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 LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
4876 In @var{add} mode, LFE channel is created from input audio and added to output.
4877 In @var{sub} mode, LFE channel is created from input audio and added to output but
4878 also all non-LFE output channels are subtracted with output LFE channel.
4881 Set front center input volume. By default, this is @var{1}.
4884 Set front center output volume. By default, this is @var{1}.
4887 Set LFE input volume. By default, this is @var{1}.
4890 Set LFE output volume. By default, this is @var{1}.
4893 Set spread usage of stereo image across X axis for all channels.
4896 Set spread usage of stereo image across Y axis for all channels.
4898 @item fcx, flx, frx, blx, brx, slx, srx, bcx
4899 Set spread usage of stereo image across X axis for each channel.
4901 @item fcy, fly, fry, bly, bry, sly, sry, bcy
4902 Set spread usage of stereo image across Y axis for each channel.
4905 Set window function.
4907 It accepts the following values:
4930 Default is @code{hann}.
4933 Set window overlap. If set to 1, the recommended overlap for selected
4934 window function will be picked. Default is @code{0.5}.
4937 @section treble, highshelf
4939 Boost or cut treble (upper) frequencies of the audio using a two-pole
4940 shelving filter with a response similar to that of a standard
4941 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4943 The filter accepts the following options:
4947 Give the gain at whichever is the lower of ~22 kHz and the
4948 Nyquist frequency. Its useful range is about -20 (for a large cut)
4949 to +20 (for a large boost). Beware of clipping when using a positive gain.
4952 Set the filter's central frequency and so can be used
4953 to extend or reduce the frequency range to be boosted or cut.
4954 The default value is @code{3000} Hz.
4957 Set method to specify band-width of filter.
4972 Determine how steep is the filter's shelf transition.
4975 Specify which channels to filter, by default all available are filtered.
4978 @subsection Commands
4980 This filter supports the following commands:
4983 Change treble frequency.
4984 Syntax for the command is : "@var{frequency}"
4987 Change treble width_type.
4988 Syntax for the command is : "@var{width_type}"
4991 Change treble width.
4992 Syntax for the command is : "@var{width}"
4996 Syntax for the command is : "@var{gain}"
5001 Sinusoidal amplitude modulation.
5003 The filter accepts the following options:
5007 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
5008 (20 Hz or lower) will result in a tremolo effect.
5009 This filter may also be used as a ring modulator by specifying
5010 a modulation frequency higher than 20 Hz.
5011 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5014 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5015 Default value is 0.5.
5020 Sinusoidal phase modulation.
5022 The filter accepts the following options:
5026 Modulation frequency in Hertz.
5027 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5030 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5031 Default value is 0.5.
5036 Adjust the input audio volume.
5038 It accepts the following parameters:
5042 Set audio volume expression.
5044 Output values are clipped to the maximum value.
5046 The output audio volume is given by the relation:
5048 @var{output_volume} = @var{volume} * @var{input_volume}
5051 The default value for @var{volume} is "1.0".
5054 This parameter represents the mathematical precision.
5056 It determines which input sample formats will be allowed, which affects the
5057 precision of the volume scaling.
5061 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
5063 32-bit floating-point; this limits input sample format to FLT. (default)
5065 64-bit floating-point; this limits input sample format to DBL.
5069 Choose the behaviour on encountering ReplayGain side data in input frames.
5073 Remove ReplayGain side data, ignoring its contents (the default).
5076 Ignore ReplayGain side data, but leave it in the frame.
5079 Prefer the track gain, if present.
5082 Prefer the album gain, if present.
5085 @item replaygain_preamp
5086 Pre-amplification gain in dB to apply to the selected replaygain gain.
5088 Default value for @var{replaygain_preamp} is 0.0.
5091 Set when the volume expression is evaluated.
5093 It accepts the following values:
5096 only evaluate expression once during the filter initialization, or
5097 when the @samp{volume} command is sent
5100 evaluate expression for each incoming frame
5103 Default value is @samp{once}.
5106 The volume expression can contain the following parameters.
5110 frame number (starting at zero)
5113 @item nb_consumed_samples
5114 number of samples consumed by the filter
5116 number of samples in the current frame
5118 original frame position in the file
5124 PTS at start of stream
5126 time at start of stream
5132 last set volume value
5135 Note that when @option{eval} is set to @samp{once} only the
5136 @var{sample_rate} and @var{tb} variables are available, all other
5137 variables will evaluate to NAN.
5139 @subsection Commands
5141 This filter supports the following commands:
5144 Modify the volume expression.
5145 The command accepts the same syntax of the corresponding option.
5147 If the specified expression is not valid, it is kept at its current
5149 @item replaygain_noclip
5150 Prevent clipping by limiting the gain applied.
5152 Default value for @var{replaygain_noclip} is 1.
5156 @subsection Examples
5160 Halve the input audio volume:
5164 volume=volume=-6.0206dB
5167 In all the above example the named key for @option{volume} can be
5168 omitted, for example like in:
5174 Increase input audio power by 6 decibels using fixed-point precision:
5176 volume=volume=6dB:precision=fixed
5180 Fade volume after time 10 with an annihilation period of 5 seconds:
5182 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5186 @section volumedetect
5188 Detect the volume of the input video.
5190 The filter has no parameters. The input is not modified. Statistics about
5191 the volume will be printed in the log when the input stream end is reached.
5193 In particular it will show the mean volume (root mean square), maximum
5194 volume (on a per-sample basis), and the beginning of a histogram of the
5195 registered volume values (from the maximum value to a cumulated 1/1000 of
5198 All volumes are in decibels relative to the maximum PCM value.
5200 @subsection Examples
5202 Here is an excerpt of the output:
5204 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5205 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5206 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5207 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5208 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5209 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5210 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5211 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5212 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5218 The mean square energy is approximately -27 dB, or 10^-2.7.
5220 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5222 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5225 In other words, raising the volume by +4 dB does not cause any clipping,
5226 raising it by +5 dB causes clipping for 6 samples, etc.
5228 @c man end AUDIO FILTERS
5230 @chapter Audio Sources
5231 @c man begin AUDIO SOURCES
5233 Below is a description of the currently available audio sources.
5237 Buffer audio frames, and make them available to the filter chain.
5239 This source is mainly intended for a programmatic use, in particular
5240 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5242 It accepts the following parameters:
5246 The timebase which will be used for timestamps of submitted frames. It must be
5247 either a floating-point number or in @var{numerator}/@var{denominator} form.
5250 The sample rate of the incoming audio buffers.
5253 The sample format of the incoming audio buffers.
5254 Either a sample format name or its corresponding integer representation from
5255 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5257 @item channel_layout
5258 The channel layout of the incoming audio buffers.
5259 Either a channel layout name from channel_layout_map in
5260 @file{libavutil/channel_layout.c} or its corresponding integer representation
5261 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5264 The number of channels of the incoming audio buffers.
5265 If both @var{channels} and @var{channel_layout} are specified, then they
5270 @subsection Examples
5273 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5276 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5277 Since the sample format with name "s16p" corresponds to the number
5278 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5281 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5286 Generate an audio signal specified by an expression.
5288 This source accepts in input one or more expressions (one for each
5289 channel), which are evaluated and used to generate a corresponding
5292 This source accepts the following options:
5296 Set the '|'-separated expressions list for each separate channel. In case the
5297 @option{channel_layout} option is not specified, the selected channel layout
5298 depends on the number of provided expressions. Otherwise the last
5299 specified expression is applied to the remaining output channels.
5301 @item channel_layout, c
5302 Set the channel layout. The number of channels in the specified layout
5303 must be equal to the number of specified expressions.
5306 Set the minimum duration of the sourced audio. See
5307 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5308 for the accepted syntax.
5309 Note that the resulting duration may be greater than the specified
5310 duration, as the generated audio is always cut at the end of a
5313 If not specified, or the expressed duration is negative, the audio is
5314 supposed to be generated forever.
5317 Set the number of samples per channel per each output frame,
5320 @item sample_rate, s
5321 Specify the sample rate, default to 44100.
5324 Each expression in @var{exprs} can contain the following constants:
5328 number of the evaluated sample, starting from 0
5331 time of the evaluated sample expressed in seconds, starting from 0
5338 @subsection Examples
5348 Generate a sin signal with frequency of 440 Hz, set sample rate to
5351 aevalsrc="sin(440*2*PI*t):s=8000"
5355 Generate a two channels signal, specify the channel layout (Front
5356 Center + Back Center) explicitly:
5358 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5362 Generate white noise:
5364 aevalsrc="-2+random(0)"
5368 Generate an amplitude modulated signal:
5370 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5374 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5376 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5383 The null audio source, return unprocessed audio frames. It is mainly useful
5384 as a template and to be employed in analysis / debugging tools, or as
5385 the source for filters which ignore the input data (for example the sox
5388 This source accepts the following options:
5392 @item channel_layout, cl
5394 Specifies the channel layout, and can be either an integer or a string
5395 representing a channel layout. The default value of @var{channel_layout}
5398 Check the channel_layout_map definition in
5399 @file{libavutil/channel_layout.c} for the mapping between strings and
5400 channel layout values.
5402 @item sample_rate, r
5403 Specifies the sample rate, and defaults to 44100.
5406 Set the number of samples per requested frames.
5410 @subsection Examples
5414 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5416 anullsrc=r=48000:cl=4
5420 Do the same operation with a more obvious syntax:
5422 anullsrc=r=48000:cl=mono
5426 All the parameters need to be explicitly defined.
5430 Synthesize a voice utterance using the libflite library.
5432 To enable compilation of this filter you need to configure FFmpeg with
5433 @code{--enable-libflite}.
5435 Note that versions of the flite library prior to 2.0 are not thread-safe.
5437 The filter accepts the following options:
5442 If set to 1, list the names of the available voices and exit
5443 immediately. Default value is 0.
5446 Set the maximum number of samples per frame. Default value is 512.
5449 Set the filename containing the text to speak.
5452 Set the text to speak.
5455 Set the voice to use for the speech synthesis. Default value is
5456 @code{kal}. See also the @var{list_voices} option.
5459 @subsection Examples
5463 Read from file @file{speech.txt}, and synthesize the text using the
5464 standard flite voice:
5466 flite=textfile=speech.txt
5470 Read the specified text selecting the @code{slt} voice:
5472 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5476 Input text to ffmpeg:
5478 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5482 Make @file{ffplay} speak the specified text, using @code{flite} and
5483 the @code{lavfi} device:
5485 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5489 For more information about libflite, check:
5490 @url{http://www.festvox.org/flite/}
5494 Generate a noise audio signal.
5496 The filter accepts the following options:
5499 @item sample_rate, r
5500 Specify the sample rate. Default value is 48000 Hz.
5503 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5507 Specify the duration of the generated audio stream. Not specifying this option
5508 results in noise with an infinite length.
5510 @item color, colour, c
5511 Specify the color of noise. Available noise colors are white, pink, brown,
5512 blue and violet. Default color is white.
5515 Specify a value used to seed the PRNG.
5518 Set the number of samples per each output frame, default is 1024.
5521 @subsection Examples
5526 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5528 anoisesrc=d=60:c=pink:r=44100:a=0.5
5534 Generate odd-tap Hilbert transform FIR coefficients.
5536 The resulting stream can be used with @ref{afir} filter for phase-shifting
5537 the signal by 90 degrees.
5539 This is used in many matrix coding schemes and for analytic signal generation.
5540 The process is often written as a multiplication by i (or j), the imaginary unit.
5542 The filter accepts the following options:
5546 @item sample_rate, s
5547 Set sample rate, default is 44100.
5550 Set length of FIR filter, default is 22051.
5553 Set number of samples per each frame.
5556 Set window function to be used when generating FIR coefficients.
5561 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5563 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5565 The filter accepts the following options:
5568 @item sample_rate, r
5569 Set sample rate, default is 44100.
5572 Set number of samples per each frame. Default is 1024.
5575 Set high-pass frequency. Default is 0.
5578 Set low-pass frequency. Default is 0.
5579 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5580 is higher than 0 then filter will create band-pass filter coefficients,
5581 otherwise band-reject filter coefficients.
5584 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5587 Set Kaiser window beta.
5590 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5593 Enable rounding, by default is disabled.
5596 Set number of taps for high-pass filter.
5599 Set number of taps for low-pass filter.
5604 Generate an audio signal made of a sine wave with amplitude 1/8.
5606 The audio signal is bit-exact.
5608 The filter accepts the following options:
5613 Set the carrier frequency. Default is 440 Hz.
5615 @item beep_factor, b
5616 Enable a periodic beep every second with frequency @var{beep_factor} times
5617 the carrier frequency. Default is 0, meaning the beep is disabled.
5619 @item sample_rate, r
5620 Specify the sample rate, default is 44100.
5623 Specify the duration of the generated audio stream.
5625 @item samples_per_frame
5626 Set the number of samples per output frame.
5628 The expression can contain the following constants:
5632 The (sequential) number of the output audio frame, starting from 0.
5635 The PTS (Presentation TimeStamp) of the output audio frame,
5636 expressed in @var{TB} units.
5639 The PTS of the output audio frame, expressed in seconds.
5642 The timebase of the output audio frames.
5645 Default is @code{1024}.
5648 @subsection Examples
5653 Generate a simple 440 Hz sine wave:
5659 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5663 sine=frequency=220:beep_factor=4:duration=5
5667 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5670 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5674 @c man end AUDIO SOURCES
5676 @chapter Audio Sinks
5677 @c man begin AUDIO SINKS
5679 Below is a description of the currently available audio sinks.
5681 @section abuffersink
5683 Buffer audio frames, and make them available to the end of filter chain.
5685 This sink is mainly intended for programmatic use, in particular
5686 through the interface defined in @file{libavfilter/buffersink.h}
5687 or the options system.
5689 It accepts a pointer to an AVABufferSinkContext structure, which
5690 defines the incoming buffers' formats, to be passed as the opaque
5691 parameter to @code{avfilter_init_filter} for initialization.
5694 Null audio sink; do absolutely nothing with the input audio. It is
5695 mainly useful as a template and for use in analysis / debugging
5698 @c man end AUDIO SINKS
5700 @chapter Video Filters
5701 @c man begin VIDEO FILTERS
5703 When you configure your FFmpeg build, you can disable any of the
5704 existing filters using @code{--disable-filters}.
5705 The configure output will show the video filters included in your
5708 Below is a description of the currently available video filters.
5710 @section alphaextract
5712 Extract the alpha component from the input as a grayscale video. This
5713 is especially useful with the @var{alphamerge} filter.
5717 Add or replace the alpha component of the primary input with the
5718 grayscale value of a second input. This is intended for use with
5719 @var{alphaextract} to allow the transmission or storage of frame
5720 sequences that have alpha in a format that doesn't support an alpha
5723 For example, to reconstruct full frames from a normal YUV-encoded video
5724 and a separate video created with @var{alphaextract}, you might use:
5726 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5729 Since this filter is designed for reconstruction, it operates on frame
5730 sequences without considering timestamps, and terminates when either
5731 input reaches end of stream. This will cause problems if your encoding
5732 pipeline drops frames. If you're trying to apply an image as an
5733 overlay to a video stream, consider the @var{overlay} filter instead.
5737 Amplify differences between current pixel and pixels of adjacent frames in
5738 same pixel location.
5740 This filter accepts the following options:
5744 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5745 For example radius of 3 will instruct filter to calculate average of 7 frames.
5748 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5751 Set threshold for difference amplification. Any difference greater or equal to
5752 this value will not alter source pixel. Default is 10.
5753 Allowed range is from 0 to 65535.
5756 Set tolerance for difference amplification. Any difference lower to
5757 this value will not alter source pixel. Default is 0.
5758 Allowed range is from 0 to 65535.
5761 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5762 This option controls maximum possible value that will decrease source pixel value.
5765 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5766 This option controls maximum possible value that will increase source pixel value.
5769 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5774 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5775 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5776 Substation Alpha) subtitles files.
5778 This filter accepts the following option in addition to the common options from
5779 the @ref{subtitles} filter:
5783 Set the shaping engine
5785 Available values are:
5788 The default libass shaping engine, which is the best available.
5790 Fast, font-agnostic shaper that can do only substitutions
5792 Slower shaper using OpenType for substitutions and positioning
5795 The default is @code{auto}.
5799 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5801 The filter accepts the following options:
5805 Set threshold A for 1st plane. Default is 0.02.
5806 Valid range is 0 to 0.3.
5809 Set threshold B for 1st plane. Default is 0.04.
5810 Valid range is 0 to 5.
5813 Set threshold A for 2nd plane. Default is 0.02.
5814 Valid range is 0 to 0.3.
5817 Set threshold B for 2nd plane. Default is 0.04.
5818 Valid range is 0 to 5.
5821 Set threshold A for 3rd plane. Default is 0.02.
5822 Valid range is 0 to 0.3.
5825 Set threshold B for 3rd plane. Default is 0.04.
5826 Valid range is 0 to 5.
5828 Threshold A is designed to react on abrupt changes in the input signal and
5829 threshold B is designed to react on continuous changes in the input signal.
5832 Set number of frames filter will use for averaging. Default is 9. Must be odd
5833 number in range [5, 129].
5836 Set what planes of frame filter will use for averaging. Default is all.
5841 Apply average blur filter.
5843 The filter accepts the following options:
5847 Set horizontal radius size.
5850 Set which planes to filter. By default all planes are filtered.
5853 Set vertical radius size, if zero it will be same as @code{sizeX}.
5854 Default is @code{0}.
5859 Compute the bounding box for the non-black pixels in the input frame
5862 This filter computes the bounding box containing all the pixels with a
5863 luminance value greater than the minimum allowed value.
5864 The parameters describing the bounding box are printed on the filter
5867 The filter accepts the following option:
5871 Set the minimal luminance value. Default is @code{16}.
5874 @section bitplanenoise
5876 Show and measure bit plane noise.
5878 The filter accepts the following options:
5882 Set which plane to analyze. Default is @code{1}.
5885 Filter out noisy pixels from @code{bitplane} set above.
5886 Default is disabled.
5889 @section blackdetect
5891 Detect video intervals that are (almost) completely black. Can be
5892 useful to detect chapter transitions, commercials, or invalid
5893 recordings. Output lines contains the time for the start, end and
5894 duration of the detected black interval expressed in seconds.
5896 In order to display the output lines, you need to set the loglevel at
5897 least to the AV_LOG_INFO value.
5899 The filter accepts the following options:
5902 @item black_min_duration, d
5903 Set the minimum detected black duration expressed in seconds. It must
5904 be a non-negative floating point number.
5906 Default value is 2.0.
5908 @item picture_black_ratio_th, pic_th
5909 Set the threshold for considering a picture "black".
5910 Express the minimum value for the ratio:
5912 @var{nb_black_pixels} / @var{nb_pixels}
5915 for which a picture is considered black.
5916 Default value is 0.98.
5918 @item pixel_black_th, pix_th
5919 Set the threshold for considering a pixel "black".
5921 The threshold expresses the maximum pixel luminance value for which a
5922 pixel is considered "black". The provided value is scaled according to
5923 the following equation:
5925 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5928 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5929 the input video format, the range is [0-255] for YUV full-range
5930 formats and [16-235] for YUV non full-range formats.
5932 Default value is 0.10.
5935 The following example sets the maximum pixel threshold to the minimum
5936 value, and detects only black intervals of 2 or more seconds:
5938 blackdetect=d=2:pix_th=0.00
5943 Detect frames that are (almost) completely black. Can be useful to
5944 detect chapter transitions or commercials. Output lines consist of
5945 the frame number of the detected frame, the percentage of blackness,
5946 the position in the file if known or -1 and the timestamp in seconds.
5948 In order to display the output lines, you need to set the loglevel at
5949 least to the AV_LOG_INFO value.
5951 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5952 The value represents the percentage of pixels in the picture that
5953 are below the threshold value.
5955 It accepts the following parameters:
5960 The percentage of the pixels that have to be below the threshold; it defaults to
5963 @item threshold, thresh
5964 The threshold below which a pixel value is considered black; it defaults to
5969 @section blend, tblend
5971 Blend two video frames into each other.
5973 The @code{blend} filter takes two input streams and outputs one
5974 stream, the first input is the "top" layer and second input is
5975 "bottom" layer. By default, the output terminates when the longest input terminates.
5977 The @code{tblend} (time blend) filter takes two consecutive frames
5978 from one single stream, and outputs the result obtained by blending
5979 the new frame on top of the old frame.
5981 A description of the accepted options follows.
5989 Set blend mode for specific pixel component or all pixel components in case
5990 of @var{all_mode}. Default value is @code{normal}.
5992 Available values for component modes are:
6034 Set blend opacity for specific pixel component or all pixel components in case
6035 of @var{all_opacity}. Only used in combination with pixel component blend modes.
6042 Set blend expression for specific pixel component or all pixel components in case
6043 of @var{all_expr}. Note that related mode options will be ignored if those are set.
6045 The expressions can use the following variables:
6049 The sequential number of the filtered frame, starting from @code{0}.
6053 the coordinates of the current sample
6057 the width and height of currently filtered plane
6061 Width and height scale for the plane being filtered. It is the
6062 ratio between the dimensions of the current plane to the luma plane,
6063 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
6064 the luma plane and @code{0.5,0.5} for the chroma planes.
6067 Time of the current frame, expressed in seconds.
6070 Value of pixel component at current location for first video frame (top layer).
6073 Value of pixel component at current location for second video frame (bottom layer).
6077 The @code{blend} filter also supports the @ref{framesync} options.
6079 @subsection Examples
6083 Apply transition from bottom layer to top layer in first 10 seconds:
6085 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6089 Apply linear horizontal transition from top layer to bottom layer:
6091 blend=all_expr='A*(X/W)+B*(1-X/W)'
6095 Apply 1x1 checkerboard effect:
6097 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6101 Apply uncover left effect:
6103 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6107 Apply uncover down effect:
6109 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6113 Apply uncover up-left effect:
6115 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6119 Split diagonally video and shows top and bottom layer on each side:
6121 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6125 Display differences between the current and the previous frame:
6127 tblend=all_mode=grainextract
6133 Denoise frames using Block-Matching 3D algorithm.
6135 The filter accepts the following options.
6139 Set denoising strength. Default value is 1.
6140 Allowed range is from 0 to 999.9.
6141 The denoising algorithm is very sensitive to sigma, so adjust it
6142 according to the source.
6145 Set local patch size. This sets dimensions in 2D.
6148 Set sliding step for processing blocks. Default value is 4.
6149 Allowed range is from 1 to 64.
6150 Smaller values allows processing more reference blocks and is slower.
6153 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6154 When set to 1, no block matching is done. Larger values allows more blocks
6156 Allowed range is from 1 to 256.
6159 Set radius for search block matching. Default is 9.
6160 Allowed range is from 1 to INT32_MAX.
6163 Set step between two search locations for block matching. Default is 1.
6164 Allowed range is from 1 to 64. Smaller is slower.
6167 Set threshold of mean square error for block matching. Valid range is 0 to
6171 Set thresholding parameter for hard thresholding in 3D transformed domain.
6172 Larger values results in stronger hard-thresholding filtering in frequency
6176 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6177 Default is @code{basic}.
6180 If enabled, filter will use 2nd stream for block matching.
6181 Default is disabled for @code{basic} value of @var{estim} option,
6182 and always enabled if value of @var{estim} is @code{final}.
6185 Set planes to filter. Default is all available except alpha.
6188 @subsection Examples
6192 Basic filtering with bm3d:
6194 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6198 Same as above, but filtering only luma:
6200 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6204 Same as above, but with both estimation modes:
6206 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
6210 Same as above, but prefilter with @ref{nlmeans} filter instead:
6212 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
6218 Apply a boxblur algorithm to the input video.
6220 It accepts the following parameters:
6224 @item luma_radius, lr
6225 @item luma_power, lp
6226 @item chroma_radius, cr
6227 @item chroma_power, cp
6228 @item alpha_radius, ar
6229 @item alpha_power, ap
6233 A description of the accepted options follows.
6236 @item luma_radius, lr
6237 @item chroma_radius, cr
6238 @item alpha_radius, ar
6239 Set an expression for the box radius in pixels used for blurring the
6240 corresponding input plane.
6242 The radius value must be a non-negative number, and must not be
6243 greater than the value of the expression @code{min(w,h)/2} for the
6244 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6247 Default value for @option{luma_radius} is "2". If not specified,
6248 @option{chroma_radius} and @option{alpha_radius} default to the
6249 corresponding value set for @option{luma_radius}.
6251 The expressions can contain the following constants:
6255 The input width and height in pixels.
6259 The input chroma image width and height in pixels.
6263 The horizontal and vertical chroma subsample values. For example, for the
6264 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6267 @item luma_power, lp
6268 @item chroma_power, cp
6269 @item alpha_power, ap
6270 Specify how many times the boxblur filter is applied to the
6271 corresponding plane.
6273 Default value for @option{luma_power} is 2. If not specified,
6274 @option{chroma_power} and @option{alpha_power} default to the
6275 corresponding value set for @option{luma_power}.
6277 A value of 0 will disable the effect.
6280 @subsection Examples
6284 Apply a boxblur filter with the luma, chroma, and alpha radii
6287 boxblur=luma_radius=2:luma_power=1
6292 Set the luma radius to 2, and alpha and chroma radius to 0:
6294 boxblur=2:1:cr=0:ar=0
6298 Set the luma and chroma radii to a fraction of the video dimension:
6300 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6306 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6307 Deinterlacing Filter").
6309 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6310 interpolation algorithms.
6311 It accepts the following parameters:
6315 The interlacing mode to adopt. It accepts one of the following values:
6319 Output one frame for each frame.
6321 Output one frame for each field.
6324 The default value is @code{send_field}.
6327 The picture field parity assumed for the input interlaced video. It accepts one
6328 of the following values:
6332 Assume the top field is first.
6334 Assume the bottom field is first.
6336 Enable automatic detection of field parity.
6339 The default value is @code{auto}.
6340 If the interlacing is unknown or the decoder does not export this information,
6341 top field first will be assumed.
6344 Specify which frames to deinterlace. Accept one of the following
6349 Deinterlace all frames.
6351 Only deinterlace frames marked as interlaced.
6354 The default value is @code{all}.
6358 Remove all color information for all colors except for certain one.
6360 The filter accepts the following options:
6364 The color which will not be replaced with neutral chroma.
6367 Similarity percentage with the above color.
6368 0.01 matches only the exact key color, while 1.0 matches everything.
6371 Signals that the color passed is already in YUV instead of RGB.
6373 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6374 This can be used to pass exact YUV values as hexadecimal numbers.
6378 YUV colorspace color/chroma keying.
6380 The filter accepts the following options:
6384 The color which will be replaced with transparency.
6387 Similarity percentage with the key color.
6389 0.01 matches only the exact key color, while 1.0 matches everything.
6394 0.0 makes pixels either fully transparent, or not transparent at all.
6396 Higher values result in semi-transparent pixels, with a higher transparency
6397 the more similar the pixels color is to the key color.
6400 Signals that the color passed is already in YUV instead of RGB.
6402 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6403 This can be used to pass exact YUV values as hexadecimal numbers.
6406 @subsection Examples
6410 Make every green pixel in the input image transparent:
6412 ffmpeg -i input.png -vf chromakey=green out.png
6416 Overlay a greenscreen-video on top of a static black background.
6418 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
6422 @section chromashift
6423 Shift chroma pixels horizontally and/or vertically.
6425 The filter accepts the following options:
6428 Set amount to shift chroma-blue horizontally.
6430 Set amount to shift chroma-blue vertically.
6432 Set amount to shift chroma-red horizontally.
6434 Set amount to shift chroma-red vertically.
6436 Set edge mode, can be @var{smear}, default, or @var{warp}.
6441 Display CIE color diagram with pixels overlaid onto it.
6443 The filter accepts the following options:
6458 @item uhdtv, rec2020
6471 Set what gamuts to draw.
6473 See @code{system} option for available values.
6476 Set ciescope size, by default set to 512.
6479 Set intensity used to map input pixel values to CIE diagram.
6482 Set contrast used to draw tongue colors that are out of active color system gamut.
6485 Correct gamma displayed on scope, by default enabled.
6488 Show white point on CIE diagram, by default disabled.
6491 Set input gamma. Used only with XYZ input color space.
6496 Visualize information exported by some codecs.
6498 Some codecs can export information through frames using side-data or other
6499 means. For example, some MPEG based codecs export motion vectors through the
6500 @var{export_mvs} flag in the codec @option{flags2} option.
6502 The filter accepts the following option:
6506 Set motion vectors to visualize.
6508 Available flags for @var{mv} are:
6512 forward predicted MVs of P-frames
6514 forward predicted MVs of B-frames
6516 backward predicted MVs of B-frames
6520 Display quantization parameters using the chroma planes.
6523 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6525 Available flags for @var{mv_type} are:
6529 forward predicted MVs
6531 backward predicted MVs
6534 @item frame_type, ft
6535 Set frame type to visualize motion vectors of.
6537 Available flags for @var{frame_type} are:
6541 intra-coded frames (I-frames)
6543 predicted frames (P-frames)
6545 bi-directionally predicted frames (B-frames)
6549 @subsection Examples
6553 Visualize forward predicted MVs of all frames using @command{ffplay}:
6555 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6559 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6561 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6565 @section colorbalance
6566 Modify intensity of primary colors (red, green and blue) of input frames.
6568 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6569 regions for the red-cyan, green-magenta or blue-yellow balance.
6571 A positive adjustment value shifts the balance towards the primary color, a negative
6572 value towards the complementary color.
6574 The filter accepts the following options:
6580 Adjust red, green and blue shadows (darkest pixels).
6585 Adjust red, green and blue midtones (medium pixels).
6590 Adjust red, green and blue highlights (brightest pixels).
6592 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6595 @subsection Examples
6599 Add red color cast to shadows:
6606 RGB colorspace color keying.
6608 The filter accepts the following options:
6612 The color which will be replaced with transparency.
6615 Similarity percentage with the key color.
6617 0.01 matches only the exact key color, while 1.0 matches everything.
6622 0.0 makes pixels either fully transparent, or not transparent at all.
6624 Higher values result in semi-transparent pixels, with a higher transparency
6625 the more similar the pixels color is to the key color.
6628 @subsection Examples
6632 Make every green pixel in the input image transparent:
6634 ffmpeg -i input.png -vf colorkey=green out.png
6638 Overlay a greenscreen-video on top of a static background image.
6640 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
6644 @section colorlevels
6646 Adjust video input frames using levels.
6648 The filter accepts the following options:
6655 Adjust red, green, blue and alpha input black point.
6656 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6662 Adjust red, green, blue and alpha input white point.
6663 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6665 Input levels are used to lighten highlights (bright tones), darken shadows
6666 (dark tones), change the balance of bright and dark tones.
6672 Adjust red, green, blue and alpha output black point.
6673 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6679 Adjust red, green, blue and alpha output white point.
6680 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6682 Output levels allows manual selection of a constrained output level range.
6685 @subsection Examples
6689 Make video output darker:
6691 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6697 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6701 Make video output lighter:
6703 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6707 Increase brightness:
6709 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6713 @section colorchannelmixer
6715 Adjust video input frames by re-mixing color channels.
6717 This filter modifies a color channel by adding the values associated to
6718 the other channels of the same pixels. For example if the value to
6719 modify is red, the output value will be:
6721 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6724 The filter accepts the following options:
6731 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6732 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6738 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6739 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6745 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6746 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6752 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6753 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6755 Allowed ranges for options are @code{[-2.0, 2.0]}.
6758 @subsection Examples
6762 Convert source to grayscale:
6764 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6767 Simulate sepia tones:
6769 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6773 @section colormatrix
6775 Convert color matrix.
6777 The filter accepts the following options:
6782 Specify the source and destination color matrix. Both values must be
6785 The accepted values are:
6813 For example to convert from BT.601 to SMPTE-240M, use the command:
6815 colormatrix=bt601:smpte240m
6820 Convert colorspace, transfer characteristics or color primaries.
6821 Input video needs to have an even size.
6823 The filter accepts the following options:
6828 Specify all color properties at once.
6830 The accepted values are:
6860 Specify output colorspace.
6862 The accepted values are:
6871 BT.470BG or BT.601-6 625
6874 SMPTE-170M or BT.601-6 525
6883 BT.2020 with non-constant luminance
6889 Specify output transfer characteristics.
6891 The accepted values are:
6903 Constant gamma of 2.2
6906 Constant gamma of 2.8
6909 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6927 BT.2020 for 10-bits content
6930 BT.2020 for 12-bits content
6936 Specify output color primaries.
6938 The accepted values are:
6947 BT.470BG or BT.601-6 625
6950 SMPTE-170M or BT.601-6 525
6974 Specify output color range.
6976 The accepted values are:
6979 TV (restricted) range
6982 MPEG (restricted) range
6993 Specify output color format.
6995 The accepted values are:
6998 YUV 4:2:0 planar 8-bits
7001 YUV 4:2:0 planar 10-bits
7004 YUV 4:2:0 planar 12-bits
7007 YUV 4:2:2 planar 8-bits
7010 YUV 4:2:2 planar 10-bits
7013 YUV 4:2:2 planar 12-bits
7016 YUV 4:4:4 planar 8-bits
7019 YUV 4:4:4 planar 10-bits
7022 YUV 4:4:4 planar 12-bits
7027 Do a fast conversion, which skips gamma/primary correction. This will take
7028 significantly less CPU, but will be mathematically incorrect. To get output
7029 compatible with that produced by the colormatrix filter, use fast=1.
7032 Specify dithering mode.
7034 The accepted values are:
7040 Floyd-Steinberg dithering
7044 Whitepoint adaptation mode.
7046 The accepted values are:
7049 Bradford whitepoint adaptation
7052 von Kries whitepoint adaptation
7055 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7059 Override all input properties at once. Same accepted values as @ref{all}.
7062 Override input colorspace. Same accepted values as @ref{space}.
7065 Override input color primaries. Same accepted values as @ref{primaries}.
7068 Override input transfer characteristics. Same accepted values as @ref{trc}.
7071 Override input color range. Same accepted values as @ref{range}.
7075 The filter converts the transfer characteristics, color space and color
7076 primaries to the specified user values. The output value, if not specified,
7077 is set to a default value based on the "all" property. If that property is
7078 also not specified, the filter will log an error. The output color range and
7079 format default to the same value as the input color range and format. The
7080 input transfer characteristics, color space, color primaries and color range
7081 should be set on the input data. If any of these are missing, the filter will
7082 log an error and no conversion will take place.
7084 For example to convert the input to SMPTE-240M, use the command:
7086 colorspace=smpte240m
7089 @section convolution
7091 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7093 The filter accepts the following options:
7100 Set matrix for each plane.
7101 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7102 and from 1 to 49 odd number of signed integers in @var{row} mode.
7108 Set multiplier for calculated value for each plane.
7109 If unset or 0, it will be sum of all matrix elements.
7115 Set bias for each plane. This value is added to the result of the multiplication.
7116 Useful for making the overall image brighter or darker. Default is 0.0.
7122 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7123 Default is @var{square}.
7126 @subsection Examples
7132 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"
7138 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"
7144 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"
7150 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"
7154 Apply laplacian edge detector which includes diagonals:
7156 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"
7162 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"
7168 Apply 2D convolution of video stream in frequency domain using second stream
7171 The filter accepts the following options:
7175 Set which planes to process.
7178 Set which impulse video frames will be processed, can be @var{first}
7179 or @var{all}. Default is @var{all}.
7182 The @code{convolve} filter also supports the @ref{framesync} options.
7186 Copy the input video source unchanged to the output. This is mainly useful for
7191 Video filtering on GPU using Apple's CoreImage API on OSX.
7193 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7194 processed by video hardware. However, software-based OpenGL implementations
7195 exist which means there is no guarantee for hardware processing. It depends on
7198 There are many filters and image generators provided by Apple that come with a
7199 large variety of options. The filter has to be referenced by its name along
7202 The coreimage filter accepts the following options:
7205 List all available filters and generators along with all their respective
7206 options as well as possible minimum and maximum values along with the default
7213 Specify all filters by their respective name and options.
7214 Use @var{list_filters} to determine all valid filter names and options.
7215 Numerical options are specified by a float value and are automatically clamped
7216 to their respective value range. Vector and color options have to be specified
7217 by a list of space separated float values. Character escaping has to be done.
7218 A special option name @code{default} is available to use default options for a
7221 It is required to specify either @code{default} or at least one of the filter options.
7222 All omitted options are used with their default values.
7223 The syntax of the filter string is as follows:
7225 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7229 Specify a rectangle where the output of the filter chain is copied into the
7230 input image. It is given by a list of space separated float values:
7232 output_rect=x\ y\ width\ height
7234 If not given, the output rectangle equals the dimensions of the input image.
7235 The output rectangle is automatically cropped at the borders of the input
7236 image. Negative values are valid for each component.
7238 output_rect=25\ 25\ 100\ 100
7242 Several filters can be chained for successive processing without GPU-HOST
7243 transfers allowing for fast processing of complex filter chains.
7244 Currently, only filters with zero (generators) or exactly one (filters) input
7245 image and one output image are supported. Also, transition filters are not yet
7248 Some filters generate output images with additional padding depending on the
7249 respective filter kernel. The padding is automatically removed to ensure the
7250 filter output has the same size as the input image.
7252 For image generators, the size of the output image is determined by the
7253 previous output image of the filter chain or the input image of the whole
7254 filterchain, respectively. The generators do not use the pixel information of
7255 this image to generate their output. However, the generated output is
7256 blended onto this image, resulting in partial or complete coverage of the
7259 The @ref{coreimagesrc} video source can be used for generating input images
7260 which are directly fed into the filter chain. By using it, providing input
7261 images by another video source or an input video is not required.
7263 @subsection Examples
7268 List all filters available:
7270 coreimage=list_filters=true
7274 Use the CIBoxBlur filter with default options to blur an image:
7276 coreimage=filter=CIBoxBlur@@default
7280 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7281 its center at 100x100 and a radius of 50 pixels:
7283 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7287 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7288 given as complete and escaped command-line for Apple's standard bash shell:
7290 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7296 Crop the input video to given dimensions.
7298 It accepts the following parameters:
7302 The width of the output video. It defaults to @code{iw}.
7303 This expression is evaluated only once during the filter
7304 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7307 The height of the output video. It defaults to @code{ih}.
7308 This expression is evaluated only once during the filter
7309 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7312 The horizontal position, in the input video, of the left edge of the output
7313 video. It defaults to @code{(in_w-out_w)/2}.
7314 This expression is evaluated per-frame.
7317 The vertical position, in the input video, of the top edge of the output video.
7318 It defaults to @code{(in_h-out_h)/2}.
7319 This expression is evaluated per-frame.
7322 If set to 1 will force the output display aspect ratio
7323 to be the same of the input, by changing the output sample aspect
7324 ratio. It defaults to 0.
7327 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7328 width/height/x/y as specified and will not be rounded to nearest smaller value.
7332 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7333 expressions containing the following constants:
7338 The computed values for @var{x} and @var{y}. They are evaluated for
7343 The input width and height.
7347 These are the same as @var{in_w} and @var{in_h}.
7351 The output (cropped) width and height.
7355 These are the same as @var{out_w} and @var{out_h}.
7358 same as @var{iw} / @var{ih}
7361 input sample aspect ratio
7364 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7368 horizontal and vertical chroma subsample values. For example for the
7369 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7372 The number of the input frame, starting from 0.
7375 the position in the file of the input frame, NAN if unknown
7378 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7382 The expression for @var{out_w} may depend on the value of @var{out_h},
7383 and the expression for @var{out_h} may depend on @var{out_w}, but they
7384 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7385 evaluated after @var{out_w} and @var{out_h}.
7387 The @var{x} and @var{y} parameters specify the expressions for the
7388 position of the top-left corner of the output (non-cropped) area. They
7389 are evaluated for each frame. If the evaluated value is not valid, it
7390 is approximated to the nearest valid value.
7392 The expression for @var{x} may depend on @var{y}, and the expression
7393 for @var{y} may depend on @var{x}.
7395 @subsection Examples
7399 Crop area with size 100x100 at position (12,34).
7404 Using named options, the example above becomes:
7406 crop=w=100:h=100:x=12:y=34
7410 Crop the central input area with size 100x100:
7416 Crop the central input area with size 2/3 of the input video:
7418 crop=2/3*in_w:2/3*in_h
7422 Crop the input video central square:
7429 Delimit the rectangle with the top-left corner placed at position
7430 100:100 and the right-bottom corner corresponding to the right-bottom
7431 corner of the input image.
7433 crop=in_w-100:in_h-100:100:100
7437 Crop 10 pixels from the left and right borders, and 20 pixels from
7438 the top and bottom borders
7440 crop=in_w-2*10:in_h-2*20
7444 Keep only the bottom right quarter of the input image:
7446 crop=in_w/2:in_h/2:in_w/2:in_h/2
7450 Crop height for getting Greek harmony:
7452 crop=in_w:1/PHI*in_w
7456 Apply trembling effect:
7458 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)
7462 Apply erratic camera effect depending on timestamp:
7464 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)"
7468 Set x depending on the value of y:
7470 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7474 @subsection Commands
7476 This filter supports the following commands:
7482 Set width/height of the output video and the horizontal/vertical position
7484 The command accepts the same syntax of the corresponding option.
7486 If the specified expression is not valid, it is kept at its current
7492 Auto-detect the crop size.
7494 It calculates the necessary cropping parameters and prints the
7495 recommended parameters via the logging system. The detected dimensions
7496 correspond to the non-black area of the input video.
7498 It accepts the following parameters:
7503 Set higher black value threshold, which can be optionally specified
7504 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7505 value greater to the set value is considered non-black. It defaults to 24.
7506 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7507 on the bitdepth of the pixel format.
7510 The value which the width/height should be divisible by. It defaults to
7511 16. The offset is automatically adjusted to center the video. Use 2 to
7512 get only even dimensions (needed for 4:2:2 video). 16 is best when
7513 encoding to most video codecs.
7515 @item reset_count, reset
7516 Set the counter that determines after how many frames cropdetect will
7517 reset the previously detected largest video area and start over to
7518 detect the current optimal crop area. Default value is 0.
7520 This can be useful when channel logos distort the video area. 0
7521 indicates 'never reset', and returns the largest area encountered during
7528 Delay video filtering until a given wallclock timestamp. The filter first
7529 passes on @option{preroll} amount of frames, then it buffers at most
7530 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7531 it forwards the buffered frames and also any subsequent frames coming in its
7534 The filter can be used synchronize the output of multiple ffmpeg processes for
7535 realtime output devices like decklink. By putting the delay in the filtering
7536 chain and pre-buffering frames the process can pass on data to output almost
7537 immediately after the target wallclock timestamp is reached.
7539 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7545 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7548 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7551 The maximum duration of content to buffer before waiting for the cue expressed
7552 in seconds. Default is 0.
7559 Apply color adjustments using curves.
7561 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7562 component (red, green and blue) has its values defined by @var{N} key points
7563 tied from each other using a smooth curve. The x-axis represents the pixel
7564 values from the input frame, and the y-axis the new pixel values to be set for
7567 By default, a component curve is defined by the two points @var{(0;0)} and
7568 @var{(1;1)}. This creates a straight line where each original pixel value is
7569 "adjusted" to its own value, which means no change to the image.
7571 The filter allows you to redefine these two points and add some more. A new
7572 curve (using a natural cubic spline interpolation) will be define to pass
7573 smoothly through all these new coordinates. The new defined points needs to be
7574 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7575 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7576 the vector spaces, the values will be clipped accordingly.
7578 The filter accepts the following options:
7582 Select one of the available color presets. This option can be used in addition
7583 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7584 options takes priority on the preset values.
7585 Available presets are:
7588 @item color_negative
7591 @item increase_contrast
7593 @item linear_contrast
7594 @item medium_contrast
7596 @item strong_contrast
7599 Default is @code{none}.
7601 Set the master key points. These points will define a second pass mapping. It
7602 is sometimes called a "luminance" or "value" mapping. It can be used with
7603 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7604 post-processing LUT.
7606 Set the key points for the red component.
7608 Set the key points for the green component.
7610 Set the key points for the blue component.
7612 Set the key points for all components (not including master).
7613 Can be used in addition to the other key points component
7614 options. In this case, the unset component(s) will fallback on this
7615 @option{all} setting.
7617 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7619 Save Gnuplot script of the curves in specified file.
7622 To avoid some filtergraph syntax conflicts, each key points list need to be
7623 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7625 @subsection Examples
7629 Increase slightly the middle level of blue:
7631 curves=blue='0/0 0.5/0.58 1/1'
7637 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'
7639 Here we obtain the following coordinates for each components:
7642 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7644 @code{(0;0) (0.50;0.48) (1;1)}
7646 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7650 The previous example can also be achieved with the associated built-in preset:
7652 curves=preset=vintage
7662 Use a Photoshop preset and redefine the points of the green component:
7664 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7668 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7669 and @command{gnuplot}:
7671 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7672 gnuplot -p /tmp/curves.plt
7678 Video data analysis filter.
7680 This filter shows hexadecimal pixel values of part of video.
7682 The filter accepts the following options:
7686 Set output video size.
7689 Set x offset from where to pick pixels.
7692 Set y offset from where to pick pixels.
7695 Set scope mode, can be one of the following:
7698 Draw hexadecimal pixel values with white color on black background.
7701 Draw hexadecimal pixel values with input video pixel color on black
7705 Draw hexadecimal pixel values on color background picked from input video,
7706 the text color is picked in such way so its always visible.
7710 Draw rows and columns numbers on left and top of video.
7713 Set background opacity.
7718 Denoise frames using 2D DCT (frequency domain filtering).
7720 This filter is not designed for real time.
7722 The filter accepts the following options:
7726 Set the noise sigma constant.
7728 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7729 coefficient (absolute value) below this threshold with be dropped.
7731 If you need a more advanced filtering, see @option{expr}.
7733 Default is @code{0}.
7736 Set number overlapping pixels for each block. Since the filter can be slow, you
7737 may want to reduce this value, at the cost of a less effective filter and the
7738 risk of various artefacts.
7740 If the overlapping value doesn't permit processing the whole input width or
7741 height, a warning will be displayed and according borders won't be denoised.
7743 Default value is @var{blocksize}-1, which is the best possible setting.
7746 Set the coefficient factor expression.
7748 For each coefficient of a DCT block, this expression will be evaluated as a
7749 multiplier value for the coefficient.
7751 If this is option is set, the @option{sigma} option will be ignored.
7753 The absolute value of the coefficient can be accessed through the @var{c}
7757 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7758 @var{blocksize}, which is the width and height of the processed blocks.
7760 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7761 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7762 on the speed processing. Also, a larger block size does not necessarily means a
7766 @subsection Examples
7768 Apply a denoise with a @option{sigma} of @code{4.5}:
7773 The same operation can be achieved using the expression system:
7775 dctdnoiz=e='gte(c, 4.5*3)'
7778 Violent denoise using a block size of @code{16x16}:
7785 Remove banding artifacts from input video.
7786 It works by replacing banded pixels with average value of referenced pixels.
7788 The filter accepts the following options:
7795 Set banding detection threshold for each plane. Default is 0.02.
7796 Valid range is 0.00003 to 0.5.
7797 If difference between current pixel and reference pixel is less than threshold,
7798 it will be considered as banded.
7801 Banding detection range in pixels. Default is 16. If positive, random number
7802 in range 0 to set value will be used. If negative, exact absolute value
7804 The range defines square of four pixels around current pixel.
7807 Set direction in radians from which four pixel will be compared. If positive,
7808 random direction from 0 to set direction will be picked. If negative, exact of
7809 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7810 will pick only pixels on same row and -PI/2 will pick only pixels on same
7814 If enabled, current pixel is compared with average value of all four
7815 surrounding pixels. The default is enabled. If disabled current pixel is
7816 compared with all four surrounding pixels. The pixel is considered banded
7817 if only all four differences with surrounding pixels are less than threshold.
7820 If enabled, current pixel is changed if and only if all pixel components are banded,
7821 e.g. banding detection threshold is triggered for all color components.
7822 The default is disabled.
7827 Remove blocking artifacts from input video.
7829 The filter accepts the following options:
7833 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7834 This controls what kind of deblocking is applied.
7837 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7843 Set blocking detection thresholds. Allowed range is 0 to 1.
7844 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7845 Using higher threshold gives more deblocking strength.
7846 Setting @var{alpha} controls threshold detection at exact edge of block.
7847 Remaining options controls threshold detection near the edge. Each one for
7848 below/above or left/right. Setting any of those to @var{0} disables
7852 Set planes to filter. Default is to filter all available planes.
7855 @subsection Examples
7859 Deblock using weak filter and block size of 4 pixels.
7861 deblock=filter=weak:block=4
7865 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7866 deblocking more edges.
7868 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7872 Similar as above, but filter only first plane.
7874 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7878 Similar as above, but filter only second and third plane.
7880 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7887 Drop duplicated frames at regular intervals.
7889 The filter accepts the following options:
7893 Set the number of frames from which one will be dropped. Setting this to
7894 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7895 Default is @code{5}.
7898 Set the threshold for duplicate detection. If the difference metric for a frame
7899 is less than or equal to this value, then it is declared as duplicate. Default
7903 Set scene change threshold. Default is @code{15}.
7907 Set the size of the x and y-axis blocks used during metric calculations.
7908 Larger blocks give better noise suppression, but also give worse detection of
7909 small movements. Must be a power of two. Default is @code{32}.
7912 Mark main input as a pre-processed input and activate clean source input
7913 stream. This allows the input to be pre-processed with various filters to help
7914 the metrics calculation while keeping the frame selection lossless. When set to
7915 @code{1}, the first stream is for the pre-processed input, and the second
7916 stream is the clean source from where the kept frames are chosen. Default is
7920 Set whether or not chroma is considered in the metric calculations. Default is
7926 Apply 2D deconvolution of video stream in frequency domain using second stream
7929 The filter accepts the following options:
7933 Set which planes to process.
7936 Set which impulse video frames will be processed, can be @var{first}
7937 or @var{all}. Default is @var{all}.
7940 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7941 and height are not same and not power of 2 or if stream prior to convolving
7945 The @code{deconvolve} filter also supports the @ref{framesync} options.
7949 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
7951 It accepts the following options:
7955 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
7956 @var{rainbows} for cross-color reduction.
7959 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
7962 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
7965 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
7968 Set temporal chroma threshold. Lower values increases reduction of cross-color.
7973 Apply deflate effect to the video.
7975 This filter replaces the pixel by the local(3x3) average by taking into account
7976 only values lower than the pixel.
7978 It accepts the following options:
7985 Limit the maximum change for each plane, default is 65535.
7986 If 0, plane will remain unchanged.
7991 Remove temporal frame luminance variations.
7993 It accepts the following options:
7997 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
8000 Set averaging mode to smooth temporal luminance variations.
8002 Available values are:
8027 Do not actually modify frame. Useful when one only wants metadata.
8032 Remove judder produced by partially interlaced telecined content.
8034 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
8035 source was partially telecined content then the output of @code{pullup,dejudder}
8036 will have a variable frame rate. May change the recorded frame rate of the
8037 container. Aside from that change, this filter will not affect constant frame
8040 The option available in this filter is:
8044 Specify the length of the window over which the judder repeats.
8046 Accepts any integer greater than 1. Useful values are:
8050 If the original was telecined from 24 to 30 fps (Film to NTSC).
8053 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8056 If a mixture of the two.
8059 The default is @samp{4}.
8064 Suppress a TV station logo by a simple interpolation of the surrounding
8065 pixels. Just set a rectangle covering the logo and watch it disappear
8066 (and sometimes something even uglier appear - your mileage may vary).
8068 It accepts the following parameters:
8073 Specify the top left corner coordinates of the logo. They must be
8078 Specify the width and height of the logo to clear. They must be
8082 Specify the thickness of the fuzzy edge of the rectangle (added to
8083 @var{w} and @var{h}). The default value is 1. This option is
8084 deprecated, setting higher values should no longer be necessary and
8088 When set to 1, a green rectangle is drawn on the screen to simplify
8089 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8090 The default value is 0.
8092 The rectangle is drawn on the outermost pixels which will be (partly)
8093 replaced with interpolated values. The values of the next pixels
8094 immediately outside this rectangle in each direction will be used to
8095 compute the interpolated pixel values inside the rectangle.
8099 @subsection Examples
8103 Set a rectangle covering the area with top left corner coordinates 0,0
8104 and size 100x77, and a band of size 10:
8106 delogo=x=0:y=0:w=100:h=77:band=10
8113 Attempt to fix small changes in horizontal and/or vertical shift. This
8114 filter helps remove camera shake from hand-holding a camera, bumping a
8115 tripod, moving on a vehicle, etc.
8117 The filter accepts the following options:
8125 Specify a rectangular area where to limit the search for motion
8127 If desired the search for motion vectors can be limited to a
8128 rectangular area of the frame defined by its top left corner, width
8129 and height. These parameters have the same meaning as the drawbox
8130 filter which can be used to visualise the position of the bounding
8133 This is useful when simultaneous movement of subjects within the frame
8134 might be confused for camera motion by the motion vector search.
8136 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8137 then the full frame is used. This allows later options to be set
8138 without specifying the bounding box for the motion vector search.
8140 Default - search the whole frame.
8144 Specify the maximum extent of movement in x and y directions in the
8145 range 0-64 pixels. Default 16.
8148 Specify how to generate pixels to fill blanks at the edge of the
8149 frame. Available values are:
8152 Fill zeroes at blank locations
8154 Original image at blank locations
8156 Extruded edge value at blank locations
8158 Mirrored edge at blank locations
8160 Default value is @samp{mirror}.
8163 Specify the blocksize to use for motion search. Range 4-128 pixels,
8167 Specify the contrast threshold for blocks. Only blocks with more than
8168 the specified contrast (difference between darkest and lightest
8169 pixels) will be considered. Range 1-255, default 125.
8172 Specify the search strategy. Available values are:
8175 Set exhaustive search
8177 Set less exhaustive search.
8179 Default value is @samp{exhaustive}.
8182 If set then a detailed log of the motion search is written to the
8189 Remove unwanted contamination of foreground colors, caused by reflected color of
8190 greenscreen or bluescreen.
8192 This filter accepts the following options:
8196 Set what type of despill to use.
8199 Set how spillmap will be generated.
8202 Set how much to get rid of still remaining spill.
8205 Controls amount of red in spill area.
8208 Controls amount of green in spill area.
8209 Should be -1 for greenscreen.
8212 Controls amount of blue in spill area.
8213 Should be -1 for bluescreen.
8216 Controls brightness of spill area, preserving colors.
8219 Modify alpha from generated spillmap.
8224 Apply an exact inverse of the telecine operation. It requires a predefined
8225 pattern specified using the pattern option which must be the same as that passed
8226 to the telecine filter.
8228 This filter accepts the following options:
8237 The default value is @code{top}.
8241 A string of numbers representing the pulldown pattern you wish to apply.
8242 The default value is @code{23}.
8245 A number representing position of the first frame with respect to the telecine
8246 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8251 Apply dilation effect to the video.
8253 This filter replaces the pixel by the local(3x3) maximum.
8255 It accepts the following options:
8262 Limit the maximum change for each plane, default is 65535.
8263 If 0, plane will remain unchanged.
8266 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8269 Flags to local 3x3 coordinates maps like this:
8278 Displace pixels as indicated by second and third input stream.
8280 It takes three input streams and outputs one stream, the first input is the
8281 source, and second and third input are displacement maps.
8283 The second input specifies how much to displace pixels along the
8284 x-axis, while the third input specifies how much to displace pixels
8286 If one of displacement map streams terminates, last frame from that
8287 displacement map will be used.
8289 Note that once generated, displacements maps can be reused over and over again.
8291 A description of the accepted options follows.
8295 Set displace behavior for pixels that are out of range.
8297 Available values are:
8300 Missing pixels are replaced by black pixels.
8303 Adjacent pixels will spread out to replace missing pixels.
8306 Out of range pixels are wrapped so they point to pixels of other side.
8309 Out of range pixels will be replaced with mirrored pixels.
8311 Default is @samp{smear}.
8315 @subsection Examples
8319 Add ripple effect to rgb input of video size hd720:
8321 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
8325 Add wave effect to rgb input of video size hd720:
8327 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
8333 Draw a colored box on the input image.
8335 It accepts the following parameters:
8340 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8344 The expressions which specify the width and height of the box; if 0 they are interpreted as
8345 the input width and height. It defaults to 0.
8348 Specify the color of the box to write. For the general syntax of this option,
8349 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8350 value @code{invert} is used, the box edge color is the same as the
8351 video with inverted luma.
8354 The expression which sets the thickness of the box edge.
8355 A value of @code{fill} will create a filled box. Default value is @code{3}.
8357 See below for the list of accepted constants.
8360 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8361 will overwrite the video's color and alpha pixels.
8362 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8365 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8366 following constants:
8370 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8374 horizontal and vertical chroma subsample values. For example for the
8375 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8379 The input width and height.
8382 The input sample aspect ratio.
8386 The x and y offset coordinates where the box is drawn.
8390 The width and height of the drawn box.
8393 The thickness of the drawn box.
8395 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8396 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8400 @subsection Examples
8404 Draw a black box around the edge of the input image:
8410 Draw a box with color red and an opacity of 50%:
8412 drawbox=10:20:200:60:red@@0.5
8415 The previous example can be specified as:
8417 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8421 Fill the box with pink color:
8423 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8427 Draw a 2-pixel red 2.40:1 mask:
8429 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
8435 Draw a grid on the input image.
8437 It accepts the following parameters:
8442 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8446 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8447 input width and height, respectively, minus @code{thickness}, so image gets
8448 framed. Default to 0.
8451 Specify the color of the grid. For the general syntax of this option,
8452 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8453 value @code{invert} is used, the grid color is the same as the
8454 video with inverted luma.
8457 The expression which sets the thickness of the grid line. Default value is @code{1}.
8459 See below for the list of accepted constants.
8462 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8463 will overwrite the video's color and alpha pixels.
8464 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8467 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8468 following constants:
8472 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8476 horizontal and vertical chroma subsample values. For example for the
8477 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8481 The input grid cell width and height.
8484 The input sample aspect ratio.
8488 The x and y coordinates of some point of grid intersection (meant to configure offset).
8492 The width and height of the drawn cell.
8495 The thickness of the drawn cell.
8497 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8498 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8502 @subsection Examples
8506 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8508 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8512 Draw a white 3x3 grid with an opacity of 50%:
8514 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8521 Draw a text string or text from a specified file on top of a video, using the
8522 libfreetype library.
8524 To enable compilation of this filter, you need to configure FFmpeg with
8525 @code{--enable-libfreetype}.
8526 To enable default font fallback and the @var{font} option you need to
8527 configure FFmpeg with @code{--enable-libfontconfig}.
8528 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8529 @code{--enable-libfribidi}.
8533 It accepts the following parameters:
8538 Used to draw a box around text using the background color.
8539 The value must be either 1 (enable) or 0 (disable).
8540 The default value of @var{box} is 0.
8543 Set the width of the border to be drawn around the box using @var{boxcolor}.
8544 The default value of @var{boxborderw} is 0.
8547 The color to be used for drawing box around text. For the syntax of this
8548 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8550 The default value of @var{boxcolor} is "white".
8553 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8554 The default value of @var{line_spacing} is 0.
8557 Set the width of the border to be drawn around the text using @var{bordercolor}.
8558 The default value of @var{borderw} is 0.
8561 Set the color to be used for drawing border around text. For the syntax of this
8562 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8564 The default value of @var{bordercolor} is "black".
8567 Select how the @var{text} is expanded. Can be either @code{none},
8568 @code{strftime} (deprecated) or
8569 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8573 Set a start time for the count. Value is in microseconds. Only applied
8574 in the deprecated strftime expansion mode. To emulate in normal expansion
8575 mode use the @code{pts} function, supplying the start time (in seconds)
8576 as the second argument.
8579 If true, check and fix text coords to avoid clipping.
8582 The color to be used for drawing fonts. For the syntax of this option, check
8583 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8585 The default value of @var{fontcolor} is "black".
8587 @item fontcolor_expr
8588 String which is expanded the same way as @var{text} to obtain dynamic
8589 @var{fontcolor} value. By default this option has empty value and is not
8590 processed. When this option is set, it overrides @var{fontcolor} option.
8593 The font family to be used for drawing text. By default Sans.
8596 The font file to be used for drawing text. The path must be included.
8597 This parameter is mandatory if the fontconfig support is disabled.
8600 Draw the text applying alpha blending. The value can
8601 be a number between 0.0 and 1.0.
8602 The expression accepts the same variables @var{x, y} as well.
8603 The default value is 1.
8604 Please see @var{fontcolor_expr}.
8607 The font size to be used for drawing text.
8608 The default value of @var{fontsize} is 16.
8611 If set to 1, attempt to shape the text (for example, reverse the order of
8612 right-to-left text and join Arabic characters) before drawing it.
8613 Otherwise, just draw the text exactly as given.
8614 By default 1 (if supported).
8617 The flags to be used for loading the fonts.
8619 The flags map the corresponding flags supported by libfreetype, and are
8620 a combination of the following values:
8627 @item vertical_layout
8628 @item force_autohint
8631 @item ignore_global_advance_width
8633 @item ignore_transform
8639 Default value is "default".
8641 For more information consult the documentation for the FT_LOAD_*
8645 The color to be used for drawing a shadow behind the drawn text. For the
8646 syntax of this option, check the @ref{color syntax,,"Color" section in the
8647 ffmpeg-utils manual,ffmpeg-utils}.
8649 The default value of @var{shadowcolor} is "black".
8653 The x and y offsets for the text shadow position with respect to the
8654 position of the text. They can be either positive or negative
8655 values. The default value for both is "0".
8658 The starting frame number for the n/frame_num variable. The default value
8662 The size in number of spaces to use for rendering the tab.
8666 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8667 format. It can be used with or without text parameter. @var{timecode_rate}
8668 option must be specified.
8670 @item timecode_rate, rate, r
8671 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8672 integer. Minimum value is "1".
8673 Drop-frame timecode is supported for frame rates 30 & 60.
8676 If set to 1, the output of the timecode option will wrap around at 24 hours.
8677 Default is 0 (disabled).
8680 The text string to be drawn. The text must be a sequence of UTF-8
8682 This parameter is mandatory if no file is specified with the parameter
8686 A text file containing text to be drawn. The text must be a sequence
8687 of UTF-8 encoded characters.
8689 This parameter is mandatory if no text string is specified with the
8690 parameter @var{text}.
8692 If both @var{text} and @var{textfile} are specified, an error is thrown.
8695 If set to 1, the @var{textfile} will be reloaded before each frame.
8696 Be sure to update it atomically, or it may be read partially, or even fail.
8700 The expressions which specify the offsets where text will be drawn
8701 within the video frame. They are relative to the top/left border of the
8704 The default value of @var{x} and @var{y} is "0".
8706 See below for the list of accepted constants and functions.
8709 The parameters for @var{x} and @var{y} are expressions containing the
8710 following constants and functions:
8714 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8718 horizontal and vertical chroma subsample values. For example for the
8719 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8722 the height of each text line
8730 @item max_glyph_a, ascent
8731 the maximum distance from the baseline to the highest/upper grid
8732 coordinate used to place a glyph outline point, for all the rendered
8734 It is a positive value, due to the grid's orientation with the Y axis
8737 @item max_glyph_d, descent
8738 the maximum distance from the baseline to the lowest grid coordinate
8739 used to place a glyph outline point, for all the rendered glyphs.
8740 This is a negative value, due to the grid's orientation, with the Y axis
8744 maximum glyph height, that is the maximum height for all the glyphs
8745 contained in the rendered text, it is equivalent to @var{ascent} -
8749 maximum glyph width, that is the maximum width for all the glyphs
8750 contained in the rendered text
8753 the number of input frame, starting from 0
8755 @item rand(min, max)
8756 return a random number included between @var{min} and @var{max}
8759 The input sample aspect ratio.
8762 timestamp expressed in seconds, NAN if the input timestamp is unknown
8765 the height of the rendered text
8768 the width of the rendered text
8772 the x and y offset coordinates where the text is drawn.
8774 These parameters allow the @var{x} and @var{y} expressions to refer
8775 each other, so you can for example specify @code{y=x/dar}.
8778 @anchor{drawtext_expansion}
8779 @subsection Text expansion
8781 If @option{expansion} is set to @code{strftime},
8782 the filter recognizes strftime() sequences in the provided text and
8783 expands them accordingly. Check the documentation of strftime(). This
8784 feature is deprecated.
8786 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8788 If @option{expansion} is set to @code{normal} (which is the default),
8789 the following expansion mechanism is used.
8791 The backslash character @samp{\}, followed by any character, always expands to
8792 the second character.
8794 Sequences of the form @code{%@{...@}} are expanded. The text between the
8795 braces is a function name, possibly followed by arguments separated by ':'.
8796 If the arguments contain special characters or delimiters (':' or '@}'),
8797 they should be escaped.
8799 Note that they probably must also be escaped as the value for the
8800 @option{text} option in the filter argument string and as the filter
8801 argument in the filtergraph description, and possibly also for the shell,
8802 that makes up to four levels of escaping; using a text file avoids these
8805 The following functions are available:
8810 The expression evaluation result.
8812 It must take one argument specifying the expression to be evaluated,
8813 which accepts the same constants and functions as the @var{x} and
8814 @var{y} values. Note that not all constants should be used, for
8815 example the text size is not known when evaluating the expression, so
8816 the constants @var{text_w} and @var{text_h} will have an undefined
8819 @item expr_int_format, eif
8820 Evaluate the expression's value and output as formatted integer.
8822 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8823 The second argument specifies the output format. Allowed values are @samp{x},
8824 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8825 @code{printf} function.
8826 The third parameter is optional and sets the number of positions taken by the output.
8827 It can be used to add padding with zeros from the left.
8830 The time at which the filter is running, expressed in UTC.
8831 It can accept an argument: a strftime() format string.
8834 The time at which the filter is running, expressed in the local time zone.
8835 It can accept an argument: a strftime() format string.
8838 Frame metadata. Takes one or two arguments.
8840 The first argument is mandatory and specifies the metadata key.
8842 The second argument is optional and specifies a default value, used when the
8843 metadata key is not found or empty.
8846 The frame number, starting from 0.
8849 A 1 character description of the current picture type.
8852 The timestamp of the current frame.
8853 It can take up to three arguments.
8855 The first argument is the format of the timestamp; it defaults to @code{flt}
8856 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8857 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8858 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8859 @code{localtime} stands for the timestamp of the frame formatted as
8860 local time zone time.
8862 The second argument is an offset added to the timestamp.
8864 If the format is set to @code{hms}, a third argument @code{24HH} may be
8865 supplied to present the hour part of the formatted timestamp in 24h format
8868 If the format is set to @code{localtime} or @code{gmtime},
8869 a third argument may be supplied: a strftime() format string.
8870 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8873 @subsection Examples
8877 Draw "Test Text" with font FreeSerif, using the default values for the
8878 optional parameters.
8881 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8885 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8886 and y=50 (counting from the top-left corner of the screen), text is
8887 yellow with a red box around it. Both the text and the box have an
8891 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8892 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8895 Note that the double quotes are not necessary if spaces are not used
8896 within the parameter list.
8899 Show the text at the center of the video frame:
8901 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8905 Show the text at a random position, switching to a new position every 30 seconds:
8907 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)"
8911 Show a text line sliding from right to left in the last row of the video
8912 frame. The file @file{LONG_LINE} is assumed to contain a single line
8915 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8919 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8921 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8925 Draw a single green letter "g", at the center of the input video.
8926 The glyph baseline is placed at half screen height.
8928 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8932 Show text for 1 second every 3 seconds:
8934 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8938 Use fontconfig to set the font. Note that the colons need to be escaped.
8940 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8944 Print the date of a real-time encoding (see strftime(3)):
8946 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8950 Show text fading in and out (appearing/disappearing):
8953 DS=1.0 # display start
8954 DE=10.0 # display end
8955 FID=1.5 # fade in duration
8956 FOD=5 # fade out duration
8957 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 @}"
8961 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8962 and the @option{fontsize} value are included in the @option{y} offset.
8964 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8965 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8970 For more information about libfreetype, check:
8971 @url{http://www.freetype.org/}.
8973 For more information about fontconfig, check:
8974 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8976 For more information about libfribidi, check:
8977 @url{http://fribidi.org/}.
8981 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8983 The filter accepts the following options:
8988 Set low and high threshold values used by the Canny thresholding
8991 The high threshold selects the "strong" edge pixels, which are then
8992 connected through 8-connectivity with the "weak" edge pixels selected
8993 by the low threshold.
8995 @var{low} and @var{high} threshold values must be chosen in the range
8996 [0,1], and @var{low} should be lesser or equal to @var{high}.
8998 Default value for @var{low} is @code{20/255}, and default value for @var{high}
9002 Define the drawing mode.
9006 Draw white/gray wires on black background.
9009 Mix the colors to create a paint/cartoon effect.
9012 Apply Canny edge detector on all selected planes.
9014 Default value is @var{wires}.
9017 Select planes for filtering. By default all available planes are filtered.
9020 @subsection Examples
9024 Standard edge detection with custom values for the hysteresis thresholding:
9026 edgedetect=low=0.1:high=0.4
9030 Painting effect without thresholding:
9032 edgedetect=mode=colormix:high=0
9037 Set brightness, contrast, saturation and approximate gamma adjustment.
9039 The filter accepts the following options:
9043 Set the contrast expression. The value must be a float value in range
9044 @code{-2.0} to @code{2.0}. The default value is "1".
9047 Set the brightness expression. The value must be a float value in
9048 range @code{-1.0} to @code{1.0}. The default value is "0".
9051 Set the saturation expression. The value must be a float in
9052 range @code{0.0} to @code{3.0}. The default value is "1".
9055 Set the gamma expression. The value must be a float in range
9056 @code{0.1} to @code{10.0}. The default value is "1".
9059 Set the gamma expression for red. The value must be a float in
9060 range @code{0.1} to @code{10.0}. The default value is "1".
9063 Set the gamma expression for green. The value must be a float in range
9064 @code{0.1} to @code{10.0}. The default value is "1".
9067 Set the gamma expression for blue. The value must be a float in range
9068 @code{0.1} to @code{10.0}. The default value is "1".
9071 Set the gamma weight expression. It can be used to reduce the effect
9072 of a high gamma value on bright image areas, e.g. keep them from
9073 getting overamplified and just plain white. The value must be a float
9074 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
9075 gamma correction all the way down while @code{1.0} leaves it at its
9076 full strength. Default is "1".
9079 Set when the expressions for brightness, contrast, saturation and
9080 gamma expressions are evaluated.
9082 It accepts the following values:
9085 only evaluate expressions once during the filter initialization or
9086 when a command is processed
9089 evaluate expressions for each incoming frame
9092 Default value is @samp{init}.
9095 The expressions accept the following parameters:
9098 frame count of the input frame starting from 0
9101 byte position of the corresponding packet in the input file, NAN if
9105 frame rate of the input video, NAN if the input frame rate is unknown
9108 timestamp expressed in seconds, NAN if the input timestamp is unknown
9111 @subsection Commands
9112 The filter supports the following commands:
9116 Set the contrast expression.
9119 Set the brightness expression.
9122 Set the saturation expression.
9125 Set the gamma expression.
9128 Set the gamma_r expression.
9131 Set gamma_g expression.
9134 Set gamma_b expression.
9137 Set gamma_weight expression.
9139 The command accepts the same syntax of the corresponding option.
9141 If the specified expression is not valid, it is kept at its current
9148 Apply erosion effect to the video.
9150 This filter replaces the pixel by the local(3x3) minimum.
9152 It accepts the following options:
9159 Limit the maximum change for each plane, default is 65535.
9160 If 0, plane will remain unchanged.
9163 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9166 Flags to local 3x3 coordinates maps like this:
9173 @section extractplanes
9175 Extract color channel components from input video stream into
9176 separate grayscale video streams.
9178 The filter accepts the following option:
9182 Set plane(s) to extract.
9184 Available values for planes are:
9195 Choosing planes not available in the input will result in an error.
9196 That means you cannot select @code{r}, @code{g}, @code{b} planes
9197 with @code{y}, @code{u}, @code{v} planes at same time.
9200 @subsection Examples
9204 Extract luma, u and v color channel component from input video frame
9205 into 3 grayscale outputs:
9207 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
9213 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9215 For each input image, the filter will compute the optimal mapping from
9216 the input to the output given the codebook length, that is the number
9217 of distinct output colors.
9219 This filter accepts the following options.
9222 @item codebook_length, l
9223 Set codebook length. The value must be a positive integer, and
9224 represents the number of distinct output colors. Default value is 256.
9227 Set the maximum number of iterations to apply for computing the optimal
9228 mapping. The higher the value the better the result and the higher the
9229 computation time. Default value is 1.
9232 Set a random seed, must be an integer included between 0 and
9233 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9234 will try to use a good random seed on a best effort basis.
9237 Set pal8 output pixel format. This option does not work with codebook
9238 length greater than 256.
9243 Measure graylevel entropy in histogram of color channels of video frames.
9245 It accepts the following parameters:
9249 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9251 @var{diff} mode measures entropy of histogram delta values, absolute differences
9252 between neighbour histogram values.
9257 Apply a fade-in/out effect to the input video.
9259 It accepts the following parameters:
9263 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9265 Default is @code{in}.
9267 @item start_frame, s
9268 Specify the number of the frame to start applying the fade
9269 effect at. Default is 0.
9272 The number of frames that the fade effect lasts. At the end of the
9273 fade-in effect, the output video will have the same intensity as the input video.
9274 At the end of the fade-out transition, the output video will be filled with the
9275 selected @option{color}.
9279 If set to 1, fade only alpha channel, if one exists on the input.
9282 @item start_time, st
9283 Specify the timestamp (in seconds) of the frame to start to apply the fade
9284 effect. If both start_frame and start_time are specified, the fade will start at
9285 whichever comes last. Default is 0.
9288 The number of seconds for which the fade effect has to last. At the end of the
9289 fade-in effect the output video will have the same intensity as the input video,
9290 at the end of the fade-out transition the output video will be filled with the
9291 selected @option{color}.
9292 If both duration and nb_frames are specified, duration is used. Default is 0
9293 (nb_frames is used by default).
9296 Specify the color of the fade. Default is "black".
9299 @subsection Examples
9303 Fade in the first 30 frames of video:
9308 The command above is equivalent to:
9314 Fade out the last 45 frames of a 200-frame video:
9317 fade=type=out:start_frame=155:nb_frames=45
9321 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9323 fade=in:0:25, fade=out:975:25
9327 Make the first 5 frames yellow, then fade in from frame 5-24:
9329 fade=in:5:20:color=yellow
9333 Fade in alpha over first 25 frames of video:
9335 fade=in:0:25:alpha=1
9339 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9341 fade=t=in:st=5.5:d=0.5
9347 Apply arbitrary expressions to samples in frequency domain
9351 Adjust the dc value (gain) of the luma plane of the image. The filter
9352 accepts an integer value in range @code{0} to @code{1000}. The default
9353 value is set to @code{0}.
9356 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9357 filter accepts an integer value in range @code{0} to @code{1000}. The
9358 default value is set to @code{0}.
9361 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9362 filter accepts an integer value in range @code{0} to @code{1000}. The
9363 default value is set to @code{0}.
9366 Set the frequency domain weight expression for the luma plane.
9369 Set the frequency domain weight expression for the 1st chroma plane.
9372 Set the frequency domain weight expression for the 2nd chroma plane.
9375 Set when the expressions are evaluated.
9377 It accepts the following values:
9380 Only evaluate expressions once during the filter initialization.
9383 Evaluate expressions for each incoming frame.
9386 Default value is @samp{init}.
9388 The filter accepts the following variables:
9391 The coordinates of the current sample.
9395 The width and height of the image.
9398 The number of input frame, starting from 0.
9401 @subsection Examples
9407 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9413 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9419 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9425 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9431 Denoise frames using 3D FFT (frequency domain filtering).
9433 The filter accepts the following options:
9437 Set the noise sigma constant. This sets denoising strength.
9438 Default value is 1. Allowed range is from 0 to 30.
9439 Using very high sigma with low overlap may give blocking artifacts.
9442 Set amount of denoising. By default all detected noise is reduced.
9443 Default value is 1. Allowed range is from 0 to 1.
9446 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9447 Actual size of block in pixels is 2 to power of @var{block}, so by default
9448 block size in pixels is 2^4 which is 16.
9451 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9454 Set number of previous frames to use for denoising. By default is set to 0.
9457 Set number of next frames to to use for denoising. By default is set to 0.
9460 Set planes which will be filtered, by default are all available filtered
9466 Extract a single field from an interlaced image using stride
9467 arithmetic to avoid wasting CPU time. The output frames are marked as
9470 The filter accepts the following options:
9474 Specify whether to extract the top (if the value is @code{0} or
9475 @code{top}) or the bottom field (if the value is @code{1} or
9481 Create new frames by copying the top and bottom fields from surrounding frames
9482 supplied as numbers by the hint file.
9486 Set file containing hints: absolute/relative frame numbers.
9488 There must be one line for each frame in a clip. Each line must contain two
9489 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9490 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9491 is current frame number for @code{absolute} mode or out of [-1, 1] range
9492 for @code{relative} mode. First number tells from which frame to pick up top
9493 field and second number tells from which frame to pick up bottom field.
9495 If optionally followed by @code{+} output frame will be marked as interlaced,
9496 else if followed by @code{-} output frame will be marked as progressive, else
9497 it will be marked same as input frame.
9498 If line starts with @code{#} or @code{;} that line is skipped.
9501 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9504 Example of first several lines of @code{hint} file for @code{relative} mode:
9507 1,0 - # second frame, use third's frame top field and second's frame bottom field
9508 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9525 Field matching filter for inverse telecine. It is meant to reconstruct the
9526 progressive frames from a telecined stream. The filter does not drop duplicated
9527 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9528 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9530 The separation of the field matching and the decimation is notably motivated by
9531 the possibility of inserting a de-interlacing filter fallback between the two.
9532 If the source has mixed telecined and real interlaced content,
9533 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9534 But these remaining combed frames will be marked as interlaced, and thus can be
9535 de-interlaced by a later filter such as @ref{yadif} before decimation.
9537 In addition to the various configuration options, @code{fieldmatch} can take an
9538 optional second stream, activated through the @option{ppsrc} option. If
9539 enabled, the frames reconstruction will be based on the fields and frames from
9540 this second stream. This allows the first input to be pre-processed in order to
9541 help the various algorithms of the filter, while keeping the output lossless
9542 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9543 or brightness/contrast adjustments can help.
9545 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9546 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9547 which @code{fieldmatch} is based on. While the semantic and usage are very
9548 close, some behaviour and options names can differ.
9550 The @ref{decimate} filter currently only works for constant frame rate input.
9551 If your input has mixed telecined (30fps) and progressive content with a lower
9552 framerate like 24fps use the following filterchain to produce the necessary cfr
9553 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9555 The filter accepts the following options:
9559 Specify the assumed field order of the input stream. Available values are:
9563 Auto detect parity (use FFmpeg's internal parity value).
9565 Assume bottom field first.
9567 Assume top field first.
9570 Note that it is sometimes recommended not to trust the parity announced by the
9573 Default value is @var{auto}.
9576 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9577 sense that it won't risk creating jerkiness due to duplicate frames when
9578 possible, but if there are bad edits or blended fields it will end up
9579 outputting combed frames when a good match might actually exist. On the other
9580 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9581 but will almost always find a good frame if there is one. The other values are
9582 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9583 jerkiness and creating duplicate frames versus finding good matches in sections
9584 with bad edits, orphaned fields, blended fields, etc.
9586 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9588 Available values are:
9592 2-way matching (p/c)
9594 2-way matching, and trying 3rd match if still combed (p/c + n)
9596 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9598 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9599 still combed (p/c + n + u/b)
9601 3-way matching (p/c/n)
9603 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9604 detected as combed (p/c/n + u/b)
9607 The parenthesis at the end indicate the matches that would be used for that
9608 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9611 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9614 Default value is @var{pc_n}.
9617 Mark the main input stream as a pre-processed input, and enable the secondary
9618 input stream as the clean source to pick the fields from. See the filter
9619 introduction for more details. It is similar to the @option{clip2} feature from
9622 Default value is @code{0} (disabled).
9625 Set the field to match from. It is recommended to set this to the same value as
9626 @option{order} unless you experience matching failures with that setting. In
9627 certain circumstances changing the field that is used to match from can have a
9628 large impact on matching performance. Available values are:
9632 Automatic (same value as @option{order}).
9634 Match from the bottom field.
9636 Match from the top field.
9639 Default value is @var{auto}.
9642 Set whether or not chroma is included during the match comparisons. In most
9643 cases it is recommended to leave this enabled. You should set this to @code{0}
9644 only if your clip has bad chroma problems such as heavy rainbowing or other
9645 artifacts. Setting this to @code{0} could also be used to speed things up at
9646 the cost of some accuracy.
9648 Default value is @code{1}.
9652 These define an exclusion band which excludes the lines between @option{y0} and
9653 @option{y1} from being included in the field matching decision. An exclusion
9654 band can be used to ignore subtitles, a logo, or other things that may
9655 interfere with the matching. @option{y0} sets the starting scan line and
9656 @option{y1} sets the ending line; all lines in between @option{y0} and
9657 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9658 @option{y0} and @option{y1} to the same value will disable the feature.
9659 @option{y0} and @option{y1} defaults to @code{0}.
9662 Set the scene change detection threshold as a percentage of maximum change on
9663 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9664 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9665 @option{scthresh} is @code{[0.0, 100.0]}.
9667 Default value is @code{12.0}.
9670 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9671 account the combed scores of matches when deciding what match to use as the
9672 final match. Available values are:
9676 No final matching based on combed scores.
9678 Combed scores are only used when a scene change is detected.
9680 Use combed scores all the time.
9683 Default is @var{sc}.
9686 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9687 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9688 Available values are:
9692 No forced calculation.
9694 Force p/c/n calculations.
9696 Force p/c/n/u/b calculations.
9699 Default value is @var{none}.
9702 This is the area combing threshold used for combed frame detection. This
9703 essentially controls how "strong" or "visible" combing must be to be detected.
9704 Larger values mean combing must be more visible and smaller values mean combing
9705 can be less visible or strong and still be detected. Valid settings are from
9706 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9707 be detected as combed). This is basically a pixel difference value. A good
9708 range is @code{[8, 12]}.
9710 Default value is @code{9}.
9713 Sets whether or not chroma is considered in the combed frame decision. Only
9714 disable this if your source has chroma problems (rainbowing, etc.) that are
9715 causing problems for the combed frame detection with chroma enabled. Actually,
9716 using @option{chroma}=@var{0} is usually more reliable, except for the case
9717 where there is chroma only combing in the source.
9719 Default value is @code{0}.
9723 Respectively set the x-axis and y-axis size of the window used during combed
9724 frame detection. This has to do with the size of the area in which
9725 @option{combpel} pixels are required to be detected as combed for a frame to be
9726 declared combed. See the @option{combpel} parameter description for more info.
9727 Possible values are any number that is a power of 2 starting at 4 and going up
9730 Default value is @code{16}.
9733 The number of combed pixels inside any of the @option{blocky} by
9734 @option{blockx} size blocks on the frame for the frame to be detected as
9735 combed. While @option{cthresh} controls how "visible" the combing must be, this
9736 setting controls "how much" combing there must be in any localized area (a
9737 window defined by the @option{blockx} and @option{blocky} settings) on the
9738 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9739 which point no frames will ever be detected as combed). This setting is known
9740 as @option{MI} in TFM/VFM vocabulary.
9742 Default value is @code{80}.
9745 @anchor{p/c/n/u/b meaning}
9746 @subsection p/c/n/u/b meaning
9748 @subsubsection p/c/n
9750 We assume the following telecined stream:
9753 Top fields: 1 2 2 3 4
9754 Bottom fields: 1 2 3 4 4
9757 The numbers correspond to the progressive frame the fields relate to. Here, the
9758 first two frames are progressive, the 3rd and 4th are combed, and so on.
9760 When @code{fieldmatch} is configured to run a matching from bottom
9761 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9766 B 1 2 3 4 4 <-- matching reference
9775 As a result of the field matching, we can see that some frames get duplicated.
9776 To perform a complete inverse telecine, you need to rely on a decimation filter
9777 after this operation. See for instance the @ref{decimate} filter.
9779 The same operation now matching from top fields (@option{field}=@var{top})
9784 T 1 2 2 3 4 <-- matching reference
9794 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9795 basically, they refer to the frame and field of the opposite parity:
9798 @item @var{p} matches the field of the opposite parity in the previous frame
9799 @item @var{c} matches the field of the opposite parity in the current frame
9800 @item @var{n} matches the field of the opposite parity in the next frame
9805 The @var{u} and @var{b} matching are a bit special in the sense that they match
9806 from the opposite parity flag. In the following examples, we assume that we are
9807 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9808 'x' is placed above and below each matched fields.
9810 With bottom matching (@option{field}=@var{bottom}):
9815 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9816 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9824 With top matching (@option{field}=@var{top}):
9829 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9830 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9838 @subsection Examples
9840 Simple IVTC of a top field first telecined stream:
9842 fieldmatch=order=tff:combmatch=none, decimate
9845 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9847 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9852 Transform the field order of the input video.
9854 It accepts the following parameters:
9859 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9860 for bottom field first.
9863 The default value is @samp{tff}.
9865 The transformation is done by shifting the picture content up or down
9866 by one line, and filling the remaining line with appropriate picture content.
9867 This method is consistent with most broadcast field order converters.
9869 If the input video is not flagged as being interlaced, or it is already
9870 flagged as being of the required output field order, then this filter does
9871 not alter the incoming video.
9873 It is very useful when converting to or from PAL DV material,
9874 which is bottom field first.
9878 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9881 @section fifo, afifo
9883 Buffer input images and send them when they are requested.
9885 It is mainly useful when auto-inserted by the libavfilter
9888 It does not take parameters.
9890 @section fillborders
9892 Fill borders of the input video, without changing video stream dimensions.
9893 Sometimes video can have garbage at the four edges and you may not want to
9894 crop video input to keep size multiple of some number.
9896 This filter accepts the following options:
9900 Number of pixels to fill from left border.
9903 Number of pixels to fill from right border.
9906 Number of pixels to fill from top border.
9909 Number of pixels to fill from bottom border.
9914 It accepts the following values:
9917 fill pixels using outermost pixels
9920 fill pixels using mirroring
9923 fill pixels with constant value
9926 Default is @var{smear}.
9929 Set color for pixels in fixed mode. Default is @var{black}.
9934 Find a rectangular object
9936 It accepts the following options:
9940 Filepath of the object image, needs to be in gray8.
9943 Detection threshold, default is 0.5.
9946 Number of mipmaps, default is 3.
9948 @item xmin, ymin, xmax, ymax
9949 Specifies the rectangle in which to search.
9952 @subsection Examples
9956 Generate a representative palette of a given video using @command{ffmpeg}:
9958 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9964 Cover a rectangular object
9966 It accepts the following options:
9970 Filepath of the optional cover image, needs to be in yuv420.
9975 It accepts the following values:
9978 cover it by the supplied image
9980 cover it by interpolating the surrounding pixels
9983 Default value is @var{blur}.
9986 @subsection Examples
9990 Generate a representative palette of a given video using @command{ffmpeg}:
9992 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9998 Flood area with values of same pixel components with another values.
10000 It accepts the following options:
10003 Set pixel x coordinate.
10006 Set pixel y coordinate.
10009 Set source #0 component value.
10012 Set source #1 component value.
10015 Set source #2 component value.
10018 Set source #3 component value.
10021 Set destination #0 component value.
10024 Set destination #1 component value.
10027 Set destination #2 component value.
10030 Set destination #3 component value.
10036 Convert the input video to one of the specified pixel formats.
10037 Libavfilter will try to pick one that is suitable as input to
10040 It accepts the following parameters:
10044 A '|'-separated list of pixel format names, such as
10045 "pix_fmts=yuv420p|monow|rgb24".
10049 @subsection Examples
10053 Convert the input video to the @var{yuv420p} format
10055 format=pix_fmts=yuv420p
10058 Convert the input video to any of the formats in the list
10060 format=pix_fmts=yuv420p|yuv444p|yuv410p
10067 Convert the video to specified constant frame rate by duplicating or dropping
10068 frames as necessary.
10070 It accepts the following parameters:
10074 The desired output frame rate. The default is @code{25}.
10077 Assume the first PTS should be the given value, in seconds. This allows for
10078 padding/trimming at the start of stream. By default, no assumption is made
10079 about the first frame's expected PTS, so no padding or trimming is done.
10080 For example, this could be set to 0 to pad the beginning with duplicates of
10081 the first frame if a video stream starts after the audio stream or to trim any
10082 frames with a negative PTS.
10085 Timestamp (PTS) rounding method.
10087 Possible values are:
10094 round towards -infinity
10096 round towards +infinity
10100 The default is @code{near}.
10103 Action performed when reading the last frame.
10105 Possible values are:
10108 Use same timestamp rounding method as used for other frames.
10110 Pass through last frame if input duration has not been reached yet.
10112 The default is @code{round}.
10116 Alternatively, the options can be specified as a flat string:
10117 @var{fps}[:@var{start_time}[:@var{round}]].
10119 See also the @ref{setpts} filter.
10121 @subsection Examples
10125 A typical usage in order to set the fps to 25:
10131 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10133 fps=fps=film:round=near
10139 Pack two different video streams into a stereoscopic video, setting proper
10140 metadata on supported codecs. The two views should have the same size and
10141 framerate and processing will stop when the shorter video ends. Please note
10142 that you may conveniently adjust view properties with the @ref{scale} and
10145 It accepts the following parameters:
10149 The desired packing format. Supported values are:
10154 The views are next to each other (default).
10157 The views are on top of each other.
10160 The views are packed by line.
10163 The views are packed by column.
10166 The views are temporally interleaved.
10175 # Convert left and right views into a frame-sequential video
10176 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10178 # Convert views into a side-by-side video with the same output resolution as the input
10179 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
10184 Change the frame rate by interpolating new video output frames from the source
10187 This filter is not designed to function correctly with interlaced media. If
10188 you wish to change the frame rate of interlaced media then you are required
10189 to deinterlace before this filter and re-interlace after this filter.
10191 A description of the accepted options follows.
10195 Specify the output frames per second. This option can also be specified
10196 as a value alone. The default is @code{50}.
10199 Specify the start of a range where the output frame will be created as a
10200 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10201 the default is @code{15}.
10204 Specify the end of a range where the output frame will be created as a
10205 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10206 the default is @code{240}.
10209 Specify the level at which a scene change is detected as a value between
10210 0 and 100 to indicate a new scene; a low value reflects a low
10211 probability for the current frame to introduce a new scene, while a higher
10212 value means the current frame is more likely to be one.
10213 The default is @code{8.2}.
10216 Specify flags influencing the filter process.
10218 Available value for @var{flags} is:
10221 @item scene_change_detect, scd
10222 Enable scene change detection using the value of the option @var{scene}.
10223 This flag is enabled by default.
10229 Select one frame every N-th frame.
10231 This filter accepts the following option:
10234 Select frame after every @code{step} frames.
10235 Allowed values are positive integers higher than 0. Default value is @code{1}.
10238 @section freezedetect
10240 Detect frozen video.
10242 This filter logs a message and sets frame metadata when it detects that the
10243 input video has no significant change in content during a specified duration.
10244 Video freeze detection calculates the mean average absolute difference of all
10245 the components of video frames and compares it to a noise floor.
10247 The printed times and duration are expressed in seconds. The
10248 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10249 whose timestamp equals or exceeds the detection duration and it contains the
10250 timestamp of the first frame of the freeze. The
10251 @code{lavfi.freezedetect.freeze_duration} and
10252 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10255 The filter accepts the following options:
10259 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10260 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10264 Set freeze duration until notification (default is 2 seconds).
10270 Apply a frei0r effect to the input video.
10272 To enable the compilation of this filter, you need to install the frei0r
10273 header and configure FFmpeg with @code{--enable-frei0r}.
10275 It accepts the following parameters:
10280 The name of the frei0r effect to load. If the environment variable
10281 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10282 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10283 Otherwise, the standard frei0r paths are searched, in this order:
10284 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10285 @file{/usr/lib/frei0r-1/}.
10287 @item filter_params
10288 A '|'-separated list of parameters to pass to the frei0r effect.
10292 A frei0r effect parameter can be a boolean (its value is either
10293 "y" or "n"), a double, a color (specified as
10294 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10295 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10296 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10297 a position (specified as @var{X}/@var{Y}, where
10298 @var{X} and @var{Y} are floating point numbers) and/or a string.
10300 The number and types of parameters depend on the loaded effect. If an
10301 effect parameter is not specified, the default value is set.
10303 @subsection Examples
10307 Apply the distort0r effect, setting the first two double parameters:
10309 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10313 Apply the colordistance effect, taking a color as the first parameter:
10315 frei0r=colordistance:0.2/0.3/0.4
10316 frei0r=colordistance:violet
10317 frei0r=colordistance:0x112233
10321 Apply the perspective effect, specifying the top left and top right image
10324 frei0r=perspective:0.2/0.2|0.8/0.2
10328 For more information, see
10329 @url{http://frei0r.dyne.org}
10333 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10335 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10336 processing filter, one of them is performed once per block, not per pixel.
10337 This allows for much higher speed.
10339 The filter accepts the following options:
10343 Set quality. This option defines the number of levels for averaging. It accepts
10344 an integer in the range 4-5. Default value is @code{4}.
10347 Force a constant quantization parameter. It accepts an integer in range 0-63.
10348 If not set, the filter will use the QP from the video stream (if available).
10351 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10352 more details but also more artifacts, while higher values make the image smoother
10353 but also blurrier. Default value is @code{0} − PSNR optimal.
10355 @item use_bframe_qp
10356 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10357 option may cause flicker since the B-Frames have often larger QP. Default is
10358 @code{0} (not enabled).
10364 Apply Gaussian blur filter.
10366 The filter accepts the following options:
10370 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10373 Set number of steps for Gaussian approximation. Default is @code{1}.
10376 Set which planes to filter. By default all planes are filtered.
10379 Set vertical sigma, if negative it will be same as @code{sigma}.
10380 Default is @code{-1}.
10385 Apply generic equation to each pixel.
10387 The filter accepts the following options:
10390 @item lum_expr, lum
10391 Set the luminance expression.
10393 Set the chrominance blue expression.
10395 Set the chrominance red expression.
10396 @item alpha_expr, a
10397 Set the alpha expression.
10399 Set the red expression.
10400 @item green_expr, g
10401 Set the green expression.
10403 Set the blue expression.
10406 The colorspace is selected according to the specified options. If one
10407 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10408 options is specified, the filter will automatically select a YCbCr
10409 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10410 @option{blue_expr} options is specified, it will select an RGB
10413 If one of the chrominance expression is not defined, it falls back on the other
10414 one. If no alpha expression is specified it will evaluate to opaque value.
10415 If none of chrominance expressions are specified, they will evaluate
10416 to the luminance expression.
10418 The expressions can use the following variables and functions:
10422 The sequential number of the filtered frame, starting from @code{0}.
10426 The coordinates of the current sample.
10430 The width and height of the image.
10434 Width and height scale depending on the currently filtered plane. It is the
10435 ratio between the corresponding luma plane number of pixels and the current
10436 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10437 @code{0.5,0.5} for chroma planes.
10440 Time of the current frame, expressed in seconds.
10443 Return the value of the pixel at location (@var{x},@var{y}) of the current
10447 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10451 Return the value of the pixel at location (@var{x},@var{y}) of the
10452 blue-difference chroma plane. Return 0 if there is no such plane.
10455 Return the value of the pixel at location (@var{x},@var{y}) of the
10456 red-difference chroma plane. Return 0 if there is no such plane.
10461 Return the value of the pixel at location (@var{x},@var{y}) of the
10462 red/green/blue component. Return 0 if there is no such component.
10465 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10466 plane. Return 0 if there is no such plane.
10469 For functions, if @var{x} and @var{y} are outside the area, the value will be
10470 automatically clipped to the closer edge.
10472 @subsection Examples
10476 Flip the image horizontally:
10482 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10483 wavelength of 100 pixels:
10485 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10489 Generate a fancy enigmatic moving light:
10491 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
10495 Generate a quick emboss effect:
10497 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10501 Modify RGB components depending on pixel position:
10503 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10507 Create a radial gradient that is the same size as the input (also see
10508 the @ref{vignette} filter):
10510 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10516 Fix the banding artifacts that are sometimes introduced into nearly flat
10517 regions by truncation to 8-bit color depth.
10518 Interpolate the gradients that should go where the bands are, and
10521 It is designed for playback only. Do not use it prior to
10522 lossy compression, because compression tends to lose the dither and
10523 bring back the bands.
10525 It accepts the following parameters:
10530 The maximum amount by which the filter will change any one pixel. This is also
10531 the threshold for detecting nearly flat regions. Acceptable values range from
10532 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10536 The neighborhood to fit the gradient to. A larger radius makes for smoother
10537 gradients, but also prevents the filter from modifying the pixels near detailed
10538 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10539 values will be clipped to the valid range.
10543 Alternatively, the options can be specified as a flat string:
10544 @var{strength}[:@var{radius}]
10546 @subsection Examples
10550 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10556 Specify radius, omitting the strength (which will fall-back to the default
10564 @section graphmonitor, agraphmonitor
10565 Show various filtergraph stats.
10567 With this filter one can debug complete filtergraph.
10568 Especially issues with links filling with queued frames.
10570 The filter accepts the following options:
10574 Set video output size. Default is @var{hd720}.
10577 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10580 Set output mode, can be @var{fulll} or @var{compact}.
10581 In @var{compact} mode only filters with some queued frames have displayed stats.
10584 Set flags which enable which stats are shown in video.
10586 Available values for flags are:
10589 Display number of queued frames in each link.
10591 @item frame_count_in
10592 Display number of frames taken from filter.
10594 @item frame_count_out
10595 Display number of frames given out from filter.
10598 Display current filtered frame pts.
10601 Display current filtered frame time.
10604 Display time base for filter link.
10607 Display used format for filter link.
10610 Display video size or number of audio channels in case of audio used by filter link.
10613 Display video frame rate or sample rate in case of audio used by filter link.
10617 Set upper limit for video rate of output stream, Default value is @var{25}.
10618 This guarantee that output video frame rate will not be higher than this value.
10622 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10623 and corrects the scene colors accordingly.
10625 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10627 The filter accepts the following options:
10631 The order of differentiation to be applied on the scene. Must be chosen in the range
10632 [0,2] and default value is 1.
10635 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10636 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10637 max value instead of calculating Minkowski distance.
10640 The standard deviation of Gaussian blur to be applied on the scene. Must be
10641 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10642 can't be equal to 0 if @var{difford} is greater than 0.
10645 @subsection Examples
10651 greyedge=difford=1:minknorm=5:sigma=2
10657 greyedge=difford=1:minknorm=0:sigma=2
10665 Apply a Hald CLUT to a video stream.
10667 First input is the video stream to process, and second one is the Hald CLUT.
10668 The Hald CLUT input can be a simple picture or a complete video stream.
10670 The filter accepts the following options:
10674 Force termination when the shortest input terminates. Default is @code{0}.
10676 Continue applying the last CLUT after the end of the stream. A value of
10677 @code{0} disable the filter after the last frame of the CLUT is reached.
10678 Default is @code{1}.
10681 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10682 filters share the same internals).
10684 More information about the Hald CLUT can be found on Eskil Steenberg's website
10685 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10687 @subsection Workflow examples
10689 @subsubsection Hald CLUT video stream
10691 Generate an identity Hald CLUT stream altered with various effects:
10693 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
10696 Note: make sure you use a lossless codec.
10698 Then use it with @code{haldclut} to apply it on some random stream:
10700 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10703 The Hald CLUT will be applied to the 10 first seconds (duration of
10704 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10705 to the remaining frames of the @code{mandelbrot} stream.
10707 @subsubsection Hald CLUT with preview
10709 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10710 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10711 biggest possible square starting at the top left of the picture. The remaining
10712 padding pixels (bottom or right) will be ignored. This area can be used to add
10713 a preview of the Hald CLUT.
10715 Typically, the following generated Hald CLUT will be supported by the
10716 @code{haldclut} filter:
10719 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10720 pad=iw+320 [padded_clut];
10721 smptebars=s=320x256, split [a][b];
10722 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10723 [main][b] overlay=W-320" -frames:v 1 clut.png
10726 It contains the original and a preview of the effect of the CLUT: SMPTE color
10727 bars are displayed on the right-top, and below the same color bars processed by
10730 Then, the effect of this Hald CLUT can be visualized with:
10732 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10737 Flip the input video horizontally.
10739 For example, to horizontally flip the input video with @command{ffmpeg}:
10741 ffmpeg -i in.avi -vf "hflip" out.avi
10745 This filter applies a global color histogram equalization on a
10748 It can be used to correct video that has a compressed range of pixel
10749 intensities. The filter redistributes the pixel intensities to
10750 equalize their distribution across the intensity range. It may be
10751 viewed as an "automatically adjusting contrast filter". This filter is
10752 useful only for correcting degraded or poorly captured source
10755 The filter accepts the following options:
10759 Determine the amount of equalization to be applied. As the strength
10760 is reduced, the distribution of pixel intensities more-and-more
10761 approaches that of the input frame. The value must be a float number
10762 in the range [0,1] and defaults to 0.200.
10765 Set the maximum intensity that can generated and scale the output
10766 values appropriately. The strength should be set as desired and then
10767 the intensity can be limited if needed to avoid washing-out. The value
10768 must be a float number in the range [0,1] and defaults to 0.210.
10771 Set the antibanding level. If enabled the filter will randomly vary
10772 the luminance of output pixels by a small amount to avoid banding of
10773 the histogram. Possible values are @code{none}, @code{weak} or
10774 @code{strong}. It defaults to @code{none}.
10779 Compute and draw a color distribution histogram for the input video.
10781 The computed histogram is a representation of the color component
10782 distribution in an image.
10784 Standard histogram displays the color components distribution in an image.
10785 Displays color graph for each color component. Shows distribution of
10786 the Y, U, V, A or R, G, B components, depending on input format, in the
10787 current frame. Below each graph a color component scale meter is shown.
10789 The filter accepts the following options:
10793 Set height of level. Default value is @code{200}.
10794 Allowed range is [50, 2048].
10797 Set height of color scale. Default value is @code{12}.
10798 Allowed range is [0, 40].
10802 It accepts the following values:
10805 Per color component graphs are placed below each other.
10808 Per color component graphs are placed side by side.
10811 Presents information identical to that in the @code{parade}, except
10812 that the graphs representing color components are superimposed directly
10815 Default is @code{stack}.
10818 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10819 Default is @code{linear}.
10822 Set what color components to display.
10823 Default is @code{7}.
10826 Set foreground opacity. Default is @code{0.7}.
10829 Set background opacity. Default is @code{0.5}.
10832 @subsection Examples
10837 Calculate and draw histogram:
10839 ffplay -i input -vf histogram
10847 This is a high precision/quality 3d denoise filter. It aims to reduce
10848 image noise, producing smooth images and making still images really
10849 still. It should enhance compressibility.
10851 It accepts the following optional parameters:
10855 A non-negative floating point number which specifies spatial luma strength.
10856 It defaults to 4.0.
10858 @item chroma_spatial
10859 A non-negative floating point number which specifies spatial chroma strength.
10860 It defaults to 3.0*@var{luma_spatial}/4.0.
10863 A floating point number which specifies luma temporal strength. It defaults to
10864 6.0*@var{luma_spatial}/4.0.
10867 A floating point number which specifies chroma temporal strength. It defaults to
10868 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10871 @anchor{hwdownload}
10872 @section hwdownload
10874 Download hardware frames to system memory.
10876 The input must be in hardware frames, and the output a non-hardware format.
10877 Not all formats will be supported on the output - it may be necessary to insert
10878 an additional @option{format} filter immediately following in the graph to get
10879 the output in a supported format.
10883 Map hardware frames to system memory or to another device.
10885 This filter has several different modes of operation; which one is used depends
10886 on the input and output formats:
10889 Hardware frame input, normal frame output
10891 Map the input frames to system memory and pass them to the output. If the
10892 original hardware frame is later required (for example, after overlaying
10893 something else on part of it), the @option{hwmap} filter can be used again
10894 in the next mode to retrieve it.
10896 Normal frame input, hardware frame output
10898 If the input is actually a software-mapped hardware frame, then unmap it -
10899 that is, return the original hardware frame.
10901 Otherwise, a device must be provided. Create new hardware surfaces on that
10902 device for the output, then map them back to the software format at the input
10903 and give those frames to the preceding filter. This will then act like the
10904 @option{hwupload} filter, but may be able to avoid an additional copy when
10905 the input is already in a compatible format.
10907 Hardware frame input and output
10909 A device must be supplied for the output, either directly or with the
10910 @option{derive_device} option. The input and output devices must be of
10911 different types and compatible - the exact meaning of this is
10912 system-dependent, but typically it means that they must refer to the same
10913 underlying hardware context (for example, refer to the same graphics card).
10915 If the input frames were originally created on the output device, then unmap
10916 to retrieve the original frames.
10918 Otherwise, map the frames to the output device - create new hardware frames
10919 on the output corresponding to the frames on the input.
10922 The following additional parameters are accepted:
10926 Set the frame mapping mode. Some combination of:
10929 The mapped frame should be readable.
10931 The mapped frame should be writeable.
10933 The mapping will always overwrite the entire frame.
10935 This may improve performance in some cases, as the original contents of the
10936 frame need not be loaded.
10938 The mapping must not involve any copying.
10940 Indirect mappings to copies of frames are created in some cases where either
10941 direct mapping is not possible or it would have unexpected properties.
10942 Setting this flag ensures that the mapping is direct and will fail if that is
10945 Defaults to @var{read+write} if not specified.
10947 @item derive_device @var{type}
10948 Rather than using the device supplied at initialisation, instead derive a new
10949 device of type @var{type} from the device the input frames exist on.
10952 In a hardware to hardware mapping, map in reverse - create frames in the sink
10953 and map them back to the source. This may be necessary in some cases where
10954 a mapping in one direction is required but only the opposite direction is
10955 supported by the devices being used.
10957 This option is dangerous - it may break the preceding filter in undefined
10958 ways if there are any additional constraints on that filter's output.
10959 Do not use it without fully understanding the implications of its use.
10965 Upload system memory frames to hardware surfaces.
10967 The device to upload to must be supplied when the filter is initialised. If
10968 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10971 @anchor{hwupload_cuda}
10972 @section hwupload_cuda
10974 Upload system memory frames to a CUDA device.
10976 It accepts the following optional parameters:
10980 The number of the CUDA device to use
10985 Apply a high-quality magnification filter designed for pixel art. This filter
10986 was originally created by Maxim Stepin.
10988 It accepts the following option:
10992 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10993 @code{hq3x} and @code{4} for @code{hq4x}.
10994 Default is @code{3}.
10998 Stack input videos horizontally.
11000 All streams must be of same pixel format and of same height.
11002 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
11003 to create same output.
11005 The filter accept the following option:
11009 Set number of input streams. Default is 2.
11012 If set to 1, force the output to terminate when the shortest input
11013 terminates. Default value is 0.
11018 Modify the hue and/or the saturation of the input.
11020 It accepts the following parameters:
11024 Specify the hue angle as a number of degrees. It accepts an expression,
11025 and defaults to "0".
11028 Specify the saturation in the [-10,10] range. It accepts an expression and
11032 Specify the hue angle as a number of radians. It accepts an
11033 expression, and defaults to "0".
11036 Specify the brightness in the [-10,10] range. It accepts an expression and
11040 @option{h} and @option{H} are mutually exclusive, and can't be
11041 specified at the same time.
11043 The @option{b}, @option{h}, @option{H} and @option{s} option values are
11044 expressions containing the following constants:
11048 frame count of the input frame starting from 0
11051 presentation timestamp of the input frame expressed in time base units
11054 frame rate of the input video, NAN if the input frame rate is unknown
11057 timestamp expressed in seconds, NAN if the input timestamp is unknown
11060 time base of the input video
11063 @subsection Examples
11067 Set the hue to 90 degrees and the saturation to 1.0:
11073 Same command but expressing the hue in radians:
11079 Rotate hue and make the saturation swing between 0
11080 and 2 over a period of 1 second:
11082 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11086 Apply a 3 seconds saturation fade-in effect starting at 0:
11088 hue="s=min(t/3\,1)"
11091 The general fade-in expression can be written as:
11093 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11097 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11099 hue="s=max(0\, min(1\, (8-t)/3))"
11102 The general fade-out expression can be written as:
11104 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11109 @subsection Commands
11111 This filter supports the following commands:
11117 Modify the hue and/or the saturation and/or brightness of the input video.
11118 The command accepts the same syntax of the corresponding option.
11120 If the specified expression is not valid, it is kept at its current
11124 @section hysteresis
11126 Grow first stream into second stream by connecting components.
11127 This makes it possible to build more robust edge masks.
11129 This filter accepts the following options:
11133 Set which planes will be processed as bitmap, unprocessed planes will be
11134 copied from first stream.
11135 By default value 0xf, all planes will be processed.
11138 Set threshold which is used in filtering. If pixel component value is higher than
11139 this value filter algorithm for connecting components is activated.
11140 By default value is 0.
11145 Detect video interlacing type.
11147 This filter tries to detect if the input frames are interlaced, progressive,
11148 top or bottom field first. It will also try to detect fields that are
11149 repeated between adjacent frames (a sign of telecine).
11151 Single frame detection considers only immediately adjacent frames when classifying each frame.
11152 Multiple frame detection incorporates the classification history of previous frames.
11154 The filter will log these metadata values:
11157 @item single.current_frame
11158 Detected type of current frame using single-frame detection. One of:
11159 ``tff'' (top field first), ``bff'' (bottom field first),
11160 ``progressive'', or ``undetermined''
11163 Cumulative number of frames detected as top field first using single-frame detection.
11166 Cumulative number of frames detected as top field first using multiple-frame detection.
11169 Cumulative number of frames detected as bottom field first using single-frame detection.
11171 @item multiple.current_frame
11172 Detected type of current frame using multiple-frame detection. One of:
11173 ``tff'' (top field first), ``bff'' (bottom field first),
11174 ``progressive'', or ``undetermined''
11177 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11179 @item single.progressive
11180 Cumulative number of frames detected as progressive using single-frame detection.
11182 @item multiple.progressive
11183 Cumulative number of frames detected as progressive using multiple-frame detection.
11185 @item single.undetermined
11186 Cumulative number of frames that could not be classified using single-frame detection.
11188 @item multiple.undetermined
11189 Cumulative number of frames that could not be classified using multiple-frame detection.
11191 @item repeated.current_frame
11192 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11194 @item repeated.neither
11195 Cumulative number of frames with no repeated field.
11198 Cumulative number of frames with the top field repeated from the previous frame's top field.
11200 @item repeated.bottom
11201 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11204 The filter accepts the following options:
11208 Set interlacing threshold.
11210 Set progressive threshold.
11212 Threshold for repeated field detection.
11214 Number of frames after which a given frame's contribution to the
11215 statistics is halved (i.e., it contributes only 0.5 to its
11216 classification). The default of 0 means that all frames seen are given
11217 full weight of 1.0 forever.
11218 @item analyze_interlaced_flag
11219 When this is not 0 then idet will use the specified number of frames to determine
11220 if the interlaced flag is accurate, it will not count undetermined frames.
11221 If the flag is found to be accurate it will be used without any further
11222 computations, if it is found to be inaccurate it will be cleared without any
11223 further computations. This allows inserting the idet filter as a low computational
11224 method to clean up the interlaced flag
11229 Deinterleave or interleave fields.
11231 This filter allows one to process interlaced images fields without
11232 deinterlacing them. Deinterleaving splits the input frame into 2
11233 fields (so called half pictures). Odd lines are moved to the top
11234 half of the output image, even lines to the bottom half.
11235 You can process (filter) them independently and then re-interleave them.
11237 The filter accepts the following options:
11241 @item chroma_mode, c
11242 @item alpha_mode, a
11243 Available values for @var{luma_mode}, @var{chroma_mode} and
11244 @var{alpha_mode} are:
11250 @item deinterleave, d
11251 Deinterleave fields, placing one above the other.
11253 @item interleave, i
11254 Interleave fields. Reverse the effect of deinterleaving.
11256 Default value is @code{none}.
11258 @item luma_swap, ls
11259 @item chroma_swap, cs
11260 @item alpha_swap, as
11261 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11266 Apply inflate effect to the video.
11268 This filter replaces the pixel by the local(3x3) average by taking into account
11269 only values higher than the pixel.
11271 It accepts the following options:
11278 Limit the maximum change for each plane, default is 65535.
11279 If 0, plane will remain unchanged.
11284 Simple interlacing filter from progressive contents. This interleaves upper (or
11285 lower) lines from odd frames with lower (or upper) lines from even frames,
11286 halving the frame rate and preserving image height.
11289 Original Original New Frame
11290 Frame 'j' Frame 'j+1' (tff)
11291 ========== =========== ==================
11292 Line 0 --------------------> Frame 'j' Line 0
11293 Line 1 Line 1 ----> Frame 'j+1' Line 1
11294 Line 2 ---------------------> Frame 'j' Line 2
11295 Line 3 Line 3 ----> Frame 'j+1' Line 3
11297 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11300 It accepts the following optional parameters:
11304 This determines whether the interlaced frame is taken from the even
11305 (tff - default) or odd (bff) lines of the progressive frame.
11308 Vertical lowpass filter to avoid twitter interlacing and
11309 reduce moire patterns.
11313 Disable vertical lowpass filter
11316 Enable linear filter (default)
11319 Enable complex filter. This will slightly less reduce twitter and moire
11320 but better retain detail and subjective sharpness impression.
11327 Deinterlace input video by applying Donald Graft's adaptive kernel
11328 deinterling. Work on interlaced parts of a video to produce
11329 progressive frames.
11331 The description of the accepted parameters follows.
11335 Set the threshold which affects the filter's tolerance when
11336 determining if a pixel line must be processed. It must be an integer
11337 in the range [0,255] and defaults to 10. A value of 0 will result in
11338 applying the process on every pixels.
11341 Paint pixels exceeding the threshold value to white if set to 1.
11345 Set the fields order. Swap fields if set to 1, leave fields alone if
11349 Enable additional sharpening if set to 1. Default is 0.
11352 Enable twoway sharpening if set to 1. Default is 0.
11355 @subsection Examples
11359 Apply default values:
11361 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11365 Enable additional sharpening:
11371 Paint processed pixels in white:
11379 Slowly update darker pixels.
11381 This filter makes short flashes of light appear longer.
11382 This filter accepts the following options:
11386 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
11389 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
11392 @section lenscorrection
11394 Correct radial lens distortion
11396 This filter can be used to correct for radial distortion as can result from the use
11397 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11398 one can use tools available for example as part of opencv or simply trial-and-error.
11399 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11400 and extract the k1 and k2 coefficients from the resulting matrix.
11402 Note that effectively the same filter is available in the open-source tools Krita and
11403 Digikam from the KDE project.
11405 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11406 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11407 brightness distribution, so you may want to use both filters together in certain
11408 cases, though you will have to take care of ordering, i.e. whether vignetting should
11409 be applied before or after lens correction.
11411 @subsection Options
11413 The filter accepts the following options:
11417 Relative x-coordinate of the focal point of the image, and thereby the center of the
11418 distortion. This value has a range [0,1] and is expressed as fractions of the image
11419 width. Default is 0.5.
11421 Relative y-coordinate of the focal point of the image, and thereby the center of the
11422 distortion. This value has a range [0,1] and is expressed as fractions of the image
11423 height. Default is 0.5.
11425 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11426 no correction. Default is 0.
11428 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11429 0 means no correction. Default is 0.
11432 The formula that generates the correction is:
11434 @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)
11436 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11437 distances from the focal point in the source and target images, respectively.
11441 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11443 The @code{lensfun} filter requires the camera make, camera model, and lens model
11444 to apply the lens correction. The filter will load the lensfun database and
11445 query it to find the corresponding camera and lens entries in the database. As
11446 long as these entries can be found with the given options, the filter can
11447 perform corrections on frames. Note that incomplete strings will result in the
11448 filter choosing the best match with the given options, and the filter will
11449 output the chosen camera and lens models (logged with level "info"). You must
11450 provide the make, camera model, and lens model as they are required.
11452 The filter accepts the following options:
11456 The make of the camera (for example, "Canon"). This option is required.
11459 The model of the camera (for example, "Canon EOS 100D"). This option is
11463 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11464 option is required.
11467 The type of correction to apply. The following values are valid options:
11471 Enables fixing lens vignetting.
11474 Enables fixing lens geometry. This is the default.
11477 Enables fixing chromatic aberrations.
11480 Enables fixing lens vignetting and lens geometry.
11483 Enables fixing lens vignetting and chromatic aberrations.
11486 Enables fixing both lens geometry and chromatic aberrations.
11489 Enables all possible corrections.
11493 The focal length of the image/video (zoom; expected constant for video). For
11494 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11495 range should be chosen when using that lens. Default 18.
11498 The aperture of the image/video (expected constant for video). Note that
11499 aperture is only used for vignetting correction. Default 3.5.
11501 @item focus_distance
11502 The focus distance of the image/video (expected constant for video). Note that
11503 focus distance is only used for vignetting and only slightly affects the
11504 vignetting correction process. If unknown, leave it at the default value (which
11508 The scale factor which is applied after transformation. After correction the
11509 video is no longer necessarily rectangular. This parameter controls how much of
11510 the resulting image is visible. The value 0 means that a value will be chosen
11511 automatically such that there is little or no unmapped area in the output
11512 image. 1.0 means that no additional scaling is done. Lower values may result
11513 in more of the corrected image being visible, while higher values may avoid
11514 unmapped areas in the output.
11516 @item target_geometry
11517 The target geometry of the output image/video. The following values are valid
11521 @item rectilinear (default)
11524 @item equirectangular
11525 @item fisheye_orthographic
11526 @item fisheye_stereographic
11527 @item fisheye_equisolid
11528 @item fisheye_thoby
11531 Apply the reverse of image correction (instead of correcting distortion, apply
11534 @item interpolation
11535 The type of interpolation used when correcting distortion. The following values
11540 @item linear (default)
11545 @subsection Examples
11549 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11550 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11554 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
11558 Apply the same as before, but only for the first 5 seconds of video.
11561 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
11568 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11569 score between two input videos.
11571 The obtained VMAF score is printed through the logging system.
11573 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11574 After installing the library it can be enabled using:
11575 @code{./configure --enable-libvmaf --enable-version3}.
11576 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11578 The filter has following options:
11582 Set the model path which is to be used for SVM.
11583 Default value: @code{"vmaf_v0.6.1.pkl"}
11586 Set the file path to be used to store logs.
11589 Set the format of the log file (xml or json).
11591 @item enable_transform
11592 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11593 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11594 Default value: @code{false}
11597 Invokes the phone model which will generate VMAF scores higher than in the
11598 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11601 Enables computing psnr along with vmaf.
11604 Enables computing ssim along with vmaf.
11607 Enables computing ms_ssim along with vmaf.
11610 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11613 Set number of threads to be used when computing vmaf.
11616 Set interval for frame subsampling used when computing vmaf.
11618 @item enable_conf_interval
11619 Enables confidence interval.
11622 This filter also supports the @ref{framesync} options.
11624 On the below examples the input file @file{main.mpg} being processed is
11625 compared with the reference file @file{ref.mpg}.
11628 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11631 Example with options:
11633 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11638 Limits the pixel components values to the specified range [min, max].
11640 The filter accepts the following options:
11644 Lower bound. Defaults to the lowest allowed value for the input.
11647 Upper bound. Defaults to the highest allowed value for the input.
11650 Specify which planes will be processed. Defaults to all available.
11657 The filter accepts the following options:
11661 Set the number of loops. Setting this value to -1 will result in infinite loops.
11665 Set maximal size in number of frames. Default is 0.
11668 Set first frame of loop. Default is 0.
11671 @subsection Examples
11675 Loop single first frame infinitely:
11677 loop=loop=-1:size=1:start=0
11681 Loop single first frame 10 times:
11683 loop=loop=10:size=1:start=0
11687 Loop 10 first frames 5 times:
11689 loop=loop=5:size=10:start=0
11695 Apply a 1D LUT to an input video.
11697 The filter accepts the following options:
11701 Set the 1D LUT file name.
11703 Currently supported formats:
11712 Select interpolation mode.
11714 Available values are:
11718 Use values from the nearest defined point.
11720 Interpolate values using the linear interpolation.
11722 Interpolate values using the cosine interpolation.
11724 Interpolate values using the cubic interpolation.
11726 Interpolate values using the spline interpolation.
11733 Apply a 3D LUT to an input video.
11735 The filter accepts the following options:
11739 Set the 3D LUT file name.
11741 Currently supported formats:
11755 Select interpolation mode.
11757 Available values are:
11761 Use values from the nearest defined point.
11763 Interpolate values using the 8 points defining a cube.
11765 Interpolate values using a tetrahedron.
11769 This filter also supports the @ref{framesync} options.
11773 Turn certain luma values into transparency.
11775 The filter accepts the following options:
11779 Set the luma which will be used as base for transparency.
11780 Default value is @code{0}.
11783 Set the range of luma values to be keyed out.
11784 Default value is @code{0}.
11787 Set the range of softness. Default value is @code{0}.
11788 Use this to control gradual transition from zero to full transparency.
11791 @section lut, lutrgb, lutyuv
11793 Compute a look-up table for binding each pixel component input value
11794 to an output value, and apply it to the input video.
11796 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11797 to an RGB input video.
11799 These filters accept the following parameters:
11802 set first pixel component expression
11804 set second pixel component expression
11806 set third pixel component expression
11808 set fourth pixel component expression, corresponds to the alpha component
11811 set red component expression
11813 set green component expression
11815 set blue component expression
11817 alpha component expression
11820 set Y/luminance component expression
11822 set U/Cb component expression
11824 set V/Cr component expression
11827 Each of them specifies the expression to use for computing the lookup table for
11828 the corresponding pixel component values.
11830 The exact component associated to each of the @var{c*} options depends on the
11833 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11834 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11836 The expressions can contain the following constants and functions:
11841 The input width and height.
11844 The input value for the pixel component.
11847 The input value, clipped to the @var{minval}-@var{maxval} range.
11850 The maximum value for the pixel component.
11853 The minimum value for the pixel component.
11856 The negated value for the pixel component value, clipped to the
11857 @var{minval}-@var{maxval} range; it corresponds to the expression
11858 "maxval-clipval+minval".
11861 The computed value in @var{val}, clipped to the
11862 @var{minval}-@var{maxval} range.
11864 @item gammaval(gamma)
11865 The computed gamma correction value of the pixel component value,
11866 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11868 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11872 All expressions default to "val".
11874 @subsection Examples
11878 Negate input video:
11880 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11881 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11884 The above is the same as:
11886 lutrgb="r=negval:g=negval:b=negval"
11887 lutyuv="y=negval:u=negval:v=negval"
11897 Remove chroma components, turning the video into a graytone image:
11899 lutyuv="u=128:v=128"
11903 Apply a luma burning effect:
11909 Remove green and blue components:
11915 Set a constant alpha channel value on input:
11917 format=rgba,lutrgb=a="maxval-minval/2"
11921 Correct luminance gamma by a factor of 0.5:
11923 lutyuv=y=gammaval(0.5)
11927 Discard least significant bits of luma:
11929 lutyuv=y='bitand(val, 128+64+32)'
11933 Technicolor like effect:
11935 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11939 @section lut2, tlut2
11941 The @code{lut2} filter takes two input streams and outputs one
11944 The @code{tlut2} (time lut2) filter takes two consecutive frames
11945 from one single stream.
11947 This filter accepts the following parameters:
11950 set first pixel component expression
11952 set second pixel component expression
11954 set third pixel component expression
11956 set fourth pixel component expression, corresponds to the alpha component
11959 set output bit depth, only available for @code{lut2} filter. By default is 0,
11960 which means bit depth is automatically picked from first input format.
11963 Each of them specifies the expression to use for computing the lookup table for
11964 the corresponding pixel component values.
11966 The exact component associated to each of the @var{c*} options depends on the
11969 The expressions can contain the following constants:
11974 The input width and height.
11977 The first input value for the pixel component.
11980 The second input value for the pixel component.
11983 The first input video bit depth.
11986 The second input video bit depth.
11989 All expressions default to "x".
11991 @subsection Examples
11995 Highlight differences between two RGB video streams:
11997 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)'
12001 Highlight differences between two YUV video streams:
12003 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)'
12007 Show max difference between two video streams:
12009 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)))'
12013 @section maskedclamp
12015 Clamp the first input stream with the second input and third input stream.
12017 Returns the value of first stream to be between second input
12018 stream - @code{undershoot} and third input stream + @code{overshoot}.
12020 This filter accepts the following options:
12023 Default value is @code{0}.
12026 Default value is @code{0}.
12029 Set which planes will be processed as bitmap, unprocessed planes will be
12030 copied from first stream.
12031 By default value 0xf, all planes will be processed.
12034 @section maskedmerge
12036 Merge the first input stream with the second input stream using per pixel
12037 weights in the third input stream.
12039 A value of 0 in the third stream pixel component means that pixel component
12040 from first stream is returned unchanged, while maximum value (eg. 255 for
12041 8-bit videos) means that pixel component from second stream is returned
12042 unchanged. Intermediate values define the amount of merging between both
12043 input stream's pixel components.
12045 This filter accepts the following options:
12048 Set which planes will be processed as bitmap, unprocessed planes will be
12049 copied from first stream.
12050 By default value 0xf, all planes will be processed.
12054 Create mask from input video.
12056 For example it is useful to create motion masks after @code{tblend} filter.
12058 This filter accepts the following options:
12062 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
12065 Set high threshold. Any pixel component higher than this value will be set to max value
12066 allowed for current pixel format.
12069 Set planes to filter, by default all available planes are filtered.
12072 Fill all frame pixels with this value.
12075 Set max average pixel value for frame. If sum of all pixel components is higher that this
12076 average, output frame will be completely filled with value set by @var{fill} option.
12077 Typically useful for scene changes when used in combination with @code{tblend} filter.
12082 Apply motion-compensation deinterlacing.
12084 It needs one field per frame as input and must thus be used together
12085 with yadif=1/3 or equivalent.
12087 This filter accepts the following options:
12090 Set the deinterlacing mode.
12092 It accepts one of the following values:
12097 use iterative motion estimation
12099 like @samp{slow}, but use multiple reference frames.
12101 Default value is @samp{fast}.
12104 Set the picture field parity assumed for the input video. It must be
12105 one of the following values:
12109 assume top field first
12111 assume bottom field first
12114 Default value is @samp{bff}.
12117 Set per-block quantization parameter (QP) used by the internal
12120 Higher values should result in a smoother motion vector field but less
12121 optimal individual vectors. Default value is 1.
12124 @section mergeplanes
12126 Merge color channel components from several video streams.
12128 The filter accepts up to 4 input streams, and merge selected input
12129 planes to the output video.
12131 This filter accepts the following options:
12134 Set input to output plane mapping. Default is @code{0}.
12136 The mappings is specified as a bitmap. It should be specified as a
12137 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12138 mapping for the first plane of the output stream. 'A' sets the number of
12139 the input stream to use (from 0 to 3), and 'a' the plane number of the
12140 corresponding input to use (from 0 to 3). The rest of the mappings is
12141 similar, 'Bb' describes the mapping for the output stream second
12142 plane, 'Cc' describes the mapping for the output stream third plane and
12143 'Dd' describes the mapping for the output stream fourth plane.
12146 Set output pixel format. Default is @code{yuva444p}.
12149 @subsection Examples
12153 Merge three gray video streams of same width and height into single video stream:
12155 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12159 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12161 [a0][a1]mergeplanes=0x00010210:yuva444p
12165 Swap Y and A plane in yuva444p stream:
12167 format=yuva444p,mergeplanes=0x03010200:yuva444p
12171 Swap U and V plane in yuv420p stream:
12173 format=yuv420p,mergeplanes=0x000201:yuv420p
12177 Cast a rgb24 clip to yuv444p:
12179 format=rgb24,mergeplanes=0x000102:yuv444p
12185 Estimate and export motion vectors using block matching algorithms.
12186 Motion vectors are stored in frame side data to be used by other filters.
12188 This filter accepts the following options:
12191 Specify the motion estimation method. Accepts one of the following values:
12195 Exhaustive search algorithm.
12197 Three step search algorithm.
12199 Two dimensional logarithmic search algorithm.
12201 New three step search algorithm.
12203 Four step search algorithm.
12205 Diamond search algorithm.
12207 Hexagon-based search algorithm.
12209 Enhanced predictive zonal search algorithm.
12211 Uneven multi-hexagon search algorithm.
12213 Default value is @samp{esa}.
12216 Macroblock size. Default @code{16}.
12219 Search parameter. Default @code{7}.
12222 @section midequalizer
12224 Apply Midway Image Equalization effect using two video streams.
12226 Midway Image Equalization adjusts a pair of images to have the same
12227 histogram, while maintaining their dynamics as much as possible. It's
12228 useful for e.g. matching exposures from a pair of stereo cameras.
12230 This filter has two inputs and one output, which must be of same pixel format, but
12231 may be of different sizes. The output of filter is first input adjusted with
12232 midway histogram of both inputs.
12234 This filter accepts the following option:
12238 Set which planes to process. Default is @code{15}, which is all available planes.
12241 @section minterpolate
12243 Convert the video to specified frame rate using motion interpolation.
12245 This filter accepts the following options:
12248 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}.
12251 Motion interpolation mode. Following values are accepted:
12254 Duplicate previous or next frame for interpolating new ones.
12256 Blend source frames. Interpolated frame is mean of previous and next frames.
12258 Motion compensated interpolation. Following options are effective when this mode is selected:
12262 Motion compensation mode. Following values are accepted:
12265 Overlapped block motion compensation.
12267 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12269 Default mode is @samp{obmc}.
12272 Motion estimation mode. Following values are accepted:
12275 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12277 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12279 Default mode is @samp{bilat}.
12282 The algorithm to be used for motion estimation. Following values are accepted:
12285 Exhaustive search algorithm.
12287 Three step search algorithm.
12289 Two dimensional logarithmic search algorithm.
12291 New three step search algorithm.
12293 Four step search algorithm.
12295 Diamond search algorithm.
12297 Hexagon-based search algorithm.
12299 Enhanced predictive zonal search algorithm.
12301 Uneven multi-hexagon search algorithm.
12303 Default algorithm is @samp{epzs}.
12306 Macroblock size. Default @code{16}.
12309 Motion estimation search parameter. Default @code{32}.
12312 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).
12317 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:
12320 Disable scene change detection.
12322 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12324 Default method is @samp{fdiff}.
12326 @item scd_threshold
12327 Scene change detection threshold. Default is @code{5.0}.
12332 Mix several video input streams into one video stream.
12334 A description of the accepted options follows.
12338 The number of inputs. If unspecified, it defaults to 2.
12341 Specify weight of each input video stream as sequence.
12342 Each weight is separated by space. If number of weights
12343 is smaller than number of @var{frames} last specified
12344 weight will be used for all remaining unset weights.
12347 Specify scale, if it is set it will be multiplied with sum
12348 of each weight multiplied with pixel values to give final destination
12349 pixel value. By default @var{scale} is auto scaled to sum of weights.
12352 Specify how end of stream is determined.
12355 The duration of the longest input. (default)
12358 The duration of the shortest input.
12361 The duration of the first input.
12365 @section mpdecimate
12367 Drop frames that do not differ greatly from the previous frame in
12368 order to reduce frame rate.
12370 The main use of this filter is for very-low-bitrate encoding
12371 (e.g. streaming over dialup modem), but it could in theory be used for
12372 fixing movies that were inverse-telecined incorrectly.
12374 A description of the accepted options follows.
12378 Set the maximum number of consecutive frames which can be dropped (if
12379 positive), or the minimum interval between dropped frames (if
12380 negative). If the value is 0, the frame is dropped disregarding the
12381 number of previous sequentially dropped frames.
12383 Default value is 0.
12388 Set the dropping threshold values.
12390 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12391 represent actual pixel value differences, so a threshold of 64
12392 corresponds to 1 unit of difference for each pixel, or the same spread
12393 out differently over the block.
12395 A frame is a candidate for dropping if no 8x8 blocks differ by more
12396 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12397 meaning the whole image) differ by more than a threshold of @option{lo}.
12399 Default value for @option{hi} is 64*12, default value for @option{lo} is
12400 64*5, and default value for @option{frac} is 0.33.
12406 Negate (invert) the input video.
12408 It accepts the following option:
12413 With value 1, it negates the alpha component, if present. Default value is 0.
12419 Denoise frames using Non-Local Means algorithm.
12421 Each pixel is adjusted by looking for other pixels with similar contexts. This
12422 context similarity is defined by comparing their surrounding patches of size
12423 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12426 Note that the research area defines centers for patches, which means some
12427 patches will be made of pixels outside that research area.
12429 The filter accepts the following options.
12433 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12436 Set patch size. Default is 7. Must be odd number in range [0, 99].
12439 Same as @option{p} but for chroma planes.
12441 The default value is @var{0} and means automatic.
12444 Set research size. Default is 15. Must be odd number in range [0, 99].
12447 Same as @option{r} but for chroma planes.
12449 The default value is @var{0} and means automatic.
12454 Deinterlace video using neural network edge directed interpolation.
12456 This filter accepts the following options:
12460 Mandatory option, without binary file filter can not work.
12461 Currently file can be found here:
12462 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12465 Set which frames to deinterlace, by default it is @code{all}.
12466 Can be @code{all} or @code{interlaced}.
12469 Set mode of operation.
12471 Can be one of the following:
12475 Use frame flags, both fields.
12477 Use frame flags, single field.
12479 Use top field only.
12481 Use bottom field only.
12483 Use both fields, top first.
12485 Use both fields, bottom first.
12489 Set which planes to process, by default filter process all frames.
12492 Set size of local neighborhood around each pixel, used by the predictor neural
12495 Can be one of the following:
12508 Set the number of neurons in predictor neural network.
12509 Can be one of the following:
12520 Controls the number of different neural network predictions that are blended
12521 together to compute the final output value. Can be @code{fast}, default or
12525 Set which set of weights to use in the predictor.
12526 Can be one of the following:
12530 weights trained to minimize absolute error
12532 weights trained to minimize squared error
12536 Controls whether or not the prescreener neural network is used to decide
12537 which pixels should be processed by the predictor neural network and which
12538 can be handled by simple cubic interpolation.
12539 The prescreener is trained to know whether cubic interpolation will be
12540 sufficient for a pixel or whether it should be predicted by the predictor nn.
12541 The computational complexity of the prescreener nn is much less than that of
12542 the predictor nn. Since most pixels can be handled by cubic interpolation,
12543 using the prescreener generally results in much faster processing.
12544 The prescreener is pretty accurate, so the difference between using it and not
12545 using it is almost always unnoticeable.
12547 Can be one of the following:
12555 Default is @code{new}.
12558 Set various debugging flags.
12563 Force libavfilter not to use any of the specified pixel formats for the
12564 input to the next filter.
12566 It accepts the following parameters:
12570 A '|'-separated list of pixel format names, such as
12571 pix_fmts=yuv420p|monow|rgb24".
12575 @subsection Examples
12579 Force libavfilter to use a format different from @var{yuv420p} for the
12580 input to the vflip filter:
12582 noformat=pix_fmts=yuv420p,vflip
12586 Convert the input video to any of the formats not contained in the list:
12588 noformat=yuv420p|yuv444p|yuv410p
12594 Add noise on video input frame.
12596 The filter accepts the following options:
12604 Set noise seed for specific pixel component or all pixel components in case
12605 of @var{all_seed}. Default value is @code{123457}.
12607 @item all_strength, alls
12608 @item c0_strength, c0s
12609 @item c1_strength, c1s
12610 @item c2_strength, c2s
12611 @item c3_strength, c3s
12612 Set noise strength for specific pixel component or all pixel components in case
12613 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12615 @item all_flags, allf
12616 @item c0_flags, c0f
12617 @item c1_flags, c1f
12618 @item c2_flags, c2f
12619 @item c3_flags, c3f
12620 Set pixel component flags or set flags for all components if @var{all_flags}.
12621 Available values for component flags are:
12624 averaged temporal noise (smoother)
12626 mix random noise with a (semi)regular pattern
12628 temporal noise (noise pattern changes between frames)
12630 uniform noise (gaussian otherwise)
12634 @subsection Examples
12636 Add temporal and uniform noise to input video:
12638 noise=alls=20:allf=t+u
12643 Normalize RGB video (aka histogram stretching, contrast stretching).
12644 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12646 For each channel of each frame, the filter computes the input range and maps
12647 it linearly to the user-specified output range. The output range defaults
12648 to the full dynamic range from pure black to pure white.
12650 Temporal smoothing can be used on the input range to reduce flickering (rapid
12651 changes in brightness) caused when small dark or bright objects enter or leave
12652 the scene. This is similar to the auto-exposure (automatic gain control) on a
12653 video camera, and, like a video camera, it may cause a period of over- or
12654 under-exposure of the video.
12656 The R,G,B channels can be normalized independently, which may cause some
12657 color shifting, or linked together as a single channel, which prevents
12658 color shifting. Linked normalization preserves hue. Independent normalization
12659 does not, so it can be used to remove some color casts. Independent and linked
12660 normalization can be combined in any ratio.
12662 The normalize filter accepts the following options:
12667 Colors which define the output range. The minimum input value is mapped to
12668 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12669 The defaults are black and white respectively. Specifying white for
12670 @var{blackpt} and black for @var{whitept} will give color-inverted,
12671 normalized video. Shades of grey can be used to reduce the dynamic range
12672 (contrast). Specifying saturated colors here can create some interesting
12676 The number of previous frames to use for temporal smoothing. The input range
12677 of each channel is smoothed using a rolling average over the current frame
12678 and the @var{smoothing} previous frames. The default is 0 (no temporal
12682 Controls the ratio of independent (color shifting) channel normalization to
12683 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12684 independent. Defaults to 1.0 (fully independent).
12687 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12688 expensive no-op. Defaults to 1.0 (full strength).
12692 @subsection Examples
12694 Stretch video contrast to use the full dynamic range, with no temporal
12695 smoothing; may flicker depending on the source content:
12697 normalize=blackpt=black:whitept=white:smoothing=0
12700 As above, but with 50 frames of temporal smoothing; flicker should be
12701 reduced, depending on the source content:
12703 normalize=blackpt=black:whitept=white:smoothing=50
12706 As above, but with hue-preserving linked channel normalization:
12708 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12711 As above, but with half strength:
12713 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12716 Map the darkest input color to red, the brightest input color to cyan:
12718 normalize=blackpt=red:whitept=cyan
12723 Pass the video source unchanged to the output.
12726 Optical Character Recognition
12728 This filter uses Tesseract for optical character recognition. To enable
12729 compilation of this filter, you need to configure FFmpeg with
12730 @code{--enable-libtesseract}.
12732 It accepts the following options:
12736 Set datapath to tesseract data. Default is to use whatever was
12737 set at installation.
12740 Set language, default is "eng".
12743 Set character whitelist.
12746 Set character blacklist.
12749 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12753 Apply a video transform using libopencv.
12755 To enable this filter, install the libopencv library and headers and
12756 configure FFmpeg with @code{--enable-libopencv}.
12758 It accepts the following parameters:
12763 The name of the libopencv filter to apply.
12765 @item filter_params
12766 The parameters to pass to the libopencv filter. If not specified, the default
12767 values are assumed.
12771 Refer to the official libopencv documentation for more precise
12773 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12775 Several libopencv filters are supported; see the following subsections.
12780 Dilate an image by using a specific structuring element.
12781 It corresponds to the libopencv function @code{cvDilate}.
12783 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12785 @var{struct_el} represents a structuring element, and has the syntax:
12786 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12788 @var{cols} and @var{rows} represent the number of columns and rows of
12789 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12790 point, and @var{shape} the shape for the structuring element. @var{shape}
12791 must be "rect", "cross", "ellipse", or "custom".
12793 If the value for @var{shape} is "custom", it must be followed by a
12794 string of the form "=@var{filename}". The file with name
12795 @var{filename} is assumed to represent a binary image, with each
12796 printable character corresponding to a bright pixel. When a custom
12797 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12798 or columns and rows of the read file are assumed instead.
12800 The default value for @var{struct_el} is "3x3+0x0/rect".
12802 @var{nb_iterations} specifies the number of times the transform is
12803 applied to the image, and defaults to 1.
12807 # Use the default values
12810 # Dilate using a structuring element with a 5x5 cross, iterating two times
12811 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12813 # Read the shape from the file diamond.shape, iterating two times.
12814 # The file diamond.shape may contain a pattern of characters like this
12820 # The specified columns and rows are ignored
12821 # but the anchor point coordinates are not
12822 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12827 Erode an image by using a specific structuring element.
12828 It corresponds to the libopencv function @code{cvErode}.
12830 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12831 with the same syntax and semantics as the @ref{dilate} filter.
12835 Smooth the input video.
12837 The filter takes the following parameters:
12838 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12840 @var{type} is the type of smooth filter to apply, and must be one of
12841 the following values: "blur", "blur_no_scale", "median", "gaussian",
12842 or "bilateral". The default value is "gaussian".
12844 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12845 depend on the smooth type. @var{param1} and
12846 @var{param2} accept integer positive values or 0. @var{param3} and
12847 @var{param4} accept floating point values.
12849 The default value for @var{param1} is 3. The default value for the
12850 other parameters is 0.
12852 These parameters correspond to the parameters assigned to the
12853 libopencv function @code{cvSmooth}.
12855 @section oscilloscope
12857 2D Video Oscilloscope.
12859 Useful to measure spatial impulse, step responses, chroma delays, etc.
12861 It accepts the following parameters:
12865 Set scope center x position.
12868 Set scope center y position.
12871 Set scope size, relative to frame diagonal.
12874 Set scope tilt/rotation.
12880 Set trace center x position.
12883 Set trace center y position.
12886 Set trace width, relative to width of frame.
12889 Set trace height, relative to height of frame.
12892 Set which components to trace. By default it traces first three components.
12895 Draw trace grid. By default is enabled.
12898 Draw some statistics. By default is enabled.
12901 Draw scope. By default is enabled.
12904 @subsection Examples
12908 Inspect full first row of video frame.
12910 oscilloscope=x=0.5:y=0:s=1
12914 Inspect full last row of video frame.
12916 oscilloscope=x=0.5:y=1:s=1
12920 Inspect full 5th line of video frame of height 1080.
12922 oscilloscope=x=0.5:y=5/1080:s=1
12926 Inspect full last column of video frame.
12928 oscilloscope=x=1:y=0.5:s=1:t=1
12936 Overlay one video on top of another.
12938 It takes two inputs and has one output. The first input is the "main"
12939 video on which the second input is overlaid.
12941 It accepts the following parameters:
12943 A description of the accepted options follows.
12948 Set the expression for the x and y coordinates of the overlaid video
12949 on the main video. Default value is "0" for both expressions. In case
12950 the expression is invalid, it is set to a huge value (meaning that the
12951 overlay will not be displayed within the output visible area).
12954 See @ref{framesync}.
12957 Set when the expressions for @option{x}, and @option{y} are evaluated.
12959 It accepts the following values:
12962 only evaluate expressions once during the filter initialization or
12963 when a command is processed
12966 evaluate expressions for each incoming frame
12969 Default value is @samp{frame}.
12972 See @ref{framesync}.
12975 Set the format for the output video.
12977 It accepts the following values:
12980 force YUV420 output
12983 force YUV422 output
12986 force YUV444 output
12989 force packed RGB output
12992 force planar RGB output
12995 automatically pick format
12998 Default value is @samp{yuv420}.
13001 See @ref{framesync}.
13004 Set format of alpha of the overlaid video, it can be @var{straight} or
13005 @var{premultiplied}. Default is @var{straight}.
13008 The @option{x}, and @option{y} expressions can contain the following
13014 The main input width and height.
13018 The overlay input width and height.
13022 The computed values for @var{x} and @var{y}. They are evaluated for
13027 horizontal and vertical chroma subsample values of the output
13028 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
13032 the number of input frame, starting from 0
13035 the position in the file of the input frame, NAN if unknown
13038 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
13042 This filter also supports the @ref{framesync} options.
13044 Note that the @var{n}, @var{pos}, @var{t} variables are available only
13045 when evaluation is done @emph{per frame}, and will evaluate to NAN
13046 when @option{eval} is set to @samp{init}.
13048 Be aware that frames are taken from each input video in timestamp
13049 order, hence, if their initial timestamps differ, it is a good idea
13050 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
13051 have them begin in the same zero timestamp, as the example for
13052 the @var{movie} filter does.
13054 You can chain together more overlays but you should test the
13055 efficiency of such approach.
13057 @subsection Commands
13059 This filter supports the following commands:
13063 Modify the x and y of the overlay input.
13064 The command accepts the same syntax of the corresponding option.
13066 If the specified expression is not valid, it is kept at its current
13070 @subsection Examples
13074 Draw the overlay at 10 pixels from the bottom right corner of the main
13077 overlay=main_w-overlay_w-10:main_h-overlay_h-10
13080 Using named options the example above becomes:
13082 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
13086 Insert a transparent PNG logo in the bottom left corner of the input,
13087 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
13089 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
13093 Insert 2 different transparent PNG logos (second logo on bottom
13094 right corner) using the @command{ffmpeg} tool:
13096 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
13100 Add a transparent color layer on top of the main video; @code{WxH}
13101 must specify the size of the main input to the overlay filter:
13103 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
13107 Play an original video and a filtered version (here with the deshake
13108 filter) side by side using the @command{ffplay} tool:
13110 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13113 The above command is the same as:
13115 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13119 Make a sliding overlay appearing from the left to the right top part of the
13120 screen starting since time 2:
13122 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13126 Compose output by putting two input videos side to side:
13128 ffmpeg -i left.avi -i right.avi -filter_complex "
13129 nullsrc=size=200x100 [background];
13130 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13131 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13132 [background][left] overlay=shortest=1 [background+left];
13133 [background+left][right] overlay=shortest=1:x=100 [left+right]
13138 Mask 10-20 seconds of a video by applying the delogo filter to a section
13140 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13141 -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]'
13146 Chain several overlays in cascade:
13148 nullsrc=s=200x200 [bg];
13149 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13150 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13151 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13152 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13153 [in3] null, [mid2] overlay=100:100 [out0]
13160 Apply Overcomplete Wavelet denoiser.
13162 The filter accepts the following options:
13168 Larger depth values will denoise lower frequency components more, but
13169 slow down filtering.
13171 Must be an int in the range 8-16, default is @code{8}.
13173 @item luma_strength, ls
13176 Must be a double value in the range 0-1000, default is @code{1.0}.
13178 @item chroma_strength, cs
13179 Set chroma strength.
13181 Must be a double value in the range 0-1000, default is @code{1.0}.
13187 Add paddings to the input image, and place the original input at the
13188 provided @var{x}, @var{y} coordinates.
13190 It accepts the following parameters:
13195 Specify an expression for the size of the output image with the
13196 paddings added. If the value for @var{width} or @var{height} is 0, the
13197 corresponding input size is used for the output.
13199 The @var{width} expression can reference the value set by the
13200 @var{height} expression, and vice versa.
13202 The default value of @var{width} and @var{height} is 0.
13206 Specify the offsets to place the input image at within the padded area,
13207 with respect to the top/left border of the output image.
13209 The @var{x} expression can reference the value set by the @var{y}
13210 expression, and vice versa.
13212 The default value of @var{x} and @var{y} is 0.
13214 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13215 so the input image is centered on the padded area.
13218 Specify the color of the padded area. For the syntax of this option,
13219 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13220 manual,ffmpeg-utils}.
13222 The default value of @var{color} is "black".
13225 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13227 It accepts the following values:
13231 Only evaluate expressions once during the filter initialization or when
13232 a command is processed.
13235 Evaluate expressions for each incoming frame.
13239 Default value is @samp{init}.
13242 Pad to aspect instead to a resolution.
13246 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13247 options are expressions containing the following constants:
13252 The input video width and height.
13256 These are the same as @var{in_w} and @var{in_h}.
13260 The output width and height (the size of the padded area), as
13261 specified by the @var{width} and @var{height} expressions.
13265 These are the same as @var{out_w} and @var{out_h}.
13269 The x and y offsets as specified by the @var{x} and @var{y}
13270 expressions, or NAN if not yet specified.
13273 same as @var{iw} / @var{ih}
13276 input sample aspect ratio
13279 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13283 The horizontal and vertical chroma subsample values. For example for the
13284 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13287 @subsection Examples
13291 Add paddings with the color "violet" to the input video. The output video
13292 size is 640x480, and the top-left corner of the input video is placed at
13295 pad=640:480:0:40:violet
13298 The example above is equivalent to the following command:
13300 pad=width=640:height=480:x=0:y=40:color=violet
13304 Pad the input to get an output with dimensions increased by 3/2,
13305 and put the input video at the center of the padded area:
13307 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13311 Pad the input to get a squared output with size equal to the maximum
13312 value between the input width and height, and put the input video at
13313 the center of the padded area:
13315 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13319 Pad the input to get a final w/h ratio of 16:9:
13321 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13325 In case of anamorphic video, in order to set the output display aspect
13326 correctly, it is necessary to use @var{sar} in the expression,
13327 according to the relation:
13329 (ih * X / ih) * sar = output_dar
13330 X = output_dar / sar
13333 Thus the previous example needs to be modified to:
13335 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13339 Double the output size and put the input video in the bottom-right
13340 corner of the output padded area:
13342 pad="2*iw:2*ih:ow-iw:oh-ih"
13346 @anchor{palettegen}
13347 @section palettegen
13349 Generate one palette for a whole video stream.
13351 It accepts the following options:
13355 Set the maximum number of colors to quantize in the palette.
13356 Note: the palette will still contain 256 colors; the unused palette entries
13359 @item reserve_transparent
13360 Create a palette of 255 colors maximum and reserve the last one for
13361 transparency. Reserving the transparency color is useful for GIF optimization.
13362 If not set, the maximum of colors in the palette will be 256. You probably want
13363 to disable this option for a standalone image.
13366 @item transparency_color
13367 Set the color that will be used as background for transparency.
13370 Set statistics mode.
13372 It accepts the following values:
13375 Compute full frame histograms.
13377 Compute histograms only for the part that differs from previous frame. This
13378 might be relevant to give more importance to the moving part of your input if
13379 the background is static.
13381 Compute new histogram for each frame.
13384 Default value is @var{full}.
13387 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13388 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13389 color quantization of the palette. This information is also visible at
13390 @var{info} logging level.
13392 @subsection Examples
13396 Generate a representative palette of a given video using @command{ffmpeg}:
13398 ffmpeg -i input.mkv -vf palettegen palette.png
13402 @section paletteuse
13404 Use a palette to downsample an input video stream.
13406 The filter takes two inputs: one video stream and a palette. The palette must
13407 be a 256 pixels image.
13409 It accepts the following options:
13413 Select dithering mode. Available algorithms are:
13416 Ordered 8x8 bayer dithering (deterministic)
13418 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13419 Note: this dithering is sometimes considered "wrong" and is included as a
13421 @item floyd_steinberg
13422 Floyd and Steingberg dithering (error diffusion)
13424 Frankie Sierra dithering v2 (error diffusion)
13426 Frankie Sierra dithering v2 "Lite" (error diffusion)
13429 Default is @var{sierra2_4a}.
13432 When @var{bayer} dithering is selected, this option defines the scale of the
13433 pattern (how much the crosshatch pattern is visible). A low value means more
13434 visible pattern for less banding, and higher value means less visible pattern
13435 at the cost of more banding.
13437 The option must be an integer value in the range [0,5]. Default is @var{2}.
13440 If set, define the zone to process
13444 Only the changing rectangle will be reprocessed. This is similar to GIF
13445 cropping/offsetting compression mechanism. This option can be useful for speed
13446 if only a part of the image is changing, and has use cases such as limiting the
13447 scope of the error diffusal @option{dither} to the rectangle that bounds the
13448 moving scene (it leads to more deterministic output if the scene doesn't change
13449 much, and as a result less moving noise and better GIF compression).
13452 Default is @var{none}.
13455 Take new palette for each output frame.
13457 @item alpha_threshold
13458 Sets the alpha threshold for transparency. Alpha values above this threshold
13459 will be treated as completely opaque, and values below this threshold will be
13460 treated as completely transparent.
13462 The option must be an integer value in the range [0,255]. Default is @var{128}.
13465 @subsection Examples
13469 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13470 using @command{ffmpeg}:
13472 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13476 @section perspective
13478 Correct perspective of video not recorded perpendicular to the screen.
13480 A description of the accepted parameters follows.
13491 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13492 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13493 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13494 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13495 then the corners of the source will be sent to the specified coordinates.
13497 The expressions can use the following variables:
13502 the width and height of video frame.
13506 Output frame count.
13509 @item interpolation
13510 Set interpolation for perspective correction.
13512 It accepts the following values:
13518 Default value is @samp{linear}.
13521 Set interpretation of coordinate options.
13523 It accepts the following values:
13527 Send point in the source specified by the given coordinates to
13528 the corners of the destination.
13530 @item 1, destination
13532 Send the corners of the source to the point in the destination specified
13533 by the given coordinates.
13535 Default value is @samp{source}.
13539 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13541 It accepts the following values:
13544 only evaluate expressions once during the filter initialization or
13545 when a command is processed
13548 evaluate expressions for each incoming frame
13551 Default value is @samp{init}.
13556 Delay interlaced video by one field time so that the field order changes.
13558 The intended use is to fix PAL movies that have been captured with the
13559 opposite field order to the film-to-video transfer.
13561 A description of the accepted parameters follows.
13567 It accepts the following values:
13570 Capture field order top-first, transfer bottom-first.
13571 Filter will delay the bottom field.
13574 Capture field order bottom-first, transfer top-first.
13575 Filter will delay the top field.
13578 Capture and transfer with the same field order. This mode only exists
13579 for the documentation of the other options to refer to, but if you
13580 actually select it, the filter will faithfully do nothing.
13583 Capture field order determined automatically by field flags, transfer
13585 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13586 basis using field flags. If no field information is available,
13587 then this works just like @samp{u}.
13590 Capture unknown or varying, transfer opposite.
13591 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13592 analyzing the images and selecting the alternative that produces best
13593 match between the fields.
13596 Capture top-first, transfer unknown or varying.
13597 Filter selects among @samp{t} and @samp{p} using image analysis.
13600 Capture bottom-first, transfer unknown or varying.
13601 Filter selects among @samp{b} and @samp{p} using image analysis.
13604 Capture determined by field flags, transfer unknown or varying.
13605 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13606 image analysis. If no field information is available, then this works just
13607 like @samp{U}. This is the default mode.
13610 Both capture and transfer unknown or varying.
13611 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13615 @section pixdesctest
13617 Pixel format descriptor test filter, mainly useful for internal
13618 testing. The output video should be equal to the input video.
13622 format=monow, pixdesctest
13625 can be used to test the monowhite pixel format descriptor definition.
13629 Display sample values of color channels. Mainly useful for checking color
13630 and levels. Minimum supported resolution is 640x480.
13632 The filters accept the following options:
13636 Set scope X position, relative offset on X axis.
13639 Set scope Y position, relative offset on Y axis.
13648 Set window opacity. This window also holds statistics about pixel area.
13651 Set window X position, relative offset on X axis.
13654 Set window Y position, relative offset on Y axis.
13659 Enable the specified chain of postprocessing subfilters using libpostproc. This
13660 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13661 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13662 Each subfilter and some options have a short and a long name that can be used
13663 interchangeably, i.e. dr/dering are the same.
13665 The filters accept the following options:
13669 Set postprocessing subfilters string.
13672 All subfilters share common options to determine their scope:
13676 Honor the quality commands for this subfilter.
13679 Do chrominance filtering, too (default).
13682 Do luminance filtering only (no chrominance).
13685 Do chrominance filtering only (no luminance).
13688 These options can be appended after the subfilter name, separated by a '|'.
13690 Available subfilters are:
13693 @item hb/hdeblock[|difference[|flatness]]
13694 Horizontal deblocking filter
13697 Difference factor where higher values mean more deblocking (default: @code{32}).
13699 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13702 @item vb/vdeblock[|difference[|flatness]]
13703 Vertical deblocking filter
13706 Difference factor where higher values mean more deblocking (default: @code{32}).
13708 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13711 @item ha/hadeblock[|difference[|flatness]]
13712 Accurate horizontal deblocking filter
13715 Difference factor where higher values mean more deblocking (default: @code{32}).
13717 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13720 @item va/vadeblock[|difference[|flatness]]
13721 Accurate vertical deblocking filter
13724 Difference factor where higher values mean more deblocking (default: @code{32}).
13726 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13730 The horizontal and vertical deblocking filters share the difference and
13731 flatness values so you cannot set different horizontal and vertical
13735 @item h1/x1hdeblock
13736 Experimental horizontal deblocking filter
13738 @item v1/x1vdeblock
13739 Experimental vertical deblocking filter
13744 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13747 larger -> stronger filtering
13749 larger -> stronger filtering
13751 larger -> stronger filtering
13754 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13757 Stretch luminance to @code{0-255}.
13760 @item lb/linblenddeint
13761 Linear blend deinterlacing filter that deinterlaces the given block by
13762 filtering all lines with a @code{(1 2 1)} filter.
13764 @item li/linipoldeint
13765 Linear interpolating deinterlacing filter that deinterlaces the given block by
13766 linearly interpolating every second line.
13768 @item ci/cubicipoldeint
13769 Cubic interpolating deinterlacing filter deinterlaces the given block by
13770 cubically interpolating every second line.
13772 @item md/mediandeint
13773 Median deinterlacing filter that deinterlaces the given block by applying a
13774 median filter to every second line.
13776 @item fd/ffmpegdeint
13777 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13778 second line with a @code{(-1 4 2 4 -1)} filter.
13781 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13782 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13784 @item fq/forceQuant[|quantizer]
13785 Overrides the quantizer table from the input with the constant quantizer you
13793 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13796 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13799 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13802 @subsection Examples
13806 Apply horizontal and vertical deblocking, deringing and automatic
13807 brightness/contrast:
13813 Apply default filters without brightness/contrast correction:
13819 Apply default filters and temporal denoiser:
13821 pp=default/tmpnoise|1|2|3
13825 Apply deblocking on luminance only, and switch vertical deblocking on or off
13826 automatically depending on available CPU time:
13833 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13834 similar to spp = 6 with 7 point DCT, where only the center sample is
13837 The filter accepts the following options:
13841 Force a constant quantization parameter. It accepts an integer in range
13842 0 to 63. If not set, the filter will use the QP from the video stream
13846 Set thresholding mode. Available modes are:
13850 Set hard thresholding.
13852 Set soft thresholding (better de-ringing effect, but likely blurrier).
13854 Set medium thresholding (good results, default).
13858 @section premultiply
13859 Apply alpha premultiply effect to input video stream using first plane
13860 of second stream as alpha.
13862 Both streams must have same dimensions and same pixel format.
13864 The filter accepts the following option:
13868 Set which planes will be processed, unprocessed planes will be copied.
13869 By default value 0xf, all planes will be processed.
13872 Do not require 2nd input for processing, instead use alpha plane from input stream.
13876 Apply prewitt operator to input video stream.
13878 The filter accepts the following option:
13882 Set which planes will be processed, unprocessed planes will be copied.
13883 By default value 0xf, all planes will be processed.
13886 Set value which will be multiplied with filtered result.
13889 Set value which will be added to filtered result.
13892 @anchor{program_opencl}
13893 @section program_opencl
13895 Filter video using an OpenCL program.
13900 OpenCL program source file.
13903 Kernel name in program.
13906 Number of inputs to the filter. Defaults to 1.
13909 Size of output frames. Defaults to the same as the first input.
13913 The program source file must contain a kernel function with the given name,
13914 which will be run once for each plane of the output. Each run on a plane
13915 gets enqueued as a separate 2D global NDRange with one work-item for each
13916 pixel to be generated. The global ID offset for each work-item is therefore
13917 the coordinates of a pixel in the destination image.
13919 The kernel function needs to take the following arguments:
13922 Destination image, @var{__write_only image2d_t}.
13924 This image will become the output; the kernel should write all of it.
13926 Frame index, @var{unsigned int}.
13928 This is a counter starting from zero and increasing by one for each frame.
13930 Source images, @var{__read_only image2d_t}.
13932 These are the most recent images on each input. The kernel may read from
13933 them to generate the output, but they can't be written to.
13940 Copy the input to the output (output must be the same size as the input).
13942 __kernel void copy(__write_only image2d_t destination,
13943 unsigned int index,
13944 __read_only image2d_t source)
13946 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13948 int2 location = (int2)(get_global_id(0), get_global_id(1));
13950 float4 value = read_imagef(source, sampler, location);
13952 write_imagef(destination, location, value);
13957 Apply a simple transformation, rotating the input by an amount increasing
13958 with the index counter. Pixel values are linearly interpolated by the
13959 sampler, and the output need not have the same dimensions as the input.
13961 __kernel void rotate_image(__write_only image2d_t dst,
13962 unsigned int index,
13963 __read_only image2d_t src)
13965 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13966 CLK_FILTER_LINEAR);
13968 float angle = (float)index / 100.0f;
13970 float2 dst_dim = convert_float2(get_image_dim(dst));
13971 float2 src_dim = convert_float2(get_image_dim(src));
13973 float2 dst_cen = dst_dim / 2.0f;
13974 float2 src_cen = src_dim / 2.0f;
13976 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13978 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13980 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13981 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13983 src_pos = src_pos * src_dim / dst_dim;
13985 float2 src_loc = src_pos + src_cen;
13987 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13988 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13989 write_imagef(dst, dst_loc, 0.5f);
13991 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13996 Blend two inputs together, with the amount of each input used varying
13997 with the index counter.
13999 __kernel void blend_images(__write_only image2d_t dst,
14000 unsigned int index,
14001 __read_only image2d_t src1,
14002 __read_only image2d_t src2)
14004 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
14005 CLK_FILTER_LINEAR);
14007 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
14009 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
14010 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
14011 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
14013 float4 val1 = read_imagef(src1, sampler, src1_loc);
14014 float4 val2 = read_imagef(src2, sampler, src2_loc);
14016 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
14022 @section pseudocolor
14024 Alter frame colors in video with pseudocolors.
14026 This filter accept the following options:
14030 set pixel first component expression
14033 set pixel second component expression
14036 set pixel third component expression
14039 set pixel fourth component expression, corresponds to the alpha component
14042 set component to use as base for altering colors
14045 Each of them specifies the expression to use for computing the lookup table for
14046 the corresponding pixel component values.
14048 The expressions can contain the following constants and functions:
14053 The input width and height.
14056 The input value for the pixel component.
14058 @item ymin, umin, vmin, amin
14059 The minimum allowed component value.
14061 @item ymax, umax, vmax, amax
14062 The maximum allowed component value.
14065 All expressions default to "val".
14067 @subsection Examples
14071 Change too high luma values to gradient:
14073 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'"
14079 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
14080 Ratio) between two input videos.
14082 This filter takes in input two input videos, the first input is
14083 considered the "main" source and is passed unchanged to the
14084 output. The second input is used as a "reference" video for computing
14087 Both video inputs must have the same resolution and pixel format for
14088 this filter to work correctly. Also it assumes that both inputs
14089 have the same number of frames, which are compared one by one.
14091 The obtained average PSNR is printed through the logging system.
14093 The filter stores the accumulated MSE (mean squared error) of each
14094 frame, and at the end of the processing it is averaged across all frames
14095 equally, and the following formula is applied to obtain the PSNR:
14098 PSNR = 10*log10(MAX^2/MSE)
14101 Where MAX is the average of the maximum values of each component of the
14104 The description of the accepted parameters follows.
14107 @item stats_file, f
14108 If specified the filter will use the named file to save the PSNR of
14109 each individual frame. When filename equals "-" the data is sent to
14112 @item stats_version
14113 Specifies which version of the stats file format to use. Details of
14114 each format are written below.
14115 Default value is 1.
14117 @item stats_add_max
14118 Determines whether the max value is output to the stats log.
14119 Default value is 0.
14120 Requires stats_version >= 2. If this is set and stats_version < 2,
14121 the filter will return an error.
14124 This filter also supports the @ref{framesync} options.
14126 The file printed if @var{stats_file} is selected, contains a sequence of
14127 key/value pairs of the form @var{key}:@var{value} for each compared
14130 If a @var{stats_version} greater than 1 is specified, a header line precedes
14131 the list of per-frame-pair stats, with key value pairs following the frame
14132 format with the following parameters:
14135 @item psnr_log_version
14136 The version of the log file format. Will match @var{stats_version}.
14139 A comma separated list of the per-frame-pair parameters included in
14143 A description of each shown per-frame-pair parameter follows:
14147 sequential number of the input frame, starting from 1
14150 Mean Square Error pixel-by-pixel average difference of the compared
14151 frames, averaged over all the image components.
14153 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14154 Mean Square Error pixel-by-pixel average difference of the compared
14155 frames for the component specified by the suffix.
14157 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14158 Peak Signal to Noise ratio of the compared frames for the component
14159 specified by the suffix.
14161 @item max_avg, max_y, max_u, max_v
14162 Maximum allowed value for each channel, and average over all
14168 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14169 [main][ref] psnr="stats_file=stats.log" [out]
14172 On this example the input file being processed is compared with the
14173 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14174 is stored in @file{stats.log}.
14179 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14180 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14183 The pullup filter is designed to take advantage of future context in making
14184 its decisions. This filter is stateless in the sense that it does not lock
14185 onto a pattern to follow, but it instead looks forward to the following
14186 fields in order to identify matches and rebuild progressive frames.
14188 To produce content with an even framerate, insert the fps filter after
14189 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14190 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14192 The filter accepts the following options:
14199 These options set the amount of "junk" to ignore at the left, right, top, and
14200 bottom of the image, respectively. Left and right are in units of 8 pixels,
14201 while top and bottom are in units of 2 lines.
14202 The default is 8 pixels on each side.
14205 Set the strict breaks. Setting this option to 1 will reduce the chances of
14206 filter generating an occasional mismatched frame, but it may also cause an
14207 excessive number of frames to be dropped during high motion sequences.
14208 Conversely, setting it to -1 will make filter match fields more easily.
14209 This may help processing of video where there is slight blurring between
14210 the fields, but may also cause there to be interlaced frames in the output.
14211 Default value is @code{0}.
14214 Set the metric plane to use. It accepts the following values:
14220 Use chroma blue plane.
14223 Use chroma red plane.
14226 This option may be set to use chroma plane instead of the default luma plane
14227 for doing filter's computations. This may improve accuracy on very clean
14228 source material, but more likely will decrease accuracy, especially if there
14229 is chroma noise (rainbow effect) or any grayscale video.
14230 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14231 load and make pullup usable in realtime on slow machines.
14234 For best results (without duplicated frames in the output file) it is
14235 necessary to change the output frame rate. For example, to inverse
14236 telecine NTSC input:
14238 ffmpeg -i input -vf pullup -r 24000/1001 ...
14243 Change video quantization parameters (QP).
14245 The filter accepts the following option:
14249 Set expression for quantization parameter.
14252 The expression is evaluated through the eval API and can contain, among others,
14253 the following constants:
14257 1 if index is not 129, 0 otherwise.
14260 Sequential index starting from -129 to 128.
14263 @subsection Examples
14267 Some equation like:
14275 Flush video frames from internal cache of frames into a random order.
14276 No frame is discarded.
14277 Inspired by @ref{frei0r} nervous filter.
14281 Set size in number of frames of internal cache, in range from @code{2} to
14282 @code{512}. Default is @code{30}.
14285 Set seed for random number generator, must be an integer included between
14286 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14287 less than @code{0}, the filter will try to use a good random seed on a
14291 @section readeia608
14293 Read closed captioning (EIA-608) information from the top lines of a video frame.
14295 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14296 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14297 with EIA-608 data (starting from 0). A description of each metadata value follows:
14300 @item lavfi.readeia608.X.cc
14301 The two bytes stored as EIA-608 data (printed in hexadecimal).
14303 @item lavfi.readeia608.X.line
14304 The number of the line on which the EIA-608 data was identified and read.
14307 This filter accepts the following options:
14311 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14314 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14317 Set minimal acceptable amplitude change for sync codes detection.
14318 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14321 Set the ratio of width reserved for sync code detection.
14322 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14325 Set the max peaks height difference for sync code detection.
14326 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14329 Set max peaks period difference for sync code detection.
14330 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14333 Set the first two max start code bits differences.
14334 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14337 Set the minimum ratio of bits height compared to 3rd start code bit.
14338 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14341 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14344 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14347 Enable checking the parity bit. In the event of a parity error, the filter will output
14348 @code{0x00} for that character. Default is false.
14351 @subsection Examples
14355 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14357 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
14363 Read vertical interval timecode (VITC) information from the top lines of a
14366 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14367 timecode value, if a valid timecode has been detected. Further metadata key
14368 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14369 timecode data has been found or not.
14371 This filter accepts the following options:
14375 Set the maximum number of lines to scan for VITC data. If the value is set to
14376 @code{-1} the full video frame is scanned. Default is @code{45}.
14379 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14380 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14383 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14384 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14387 @subsection Examples
14391 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14392 draw @code{--:--:--:--} as a placeholder:
14394 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14400 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14402 Destination pixel at position (X, Y) will be picked from source (x, y) position
14403 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14404 value for pixel will be used for destination pixel.
14406 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14407 will have Xmap/Ymap video stream dimensions.
14408 Xmap and Ymap input video streams are 16bit depth, single channel.
14410 @section removegrain
14412 The removegrain filter is a spatial denoiser for progressive video.
14416 Set mode for the first plane.
14419 Set mode for the second plane.
14422 Set mode for the third plane.
14425 Set mode for the fourth plane.
14428 Range of mode is from 0 to 24. Description of each mode follows:
14432 Leave input plane unchanged. Default.
14435 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14438 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14441 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14444 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14445 This is equivalent to a median filter.
14448 Line-sensitive clipping giving the minimal change.
14451 Line-sensitive clipping, intermediate.
14454 Line-sensitive clipping, intermediate.
14457 Line-sensitive clipping, intermediate.
14460 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14463 Replaces the target pixel with the closest neighbour.
14466 [1 2 1] horizontal and vertical kernel blur.
14472 Bob mode, interpolates top field from the line where the neighbours
14473 pixels are the closest.
14476 Bob mode, interpolates bottom field from the line where the neighbours
14477 pixels are the closest.
14480 Bob mode, interpolates top field. Same as 13 but with a more complicated
14481 interpolation formula.
14484 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14485 interpolation formula.
14488 Clips the pixel with the minimum and maximum of respectively the maximum and
14489 minimum of each pair of opposite neighbour pixels.
14492 Line-sensitive clipping using opposite neighbours whose greatest distance from
14493 the current pixel is minimal.
14496 Replaces the pixel with the average of its 8 neighbours.
14499 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14502 Clips pixels using the averages of opposite neighbour.
14505 Same as mode 21 but simpler and faster.
14508 Small edge and halo removal, but reputed useless.
14514 @section removelogo
14516 Suppress a TV station logo, using an image file to determine which
14517 pixels comprise the logo. It works by filling in the pixels that
14518 comprise the logo with neighboring pixels.
14520 The filter accepts the following options:
14524 Set the filter bitmap file, which can be any image format supported by
14525 libavformat. The width and height of the image file must match those of the
14526 video stream being processed.
14529 Pixels in the provided bitmap image with a value of zero are not
14530 considered part of the logo, non-zero pixels are considered part of
14531 the logo. If you use white (255) for the logo and black (0) for the
14532 rest, you will be safe. For making the filter bitmap, it is
14533 recommended to take a screen capture of a black frame with the logo
14534 visible, and then using a threshold filter followed by the erode
14535 filter once or twice.
14537 If needed, little splotches can be fixed manually. Remember that if
14538 logo pixels are not covered, the filter quality will be much
14539 reduced. Marking too many pixels as part of the logo does not hurt as
14540 much, but it will increase the amount of blurring needed to cover over
14541 the image and will destroy more information than necessary, and extra
14542 pixels will slow things down on a large logo.
14544 @section repeatfields
14546 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14547 fields based on its value.
14551 Reverse a video clip.
14553 Warning: This filter requires memory to buffer the entire clip, so trimming
14556 @subsection Examples
14560 Take the first 5 seconds of a clip, and reverse it.
14567 Shift R/G/B/A pixels horizontally and/or vertically.
14569 The filter accepts the following options:
14572 Set amount to shift red horizontally.
14574 Set amount to shift red vertically.
14576 Set amount to shift green horizontally.
14578 Set amount to shift green vertically.
14580 Set amount to shift blue horizontally.
14582 Set amount to shift blue vertically.
14584 Set amount to shift alpha horizontally.
14586 Set amount to shift alpha vertically.
14588 Set edge mode, can be @var{smear}, default, or @var{warp}.
14592 Apply roberts cross operator to input video stream.
14594 The filter accepts the following option:
14598 Set which planes will be processed, unprocessed planes will be copied.
14599 By default value 0xf, all planes will be processed.
14602 Set value which will be multiplied with filtered result.
14605 Set value which will be added to filtered result.
14610 Rotate video by an arbitrary angle expressed in radians.
14612 The filter accepts the following options:
14614 A description of the optional parameters follows.
14617 Set an expression for the angle by which to rotate the input video
14618 clockwise, expressed as a number of radians. A negative value will
14619 result in a counter-clockwise rotation. By default it is set to "0".
14621 This expression is evaluated for each frame.
14624 Set the output width expression, default value is "iw".
14625 This expression is evaluated just once during configuration.
14628 Set the output height expression, default value is "ih".
14629 This expression is evaluated just once during configuration.
14632 Enable bilinear interpolation if set to 1, a value of 0 disables
14633 it. Default value is 1.
14636 Set the color used to fill the output area not covered by the rotated
14637 image. For the general syntax of this option, check the
14638 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14639 If the special value "none" is selected then no
14640 background is printed (useful for example if the background is never shown).
14642 Default value is "black".
14645 The expressions for the angle and the output size can contain the
14646 following constants and functions:
14650 sequential number of the input frame, starting from 0. It is always NAN
14651 before the first frame is filtered.
14654 time in seconds of the input frame, it is set to 0 when the filter is
14655 configured. It is always NAN before the first frame is filtered.
14659 horizontal and vertical chroma subsample values. For example for the
14660 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14664 the input video width and height
14668 the output width and height, that is the size of the padded area as
14669 specified by the @var{width} and @var{height} expressions
14673 the minimal width/height required for completely containing the input
14674 video rotated by @var{a} radians.
14676 These are only available when computing the @option{out_w} and
14677 @option{out_h} expressions.
14680 @subsection Examples
14684 Rotate the input by PI/6 radians clockwise:
14690 Rotate the input by PI/6 radians counter-clockwise:
14696 Rotate the input by 45 degrees clockwise:
14702 Apply a constant rotation with period T, starting from an angle of PI/3:
14704 rotate=PI/3+2*PI*t/T
14708 Make the input video rotation oscillating with a period of T
14709 seconds and an amplitude of A radians:
14711 rotate=A*sin(2*PI/T*t)
14715 Rotate the video, output size is chosen so that the whole rotating
14716 input video is always completely contained in the output:
14718 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14722 Rotate the video, reduce the output size so that no background is ever
14725 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14729 @subsection Commands
14731 The filter supports the following commands:
14735 Set the angle expression.
14736 The command accepts the same syntax of the corresponding option.
14738 If the specified expression is not valid, it is kept at its current
14744 Apply Shape Adaptive Blur.
14746 The filter accepts the following options:
14749 @item luma_radius, lr
14750 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14751 value is 1.0. A greater value will result in a more blurred image, and
14752 in slower processing.
14754 @item luma_pre_filter_radius, lpfr
14755 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14758 @item luma_strength, ls
14759 Set luma maximum difference between pixels to still be considered, must
14760 be a value in the 0.1-100.0 range, default value is 1.0.
14762 @item chroma_radius, cr
14763 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14764 greater value will result in a more blurred image, and in slower
14767 @item chroma_pre_filter_radius, cpfr
14768 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14770 @item chroma_strength, cs
14771 Set chroma maximum difference between pixels to still be considered,
14772 must be a value in the -0.9-100.0 range.
14775 Each chroma option value, if not explicitly specified, is set to the
14776 corresponding luma option value.
14781 Scale (resize) the input video, using the libswscale library.
14783 The scale filter forces the output display aspect ratio to be the same
14784 of the input, by changing the output sample aspect ratio.
14786 If the input image format is different from the format requested by
14787 the next filter, the scale filter will convert the input to the
14790 @subsection Options
14791 The filter accepts the following options, or any of the options
14792 supported by the libswscale scaler.
14794 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14795 the complete list of scaler options.
14800 Set the output video dimension expression. Default value is the input
14803 If the @var{width} or @var{w} value is 0, the input width is used for
14804 the output. If the @var{height} or @var{h} value is 0, the input height
14805 is used for the output.
14807 If one and only one of the values is -n with n >= 1, the scale filter
14808 will use a value that maintains the aspect ratio of the input image,
14809 calculated from the other specified dimension. After that it will,
14810 however, make sure that the calculated dimension is divisible by n and
14811 adjust the value if necessary.
14813 If both values are -n with n >= 1, the behavior will be identical to
14814 both values being set to 0 as previously detailed.
14816 See below for the list of accepted constants for use in the dimension
14820 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14824 Only evaluate expressions once during the filter initialization or when a command is processed.
14827 Evaluate expressions for each incoming frame.
14831 Default value is @samp{init}.
14835 Set the interlacing mode. It accepts the following values:
14839 Force interlaced aware scaling.
14842 Do not apply interlaced scaling.
14845 Select interlaced aware scaling depending on whether the source frames
14846 are flagged as interlaced or not.
14849 Default value is @samp{0}.
14852 Set libswscale scaling flags. See
14853 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14854 complete list of values. If not explicitly specified the filter applies
14858 @item param0, param1
14859 Set libswscale input parameters for scaling algorithms that need them. See
14860 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14861 complete documentation. If not explicitly specified the filter applies
14867 Set the video size. For the syntax of this option, check the
14868 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14870 @item in_color_matrix
14871 @item out_color_matrix
14872 Set in/output YCbCr color space type.
14874 This allows the autodetected value to be overridden as well as allows forcing
14875 a specific value used for the output and encoder.
14877 If not specified, the color space type depends on the pixel format.
14883 Choose automatically.
14886 Format conforming to International Telecommunication Union (ITU)
14887 Recommendation BT.709.
14890 Set color space conforming to the United States Federal Communications
14891 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14894 Set color space conforming to:
14898 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14901 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14904 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14909 Set color space conforming to SMPTE ST 240:1999.
14914 Set in/output YCbCr sample range.
14916 This allows the autodetected value to be overridden as well as allows forcing
14917 a specific value used for the output and encoder. If not specified, the
14918 range depends on the pixel format. Possible values:
14922 Choose automatically.
14925 Set full range (0-255 in case of 8-bit luma).
14927 @item mpeg/limited/tv
14928 Set "MPEG" range (16-235 in case of 8-bit luma).
14931 @item force_original_aspect_ratio
14932 Enable decreasing or increasing output video width or height if necessary to
14933 keep the original aspect ratio. Possible values:
14937 Scale the video as specified and disable this feature.
14940 The output video dimensions will automatically be decreased if needed.
14943 The output video dimensions will automatically be increased if needed.
14947 One useful instance of this option is that when you know a specific device's
14948 maximum allowed resolution, you can use this to limit the output video to
14949 that, while retaining the aspect ratio. For example, device A allows
14950 1280x720 playback, and your video is 1920x800. Using this option (set it to
14951 decrease) and specifying 1280x720 to the command line makes the output
14954 Please note that this is a different thing than specifying -1 for @option{w}
14955 or @option{h}, you still need to specify the output resolution for this option
14960 The values of the @option{w} and @option{h} options are expressions
14961 containing the following constants:
14966 The input width and height
14970 These are the same as @var{in_w} and @var{in_h}.
14974 The output (scaled) width and height
14978 These are the same as @var{out_w} and @var{out_h}
14981 The same as @var{iw} / @var{ih}
14984 input sample aspect ratio
14987 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14991 horizontal and vertical input chroma subsample values. For example for the
14992 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14996 horizontal and vertical output chroma subsample values. For example for the
14997 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15000 @subsection Examples
15004 Scale the input video to a size of 200x100
15009 This is equivalent to:
15020 Specify a size abbreviation for the output size:
15025 which can also be written as:
15031 Scale the input to 2x:
15033 scale=w=2*iw:h=2*ih
15037 The above is the same as:
15039 scale=2*in_w:2*in_h
15043 Scale the input to 2x with forced interlaced scaling:
15045 scale=2*iw:2*ih:interl=1
15049 Scale the input to half size:
15051 scale=w=iw/2:h=ih/2
15055 Increase the width, and set the height to the same size:
15061 Seek Greek harmony:
15068 Increase the height, and set the width to 3/2 of the height:
15070 scale=w=3/2*oh:h=3/5*ih
15074 Increase the size, making the size a multiple of the chroma
15077 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
15081 Increase the width to a maximum of 500 pixels,
15082 keeping the same aspect ratio as the input:
15084 scale=w='min(500\, iw*3/2):h=-1'
15088 Make pixels square by combining scale and setsar:
15090 scale='trunc(ih*dar):ih',setsar=1/1
15094 Make pixels square by combining scale and setsar,
15095 making sure the resulting resolution is even (required by some codecs):
15097 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
15101 @subsection Commands
15103 This filter supports the following commands:
15107 Set the output video dimension expression.
15108 The command accepts the same syntax of the corresponding option.
15110 If the specified expression is not valid, it is kept at its current
15116 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15117 format conversion on CUDA video frames. Setting the output width and height
15118 works in the same way as for the @var{scale} filter.
15120 The following additional options are accepted:
15123 The pixel format of the output CUDA frames. If set to the string "same" (the
15124 default), the input format will be kept. Note that automatic format negotiation
15125 and conversion is not yet supported for hardware frames
15128 The interpolation algorithm used for resizing. One of the following:
15135 @item cubic2p_bspline
15136 2-parameter cubic (B=1, C=0)
15138 @item cubic2p_catmullrom
15139 2-parameter cubic (B=0, C=1/2)
15141 @item cubic2p_b05c03
15142 2-parameter cubic (B=1/2, C=3/10)
15154 Scale (resize) the input video, based on a reference video.
15156 See the scale filter for available options, scale2ref supports the same but
15157 uses the reference video instead of the main input as basis. scale2ref also
15158 supports the following additional constants for the @option{w} and
15159 @option{h} options:
15164 The main input video's width and height
15167 The same as @var{main_w} / @var{main_h}
15170 The main input video's sample aspect ratio
15172 @item main_dar, mdar
15173 The main input video's display aspect ratio. Calculated from
15174 @code{(main_w / main_h) * main_sar}.
15178 The main input video's horizontal and vertical chroma subsample values.
15179 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15183 @subsection Examples
15187 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15189 'scale2ref[b][a];[a][b]overlay'
15193 @anchor{selectivecolor}
15194 @section selectivecolor
15196 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15197 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15198 by the "purity" of the color (that is, how saturated it already is).
15200 This filter is similar to the Adobe Photoshop Selective Color tool.
15202 The filter accepts the following options:
15205 @item correction_method
15206 Select color correction method.
15208 Available values are:
15211 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15214 Specified adjustments are relative to the original component value.
15216 Default is @code{absolute}.
15218 Adjustments for red pixels (pixels where the red component is the maximum)
15220 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15222 Adjustments for green pixels (pixels where the green component is the maximum)
15224 Adjustments for cyan pixels (pixels where the red component is the minimum)
15226 Adjustments for blue pixels (pixels where the blue component is the maximum)
15228 Adjustments for magenta pixels (pixels where the green component is the minimum)
15230 Adjustments for white pixels (pixels where all components are greater than 128)
15232 Adjustments for all pixels except pure black and pure white
15234 Adjustments for black pixels (pixels where all components are lesser than 128)
15236 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15239 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15240 4 space separated floating point adjustment values in the [-1,1] range,
15241 respectively to adjust the amount of cyan, magenta, yellow and black for the
15242 pixels of its range.
15244 @subsection Examples
15248 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15249 increase magenta by 27% in blue areas:
15251 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15255 Use a Photoshop selective color preset:
15257 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15261 @anchor{separatefields}
15262 @section separatefields
15264 The @code{separatefields} takes a frame-based video input and splits
15265 each frame into its components fields, producing a new half height clip
15266 with twice the frame rate and twice the frame count.
15268 This filter use field-dominance information in frame to decide which
15269 of each pair of fields to place first in the output.
15270 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15272 @section setdar, setsar
15274 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15277 This is done by changing the specified Sample (aka Pixel) Aspect
15278 Ratio, according to the following equation:
15280 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15283 Keep in mind that the @code{setdar} filter does not modify the pixel
15284 dimensions of the video frame. Also, the display aspect ratio set by
15285 this filter may be changed by later filters in the filterchain,
15286 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15289 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15290 the filter output video.
15292 Note that as a consequence of the application of this filter, the
15293 output display aspect ratio will change according to the equation
15296 Keep in mind that the sample aspect ratio set by the @code{setsar}
15297 filter may be changed by later filters in the filterchain, e.g. if
15298 another "setsar" or a "setdar" filter is applied.
15300 It accepts the following parameters:
15303 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15304 Set the aspect ratio used by the filter.
15306 The parameter can be a floating point number string, an expression, or
15307 a string of the form @var{num}:@var{den}, where @var{num} and
15308 @var{den} are the numerator and denominator of the aspect ratio. If
15309 the parameter is not specified, it is assumed the value "0".
15310 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15314 Set the maximum integer value to use for expressing numerator and
15315 denominator when reducing the expressed aspect ratio to a rational.
15316 Default value is @code{100}.
15320 The parameter @var{sar} is an expression containing
15321 the following constants:
15325 These are approximated values for the mathematical constants e
15326 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15329 The input width and height.
15332 These are the same as @var{w} / @var{h}.
15335 The input sample aspect ratio.
15338 The input display aspect ratio. It is the same as
15339 (@var{w} / @var{h}) * @var{sar}.
15342 Horizontal and vertical chroma subsample values. For example, for the
15343 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15346 @subsection Examples
15351 To change the display aspect ratio to 16:9, specify one of the following:
15358 To change the sample aspect ratio to 10:11, specify:
15364 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15365 1000 in the aspect ratio reduction, use the command:
15367 setdar=ratio=16/9:max=1000
15375 Force field for the output video frame.
15377 The @code{setfield} filter marks the interlace type field for the
15378 output frames. It does not change the input frame, but only sets the
15379 corresponding property, which affects how the frame is treated by
15380 following filters (e.g. @code{fieldorder} or @code{yadif}).
15382 The filter accepts the following options:
15387 Available values are:
15391 Keep the same field property.
15394 Mark the frame as bottom-field-first.
15397 Mark the frame as top-field-first.
15400 Mark the frame as progressive.
15407 Force frame parameter for the output video frame.
15409 The @code{setparams} filter marks interlace and color range for the
15410 output frames. It does not change the input frame, but only sets the
15411 corresponding property, which affects how the frame is treated by
15416 Available values are:
15420 Keep the same field property (default).
15423 Mark the frame as bottom-field-first.
15426 Mark the frame as top-field-first.
15429 Mark the frame as progressive.
15433 Available values are:
15437 Keep the same color range property (default).
15439 @item unspecified, unknown
15440 Mark the frame as unspecified color range.
15442 @item limited, tv, mpeg
15443 Mark the frame as limited range.
15445 @item full, pc, jpeg
15446 Mark the frame as full range.
15449 @item color_primaries
15450 Set the color primaries.
15451 Available values are:
15455 Keep the same color primaries property (default).
15472 Set the color transfer.
15473 Available values are:
15477 Keep the same color trc property (default).
15499 Set the colorspace.
15500 Available values are:
15504 Keep the same colorspace property (default).
15517 @item chroma-derived-nc
15518 @item chroma-derived-c
15525 Show a line containing various information for each input video frame.
15526 The input video is not modified.
15528 This filter supports the following options:
15532 Calculate checksums of each plane. By default enabled.
15535 The shown line contains a sequence of key/value pairs of the form
15536 @var{key}:@var{value}.
15538 The following values are shown in the output:
15542 The (sequential) number of the input frame, starting from 0.
15545 The Presentation TimeStamp of the input frame, expressed as a number of
15546 time base units. The time base unit depends on the filter input pad.
15549 The Presentation TimeStamp of the input frame, expressed as a number of
15553 The position of the frame in the input stream, or -1 if this information is
15554 unavailable and/or meaningless (for example in case of synthetic video).
15557 The pixel format name.
15560 The sample aspect ratio of the input frame, expressed in the form
15561 @var{num}/@var{den}.
15564 The size of the input frame. For the syntax of this option, check the
15565 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15568 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15569 for bottom field first).
15572 This is 1 if the frame is a key frame, 0 otherwise.
15575 The picture type of the input frame ("I" for an I-frame, "P" for a
15576 P-frame, "B" for a B-frame, or "?" for an unknown type).
15577 Also refer to the documentation of the @code{AVPictureType} enum and of
15578 the @code{av_get_picture_type_char} function defined in
15579 @file{libavutil/avutil.h}.
15582 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15584 @item plane_checksum
15585 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15586 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15589 @section showpalette
15591 Displays the 256 colors palette of each frame. This filter is only relevant for
15592 @var{pal8} pixel format frames.
15594 It accepts the following option:
15598 Set the size of the box used to represent one palette color entry. Default is
15599 @code{30} (for a @code{30x30} pixel box).
15602 @section shuffleframes
15604 Reorder and/or duplicate and/or drop video frames.
15606 It accepts the following parameters:
15610 Set the destination indexes of input frames.
15611 This is space or '|' separated list of indexes that maps input frames to output
15612 frames. Number of indexes also sets maximal value that each index may have.
15613 '-1' index have special meaning and that is to drop frame.
15616 The first frame has the index 0. The default is to keep the input unchanged.
15618 @subsection Examples
15622 Swap second and third frame of every three frames of the input:
15624 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15628 Swap 10th and 1st frame of every ten frames of the input:
15630 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15634 @section shuffleplanes
15636 Reorder and/or duplicate video planes.
15638 It accepts the following parameters:
15643 The index of the input plane to be used as the first output plane.
15646 The index of the input plane to be used as the second output plane.
15649 The index of the input plane to be used as the third output plane.
15652 The index of the input plane to be used as the fourth output plane.
15656 The first plane has the index 0. The default is to keep the input unchanged.
15658 @subsection Examples
15662 Swap the second and third planes of the input:
15664 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15668 @anchor{signalstats}
15669 @section signalstats
15670 Evaluate various visual metrics that assist in determining issues associated
15671 with the digitization of analog video media.
15673 By default the filter will log these metadata values:
15677 Display the minimal Y value contained within the input frame. Expressed in
15681 Display the Y value at the 10% percentile within the input frame. Expressed in
15685 Display the average Y value within the input frame. Expressed in range of
15689 Display the Y value at the 90% percentile within the input frame. Expressed in
15693 Display the maximum Y value contained within the input frame. Expressed in
15697 Display the minimal U value contained within the input frame. Expressed in
15701 Display the U value at the 10% percentile within the input frame. Expressed in
15705 Display the average U value within the input frame. Expressed in range of
15709 Display the U value at the 90% percentile within the input frame. Expressed in
15713 Display the maximum U value contained within the input frame. Expressed in
15717 Display the minimal V value contained within the input frame. Expressed in
15721 Display the V value at the 10% percentile within the input frame. Expressed in
15725 Display the average V value within the input frame. Expressed in range of
15729 Display the V value at the 90% percentile within the input frame. Expressed in
15733 Display the maximum V value contained within the input frame. Expressed in
15737 Display the minimal saturation value contained within the input frame.
15738 Expressed in range of [0-~181.02].
15741 Display the saturation value at the 10% percentile within the input frame.
15742 Expressed in range of [0-~181.02].
15745 Display the average saturation value within the input frame. Expressed in range
15749 Display the saturation value at the 90% percentile within the input frame.
15750 Expressed in range of [0-~181.02].
15753 Display the maximum saturation value contained within the input frame.
15754 Expressed in range of [0-~181.02].
15757 Display the median value for hue within the input frame. Expressed in range of
15761 Display the average value for hue within the input frame. Expressed in range of
15765 Display the average of sample value difference between all values of the Y
15766 plane in the current frame and corresponding values of the previous input frame.
15767 Expressed in range of [0-255].
15770 Display the average of sample value difference between all values of the U
15771 plane in the current frame and corresponding values of the previous input frame.
15772 Expressed in range of [0-255].
15775 Display the average of sample value difference between all values of the V
15776 plane in the current frame and corresponding values of the previous input frame.
15777 Expressed in range of [0-255].
15780 Display bit depth of Y plane in current frame.
15781 Expressed in range of [0-16].
15784 Display bit depth of U plane in current frame.
15785 Expressed in range of [0-16].
15788 Display bit depth of V plane in current frame.
15789 Expressed in range of [0-16].
15792 The filter accepts the following options:
15798 @option{stat} specify an additional form of image analysis.
15799 @option{out} output video with the specified type of pixel highlighted.
15801 Both options accept the following values:
15805 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15806 unlike the neighboring pixels of the same field. Examples of temporal outliers
15807 include the results of video dropouts, head clogs, or tape tracking issues.
15810 Identify @var{vertical line repetition}. Vertical line repetition includes
15811 similar rows of pixels within a frame. In born-digital video vertical line
15812 repetition is common, but this pattern is uncommon in video digitized from an
15813 analog source. When it occurs in video that results from the digitization of an
15814 analog source it can indicate concealment from a dropout compensator.
15817 Identify pixels that fall outside of legal broadcast range.
15821 Set the highlight color for the @option{out} option. The default color is
15825 @subsection Examples
15829 Output data of various video metrics:
15831 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15835 Output specific data about the minimum and maximum values of the Y plane per frame:
15837 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15841 Playback video while highlighting pixels that are outside of broadcast range in red.
15843 ffplay example.mov -vf signalstats="out=brng:color=red"
15847 Playback video with signalstats metadata drawn over the frame.
15849 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15852 The contents of signalstat_drawtext.txt used in the command are:
15855 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15856 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15857 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15858 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15866 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15867 input. In this case the matching between the inputs can be calculated additionally.
15868 The filter always passes through the first input. The signature of each stream can
15869 be written into a file.
15871 It accepts the following options:
15875 Enable or disable the matching process.
15877 Available values are:
15881 Disable the calculation of a matching (default).
15883 Calculate the matching for the whole video and output whether the whole video
15884 matches or only parts.
15886 Calculate only until a matching is found or the video ends. Should be faster in
15891 Set the number of inputs. The option value must be a non negative integer.
15892 Default value is 1.
15895 Set the path to which the output is written. If there is more than one input,
15896 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15897 integer), that will be replaced with the input number. If no filename is
15898 specified, no output will be written. This is the default.
15901 Choose the output format.
15903 Available values are:
15907 Use the specified binary representation (default).
15909 Use the specified xml representation.
15913 Set threshold to detect one word as similar. The option value must be an integer
15914 greater than zero. The default value is 9000.
15917 Set threshold to detect all words as similar. The option value must be an integer
15918 greater than zero. The default value is 60000.
15921 Set threshold to detect frames as similar. The option value must be an integer
15922 greater than zero. The default value is 116.
15925 Set the minimum length of a sequence in frames to recognize it as matching
15926 sequence. The option value must be a non negative integer value.
15927 The default value is 0.
15930 Set the minimum relation, that matching frames to all frames must have.
15931 The option value must be a double value between 0 and 1. The default value is 0.5.
15934 @subsection Examples
15938 To calculate the signature of an input video and store it in signature.bin:
15940 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15944 To detect whether two videos match and store the signatures in XML format in
15945 signature0.xml and signature1.xml:
15947 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 -
15955 Blur the input video without impacting the outlines.
15957 It accepts the following options:
15960 @item luma_radius, lr
15961 Set the luma radius. The option value must be a float number in
15962 the range [0.1,5.0] that specifies the variance of the gaussian filter
15963 used to blur the image (slower if larger). Default value is 1.0.
15965 @item luma_strength, ls
15966 Set the luma strength. The option value must be a float number
15967 in the range [-1.0,1.0] that configures the blurring. A value included
15968 in [0.0,1.0] will blur the image whereas a value included in
15969 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15971 @item luma_threshold, lt
15972 Set the luma threshold used as a coefficient to determine
15973 whether a pixel should be blurred or not. The option value must be an
15974 integer in the range [-30,30]. A value of 0 will filter all the image,
15975 a value included in [0,30] will filter flat areas and a value included
15976 in [-30,0] will filter edges. Default value is 0.
15978 @item chroma_radius, cr
15979 Set the chroma radius. The option value must be a float number in
15980 the range [0.1,5.0] that specifies the variance of the gaussian filter
15981 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15983 @item chroma_strength, cs
15984 Set the chroma strength. The option value must be a float number
15985 in the range [-1.0,1.0] that configures the blurring. A value included
15986 in [0.0,1.0] will blur the image whereas a value included in
15987 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15989 @item chroma_threshold, ct
15990 Set the chroma threshold used as a coefficient to determine
15991 whether a pixel should be blurred or not. The option value must be an
15992 integer in the range [-30,30]. A value of 0 will filter all the image,
15993 a value included in [0,30] will filter flat areas and a value included
15994 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15997 If a chroma option is not explicitly set, the corresponding luma value
16002 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
16004 This filter takes in input two input videos, the first input is
16005 considered the "main" source and is passed unchanged to the
16006 output. The second input is used as a "reference" video for computing
16009 Both video inputs must have the same resolution and pixel format for
16010 this filter to work correctly. Also it assumes that both inputs
16011 have the same number of frames, which are compared one by one.
16013 The filter stores the calculated SSIM of each frame.
16015 The description of the accepted parameters follows.
16018 @item stats_file, f
16019 If specified the filter will use the named file to save the SSIM of
16020 each individual frame. When filename equals "-" the data is sent to
16024 The file printed if @var{stats_file} is selected, contains a sequence of
16025 key/value pairs of the form @var{key}:@var{value} for each compared
16028 A description of each shown parameter follows:
16032 sequential number of the input frame, starting from 1
16034 @item Y, U, V, R, G, B
16035 SSIM of the compared frames for the component specified by the suffix.
16038 SSIM of the compared frames for the whole frame.
16041 Same as above but in dB representation.
16044 This filter also supports the @ref{framesync} options.
16048 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16049 [main][ref] ssim="stats_file=stats.log" [out]
16052 On this example the input file being processed is compared with the
16053 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
16054 is stored in @file{stats.log}.
16056 Another example with both psnr and ssim at same time:
16058 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
16063 Convert between different stereoscopic image formats.
16065 The filters accept the following options:
16069 Set stereoscopic image format of input.
16071 Available values for input image formats are:
16074 side by side parallel (left eye left, right eye right)
16077 side by side crosseye (right eye left, left eye right)
16080 side by side parallel with half width resolution
16081 (left eye left, right eye right)
16084 side by side crosseye with half width resolution
16085 (right eye left, left eye right)
16088 above-below (left eye above, right eye below)
16091 above-below (right eye above, left eye below)
16094 above-below with half height resolution
16095 (left eye above, right eye below)
16098 above-below with half height resolution
16099 (right eye above, left eye below)
16102 alternating frames (left eye first, right eye second)
16105 alternating frames (right eye first, left eye second)
16108 interleaved rows (left eye has top row, right eye starts on next row)
16111 interleaved rows (right eye has top row, left eye starts on next row)
16114 interleaved columns, left eye first
16117 interleaved columns, right eye first
16119 Default value is @samp{sbsl}.
16123 Set stereoscopic image format of output.
16127 side by side parallel (left eye left, right eye right)
16130 side by side crosseye (right eye left, left eye right)
16133 side by side parallel with half width resolution
16134 (left eye left, right eye right)
16137 side by side crosseye with half width resolution
16138 (right eye left, left eye right)
16141 above-below (left eye above, right eye below)
16144 above-below (right eye above, left eye below)
16147 above-below with half height resolution
16148 (left eye above, right eye below)
16151 above-below with half height resolution
16152 (right eye above, left eye below)
16155 alternating frames (left eye first, right eye second)
16158 alternating frames (right eye first, left eye second)
16161 interleaved rows (left eye has top row, right eye starts on next row)
16164 interleaved rows (right eye has top row, left eye starts on next row)
16167 anaglyph red/blue gray
16168 (red filter on left eye, blue filter on right eye)
16171 anaglyph red/green gray
16172 (red filter on left eye, green filter on right eye)
16175 anaglyph red/cyan gray
16176 (red filter on left eye, cyan filter on right eye)
16179 anaglyph red/cyan half colored
16180 (red filter on left eye, cyan filter on right eye)
16183 anaglyph red/cyan color
16184 (red filter on left eye, cyan filter on right eye)
16187 anaglyph red/cyan color optimized with the least squares projection of dubois
16188 (red filter on left eye, cyan filter on right eye)
16191 anaglyph green/magenta gray
16192 (green filter on left eye, magenta filter on right eye)
16195 anaglyph green/magenta half colored
16196 (green filter on left eye, magenta filter on right eye)
16199 anaglyph green/magenta colored
16200 (green filter on left eye, magenta filter on right eye)
16203 anaglyph green/magenta color optimized with the least squares projection of dubois
16204 (green filter on left eye, magenta filter on right eye)
16207 anaglyph yellow/blue gray
16208 (yellow filter on left eye, blue filter on right eye)
16211 anaglyph yellow/blue half colored
16212 (yellow filter on left eye, blue filter on right eye)
16215 anaglyph yellow/blue colored
16216 (yellow filter on left eye, blue filter on right eye)
16219 anaglyph yellow/blue color optimized with the least squares projection of dubois
16220 (yellow filter on left eye, blue filter on right eye)
16223 mono output (left eye only)
16226 mono output (right eye only)
16229 checkerboard, left eye first
16232 checkerboard, right eye first
16235 interleaved columns, left eye first
16238 interleaved columns, right eye first
16244 Default value is @samp{arcd}.
16247 @subsection Examples
16251 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16257 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16263 @section streamselect, astreamselect
16264 Select video or audio streams.
16266 The filter accepts the following options:
16270 Set number of inputs. Default is 2.
16273 Set input indexes to remap to outputs.
16276 @subsection Commands
16278 The @code{streamselect} and @code{astreamselect} filter supports the following
16283 Set input indexes to remap to outputs.
16286 @subsection Examples
16290 Select first 5 seconds 1st stream and rest of time 2nd stream:
16292 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16296 Same as above, but for audio:
16298 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16303 Apply sobel operator to input video stream.
16305 The filter accepts the following option:
16309 Set which planes will be processed, unprocessed planes will be copied.
16310 By default value 0xf, all planes will be processed.
16313 Set value which will be multiplied with filtered result.
16316 Set value which will be added to filtered result.
16322 Apply a simple postprocessing filter that compresses and decompresses the image
16323 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16324 and average the results.
16326 The filter accepts the following options:
16330 Set quality. This option defines the number of levels for averaging. It accepts
16331 an integer in the range 0-6. If set to @code{0}, the filter will have no
16332 effect. A value of @code{6} means the higher quality. For each increment of
16333 that value the speed drops by a factor of approximately 2. Default value is
16337 Force a constant quantization parameter. If not set, the filter will use the QP
16338 from the video stream (if available).
16341 Set thresholding mode. Available modes are:
16345 Set hard thresholding (default).
16347 Set soft thresholding (better de-ringing effect, but likely blurrier).
16350 @item use_bframe_qp
16351 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16352 option may cause flicker since the B-Frames have often larger QP. Default is
16353 @code{0} (not enabled).
16358 Scale the input by applying one of the super-resolution methods based on
16359 convolutional neural networks. Supported models:
16363 Super-Resolution Convolutional Neural Network model (SRCNN).
16364 See @url{https://arxiv.org/abs/1501.00092}.
16367 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16368 See @url{https://arxiv.org/abs/1609.05158}.
16371 Training scripts as well as scripts for model generation are provided in
16372 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16374 The filter accepts the following options:
16378 Specify which DNN backend to use for model loading and execution. This option accepts
16379 the following values:
16383 Native implementation of DNN loading and execution.
16386 TensorFlow backend. To enable this backend you
16387 need to install the TensorFlow for C library (see
16388 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16389 @code{--enable-libtensorflow}
16392 Default value is @samp{native}.
16395 Set path to model file specifying network architecture and its parameters.
16396 Note that different backends use different file formats. TensorFlow backend
16397 can load files for both formats, while native backend can load files for only
16401 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16402 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16403 input upscaled using bicubic upscaling with proper scale factor.
16409 Draw subtitles on top of input video using the libass library.
16411 To enable compilation of this filter you need to configure FFmpeg with
16412 @code{--enable-libass}. This filter also requires a build with libavcodec and
16413 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16414 Alpha) subtitles format.
16416 The filter accepts the following options:
16420 Set the filename of the subtitle file to read. It must be specified.
16422 @item original_size
16423 Specify the size of the original video, the video for which the ASS file
16424 was composed. For the syntax of this option, check the
16425 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16426 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16427 correctly scale the fonts if the aspect ratio has been changed.
16430 Set a directory path containing fonts that can be used by the filter.
16431 These fonts will be used in addition to whatever the font provider uses.
16434 Process alpha channel, by default alpha channel is untouched.
16437 Set subtitles input character encoding. @code{subtitles} filter only. Only
16438 useful if not UTF-8.
16440 @item stream_index, si
16441 Set subtitles stream index. @code{subtitles} filter only.
16444 Override default style or script info parameters of the subtitles. It accepts a
16445 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16448 If the first key is not specified, it is assumed that the first value
16449 specifies the @option{filename}.
16451 For example, to render the file @file{sub.srt} on top of the input
16452 video, use the command:
16457 which is equivalent to:
16459 subtitles=filename=sub.srt
16462 To render the default subtitles stream from file @file{video.mkv}, use:
16464 subtitles=video.mkv
16467 To render the second subtitles stream from that file, use:
16469 subtitles=video.mkv:si=1
16472 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16473 @code{DejaVu Serif}, use:
16475 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16478 @section super2xsai
16480 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16481 Interpolate) pixel art scaling algorithm.
16483 Useful for enlarging pixel art images without reducing sharpness.
16487 Swap two rectangular objects in video.
16489 This filter accepts the following options:
16499 Set 1st rect x coordinate.
16502 Set 1st rect y coordinate.
16505 Set 2nd rect x coordinate.
16508 Set 2nd rect y coordinate.
16510 All expressions are evaluated once for each frame.
16513 The all options are expressions containing the following constants:
16518 The input width and height.
16521 same as @var{w} / @var{h}
16524 input sample aspect ratio
16527 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16530 The number of the input frame, starting from 0.
16533 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16536 the position in the file of the input frame, NAN if unknown
16544 Apply telecine process to the video.
16546 This filter accepts the following options:
16555 The default value is @code{top}.
16559 A string of numbers representing the pulldown pattern you wish to apply.
16560 The default value is @code{23}.
16564 Some typical patterns:
16569 24p: 2332 (preferred)
16576 24p: 222222222223 ("Euro pulldown")
16583 Apply threshold effect to video stream.
16585 This filter needs four video streams to perform thresholding.
16586 First stream is stream we are filtering.
16587 Second stream is holding threshold values, third stream is holding min values,
16588 and last, fourth stream is holding max values.
16590 The filter accepts the following option:
16594 Set which planes will be processed, unprocessed planes will be copied.
16595 By default value 0xf, all planes will be processed.
16598 For example if first stream pixel's component value is less then threshold value
16599 of pixel component from 2nd threshold stream, third stream value will picked,
16600 otherwise fourth stream pixel component value will be picked.
16602 Using color source filter one can perform various types of thresholding:
16604 @subsection Examples
16608 Binary threshold, using gray color as threshold:
16610 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16614 Inverted binary threshold, using gray color as threshold:
16616 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16620 Truncate binary threshold, using gray color as threshold:
16622 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16626 Threshold to zero, using gray color as threshold:
16628 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16632 Inverted threshold to zero, using gray color as threshold:
16634 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16639 Select the most representative frame in a given sequence of consecutive frames.
16641 The filter accepts the following options:
16645 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16646 will pick one of them, and then handle the next batch of @var{n} frames until
16647 the end. Default is @code{100}.
16650 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16651 value will result in a higher memory usage, so a high value is not recommended.
16653 @subsection Examples
16657 Extract one picture each 50 frames:
16663 Complete example of a thumbnail creation with @command{ffmpeg}:
16665 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16671 Tile several successive frames together.
16673 The filter accepts the following options:
16678 Set the grid size (i.e. the number of lines and columns). For the syntax of
16679 this option, check the
16680 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16683 Set the maximum number of frames to render in the given area. It must be less
16684 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16685 the area will be used.
16688 Set the outer border margin in pixels.
16691 Set the inner border thickness (i.e. the number of pixels between frames). For
16692 more advanced padding options (such as having different values for the edges),
16693 refer to the pad video filter.
16696 Specify the color of the unused area. For the syntax of this option, check the
16697 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16698 The default value of @var{color} is "black".
16701 Set the number of frames to overlap when tiling several successive frames together.
16702 The value must be between @code{0} and @var{nb_frames - 1}.
16705 Set the number of frames to initially be empty before displaying first output frame.
16706 This controls how soon will one get first output frame.
16707 The value must be between @code{0} and @var{nb_frames - 1}.
16710 @subsection Examples
16714 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16716 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16718 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16719 duplicating each output frame to accommodate the originally detected frame
16723 Display @code{5} pictures in an area of @code{3x2} frames,
16724 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16725 mixed flat and named options:
16727 tile=3x2:nb_frames=5:padding=7:margin=2
16731 @section tinterlace
16733 Perform various types of temporal field interlacing.
16735 Frames are counted starting from 1, so the first input frame is
16738 The filter accepts the following options:
16743 Specify the mode of the interlacing. This option can also be specified
16744 as a value alone. See below for a list of values for this option.
16746 Available values are:
16750 Move odd frames into the upper field, even into the lower field,
16751 generating a double height frame at half frame rate.
16755 Frame 1 Frame 2 Frame 3 Frame 4
16757 11111 22222 33333 44444
16758 11111 22222 33333 44444
16759 11111 22222 33333 44444
16760 11111 22222 33333 44444
16774 Only output odd frames, even frames are dropped, generating a frame with
16775 unchanged height at half frame rate.
16780 Frame 1 Frame 2 Frame 3 Frame 4
16782 11111 22222 33333 44444
16783 11111 22222 33333 44444
16784 11111 22222 33333 44444
16785 11111 22222 33333 44444
16795 Only output even frames, odd frames are dropped, generating a frame with
16796 unchanged height at half frame rate.
16801 Frame 1 Frame 2 Frame 3 Frame 4
16803 11111 22222 33333 44444
16804 11111 22222 33333 44444
16805 11111 22222 33333 44444
16806 11111 22222 33333 44444
16816 Expand each frame to full height, but pad alternate lines with black,
16817 generating a frame with double height at the same input frame rate.
16822 Frame 1 Frame 2 Frame 3 Frame 4
16824 11111 22222 33333 44444
16825 11111 22222 33333 44444
16826 11111 22222 33333 44444
16827 11111 22222 33333 44444
16830 11111 ..... 33333 .....
16831 ..... 22222 ..... 44444
16832 11111 ..... 33333 .....
16833 ..... 22222 ..... 44444
16834 11111 ..... 33333 .....
16835 ..... 22222 ..... 44444
16836 11111 ..... 33333 .....
16837 ..... 22222 ..... 44444
16841 @item interleave_top, 4
16842 Interleave the upper field from odd frames with the lower field from
16843 even frames, generating a frame with unchanged height at half frame rate.
16848 Frame 1 Frame 2 Frame 3 Frame 4
16850 11111<- 22222 33333<- 44444
16851 11111 22222<- 33333 44444<-
16852 11111<- 22222 33333<- 44444
16853 11111 22222<- 33333 44444<-
16863 @item interleave_bottom, 5
16864 Interleave the lower field from odd frames with the upper field from
16865 even frames, generating a frame with unchanged height at half frame rate.
16870 Frame 1 Frame 2 Frame 3 Frame 4
16872 11111 22222<- 33333 44444<-
16873 11111<- 22222 33333<- 44444
16874 11111 22222<- 33333 44444<-
16875 11111<- 22222 33333<- 44444
16885 @item interlacex2, 6
16886 Double frame rate with unchanged height. Frames are inserted each
16887 containing the second temporal field from the previous input frame and
16888 the first temporal field from the next input frame. This mode relies on
16889 the top_field_first flag. Useful for interlaced video displays with no
16890 field synchronisation.
16895 Frame 1 Frame 2 Frame 3 Frame 4
16897 11111 22222 33333 44444
16898 11111 22222 33333 44444
16899 11111 22222 33333 44444
16900 11111 22222 33333 44444
16903 11111 22222 22222 33333 33333 44444 44444
16904 11111 11111 22222 22222 33333 33333 44444
16905 11111 22222 22222 33333 33333 44444 44444
16906 11111 11111 22222 22222 33333 33333 44444
16911 Move odd frames into the upper field, even into the lower field,
16912 generating a double height frame at same frame rate.
16917 Frame 1 Frame 2 Frame 3 Frame 4
16919 11111 22222 33333 44444
16920 11111 22222 33333 44444
16921 11111 22222 33333 44444
16922 11111 22222 33333 44444
16925 11111 33333 33333 55555
16926 22222 22222 44444 44444
16927 11111 33333 33333 55555
16928 22222 22222 44444 44444
16929 11111 33333 33333 55555
16930 22222 22222 44444 44444
16931 11111 33333 33333 55555
16932 22222 22222 44444 44444
16937 Numeric values are deprecated but are accepted for backward
16938 compatibility reasons.
16940 Default mode is @code{merge}.
16943 Specify flags influencing the filter process.
16945 Available value for @var{flags} is:
16948 @item low_pass_filter, vlfp
16949 Enable linear vertical low-pass filtering in the filter.
16950 Vertical low-pass filtering is required when creating an interlaced
16951 destination from a progressive source which contains high-frequency
16952 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16955 @item complex_filter, cvlfp
16956 Enable complex vertical low-pass filtering.
16957 This will slightly less reduce interlace 'twitter' and Moire
16958 patterning but better retain detail and subjective sharpness impression.
16962 Vertical low-pass filtering can only be enabled for @option{mode}
16963 @var{interleave_top} and @var{interleave_bottom}.
16969 Mix successive video frames.
16971 A description of the accepted options follows.
16975 The number of successive frames to mix. If unspecified, it defaults to 3.
16978 Specify weight of each input video frame.
16979 Each weight is separated by space. If number of weights is smaller than
16980 number of @var{frames} last specified weight will be used for all remaining
16984 Specify scale, if it is set it will be multiplied with sum
16985 of each weight multiplied with pixel values to give final destination
16986 pixel value. By default @var{scale} is auto scaled to sum of weights.
16989 @subsection Examples
16993 Average 7 successive frames:
16995 tmix=frames=7:weights="1 1 1 1 1 1 1"
16999 Apply simple temporal convolution:
17001 tmix=frames=3:weights="-1 3 -1"
17005 Similar as above but only showing temporal differences:
17007 tmix=frames=3:weights="-1 2 -1":scale=1
17013 Tone map colors from different dynamic ranges.
17015 This filter expects data in single precision floating point, as it needs to
17016 operate on (and can output) out-of-range values. Another filter, such as
17017 @ref{zscale}, is needed to convert the resulting frame to a usable format.
17019 The tonemapping algorithms implemented only work on linear light, so input
17020 data should be linearized beforehand (and possibly correctly tagged).
17023 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
17026 @subsection Options
17027 The filter accepts the following options.
17031 Set the tone map algorithm to use.
17033 Possible values are:
17036 Do not apply any tone map, only desaturate overbright pixels.
17039 Hard-clip any out-of-range values. Use it for perfect color accuracy for
17040 in-range values, while distorting out-of-range values.
17043 Stretch the entire reference gamut to a linear multiple of the display.
17046 Fit a logarithmic transfer between the tone curves.
17049 Preserve overall image brightness with a simple curve, using nonlinear
17050 contrast, which results in flattening details and degrading color accuracy.
17053 Preserve both dark and bright details better than @var{reinhard}, at the cost
17054 of slightly darkening everything. Use it when detail preservation is more
17055 important than color and brightness accuracy.
17058 Smoothly map out-of-range values, while retaining contrast and colors for
17059 in-range material as much as possible. Use it when color accuracy is more
17060 important than detail preservation.
17066 Tune the tone mapping algorithm.
17068 This affects the following algorithms:
17074 Specifies the scale factor to use while stretching.
17078 Specifies the exponent of the function.
17082 Specify an extra linear coefficient to multiply into the signal before clipping.
17086 Specify the local contrast coefficient at the display peak.
17087 Default to 0.5, which means that in-gamut values will be about half as bright
17094 Specify the transition point from linear to mobius transform. Every value
17095 below this point is guaranteed to be mapped 1:1. The higher the value, the
17096 more accurate the result will be, at the cost of losing bright details.
17097 Default to 0.3, which due to the steep initial slope still preserves in-range
17098 colors fairly accurately.
17102 Apply desaturation for highlights that exceed this level of brightness. The
17103 higher the parameter, the more color information will be preserved. This
17104 setting helps prevent unnaturally blown-out colors for super-highlights, by
17105 (smoothly) turning into white instead. This makes images feel more natural,
17106 at the cost of reducing information about out-of-range colors.
17108 The default of 2.0 is somewhat conservative and will mostly just apply to
17109 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17111 This option works only if the input frame has a supported color tag.
17114 Override signal/nominal/reference peak with this value. Useful when the
17115 embedded peak information in display metadata is not reliable or when tone
17116 mapping from a lower range to a higher range.
17121 Temporarily pad video frames.
17123 The filter accepts the following options:
17127 Specify number of delay frames before input video stream.
17130 Specify number of padding frames after input video stream.
17131 Set to -1 to pad indefinitely.
17134 Set kind of frames added to beginning of stream.
17135 Can be either @var{add} or @var{clone}.
17136 With @var{add} frames of solid-color are added.
17137 With @var{clone} frames are clones of first frame.
17140 Set kind of frames added to end of stream.
17141 Can be either @var{add} or @var{clone}.
17142 With @var{add} frames of solid-color are added.
17143 With @var{clone} frames are clones of last frame.
17145 @item start_duration, stop_duration
17146 Specify the duration of the start/stop delay. See
17147 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17148 for the accepted syntax.
17149 These options override @var{start} and @var{stop}.
17152 Specify the color of the padded area. For the syntax of this option,
17153 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17154 manual,ffmpeg-utils}.
17156 The default value of @var{color} is "black".
17162 Transpose rows with columns in the input video and optionally flip it.
17164 It accepts the following parameters:
17169 Specify the transposition direction.
17171 Can assume the following values:
17173 @item 0, 4, cclock_flip
17174 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17182 Rotate by 90 degrees clockwise, that is:
17190 Rotate by 90 degrees counterclockwise, that is:
17197 @item 3, 7, clock_flip
17198 Rotate by 90 degrees clockwise and vertically flip, that is:
17206 For values between 4-7, the transposition is only done if the input
17207 video geometry is portrait and not landscape. These values are
17208 deprecated, the @code{passthrough} option should be used instead.
17210 Numerical values are deprecated, and should be dropped in favor of
17211 symbolic constants.
17214 Do not apply the transposition if the input geometry matches the one
17215 specified by the specified value. It accepts the following values:
17218 Always apply transposition.
17220 Preserve portrait geometry (when @var{height} >= @var{width}).
17222 Preserve landscape geometry (when @var{width} >= @var{height}).
17225 Default value is @code{none}.
17228 For example to rotate by 90 degrees clockwise and preserve portrait
17231 transpose=dir=1:passthrough=portrait
17234 The command above can also be specified as:
17236 transpose=1:portrait
17239 @section transpose_npp
17241 Transpose rows with columns in the input video and optionally flip it.
17242 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17244 It accepts the following parameters:
17249 Specify the transposition direction.
17251 Can assume the following values:
17254 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17257 Rotate by 90 degrees clockwise.
17260 Rotate by 90 degrees counterclockwise.
17263 Rotate by 90 degrees clockwise and vertically flip.
17267 Do not apply the transposition if the input geometry matches the one
17268 specified by the specified value. It accepts the following values:
17271 Always apply transposition. (default)
17273 Preserve portrait geometry (when @var{height} >= @var{width}).
17275 Preserve landscape geometry (when @var{width} >= @var{height}).
17281 Trim the input so that the output contains one continuous subpart of the input.
17283 It accepts the following parameters:
17286 Specify the time of the start of the kept section, i.e. the frame with the
17287 timestamp @var{start} will be the first frame in the output.
17290 Specify the time of the first frame that will be dropped, i.e. the frame
17291 immediately preceding the one with the timestamp @var{end} will be the last
17292 frame in the output.
17295 This is the same as @var{start}, except this option sets the start timestamp
17296 in timebase units instead of seconds.
17299 This is the same as @var{end}, except this option sets the end timestamp
17300 in timebase units instead of seconds.
17303 The maximum duration of the output in seconds.
17306 The number of the first frame that should be passed to the output.
17309 The number of the first frame that should be dropped.
17312 @option{start}, @option{end}, and @option{duration} are expressed as time
17313 duration specifications; see
17314 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17315 for the accepted syntax.
17317 Note that the first two sets of the start/end options and the @option{duration}
17318 option look at the frame timestamp, while the _frame variants simply count the
17319 frames that pass through the filter. Also note that this filter does not modify
17320 the timestamps. If you wish for the output timestamps to start at zero, insert a
17321 setpts filter after the trim filter.
17323 If multiple start or end options are set, this filter tries to be greedy and
17324 keep all the frames that match at least one of the specified constraints. To keep
17325 only the part that matches all the constraints at once, chain multiple trim
17328 The defaults are such that all the input is kept. So it is possible to set e.g.
17329 just the end values to keep everything before the specified time.
17334 Drop everything except the second minute of input:
17336 ffmpeg -i INPUT -vf trim=60:120
17340 Keep only the first second:
17342 ffmpeg -i INPUT -vf trim=duration=1
17347 @section unpremultiply
17348 Apply alpha unpremultiply effect to input video stream using first plane
17349 of second stream as alpha.
17351 Both streams must have same dimensions and same pixel format.
17353 The filter accepts the following option:
17357 Set which planes will be processed, unprocessed planes will be copied.
17358 By default value 0xf, all planes will be processed.
17360 If the format has 1 or 2 components, then luma is bit 0.
17361 If the format has 3 or 4 components:
17362 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17363 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17364 If present, the alpha channel is always the last bit.
17367 Do not require 2nd input for processing, instead use alpha plane from input stream.
17373 Sharpen or blur the input video.
17375 It accepts the following parameters:
17378 @item luma_msize_x, lx
17379 Set the luma matrix horizontal size. It must be an odd integer between
17380 3 and 23. The default value is 5.
17382 @item luma_msize_y, ly
17383 Set the luma matrix vertical size. It must be an odd integer between 3
17384 and 23. The default value is 5.
17386 @item luma_amount, la
17387 Set the luma effect strength. It must be a floating point number, reasonable
17388 values lay between -1.5 and 1.5.
17390 Negative values will blur the input video, while positive values will
17391 sharpen it, a value of zero will disable the effect.
17393 Default value is 1.0.
17395 @item chroma_msize_x, cx
17396 Set the chroma matrix horizontal size. It must be an odd integer
17397 between 3 and 23. The default value is 5.
17399 @item chroma_msize_y, cy
17400 Set the chroma matrix vertical size. It must be an odd integer
17401 between 3 and 23. The default value is 5.
17403 @item chroma_amount, ca
17404 Set the chroma effect strength. It must be a floating point number, reasonable
17405 values lay between -1.5 and 1.5.
17407 Negative values will blur the input video, while positive values will
17408 sharpen it, a value of zero will disable the effect.
17410 Default value is 0.0.
17414 All parameters are optional and default to the equivalent of the
17415 string '5:5:1.0:5:5:0.0'.
17417 @subsection Examples
17421 Apply strong luma sharpen effect:
17423 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17427 Apply a strong blur of both luma and chroma parameters:
17429 unsharp=7:7:-2:7:7:-2
17435 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17436 the image at several (or - in the case of @option{quality} level @code{8} - all)
17437 shifts and average the results.
17439 The way this differs from the behavior of spp is that uspp actually encodes &
17440 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17441 DCT similar to MJPEG.
17443 The filter accepts the following options:
17447 Set quality. This option defines the number of levels for averaging. It accepts
17448 an integer in the range 0-8. If set to @code{0}, the filter will have no
17449 effect. A value of @code{8} means the higher quality. For each increment of
17450 that value the speed drops by a factor of approximately 2. Default value is
17454 Force a constant quantization parameter. If not set, the filter will use the QP
17455 from the video stream (if available).
17458 @section vaguedenoiser
17460 Apply a wavelet based denoiser.
17462 It transforms each frame from the video input into the wavelet domain,
17463 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17464 the obtained coefficients. It does an inverse wavelet transform after.
17465 Due to wavelet properties, it should give a nice smoothed result, and
17466 reduced noise, without blurring picture features.
17468 This filter accepts the following options:
17472 The filtering strength. The higher, the more filtered the video will be.
17473 Hard thresholding can use a higher threshold than soft thresholding
17474 before the video looks overfiltered. Default value is 2.
17477 The filtering method the filter will use.
17479 It accepts the following values:
17482 All values under the threshold will be zeroed.
17485 All values under the threshold will be zeroed. All values above will be
17486 reduced by the threshold.
17489 Scales or nullifies coefficients - intermediary between (more) soft and
17490 (less) hard thresholding.
17493 Default is garrote.
17496 Number of times, the wavelet will decompose the picture. Picture can't
17497 be decomposed beyond a particular point (typically, 8 for a 640x480
17498 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17501 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17504 A list of the planes to process. By default all planes are processed.
17507 @section vectorscope
17509 Display 2 color component values in the two dimensional graph (which is called
17512 This filter accepts the following options:
17516 Set vectorscope mode.
17518 It accepts the following values:
17521 Gray values are displayed on graph, higher brightness means more pixels have
17522 same component color value on location in graph. This is the default mode.
17525 Gray values are displayed on graph. Surrounding pixels values which are not
17526 present in video frame are drawn in gradient of 2 color components which are
17527 set by option @code{x} and @code{y}. The 3rd color component is static.
17530 Actual color components values present in video frame are displayed on graph.
17533 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17534 on graph increases value of another color component, which is luminance by
17535 default values of @code{x} and @code{y}.
17538 Actual colors present in video frame are displayed on graph. If two different
17539 colors map to same position on graph then color with higher value of component
17540 not present in graph is picked.
17543 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17544 component picked from radial gradient.
17548 Set which color component will be represented on X-axis. Default is @code{1}.
17551 Set which color component will be represented on Y-axis. Default is @code{2}.
17554 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17555 of color component which represents frequency of (X, Y) location in graph.
17560 No envelope, this is default.
17563 Instant envelope, even darkest single pixel will be clearly highlighted.
17566 Hold maximum and minimum values presented in graph over time. This way you
17567 can still spot out of range values without constantly looking at vectorscope.
17570 Peak and instant envelope combined together.
17574 Set what kind of graticule to draw.
17582 Set graticule opacity.
17585 Set graticule flags.
17589 Draw graticule for white point.
17592 Draw graticule for black point.
17595 Draw color points short names.
17599 Set background opacity.
17601 @item lthreshold, l
17602 Set low threshold for color component not represented on X or Y axis.
17603 Values lower than this value will be ignored. Default is 0.
17604 Note this value is multiplied with actual max possible value one pixel component
17605 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17608 @item hthreshold, h
17609 Set high threshold for color component not represented on X or Y axis.
17610 Values higher than this value will be ignored. Default is 1.
17611 Note this value is multiplied with actual max possible value one pixel component
17612 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17613 is 0.9 * 255 = 230.
17615 @item colorspace, c
17616 Set what kind of colorspace to use when drawing graticule.
17625 @anchor{vidstabdetect}
17626 @section vidstabdetect
17628 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17629 @ref{vidstabtransform} for pass 2.
17631 This filter generates a file with relative translation and rotation
17632 transform information about subsequent frames, which is then used by
17633 the @ref{vidstabtransform} filter.
17635 To enable compilation of this filter you need to configure FFmpeg with
17636 @code{--enable-libvidstab}.
17638 This filter accepts the following options:
17642 Set the path to the file used to write the transforms information.
17643 Default value is @file{transforms.trf}.
17646 Set how shaky the video is and how quick the camera is. It accepts an
17647 integer in the range 1-10, a value of 1 means little shakiness, a
17648 value of 10 means strong shakiness. Default value is 5.
17651 Set the accuracy of the detection process. It must be a value in the
17652 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17653 accuracy. Default value is 15.
17656 Set stepsize of the search process. The region around minimum is
17657 scanned with 1 pixel resolution. Default value is 6.
17660 Set minimum contrast. Below this value a local measurement field is
17661 discarded. Must be a floating point value in the range 0-1. Default
17665 Set reference frame number for tripod mode.
17667 If enabled, the motion of the frames is compared to a reference frame
17668 in the filtered stream, identified by the specified number. The idea
17669 is to compensate all movements in a more-or-less static scene and keep
17670 the camera view absolutely still.
17672 If set to 0, it is disabled. The frames are counted starting from 1.
17675 Show fields and transforms in the resulting frames. It accepts an
17676 integer in the range 0-2. Default value is 0, which disables any
17680 @subsection Examples
17684 Use default values:
17690 Analyze strongly shaky movie and put the results in file
17691 @file{mytransforms.trf}:
17693 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17697 Visualize the result of internal transformations in the resulting
17700 vidstabdetect=show=1
17704 Analyze a video with medium shakiness using @command{ffmpeg}:
17706 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17710 @anchor{vidstabtransform}
17711 @section vidstabtransform
17713 Video stabilization/deshaking: pass 2 of 2,
17714 see @ref{vidstabdetect} for pass 1.
17716 Read a file with transform information for each frame and
17717 apply/compensate them. Together with the @ref{vidstabdetect}
17718 filter this can be used to deshake videos. See also
17719 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17720 the @ref{unsharp} filter, see below.
17722 To enable compilation of this filter you need to configure FFmpeg with
17723 @code{--enable-libvidstab}.
17725 @subsection Options
17729 Set path to the file used to read the transforms. Default value is
17730 @file{transforms.trf}.
17733 Set the number of frames (value*2 + 1) used for lowpass filtering the
17734 camera movements. Default value is 10.
17736 For example a number of 10 means that 21 frames are used (10 in the
17737 past and 10 in the future) to smoothen the motion in the video. A
17738 larger value leads to a smoother video, but limits the acceleration of
17739 the camera (pan/tilt movements). 0 is a special case where a static
17740 camera is simulated.
17743 Set the camera path optimization algorithm.
17745 Accepted values are:
17748 gaussian kernel low-pass filter on camera motion (default)
17750 averaging on transformations
17754 Set maximal number of pixels to translate frames. Default value is -1,
17758 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17759 value is -1, meaning no limit.
17762 Specify how to deal with borders that may be visible due to movement
17765 Available values are:
17768 keep image information from previous frame (default)
17770 fill the border black
17774 Invert transforms if set to 1. Default value is 0.
17777 Consider transforms as relative to previous frame if set to 1,
17778 absolute if set to 0. Default value is 0.
17781 Set percentage to zoom. A positive value will result in a zoom-in
17782 effect, a negative value in a zoom-out effect. Default value is 0 (no
17786 Set optimal zooming to avoid borders.
17788 Accepted values are:
17793 optimal static zoom value is determined (only very strong movements
17794 will lead to visible borders) (default)
17796 optimal adaptive zoom value is determined (no borders will be
17797 visible), see @option{zoomspeed}
17800 Note that the value given at zoom is added to the one calculated here.
17803 Set percent to zoom maximally each frame (enabled when
17804 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17808 Specify type of interpolation.
17810 Available values are:
17815 linear only horizontal
17817 linear in both directions (default)
17819 cubic in both directions (slow)
17823 Enable virtual tripod mode if set to 1, which is equivalent to
17824 @code{relative=0:smoothing=0}. Default value is 0.
17826 Use also @code{tripod} option of @ref{vidstabdetect}.
17829 Increase log verbosity if set to 1. Also the detected global motions
17830 are written to the temporary file @file{global_motions.trf}. Default
17834 @subsection Examples
17838 Use @command{ffmpeg} for a typical stabilization with default values:
17840 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17843 Note the use of the @ref{unsharp} filter which is always recommended.
17846 Zoom in a bit more and load transform data from a given file:
17848 vidstabtransform=zoom=5:input="mytransforms.trf"
17852 Smoothen the video even more:
17854 vidstabtransform=smoothing=30
17860 Flip the input video vertically.
17862 For example, to vertically flip a video with @command{ffmpeg}:
17864 ffmpeg -i in.avi -vf "vflip" out.avi
17869 Detect variable frame rate video.
17871 This filter tries to detect if the input is variable or constant frame rate.
17873 At end it will output number of frames detected as having variable delta pts,
17874 and ones with constant delta pts.
17875 If there was frames with variable delta, than it will also show min and max delta
17880 Boost or alter saturation.
17882 The filter accepts the following options:
17885 Set strength of boost if positive value or strength of alter if negative value.
17886 Default is 0. Allowed range is from -2 to 2.
17889 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17892 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17895 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17898 Set the red luma coefficient.
17901 Set the green luma coefficient.
17904 Set the blue luma coefficient.
17910 Make or reverse a natural vignetting effect.
17912 The filter accepts the following options:
17916 Set lens angle expression as a number of radians.
17918 The value is clipped in the @code{[0,PI/2]} range.
17920 Default value: @code{"PI/5"}
17924 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17928 Set forward/backward mode.
17930 Available modes are:
17933 The larger the distance from the central point, the darker the image becomes.
17936 The larger the distance from the central point, the brighter the image becomes.
17937 This can be used to reverse a vignette effect, though there is no automatic
17938 detection to extract the lens @option{angle} and other settings (yet). It can
17939 also be used to create a burning effect.
17942 Default value is @samp{forward}.
17945 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17947 It accepts the following values:
17950 Evaluate expressions only once during the filter initialization.
17953 Evaluate expressions for each incoming frame. This is way slower than the
17954 @samp{init} mode since it requires all the scalers to be re-computed, but it
17955 allows advanced dynamic expressions.
17958 Default value is @samp{init}.
17961 Set dithering to reduce the circular banding effects. Default is @code{1}
17965 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17966 Setting this value to the SAR of the input will make a rectangular vignetting
17967 following the dimensions of the video.
17969 Default is @code{1/1}.
17972 @subsection Expressions
17974 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17975 following parameters.
17980 input width and height
17983 the number of input frame, starting from 0
17986 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17987 @var{TB} units, NAN if undefined
17990 frame rate of the input video, NAN if the input frame rate is unknown
17993 the PTS (Presentation TimeStamp) of the filtered video frame,
17994 expressed in seconds, NAN if undefined
17997 time base of the input video
18001 @subsection Examples
18005 Apply simple strong vignetting effect:
18011 Make a flickering vignetting:
18013 vignette='PI/4+random(1)*PI/50':eval=frame
18018 @section vmafmotion
18020 Obtain the average vmaf motion score of a video.
18021 It is one of the component filters of VMAF.
18023 The obtained average motion score is printed through the logging system.
18025 In the below example the input file @file{ref.mpg} is being processed and score
18029 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
18033 Stack input videos vertically.
18035 All streams must be of same pixel format and of same width.
18037 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
18038 to create same output.
18040 The filter accept the following option:
18044 Set number of input streams. Default is 2.
18047 If set to 1, force the output to terminate when the shortest input
18048 terminates. Default value is 0.
18053 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
18054 Deinterlacing Filter").
18056 Based on the process described by Martin Weston for BBC R&D, and
18057 implemented based on the de-interlace algorithm written by Jim
18058 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
18059 uses filter coefficients calculated by BBC R&D.
18061 There are two sets of filter coefficients, so called "simple":
18062 and "complex". Which set of filter coefficients is used can
18063 be set by passing an optional parameter:
18067 Set the interlacing filter coefficients. Accepts one of the following values:
18071 Simple filter coefficient set.
18073 More-complex filter coefficient set.
18075 Default value is @samp{complex}.
18078 Specify which frames to deinterlace. Accept one of the following values:
18082 Deinterlace all frames,
18084 Only deinterlace frames marked as interlaced.
18087 Default value is @samp{all}.
18091 Video waveform monitor.
18093 The waveform monitor plots color component intensity. By default luminance
18094 only. Each column of the waveform corresponds to a column of pixels in the
18097 It accepts the following options:
18101 Can be either @code{row}, or @code{column}. Default is @code{column}.
18102 In row mode, the graph on the left side represents color component value 0 and
18103 the right side represents value = 255. In column mode, the top side represents
18104 color component value = 0 and bottom side represents value = 255.
18107 Set intensity. Smaller values are useful to find out how many values of the same
18108 luminance are distributed across input rows/columns.
18109 Default value is @code{0.04}. Allowed range is [0, 1].
18112 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18113 In mirrored mode, higher values will be represented on the left
18114 side for @code{row} mode and at the top for @code{column} mode. Default is
18115 @code{1} (mirrored).
18119 It accepts the following values:
18122 Presents information identical to that in the @code{parade}, except
18123 that the graphs representing color components are superimposed directly
18126 This display mode makes it easier to spot relative differences or similarities
18127 in overlapping areas of the color components that are supposed to be identical,
18128 such as neutral whites, grays, or blacks.
18131 Display separate graph for the color components side by side in
18132 @code{row} mode or one below the other in @code{column} mode.
18135 Display separate graph for the color components side by side in
18136 @code{column} mode or one below the other in @code{row} mode.
18138 Using this display mode makes it easy to spot color casts in the highlights
18139 and shadows of an image, by comparing the contours of the top and the bottom
18140 graphs of each waveform. Since whites, grays, and blacks are characterized
18141 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18142 should display three waveforms of roughly equal width/height. If not, the
18143 correction is easy to perform by making level adjustments the three waveforms.
18145 Default is @code{stack}.
18147 @item components, c
18148 Set which color components to display. Default is 1, which means only luminance
18149 or red color component if input is in RGB colorspace. If is set for example to
18150 7 it will display all 3 (if) available color components.
18155 No envelope, this is default.
18158 Instant envelope, minimum and maximum values presented in graph will be easily
18159 visible even with small @code{step} value.
18162 Hold minimum and maximum values presented in graph across time. This way you
18163 can still spot out of range values without constantly looking at waveforms.
18166 Peak and instant envelope combined together.
18172 No filtering, this is default.
18175 Luma and chroma combined together.
18178 Similar as above, but shows difference between blue and red chroma.
18181 Similar as above, but use different colors.
18184 Displays only chroma.
18187 Displays actual color value on waveform.
18190 Similar as above, but with luma showing frequency of chroma values.
18194 Set which graticule to display.
18198 Do not display graticule.
18201 Display green graticule showing legal broadcast ranges.
18204 Display orange graticule showing legal broadcast ranges.
18208 Set graticule opacity.
18211 Set graticule flags.
18215 Draw numbers above lines. By default enabled.
18218 Draw dots instead of lines.
18222 Set scale used for displaying graticule.
18229 Default is digital.
18232 Set background opacity.
18235 @section weave, doubleweave
18237 The @code{weave} takes a field-based video input and join
18238 each two sequential fields into single frame, producing a new double
18239 height clip with half the frame rate and half the frame count.
18241 The @code{doubleweave} works same as @code{weave} but without
18242 halving frame rate and frame count.
18244 It accepts the following option:
18248 Set first field. Available values are:
18252 Set the frame as top-field-first.
18255 Set the frame as bottom-field-first.
18259 @subsection Examples
18263 Interlace video using @ref{select} and @ref{separatefields} filter:
18265 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18270 Apply the xBR high-quality magnification filter which is designed for pixel
18271 art. It follows a set of edge-detection rules, see
18272 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18274 It accepts the following option:
18278 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18279 @code{3xBR} and @code{4} for @code{4xBR}.
18280 Default is @code{3}.
18284 Stack video inputs into custom layout.
18286 All streams must be of same pixel format.
18288 The filter accept the following option:
18292 Set number of input streams. Default is 2.
18295 Specify layout of inputs.
18296 This option requires the desired layout configuration to be explicitly set by the user.
18297 This sets position of each video input in output. Each input
18298 is separated by '|'.
18299 The first number represents the column, and the second number represents the row.
18300 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18301 where X is video input from which to take width or height.
18302 Multiple values can be used when separated by '+'. In such
18303 case values are summed together.
18306 If set to 1, force the output to terminate when the shortest input
18307 terminates. Default value is 0.
18310 @subsection Examples
18314 Display 4 inputs into 2x2 grid,
18315 note that if inputs are of different sizes unused gaps might appear,
18316 as not all of output video is used.
18318 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18322 Display 4 inputs into 1x4 grid,
18323 note that if inputs are of different sizes unused gaps might appear,
18324 as not all of output video is used.
18326 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18330 Display 9 inputs into 3x3 grid,
18331 note that if inputs are of different sizes unused gaps might appear,
18332 as not all of output video is used.
18334 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
18341 Deinterlace the input video ("yadif" means "yet another deinterlacing
18344 It accepts the following parameters:
18350 The interlacing mode to adopt. It accepts one of the following values:
18353 @item 0, send_frame
18354 Output one frame for each frame.
18355 @item 1, send_field
18356 Output one frame for each field.
18357 @item 2, send_frame_nospatial
18358 Like @code{send_frame}, but it skips the spatial interlacing check.
18359 @item 3, send_field_nospatial
18360 Like @code{send_field}, but it skips the spatial interlacing check.
18363 The default value is @code{send_frame}.
18366 The picture field parity assumed for the input interlaced video. It accepts one
18367 of the following values:
18371 Assume the top field is first.
18373 Assume the bottom field is first.
18375 Enable automatic detection of field parity.
18378 The default value is @code{auto}.
18379 If the interlacing is unknown or the decoder does not export this information,
18380 top field first will be assumed.
18383 Specify which frames to deinterlace. Accept one of the following
18388 Deinterlace all frames.
18389 @item 1, interlaced
18390 Only deinterlace frames marked as interlaced.
18393 The default value is @code{all}.
18396 @section yadif_cuda
18398 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18399 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18402 It accepts the following parameters:
18408 The interlacing mode to adopt. It accepts one of the following values:
18411 @item 0, send_frame
18412 Output one frame for each frame.
18413 @item 1, send_field
18414 Output one frame for each field.
18415 @item 2, send_frame_nospatial
18416 Like @code{send_frame}, but it skips the spatial interlacing check.
18417 @item 3, send_field_nospatial
18418 Like @code{send_field}, but it skips the spatial interlacing check.
18421 The default value is @code{send_frame}.
18424 The picture field parity assumed for the input interlaced video. It accepts one
18425 of the following values:
18429 Assume the top field is first.
18431 Assume the bottom field is first.
18433 Enable automatic detection of field parity.
18436 The default value is @code{auto}.
18437 If the interlacing is unknown or the decoder does not export this information,
18438 top field first will be assumed.
18441 Specify which frames to deinterlace. Accept one of the following
18446 Deinterlace all frames.
18447 @item 1, interlaced
18448 Only deinterlace frames marked as interlaced.
18451 The default value is @code{all}.
18456 Apply Zoom & Pan effect.
18458 This filter accepts the following options:
18462 Set the zoom expression. Range is 1-10. Default is 1.
18466 Set the x and y expression. Default is 0.
18469 Set the duration expression in number of frames.
18470 This sets for how many number of frames effect will last for
18471 single input image.
18474 Set the output image size, default is 'hd720'.
18477 Set the output frame rate, default is '25'.
18480 Each expression can contain the following constants:
18499 Output frame count.
18503 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18504 for current input frame.
18508 'x' and 'y' of last output frame of previous input frame or 0 when there was
18509 not yet such frame (first input frame).
18512 Last calculated zoom from 'z' expression for current input frame.
18515 Last calculated zoom of last output frame of previous input frame.
18518 Number of output frames for current input frame. Calculated from 'd' expression
18519 for each input frame.
18522 number of output frames created for previous input frame
18525 Rational number: input width / input height
18528 sample aspect ratio
18531 display aspect ratio
18535 @subsection Examples
18539 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18541 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
18545 Zoom-in up to 1.5 and pan always at center of picture:
18547 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18551 Same as above but without pausing:
18553 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18559 Scale (resize) the input video, using the z.lib library:
18560 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18561 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18563 The zscale filter forces the output display aspect ratio to be the same
18564 as the input, by changing the output sample aspect ratio.
18566 If the input image format is different from the format requested by
18567 the next filter, the zscale filter will convert the input to the
18570 @subsection Options
18571 The filter accepts the following options.
18576 Set the output video dimension expression. Default value is the input
18579 If the @var{width} or @var{w} value is 0, the input width is used for
18580 the output. If the @var{height} or @var{h} value is 0, the input height
18581 is used for the output.
18583 If one and only one of the values is -n with n >= 1, the zscale filter
18584 will use a value that maintains the aspect ratio of the input image,
18585 calculated from the other specified dimension. After that it will,
18586 however, make sure that the calculated dimension is divisible by n and
18587 adjust the value if necessary.
18589 If both values are -n with n >= 1, the behavior will be identical to
18590 both values being set to 0 as previously detailed.
18592 See below for the list of accepted constants for use in the dimension
18596 Set the video size. For the syntax of this option, check the
18597 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18600 Set the dither type.
18602 Possible values are:
18607 @item error_diffusion
18613 Set the resize filter type.
18615 Possible values are:
18625 Default is bilinear.
18628 Set the color range.
18630 Possible values are:
18637 Default is same as input.
18640 Set the color primaries.
18642 Possible values are:
18652 Default is same as input.
18655 Set the transfer characteristics.
18657 Possible values are:
18671 Default is same as input.
18674 Set the colorspace matrix.
18676 Possible value are:
18687 Default is same as input.
18690 Set the input color range.
18692 Possible values are:
18699 Default is same as input.
18701 @item primariesin, pin
18702 Set the input color primaries.
18704 Possible values are:
18714 Default is same as input.
18716 @item transferin, tin
18717 Set the input transfer characteristics.
18719 Possible values are:
18730 Default is same as input.
18732 @item matrixin, min
18733 Set the input colorspace matrix.
18735 Possible value are:
18747 Set the output chroma location.
18749 Possible values are:
18760 @item chromalin, cin
18761 Set the input chroma location.
18763 Possible values are:
18775 Set the nominal peak luminance.
18778 The values of the @option{w} and @option{h} options are expressions
18779 containing the following constants:
18784 The input width and height
18788 These are the same as @var{in_w} and @var{in_h}.
18792 The output (scaled) width and height
18796 These are the same as @var{out_w} and @var{out_h}
18799 The same as @var{iw} / @var{ih}
18802 input sample aspect ratio
18805 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18809 horizontal and vertical input chroma subsample values. For example for the
18810 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18814 horizontal and vertical output chroma subsample values. For example for the
18815 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18821 @c man end VIDEO FILTERS
18823 @chapter OpenCL Video Filters
18824 @c man begin OPENCL VIDEO FILTERS
18826 Below is a description of the currently available OpenCL video filters.
18828 To enable compilation of these filters you need to configure FFmpeg with
18829 @code{--enable-opencl}.
18831 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18834 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18835 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18836 given device parameters.
18838 @item -filter_hw_device @var{name}
18839 Pass the hardware device called @var{name} to all filters in any filter graph.
18843 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18847 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18849 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18853 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.
18855 @section avgblur_opencl
18857 Apply average blur filter.
18859 The filter accepts the following options:
18863 Set horizontal radius size.
18864 Range is @code{[1, 1024]} and default value is @code{1}.
18867 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18870 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18873 @subsection Example
18877 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.
18879 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18883 @section boxblur_opencl
18885 Apply a boxblur algorithm to the input video.
18887 It accepts the following parameters:
18891 @item luma_radius, lr
18892 @item luma_power, lp
18893 @item chroma_radius, cr
18894 @item chroma_power, cp
18895 @item alpha_radius, ar
18896 @item alpha_power, ap
18900 A description of the accepted options follows.
18903 @item luma_radius, lr
18904 @item chroma_radius, cr
18905 @item alpha_radius, ar
18906 Set an expression for the box radius in pixels used for blurring the
18907 corresponding input plane.
18909 The radius value must be a non-negative number, and must not be
18910 greater than the value of the expression @code{min(w,h)/2} for the
18911 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18914 Default value for @option{luma_radius} is "2". If not specified,
18915 @option{chroma_radius} and @option{alpha_radius} default to the
18916 corresponding value set for @option{luma_radius}.
18918 The expressions can contain the following constants:
18922 The input width and height in pixels.
18926 The input chroma image width and height in pixels.
18930 The horizontal and vertical chroma subsample values. For example, for the
18931 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
18934 @item luma_power, lp
18935 @item chroma_power, cp
18936 @item alpha_power, ap
18937 Specify how many times the boxblur filter is applied to the
18938 corresponding plane.
18940 Default value for @option{luma_power} is 2. If not specified,
18941 @option{chroma_power} and @option{alpha_power} default to the
18942 corresponding value set for @option{luma_power}.
18944 A value of 0 will disable the effect.
18947 @subsection Examples
18949 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.
18953 Apply a boxblur filter with the luma, chroma, and alpha radius
18954 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.
18956 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
18957 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
18961 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.
18963 For the luma plane, a 2x2 box radius will be run once.
18965 For the chroma plane, a 4x4 box radius will be run 5 times.
18967 For the alpha plane, a 3x3 box radius will be run 7 times.
18969 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
18973 @section convolution_opencl
18975 Apply convolution of 3x3, 5x5, 7x7 matrix.
18977 The filter accepts the following options:
18984 Set matrix for each plane.
18985 Matrix is sequence of 9, 25 or 49 signed numbers.
18986 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
18992 Set multiplier for calculated value for each plane.
18993 If unset or 0, it will be sum of all matrix elements.
18994 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
19000 Set bias for each plane. This value is added to the result of the multiplication.
19001 Useful for making the overall image brighter or darker.
19002 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
19006 @subsection Examples
19012 -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
19018 -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
19022 Apply edge enhance:
19024 -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
19030 -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
19034 Apply laplacian edge detector which includes diagonals:
19036 -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
19042 -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
19046 @section dilation_opencl
19048 Apply dilation effect to the video.
19050 This filter replaces the pixel by the local(3x3) maximum.
19052 It accepts the following options:
19059 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19060 If @code{0}, plane will remain unchanged.
19063 Flag which specifies the pixel to refer to.
19064 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19066 Flags to local 3x3 coordinates region centered on @code{x}:
19075 @subsection Example
19079 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.
19081 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19085 @section erosion_opencl
19087 Apply erosion effect to the video.
19089 This filter replaces the pixel by the local(3x3) minimum.
19091 It accepts the following options:
19098 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19099 If @code{0}, plane will remain unchanged.
19102 Flag which specifies the pixel to refer to.
19103 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19105 Flags to local 3x3 coordinates region centered on @code{x}:
19114 @subsection Example
19118 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.
19120 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19124 @section colorkey_opencl
19125 RGB colorspace color keying.
19127 The filter accepts the following options:
19131 The color which will be replaced with transparency.
19134 Similarity percentage with the key color.
19136 0.01 matches only the exact key color, while 1.0 matches everything.
19141 0.0 makes pixels either fully transparent, or not transparent at all.
19143 Higher values result in semi-transparent pixels, with a higher transparency
19144 the more similar the pixels color is to the key color.
19147 @subsection Examples
19151 Make every semi-green pixel in the input transparent with some slight blending:
19153 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
19157 @section overlay_opencl
19159 Overlay one video on top of another.
19161 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19162 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19164 The filter accepts the following options:
19169 Set the x coordinate of the overlaid video on the main video.
19170 Default value is @code{0}.
19173 Set the x coordinate of the overlaid video on the main video.
19174 Default value is @code{0}.
19178 @subsection Examples
19182 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19184 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19187 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19189 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19194 @section prewitt_opencl
19196 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19198 The filter accepts the following option:
19202 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19205 Set value which will be multiplied with filtered result.
19206 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19209 Set value which will be added to filtered result.
19210 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19213 @subsection Example
19217 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19219 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19223 @section roberts_opencl
19224 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19226 The filter accepts the following option:
19230 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19233 Set value which will be multiplied with filtered result.
19234 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19237 Set value which will be added to filtered result.
19238 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19241 @subsection Example
19245 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19247 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19251 @section sobel_opencl
19253 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19255 The filter accepts the following option:
19259 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19262 Set value which will be multiplied with filtered result.
19263 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19266 Set value which will be added to filtered result.
19267 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19270 @subsection Example
19274 Apply sobel operator with scale set to 2 and delta set to 10
19276 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19280 @section tonemap_opencl
19282 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19284 It accepts the following parameters:
19288 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19291 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19294 Apply desaturation for highlights that exceed this level of brightness. The
19295 higher the parameter, the more color information will be preserved. This
19296 setting helps prevent unnaturally blown-out colors for super-highlights, by
19297 (smoothly) turning into white instead. This makes images feel more natural,
19298 at the cost of reducing information about out-of-range colors.
19300 The default value is 0.5, and the algorithm here is a little different from
19301 the cpu version tonemap currently. A setting of 0.0 disables this option.
19304 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19305 is used to detect whether the scene has changed or not. If the distance between
19306 the current frame average brightness and the current running average exceeds
19307 a threshold value, we would re-calculate scene average and peak brightness.
19308 The default value is 0.2.
19311 Specify the output pixel format.
19313 Currently supported formats are:
19320 Set the output color range.
19322 Possible values are:
19328 Default is same as input.
19331 Set the output color primaries.
19333 Possible values are:
19339 Default is same as input.
19342 Set the output transfer characteristics.
19344 Possible values are:
19353 Set the output colorspace matrix.
19355 Possible value are:
19361 Default is same as input.
19365 @subsection Example
19369 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19371 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19375 @section unsharp_opencl
19377 Sharpen or blur the input video.
19379 It accepts the following parameters:
19382 @item luma_msize_x, lx
19383 Set the luma matrix horizontal size.
19384 Range is @code{[1, 23]} and default value is @code{5}.
19386 @item luma_msize_y, ly
19387 Set the luma matrix vertical size.
19388 Range is @code{[1, 23]} and default value is @code{5}.
19390 @item luma_amount, la
19391 Set the luma effect strength.
19392 Range is @code{[-10, 10]} and default value is @code{1.0}.
19394 Negative values will blur the input video, while positive values will
19395 sharpen it, a value of zero will disable the effect.
19397 @item chroma_msize_x, cx
19398 Set the chroma matrix horizontal size.
19399 Range is @code{[1, 23]} and default value is @code{5}.
19401 @item chroma_msize_y, cy
19402 Set the chroma matrix vertical size.
19403 Range is @code{[1, 23]} and default value is @code{5}.
19405 @item chroma_amount, ca
19406 Set the chroma effect strength.
19407 Range is @code{[-10, 10]} and default value is @code{0.0}.
19409 Negative values will blur the input video, while positive values will
19410 sharpen it, a value of zero will disable the effect.
19414 All parameters are optional and default to the equivalent of the
19415 string '5:5:1.0:5:5:0.0'.
19417 @subsection Examples
19421 Apply strong luma sharpen effect:
19423 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19427 Apply a strong blur of both luma and chroma parameters:
19429 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19433 @c man end OPENCL VIDEO FILTERS
19435 @chapter Video Sources
19436 @c man begin VIDEO SOURCES
19438 Below is a description of the currently available video sources.
19442 Buffer video frames, and make them available to the filter chain.
19444 This source is mainly intended for a programmatic use, in particular
19445 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19447 It accepts the following parameters:
19452 Specify the size (width and height) of the buffered video frames. For the
19453 syntax of this option, check the
19454 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19457 The input video width.
19460 The input video height.
19463 A string representing the pixel format of the buffered video frames.
19464 It may be a number corresponding to a pixel format, or a pixel format
19468 Specify the timebase assumed by the timestamps of the buffered frames.
19471 Specify the frame rate expected for the video stream.
19473 @item pixel_aspect, sar
19474 The sample (pixel) aspect ratio of the input video.
19477 Specify the optional parameters to be used for the scale filter which
19478 is automatically inserted when an input change is detected in the
19479 input size or format.
19481 @item hw_frames_ctx
19482 When using a hardware pixel format, this should be a reference to an
19483 AVHWFramesContext describing input frames.
19488 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19491 will instruct the source to accept video frames with size 320x240 and
19492 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19493 square pixels (1:1 sample aspect ratio).
19494 Since the pixel format with name "yuv410p" corresponds to the number 6
19495 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19496 this example corresponds to:
19498 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19501 Alternatively, the options can be specified as a flat string, but this
19502 syntax is deprecated:
19504 @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}]
19508 Create a pattern generated by an elementary cellular automaton.
19510 The initial state of the cellular automaton can be defined through the
19511 @option{filename} and @option{pattern} options. If such options are
19512 not specified an initial state is created randomly.
19514 At each new frame a new row in the video is filled with the result of
19515 the cellular automaton next generation. The behavior when the whole
19516 frame is filled is defined by the @option{scroll} option.
19518 This source accepts the following options:
19522 Read the initial cellular automaton state, i.e. the starting row, from
19523 the specified file.
19524 In the file, each non-whitespace character is considered an alive
19525 cell, a newline will terminate the row, and further characters in the
19526 file will be ignored.
19529 Read the initial cellular automaton state, i.e. the starting row, from
19530 the specified string.
19532 Each non-whitespace character in the string is considered an alive
19533 cell, a newline will terminate the row, and further characters in the
19534 string will be ignored.
19537 Set the video rate, that is the number of frames generated per second.
19540 @item random_fill_ratio, ratio
19541 Set the random fill ratio for the initial cellular automaton row. It
19542 is a floating point number value ranging from 0 to 1, defaults to
19545 This option is ignored when a file or a pattern is specified.
19547 @item random_seed, seed
19548 Set the seed for filling randomly the initial row, must be an integer
19549 included between 0 and UINT32_MAX. If not specified, or if explicitly
19550 set to -1, the filter will try to use a good random seed on a best
19554 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19555 Default value is 110.
19558 Set the size of the output video. For the syntax of this option, check the
19559 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19561 If @option{filename} or @option{pattern} is specified, the size is set
19562 by default to the width of the specified initial state row, and the
19563 height is set to @var{width} * PHI.
19565 If @option{size} is set, it must contain the width of the specified
19566 pattern string, and the specified pattern will be centered in the
19569 If a filename or a pattern string is not specified, the size value
19570 defaults to "320x518" (used for a randomly generated initial state).
19573 If set to 1, scroll the output upward when all the rows in the output
19574 have been already filled. If set to 0, the new generated row will be
19575 written over the top row just after the bottom row is filled.
19578 @item start_full, full
19579 If set to 1, completely fill the output with generated rows before
19580 outputting the first frame.
19581 This is the default behavior, for disabling set the value to 0.
19584 If set to 1, stitch the left and right row edges together.
19585 This is the default behavior, for disabling set the value to 0.
19588 @subsection Examples
19592 Read the initial state from @file{pattern}, and specify an output of
19595 cellauto=f=pattern:s=200x400
19599 Generate a random initial row with a width of 200 cells, with a fill
19602 cellauto=ratio=2/3:s=200x200
19606 Create a pattern generated by rule 18 starting by a single alive cell
19607 centered on an initial row with width 100:
19609 cellauto=p=@@:s=100x400:full=0:rule=18
19613 Specify a more elaborated initial pattern:
19615 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19620 @anchor{coreimagesrc}
19621 @section coreimagesrc
19622 Video source generated on GPU using Apple's CoreImage API on OSX.
19624 This video source is a specialized version of the @ref{coreimage} video filter.
19625 Use a core image generator at the beginning of the applied filterchain to
19626 generate the content.
19628 The coreimagesrc video source accepts the following options:
19630 @item list_generators
19631 List all available generators along with all their respective options as well as
19632 possible minimum and maximum values along with the default values.
19634 list_generators=true
19638 Specify the size of the sourced video. For the syntax of this option, check the
19639 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19640 The default value is @code{320x240}.
19643 Specify the frame rate of the sourced video, as the number of frames
19644 generated per second. It has to be a string in the format
19645 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19646 number or a valid video frame rate abbreviation. The default value is
19650 Set the sample aspect ratio of the sourced video.
19653 Set the duration of the sourced video. See
19654 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19655 for the accepted syntax.
19657 If not specified, or the expressed duration is negative, the video is
19658 supposed to be generated forever.
19661 Additionally, all options of the @ref{coreimage} video filter are accepted.
19662 A complete filterchain can be used for further processing of the
19663 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19664 and examples for details.
19666 @subsection Examples
19671 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19672 given as complete and escaped command-line for Apple's standard bash shell:
19674 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19676 This example is equivalent to the QRCode example of @ref{coreimage} without the
19677 need for a nullsrc video source.
19681 @section mandelbrot
19683 Generate a Mandelbrot set fractal, and progressively zoom towards the
19684 point specified with @var{start_x} and @var{start_y}.
19686 This source accepts the following options:
19691 Set the terminal pts value. Default value is 400.
19694 Set the terminal scale value.
19695 Must be a floating point value. Default value is 0.3.
19698 Set the inner coloring mode, that is the algorithm used to draw the
19699 Mandelbrot fractal internal region.
19701 It shall assume one of the following values:
19706 Show time until convergence.
19708 Set color based on point closest to the origin of the iterations.
19713 Default value is @var{mincol}.
19716 Set the bailout value. Default value is 10.0.
19719 Set the maximum of iterations performed by the rendering
19720 algorithm. Default value is 7189.
19723 Set outer coloring mode.
19724 It shall assume one of following values:
19726 @item iteration_count
19727 Set iteration count mode.
19728 @item normalized_iteration_count
19729 set normalized iteration count mode.
19731 Default value is @var{normalized_iteration_count}.
19734 Set frame rate, expressed as number of frames per second. Default
19738 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19739 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19742 Set the initial scale value. Default value is 3.0.
19745 Set the initial x position. Must be a floating point value between
19746 -100 and 100. Default value is -0.743643887037158704752191506114774.
19749 Set the initial y position. Must be a floating point value between
19750 -100 and 100. Default value is -0.131825904205311970493132056385139.
19755 Generate various test patterns, as generated by the MPlayer test filter.
19757 The size of the generated video is fixed, and is 256x256.
19758 This source is useful in particular for testing encoding features.
19760 This source accepts the following options:
19765 Specify the frame rate of the sourced video, as the number of frames
19766 generated per second. It has to be a string in the format
19767 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19768 number or a valid video frame rate abbreviation. The default value is
19772 Set the duration of the sourced video. See
19773 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19774 for the accepted syntax.
19776 If not specified, or the expressed duration is negative, the video is
19777 supposed to be generated forever.
19781 Set the number or the name of the test to perform. Supported tests are:
19797 Default value is "all", which will cycle through the list of all tests.
19802 mptestsrc=t=dc_luma
19805 will generate a "dc_luma" test pattern.
19807 @section frei0r_src
19809 Provide a frei0r source.
19811 To enable compilation of this filter you need to install the frei0r
19812 header and configure FFmpeg with @code{--enable-frei0r}.
19814 This source accepts the following parameters:
19819 The size of the video to generate. For the syntax of this option, check the
19820 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19823 The framerate of the generated video. It may be a string of the form
19824 @var{num}/@var{den} or a frame rate abbreviation.
19827 The name to the frei0r source to load. For more information regarding frei0r and
19828 how to set the parameters, read the @ref{frei0r} section in the video filters
19831 @item filter_params
19832 A '|'-separated list of parameters to pass to the frei0r source.
19836 For example, to generate a frei0r partik0l source with size 200x200
19837 and frame rate 10 which is overlaid on the overlay filter main input:
19839 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19844 Generate a life pattern.
19846 This source is based on a generalization of John Conway's life game.
19848 The sourced input represents a life grid, each pixel represents a cell
19849 which can be in one of two possible states, alive or dead. Every cell
19850 interacts with its eight neighbours, which are the cells that are
19851 horizontally, vertically, or diagonally adjacent.
19853 At each interaction the grid evolves according to the adopted rule,
19854 which specifies the number of neighbor alive cells which will make a
19855 cell stay alive or born. The @option{rule} option allows one to specify
19858 This source accepts the following options:
19862 Set the file from which to read the initial grid state. In the file,
19863 each non-whitespace character is considered an alive cell, and newline
19864 is used to delimit the end of each row.
19866 If this option is not specified, the initial grid is generated
19870 Set the video rate, that is the number of frames generated per second.
19873 @item random_fill_ratio, ratio
19874 Set the random fill ratio for the initial random grid. It is a
19875 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19876 It is ignored when a file is specified.
19878 @item random_seed, seed
19879 Set the seed for filling the initial random grid, must be an integer
19880 included between 0 and UINT32_MAX. If not specified, or if explicitly
19881 set to -1, the filter will try to use a good random seed on a best
19887 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19888 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19889 @var{NS} specifies the number of alive neighbor cells which make a
19890 live cell stay alive, and @var{NB} the number of alive neighbor cells
19891 which make a dead cell to become alive (i.e. to "born").
19892 "s" and "b" can be used in place of "S" and "B", respectively.
19894 Alternatively a rule can be specified by an 18-bits integer. The 9
19895 high order bits are used to encode the next cell state if it is alive
19896 for each number of neighbor alive cells, the low order bits specify
19897 the rule for "borning" new cells. Higher order bits encode for an
19898 higher number of neighbor cells.
19899 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19900 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19902 Default value is "S23/B3", which is the original Conway's game of life
19903 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19904 cells, and will born a new cell if there are three alive cells around
19908 Set the size of the output video. For the syntax of this option, check the
19909 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19911 If @option{filename} is specified, the size is set by default to the
19912 same size of the input file. If @option{size} is set, it must contain
19913 the size specified in the input file, and the initial grid defined in
19914 that file is centered in the larger resulting area.
19916 If a filename is not specified, the size value defaults to "320x240"
19917 (used for a randomly generated initial grid).
19920 If set to 1, stitch the left and right grid edges together, and the
19921 top and bottom edges also. Defaults to 1.
19924 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
19925 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
19926 value from 0 to 255.
19929 Set the color of living (or new born) cells.
19932 Set the color of dead cells. If @option{mold} is set, this is the first color
19933 used to represent a dead cell.
19936 Set mold color, for definitely dead and moldy cells.
19938 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
19939 ffmpeg-utils manual,ffmpeg-utils}.
19942 @subsection Examples
19946 Read a grid from @file{pattern}, and center it on a grid of size
19949 life=f=pattern:s=300x300
19953 Generate a random grid of size 200x200, with a fill ratio of 2/3:
19955 life=ratio=2/3:s=200x200
19959 Specify a custom rule for evolving a randomly generated grid:
19965 Full example with slow death effect (mold) using @command{ffplay}:
19967 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
19974 @anchor{haldclutsrc}
19977 @anchor{pal100bars}
19978 @anchor{rgbtestsrc}
19980 @anchor{smptehdbars}
19983 @anchor{yuvtestsrc}
19984 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
19986 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
19988 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
19990 The @code{color} source provides an uniformly colored input.
19992 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
19993 @ref{haldclut} filter.
19995 The @code{nullsrc} source returns unprocessed video frames. It is
19996 mainly useful to be employed in analysis / debugging tools, or as the
19997 source for filters which ignore the input data.
19999 The @code{pal75bars} source generates a color bars pattern, based on
20000 EBU PAL recommendations with 75% color levels.
20002 The @code{pal100bars} source generates a color bars pattern, based on
20003 EBU PAL recommendations with 100% color levels.
20005 The @code{rgbtestsrc} source generates an RGB test pattern useful for
20006 detecting RGB vs BGR issues. You should see a red, green and blue
20007 stripe from top to bottom.
20009 The @code{smptebars} source generates a color bars pattern, based on
20010 the SMPTE Engineering Guideline EG 1-1990.
20012 The @code{smptehdbars} source generates a color bars pattern, based on
20013 the SMPTE RP 219-2002.
20015 The @code{testsrc} source generates a test video pattern, showing a
20016 color pattern, a scrolling gradient and a timestamp. This is mainly
20017 intended for testing purposes.
20019 The @code{testsrc2} source is similar to testsrc, but supports more
20020 pixel formats instead of just @code{rgb24}. This allows using it as an
20021 input for other tests without requiring a format conversion.
20023 The @code{yuvtestsrc} source generates an YUV test pattern. You should
20024 see a y, cb and cr stripe from top to bottom.
20026 The sources accept the following parameters:
20031 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
20032 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
20033 pixels to be used as identity matrix for 3D lookup tables. Each component is
20034 coded on a @code{1/(N*N)} scale.
20037 Specify the color of the source, only available in the @code{color}
20038 source. For the syntax of this option, check the
20039 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
20042 Specify the size of the sourced video. For the syntax of this option, check the
20043 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20044 The default value is @code{320x240}.
20046 This option is not available with the @code{allrgb}, @code{allyuv}, and
20047 @code{haldclutsrc} filters.
20050 Specify the frame rate of the sourced video, as the number of frames
20051 generated per second. It has to be a string in the format
20052 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
20053 number or a valid video frame rate abbreviation. The default value is
20057 Set the duration of the sourced video. See
20058 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20059 for the accepted syntax.
20061 If not specified, or the expressed duration is negative, the video is
20062 supposed to be generated forever.
20065 Set the sample aspect ratio of the sourced video.
20068 Specify the alpha (opacity) of the background, only available in the
20069 @code{testsrc2} source. The value must be between 0 (fully transparent) and
20070 255 (fully opaque, the default).
20073 Set the number of decimals to show in the timestamp, only available in the
20074 @code{testsrc} source.
20076 The displayed timestamp value will correspond to the original
20077 timestamp value multiplied by the power of 10 of the specified
20078 value. Default value is 0.
20081 @subsection Examples
20085 Generate a video with a duration of 5.3 seconds, with size
20086 176x144 and a frame rate of 10 frames per second:
20088 testsrc=duration=5.3:size=qcif:rate=10
20092 The following graph description will generate a red source
20093 with an opacity of 0.2, with size "qcif" and a frame rate of 10
20096 color=c=red@@0.2:s=qcif:r=10
20100 If the input content is to be ignored, @code{nullsrc} can be used. The
20101 following command generates noise in the luminance plane by employing
20102 the @code{geq} filter:
20104 nullsrc=s=256x256, geq=random(1)*255:128:128
20108 @subsection Commands
20110 The @code{color} source supports the following commands:
20114 Set the color of the created image. Accepts the same syntax of the
20115 corresponding @option{color} option.
20120 Generate video using an OpenCL program.
20125 OpenCL program source file.
20128 Kernel name in program.
20131 Size of frames to generate. This must be set.
20134 Pixel format to use for the generated frames. This must be set.
20137 Number of frames generated every second. Default value is '25'.
20141 For details of how the program loading works, see the @ref{program_opencl}
20148 Generate a colour ramp by setting pixel values from the position of the pixel
20149 in the output image. (Note that this will work with all pixel formats, but
20150 the generated output will not be the same.)
20152 __kernel void ramp(__write_only image2d_t dst,
20153 unsigned int index)
20155 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20158 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20160 write_imagef(dst, loc, val);
20165 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20167 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20168 unsigned int index)
20170 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20172 float4 value = 0.0f;
20173 int x = loc.x + index;
20174 int y = loc.y + index;
20175 while (x > 0 || y > 0) {
20176 if (x % 3 == 1 && y % 3 == 1) {
20184 write_imagef(dst, loc, value);
20190 @c man end VIDEO SOURCES
20192 @chapter Video Sinks
20193 @c man begin VIDEO SINKS
20195 Below is a description of the currently available video sinks.
20197 @section buffersink
20199 Buffer video frames, and make them available to the end of the filter
20202 This sink is mainly intended for programmatic use, in particular
20203 through the interface defined in @file{libavfilter/buffersink.h}
20204 or the options system.
20206 It accepts a pointer to an AVBufferSinkContext structure, which
20207 defines the incoming buffers' formats, to be passed as the opaque
20208 parameter to @code{avfilter_init_filter} for initialization.
20212 Null video sink: do absolutely nothing with the input video. It is
20213 mainly useful as a template and for use in analysis / debugging
20216 @c man end VIDEO SINKS
20218 @chapter Multimedia Filters
20219 @c man begin MULTIMEDIA FILTERS
20221 Below is a description of the currently available multimedia filters.
20225 Convert input audio to a video output, displaying the audio bit scope.
20227 The filter accepts the following options:
20231 Set frame rate, expressed as number of frames per second. Default
20235 Specify the video size for the output. For the syntax of this option, check the
20236 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20237 Default value is @code{1024x256}.
20240 Specify list of colors separated by space or by '|' which will be used to
20241 draw channels. Unrecognized or missing colors will be replaced
20245 @section ahistogram
20247 Convert input audio to a video output, displaying the volume histogram.
20249 The filter accepts the following options:
20253 Specify how histogram is calculated.
20255 It accepts the following values:
20258 Use single histogram for all channels.
20260 Use separate histogram for each channel.
20262 Default is @code{single}.
20265 Set frame rate, expressed as number of frames per second. Default
20269 Specify the video size for the output. For the syntax of this option, check the
20270 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20271 Default value is @code{hd720}.
20276 It accepts the following values:
20287 reverse logarithmic
20289 Default is @code{log}.
20292 Set amplitude scale.
20294 It accepts the following values:
20301 Default is @code{log}.
20304 Set how much frames to accumulate in histogram.
20305 Default is 1. Setting this to -1 accumulates all frames.
20308 Set histogram ratio of window height.
20311 Set sonogram sliding.
20313 It accepts the following values:
20316 replace old rows with new ones.
20318 scroll from top to bottom.
20320 Default is @code{replace}.
20323 @section aphasemeter
20325 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20326 representing mean phase of current audio frame. A video output can also be produced and is
20327 enabled by default. The audio is passed through as first output.
20329 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20330 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20331 and @code{1} means channels are in phase.
20333 The filter accepts the following options, all related to its video output:
20337 Set the output frame rate. Default value is @code{25}.
20340 Set the video size for the output. For the syntax of this option, check the
20341 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20342 Default value is @code{800x400}.
20347 Specify the red, green, blue contrast. Default values are @code{2},
20348 @code{7} and @code{1}.
20349 Allowed range is @code{[0, 255]}.
20352 Set color which will be used for drawing median phase. If color is
20353 @code{none} which is default, no median phase value will be drawn.
20356 Enable video output. Default is enabled.
20359 @section avectorscope
20361 Convert input audio to a video output, representing the audio vector
20364 The filter is used to measure the difference between channels of stereo
20365 audio stream. A monoaural signal, consisting of identical left and right
20366 signal, results in straight vertical line. Any stereo separation is visible
20367 as a deviation from this line, creating a Lissajous figure.
20368 If the straight (or deviation from it) but horizontal line appears this
20369 indicates that the left and right channels are out of phase.
20371 The filter accepts the following options:
20375 Set the vectorscope mode.
20377 Available values are:
20380 Lissajous rotated by 45 degrees.
20383 Same as above but not rotated.
20386 Shape resembling half of circle.
20389 Default value is @samp{lissajous}.
20392 Set the video size for the output. For the syntax of this option, check the
20393 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20394 Default value is @code{400x400}.
20397 Set the output frame rate. Default value is @code{25}.
20403 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20404 @code{160}, @code{80} and @code{255}.
20405 Allowed range is @code{[0, 255]}.
20411 Specify the red, green, blue and alpha fade. Default values are @code{15},
20412 @code{10}, @code{5} and @code{5}.
20413 Allowed range is @code{[0, 255]}.
20416 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20417 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20420 Set the vectorscope drawing mode.
20422 Available values are:
20425 Draw dot for each sample.
20428 Draw line between previous and current sample.
20431 Default value is @samp{dot}.
20434 Specify amplitude scale of audio samples.
20436 Available values are:
20452 Swap left channel axis with right channel axis.
20462 Mirror only x axis.
20465 Mirror only y axis.
20473 @subsection Examples
20477 Complete example using @command{ffplay}:
20479 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20480 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20484 @section bench, abench
20486 Benchmark part of a filtergraph.
20488 The filter accepts the following options:
20492 Start or stop a timer.
20494 Available values are:
20497 Get the current time, set it as frame metadata (using the key
20498 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20501 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20502 the input frame metadata to get the time difference. Time difference, average,
20503 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20504 @code{min}) are then printed. The timestamps are expressed in seconds.
20508 @subsection Examples
20512 Benchmark @ref{selectivecolor} filter:
20514 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20520 Concatenate audio and video streams, joining them together one after the
20523 The filter works on segments of synchronized video and audio streams. All
20524 segments must have the same number of streams of each type, and that will
20525 also be the number of streams at output.
20527 The filter accepts the following options:
20532 Set the number of segments. Default is 2.
20535 Set the number of output video streams, that is also the number of video
20536 streams in each segment. Default is 1.
20539 Set the number of output audio streams, that is also the number of audio
20540 streams in each segment. Default is 0.
20543 Activate unsafe mode: do not fail if segments have a different format.
20547 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20548 @var{a} audio outputs.
20550 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20551 segment, in the same order as the outputs, then the inputs for the second
20554 Related streams do not always have exactly the same duration, for various
20555 reasons including codec frame size or sloppy authoring. For that reason,
20556 related synchronized streams (e.g. a video and its audio track) should be
20557 concatenated at once. The concat filter will use the duration of the longest
20558 stream in each segment (except the last one), and if necessary pad shorter
20559 audio streams with silence.
20561 For this filter to work correctly, all segments must start at timestamp 0.
20563 All corresponding streams must have the same parameters in all segments; the
20564 filtering system will automatically select a common pixel format for video
20565 streams, and a common sample format, sample rate and channel layout for
20566 audio streams, but other settings, such as resolution, must be converted
20567 explicitly by the user.
20569 Different frame rates are acceptable but will result in variable frame rate
20570 at output; be sure to configure the output file to handle it.
20572 @subsection Examples
20576 Concatenate an opening, an episode and an ending, all in bilingual version
20577 (video in stream 0, audio in streams 1 and 2):
20579 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20580 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20581 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20582 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20586 Concatenate two parts, handling audio and video separately, using the
20587 (a)movie sources, and adjusting the resolution:
20589 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20590 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20591 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20593 Note that a desync will happen at the stitch if the audio and video streams
20594 do not have exactly the same duration in the first file.
20598 @subsection Commands
20600 This filter supports the following commands:
20603 Close the current segment and step to the next one
20606 @section drawgraph, adrawgraph
20608 Draw a graph using input video or audio metadata.
20610 It accepts the following parameters:
20614 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20617 Set 1st foreground color expression.
20620 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20623 Set 2nd foreground color expression.
20626 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20629 Set 3rd foreground color expression.
20632 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20635 Set 4th foreground color expression.
20638 Set minimal value of metadata value.
20641 Set maximal value of metadata value.
20644 Set graph background color. Default is white.
20649 Available values for mode is:
20656 Default is @code{line}.
20661 Available values for slide is:
20664 Draw new frame when right border is reached.
20667 Replace old columns with new ones.
20670 Scroll from right to left.
20673 Scroll from left to right.
20676 Draw single picture.
20679 Default is @code{frame}.
20682 Set size of graph video. For the syntax of this option, check the
20683 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20684 The default value is @code{900x256}.
20686 The foreground color expressions can use the following variables:
20689 Minimal value of metadata value.
20692 Maximal value of metadata value.
20695 Current metadata key value.
20698 The color is defined as 0xAABBGGRR.
20701 Example using metadata from @ref{signalstats} filter:
20703 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20706 Example using metadata from @ref{ebur128} filter:
20708 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20714 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
20715 level. By default, it logs a message at a frequency of 10Hz with the
20716 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20717 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20719 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
20720 sample format is double-precision floating point. The input stream will be converted to
20721 this specification, if needed. Users may need to insert aformat and/or aresample filters
20722 after this filter to obtain the original parameters.
20724 The filter also has a video output (see the @var{video} option) with a real
20725 time graph to observe the loudness evolution. The graphic contains the logged
20726 message mentioned above, so it is not printed anymore when this option is set,
20727 unless the verbose logging is set. The main graphing area contains the
20728 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20729 the momentary loudness (400 milliseconds), but can optionally be configured
20730 to instead display short-term loudness (see @var{gauge}).
20732 The green area marks a +/- 1LU target range around the target loudness
20733 (-23LUFS by default, unless modified through @var{target}).
20735 More information about the Loudness Recommendation EBU R128 on
20736 @url{http://tech.ebu.ch/loudness}.
20738 The filter accepts the following options:
20743 Activate the video output. The audio stream is passed unchanged whether this
20744 option is set or no. The video stream will be the first output stream if
20745 activated. Default is @code{0}.
20748 Set the video size. This option is for video only. For the syntax of this
20750 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20751 Default and minimum resolution is @code{640x480}.
20754 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20755 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20756 other integer value between this range is allowed.
20759 Set metadata injection. If set to @code{1}, the audio input will be segmented
20760 into 100ms output frames, each of them containing various loudness information
20761 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20763 Default is @code{0}.
20766 Force the frame logging level.
20768 Available values are:
20771 information logging level
20773 verbose logging level
20776 By default, the logging level is set to @var{info}. If the @option{video} or
20777 the @option{metadata} options are set, it switches to @var{verbose}.
20782 Available modes can be cumulated (the option is a @code{flag} type). Possible
20786 Disable any peak mode (default).
20788 Enable sample-peak mode.
20790 Simple peak mode looking for the higher sample value. It logs a message
20791 for sample-peak (identified by @code{SPK}).
20793 Enable true-peak mode.
20795 If enabled, the peak lookup is done on an over-sampled version of the input
20796 stream for better peak accuracy. It logs a message for true-peak.
20797 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20798 This mode requires a build with @code{libswresample}.
20802 Treat mono input files as "dual mono". If a mono file is intended for playback
20803 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20804 If set to @code{true}, this option will compensate for this effect.
20805 Multi-channel input files are not affected by this option.
20808 Set a specific pan law to be used for the measurement of dual mono files.
20809 This parameter is optional, and has a default value of -3.01dB.
20812 Set a specific target level (in LUFS) used as relative zero in the visualization.
20813 This parameter is optional and has a default value of -23LUFS as specified
20814 by EBU R128. However, material published online may prefer a level of -16LUFS
20815 (e.g. for use with podcasts or video platforms).
20818 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20819 @code{shortterm}. By default the momentary value will be used, but in certain
20820 scenarios it may be more useful to observe the short term value instead (e.g.
20824 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20825 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20826 video output, not the summary or continuous log output.
20829 @subsection Examples
20833 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20835 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20839 Run an analysis with @command{ffmpeg}:
20841 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20845 @section interleave, ainterleave
20847 Temporally interleave frames from several inputs.
20849 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20851 These filters read frames from several inputs and send the oldest
20852 queued frame to the output.
20854 Input streams must have well defined, monotonically increasing frame
20857 In order to submit one frame to output, these filters need to enqueue
20858 at least one frame for each input, so they cannot work in case one
20859 input is not yet terminated and will not receive incoming frames.
20861 For example consider the case when one input is a @code{select} filter
20862 which always drops input frames. The @code{interleave} filter will keep
20863 reading from that input, but it will never be able to send new frames
20864 to output until the input sends an end-of-stream signal.
20866 Also, depending on inputs synchronization, the filters will drop
20867 frames in case one input receives more frames than the other ones, and
20868 the queue is already filled.
20870 These filters accept the following options:
20874 Set the number of different inputs, it is 2 by default.
20877 @subsection Examples
20881 Interleave frames belonging to different streams using @command{ffmpeg}:
20883 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20887 Add flickering blur effect:
20889 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20893 @section metadata, ametadata
20895 Manipulate frame metadata.
20897 This filter accepts the following options:
20901 Set mode of operation of the filter.
20903 Can be one of the following:
20907 If both @code{value} and @code{key} is set, select frames
20908 which have such metadata. If only @code{key} is set, select
20909 every frame that has such key in metadata.
20912 Add new metadata @code{key} and @code{value}. If key is already available
20916 Modify value of already present key.
20919 If @code{value} is set, delete only keys that have such value.
20920 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
20924 Print key and its value if metadata was found. If @code{key} is not set print all
20925 metadata values available in frame.
20929 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
20932 Set metadata value which will be used. This option is mandatory for
20933 @code{modify} and @code{add} mode.
20936 Which function to use when comparing metadata value and @code{value}.
20938 Can be one of following:
20942 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
20945 Values are interpreted as strings, returns true if metadata value starts with
20946 the @code{value} option string.
20949 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
20952 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
20955 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
20958 Values are interpreted as floats, returns true if expression from option @code{expr}
20963 Set expression which is used when @code{function} is set to @code{expr}.
20964 The expression is evaluated through the eval API and can contain the following
20969 Float representation of @code{value} from metadata key.
20972 Float representation of @code{value} as supplied by user in @code{value} option.
20976 If specified in @code{print} mode, output is written to the named file. Instead of
20977 plain filename any writable url can be specified. Filename ``-'' is a shorthand
20978 for standard output. If @code{file} option is not set, output is written to the log
20979 with AV_LOG_INFO loglevel.
20983 @subsection Examples
20987 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
20990 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
20993 Print silencedetect output to file @file{metadata.txt}.
20995 silencedetect,ametadata=mode=print:file=metadata.txt
20998 Direct all metadata to a pipe with file descriptor 4.
21000 metadata=mode=print:file='pipe\:4'
21004 @section perms, aperms
21006 Set read/write permissions for the output frames.
21008 These filters are mainly aimed at developers to test direct path in the
21009 following filter in the filtergraph.
21011 The filters accept the following options:
21015 Select the permissions mode.
21017 It accepts the following values:
21020 Do nothing. This is the default.
21022 Set all the output frames read-only.
21024 Set all the output frames directly writable.
21026 Make the frame read-only if writable, and writable if read-only.
21028 Set each output frame read-only or writable randomly.
21032 Set the seed for the @var{random} mode, must be an integer included between
21033 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
21034 @code{-1}, the filter will try to use a good random seed on a best effort
21038 Note: in case of auto-inserted filter between the permission filter and the
21039 following one, the permission might not be received as expected in that
21040 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
21041 perms/aperms filter can avoid this problem.
21043 @section realtime, arealtime
21045 Slow down filtering to match real time approximately.
21047 These filters will pause the filtering for a variable amount of time to
21048 match the output rate with the input timestamps.
21049 They are similar to the @option{re} option to @code{ffmpeg}.
21051 They accept the following options:
21055 Time limit for the pauses. Any pause longer than that will be considered
21056 a timestamp discontinuity and reset the timer. Default is 2 seconds.
21060 @section select, aselect
21062 Select frames to pass in output.
21064 This filter accepts the following options:
21069 Set expression, which is evaluated for each input frame.
21071 If the expression is evaluated to zero, the frame is discarded.
21073 If the evaluation result is negative or NaN, the frame is sent to the
21074 first output; otherwise it is sent to the output with index
21075 @code{ceil(val)-1}, assuming that the input index starts from 0.
21077 For example a value of @code{1.2} corresponds to the output with index
21078 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
21081 Set the number of outputs. The output to which to send the selected
21082 frame is based on the result of the evaluation. Default value is 1.
21085 The expression can contain the following constants:
21089 The (sequential) number of the filtered frame, starting from 0.
21092 The (sequential) number of the selected frame, starting from 0.
21094 @item prev_selected_n
21095 The sequential number of the last selected frame. It's NAN if undefined.
21098 The timebase of the input timestamps.
21101 The PTS (Presentation TimeStamp) of the filtered video frame,
21102 expressed in @var{TB} units. It's NAN if undefined.
21105 The PTS of the filtered video frame,
21106 expressed in seconds. It's NAN if undefined.
21109 The PTS of the previously filtered video frame. It's NAN if undefined.
21111 @item prev_selected_pts
21112 The PTS of the last previously filtered video frame. It's NAN if undefined.
21114 @item prev_selected_t
21115 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
21118 The PTS of the first video frame in the video. It's NAN if undefined.
21121 The time of the first video frame in the video. It's NAN if undefined.
21123 @item pict_type @emph{(video only)}
21124 The type of the filtered frame. It can assume one of the following
21136 @item interlace_type @emph{(video only)}
21137 The frame interlace type. It can assume one of the following values:
21140 The frame is progressive (not interlaced).
21142 The frame is top-field-first.
21144 The frame is bottom-field-first.
21147 @item consumed_sample_n @emph{(audio only)}
21148 the number of selected samples before the current frame
21150 @item samples_n @emph{(audio only)}
21151 the number of samples in the current frame
21153 @item sample_rate @emph{(audio only)}
21154 the input sample rate
21157 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21160 the position in the file of the filtered frame, -1 if the information
21161 is not available (e.g. for synthetic video)
21163 @item scene @emph{(video only)}
21164 value between 0 and 1 to indicate a new scene; a low value reflects a low
21165 probability for the current frame to introduce a new scene, while a higher
21166 value means the current frame is more likely to be one (see the example below)
21168 @item concatdec_select
21169 The concat demuxer can select only part of a concat input file by setting an
21170 inpoint and an outpoint, but the output packets may not be entirely contained
21171 in the selected interval. By using this variable, it is possible to skip frames
21172 generated by the concat demuxer which are not exactly contained in the selected
21175 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21176 and the @var{lavf.concat.duration} packet metadata values which are also
21177 present in the decoded frames.
21179 The @var{concatdec_select} variable is -1 if the frame pts is at least
21180 start_time and either the duration metadata is missing or the frame pts is less
21181 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21184 That basically means that an input frame is selected if its pts is within the
21185 interval set by the concat demuxer.
21189 The default value of the select expression is "1".
21191 @subsection Examples
21195 Select all frames in input:
21200 The example above is the same as:
21212 Select only I-frames:
21214 select='eq(pict_type\,I)'
21218 Select one frame every 100:
21220 select='not(mod(n\,100))'
21224 Select only frames contained in the 10-20 time interval:
21226 select=between(t\,10\,20)
21230 Select only I-frames contained in the 10-20 time interval:
21232 select=between(t\,10\,20)*eq(pict_type\,I)
21236 Select frames with a minimum distance of 10 seconds:
21238 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21242 Use aselect to select only audio frames with samples number > 100:
21244 aselect='gt(samples_n\,100)'
21248 Create a mosaic of the first scenes:
21250 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21253 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21257 Send even and odd frames to separate outputs, and compose them:
21259 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21263 Select useful frames from an ffconcat file which is using inpoints and
21264 outpoints but where the source files are not intra frame only.
21266 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21270 @section sendcmd, asendcmd
21272 Send commands to filters in the filtergraph.
21274 These filters read commands to be sent to other filters in the
21277 @code{sendcmd} must be inserted between two video filters,
21278 @code{asendcmd} must be inserted between two audio filters, but apart
21279 from that they act the same way.
21281 The specification of commands can be provided in the filter arguments
21282 with the @var{commands} option, or in a file specified by the
21283 @var{filename} option.
21285 These filters accept the following options:
21288 Set the commands to be read and sent to the other filters.
21290 Set the filename of the commands to be read and sent to the other
21294 @subsection Commands syntax
21296 A commands description consists of a sequence of interval
21297 specifications, comprising a list of commands to be executed when a
21298 particular event related to that interval occurs. The occurring event
21299 is typically the current frame time entering or leaving a given time
21302 An interval is specified by the following syntax:
21304 @var{START}[-@var{END}] @var{COMMANDS};
21307 The time interval is specified by the @var{START} and @var{END} times.
21308 @var{END} is optional and defaults to the maximum time.
21310 The current frame time is considered within the specified interval if
21311 it is included in the interval [@var{START}, @var{END}), that is when
21312 the time is greater or equal to @var{START} and is lesser than
21315 @var{COMMANDS} consists of a sequence of one or more command
21316 specifications, separated by ",", relating to that interval. The
21317 syntax of a command specification is given by:
21319 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21322 @var{FLAGS} is optional and specifies the type of events relating to
21323 the time interval which enable sending the specified command, and must
21324 be a non-null sequence of identifier flags separated by "+" or "|" and
21325 enclosed between "[" and "]".
21327 The following flags are recognized:
21330 The command is sent when the current frame timestamp enters the
21331 specified interval. In other words, the command is sent when the
21332 previous frame timestamp was not in the given interval, and the
21336 The command is sent when the current frame timestamp leaves the
21337 specified interval. In other words, the command is sent when the
21338 previous frame timestamp was in the given interval, and the
21342 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21345 @var{TARGET} specifies the target of the command, usually the name of
21346 the filter class or a specific filter instance name.
21348 @var{COMMAND} specifies the name of the command for the target filter.
21350 @var{ARG} is optional and specifies the optional list of argument for
21351 the given @var{COMMAND}.
21353 Between one interval specification and another, whitespaces, or
21354 sequences of characters starting with @code{#} until the end of line,
21355 are ignored and can be used to annotate comments.
21357 A simplified BNF description of the commands specification syntax
21360 @var{COMMAND_FLAG} ::= "enter" | "leave"
21361 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21362 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21363 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21364 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21365 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21368 @subsection Examples
21372 Specify audio tempo change at second 4:
21374 asendcmd=c='4.0 atempo tempo 1.5',atempo
21378 Target a specific filter instance:
21380 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21384 Specify a list of drawtext and hue commands in a file.
21386 # show text in the interval 5-10
21387 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21388 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21390 # desaturate the image in the interval 15-20
21391 15.0-20.0 [enter] hue s 0,
21392 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21394 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21396 # apply an exponential saturation fade-out effect, starting from time 25
21397 25 [enter] hue s exp(25-t)
21400 A filtergraph allowing to read and process the above command list
21401 stored in a file @file{test.cmd}, can be specified with:
21403 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21408 @section setpts, asetpts
21410 Change the PTS (presentation timestamp) of the input frames.
21412 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21414 This filter accepts the following options:
21419 The expression which is evaluated for each frame to construct its timestamp.
21423 The expression is evaluated through the eval API and can contain the following
21427 @item FRAME_RATE, FR
21428 frame rate, only defined for constant frame-rate video
21431 The presentation timestamp in input
21434 The count of the input frame for video or the number of consumed samples,
21435 not including the current frame for audio, starting from 0.
21437 @item NB_CONSUMED_SAMPLES
21438 The number of consumed samples, not including the current frame (only
21441 @item NB_SAMPLES, S
21442 The number of samples in the current frame (only audio)
21444 @item SAMPLE_RATE, SR
21445 The audio sample rate.
21448 The PTS of the first frame.
21451 the time in seconds of the first frame
21454 State whether the current frame is interlaced.
21457 the time in seconds of the current frame
21460 original position in the file of the frame, or undefined if undefined
21461 for the current frame
21464 The previous input PTS.
21467 previous input time in seconds
21470 The previous output PTS.
21473 previous output time in seconds
21476 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21480 The wallclock (RTC) time at the start of the movie in microseconds.
21483 The timebase of the input timestamps.
21487 @subsection Examples
21491 Start counting PTS from zero
21493 setpts=PTS-STARTPTS
21497 Apply fast motion effect:
21503 Apply slow motion effect:
21509 Set fixed rate of 25 frames per second:
21515 Set fixed rate 25 fps with some jitter:
21517 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21521 Apply an offset of 10 seconds to the input PTS:
21527 Generate timestamps from a "live source" and rebase onto the current timebase:
21529 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21533 Generate timestamps by counting samples:
21542 Force color range for the output video frame.
21544 The @code{setrange} filter marks the color range property for the
21545 output frames. It does not change the input frame, but only sets the
21546 corresponding property, which affects how the frame is treated by
21549 The filter accepts the following options:
21554 Available values are:
21558 Keep the same color range property.
21560 @item unspecified, unknown
21561 Set the color range as unspecified.
21563 @item limited, tv, mpeg
21564 Set the color range as limited.
21566 @item full, pc, jpeg
21567 Set the color range as full.
21571 @section settb, asettb
21573 Set the timebase to use for the output frames timestamps.
21574 It is mainly useful for testing timebase configuration.
21576 It accepts the following parameters:
21581 The expression which is evaluated into the output timebase.
21585 The value for @option{tb} is an arithmetic expression representing a
21586 rational. The expression can contain the constants "AVTB" (the default
21587 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21588 audio only). Default value is "intb".
21590 @subsection Examples
21594 Set the timebase to 1/25:
21600 Set the timebase to 1/10:
21606 Set the timebase to 1001/1000:
21612 Set the timebase to 2*intb:
21618 Set the default timebase value:
21625 Convert input audio to a video output representing frequency spectrum
21626 logarithmically using Brown-Puckette constant Q transform algorithm with
21627 direct frequency domain coefficient calculation (but the transform itself
21628 is not really constant Q, instead the Q factor is actually variable/clamped),
21629 with musical tone scale, from E0 to D#10.
21631 The filter accepts the following options:
21635 Specify the video size for the output. It must be even. For the syntax of this option,
21636 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21637 Default value is @code{1920x1080}.
21640 Set the output frame rate. Default value is @code{25}.
21643 Set the bargraph height. It must be even. Default value is @code{-1} which
21644 computes the bargraph height automatically.
21647 Set the axis height. It must be even. Default value is @code{-1} which computes
21648 the axis height automatically.
21651 Set the sonogram height. It must be even. Default value is @code{-1} which
21652 computes the sonogram height automatically.
21655 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21656 instead. Default value is @code{1}.
21658 @item sono_v, volume
21659 Specify the sonogram volume expression. It can contain variables:
21662 the @var{bar_v} evaluated expression
21663 @item frequency, freq, f
21664 the frequency where it is evaluated
21665 @item timeclamp, tc
21666 the value of @var{timeclamp} option
21670 @item a_weighting(f)
21671 A-weighting of equal loudness
21672 @item b_weighting(f)
21673 B-weighting of equal loudness
21674 @item c_weighting(f)
21675 C-weighting of equal loudness.
21677 Default value is @code{16}.
21679 @item bar_v, volume2
21680 Specify the bargraph volume expression. It can contain variables:
21683 the @var{sono_v} evaluated expression
21684 @item frequency, freq, f
21685 the frequency where it is evaluated
21686 @item timeclamp, tc
21687 the value of @var{timeclamp} option
21691 @item a_weighting(f)
21692 A-weighting of equal loudness
21693 @item b_weighting(f)
21694 B-weighting of equal loudness
21695 @item c_weighting(f)
21696 C-weighting of equal loudness.
21698 Default value is @code{sono_v}.
21700 @item sono_g, gamma
21701 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21702 higher gamma makes the spectrum having more range. Default value is @code{3}.
21703 Acceptable range is @code{[1, 7]}.
21705 @item bar_g, gamma2
21706 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21710 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21711 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21713 @item timeclamp, tc
21714 Specify the transform timeclamp. At low frequency, there is trade-off between
21715 accuracy in time domain and frequency domain. If timeclamp is lower,
21716 event in time domain is represented more accurately (such as fast bass drum),
21717 otherwise event in frequency domain is represented more accurately
21718 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21721 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21722 limits future samples by applying asymmetric windowing in time domain, useful
21723 when low latency is required. Accepted range is @code{[0, 1]}.
21726 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21727 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21730 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21731 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21734 This option is deprecated and ignored.
21737 Specify the transform length in time domain. Use this option to control accuracy
21738 trade-off between time domain and frequency domain at every frequency sample.
21739 It can contain variables:
21741 @item frequency, freq, f
21742 the frequency where it is evaluated
21743 @item timeclamp, tc
21744 the value of @var{timeclamp} option.
21746 Default value is @code{384*tc/(384+tc*f)}.
21749 Specify the transform count for every video frame. Default value is @code{6}.
21750 Acceptable range is @code{[1, 30]}.
21753 Specify the transform count for every single pixel. Default value is @code{0},
21754 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21757 Specify font file for use with freetype to draw the axis. If not specified,
21758 use embedded font. Note that drawing with font file or embedded font is not
21759 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21763 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21764 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21767 Specify font color expression. This is arithmetic expression that should return
21768 integer value 0xRRGGBB. It can contain variables:
21770 @item frequency, freq, f
21771 the frequency where it is evaluated
21772 @item timeclamp, tc
21773 the value of @var{timeclamp} option
21778 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21779 @item r(x), g(x), b(x)
21780 red, green, and blue value of intensity x.
21782 Default value is @code{st(0, (midi(f)-59.5)/12);
21783 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21784 r(1-ld(1)) + b(ld(1))}.
21787 Specify image file to draw the axis. This option override @var{fontfile} and
21788 @var{fontcolor} option.
21791 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21792 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21793 Default value is @code{1}.
21796 Set colorspace. The accepted values are:
21799 Unspecified (default)
21808 BT.470BG or BT.601-6 625
21811 SMPTE-170M or BT.601-6 525
21817 BT.2020 with non-constant luminance
21822 Set spectrogram color scheme. This is list of floating point values with format
21823 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21824 The default is @code{1|0.5|0|0|0.5|1}.
21828 @subsection Examples
21832 Playing audio while showing the spectrum:
21834 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21838 Same as above, but with frame rate 30 fps:
21840 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21844 Playing at 1280x720:
21846 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21850 Disable sonogram display:
21856 A1 and its harmonics: A1, A2, (near)E3, A3:
21858 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),
21859 asplit[a][out1]; [a] showcqt [out0]'
21863 Same as above, but with more accuracy in frequency domain:
21865 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),
21866 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21872 bar_v=10:sono_v=bar_v*a_weighting(f)
21876 Custom gamma, now spectrum is linear to the amplitude.
21882 Custom tlength equation:
21884 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)))'
21888 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21890 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21894 Custom font using fontconfig:
21896 font='Courier New,Monospace,mono|bold'
21900 Custom frequency range with custom axis using image file:
21902 axisfile=myaxis.png:basefreq=40:endfreq=10000
21908 Convert input audio to video output representing the audio power spectrum.
21909 Audio amplitude is on Y-axis while frequency is on X-axis.
21911 The filter accepts the following options:
21915 Specify size of video. For the syntax of this option, check the
21916 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21917 Default is @code{1024x512}.
21921 This set how each frequency bin will be represented.
21923 It accepts the following values:
21929 Default is @code{bar}.
21932 Set amplitude scale.
21934 It accepts the following values:
21948 Default is @code{log}.
21951 Set frequency scale.
21953 It accepts the following values:
21962 Reverse logarithmic scale.
21964 Default is @code{lin}.
21969 It accepts the following values:
21985 Default is @code{w2048}
21988 Set windowing function.
21990 It accepts the following values:
22013 Default is @code{hanning}.
22016 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22017 which means optimal overlap for selected window function will be picked.
22020 Set time averaging. Setting this to 0 will display current maximal peaks.
22021 Default is @code{1}, which means time averaging is disabled.
22024 Specify list of colors separated by space or by '|' which will be used to
22025 draw channel frequencies. Unrecognized or missing colors will be replaced
22029 Set channel display mode.
22031 It accepts the following values:
22036 Default is @code{combined}.
22039 Set minimum amplitude used in @code{log} amplitude scaler.
22043 @anchor{showspectrum}
22044 @section showspectrum
22046 Convert input audio to a video output, representing the audio frequency
22049 The filter accepts the following options:
22053 Specify the video size for the output. For the syntax of this option, check the
22054 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22055 Default value is @code{640x512}.
22058 Specify how the spectrum should slide along the window.
22060 It accepts the following values:
22063 the samples start again on the left when they reach the right
22065 the samples scroll from right to left
22067 frames are only produced when the samples reach the right
22069 the samples scroll from left to right
22072 Default value is @code{replace}.
22075 Specify display mode.
22077 It accepts the following values:
22080 all channels are displayed in the same row
22082 all channels are displayed in separate rows
22085 Default value is @samp{combined}.
22088 Specify display color mode.
22090 It accepts the following values:
22093 each channel is displayed in a separate color
22095 each channel is displayed using the same color scheme
22097 each channel is displayed using the rainbow color scheme
22099 each channel is displayed using the moreland color scheme
22101 each channel is displayed using the nebulae color scheme
22103 each channel is displayed using the fire color scheme
22105 each channel is displayed using the fiery color scheme
22107 each channel is displayed using the fruit color scheme
22109 each channel is displayed using the cool color scheme
22111 each channel is displayed using the magma color scheme
22113 each channel is displayed using the green color scheme
22115 each channel is displayed using the viridis color scheme
22117 each channel is displayed using the plasma color scheme
22119 each channel is displayed using the cividis color scheme
22121 each channel is displayed using the terrain color scheme
22124 Default value is @samp{channel}.
22127 Specify scale used for calculating intensity color values.
22129 It accepts the following values:
22134 square root, default
22145 Default value is @samp{sqrt}.
22148 Set saturation modifier for displayed colors. Negative values provide
22149 alternative color scheme. @code{0} is no saturation at all.
22150 Saturation must be in [-10.0, 10.0] range.
22151 Default value is @code{1}.
22154 Set window function.
22156 It accepts the following values:
22181 Default value is @code{hann}.
22184 Set orientation of time vs frequency axis. Can be @code{vertical} or
22185 @code{horizontal}. Default is @code{vertical}.
22188 Set ratio of overlap window. Default value is @code{0}.
22189 When value is @code{1} overlap is set to recommended size for specific
22190 window function currently used.
22193 Set scale gain for calculating intensity color values.
22194 Default value is @code{1}.
22197 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22200 Set color rotation, must be in [-1.0, 1.0] range.
22201 Default value is @code{0}.
22204 Set start frequency from which to display spectrogram. Default is @code{0}.
22207 Set stop frequency to which to display spectrogram. Default is @code{0}.
22210 Set upper frame rate limit. Default is @code{auto}, unlimited.
22213 Draw time and frequency axes and legends. Default is disabled.
22216 The usage is very similar to the showwaves filter; see the examples in that
22219 @subsection Examples
22223 Large window with logarithmic color scaling:
22225 showspectrum=s=1280x480:scale=log
22229 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22231 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22232 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22236 @section showspectrumpic
22238 Convert input audio to a single video frame, representing the audio frequency
22241 The filter accepts the following options:
22245 Specify the video size for the output. For the syntax of this option, check the
22246 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22247 Default value is @code{4096x2048}.
22250 Specify display mode.
22252 It accepts the following values:
22255 all channels are displayed in the same row
22257 all channels are displayed in separate rows
22259 Default value is @samp{combined}.
22262 Specify display color mode.
22264 It accepts the following values:
22267 each channel is displayed in a separate color
22269 each channel is displayed using the same color scheme
22271 each channel is displayed using the rainbow color scheme
22273 each channel is displayed using the moreland color scheme
22275 each channel is displayed using the nebulae color scheme
22277 each channel is displayed using the fire color scheme
22279 each channel is displayed using the fiery color scheme
22281 each channel is displayed using the fruit color scheme
22283 each channel is displayed using the cool color scheme
22285 each channel is displayed using the magma color scheme
22287 each channel is displayed using the green color scheme
22289 each channel is displayed using the viridis color scheme
22291 each channel is displayed using the plasma color scheme
22293 each channel is displayed using the cividis color scheme
22295 each channel is displayed using the terrain color scheme
22297 Default value is @samp{intensity}.
22300 Specify scale used for calculating intensity color values.
22302 It accepts the following values:
22307 square root, default
22317 Default value is @samp{log}.
22320 Set saturation modifier for displayed colors. Negative values provide
22321 alternative color scheme. @code{0} is no saturation at all.
22322 Saturation must be in [-10.0, 10.0] range.
22323 Default value is @code{1}.
22326 Set window function.
22328 It accepts the following values:
22352 Default value is @code{hann}.
22355 Set orientation of time vs frequency axis. Can be @code{vertical} or
22356 @code{horizontal}. Default is @code{vertical}.
22359 Set scale gain for calculating intensity color values.
22360 Default value is @code{1}.
22363 Draw time and frequency axes and legends. Default is enabled.
22366 Set color rotation, must be in [-1.0, 1.0] range.
22367 Default value is @code{0}.
22370 Set start frequency from which to display spectrogram. Default is @code{0}.
22373 Set stop frequency to which to display spectrogram. Default is @code{0}.
22376 @subsection Examples
22380 Extract an audio spectrogram of a whole audio track
22381 in a 1024x1024 picture using @command{ffmpeg}:
22383 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22387 @section showvolume
22389 Convert input audio volume to a video output.
22391 The filter accepts the following options:
22398 Set border width, allowed range is [0, 5]. Default is 1.
22401 Set channel width, allowed range is [80, 8192]. Default is 400.
22404 Set channel height, allowed range is [1, 900]. Default is 20.
22407 Set fade, allowed range is [0, 1]. Default is 0.95.
22410 Set volume color expression.
22412 The expression can use the following variables:
22416 Current max volume of channel in dB.
22422 Current channel number, starting from 0.
22426 If set, displays channel names. Default is enabled.
22429 If set, displays volume values. Default is enabled.
22432 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22433 default is @code{h}.
22436 Set step size, allowed range is [0, 5]. Default is 0, which means
22440 Set background opacity, allowed range is [0, 1]. Default is 0.
22443 Set metering mode, can be peak: @code{p} or rms: @code{r},
22444 default is @code{p}.
22447 Set display scale, can be linear: @code{lin} or log: @code{log},
22448 default is @code{lin}.
22452 If set to > 0., display a line for the max level
22453 in the previous seconds.
22454 default is disabled: @code{0.}
22457 The color of the max line. Use when @code{dm} option is set to > 0.
22458 default is: @code{orange}
22463 Convert input audio to a video output, representing the samples waves.
22465 The filter accepts the following options:
22469 Specify the video size for the output. For the syntax of this option, check the
22470 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22471 Default value is @code{600x240}.
22476 Available values are:
22479 Draw a point for each sample.
22482 Draw a vertical line for each sample.
22485 Draw a point for each sample and a line between them.
22488 Draw a centered vertical line for each sample.
22491 Default value is @code{point}.
22494 Set the number of samples which are printed on the same column. A
22495 larger value will decrease the frame rate. Must be a positive
22496 integer. This option can be set only if the value for @var{rate}
22497 is not explicitly specified.
22500 Set the (approximate) output frame rate. This is done by setting the
22501 option @var{n}. Default value is "25".
22503 @item split_channels
22504 Set if channels should be drawn separately or overlap. Default value is 0.
22507 Set colors separated by '|' which are going to be used for drawing of each channel.
22510 Set amplitude scale.
22512 Available values are:
22530 Set the draw mode. This is mostly useful to set for high @var{n}.
22532 Available values are:
22535 Scale pixel values for each drawn sample.
22538 Draw every sample directly.
22541 Default value is @code{scale}.
22544 @subsection Examples
22548 Output the input file audio and the corresponding video representation
22551 amovie=a.mp3,asplit[out0],showwaves[out1]
22555 Create a synthetic signal and show it with showwaves, forcing a
22556 frame rate of 30 frames per second:
22558 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22562 @section showwavespic
22564 Convert input audio to a single video frame, representing the samples waves.
22566 The filter accepts the following options:
22570 Specify the video size for the output. For the syntax of this option, check the
22571 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22572 Default value is @code{600x240}.
22574 @item split_channels
22575 Set if channels should be drawn separately or overlap. Default value is 0.
22578 Set colors separated by '|' which are going to be used for drawing of each channel.
22581 Set amplitude scale.
22583 Available values are:
22601 @subsection Examples
22605 Extract a channel split representation of the wave form of a whole audio track
22606 in a 1024x800 picture using @command{ffmpeg}:
22608 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22612 @section sidedata, asidedata
22614 Delete frame side data, or select frames based on it.
22616 This filter accepts the following options:
22620 Set mode of operation of the filter.
22622 Can be one of the following:
22626 Select every frame with side data of @code{type}.
22629 Delete side data of @code{type}. If @code{type} is not set, delete all side
22635 Set side data type used with all modes. Must be set for @code{select} mode. For
22636 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22637 in @file{libavutil/frame.h}. For example, to choose
22638 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22642 @section spectrumsynth
22644 Sythesize audio from 2 input video spectrums, first input stream represents
22645 magnitude across time and second represents phase across time.
22646 The filter will transform from frequency domain as displayed in videos back
22647 to time domain as presented in audio output.
22649 This filter is primarily created for reversing processed @ref{showspectrum}
22650 filter outputs, but can synthesize sound from other spectrograms too.
22651 But in such case results are going to be poor if the phase data is not
22652 available, because in such cases phase data need to be recreated, usually
22653 it's just recreated from random noise.
22654 For best results use gray only output (@code{channel} color mode in
22655 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22656 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22657 @code{data} option. Inputs videos should generally use @code{fullframe}
22658 slide mode as that saves resources needed for decoding video.
22660 The filter accepts the following options:
22664 Specify sample rate of output audio, the sample rate of audio from which
22665 spectrum was generated may differ.
22668 Set number of channels represented in input video spectrums.
22671 Set scale which was used when generating magnitude input spectrum.
22672 Can be @code{lin} or @code{log}. Default is @code{log}.
22675 Set slide which was used when generating inputs spectrums.
22676 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22677 Default is @code{fullframe}.
22680 Set window function used for resynthesis.
22683 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22684 which means optimal overlap for selected window function will be picked.
22687 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22688 Default is @code{vertical}.
22691 @subsection Examples
22695 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22696 then resynthesize videos back to audio with spectrumsynth:
22698 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
22699 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
22700 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22704 @section split, asplit
22706 Split input into several identical outputs.
22708 @code{asplit} works with audio input, @code{split} with video.
22710 The filter accepts a single parameter which specifies the number of outputs. If
22711 unspecified, it defaults to 2.
22713 @subsection Examples
22717 Create two separate outputs from the same input:
22719 [in] split [out0][out1]
22723 To create 3 or more outputs, you need to specify the number of
22726 [in] asplit=3 [out0][out1][out2]
22730 Create two separate outputs from the same input, one cropped and
22733 [in] split [splitout1][splitout2];
22734 [splitout1] crop=100:100:0:0 [cropout];
22735 [splitout2] pad=200:200:100:100 [padout];
22739 Create 5 copies of the input audio with @command{ffmpeg}:
22741 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22747 Receive commands sent through a libzmq client, and forward them to
22748 filters in the filtergraph.
22750 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22751 must be inserted between two video filters, @code{azmq} between two
22752 audio filters. Both are capable to send messages to any filter type.
22754 To enable these filters you need to install the libzmq library and
22755 headers and configure FFmpeg with @code{--enable-libzmq}.
22757 For more information about libzmq see:
22758 @url{http://www.zeromq.org/}
22760 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22761 receives messages sent through a network interface defined by the
22762 @option{bind_address} (or the abbreviation "@option{b}") option.
22763 Default value of this option is @file{tcp://localhost:5555}. You may
22764 want to alter this value to your needs, but do not forget to escape any
22765 ':' signs (see @ref{filtergraph escaping}).
22767 The received message must be in the form:
22769 @var{TARGET} @var{COMMAND} [@var{ARG}]
22772 @var{TARGET} specifies the target of the command, usually the name of
22773 the filter class or a specific filter instance name. The default
22774 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22775 but you can override this by using the @samp{filter_name@@id} syntax
22776 (see @ref{Filtergraph syntax}).
22778 @var{COMMAND} specifies the name of the command for the target filter.
22780 @var{ARG} is optional and specifies the optional argument list for the
22781 given @var{COMMAND}.
22783 Upon reception, the message is processed and the corresponding command
22784 is injected into the filtergraph. Depending on the result, the filter
22785 will send a reply to the client, adopting the format:
22787 @var{ERROR_CODE} @var{ERROR_REASON}
22791 @var{MESSAGE} is optional.
22793 @subsection Examples
22795 Look at @file{tools/zmqsend} for an example of a zmq client which can
22796 be used to send commands processed by these filters.
22798 Consider the following filtergraph generated by @command{ffplay}.
22799 In this example the last overlay filter has an instance name. All other
22800 filters will have default instance names.
22803 ffplay -dumpgraph 1 -f lavfi "
22804 color=s=100x100:c=red [l];
22805 color=s=100x100:c=blue [r];
22806 nullsrc=s=200x100, zmq [bg];
22807 [bg][l] overlay [bg+l];
22808 [bg+l][r] overlay@@my=x=100 "
22811 To change the color of the left side of the video, the following
22812 command can be used:
22814 echo Parsed_color_0 c yellow | tools/zmqsend
22817 To change the right side:
22819 echo Parsed_color_1 c pink | tools/zmqsend
22822 To change the position of the right side:
22824 echo overlay@@my x 150 | tools/zmqsend
22828 @c man end MULTIMEDIA FILTERS
22830 @chapter Multimedia Sources
22831 @c man begin MULTIMEDIA SOURCES
22833 Below is a description of the currently available multimedia sources.
22837 This is the same as @ref{movie} source, except it selects an audio
22843 Read audio and/or video stream(s) from a movie container.
22845 It accepts the following parameters:
22849 The name of the resource to read (not necessarily a file; it can also be a
22850 device or a stream accessed through some protocol).
22852 @item format_name, f
22853 Specifies the format assumed for the movie to read, and can be either
22854 the name of a container or an input device. If not specified, the
22855 format is guessed from @var{movie_name} or by probing.
22857 @item seek_point, sp
22858 Specifies the seek point in seconds. The frames will be output
22859 starting from this seek point. The parameter is evaluated with
22860 @code{av_strtod}, so the numerical value may be suffixed by an IS
22861 postfix. The default value is "0".
22864 Specifies the streams to read. Several streams can be specified,
22865 separated by "+". The source will then have as many outputs, in the
22866 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22867 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22868 respectively the default (best suited) video and audio stream. Default
22869 is "dv", or "da" if the filter is called as "amovie".
22871 @item stream_index, si
22872 Specifies the index of the video stream to read. If the value is -1,
22873 the most suitable video stream will be automatically selected. The default
22874 value is "-1". Deprecated. If the filter is called "amovie", it will select
22875 audio instead of video.
22878 Specifies how many times to read the stream in sequence.
22879 If the value is 0, the stream will be looped infinitely.
22880 Default value is "1".
22882 Note that when the movie is looped the source timestamps are not
22883 changed, so it will generate non monotonically increasing timestamps.
22885 @item discontinuity
22886 Specifies the time difference between frames above which the point is
22887 considered a timestamp discontinuity which is removed by adjusting the later
22891 It allows overlaying a second video on top of the main input of
22892 a filtergraph, as shown in this graph:
22894 input -----------> deltapts0 --> overlay --> output
22897 movie --> scale--> deltapts1 -------+
22899 @subsection Examples
22903 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
22904 on top of the input labelled "in":
22906 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
22907 [in] setpts=PTS-STARTPTS [main];
22908 [main][over] overlay=16:16 [out]
22912 Read from a video4linux2 device, and overlay it on top of the input
22915 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
22916 [in] setpts=PTS-STARTPTS [main];
22917 [main][over] overlay=16:16 [out]
22921 Read the first video stream and the audio stream with id 0x81 from
22922 dvd.vob; the video is connected to the pad named "video" and the audio is
22923 connected to the pad named "audio":
22925 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
22929 @subsection Commands
22931 Both movie and amovie support the following commands:
22934 Perform seek using "av_seek_frame".
22935 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
22938 @var{stream_index}: If stream_index is -1, a default
22939 stream is selected, and @var{timestamp} is automatically converted
22940 from AV_TIME_BASE units to the stream specific time_base.
22942 @var{timestamp}: Timestamp in AVStream.time_base units
22943 or, if no stream is specified, in AV_TIME_BASE units.
22945 @var{flags}: Flags which select direction and seeking mode.
22949 Get movie duration in AV_TIME_BASE units.
22953 @c man end MULTIMEDIA SOURCES