is linear to the amplitude.
21787 Custom tlength equation:
21789 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)))'
21793 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21795 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21799 Custom font using fontconfig:
21801 font='Courier New,Monospace,mono|bold'
21805 Custom frequency range with custom axis using image file:
21807 axisfile=myaxis.png:basefreq=40:endfreq=10000
21813 Convert input audio to video output representing the audio power spectrum.
21814 Audio amplitude is on Y-axis while frequency is on X-axis.
21816 The filter accepts the following options:
21820 Specify size of video. For the syntax of this option, check the
21821 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21822 Default is @code{1024x512}.
21826 This set how each frequency bin will be represented.
21828 It accepts the following values:
21834 Default is @code{bar}.
21837 Set amplitude scale.
21839 It accepts the following values:
21853 Default is @code{log}.
21856 Set frequency scale.
21858 It accepts the following values:
21867 Reverse logarithmic scale.
21869 Default is @code{lin}.
21874 It accepts the following values:
21890 Default is @code{w2048}
21893 Set windowing function.
21895 It accepts the following values:
21918 Default is @code{hanning}.
21921 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
21922 which means optimal overlap for selected window function will be picked.
21925 Set time averaging. Setting this to 0 will display current maximal peaks.
21926 Default is @code{1}, which means time averaging is disabled.
21929 Specify list of colors separated by space or by '|' which will be used to
21930 draw channel frequencies. Unrecognized or missing colors will be replaced
21934 Set channel display mode.
21936 It accepts the following values:
21941 Default is @code{combined}.
21944 Set minimum amplitude used in @code{log} amplitude scaler.
21948 @anchor{showspectrum}
21949 @section showspectrum
21951 Convert input audio to a video output, representing the audio frequency
21954 The filter accepts the following options:
21958 Specify the video size for the output. For the syntax of this option, check the
21959 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21960 Default value is @code{640x512}.
21963 Specify how the spectrum should slide along the window.
21965 It accepts the following values:
21968 the samples start again on the left when they reach the right
21970 the samples scroll from right to left
21972 frames are only produced when the samples reach the right
21974 the samples scroll from left to right
21977 Default value is @code{replace}.
21980 Specify display mode.
21982 It accepts the following values:
21985 all channels are displayed in the same row
21987 all channels are displayed in separate rows
21990 Default value is @samp{combined}.
21993 Specify display color mode.
21995 It accepts the following values:
21998 each channel is displayed in a separate color
22000 each channel is displayed using the same color scheme
22002 each channel is displayed using the rainbow color scheme
22004 each channel is displayed using the moreland color scheme
22006 each channel is displayed using the nebulae color scheme
22008 each channel is displayed using the fire color scheme
22010 each channel is displayed using the fiery color scheme
22012 each channel is displayed using the fruit color scheme
22014 each channel is displayed using the cool color scheme
22016 each channel is displayed using the magma color scheme
22018 each channel is displayed using the green color scheme
22020 each channel is displayed using the viridis color scheme
22022 each channel is displayed using the plasma color scheme
22024 each channel is displayed using the cividis color scheme
22026 each channel is displayed using the terrain color scheme
22029 Default value is @samp{channel}.
22032 Specify scale used for calculating intensity color values.
22034 It accepts the following values:
22039 square root, default
22050 Default value is @samp{sqrt}.
22053 Set saturation modifier for displayed colors. Negative values provide
22054 alternative color scheme. @code{0} is no saturation at all.
22055 Saturation must be in [-10.0, 10.0] range.
22056 Default value is @code{1}.
22059 Set window function.
22061 It accepts the following values:
22086 Default value is @code{hann}.
22089 Set orientation of time vs frequency axis. Can be @code{vertical} or
22090 @code{horizontal}. Default is @code{vertical}.
22093 Set ratio of overlap window. Default value is @code{0}.
22094 When value is @code{1} overlap is set to recommended size for specific
22095 window function currently used.
22098 Set scale gain for calculating intensity color values.
22099 Default value is @code{1}.
22102 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22105 Set color rotation, must be in [-1.0, 1.0] range.
22106 Default value is @code{0}.
22109 Set start frequency from which to display spectrogram. Default is @code{0}.
22112 Set stop frequency to which to display spectrogram. Default is @code{0}.
22115 Set upper frame rate limit. Default is @code{auto}, unlimited.
22118 Draw time and frequency axes and legends. Default is disabled.
22121 The usage is very similar to the showwaves filter; see the examples in that
22124 @subsection Examples
22128 Large window with logarithmic color scaling:
22130 showspectrum=s=1280x480:scale=log
22134 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22136 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22137 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22141 @section showspectrumpic
22143 Convert input audio to a single video frame, representing the audio frequency
22146 The filter accepts the following options:
22150 Specify the video size for the output. For the syntax of this option, check the
22151 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22152 Default value is @code{4096x2048}.
22155 Specify display mode.
22157 It accepts the following values:
22160 all channels are displayed in the same row
22162 all channels are displayed in separate rows
22164 Default value is @samp{combined}.
22167 Specify display color mode.
22169 It accepts the following values:
22172 each channel is displayed in a separate color
22174 each channel is displayed using the same color scheme
22176 each channel is displayed using the rainbow color scheme
22178 each channel is displayed using the moreland color scheme
22180 each channel is displayed using the nebulae color scheme
22182 each channel is displayed using the fire color scheme
22184 each channel is displayed using the fiery color scheme
22186 each channel is displayed using the fruit color scheme
22188 each channel is displayed using the cool color scheme
22190 each channel is displayed using the magma color scheme
22192 each channel is displayed using the green color scheme
22194 each channel is displayed using the viridis color scheme
22196 each channel is displayed using the plasma color scheme
22198 each channel is displayed using the cividis color scheme
22200 each channel is displayed using the terrain color scheme
22202 Default value is @samp{intensity}.
22205 Specify scale used for calculating intensity color values.
22207 It accepts the following values:
22212 square root, default
22222 Default value is @samp{log}.
22225 Set saturation modifier for displayed colors. Negative values provide
22226 alternative color scheme. @code{0} is no saturation at all.
22227 Saturation must be in [-10.0, 10.0] range.
22228 Default value is @code{1}.
22231 Set window function.
22233 It accepts the following values:
22257 Default value is @code{hann}.
22260 Set orientation of time vs frequency axis. Can be @code{vertical} or
22261 @code{horizontal}. Default is @code{vertical}.
22264 Set scale gain for calculating intensity color values.
22265 Default value is @code{1}.
22268 Draw time and frequency axes and legends. Default is enabled.
22271 Set color rotation, must be in [-1.0, 1.0] range.
22272 Default value is @code{0}.
22275 Set start frequency from which to display spectrogram. Default is @code{0}.
22278 Set stop frequency to which to display spectrogram. Default is @code{0}.
22281 @subsection Examples
22285 Extract an audio spectrogram of a whole audio track
22286 in a 1024x1024 picture using @command{ffmpeg}:
22288 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22292 @section showvolume
22294 Convert input audio volume to a video output.
22296 The filter accepts the following options:
22303 Set border width, allowed range is [0, 5]. Default is 1.
22306 Set channel width, allowed range is [80, 8192]. Default is 400.
22309 Set channel height, allowed range is [1, 900]. Default is 20.
22312 Set fade, allowed range is [0, 1]. Default is 0.95.
22315 Set volume color expression.
22317 The expression can use the following variables:
22321 Current max volume of channel in dB.
22327 Current channel number, starting from 0.
22331 If set, displays channel names. Default is enabled.
22334 If set, displays volume values. Default is enabled.
22337 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22338 default is @code{h}.
22341 Set step size, allowed range is [0, 5]. Default is 0, which means
22345 Set background opacity, allowed range is [0, 1]. Default is 0.
22348 Set metering mode, can be peak: @code{p} or rms: @code{r},
22349 default is @code{p}.
22352 Set display scale, can be linear: @code{lin} or log: @code{log},
22353 default is @code{lin}.
22357 If set to > 0., display a line for the max level
22358 in the previous seconds.
22359 default is disabled: @code{0.}
22362 The color of the max line. Use when @code{dm} option is set to > 0.
22363 default is: @code{orange}
22368 Convert input audio to a video output, representing the samples waves.
22370 The filter accepts the following options:
22374 Specify the video size for the output. For the syntax of this option, check the
22375 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22376 Default value is @code{600x240}.
22381 Available values are:
22384 Draw a point for each sample.
22387 Draw a vertical line for each sample.
22390 Draw a point for each sample and a line between them.
22393 Draw a centered vertical line for each sample.
22396 Default value is @code{point}.
22399 Set the number of samples which are printed on the same column. A
22400 larger value will decrease the frame rate. Must be a positive
22401 integer. This option can be set only if the value for @var{rate}
22402 is not explicitly specified.
22405 Set the (approximate) output frame rate. This is done by setting the
22406 option @var{n}. Default value is "25".
22408 @item split_channels
22409 Set if channels should be drawn separately or overlap. Default value is 0.
22412 Set colors separated by '|' which are going to be used for drawing of each channel.
22415 Set amplitude scale.
22417 Available values are:
22435 Set the draw mode. This is mostly useful to set for high @var{n}.
22437 Available values are:
22440 Scale pixel values for each drawn sample.
22443 Draw every sample directly.
22446 Default value is @code{scale}.
22449 @subsection Examples
22453 Output the input file audio and the corresponding video representation
22456 amovie=a.mp3,asplit[out0],showwaves[out1]
22460 Create a synthetic signal and show it with showwaves, forcing a
22461 frame rate of 30 frames per second:
22463 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22467 @section showwavespic
22469 Convert input audio to a single video frame, representing the samples waves.
22471 The filter accepts the following options:
22475 Specify the video size for the output. For the syntax of this option, check the
22476 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22477 Default value is @code{600x240}.
22479 @item split_channels
22480 Set if channels should be drawn separately or overlap. Default value is 0.
22483 Set colors separated by '|' which are going to be used for drawing of each channel.
22486 Set amplitude scale.
22488 Available values are:
22506 @subsection Examples
22510 Extract a channel split representation of the wave form of a whole audio track
22511 in a 1024x800 picture using @command{ffmpeg}:
22513 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22517 @section sidedata, asidedata
22519 Delete frame side data, or select frames based on it.
22521 This filter accepts the following options:
22525 Set mode of operation of the filter.
22527 Can be one of the following:
22531 Select every frame with side data of @code{type}.
22534 Delete side data of @code{type}. If @code{type} is not set, delete all side
22540 Set side data type used with all modes. Must be set for @code{select} mode. For
22541 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22542 in @file{libavutil/frame.h}. For example, to choose
22543 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22547 @section spectrumsynth
22549 Sythesize audio from 2 input video spectrums, first input stream represents
22550 magnitude across time and second represents phase across time.
22551 The filter will transform from frequency domain as displayed in videos back
22552 to time domain as presented in audio output.
22554 This filter is primarily created for reversing processed @ref{showspectrum}
22555 filter outputs, but can synthesize sound from other spectrograms too.
22556 But in such case results are going to be poor if the phase data is not
22557 available, because in such cases phase data need to be recreated, usually
22558 it's just recreated from random noise.
22559 For best results use gray only output (@code{channel} color mode in
22560 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22561 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22562 @code{data} option. Inputs videos should generally use @code{fullframe}
22563 slide mode as that saves resources needed for decoding video.
22565 The filter accepts the following options:
22569 Specify sample rate of output audio, the sample rate of audio from which
22570 spectrum was generated may differ.
22573 Set number of channels represented in input video spectrums.
22576 Set scale which was used when generating magnitude input spectrum.
22577 Can be @code{lin} or @code{log}. Default is @code{log}.
22580 Set slide which was used when generating inputs spectrums.
22581 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22582 Default is @code{fullframe}.
22585 Set window function used for resynthesis.
22588 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22589 which means optimal overlap for selected window function will be picked.
22592 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22593 Default is @code{vertical}.
22596 @subsection Examples
22600 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22601 then resynthesize videos back to audio with spectrumsynth:
22603 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
22604 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
22605 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22609 @section split, asplit
22611 Split input into several identical outputs.
22613 @code{asplit} works with audio input, @code{split} with video.
22615 The filter accepts a single parameter which specifies the number of outputs. If
22616 unspecified, it defaults to 2.
22618 @subsection Examples
22622 Create two separate outputs from the same input:
22624 [in] split [out0][out1]
22628 To create 3 or more outputs, you need to specify the number of
22631 [in] asplit=3 [out0][out1][out2]
22635 Create two separate outputs from the same input, one cropped and
22638 [in] split [splitout1][splitout2];
22639 [splitout1] crop=100:100:0:0 [cropout];
22640 [splitout2] pad=200:200:100:100 [padout];
22644 Create 5 copies of the input audio with @command{ffmpeg}:
22646 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22652 Receive commands sent through a libzmq client, and forward them to
22653 filters in the filtergraph.
22655 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22656 must be inserted between two video filters, @code{azmq} between two
22657 audio filters. Both are capable to send messages to any filter type.
22659 To enable these filters you need to install the libzmq library and
22660 headers and configure FFmpeg with @code{--enable-libzmq}.
22662 For more information about libzmq see:
22663 @url{http://www.zeromq.org/}
22665 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22666 receives messages sent through a network interface defined by the
22667 @option{bind_address} (or the abbreviation "@option{b}") option.
22668 Default value of this option is @file{tcp://localhost:5555}. You may
22669 want to alter this value to your needs, but do not forget to escape any
22670 ':' signs (see @ref{filtergraph escaping}).
22672 The received message must be in the form:
22674 @var{TARGET} @var{COMMAND} [@var{ARG}]
22677 @var{TARGET} specifies the target of the command, usually the name of
22678 the filter class or a specific filter instance name. The default
22679 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22680 but you can override this by using the @samp{filter_name@@id} syntax
22681 (see @ref{Filtergraph syntax}).
22683 @var{COMMAND} specifies the name of the command for the target filter.
22685 @var{ARG} is optional and specifies the optional argument list for the
22686 given @var{COMMAND}.
22688 Upon reception, the message is processed and the corresponding command
22689 is injected into the filtergraph. Depending on the result, the filter
22690 will send a reply to the client, adopting the format:
22692 @var{ERROR_CODE} @var{ERROR_REASON}
22696 @var{MESSAGE} is optional.
22698 @subsection Examples
22700 Look at @file{tools/zmqsend} for an example of a zmq client which can
22701 be used to send commands processed by these filters.
22703 Consider the following filtergraph generated by @command{ffplay}.
22704 In this example the last overlay filter has an instance name. All other
22705 filters will have default instance names.
22708 ffplay -dumpgraph 1 -f lavfi "
22709 color=s=100x100:c=red [l];
22710 color=s=100x100:c=blue [r];
22711 nullsrc=s=200x100, zmq [bg];
22712 [bg][l] overlay [bg+l];
22713 [bg+l][r] overlay@@my=x=100 "
22716 To change the color of the left side of the video, the following
22717 command can be used:
22719 echo Parsed_color_0 c yellow | tools/zmqsend
22722 To change the right side:
22724 echo Parsed_color_1 c pink | tools/zmqsend
22727 To change the position of the right side:
22729 echo overlay@@my x 150 | tools/zmqsend
22733 @c man end MULTIMEDIA FILTERS
22735 @chapter Multimedia Sources
22736 @c man begin MULTIMEDIA SOURCES
22738 Below is a description of the currently available multimedia sources.
22742 This is the same as @ref{movie} source, except it selects an audio
22748 Read audio and/or video stream(s) from a movie container.
22750 It accepts the following parameters:
22754 The name of the resource to read (not necessarily a file; it can also be a
22755 device or a stream accessed through some protocol).
22757 @item format_name, f
22758 Specifies the format assumed for the movie to read, and can be either
22759 the name of a container or an input device. If not specified, the
22760 format is guessed from @var{movie_name} or by probing.
22762 @item seek_point, sp
22763 Specifies the seek point in seconds. The frames will be output
22764 starting from this seek point. The parameter is evaluated with
22765 @code{av_strtod}, so the numerical value may be suffixed by an IS
22766 postfix. The default value is "0".
22769 Specifies the streams to read. Several streams can be specified,
22770 separated by "+". The source will then have as many outputs, in the
22771 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22772 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22773 respectively the default (best suited) video and audio stream. Default
22774 is "dv", or "da" if the filter is called as "amovie".
22776 @item stream_index, si
22777 Specifies the index of the video stream to read. If the value is -1,
22778 the most suitable video stream will be automatically selected. The default
22779 value is "-1". Deprecated. If the filter is called "amovie", it will select
22780 audio instead of video.
22783 Specifies how many times to read the stream in sequence.
22784 If the value is 0, the stream will be looped infinitely.
22785 Default value is "1".
22787 Note that when the movie is looped the source timestamps are not
22788 changed, so it will generate non monotonically increasing timestamps.
22790 @item discontinuity
22791 Specifies the time difference between frames above which the point is
22792 considered a timestamp discontinuity which is removed by adjusting the later
22796 It allows overlaying a second video on top of the main input of
22797 a filtergraph, as shown in this graph:
22799 input -----------> deltapts0 --> overlay --> output
22802 movie --> scale--> deltapts1 -------+
22804 @subsection Examples
22808 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
22809 on top of the input labelled "in":
22811 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
22812 [in] setpts=PTS-STARTPTS [main];
22813 [main][over] overlay=16:16 [out]
22817 Read from a video4linux2 device, and overlay it on top of the input
22820 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
22821 [in] setpts=PTS-STARTPTS [main];
22822 [main][over] overlay=16:16 [out]
22826 Read the first video stream and the audio stream with id 0x81 from
22827 dvd.vob; the video is connected to the pad named "video" and the audio is
22828 connected to the pad named "audio":
22830 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
22834 @subsection Commands
22836 Both movie and amovie support the following commands:
22839 Perform seek using "av_seek_frame".
22840 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
22843 @var{stream_index}: If stream_index is -1, a default
22844 stream is selected, and @var{timestamp} is automatically converted
22845 from AV_TIME_BASE units to the stream specific time_base.
22847 @var{timestamp}: Timestamp in AVStream.time_base units
22848 or, if no stream is specified, in AV_TIME_BASE units.
22850 @var{flags}: Flags which select direction and seeking mode.
22854 Get movie duration in AV_TIME_BASE units.
22858 @c man end MULTIMEDIA SOURCES