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.
2108 Apply audio soft clipping.
2110 Soft clipping is a type of distortion effect where the amplitude of a signal is saturated
2111 along a smooth curve, rather than the abrupt shape of hard-clipping.
2113 This filter accepts the following options:
2117 Set type of soft-clipping.
2119 It accepts the following values:
2131 Set additional parameter which controls sigmoid function.
2137 Display time domain statistical information about the audio channels.
2138 Statistics are calculated and displayed for each audio channel and,
2139 where applicable, an overall figure is also given.
2141 It accepts the following option:
2144 Short window length in seconds, used for peak and trough RMS measurement.
2145 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2149 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2150 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2153 Available keys for each channel are:
2195 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2196 this @code{lavfi.astats.Overall.Peak_count}.
2198 For description what each key means read below.
2201 Set number of frame after which stats are going to be recalculated.
2202 Default is disabled.
2204 @item measure_perchannel
2205 Select the entries which need to be measured per channel. The metadata keys can
2206 be used as flags, default is @option{all} which measures everything.
2207 @option{none} disables all per channel measurement.
2209 @item measure_overall
2210 Select the entries which need to be measured overall. The metadata keys can
2211 be used as flags, default is @option{all} which measures everything.
2212 @option{none} disables all overall measurement.
2216 A description of each shown parameter follows:
2220 Mean amplitude displacement from zero.
2223 Minimal sample level.
2226 Maximal sample level.
2228 @item Min difference
2229 Minimal difference between two consecutive samples.
2231 @item Max difference
2232 Maximal difference between two consecutive samples.
2234 @item Mean difference
2235 Mean difference between two consecutive samples.
2236 The average of each difference between two consecutive samples.
2238 @item RMS difference
2239 Root Mean Square difference between two consecutive samples.
2243 Standard peak and RMS level measured in dBFS.
2247 Peak and trough values for RMS level measured over a short window.
2250 Standard ratio of peak to RMS level (note: not in dB).
2253 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2254 (i.e. either @var{Min level} or @var{Max level}).
2257 Number of occasions (not the number of samples) that the signal attained either
2258 @var{Min level} or @var{Max level}.
2261 Overall bit depth of audio. Number of bits used for each sample.
2264 Measured dynamic range of audio in dB.
2266 @item Zero crossings
2267 Number of points where the waveform crosses the zero level axis.
2269 @item Zero crossings rate
2270 Rate of Zero crossings and number of audio samples.
2277 The filter accepts exactly one parameter, the audio tempo. If not
2278 specified then the filter will assume nominal 1.0 tempo. Tempo must
2279 be in the [0.5, 100.0] range.
2281 Note that tempo greater than 2 will skip some samples rather than
2282 blend them in. If for any reason this is a concern it is always
2283 possible to daisy-chain several instances of atempo to achieve the
2284 desired product tempo.
2286 @subsection Examples
2290 Slow down audio to 80% tempo:
2296 To speed up audio to 300% tempo:
2302 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2304 atempo=sqrt(3),atempo=sqrt(3)
2310 Trim the input so that the output contains one continuous subpart of the input.
2312 It accepts the following parameters:
2315 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2316 sample with the timestamp @var{start} will be the first sample in the output.
2319 Specify time of the first audio sample that will be dropped, i.e. the
2320 audio sample immediately preceding the one with the timestamp @var{end} will be
2321 the last sample in the output.
2324 Same as @var{start}, except this option sets the start timestamp in samples
2328 Same as @var{end}, except this option sets the end timestamp in samples instead
2332 The maximum duration of the output in seconds.
2335 The number of the first sample that should be output.
2338 The number of the first sample that should be dropped.
2341 @option{start}, @option{end}, and @option{duration} are expressed as time
2342 duration specifications; see
2343 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2345 Note that the first two sets of the start/end options and the @option{duration}
2346 option look at the frame timestamp, while the _sample options simply count the
2347 samples that pass through the filter. So start/end_pts and start/end_sample will
2348 give different results when the timestamps are wrong, inexact or do not start at
2349 zero. Also note that this filter does not modify the timestamps. If you wish
2350 to have the output timestamps start at zero, insert the asetpts filter after the
2353 If multiple start or end options are set, this filter tries to be greedy and
2354 keep all samples that match at least one of the specified constraints. To keep
2355 only the part that matches all the constraints at once, chain multiple atrim
2358 The defaults are such that all the input is kept. So it is possible to set e.g.
2359 just the end values to keep everything before the specified time.
2364 Drop everything except the second minute of input:
2366 ffmpeg -i INPUT -af atrim=60:120
2370 Keep only the first 1000 samples:
2372 ffmpeg -i INPUT -af atrim=end_sample=1000
2379 Apply a two-pole Butterworth band-pass filter with central
2380 frequency @var{frequency}, and (3dB-point) band-width width.
2381 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2382 instead of the default: constant 0dB peak gain.
2383 The filter roll off at 6dB per octave (20dB per decade).
2385 The filter accepts the following options:
2389 Set the filter's central frequency. Default is @code{3000}.
2392 Constant skirt gain if set to 1. Defaults to 0.
2395 Set method to specify band-width of filter.
2410 Specify the band-width of a filter in width_type units.
2413 Specify which channels to filter, by default all available are filtered.
2416 @subsection Commands
2418 This filter supports the following commands:
2421 Change bandpass frequency.
2422 Syntax for the command is : "@var{frequency}"
2425 Change bandpass width_type.
2426 Syntax for the command is : "@var{width_type}"
2429 Change bandpass width.
2430 Syntax for the command is : "@var{width}"
2435 Apply a two-pole Butterworth band-reject filter with central
2436 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2437 The filter roll off at 6dB per octave (20dB per decade).
2439 The filter accepts the following options:
2443 Set the filter's central frequency. Default is @code{3000}.
2446 Set method to specify band-width of filter.
2461 Specify the band-width of a filter in width_type units.
2464 Specify which channels to filter, by default all available are filtered.
2467 @subsection Commands
2469 This filter supports the following commands:
2472 Change bandreject frequency.
2473 Syntax for the command is : "@var{frequency}"
2476 Change bandreject width_type.
2477 Syntax for the command is : "@var{width_type}"
2480 Change bandreject width.
2481 Syntax for the command is : "@var{width}"
2484 @section bass, lowshelf
2486 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2487 shelving filter with a response similar to that of a standard
2488 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2490 The filter accepts the following options:
2494 Give the gain at 0 Hz. Its useful range is about -20
2495 (for a large cut) to +20 (for a large boost).
2496 Beware of clipping when using a positive gain.
2499 Set the filter's central frequency and so can be used
2500 to extend or reduce the frequency range to be boosted or cut.
2501 The default value is @code{100} Hz.
2504 Set method to specify band-width of filter.
2519 Determine how steep is the filter's shelf transition.
2522 Specify which channels to filter, by default all available are filtered.
2525 @subsection Commands
2527 This filter supports the following commands:
2530 Change bass frequency.
2531 Syntax for the command is : "@var{frequency}"
2534 Change bass width_type.
2535 Syntax for the command is : "@var{width_type}"
2539 Syntax for the command is : "@var{width}"
2543 Syntax for the command is : "@var{gain}"
2548 Apply a biquad IIR filter with the given coefficients.
2549 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2550 are the numerator and denominator coefficients respectively.
2551 and @var{channels}, @var{c} specify which channels to filter, by default all
2552 available are filtered.
2554 @subsection Commands
2556 This filter supports the following commands:
2564 Change biquad parameter.
2565 Syntax for the command is : "@var{value}"
2569 Bauer stereo to binaural transformation, which improves headphone listening of
2570 stereo audio records.
2572 To enable compilation of this filter you need to configure FFmpeg with
2573 @code{--enable-libbs2b}.
2575 It accepts the following parameters:
2579 Pre-defined crossfeed level.
2583 Default level (fcut=700, feed=50).
2586 Chu Moy circuit (fcut=700, feed=60).
2589 Jan Meier circuit (fcut=650, feed=95).
2594 Cut frequency (in Hz).
2603 Remap input channels to new locations.
2605 It accepts the following parameters:
2608 Map channels from input to output. The argument is a '|'-separated list of
2609 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2610 @var{in_channel} form. @var{in_channel} can be either the name of the input
2611 channel (e.g. FL for front left) or its index in the input channel layout.
2612 @var{out_channel} is the name of the output channel or its index in the output
2613 channel layout. If @var{out_channel} is not given then it is implicitly an
2614 index, starting with zero and increasing by one for each mapping.
2616 @item channel_layout
2617 The channel layout of the output stream.
2620 If no mapping is present, the filter will implicitly map input channels to
2621 output channels, preserving indices.
2623 @subsection Examples
2627 For example, assuming a 5.1+downmix input MOV file,
2629 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2631 will create an output WAV file tagged as stereo from the downmix channels of
2635 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2637 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2641 @section channelsplit
2643 Split each channel from an input audio stream into a separate output stream.
2645 It accepts the following parameters:
2647 @item channel_layout
2648 The channel layout of the input stream. The default is "stereo".
2650 A channel layout describing the channels to be extracted as separate output streams
2651 or "all" to extract each input channel as a separate stream. The default is "all".
2653 Choosing channels not present in channel layout in the input will result in an error.
2656 @subsection Examples
2660 For example, assuming a stereo input MP3 file,
2662 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2664 will create an output Matroska file with two audio streams, one containing only
2665 the left channel and the other the right channel.
2668 Split a 5.1 WAV file into per-channel files:
2670 ffmpeg -i in.wav -filter_complex
2671 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2672 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2673 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2678 Extract only LFE from a 5.1 WAV file:
2680 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2681 -map '[LFE]' lfe.wav
2686 Add a chorus effect to the audio.
2688 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2690 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2691 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2692 The modulation depth defines the range the modulated delay is played before or after
2693 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2694 sound tuned around the original one, like in a chorus where some vocals are slightly
2697 It accepts the following parameters:
2700 Set input gain. Default is 0.4.
2703 Set output gain. Default is 0.4.
2706 Set delays. A typical delay is around 40ms to 60ms.
2718 @subsection Examples
2724 chorus=0.7:0.9:55:0.4:0.25:2
2730 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2734 Fuller sounding chorus with three delays:
2736 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
2741 Compress or expand the audio's dynamic range.
2743 It accepts the following parameters:
2749 A list of times in seconds for each channel over which the instantaneous level
2750 of the input signal is averaged to determine its volume. @var{attacks} refers to
2751 increase of volume and @var{decays} refers to decrease of volume. For most
2752 situations, the attack time (response to the audio getting louder) should be
2753 shorter than the decay time, because the human ear is more sensitive to sudden
2754 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2755 a typical value for decay is 0.8 seconds.
2756 If specified number of attacks & decays is lower than number of channels, the last
2757 set attack/decay will be used for all remaining channels.
2760 A list of points for the transfer function, specified in dB relative to the
2761 maximum possible signal amplitude. Each key points list must be defined using
2762 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2763 @code{x0/y0 x1/y1 x2/y2 ....}
2765 The input values must be in strictly increasing order but the transfer function
2766 does not have to be monotonically rising. The point @code{0/0} is assumed but
2767 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2768 function are @code{-70/-70|-60/-20|1/0}.
2771 Set the curve radius in dB for all joints. It defaults to 0.01.
2774 Set the additional gain in dB to be applied at all points on the transfer
2775 function. This allows for easy adjustment of the overall gain.
2779 Set an initial volume, in dB, to be assumed for each channel when filtering
2780 starts. This permits the user to supply a nominal level initially, so that, for
2781 example, a very large gain is not applied to initial signal levels before the
2782 companding has begun to operate. A typical value for audio which is initially
2783 quiet is -90 dB. It defaults to 0.
2786 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2787 delayed before being fed to the volume adjuster. Specifying a delay
2788 approximately equal to the attack/decay times allows the filter to effectively
2789 operate in predictive rather than reactive mode. It defaults to 0.
2793 @subsection Examples
2797 Make music with both quiet and loud passages suitable for listening to in a
2800 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2803 Another example for audio with whisper and explosion parts:
2805 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2809 A noise gate for when the noise is at a lower level than the signal:
2811 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2815 Here is another noise gate, this time for when the noise is at a higher level
2816 than the signal (making it, in some ways, similar to squelch):
2818 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2822 2:1 compression starting at -6dB:
2824 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2828 2:1 compression starting at -9dB:
2830 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2834 2:1 compression starting at -12dB:
2836 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2840 2:1 compression starting at -18dB:
2842 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2846 3:1 compression starting at -15dB:
2848 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2854 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2860 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
2864 Hard limiter at -6dB:
2866 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2870 Hard limiter at -12dB:
2872 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2876 Hard noise gate at -35 dB:
2878 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2884 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2888 @section compensationdelay
2890 Compensation Delay Line is a metric based delay to compensate differing
2891 positions of microphones or speakers.
2893 For example, you have recorded guitar with two microphones placed in
2894 different location. Because the front of sound wave has fixed speed in
2895 normal conditions, the phasing of microphones can vary and depends on
2896 their location and interposition. The best sound mix can be achieved when
2897 these microphones are in phase (synchronized). Note that distance of
2898 ~30 cm between microphones makes one microphone to capture signal in
2899 antiphase to another microphone. That makes the final mix sounding moody.
2900 This filter helps to solve phasing problems by adding different delays
2901 to each microphone track and make them synchronized.
2903 The best result can be reached when you take one track as base and
2904 synchronize other tracks one by one with it.
2905 Remember that synchronization/delay tolerance depends on sample rate, too.
2906 Higher sample rates will give more tolerance.
2908 It accepts the following parameters:
2912 Set millimeters distance. This is compensation distance for fine tuning.
2916 Set cm distance. This is compensation distance for tightening distance setup.
2920 Set meters distance. This is compensation distance for hard distance setup.
2924 Set dry amount. Amount of unprocessed (dry) signal.
2928 Set wet amount. Amount of processed (wet) signal.
2932 Set temperature degree in Celsius. This is the temperature of the environment.
2937 Apply headphone crossfeed filter.
2939 Crossfeed is the process of blending the left and right channels of stereo
2941 It is mainly used to reduce extreme stereo separation of low frequencies.
2943 The intent is to produce more speaker like sound to the listener.
2945 The filter accepts the following options:
2949 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2950 This sets gain of low shelf filter for side part of stereo image.
2951 Default is -6dB. Max allowed is -30db when strength is set to 1.
2954 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2955 This sets cut off frequency of low shelf filter. Default is cut off near
2956 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2959 Set input gain. Default is 0.9.
2962 Set output gain. Default is 1.
2965 @section crystalizer
2966 Simple algorithm to expand audio dynamic range.
2968 The filter accepts the following options:
2972 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2973 (unchanged sound) to 10.0 (maximum effect).
2976 Enable clipping. By default is enabled.
2980 Apply a DC shift to the audio.
2982 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2983 in the recording chain) from the audio. The effect of a DC offset is reduced
2984 headroom and hence volume. The @ref{astats} filter can be used to determine if
2985 a signal has a DC offset.
2989 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2993 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2994 used to prevent clipping.
2998 Measure audio dynamic range.
3000 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
3001 is found in transition material. And anything less that 8 have very poor dynamics
3002 and is very compressed.
3004 The filter accepts the following options:
3008 Set window length in seconds used to split audio into segments of equal length.
3009 Default is 3 seconds.
3013 Dynamic Audio Normalizer.
3015 This filter applies a certain amount of gain to the input audio in order
3016 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
3017 contrast to more "simple" normalization algorithms, the Dynamic Audio
3018 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
3019 This allows for applying extra gain to the "quiet" sections of the audio
3020 while avoiding distortions or clipping the "loud" sections. In other words:
3021 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
3022 sections, in the sense that the volume of each section is brought to the
3023 same target level. Note, however, that the Dynamic Audio Normalizer achieves
3024 this goal *without* applying "dynamic range compressing". It will retain 100%
3025 of the dynamic range *within* each section of the audio file.
3029 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
3030 Default is 500 milliseconds.
3031 The Dynamic Audio Normalizer processes the input audio in small chunks,
3032 referred to as frames. This is required, because a peak magnitude has no
3033 meaning for just a single sample value. Instead, we need to determine the
3034 peak magnitude for a contiguous sequence of sample values. While a "standard"
3035 normalizer would simply use the peak magnitude of the complete file, the
3036 Dynamic Audio Normalizer determines the peak magnitude individually for each
3037 frame. The length of a frame is specified in milliseconds. By default, the
3038 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3039 been found to give good results with most files.
3040 Note that the exact frame length, in number of samples, will be determined
3041 automatically, based on the sampling rate of the individual input audio file.
3044 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3045 number. Default is 31.
3046 Probably the most important parameter of the Dynamic Audio Normalizer is the
3047 @code{window size} of the Gaussian smoothing filter. The filter's window size
3048 is specified in frames, centered around the current frame. For the sake of
3049 simplicity, this must be an odd number. Consequently, the default value of 31
3050 takes into account the current frame, as well as the 15 preceding frames and
3051 the 15 subsequent frames. Using a larger window results in a stronger
3052 smoothing effect and thus in less gain variation, i.e. slower gain
3053 adaptation. Conversely, using a smaller window results in a weaker smoothing
3054 effect and thus in more gain variation, i.e. faster gain adaptation.
3055 In other words, the more you increase this value, the more the Dynamic Audio
3056 Normalizer will behave like a "traditional" normalization filter. On the
3057 contrary, the more you decrease this value, the more the Dynamic Audio
3058 Normalizer will behave like a dynamic range compressor.
3061 Set the target peak value. This specifies the highest permissible magnitude
3062 level for the normalized audio input. This filter will try to approach the
3063 target peak magnitude as closely as possible, but at the same time it also
3064 makes sure that the normalized signal will never exceed the peak magnitude.
3065 A frame's maximum local gain factor is imposed directly by the target peak
3066 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3067 It is not recommended to go above this value.
3070 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3071 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3072 factor for each input frame, i.e. the maximum gain factor that does not
3073 result in clipping or distortion. The maximum gain factor is determined by
3074 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3075 additionally bounds the frame's maximum gain factor by a predetermined
3076 (global) maximum gain factor. This is done in order to avoid excessive gain
3077 factors in "silent" or almost silent frames. By default, the maximum gain
3078 factor is 10.0, For most inputs the default value should be sufficient and
3079 it usually is not recommended to increase this value. Though, for input
3080 with an extremely low overall volume level, it may be necessary to allow even
3081 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3082 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3083 Instead, a "sigmoid" threshold function will be applied. This way, the
3084 gain factors will smoothly approach the threshold value, but never exceed that
3088 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3089 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3090 This means that the maximum local gain factor for each frame is defined
3091 (only) by the frame's highest magnitude sample. This way, the samples can
3092 be amplified as much as possible without exceeding the maximum signal
3093 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3094 Normalizer can also take into account the frame's root mean square,
3095 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3096 determine the power of a time-varying signal. It is therefore considered
3097 that the RMS is a better approximation of the "perceived loudness" than
3098 just looking at the signal's peak magnitude. Consequently, by adjusting all
3099 frames to a constant RMS value, a uniform "perceived loudness" can be
3100 established. If a target RMS value has been specified, a frame's local gain
3101 factor is defined as the factor that would result in exactly that RMS value.
3102 Note, however, that the maximum local gain factor is still restricted by the
3103 frame's highest magnitude sample, in order to prevent clipping.
3106 Enable channels coupling. By default is enabled.
3107 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3108 amount. This means the same gain factor will be applied to all channels, i.e.
3109 the maximum possible gain factor is determined by the "loudest" channel.
3110 However, in some recordings, it may happen that the volume of the different
3111 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3112 In this case, this option can be used to disable the channel coupling. This way,
3113 the gain factor will be determined independently for each channel, depending
3114 only on the individual channel's highest magnitude sample. This allows for
3115 harmonizing the volume of the different channels.
3118 Enable DC bias correction. By default is disabled.
3119 An audio signal (in the time domain) is a sequence of sample values.
3120 In the Dynamic Audio Normalizer these sample values are represented in the
3121 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3122 audio signal, or "waveform", should be centered around the zero point.
3123 That means if we calculate the mean value of all samples in a file, or in a
3124 single frame, then the result should be 0.0 or at least very close to that
3125 value. If, however, there is a significant deviation of the mean value from
3126 0.0, in either positive or negative direction, this is referred to as a
3127 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3128 Audio Normalizer provides optional DC bias correction.
3129 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3130 the mean value, or "DC correction" offset, of each input frame and subtract
3131 that value from all of the frame's sample values which ensures those samples
3132 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3133 boundaries, the DC correction offset values will be interpolated smoothly
3134 between neighbouring frames.
3137 Enable alternative boundary mode. By default is disabled.
3138 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3139 around each frame. This includes the preceding frames as well as the
3140 subsequent frames. However, for the "boundary" frames, located at the very
3141 beginning and at the very end of the audio file, not all neighbouring
3142 frames are available. In particular, for the first few frames in the audio
3143 file, the preceding frames are not known. And, similarly, for the last few
3144 frames in the audio file, the subsequent frames are not known. Thus, the
3145 question arises which gain factors should be assumed for the missing frames
3146 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3147 to deal with this situation. The default boundary mode assumes a gain factor
3148 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3149 "fade out" at the beginning and at the end of the input, respectively.
3152 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3153 By default, the Dynamic Audio Normalizer does not apply "traditional"
3154 compression. This means that signal peaks will not be pruned and thus the
3155 full dynamic range will be retained within each local neighbourhood. However,
3156 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3157 normalization algorithm with a more "traditional" compression.
3158 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3159 (thresholding) function. If (and only if) the compression feature is enabled,
3160 all input frames will be processed by a soft knee thresholding function prior
3161 to the actual normalization process. Put simply, the thresholding function is
3162 going to prune all samples whose magnitude exceeds a certain threshold value.
3163 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3164 value. Instead, the threshold value will be adjusted for each individual
3166 In general, smaller parameters result in stronger compression, and vice versa.
3167 Values below 3.0 are not recommended, because audible distortion may appear.
3172 Make audio easier to listen to on headphones.
3174 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3175 so that when listened to on headphones the stereo image is moved from
3176 inside your head (standard for headphones) to outside and in front of
3177 the listener (standard for speakers).
3183 Apply a two-pole peaking equalisation (EQ) filter. With this
3184 filter, the signal-level at and around a selected frequency can
3185 be increased or decreased, whilst (unlike bandpass and bandreject
3186 filters) that at all other frequencies is unchanged.
3188 In order to produce complex equalisation curves, this filter can
3189 be given several times, each with a different central frequency.
3191 The filter accepts the following options:
3195 Set the filter's central frequency in Hz.
3198 Set method to specify band-width of filter.
3213 Specify the band-width of a filter in width_type units.
3216 Set the required gain or attenuation in dB.
3217 Beware of clipping when using a positive gain.
3220 Specify which channels to filter, by default all available are filtered.
3223 @subsection Examples
3226 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3228 equalizer=f=1000:t=h:width=200:g=-10
3232 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3234 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3238 @subsection Commands
3240 This filter supports the following commands:
3243 Change equalizer frequency.
3244 Syntax for the command is : "@var{frequency}"
3247 Change equalizer width_type.
3248 Syntax for the command is : "@var{width_type}"
3251 Change equalizer width.
3252 Syntax for the command is : "@var{width}"
3255 Change equalizer gain.
3256 Syntax for the command is : "@var{gain}"
3259 @section extrastereo
3261 Linearly increases the difference between left and right channels which
3262 adds some sort of "live" effect to playback.
3264 The filter accepts the following options:
3268 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3269 (average of both channels), with 1.0 sound will be unchanged, with
3270 -1.0 left and right channels will be swapped.
3273 Enable clipping. By default is enabled.
3276 @section firequalizer
3277 Apply FIR Equalization using arbitrary frequency response.
3279 The filter accepts the following option:
3283 Set gain curve equation (in dB). The expression can contain variables:
3286 the evaluated frequency
3290 channel number, set to 0 when multichannels evaluation is disabled
3292 channel id, see libavutil/channel_layout.h, set to the first channel id when
3293 multichannels evaluation is disabled
3297 channel_layout, see libavutil/channel_layout.h
3302 @item gain_interpolate(f)
3303 interpolate gain on frequency f based on gain_entry
3304 @item cubic_interpolate(f)
3305 same as gain_interpolate, but smoother
3307 This option is also available as command. Default is @code{gain_interpolate(f)}.
3310 Set gain entry for gain_interpolate function. The expression can
3314 store gain entry at frequency f with value g
3316 This option is also available as command.
3319 Set filter delay in seconds. Higher value means more accurate.
3320 Default is @code{0.01}.
3323 Set filter accuracy in Hz. Lower value means more accurate.
3324 Default is @code{5}.
3327 Set window function. Acceptable values are:
3330 rectangular window, useful when gain curve is already smooth
3332 hann window (default)
3338 3-terms continuous 1st derivative nuttall window
3340 minimum 3-terms discontinuous nuttall window
3342 4-terms continuous 1st derivative nuttall window
3344 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3346 blackman-harris window
3352 If enabled, use fixed number of audio samples. This improves speed when
3353 filtering with large delay. Default is disabled.
3356 Enable multichannels evaluation on gain. Default is disabled.
3359 Enable zero phase mode by subtracting timestamp to compensate delay.
3360 Default is disabled.
3363 Set scale used by gain. Acceptable values are:
3366 linear frequency, linear gain
3368 linear frequency, logarithmic (in dB) gain (default)
3370 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3372 logarithmic frequency, logarithmic gain
3376 Set file for dumping, suitable for gnuplot.
3379 Set scale for dumpfile. Acceptable values are same with scale option.
3383 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3384 Default is disabled.
3387 Enable minimum phase impulse response. Default is disabled.
3390 @subsection Examples
3395 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3398 lowpass at 1000 Hz with gain_entry:
3400 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3403 custom equalization:
3405 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3408 higher delay with zero phase to compensate delay:
3410 firequalizer=delay=0.1:fixed=on:zero_phase=on
3413 lowpass on left channel, highpass on right channel:
3415 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3416 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3421 Apply a flanging effect to the audio.
3423 The filter accepts the following options:
3427 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3430 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3433 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3437 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3438 Default value is 71.
3441 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3444 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3445 Default value is @var{sinusoidal}.
3448 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3449 Default value is 25.
3452 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3453 Default is @var{linear}.
3457 Apply Haas effect to audio.
3459 Note that this makes most sense to apply on mono signals.
3460 With this filter applied to mono signals it give some directionality and
3461 stretches its stereo image.
3463 The filter accepts the following options:
3467 Set input level. By default is @var{1}, or 0dB
3470 Set output level. By default is @var{1}, or 0dB.
3473 Set gain applied to side part of signal. By default is @var{1}.
3476 Set kind of middle source. Can be one of the following:
3486 Pick middle part signal of stereo image.
3489 Pick side part signal of stereo image.
3493 Change middle phase. By default is disabled.
3496 Set left channel delay. By default is @var{2.05} milliseconds.
3499 Set left channel balance. By default is @var{-1}.
3502 Set left channel gain. By default is @var{1}.
3505 Change left phase. By default is disabled.
3508 Set right channel delay. By defaults is @var{2.12} milliseconds.
3511 Set right channel balance. By default is @var{1}.
3514 Set right channel gain. By default is @var{1}.
3517 Change right phase. By default is enabled.
3522 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3523 embedded HDCD codes is expanded into a 20-bit PCM stream.
3525 The filter supports the Peak Extend and Low-level Gain Adjustment features
3526 of HDCD, and detects the Transient Filter flag.
3529 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3532 When using the filter with wav, note the default encoding for wav is 16-bit,
3533 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3534 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3536 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3537 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3540 The filter accepts the following options:
3543 @item disable_autoconvert
3544 Disable any automatic format conversion or resampling in the filter graph.
3546 @item process_stereo
3547 Process the stereo channels together. If target_gain does not match between
3548 channels, consider it invalid and use the last valid target_gain.
3551 Set the code detect timer period in ms.
3554 Always extend peaks above -3dBFS even if PE isn't signaled.
3557 Replace audio with a solid tone and adjust the amplitude to signal some
3558 specific aspect of the decoding process. The output file can be loaded in
3559 an audio editor alongside the original to aid analysis.
3561 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3568 Gain adjustment level at each sample
3570 Samples where peak extend occurs
3572 Samples where the code detect timer is active
3574 Samples where the target gain does not match between channels
3580 Apply head-related transfer functions (HRTFs) to create virtual
3581 loudspeakers around the user for binaural listening via headphones.
3582 The HRIRs are provided via additional streams, for each channel
3583 one stereo input stream is needed.
3585 The filter accepts the following options:
3589 Set mapping of input streams for convolution.
3590 The argument is a '|'-separated list of channel names in order as they
3591 are given as additional stream inputs for filter.
3592 This also specify number of input streams. Number of input streams
3593 must be not less than number of channels in first stream plus one.
3596 Set gain applied to audio. Value is in dB. Default is 0.
3599 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3600 processing audio in time domain which is slow.
3601 @var{freq} is processing audio in frequency domain which is fast.
3602 Default is @var{freq}.
3605 Set custom gain for LFE channels. Value is in dB. Default is 0.
3608 Set size of frame in number of samples which will be processed at once.
3609 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3612 Set format of hrir stream.
3613 Default value is @var{stereo}. Alternative value is @var{multich}.
3614 If value is set to @var{stereo}, number of additional streams should
3615 be greater or equal to number of input channels in first input stream.
3616 Also each additional stream should have stereo number of channels.
3617 If value is set to @var{multich}, number of additional streams should
3618 be exactly one. Also number of input channels of additional stream
3619 should be equal or greater than twice number of channels of first input
3623 @subsection Examples
3627 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3628 each amovie filter use stereo file with IR coefficients as input.
3629 The files give coefficients for each position of virtual loudspeaker:
3632 -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"
3637 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3638 but now in @var{multich} @var{hrir} format.
3640 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"
3647 Apply a high-pass filter with 3dB point frequency.
3648 The filter can be either single-pole, or double-pole (the default).
3649 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3651 The filter accepts the following options:
3655 Set frequency in Hz. Default is 3000.
3658 Set number of poles. Default is 2.
3661 Set method to specify band-width of filter.
3676 Specify the band-width of a filter in width_type units.
3677 Applies only to double-pole filter.
3678 The default is 0.707q and gives a Butterworth response.
3681 Specify which channels to filter, by default all available are filtered.
3684 @subsection Commands
3686 This filter supports the following commands:
3689 Change highpass frequency.
3690 Syntax for the command is : "@var{frequency}"
3693 Change highpass width_type.
3694 Syntax for the command is : "@var{width_type}"
3697 Change highpass width.
3698 Syntax for the command is : "@var{width}"
3703 Join multiple input streams into one multi-channel stream.
3705 It accepts the following parameters:
3709 The number of input streams. It defaults to 2.
3711 @item channel_layout
3712 The desired output channel layout. It defaults to stereo.
3715 Map channels from inputs to output. The argument is a '|'-separated list of
3716 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3717 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3718 can be either the name of the input channel (e.g. FL for front left) or its
3719 index in the specified input stream. @var{out_channel} is the name of the output
3723 The filter will attempt to guess the mappings when they are not specified
3724 explicitly. It does so by first trying to find an unused matching input channel
3725 and if that fails it picks the first unused input channel.
3727 Join 3 inputs (with properly set channel layouts):
3729 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3732 Build a 5.1 output from 6 single-channel streams:
3734 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3735 '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'
3741 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3743 To enable compilation of this filter you need to configure FFmpeg with
3744 @code{--enable-ladspa}.
3748 Specifies the name of LADSPA plugin library to load. If the environment
3749 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3750 each one of the directories specified by the colon separated list in
3751 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3752 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3753 @file{/usr/lib/ladspa/}.
3756 Specifies the plugin within the library. Some libraries contain only
3757 one plugin, but others contain many of them. If this is not set filter
3758 will list all available plugins within the specified library.
3761 Set the '|' separated list of controls which are zero or more floating point
3762 values that determine the behavior of the loaded plugin (for example delay,
3764 Controls need to be defined using the following syntax:
3765 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3766 @var{valuei} is the value set on the @var{i}-th control.
3767 Alternatively they can be also defined using the following syntax:
3768 @var{value0}|@var{value1}|@var{value2}|..., where
3769 @var{valuei} is the value set on the @var{i}-th control.
3770 If @option{controls} is set to @code{help}, all available controls and
3771 their valid ranges are printed.
3773 @item sample_rate, s
3774 Specify the sample rate, default to 44100. Only used if plugin have
3778 Set the number of samples per channel per each output frame, default
3779 is 1024. Only used if plugin have zero inputs.
3782 Set the minimum duration of the sourced audio. See
3783 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3784 for the accepted syntax.
3785 Note that the resulting duration may be greater than the specified duration,
3786 as the generated audio is always cut at the end of a complete frame.
3787 If not specified, or the expressed duration is negative, the audio is
3788 supposed to be generated forever.
3789 Only used if plugin have zero inputs.
3793 @subsection Examples
3797 List all available plugins within amp (LADSPA example plugin) library:
3803 List all available controls and their valid ranges for @code{vcf_notch}
3804 plugin from @code{VCF} library:
3806 ladspa=f=vcf:p=vcf_notch:c=help
3810 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3813 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3817 Add reverberation to the audio using TAP-plugins
3818 (Tom's Audio Processing plugins):
3820 ladspa=file=tap_reverb:tap_reverb
3824 Generate white noise, with 0.2 amplitude:
3826 ladspa=file=cmt:noise_source_white:c=c0=.2
3830 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3831 @code{C* Audio Plugin Suite} (CAPS) library:
3833 ladspa=file=caps:Click:c=c1=20'
3837 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3839 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3843 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3844 @code{SWH Plugins} collection:
3846 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3850 Attenuate low frequencies using Multiband EQ from Steve Harris
3851 @code{SWH Plugins} collection:
3853 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3857 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3860 ladspa=caps:Narrower
3864 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3866 ladspa=caps:White:.2
3870 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3872 ladspa=caps:Fractal:c=c1=1
3876 Dynamic volume normalization using @code{VLevel} plugin:
3878 ladspa=vlevel-ladspa:vlevel_mono
3882 @subsection Commands
3884 This filter supports the following commands:
3887 Modify the @var{N}-th control value.
3889 If the specified value is not valid, it is ignored and prior one is kept.
3894 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3895 Support for both single pass (livestreams, files) and double pass (files) modes.
3896 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3897 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3898 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3900 The filter accepts the following options:
3904 Set integrated loudness target.
3905 Range is -70.0 - -5.0. Default value is -24.0.
3908 Set loudness range target.
3909 Range is 1.0 - 20.0. Default value is 7.0.
3912 Set maximum true peak.
3913 Range is -9.0 - +0.0. Default value is -2.0.
3915 @item measured_I, measured_i
3916 Measured IL of input file.
3917 Range is -99.0 - +0.0.
3919 @item measured_LRA, measured_lra
3920 Measured LRA of input file.
3921 Range is 0.0 - 99.0.
3923 @item measured_TP, measured_tp
3924 Measured true peak of input file.
3925 Range is -99.0 - +99.0.
3927 @item measured_thresh
3928 Measured threshold of input file.
3929 Range is -99.0 - +0.0.
3932 Set offset gain. Gain is applied before the true-peak limiter.
3933 Range is -99.0 - +99.0. Default is +0.0.
3936 Normalize linearly if possible.
3937 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3938 to be specified in order to use this mode.
3939 Options are true or false. Default is true.
3942 Treat mono input files as "dual-mono". If a mono file is intended for playback
3943 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3944 If set to @code{true}, this option will compensate for this effect.
3945 Multi-channel input files are not affected by this option.
3946 Options are true or false. Default is false.
3949 Set print format for stats. Options are summary, json, or none.
3950 Default value is none.
3955 Apply a low-pass filter with 3dB point frequency.
3956 The filter can be either single-pole or double-pole (the default).
3957 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3959 The filter accepts the following options:
3963 Set frequency in Hz. Default is 500.
3966 Set number of poles. Default is 2.
3969 Set method to specify band-width of filter.
3984 Specify the band-width of a filter in width_type units.
3985 Applies only to double-pole filter.
3986 The default is 0.707q and gives a Butterworth response.
3989 Specify which channels to filter, by default all available are filtered.
3992 @subsection Examples
3995 Lowpass only LFE channel, it LFE is not present it does nothing:
4001 @subsection Commands
4003 This filter supports the following commands:
4006 Change lowpass frequency.
4007 Syntax for the command is : "@var{frequency}"
4010 Change lowpass width_type.
4011 Syntax for the command is : "@var{width_type}"
4014 Change lowpass width.
4015 Syntax for the command is : "@var{width}"
4020 Load a LV2 (LADSPA Version 2) plugin.
4022 To enable compilation of this filter you need to configure FFmpeg with
4023 @code{--enable-lv2}.
4027 Specifies the plugin URI. You may need to escape ':'.
4030 Set the '|' separated list of controls which are zero or more floating point
4031 values that determine the behavior of the loaded plugin (for example delay,
4033 If @option{controls} is set to @code{help}, all available controls and
4034 their valid ranges are printed.
4036 @item sample_rate, s
4037 Specify the sample rate, default to 44100. Only used if plugin have
4041 Set the number of samples per channel per each output frame, default
4042 is 1024. Only used if plugin have zero inputs.
4045 Set the minimum duration of the sourced audio. See
4046 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4047 for the accepted syntax.
4048 Note that the resulting duration may be greater than the specified duration,
4049 as the generated audio is always cut at the end of a complete frame.
4050 If not specified, or the expressed duration is negative, the audio is
4051 supposed to be generated forever.
4052 Only used if plugin have zero inputs.
4055 @subsection Examples
4059 Apply bass enhancer plugin from Calf:
4061 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4065 Apply vinyl plugin from Calf:
4067 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4071 Apply bit crusher plugin from ArtyFX:
4073 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4078 Multiband Compress or expand the audio's dynamic range.
4080 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4081 This is akin to the crossover of a loudspeaker, and results in flat frequency
4082 response when absent compander action.
4084 It accepts the following parameters:
4088 This option syntax is:
4089 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4090 For explanation of each item refer to compand filter documentation.
4096 Mix channels with specific gain levels. The filter accepts the output
4097 channel layout followed by a set of channels definitions.
4099 This filter is also designed to efficiently remap the channels of an audio
4102 The filter accepts parameters of the form:
4103 "@var{l}|@var{outdef}|@var{outdef}|..."
4107 output channel layout or number of channels
4110 output channel specification, of the form:
4111 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4114 output channel to define, either a channel name (FL, FR, etc.) or a channel
4115 number (c0, c1, etc.)
4118 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4121 input channel to use, see out_name for details; it is not possible to mix
4122 named and numbered input channels
4125 If the `=' in a channel specification is replaced by `<', then the gains for
4126 that specification will be renormalized so that the total is 1, thus
4127 avoiding clipping noise.
4129 @subsection Mixing examples
4131 For example, if you want to down-mix from stereo to mono, but with a bigger
4132 factor for the left channel:
4134 pan=1c|c0=0.9*c0+0.1*c1
4137 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4138 7-channels surround:
4140 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4143 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4144 that should be preferred (see "-ac" option) unless you have very specific
4147 @subsection Remapping examples
4149 The channel remapping will be effective if, and only if:
4152 @item gain coefficients are zeroes or ones,
4153 @item only one input per channel output,
4156 If all these conditions are satisfied, the filter will notify the user ("Pure
4157 channel mapping detected"), and use an optimized and lossless method to do the
4160 For example, if you have a 5.1 source and want a stereo audio stream by
4161 dropping the extra channels:
4163 pan="stereo| c0=FL | c1=FR"
4166 Given the same source, you can also switch front left and front right channels
4167 and keep the input channel layout:
4169 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4172 If the input is a stereo audio stream, you can mute the front left channel (and
4173 still keep the stereo channel layout) with:
4178 Still with a stereo audio stream input, you can copy the right channel in both
4179 front left and right:
4181 pan="stereo| c0=FR | c1=FR"
4186 ReplayGain scanner filter. This filter takes an audio stream as an input and
4187 outputs it unchanged.
4188 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4192 Convert the audio sample format, sample rate and channel layout. It is
4193 not meant to be used directly.
4196 Apply time-stretching and pitch-shifting with librubberband.
4198 To enable compilation of this filter, you need to configure FFmpeg with
4199 @code{--enable-librubberband}.
4201 The filter accepts the following options:
4205 Set tempo scale factor.
4208 Set pitch scale factor.
4211 Set transients detector.
4212 Possible values are:
4221 Possible values are:
4230 Possible values are:
4237 Set processing window size.
4238 Possible values are:
4247 Possible values are:
4254 Enable formant preservation when shift pitching.
4255 Possible values are:
4263 Possible values are:
4272 Possible values are:
4279 @section sidechaincompress
4281 This filter acts like normal compressor but has the ability to compress
4282 detected signal using second input signal.
4283 It needs two input streams and returns one output stream.
4284 First input stream will be processed depending on second stream signal.
4285 The filtered signal then can be filtered with other filters in later stages of
4286 processing. See @ref{pan} and @ref{amerge} filter.
4288 The filter accepts the following options:
4292 Set input gain. Default is 1. Range is between 0.015625 and 64.
4295 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
4296 Default is @code{downward}.
4299 If a signal of second stream raises above this level it will affect the gain
4300 reduction of first stream.
4301 By default is 0.125. Range is between 0.00097563 and 1.
4304 Set a ratio about which the signal is reduced. 1:2 means that if the level
4305 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4306 Default is 2. Range is between 1 and 20.
4309 Amount of milliseconds the signal has to rise above the threshold before gain
4310 reduction starts. Default is 20. Range is between 0.01 and 2000.
4313 Amount of milliseconds the signal has to fall below the threshold before
4314 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4317 Set the amount by how much signal will be amplified after processing.
4318 Default is 1. Range is from 1 to 64.
4321 Curve the sharp knee around the threshold to enter gain reduction more softly.
4322 Default is 2.82843. Range is between 1 and 8.
4325 Choose if the @code{average} level between all channels of side-chain stream
4326 or the louder(@code{maximum}) channel of side-chain stream affects the
4327 reduction. Default is @code{average}.
4330 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4331 of @code{rms}. Default is @code{rms} which is mainly smoother.
4334 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4337 How much to use compressed signal in output. Default is 1.
4338 Range is between 0 and 1.
4341 @subsection Examples
4345 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4346 depending on the signal of 2nd input and later compressed signal to be
4347 merged with 2nd input:
4349 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4353 @section sidechaingate
4355 A sidechain gate acts like a normal (wideband) gate but has the ability to
4356 filter the detected signal before sending it to the gain reduction stage.
4357 Normally a gate uses the full range signal to detect a level above the
4359 For example: If you cut all lower frequencies from your sidechain signal
4360 the gate will decrease the volume of your track only if not enough highs
4361 appear. With this technique you are able to reduce the resonation of a
4362 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4364 It needs two input streams and returns one output stream.
4365 First input stream will be processed depending on second stream signal.
4367 The filter accepts the following options:
4371 Set input level before filtering.
4372 Default is 1. Allowed range is from 0.015625 to 64.
4375 Set the mode of operation. Can be @code{upward} or @code{downward}.
4376 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
4377 will be amplified, expanding dynamic range in upward direction.
4378 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
4381 Set the level of gain reduction when the signal is below the threshold.
4382 Default is 0.06125. Allowed range is from 0 to 1.
4383 Setting this to 0 disables reduction and then filter behaves like expander.
4386 If a signal rises above this level the gain reduction is released.
4387 Default is 0.125. Allowed range is from 0 to 1.
4390 Set a ratio about which the signal is reduced.
4391 Default is 2. Allowed range is from 1 to 9000.
4394 Amount of milliseconds the signal has to rise above the threshold before gain
4396 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4399 Amount of milliseconds the signal has to fall below the threshold before the
4400 reduction is increased again. Default is 250 milliseconds.
4401 Allowed range is from 0.01 to 9000.
4404 Set amount of amplification of signal after processing.
4405 Default is 1. Allowed range is from 1 to 64.
4408 Curve the sharp knee around the threshold to enter gain reduction more softly.
4409 Default is 2.828427125. Allowed range is from 1 to 8.
4412 Choose if exact signal should be taken for detection or an RMS like one.
4413 Default is rms. Can be peak or rms.
4416 Choose if the average level between all channels or the louder channel affects
4418 Default is average. Can be average or maximum.
4421 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4424 @section silencedetect
4426 Detect silence in an audio stream.
4428 This filter logs a message when it detects that the input audio volume is less
4429 or equal to a noise tolerance value for a duration greater or equal to the
4430 minimum detected noise duration.
4432 The printed times and duration are expressed in seconds.
4434 The filter accepts the following options:
4438 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4439 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4442 Set silence duration until notification (default is 2 seconds).
4445 Process each channel separately, instead of combined. By default is disabled.
4448 @subsection Examples
4452 Detect 5 seconds of silence with -50dB noise tolerance:
4454 silencedetect=n=-50dB:d=5
4458 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4459 tolerance in @file{silence.mp3}:
4461 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4465 @section silenceremove
4467 Remove silence from the beginning, middle or end of the audio.
4469 The filter accepts the following options:
4473 This value is used to indicate if audio should be trimmed at beginning of
4474 the audio. A value of zero indicates no silence should be trimmed from the
4475 beginning. When specifying a non-zero value, it trims audio up until it
4476 finds non-silence. Normally, when trimming silence from beginning of audio
4477 the @var{start_periods} will be @code{1} but it can be increased to higher
4478 values to trim all audio up to specific count of non-silence periods.
4479 Default value is @code{0}.
4481 @item start_duration
4482 Specify the amount of time that non-silence must be detected before it stops
4483 trimming audio. By increasing the duration, bursts of noises can be treated
4484 as silence and trimmed off. Default value is @code{0}.
4486 @item start_threshold
4487 This indicates what sample value should be treated as silence. For digital
4488 audio, a value of @code{0} may be fine but for audio recorded from analog,
4489 you may wish to increase the value to account for background noise.
4490 Can be specified in dB (in case "dB" is appended to the specified value)
4491 or amplitude ratio. Default value is @code{0}.
4494 Specify max duration of silence at beginning that will be kept after
4495 trimming. Default is 0, which is equal to trimming all samples detected
4499 Specify mode of detection of silence end in start of multi-channel audio.
4500 Can be @var{any} or @var{all}. Default is @var{any}.
4501 With @var{any}, any sample that is detected as non-silence will cause
4502 stopped trimming of silence.
4503 With @var{all}, only if all channels are detected as non-silence will cause
4504 stopped trimming of silence.
4507 Set the count for trimming silence from the end of audio.
4508 To remove silence from the middle of a file, specify a @var{stop_periods}
4509 that is negative. This value is then treated as a positive value and is
4510 used to indicate the effect should restart processing as specified by
4511 @var{start_periods}, making it suitable for removing periods of silence
4512 in the middle of the audio.
4513 Default value is @code{0}.
4516 Specify a duration of silence that must exist before audio is not copied any
4517 more. By specifying a higher duration, silence that is wanted can be left in
4519 Default value is @code{0}.
4521 @item stop_threshold
4522 This is the same as @option{start_threshold} but for trimming silence from
4524 Can be specified in dB (in case "dB" is appended to the specified value)
4525 or amplitude ratio. Default value is @code{0}.
4528 Specify max duration of silence at end that will be kept after
4529 trimming. Default is 0, which is equal to trimming all samples detected
4533 Specify mode of detection of silence start in end of multi-channel audio.
4534 Can be @var{any} or @var{all}. Default is @var{any}.
4535 With @var{any}, any sample that is detected as non-silence will cause
4536 stopped trimming of silence.
4537 With @var{all}, only if all channels are detected as non-silence will cause
4538 stopped trimming of silence.
4541 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4542 and works better with digital silence which is exactly 0.
4543 Default value is @code{rms}.
4546 Set duration in number of seconds used to calculate size of window in number
4547 of samples for detecting silence.
4548 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4551 @subsection Examples
4555 The following example shows how this filter can be used to start a recording
4556 that does not contain the delay at the start which usually occurs between
4557 pressing the record button and the start of the performance:
4559 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4563 Trim all silence encountered from beginning to end where there is more than 1
4564 second of silence in audio:
4566 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4572 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4573 loudspeakers around the user for binaural listening via headphones (audio
4574 formats up to 9 channels supported).
4575 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4576 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4577 Austrian Academy of Sciences.
4579 To enable compilation of this filter you need to configure FFmpeg with
4580 @code{--enable-libmysofa}.
4582 The filter accepts the following options:
4586 Set the SOFA file used for rendering.
4589 Set gain applied to audio. Value is in dB. Default is 0.
4592 Set rotation of virtual loudspeakers in deg. Default is 0.
4595 Set elevation of virtual speakers in deg. Default is 0.
4598 Set distance in meters between loudspeakers and the listener with near-field
4599 HRTFs. Default is 1.
4602 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4603 processing audio in time domain which is slow.
4604 @var{freq} is processing audio in frequency domain which is fast.
4605 Default is @var{freq}.
4608 Set custom positions of virtual loudspeakers. Syntax for this option is:
4609 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4610 Each virtual loudspeaker is described with short channel name following with
4611 azimuth and elevation in degrees.
4612 Each virtual loudspeaker description is separated by '|'.
4613 For example to override front left and front right channel positions use:
4614 'speakers=FL 45 15|FR 345 15'.
4615 Descriptions with unrecognised channel names are ignored.
4618 Set custom gain for LFE channels. Value is in dB. Default is 0.
4621 Set custom frame size in number of samples. Default is 1024.
4622 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4623 is set to @var{freq}.
4626 Should all IRs be normalized upon importing SOFA file.
4627 By default is enabled.
4630 Should nearest IRs be interpolated with neighbor IRs if exact position
4631 does not match. By default is disabled.
4634 Minphase all IRs upon loading of SOFA file. By default is disabled.
4637 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4640 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4643 @subsection Examples
4647 Using ClubFritz6 sofa file:
4649 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4653 Using ClubFritz12 sofa file and bigger radius with small rotation:
4655 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4659 Similar as above but with custom speaker positions for front left, front right, back left and back right
4660 and also with custom gain:
4662 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4666 @section stereotools
4668 This filter has some handy utilities to manage stereo signals, for converting
4669 M/S stereo recordings to L/R signal while having control over the parameters
4670 or spreading the stereo image of master track.
4672 The filter accepts the following options:
4676 Set input level before filtering for both channels. Defaults is 1.
4677 Allowed range is from 0.015625 to 64.
4680 Set output level after filtering for both channels. Defaults is 1.
4681 Allowed range is from 0.015625 to 64.
4684 Set input balance between both channels. Default is 0.
4685 Allowed range is from -1 to 1.
4688 Set output balance between both channels. Default is 0.
4689 Allowed range is from -1 to 1.
4692 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4693 clipping. Disabled by default.
4696 Mute the left channel. Disabled by default.
4699 Mute the right channel. Disabled by default.
4702 Change the phase of the left channel. Disabled by default.
4705 Change the phase of the right channel. Disabled by default.
4708 Set stereo mode. Available values are:
4712 Left/Right to Left/Right, this is default.
4715 Left/Right to Mid/Side.
4718 Mid/Side to Left/Right.
4721 Left/Right to Left/Left.
4724 Left/Right to Right/Right.
4727 Left/Right to Left + Right.
4730 Left/Right to Right/Left.
4733 Mid/Side to Left/Left.
4736 Mid/Side to Right/Right.
4740 Set level of side signal. Default is 1.
4741 Allowed range is from 0.015625 to 64.
4744 Set balance of side signal. Default is 0.
4745 Allowed range is from -1 to 1.
4748 Set level of the middle signal. Default is 1.
4749 Allowed range is from 0.015625 to 64.
4752 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4755 Set stereo base between mono and inversed channels. Default is 0.
4756 Allowed range is from -1 to 1.
4759 Set delay in milliseconds how much to delay left from right channel and
4760 vice versa. Default is 0. Allowed range is from -20 to 20.
4763 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4766 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4768 @item bmode_in, bmode_out
4769 Set balance mode for balance_in/balance_out option.
4771 Can be one of the following:
4775 Classic balance mode. Attenuate one channel at time.
4776 Gain is raised up to 1.
4779 Similar as classic mode above but gain is raised up to 2.
4782 Equal power distribution, from -6dB to +6dB range.
4786 @subsection Examples
4790 Apply karaoke like effect:
4792 stereotools=mlev=0.015625
4796 Convert M/S signal to L/R:
4798 "stereotools=mode=ms>lr"
4802 @section stereowiden
4804 This filter enhance the stereo effect by suppressing signal common to both
4805 channels and by delaying the signal of left into right and vice versa,
4806 thereby widening the stereo effect.
4808 The filter accepts the following options:
4812 Time in milliseconds of the delay of left signal into right and vice versa.
4813 Default is 20 milliseconds.
4816 Amount of gain in delayed signal into right and vice versa. Gives a delay
4817 effect of left signal in right output and vice versa which gives widening
4818 effect. Default is 0.3.
4821 Cross feed of left into right with inverted phase. This helps in suppressing
4822 the mono. If the value is 1 it will cancel all the signal common to both
4823 channels. Default is 0.3.
4826 Set level of input signal of original channel. Default is 0.8.
4829 @section superequalizer
4830 Apply 18 band equalizer.
4832 The filter accepts the following options:
4839 Set 131Hz band gain.
4841 Set 185Hz band gain.
4843 Set 262Hz band gain.
4845 Set 370Hz band gain.
4847 Set 523Hz band gain.
4849 Set 740Hz band gain.
4851 Set 1047Hz band gain.
4853 Set 1480Hz band gain.
4855 Set 2093Hz band gain.
4857 Set 2960Hz band gain.
4859 Set 4186Hz band gain.
4861 Set 5920Hz band gain.
4863 Set 8372Hz band gain.
4865 Set 11840Hz band gain.
4867 Set 16744Hz band gain.
4869 Set 20000Hz band gain.
4873 Apply audio surround upmix filter.
4875 This filter allows to produce multichannel output from audio stream.
4877 The filter accepts the following options:
4881 Set output channel layout. By default, this is @var{5.1}.
4883 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4884 for the required syntax.
4887 Set input channel layout. By default, this is @var{stereo}.
4889 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4890 for the required syntax.
4893 Set input volume level. By default, this is @var{1}.
4896 Set output volume level. By default, this is @var{1}.
4899 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4902 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4905 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4908 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
4909 In @var{add} mode, LFE channel is created from input audio and added to output.
4910 In @var{sub} mode, LFE channel is created from input audio and added to output but
4911 also all non-LFE output channels are subtracted with output LFE channel.
4914 Set angle of stereo surround transform, Allowed range is from @var{0} to @var{360}.
4915 Default is @var{90}.
4918 Set front center input volume. By default, this is @var{1}.
4921 Set front center output volume. By default, this is @var{1}.
4924 Set front left input volume. By default, this is @var{1}.
4927 Set front left output volume. By default, this is @var{1}.
4930 Set front right input volume. By default, this is @var{1}.
4933 Set front right output volume. By default, this is @var{1}.
4936 Set side left input volume. By default, this is @var{1}.
4939 Set side left output volume. By default, this is @var{1}.
4942 Set side right input volume. By default, this is @var{1}.
4945 Set side right output volume. By default, this is @var{1}.
4948 Set back left input volume. By default, this is @var{1}.
4951 Set back left output volume. By default, this is @var{1}.
4954 Set back right input volume. By default, this is @var{1}.
4957 Set back right output volume. By default, this is @var{1}.
4960 Set back center input volume. By default, this is @var{1}.
4963 Set back center output volume. By default, this is @var{1}.
4966 Set LFE input volume. By default, this is @var{1}.
4969 Set LFE output volume. By default, this is @var{1}.
4972 Set spread usage of stereo image across X axis for all channels.
4975 Set spread usage of stereo image across Y axis for all channels.
4977 @item fcx, flx, frx, blx, brx, slx, srx, bcx
4978 Set spread usage of stereo image across X axis for each channel.
4980 @item fcy, fly, fry, bly, bry, sly, sry, bcy
4981 Set spread usage of stereo image across Y axis for each channel.
4984 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
4987 Set window function.
4989 It accepts the following values:
5012 Default is @code{hann}.
5015 Set window overlap. If set to 1, the recommended overlap for selected
5016 window function will be picked. Default is @code{0.5}.
5019 @section treble, highshelf
5021 Boost or cut treble (upper) frequencies of the audio using a two-pole
5022 shelving filter with a response similar to that of a standard
5023 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
5025 The filter accepts the following options:
5029 Give the gain at whichever is the lower of ~22 kHz and the
5030 Nyquist frequency. Its useful range is about -20 (for a large cut)
5031 to +20 (for a large boost). Beware of clipping when using a positive gain.
5034 Set the filter's central frequency and so can be used
5035 to extend or reduce the frequency range to be boosted or cut.
5036 The default value is @code{3000} Hz.
5039 Set method to specify band-width of filter.
5054 Determine how steep is the filter's shelf transition.
5057 Specify which channels to filter, by default all available are filtered.
5060 @subsection Commands
5062 This filter supports the following commands:
5065 Change treble frequency.
5066 Syntax for the command is : "@var{frequency}"
5069 Change treble width_type.
5070 Syntax for the command is : "@var{width_type}"
5073 Change treble width.
5074 Syntax for the command is : "@var{width}"
5078 Syntax for the command is : "@var{gain}"
5083 Sinusoidal amplitude modulation.
5085 The filter accepts the following options:
5089 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
5090 (20 Hz or lower) will result in a tremolo effect.
5091 This filter may also be used as a ring modulator by specifying
5092 a modulation frequency higher than 20 Hz.
5093 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5096 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5097 Default value is 0.5.
5102 Sinusoidal phase modulation.
5104 The filter accepts the following options:
5108 Modulation frequency in Hertz.
5109 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5112 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5113 Default value is 0.5.
5118 Adjust the input audio volume.
5120 It accepts the following parameters:
5124 Set audio volume expression.
5126 Output values are clipped to the maximum value.
5128 The output audio volume is given by the relation:
5130 @var{output_volume} = @var{volume} * @var{input_volume}
5133 The default value for @var{volume} is "1.0".
5136 This parameter represents the mathematical precision.
5138 It determines which input sample formats will be allowed, which affects the
5139 precision of the volume scaling.
5143 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
5145 32-bit floating-point; this limits input sample format to FLT. (default)
5147 64-bit floating-point; this limits input sample format to DBL.
5151 Choose the behaviour on encountering ReplayGain side data in input frames.
5155 Remove ReplayGain side data, ignoring its contents (the default).
5158 Ignore ReplayGain side data, but leave it in the frame.
5161 Prefer the track gain, if present.
5164 Prefer the album gain, if present.
5167 @item replaygain_preamp
5168 Pre-amplification gain in dB to apply to the selected replaygain gain.
5170 Default value for @var{replaygain_preamp} is 0.0.
5173 Set when the volume expression is evaluated.
5175 It accepts the following values:
5178 only evaluate expression once during the filter initialization, or
5179 when the @samp{volume} command is sent
5182 evaluate expression for each incoming frame
5185 Default value is @samp{once}.
5188 The volume expression can contain the following parameters.
5192 frame number (starting at zero)
5195 @item nb_consumed_samples
5196 number of samples consumed by the filter
5198 number of samples in the current frame
5200 original frame position in the file
5206 PTS at start of stream
5208 time at start of stream
5214 last set volume value
5217 Note that when @option{eval} is set to @samp{once} only the
5218 @var{sample_rate} and @var{tb} variables are available, all other
5219 variables will evaluate to NAN.
5221 @subsection Commands
5223 This filter supports the following commands:
5226 Modify the volume expression.
5227 The command accepts the same syntax of the corresponding option.
5229 If the specified expression is not valid, it is kept at its current
5231 @item replaygain_noclip
5232 Prevent clipping by limiting the gain applied.
5234 Default value for @var{replaygain_noclip} is 1.
5238 @subsection Examples
5242 Halve the input audio volume:
5246 volume=volume=-6.0206dB
5249 In all the above example the named key for @option{volume} can be
5250 omitted, for example like in:
5256 Increase input audio power by 6 decibels using fixed-point precision:
5258 volume=volume=6dB:precision=fixed
5262 Fade volume after time 10 with an annihilation period of 5 seconds:
5264 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5268 @section volumedetect
5270 Detect the volume of the input video.
5272 The filter has no parameters. The input is not modified. Statistics about
5273 the volume will be printed in the log when the input stream end is reached.
5275 In particular it will show the mean volume (root mean square), maximum
5276 volume (on a per-sample basis), and the beginning of a histogram of the
5277 registered volume values (from the maximum value to a cumulated 1/1000 of
5280 All volumes are in decibels relative to the maximum PCM value.
5282 @subsection Examples
5284 Here is an excerpt of the output:
5286 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5287 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5288 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5289 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5290 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5291 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5292 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5293 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5294 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5300 The mean square energy is approximately -27 dB, or 10^-2.7.
5302 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5304 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5307 In other words, raising the volume by +4 dB does not cause any clipping,
5308 raising it by +5 dB causes clipping for 6 samples, etc.
5310 @c man end AUDIO FILTERS
5312 @chapter Audio Sources
5313 @c man begin AUDIO SOURCES
5315 Below is a description of the currently available audio sources.
5319 Buffer audio frames, and make them available to the filter chain.
5321 This source is mainly intended for a programmatic use, in particular
5322 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5324 It accepts the following parameters:
5328 The timebase which will be used for timestamps of submitted frames. It must be
5329 either a floating-point number or in @var{numerator}/@var{denominator} form.
5332 The sample rate of the incoming audio buffers.
5335 The sample format of the incoming audio buffers.
5336 Either a sample format name or its corresponding integer representation from
5337 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5339 @item channel_layout
5340 The channel layout of the incoming audio buffers.
5341 Either a channel layout name from channel_layout_map in
5342 @file{libavutil/channel_layout.c} or its corresponding integer representation
5343 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5346 The number of channels of the incoming audio buffers.
5347 If both @var{channels} and @var{channel_layout} are specified, then they
5352 @subsection Examples
5355 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5358 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5359 Since the sample format with name "s16p" corresponds to the number
5360 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5363 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5368 Generate an audio signal specified by an expression.
5370 This source accepts in input one or more expressions (one for each
5371 channel), which are evaluated and used to generate a corresponding
5374 This source accepts the following options:
5378 Set the '|'-separated expressions list for each separate channel. In case the
5379 @option{channel_layout} option is not specified, the selected channel layout
5380 depends on the number of provided expressions. Otherwise the last
5381 specified expression is applied to the remaining output channels.
5383 @item channel_layout, c
5384 Set the channel layout. The number of channels in the specified layout
5385 must be equal to the number of specified expressions.
5388 Set the minimum duration of the sourced audio. See
5389 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5390 for the accepted syntax.
5391 Note that the resulting duration may be greater than the specified
5392 duration, as the generated audio is always cut at the end of a
5395 If not specified, or the expressed duration is negative, the audio is
5396 supposed to be generated forever.
5399 Set the number of samples per channel per each output frame,
5402 @item sample_rate, s
5403 Specify the sample rate, default to 44100.
5406 Each expression in @var{exprs} can contain the following constants:
5410 number of the evaluated sample, starting from 0
5413 time of the evaluated sample expressed in seconds, starting from 0
5420 @subsection Examples
5430 Generate a sin signal with frequency of 440 Hz, set sample rate to
5433 aevalsrc="sin(440*2*PI*t):s=8000"
5437 Generate a two channels signal, specify the channel layout (Front
5438 Center + Back Center) explicitly:
5440 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5444 Generate white noise:
5446 aevalsrc="-2+random(0)"
5450 Generate an amplitude modulated signal:
5452 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5456 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5458 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5465 The null audio source, return unprocessed audio frames. It is mainly useful
5466 as a template and to be employed in analysis / debugging tools, or as
5467 the source for filters which ignore the input data (for example the sox
5470 This source accepts the following options:
5474 @item channel_layout, cl
5476 Specifies the channel layout, and can be either an integer or a string
5477 representing a channel layout. The default value of @var{channel_layout}
5480 Check the channel_layout_map definition in
5481 @file{libavutil/channel_layout.c} for the mapping between strings and
5482 channel layout values.
5484 @item sample_rate, r
5485 Specifies the sample rate, and defaults to 44100.
5488 Set the number of samples per requested frames.
5492 @subsection Examples
5496 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5498 anullsrc=r=48000:cl=4
5502 Do the same operation with a more obvious syntax:
5504 anullsrc=r=48000:cl=mono
5508 All the parameters need to be explicitly defined.
5512 Synthesize a voice utterance using the libflite library.
5514 To enable compilation of this filter you need to configure FFmpeg with
5515 @code{--enable-libflite}.
5517 Note that versions of the flite library prior to 2.0 are not thread-safe.
5519 The filter accepts the following options:
5524 If set to 1, list the names of the available voices and exit
5525 immediately. Default value is 0.
5528 Set the maximum number of samples per frame. Default value is 512.
5531 Set the filename containing the text to speak.
5534 Set the text to speak.
5537 Set the voice to use for the speech synthesis. Default value is
5538 @code{kal}. See also the @var{list_voices} option.
5541 @subsection Examples
5545 Read from file @file{speech.txt}, and synthesize the text using the
5546 standard flite voice:
5548 flite=textfile=speech.txt
5552 Read the specified text selecting the @code{slt} voice:
5554 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5558 Input text to ffmpeg:
5560 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5564 Make @file{ffplay} speak the specified text, using @code{flite} and
5565 the @code{lavfi} device:
5567 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5571 For more information about libflite, check:
5572 @url{http://www.festvox.org/flite/}
5576 Generate a noise audio signal.
5578 The filter accepts the following options:
5581 @item sample_rate, r
5582 Specify the sample rate. Default value is 48000 Hz.
5585 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5589 Specify the duration of the generated audio stream. Not specifying this option
5590 results in noise with an infinite length.
5592 @item color, colour, c
5593 Specify the color of noise. Available noise colors are white, pink, brown,
5594 blue and violet. Default color is white.
5597 Specify a value used to seed the PRNG.
5600 Set the number of samples per each output frame, default is 1024.
5603 @subsection Examples
5608 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5610 anoisesrc=d=60:c=pink:r=44100:a=0.5
5616 Generate odd-tap Hilbert transform FIR coefficients.
5618 The resulting stream can be used with @ref{afir} filter for phase-shifting
5619 the signal by 90 degrees.
5621 This is used in many matrix coding schemes and for analytic signal generation.
5622 The process is often written as a multiplication by i (or j), the imaginary unit.
5624 The filter accepts the following options:
5628 @item sample_rate, s
5629 Set sample rate, default is 44100.
5632 Set length of FIR filter, default is 22051.
5635 Set number of samples per each frame.
5638 Set window function to be used when generating FIR coefficients.
5643 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5645 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5647 The filter accepts the following options:
5650 @item sample_rate, r
5651 Set sample rate, default is 44100.
5654 Set number of samples per each frame. Default is 1024.
5657 Set high-pass frequency. Default is 0.
5660 Set low-pass frequency. Default is 0.
5661 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5662 is higher than 0 then filter will create band-pass filter coefficients,
5663 otherwise band-reject filter coefficients.
5666 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5669 Set Kaiser window beta.
5672 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5675 Enable rounding, by default is disabled.
5678 Set number of taps for high-pass filter.
5681 Set number of taps for low-pass filter.
5686 Generate an audio signal made of a sine wave with amplitude 1/8.
5688 The audio signal is bit-exact.
5690 The filter accepts the following options:
5695 Set the carrier frequency. Default is 440 Hz.
5697 @item beep_factor, b
5698 Enable a periodic beep every second with frequency @var{beep_factor} times
5699 the carrier frequency. Default is 0, meaning the beep is disabled.
5701 @item sample_rate, r
5702 Specify the sample rate, default is 44100.
5705 Specify the duration of the generated audio stream.
5707 @item samples_per_frame
5708 Set the number of samples per output frame.
5710 The expression can contain the following constants:
5714 The (sequential) number of the output audio frame, starting from 0.
5717 The PTS (Presentation TimeStamp) of the output audio frame,
5718 expressed in @var{TB} units.
5721 The PTS of the output audio frame, expressed in seconds.
5724 The timebase of the output audio frames.
5727 Default is @code{1024}.
5730 @subsection Examples
5735 Generate a simple 440 Hz sine wave:
5741 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5745 sine=frequency=220:beep_factor=4:duration=5
5749 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5752 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5756 @c man end AUDIO SOURCES
5758 @chapter Audio Sinks
5759 @c man begin AUDIO SINKS
5761 Below is a description of the currently available audio sinks.
5763 @section abuffersink
5765 Buffer audio frames, and make them available to the end of filter chain.
5767 This sink is mainly intended for programmatic use, in particular
5768 through the interface defined in @file{libavfilter/buffersink.h}
5769 or the options system.
5771 It accepts a pointer to an AVABufferSinkContext structure, which
5772 defines the incoming buffers' formats, to be passed as the opaque
5773 parameter to @code{avfilter_init_filter} for initialization.
5776 Null audio sink; do absolutely nothing with the input audio. It is
5777 mainly useful as a template and for use in analysis / debugging
5780 @c man end AUDIO SINKS
5782 @chapter Video Filters
5783 @c man begin VIDEO FILTERS
5785 When you configure your FFmpeg build, you can disable any of the
5786 existing filters using @code{--disable-filters}.
5787 The configure output will show the video filters included in your
5790 Below is a description of the currently available video filters.
5792 @section alphaextract
5794 Extract the alpha component from the input as a grayscale video. This
5795 is especially useful with the @var{alphamerge} filter.
5799 Add or replace the alpha component of the primary input with the
5800 grayscale value of a second input. This is intended for use with
5801 @var{alphaextract} to allow the transmission or storage of frame
5802 sequences that have alpha in a format that doesn't support an alpha
5805 For example, to reconstruct full frames from a normal YUV-encoded video
5806 and a separate video created with @var{alphaextract}, you might use:
5808 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5811 Since this filter is designed for reconstruction, it operates on frame
5812 sequences without considering timestamps, and terminates when either
5813 input reaches end of stream. This will cause problems if your encoding
5814 pipeline drops frames. If you're trying to apply an image as an
5815 overlay to a video stream, consider the @var{overlay} filter instead.
5819 Amplify differences between current pixel and pixels of adjacent frames in
5820 same pixel location.
5822 This filter accepts the following options:
5826 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5827 For example radius of 3 will instruct filter to calculate average of 7 frames.
5830 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5833 Set threshold for difference amplification. Any difference greater or equal to
5834 this value will not alter source pixel. Default is 10.
5835 Allowed range is from 0 to 65535.
5838 Set tolerance for difference amplification. Any difference lower to
5839 this value will not alter source pixel. Default is 0.
5840 Allowed range is from 0 to 65535.
5843 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5844 This option controls maximum possible value that will decrease source pixel value.
5847 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5848 This option controls maximum possible value that will increase source pixel value.
5851 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5856 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5857 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5858 Substation Alpha) subtitles files.
5860 This filter accepts the following option in addition to the common options from
5861 the @ref{subtitles} filter:
5865 Set the shaping engine
5867 Available values are:
5870 The default libass shaping engine, which is the best available.
5872 Fast, font-agnostic shaper that can do only substitutions
5874 Slower shaper using OpenType for substitutions and positioning
5877 The default is @code{auto}.
5881 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5883 The filter accepts the following options:
5887 Set threshold A for 1st plane. Default is 0.02.
5888 Valid range is 0 to 0.3.
5891 Set threshold B for 1st plane. Default is 0.04.
5892 Valid range is 0 to 5.
5895 Set threshold A for 2nd plane. Default is 0.02.
5896 Valid range is 0 to 0.3.
5899 Set threshold B for 2nd plane. Default is 0.04.
5900 Valid range is 0 to 5.
5903 Set threshold A for 3rd plane. Default is 0.02.
5904 Valid range is 0 to 0.3.
5907 Set threshold B for 3rd plane. Default is 0.04.
5908 Valid range is 0 to 5.
5910 Threshold A is designed to react on abrupt changes in the input signal and
5911 threshold B is designed to react on continuous changes in the input signal.
5914 Set number of frames filter will use for averaging. Default is 9. Must be odd
5915 number in range [5, 129].
5918 Set what planes of frame filter will use for averaging. Default is all.
5923 Apply average blur filter.
5925 The filter accepts the following options:
5929 Set horizontal radius size.
5932 Set which planes to filter. By default all planes are filtered.
5935 Set vertical radius size, if zero it will be same as @code{sizeX}.
5936 Default is @code{0}.
5941 Compute the bounding box for the non-black pixels in the input frame
5944 This filter computes the bounding box containing all the pixels with a
5945 luminance value greater than the minimum allowed value.
5946 The parameters describing the bounding box are printed on the filter
5949 The filter accepts the following option:
5953 Set the minimal luminance value. Default is @code{16}.
5956 @section bitplanenoise
5958 Show and measure bit plane noise.
5960 The filter accepts the following options:
5964 Set which plane to analyze. Default is @code{1}.
5967 Filter out noisy pixels from @code{bitplane} set above.
5968 Default is disabled.
5971 @section blackdetect
5973 Detect video intervals that are (almost) completely black. Can be
5974 useful to detect chapter transitions, commercials, or invalid
5975 recordings. Output lines contains the time for the start, end and
5976 duration of the detected black interval expressed in seconds.
5978 In order to display the output lines, you need to set the loglevel at
5979 least to the AV_LOG_INFO value.
5981 The filter accepts the following options:
5984 @item black_min_duration, d
5985 Set the minimum detected black duration expressed in seconds. It must
5986 be a non-negative floating point number.
5988 Default value is 2.0.
5990 @item picture_black_ratio_th, pic_th
5991 Set the threshold for considering a picture "black".
5992 Express the minimum value for the ratio:
5994 @var{nb_black_pixels} / @var{nb_pixels}
5997 for which a picture is considered black.
5998 Default value is 0.98.
6000 @item pixel_black_th, pix_th
6001 Set the threshold for considering a pixel "black".
6003 The threshold expresses the maximum pixel luminance value for which a
6004 pixel is considered "black". The provided value is scaled according to
6005 the following equation:
6007 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
6010 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
6011 the input video format, the range is [0-255] for YUV full-range
6012 formats and [16-235] for YUV non full-range formats.
6014 Default value is 0.10.
6017 The following example sets the maximum pixel threshold to the minimum
6018 value, and detects only black intervals of 2 or more seconds:
6020 blackdetect=d=2:pix_th=0.00
6025 Detect frames that are (almost) completely black. Can be useful to
6026 detect chapter transitions or commercials. Output lines consist of
6027 the frame number of the detected frame, the percentage of blackness,
6028 the position in the file if known or -1 and the timestamp in seconds.
6030 In order to display the output lines, you need to set the loglevel at
6031 least to the AV_LOG_INFO value.
6033 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
6034 The value represents the percentage of pixels in the picture that
6035 are below the threshold value.
6037 It accepts the following parameters:
6042 The percentage of the pixels that have to be below the threshold; it defaults to
6045 @item threshold, thresh
6046 The threshold below which a pixel value is considered black; it defaults to
6051 @section blend, tblend
6053 Blend two video frames into each other.
6055 The @code{blend} filter takes two input streams and outputs one
6056 stream, the first input is the "top" layer and second input is
6057 "bottom" layer. By default, the output terminates when the longest input terminates.
6059 The @code{tblend} (time blend) filter takes two consecutive frames
6060 from one single stream, and outputs the result obtained by blending
6061 the new frame on top of the old frame.
6063 A description of the accepted options follows.
6071 Set blend mode for specific pixel component or all pixel components in case
6072 of @var{all_mode}. Default value is @code{normal}.
6074 Available values for component modes are:
6116 Set blend opacity for specific pixel component or all pixel components in case
6117 of @var{all_opacity}. Only used in combination with pixel component blend modes.
6124 Set blend expression for specific pixel component or all pixel components in case
6125 of @var{all_expr}. Note that related mode options will be ignored if those are set.
6127 The expressions can use the following variables:
6131 The sequential number of the filtered frame, starting from @code{0}.
6135 the coordinates of the current sample
6139 the width and height of currently filtered plane
6143 Width and height scale for the plane being filtered. It is the
6144 ratio between the dimensions of the current plane to the luma plane,
6145 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
6146 the luma plane and @code{0.5,0.5} for the chroma planes.
6149 Time of the current frame, expressed in seconds.
6152 Value of pixel component at current location for first video frame (top layer).
6155 Value of pixel component at current location for second video frame (bottom layer).
6159 The @code{blend} filter also supports the @ref{framesync} options.
6161 @subsection Examples
6165 Apply transition from bottom layer to top layer in first 10 seconds:
6167 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6171 Apply linear horizontal transition from top layer to bottom layer:
6173 blend=all_expr='A*(X/W)+B*(1-X/W)'
6177 Apply 1x1 checkerboard effect:
6179 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6183 Apply uncover left effect:
6185 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6189 Apply uncover down effect:
6191 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6195 Apply uncover up-left effect:
6197 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6201 Split diagonally video and shows top and bottom layer on each side:
6203 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6207 Display differences between the current and the previous frame:
6209 tblend=all_mode=grainextract
6215 Denoise frames using Block-Matching 3D algorithm.
6217 The filter accepts the following options.
6221 Set denoising strength. Default value is 1.
6222 Allowed range is from 0 to 999.9.
6223 The denoising algorithm is very sensitive to sigma, so adjust it
6224 according to the source.
6227 Set local patch size. This sets dimensions in 2D.
6230 Set sliding step for processing blocks. Default value is 4.
6231 Allowed range is from 1 to 64.
6232 Smaller values allows processing more reference blocks and is slower.
6235 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6236 When set to 1, no block matching is done. Larger values allows more blocks
6238 Allowed range is from 1 to 256.
6241 Set radius for search block matching. Default is 9.
6242 Allowed range is from 1 to INT32_MAX.
6245 Set step between two search locations for block matching. Default is 1.
6246 Allowed range is from 1 to 64. Smaller is slower.
6249 Set threshold of mean square error for block matching. Valid range is 0 to
6253 Set thresholding parameter for hard thresholding in 3D transformed domain.
6254 Larger values results in stronger hard-thresholding filtering in frequency
6258 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6259 Default is @code{basic}.
6262 If enabled, filter will use 2nd stream for block matching.
6263 Default is disabled for @code{basic} value of @var{estim} option,
6264 and always enabled if value of @var{estim} is @code{final}.
6267 Set planes to filter. Default is all available except alpha.
6270 @subsection Examples
6274 Basic filtering with bm3d:
6276 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6280 Same as above, but filtering only luma:
6282 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6286 Same as above, but with both estimation modes:
6288 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
6292 Same as above, but prefilter with @ref{nlmeans} filter instead:
6294 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
6300 Apply a boxblur algorithm to the input video.
6302 It accepts the following parameters:
6306 @item luma_radius, lr
6307 @item luma_power, lp
6308 @item chroma_radius, cr
6309 @item chroma_power, cp
6310 @item alpha_radius, ar
6311 @item alpha_power, ap
6315 A description of the accepted options follows.
6318 @item luma_radius, lr
6319 @item chroma_radius, cr
6320 @item alpha_radius, ar
6321 Set an expression for the box radius in pixels used for blurring the
6322 corresponding input plane.
6324 The radius value must be a non-negative number, and must not be
6325 greater than the value of the expression @code{min(w,h)/2} for the
6326 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6329 Default value for @option{luma_radius} is "2". If not specified,
6330 @option{chroma_radius} and @option{alpha_radius} default to the
6331 corresponding value set for @option{luma_radius}.
6333 The expressions can contain the following constants:
6337 The input width and height in pixels.
6341 The input chroma image width and height in pixels.
6345 The horizontal and vertical chroma subsample values. For example, for the
6346 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6349 @item luma_power, lp
6350 @item chroma_power, cp
6351 @item alpha_power, ap
6352 Specify how many times the boxblur filter is applied to the
6353 corresponding plane.
6355 Default value for @option{luma_power} is 2. If not specified,
6356 @option{chroma_power} and @option{alpha_power} default to the
6357 corresponding value set for @option{luma_power}.
6359 A value of 0 will disable the effect.
6362 @subsection Examples
6366 Apply a boxblur filter with the luma, chroma, and alpha radii
6369 boxblur=luma_radius=2:luma_power=1
6374 Set the luma radius to 2, and alpha and chroma radius to 0:
6376 boxblur=2:1:cr=0:ar=0
6380 Set the luma and chroma radii to a fraction of the video dimension:
6382 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6388 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6389 Deinterlacing Filter").
6391 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6392 interpolation algorithms.
6393 It accepts the following parameters:
6397 The interlacing mode to adopt. It accepts one of the following values:
6401 Output one frame for each frame.
6403 Output one frame for each field.
6406 The default value is @code{send_field}.
6409 The picture field parity assumed for the input interlaced video. It accepts one
6410 of the following values:
6414 Assume the top field is first.
6416 Assume the bottom field is first.
6418 Enable automatic detection of field parity.
6421 The default value is @code{auto}.
6422 If the interlacing is unknown or the decoder does not export this information,
6423 top field first will be assumed.
6426 Specify which frames to deinterlace. Accept one of the following
6431 Deinterlace all frames.
6433 Only deinterlace frames marked as interlaced.
6436 The default value is @code{all}.
6440 Remove all color information for all colors except for certain one.
6442 The filter accepts the following options:
6446 The color which will not be replaced with neutral chroma.
6449 Similarity percentage with the above color.
6450 0.01 matches only the exact key color, while 1.0 matches everything.
6453 Signals that the color passed is already in YUV instead of RGB.
6455 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6456 This can be used to pass exact YUV values as hexadecimal numbers.
6460 YUV colorspace color/chroma keying.
6462 The filter accepts the following options:
6466 The color which will be replaced with transparency.
6469 Similarity percentage with the key color.
6471 0.01 matches only the exact key color, while 1.0 matches everything.
6476 0.0 makes pixels either fully transparent, or not transparent at all.
6478 Higher values result in semi-transparent pixels, with a higher transparency
6479 the more similar the pixels color is to the key color.
6482 Signals that the color passed is already in YUV instead of RGB.
6484 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6485 This can be used to pass exact YUV values as hexadecimal numbers.
6488 @subsection Examples
6492 Make every green pixel in the input image transparent:
6494 ffmpeg -i input.png -vf chromakey=green out.png
6498 Overlay a greenscreen-video on top of a static black background.
6500 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
6504 @section chromashift
6505 Shift chroma pixels horizontally and/or vertically.
6507 The filter accepts the following options:
6510 Set amount to shift chroma-blue horizontally.
6512 Set amount to shift chroma-blue vertically.
6514 Set amount to shift chroma-red horizontally.
6516 Set amount to shift chroma-red vertically.
6518 Set edge mode, can be @var{smear}, default, or @var{warp}.
6523 Display CIE color diagram with pixels overlaid onto it.
6525 The filter accepts the following options:
6540 @item uhdtv, rec2020
6553 Set what gamuts to draw.
6555 See @code{system} option for available values.
6558 Set ciescope size, by default set to 512.
6561 Set intensity used to map input pixel values to CIE diagram.
6564 Set contrast used to draw tongue colors that are out of active color system gamut.
6567 Correct gamma displayed on scope, by default enabled.
6570 Show white point on CIE diagram, by default disabled.
6573 Set input gamma. Used only with XYZ input color space.
6578 Visualize information exported by some codecs.
6580 Some codecs can export information through frames using side-data or other
6581 means. For example, some MPEG based codecs export motion vectors through the
6582 @var{export_mvs} flag in the codec @option{flags2} option.
6584 The filter accepts the following option:
6588 Set motion vectors to visualize.
6590 Available flags for @var{mv} are:
6594 forward predicted MVs of P-frames
6596 forward predicted MVs of B-frames
6598 backward predicted MVs of B-frames
6602 Display quantization parameters using the chroma planes.
6605 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6607 Available flags for @var{mv_type} are:
6611 forward predicted MVs
6613 backward predicted MVs
6616 @item frame_type, ft
6617 Set frame type to visualize motion vectors of.
6619 Available flags for @var{frame_type} are:
6623 intra-coded frames (I-frames)
6625 predicted frames (P-frames)
6627 bi-directionally predicted frames (B-frames)
6631 @subsection Examples
6635 Visualize forward predicted MVs of all frames using @command{ffplay}:
6637 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6641 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6643 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6647 @section colorbalance
6648 Modify intensity of primary colors (red, green and blue) of input frames.
6650 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6651 regions for the red-cyan, green-magenta or blue-yellow balance.
6653 A positive adjustment value shifts the balance towards the primary color, a negative
6654 value towards the complementary color.
6656 The filter accepts the following options:
6662 Adjust red, green and blue shadows (darkest pixels).
6667 Adjust red, green and blue midtones (medium pixels).
6672 Adjust red, green and blue highlights (brightest pixels).
6674 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6677 @subsection Examples
6681 Add red color cast to shadows:
6688 RGB colorspace color keying.
6690 The filter accepts the following options:
6694 The color which will be replaced with transparency.
6697 Similarity percentage with the key color.
6699 0.01 matches only the exact key color, while 1.0 matches everything.
6704 0.0 makes pixels either fully transparent, or not transparent at all.
6706 Higher values result in semi-transparent pixels, with a higher transparency
6707 the more similar the pixels color is to the key color.
6710 @subsection Examples
6714 Make every green pixel in the input image transparent:
6716 ffmpeg -i input.png -vf colorkey=green out.png
6720 Overlay a greenscreen-video on top of a static background image.
6722 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
6726 @section colorlevels
6728 Adjust video input frames using levels.
6730 The filter accepts the following options:
6737 Adjust red, green, blue and alpha input black point.
6738 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6744 Adjust red, green, blue and alpha input white point.
6745 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6747 Input levels are used to lighten highlights (bright tones), darken shadows
6748 (dark tones), change the balance of bright and dark tones.
6754 Adjust red, green, blue and alpha output black point.
6755 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6761 Adjust red, green, blue and alpha output white point.
6762 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6764 Output levels allows manual selection of a constrained output level range.
6767 @subsection Examples
6771 Make video output darker:
6773 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6779 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6783 Make video output lighter:
6785 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6789 Increase brightness:
6791 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6795 @section colorchannelmixer
6797 Adjust video input frames by re-mixing color channels.
6799 This filter modifies a color channel by adding the values associated to
6800 the other channels of the same pixels. For example if the value to
6801 modify is red, the output value will be:
6803 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6806 The filter accepts the following options:
6813 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6814 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6820 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6821 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6827 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6828 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6834 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6835 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6837 Allowed ranges for options are @code{[-2.0, 2.0]}.
6840 @subsection Examples
6844 Convert source to grayscale:
6846 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6849 Simulate sepia tones:
6851 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6855 @section colormatrix
6857 Convert color matrix.
6859 The filter accepts the following options:
6864 Specify the source and destination color matrix. Both values must be
6867 The accepted values are:
6895 For example to convert from BT.601 to SMPTE-240M, use the command:
6897 colormatrix=bt601:smpte240m
6902 Convert colorspace, transfer characteristics or color primaries.
6903 Input video needs to have an even size.
6905 The filter accepts the following options:
6910 Specify all color properties at once.
6912 The accepted values are:
6942 Specify output colorspace.
6944 The accepted values are:
6953 BT.470BG or BT.601-6 625
6956 SMPTE-170M or BT.601-6 525
6965 BT.2020 with non-constant luminance
6971 Specify output transfer characteristics.
6973 The accepted values are:
6985 Constant gamma of 2.2
6988 Constant gamma of 2.8
6991 SMPTE-170M, BT.601-6 625 or BT.601-6 525
7009 BT.2020 for 10-bits content
7012 BT.2020 for 12-bits content
7018 Specify output color primaries.
7020 The accepted values are:
7029 BT.470BG or BT.601-6 625
7032 SMPTE-170M or BT.601-6 525
7056 Specify output color range.
7058 The accepted values are:
7061 TV (restricted) range
7064 MPEG (restricted) range
7075 Specify output color format.
7077 The accepted values are:
7080 YUV 4:2:0 planar 8-bits
7083 YUV 4:2:0 planar 10-bits
7086 YUV 4:2:0 planar 12-bits
7089 YUV 4:2:2 planar 8-bits
7092 YUV 4:2:2 planar 10-bits
7095 YUV 4:2:2 planar 12-bits
7098 YUV 4:4:4 planar 8-bits
7101 YUV 4:4:4 planar 10-bits
7104 YUV 4:4:4 planar 12-bits
7109 Do a fast conversion, which skips gamma/primary correction. This will take
7110 significantly less CPU, but will be mathematically incorrect. To get output
7111 compatible with that produced by the colormatrix filter, use fast=1.
7114 Specify dithering mode.
7116 The accepted values are:
7122 Floyd-Steinberg dithering
7126 Whitepoint adaptation mode.
7128 The accepted values are:
7131 Bradford whitepoint adaptation
7134 von Kries whitepoint adaptation
7137 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7141 Override all input properties at once. Same accepted values as @ref{all}.
7144 Override input colorspace. Same accepted values as @ref{space}.
7147 Override input color primaries. Same accepted values as @ref{primaries}.
7150 Override input transfer characteristics. Same accepted values as @ref{trc}.
7153 Override input color range. Same accepted values as @ref{range}.
7157 The filter converts the transfer characteristics, color space and color
7158 primaries to the specified user values. The output value, if not specified,
7159 is set to a default value based on the "all" property. If that property is
7160 also not specified, the filter will log an error. The output color range and
7161 format default to the same value as the input color range and format. The
7162 input transfer characteristics, color space, color primaries and color range
7163 should be set on the input data. If any of these are missing, the filter will
7164 log an error and no conversion will take place.
7166 For example to convert the input to SMPTE-240M, use the command:
7168 colorspace=smpte240m
7171 @section convolution
7173 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7175 The filter accepts the following options:
7182 Set matrix for each plane.
7183 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7184 and from 1 to 49 odd number of signed integers in @var{row} mode.
7190 Set multiplier for calculated value for each plane.
7191 If unset or 0, it will be sum of all matrix elements.
7197 Set bias for each plane. This value is added to the result of the multiplication.
7198 Useful for making the overall image brighter or darker. Default is 0.0.
7204 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7205 Default is @var{square}.
7208 @subsection Examples
7214 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"
7220 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"
7226 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"
7232 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"
7236 Apply laplacian edge detector which includes diagonals:
7238 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"
7244 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"
7250 Apply 2D convolution of video stream in frequency domain using second stream
7253 The filter accepts the following options:
7257 Set which planes to process.
7260 Set which impulse video frames will be processed, can be @var{first}
7261 or @var{all}. Default is @var{all}.
7264 The @code{convolve} filter also supports the @ref{framesync} options.
7268 Copy the input video source unchanged to the output. This is mainly useful for
7273 Video filtering on GPU using Apple's CoreImage API on OSX.
7275 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7276 processed by video hardware. However, software-based OpenGL implementations
7277 exist which means there is no guarantee for hardware processing. It depends on
7280 There are many filters and image generators provided by Apple that come with a
7281 large variety of options. The filter has to be referenced by its name along
7284 The coreimage filter accepts the following options:
7287 List all available filters and generators along with all their respective
7288 options as well as possible minimum and maximum values along with the default
7295 Specify all filters by their respective name and options.
7296 Use @var{list_filters} to determine all valid filter names and options.
7297 Numerical options are specified by a float value and are automatically clamped
7298 to their respective value range. Vector and color options have to be specified
7299 by a list of space separated float values. Character escaping has to be done.
7300 A special option name @code{default} is available to use default options for a
7303 It is required to specify either @code{default} or at least one of the filter options.
7304 All omitted options are used with their default values.
7305 The syntax of the filter string is as follows:
7307 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7311 Specify a rectangle where the output of the filter chain is copied into the
7312 input image. It is given by a list of space separated float values:
7314 output_rect=x\ y\ width\ height
7316 If not given, the output rectangle equals the dimensions of the input image.
7317 The output rectangle is automatically cropped at the borders of the input
7318 image. Negative values are valid for each component.
7320 output_rect=25\ 25\ 100\ 100
7324 Several filters can be chained for successive processing without GPU-HOST
7325 transfers allowing for fast processing of complex filter chains.
7326 Currently, only filters with zero (generators) or exactly one (filters) input
7327 image and one output image are supported. Also, transition filters are not yet
7330 Some filters generate output images with additional padding depending on the
7331 respective filter kernel. The padding is automatically removed to ensure the
7332 filter output has the same size as the input image.
7334 For image generators, the size of the output image is determined by the
7335 previous output image of the filter chain or the input image of the whole
7336 filterchain, respectively. The generators do not use the pixel information of
7337 this image to generate their output. However, the generated output is
7338 blended onto this image, resulting in partial or complete coverage of the
7341 The @ref{coreimagesrc} video source can be used for generating input images
7342 which are directly fed into the filter chain. By using it, providing input
7343 images by another video source or an input video is not required.
7345 @subsection Examples
7350 List all filters available:
7352 coreimage=list_filters=true
7356 Use the CIBoxBlur filter with default options to blur an image:
7358 coreimage=filter=CIBoxBlur@@default
7362 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7363 its center at 100x100 and a radius of 50 pixels:
7365 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7369 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7370 given as complete and escaped command-line for Apple's standard bash shell:
7372 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7378 Crop the input video to given dimensions.
7380 It accepts the following parameters:
7384 The width of the output video. It defaults to @code{iw}.
7385 This expression is evaluated only once during the filter
7386 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7389 The height of the output video. It defaults to @code{ih}.
7390 This expression is evaluated only once during the filter
7391 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7394 The horizontal position, in the input video, of the left edge of the output
7395 video. It defaults to @code{(in_w-out_w)/2}.
7396 This expression is evaluated per-frame.
7399 The vertical position, in the input video, of the top edge of the output video.
7400 It defaults to @code{(in_h-out_h)/2}.
7401 This expression is evaluated per-frame.
7404 If set to 1 will force the output display aspect ratio
7405 to be the same of the input, by changing the output sample aspect
7406 ratio. It defaults to 0.
7409 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7410 width/height/x/y as specified and will not be rounded to nearest smaller value.
7414 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7415 expressions containing the following constants:
7420 The computed values for @var{x} and @var{y}. They are evaluated for
7425 The input width and height.
7429 These are the same as @var{in_w} and @var{in_h}.
7433 The output (cropped) width and height.
7437 These are the same as @var{out_w} and @var{out_h}.
7440 same as @var{iw} / @var{ih}
7443 input sample aspect ratio
7446 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7450 horizontal and vertical chroma subsample values. For example for the
7451 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7454 The number of the input frame, starting from 0.
7457 the position in the file of the input frame, NAN if unknown
7460 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7464 The expression for @var{out_w} may depend on the value of @var{out_h},
7465 and the expression for @var{out_h} may depend on @var{out_w}, but they
7466 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7467 evaluated after @var{out_w} and @var{out_h}.
7469 The @var{x} and @var{y} parameters specify the expressions for the
7470 position of the top-left corner of the output (non-cropped) area. They
7471 are evaluated for each frame. If the evaluated value is not valid, it
7472 is approximated to the nearest valid value.
7474 The expression for @var{x} may depend on @var{y}, and the expression
7475 for @var{y} may depend on @var{x}.
7477 @subsection Examples
7481 Crop area with size 100x100 at position (12,34).
7486 Using named options, the example above becomes:
7488 crop=w=100:h=100:x=12:y=34
7492 Crop the central input area with size 100x100:
7498 Crop the central input area with size 2/3 of the input video:
7500 crop=2/3*in_w:2/3*in_h
7504 Crop the input video central square:
7511 Delimit the rectangle with the top-left corner placed at position
7512 100:100 and the right-bottom corner corresponding to the right-bottom
7513 corner of the input image.
7515 crop=in_w-100:in_h-100:100:100
7519 Crop 10 pixels from the left and right borders, and 20 pixels from
7520 the top and bottom borders
7522 crop=in_w-2*10:in_h-2*20
7526 Keep only the bottom right quarter of the input image:
7528 crop=in_w/2:in_h/2:in_w/2:in_h/2
7532 Crop height for getting Greek harmony:
7534 crop=in_w:1/PHI*in_w
7538 Apply trembling effect:
7540 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)
7544 Apply erratic camera effect depending on timestamp:
7546 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)"
7550 Set x depending on the value of y:
7552 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7556 @subsection Commands
7558 This filter supports the following commands:
7564 Set width/height of the output video and the horizontal/vertical position
7566 The command accepts the same syntax of the corresponding option.
7568 If the specified expression is not valid, it is kept at its current
7574 Auto-detect the crop size.
7576 It calculates the necessary cropping parameters and prints the
7577 recommended parameters via the logging system. The detected dimensions
7578 correspond to the non-black area of the input video.
7580 It accepts the following parameters:
7585 Set higher black value threshold, which can be optionally specified
7586 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7587 value greater to the set value is considered non-black. It defaults to 24.
7588 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7589 on the bitdepth of the pixel format.
7592 The value which the width/height should be divisible by. It defaults to
7593 16. The offset is automatically adjusted to center the video. Use 2 to
7594 get only even dimensions (needed for 4:2:2 video). 16 is best when
7595 encoding to most video codecs.
7597 @item reset_count, reset
7598 Set the counter that determines after how many frames cropdetect will
7599 reset the previously detected largest video area and start over to
7600 detect the current optimal crop area. Default value is 0.
7602 This can be useful when channel logos distort the video area. 0
7603 indicates 'never reset', and returns the largest area encountered during
7610 Delay video filtering until a given wallclock timestamp. The filter first
7611 passes on @option{preroll} amount of frames, then it buffers at most
7612 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7613 it forwards the buffered frames and also any subsequent frames coming in its
7616 The filter can be used synchronize the output of multiple ffmpeg processes for
7617 realtime output devices like decklink. By putting the delay in the filtering
7618 chain and pre-buffering frames the process can pass on data to output almost
7619 immediately after the target wallclock timestamp is reached.
7621 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7627 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7630 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7633 The maximum duration of content to buffer before waiting for the cue expressed
7634 in seconds. Default is 0.
7641 Apply color adjustments using curves.
7643 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7644 component (red, green and blue) has its values defined by @var{N} key points
7645 tied from each other using a smooth curve. The x-axis represents the pixel
7646 values from the input frame, and the y-axis the new pixel values to be set for
7649 By default, a component curve is defined by the two points @var{(0;0)} and
7650 @var{(1;1)}. This creates a straight line where each original pixel value is
7651 "adjusted" to its own value, which means no change to the image.
7653 The filter allows you to redefine these two points and add some more. A new
7654 curve (using a natural cubic spline interpolation) will be define to pass
7655 smoothly through all these new coordinates. The new defined points needs to be
7656 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7657 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7658 the vector spaces, the values will be clipped accordingly.
7660 The filter accepts the following options:
7664 Select one of the available color presets. This option can be used in addition
7665 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7666 options takes priority on the preset values.
7667 Available presets are:
7670 @item color_negative
7673 @item increase_contrast
7675 @item linear_contrast
7676 @item medium_contrast
7678 @item strong_contrast
7681 Default is @code{none}.
7683 Set the master key points. These points will define a second pass mapping. It
7684 is sometimes called a "luminance" or "value" mapping. It can be used with
7685 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7686 post-processing LUT.
7688 Set the key points for the red component.
7690 Set the key points for the green component.
7692 Set the key points for the blue component.
7694 Set the key points for all components (not including master).
7695 Can be used in addition to the other key points component
7696 options. In this case, the unset component(s) will fallback on this
7697 @option{all} setting.
7699 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7701 Save Gnuplot script of the curves in specified file.
7704 To avoid some filtergraph syntax conflicts, each key points list need to be
7705 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7707 @subsection Examples
7711 Increase slightly the middle level of blue:
7713 curves=blue='0/0 0.5/0.58 1/1'
7719 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'
7721 Here we obtain the following coordinates for each components:
7724 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7726 @code{(0;0) (0.50;0.48) (1;1)}
7728 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7732 The previous example can also be achieved with the associated built-in preset:
7734 curves=preset=vintage
7744 Use a Photoshop preset and redefine the points of the green component:
7746 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7750 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7751 and @command{gnuplot}:
7753 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7754 gnuplot -p /tmp/curves.plt
7760 Video data analysis filter.
7762 This filter shows hexadecimal pixel values of part of video.
7764 The filter accepts the following options:
7768 Set output video size.
7771 Set x offset from where to pick pixels.
7774 Set y offset from where to pick pixels.
7777 Set scope mode, can be one of the following:
7780 Draw hexadecimal pixel values with white color on black background.
7783 Draw hexadecimal pixel values with input video pixel color on black
7787 Draw hexadecimal pixel values on color background picked from input video,
7788 the text color is picked in such way so its always visible.
7792 Draw rows and columns numbers on left and top of video.
7795 Set background opacity.
7800 Denoise frames using 2D DCT (frequency domain filtering).
7802 This filter is not designed for real time.
7804 The filter accepts the following options:
7808 Set the noise sigma constant.
7810 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7811 coefficient (absolute value) below this threshold with be dropped.
7813 If you need a more advanced filtering, see @option{expr}.
7815 Default is @code{0}.
7818 Set number overlapping pixels for each block. Since the filter can be slow, you
7819 may want to reduce this value, at the cost of a less effective filter and the
7820 risk of various artefacts.
7822 If the overlapping value doesn't permit processing the whole input width or
7823 height, a warning will be displayed and according borders won't be denoised.
7825 Default value is @var{blocksize}-1, which is the best possible setting.
7828 Set the coefficient factor expression.
7830 For each coefficient of a DCT block, this expression will be evaluated as a
7831 multiplier value for the coefficient.
7833 If this is option is set, the @option{sigma} option will be ignored.
7835 The absolute value of the coefficient can be accessed through the @var{c}
7839 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7840 @var{blocksize}, which is the width and height of the processed blocks.
7842 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7843 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7844 on the speed processing. Also, a larger block size does not necessarily means a
7848 @subsection Examples
7850 Apply a denoise with a @option{sigma} of @code{4.5}:
7855 The same operation can be achieved using the expression system:
7857 dctdnoiz=e='gte(c, 4.5*3)'
7860 Violent denoise using a block size of @code{16x16}:
7867 Remove banding artifacts from input video.
7868 It works by replacing banded pixels with average value of referenced pixels.
7870 The filter accepts the following options:
7877 Set banding detection threshold for each plane. Default is 0.02.
7878 Valid range is 0.00003 to 0.5.
7879 If difference between current pixel and reference pixel is less than threshold,
7880 it will be considered as banded.
7883 Banding detection range in pixels. Default is 16. If positive, random number
7884 in range 0 to set value will be used. If negative, exact absolute value
7886 The range defines square of four pixels around current pixel.
7889 Set direction in radians from which four pixel will be compared. If positive,
7890 random direction from 0 to set direction will be picked. If negative, exact of
7891 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7892 will pick only pixels on same row and -PI/2 will pick only pixels on same
7896 If enabled, current pixel is compared with average value of all four
7897 surrounding pixels. The default is enabled. If disabled current pixel is
7898 compared with all four surrounding pixels. The pixel is considered banded
7899 if only all four differences with surrounding pixels are less than threshold.
7902 If enabled, current pixel is changed if and only if all pixel components are banded,
7903 e.g. banding detection threshold is triggered for all color components.
7904 The default is disabled.
7909 Remove blocking artifacts from input video.
7911 The filter accepts the following options:
7915 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7916 This controls what kind of deblocking is applied.
7919 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7925 Set blocking detection thresholds. Allowed range is 0 to 1.
7926 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7927 Using higher threshold gives more deblocking strength.
7928 Setting @var{alpha} controls threshold detection at exact edge of block.
7929 Remaining options controls threshold detection near the edge. Each one for
7930 below/above or left/right. Setting any of those to @var{0} disables
7934 Set planes to filter. Default is to filter all available planes.
7937 @subsection Examples
7941 Deblock using weak filter and block size of 4 pixels.
7943 deblock=filter=weak:block=4
7947 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7948 deblocking more edges.
7950 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7954 Similar as above, but filter only first plane.
7956 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7960 Similar as above, but filter only second and third plane.
7962 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7969 Drop duplicated frames at regular intervals.
7971 The filter accepts the following options:
7975 Set the number of frames from which one will be dropped. Setting this to
7976 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7977 Default is @code{5}.
7980 Set the threshold for duplicate detection. If the difference metric for a frame
7981 is less than or equal to this value, then it is declared as duplicate. Default
7985 Set scene change threshold. Default is @code{15}.
7989 Set the size of the x and y-axis blocks used during metric calculations.
7990 Larger blocks give better noise suppression, but also give worse detection of
7991 small movements. Must be a power of two. Default is @code{32}.
7994 Mark main input as a pre-processed input and activate clean source input
7995 stream. This allows the input to be pre-processed with various filters to help
7996 the metrics calculation while keeping the frame selection lossless. When set to
7997 @code{1}, the first stream is for the pre-processed input, and the second
7998 stream is the clean source from where the kept frames are chosen. Default is
8002 Set whether or not chroma is considered in the metric calculations. Default is
8008 Apply 2D deconvolution of video stream in frequency domain using second stream
8011 The filter accepts the following options:
8015 Set which planes to process.
8018 Set which impulse video frames will be processed, can be @var{first}
8019 or @var{all}. Default is @var{all}.
8022 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
8023 and height are not same and not power of 2 or if stream prior to convolving
8027 The @code{deconvolve} filter also supports the @ref{framesync} options.
8031 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
8033 It accepts the following options:
8037 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
8038 @var{rainbows} for cross-color reduction.
8041 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
8044 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
8047 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
8050 Set temporal chroma threshold. Lower values increases reduction of cross-color.
8055 Apply deflate effect to the video.
8057 This filter replaces the pixel by the local(3x3) average by taking into account
8058 only values lower than the pixel.
8060 It accepts the following options:
8067 Limit the maximum change for each plane, default is 65535.
8068 If 0, plane will remain unchanged.
8073 Remove temporal frame luminance variations.
8075 It accepts the following options:
8079 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
8082 Set averaging mode to smooth temporal luminance variations.
8084 Available values are:
8109 Do not actually modify frame. Useful when one only wants metadata.
8114 Remove judder produced by partially interlaced telecined content.
8116 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
8117 source was partially telecined content then the output of @code{pullup,dejudder}
8118 will have a variable frame rate. May change the recorded frame rate of the
8119 container. Aside from that change, this filter will not affect constant frame
8122 The option available in this filter is:
8126 Specify the length of the window over which the judder repeats.
8128 Accepts any integer greater than 1. Useful values are:
8132 If the original was telecined from 24 to 30 fps (Film to NTSC).
8135 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8138 If a mixture of the two.
8141 The default is @samp{4}.
8146 Suppress a TV station logo by a simple interpolation of the surrounding
8147 pixels. Just set a rectangle covering the logo and watch it disappear
8148 (and sometimes something even uglier appear - your mileage may vary).
8150 It accepts the following parameters:
8155 Specify the top left corner coordinates of the logo. They must be
8160 Specify the width and height of the logo to clear. They must be
8164 Specify the thickness of the fuzzy edge of the rectangle (added to
8165 @var{w} and @var{h}). The default value is 1. This option is
8166 deprecated, setting higher values should no longer be necessary and
8170 When set to 1, a green rectangle is drawn on the screen to simplify
8171 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8172 The default value is 0.
8174 The rectangle is drawn on the outermost pixels which will be (partly)
8175 replaced with interpolated values. The values of the next pixels
8176 immediately outside this rectangle in each direction will be used to
8177 compute the interpolated pixel values inside the rectangle.
8181 @subsection Examples
8185 Set a rectangle covering the area with top left corner coordinates 0,0
8186 and size 100x77, and a band of size 10:
8188 delogo=x=0:y=0:w=100:h=77:band=10
8195 Attempt to fix small changes in horizontal and/or vertical shift. This
8196 filter helps remove camera shake from hand-holding a camera, bumping a
8197 tripod, moving on a vehicle, etc.
8199 The filter accepts the following options:
8207 Specify a rectangular area where to limit the search for motion
8209 If desired the search for motion vectors can be limited to a
8210 rectangular area of the frame defined by its top left corner, width
8211 and height. These parameters have the same meaning as the drawbox
8212 filter which can be used to visualise the position of the bounding
8215 This is useful when simultaneous movement of subjects within the frame
8216 might be confused for camera motion by the motion vector search.
8218 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8219 then the full frame is used. This allows later options to be set
8220 without specifying the bounding box for the motion vector search.
8222 Default - search the whole frame.
8226 Specify the maximum extent of movement in x and y directions in the
8227 range 0-64 pixels. Default 16.
8230 Specify how to generate pixels to fill blanks at the edge of the
8231 frame. Available values are:
8234 Fill zeroes at blank locations
8236 Original image at blank locations
8238 Extruded edge value at blank locations
8240 Mirrored edge at blank locations
8242 Default value is @samp{mirror}.
8245 Specify the blocksize to use for motion search. Range 4-128 pixels,
8249 Specify the contrast threshold for blocks. Only blocks with more than
8250 the specified contrast (difference between darkest and lightest
8251 pixels) will be considered. Range 1-255, default 125.
8254 Specify the search strategy. Available values are:
8257 Set exhaustive search
8259 Set less exhaustive search.
8261 Default value is @samp{exhaustive}.
8264 If set then a detailed log of the motion search is written to the
8271 Remove unwanted contamination of foreground colors, caused by reflected color of
8272 greenscreen or bluescreen.
8274 This filter accepts the following options:
8278 Set what type of despill to use.
8281 Set how spillmap will be generated.
8284 Set how much to get rid of still remaining spill.
8287 Controls amount of red in spill area.
8290 Controls amount of green in spill area.
8291 Should be -1 for greenscreen.
8294 Controls amount of blue in spill area.
8295 Should be -1 for bluescreen.
8298 Controls brightness of spill area, preserving colors.
8301 Modify alpha from generated spillmap.
8306 Apply an exact inverse of the telecine operation. It requires a predefined
8307 pattern specified using the pattern option which must be the same as that passed
8308 to the telecine filter.
8310 This filter accepts the following options:
8319 The default value is @code{top}.
8323 A string of numbers representing the pulldown pattern you wish to apply.
8324 The default value is @code{23}.
8327 A number representing position of the first frame with respect to the telecine
8328 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8333 Apply dilation effect to the video.
8335 This filter replaces the pixel by the local(3x3) maximum.
8337 It accepts the following options:
8344 Limit the maximum change for each plane, default is 65535.
8345 If 0, plane will remain unchanged.
8348 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8351 Flags to local 3x3 coordinates maps like this:
8360 Displace pixels as indicated by second and third input stream.
8362 It takes three input streams and outputs one stream, the first input is the
8363 source, and second and third input are displacement maps.
8365 The second input specifies how much to displace pixels along the
8366 x-axis, while the third input specifies how much to displace pixels
8368 If one of displacement map streams terminates, last frame from that
8369 displacement map will be used.
8371 Note that once generated, displacements maps can be reused over and over again.
8373 A description of the accepted options follows.
8377 Set displace behavior for pixels that are out of range.
8379 Available values are:
8382 Missing pixels are replaced by black pixels.
8385 Adjacent pixels will spread out to replace missing pixels.
8388 Out of range pixels are wrapped so they point to pixels of other side.
8391 Out of range pixels will be replaced with mirrored pixels.
8393 Default is @samp{smear}.
8397 @subsection Examples
8401 Add ripple effect to rgb input of video size hd720:
8403 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
8407 Add wave effect to rgb input of video size hd720:
8409 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
8415 Draw a colored box on the input image.
8417 It accepts the following parameters:
8422 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8426 The expressions which specify the width and height of the box; if 0 they are interpreted as
8427 the input width and height. It defaults to 0.
8430 Specify the color of the box to write. For the general syntax of this option,
8431 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8432 value @code{invert} is used, the box edge color is the same as the
8433 video with inverted luma.
8436 The expression which sets the thickness of the box edge.
8437 A value of @code{fill} will create a filled box. Default value is @code{3}.
8439 See below for the list of accepted constants.
8442 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8443 will overwrite the video's color and alpha pixels.
8444 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8447 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8448 following constants:
8452 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8456 horizontal and vertical chroma subsample values. For example for the
8457 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8461 The input width and height.
8464 The input sample aspect ratio.
8468 The x and y offset coordinates where the box is drawn.
8472 The width and height of the drawn box.
8475 The thickness of the drawn box.
8477 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8478 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8482 @subsection Examples
8486 Draw a black box around the edge of the input image:
8492 Draw a box with color red and an opacity of 50%:
8494 drawbox=10:20:200:60:red@@0.5
8497 The previous example can be specified as:
8499 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8503 Fill the box with pink color:
8505 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8509 Draw a 2-pixel red 2.40:1 mask:
8511 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
8517 Draw a grid on the input image.
8519 It accepts the following parameters:
8524 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8528 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8529 input width and height, respectively, minus @code{thickness}, so image gets
8530 framed. Default to 0.
8533 Specify the color of the grid. For the general syntax of this option,
8534 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8535 value @code{invert} is used, the grid color is the same as the
8536 video with inverted luma.
8539 The expression which sets the thickness of the grid line. Default value is @code{1}.
8541 See below for the list of accepted constants.
8544 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8545 will overwrite the video's color and alpha pixels.
8546 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8549 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8550 following constants:
8554 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8558 horizontal and vertical chroma subsample values. For example for the
8559 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8563 The input grid cell width and height.
8566 The input sample aspect ratio.
8570 The x and y coordinates of some point of grid intersection (meant to configure offset).
8574 The width and height of the drawn cell.
8577 The thickness of the drawn cell.
8579 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8580 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8584 @subsection Examples
8588 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8590 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8594 Draw a white 3x3 grid with an opacity of 50%:
8596 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8603 Draw a text string or text from a specified file on top of a video, using the
8604 libfreetype library.
8606 To enable compilation of this filter, you need to configure FFmpeg with
8607 @code{--enable-libfreetype}.
8608 To enable default font fallback and the @var{font} option you need to
8609 configure FFmpeg with @code{--enable-libfontconfig}.
8610 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8611 @code{--enable-libfribidi}.
8615 It accepts the following parameters:
8620 Used to draw a box around text using the background color.
8621 The value must be either 1 (enable) or 0 (disable).
8622 The default value of @var{box} is 0.
8625 Set the width of the border to be drawn around the box using @var{boxcolor}.
8626 The default value of @var{boxborderw} is 0.
8629 The color to be used for drawing box around text. For the syntax of this
8630 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8632 The default value of @var{boxcolor} is "white".
8635 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8636 The default value of @var{line_spacing} is 0.
8639 Set the width of the border to be drawn around the text using @var{bordercolor}.
8640 The default value of @var{borderw} is 0.
8643 Set the color to be used for drawing border around text. For the syntax of this
8644 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8646 The default value of @var{bordercolor} is "black".
8649 Select how the @var{text} is expanded. Can be either @code{none},
8650 @code{strftime} (deprecated) or
8651 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8655 Set a start time for the count. Value is in microseconds. Only applied
8656 in the deprecated strftime expansion mode. To emulate in normal expansion
8657 mode use the @code{pts} function, supplying the start time (in seconds)
8658 as the second argument.
8661 If true, check and fix text coords to avoid clipping.
8664 The color to be used for drawing fonts. For the syntax of this option, check
8665 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8667 The default value of @var{fontcolor} is "black".
8669 @item fontcolor_expr
8670 String which is expanded the same way as @var{text} to obtain dynamic
8671 @var{fontcolor} value. By default this option has empty value and is not
8672 processed. When this option is set, it overrides @var{fontcolor} option.
8675 The font family to be used for drawing text. By default Sans.
8678 The font file to be used for drawing text. The path must be included.
8679 This parameter is mandatory if the fontconfig support is disabled.
8682 Draw the text applying alpha blending. The value can
8683 be a number between 0.0 and 1.0.
8684 The expression accepts the same variables @var{x, y} as well.
8685 The default value is 1.
8686 Please see @var{fontcolor_expr}.
8689 The font size to be used for drawing text.
8690 The default value of @var{fontsize} is 16.
8693 If set to 1, attempt to shape the text (for example, reverse the order of
8694 right-to-left text and join Arabic characters) before drawing it.
8695 Otherwise, just draw the text exactly as given.
8696 By default 1 (if supported).
8699 The flags to be used for loading the fonts.
8701 The flags map the corresponding flags supported by libfreetype, and are
8702 a combination of the following values:
8709 @item vertical_layout
8710 @item force_autohint
8713 @item ignore_global_advance_width
8715 @item ignore_transform
8721 Default value is "default".
8723 For more information consult the documentation for the FT_LOAD_*
8727 The color to be used for drawing a shadow behind the drawn text. For the
8728 syntax of this option, check the @ref{color syntax,,"Color" section in the
8729 ffmpeg-utils manual,ffmpeg-utils}.
8731 The default value of @var{shadowcolor} is "black".
8735 The x and y offsets for the text shadow position with respect to the
8736 position of the text. They can be either positive or negative
8737 values. The default value for both is "0".
8740 The starting frame number for the n/frame_num variable. The default value
8744 The size in number of spaces to use for rendering the tab.
8748 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8749 format. It can be used with or without text parameter. @var{timecode_rate}
8750 option must be specified.
8752 @item timecode_rate, rate, r
8753 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8754 integer. Minimum value is "1".
8755 Drop-frame timecode is supported for frame rates 30 & 60.
8758 If set to 1, the output of the timecode option will wrap around at 24 hours.
8759 Default is 0 (disabled).
8762 The text string to be drawn. The text must be a sequence of UTF-8
8764 This parameter is mandatory if no file is specified with the parameter
8768 A text file containing text to be drawn. The text must be a sequence
8769 of UTF-8 encoded characters.
8771 This parameter is mandatory if no text string is specified with the
8772 parameter @var{text}.
8774 If both @var{text} and @var{textfile} are specified, an error is thrown.
8777 If set to 1, the @var{textfile} will be reloaded before each frame.
8778 Be sure to update it atomically, or it may be read partially, or even fail.
8782 The expressions which specify the offsets where text will be drawn
8783 within the video frame. They are relative to the top/left border of the
8786 The default value of @var{x} and @var{y} is "0".
8788 See below for the list of accepted constants and functions.
8791 The parameters for @var{x} and @var{y} are expressions containing the
8792 following constants and functions:
8796 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8800 horizontal and vertical chroma subsample values. For example for the
8801 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8804 the height of each text line
8812 @item max_glyph_a, ascent
8813 the maximum distance from the baseline to the highest/upper grid
8814 coordinate used to place a glyph outline point, for all the rendered
8816 It is a positive value, due to the grid's orientation with the Y axis
8819 @item max_glyph_d, descent
8820 the maximum distance from the baseline to the lowest grid coordinate
8821 used to place a glyph outline point, for all the rendered glyphs.
8822 This is a negative value, due to the grid's orientation, with the Y axis
8826 maximum glyph height, that is the maximum height for all the glyphs
8827 contained in the rendered text, it is equivalent to @var{ascent} -
8831 maximum glyph width, that is the maximum width for all the glyphs
8832 contained in the rendered text
8835 the number of input frame, starting from 0
8837 @item rand(min, max)
8838 return a random number included between @var{min} and @var{max}
8841 The input sample aspect ratio.
8844 timestamp expressed in seconds, NAN if the input timestamp is unknown
8847 the height of the rendered text
8850 the width of the rendered text
8854 the x and y offset coordinates where the text is drawn.
8856 These parameters allow the @var{x} and @var{y} expressions to refer
8857 each other, so you can for example specify @code{y=x/dar}.
8860 @anchor{drawtext_expansion}
8861 @subsection Text expansion
8863 If @option{expansion} is set to @code{strftime},
8864 the filter recognizes strftime() sequences in the provided text and
8865 expands them accordingly. Check the documentation of strftime(). This
8866 feature is deprecated.
8868 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8870 If @option{expansion} is set to @code{normal} (which is the default),
8871 the following expansion mechanism is used.
8873 The backslash character @samp{\}, followed by any character, always expands to
8874 the second character.
8876 Sequences of the form @code{%@{...@}} are expanded. The text between the
8877 braces is a function name, possibly followed by arguments separated by ':'.
8878 If the arguments contain special characters or delimiters (':' or '@}'),
8879 they should be escaped.
8881 Note that they probably must also be escaped as the value for the
8882 @option{text} option in the filter argument string and as the filter
8883 argument in the filtergraph description, and possibly also for the shell,
8884 that makes up to four levels of escaping; using a text file avoids these
8887 The following functions are available:
8892 The expression evaluation result.
8894 It must take one argument specifying the expression to be evaluated,
8895 which accepts the same constants and functions as the @var{x} and
8896 @var{y} values. Note that not all constants should be used, for
8897 example the text size is not known when evaluating the expression, so
8898 the constants @var{text_w} and @var{text_h} will have an undefined
8901 @item expr_int_format, eif
8902 Evaluate the expression's value and output as formatted integer.
8904 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8905 The second argument specifies the output format. Allowed values are @samp{x},
8906 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8907 @code{printf} function.
8908 The third parameter is optional and sets the number of positions taken by the output.
8909 It can be used to add padding with zeros from the left.
8912 The time at which the filter is running, expressed in UTC.
8913 It can accept an argument: a strftime() format string.
8916 The time at which the filter is running, expressed in the local time zone.
8917 It can accept an argument: a strftime() format string.
8920 Frame metadata. Takes one or two arguments.
8922 The first argument is mandatory and specifies the metadata key.
8924 The second argument is optional and specifies a default value, used when the
8925 metadata key is not found or empty.
8928 The frame number, starting from 0.
8931 A 1 character description of the current picture type.
8934 The timestamp of the current frame.
8935 It can take up to three arguments.
8937 The first argument is the format of the timestamp; it defaults to @code{flt}
8938 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8939 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8940 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8941 @code{localtime} stands for the timestamp of the frame formatted as
8942 local time zone time.
8944 The second argument is an offset added to the timestamp.
8946 If the format is set to @code{hms}, a third argument @code{24HH} may be
8947 supplied to present the hour part of the formatted timestamp in 24h format
8950 If the format is set to @code{localtime} or @code{gmtime},
8951 a third argument may be supplied: a strftime() format string.
8952 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8955 @subsection Examples
8959 Draw "Test Text" with font FreeSerif, using the default values for the
8960 optional parameters.
8963 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8967 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8968 and y=50 (counting from the top-left corner of the screen), text is
8969 yellow with a red box around it. Both the text and the box have an
8973 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8974 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8977 Note that the double quotes are not necessary if spaces are not used
8978 within the parameter list.
8981 Show the text at the center of the video frame:
8983 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8987 Show the text at a random position, switching to a new position every 30 seconds:
8989 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)"
8993 Show a text line sliding from right to left in the last row of the video
8994 frame. The file @file{LONG_LINE} is assumed to contain a single line
8997 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
9001 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
9003 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
9007 Draw a single green letter "g", at the center of the input video.
9008 The glyph baseline is placed at half screen height.
9010 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
9014 Show text for 1 second every 3 seconds:
9016 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
9020 Use fontconfig to set the font. Note that the colons need to be escaped.
9022 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
9026 Print the date of a real-time encoding (see strftime(3)):
9028 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
9032 Show text fading in and out (appearing/disappearing):
9035 DS=1.0 # display start
9036 DE=10.0 # display end
9037 FID=1.5 # fade in duration
9038 FOD=5 # fade out duration
9039 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 @}"
9043 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
9044 and the @option{fontsize} value are included in the @option{y} offset.
9046 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
9047 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
9052 For more information about libfreetype, check:
9053 @url{http://www.freetype.org/}.
9055 For more information about fontconfig, check:
9056 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
9058 For more information about libfribidi, check:
9059 @url{http://fribidi.org/}.
9063 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
9065 The filter accepts the following options:
9070 Set low and high threshold values used by the Canny thresholding
9073 The high threshold selects the "strong" edge pixels, which are then
9074 connected through 8-connectivity with the "weak" edge pixels selected
9075 by the low threshold.
9077 @var{low} and @var{high} threshold values must be chosen in the range
9078 [0,1], and @var{low} should be lesser or equal to @var{high}.
9080 Default value for @var{low} is @code{20/255}, and default value for @var{high}
9084 Define the drawing mode.
9088 Draw white/gray wires on black background.
9091 Mix the colors to create a paint/cartoon effect.
9094 Apply Canny edge detector on all selected planes.
9096 Default value is @var{wires}.
9099 Select planes for filtering. By default all available planes are filtered.
9102 @subsection Examples
9106 Standard edge detection with custom values for the hysteresis thresholding:
9108 edgedetect=low=0.1:high=0.4
9112 Painting effect without thresholding:
9114 edgedetect=mode=colormix:high=0
9119 Set brightness, contrast, saturation and approximate gamma adjustment.
9121 The filter accepts the following options:
9125 Set the contrast expression. The value must be a float value in range
9126 @code{-2.0} to @code{2.0}. The default value is "1".
9129 Set the brightness expression. The value must be a float value in
9130 range @code{-1.0} to @code{1.0}. The default value is "0".
9133 Set the saturation expression. The value must be a float in
9134 range @code{0.0} to @code{3.0}. The default value is "1".
9137 Set the gamma expression. The value must be a float in range
9138 @code{0.1} to @code{10.0}. The default value is "1".
9141 Set the gamma expression for red. The value must be a float in
9142 range @code{0.1} to @code{10.0}. The default value is "1".
9145 Set the gamma expression for green. The value must be a float in range
9146 @code{0.1} to @code{10.0}. The default value is "1".
9149 Set the gamma expression for blue. The value must be a float in range
9150 @code{0.1} to @code{10.0}. The default value is "1".
9153 Set the gamma weight expression. It can be used to reduce the effect
9154 of a high gamma value on bright image areas, e.g. keep them from
9155 getting overamplified and just plain white. The value must be a float
9156 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
9157 gamma correction all the way down while @code{1.0} leaves it at its
9158 full strength. Default is "1".
9161 Set when the expressions for brightness, contrast, saturation and
9162 gamma expressions are evaluated.
9164 It accepts the following values:
9167 only evaluate expressions once during the filter initialization or
9168 when a command is processed
9171 evaluate expressions for each incoming frame
9174 Default value is @samp{init}.
9177 The expressions accept the following parameters:
9180 frame count of the input frame starting from 0
9183 byte position of the corresponding packet in the input file, NAN if
9187 frame rate of the input video, NAN if the input frame rate is unknown
9190 timestamp expressed in seconds, NAN if the input timestamp is unknown
9193 @subsection Commands
9194 The filter supports the following commands:
9198 Set the contrast expression.
9201 Set the brightness expression.
9204 Set the saturation expression.
9207 Set the gamma expression.
9210 Set the gamma_r expression.
9213 Set gamma_g expression.
9216 Set gamma_b expression.
9219 Set gamma_weight expression.
9221 The command accepts the same syntax of the corresponding option.
9223 If the specified expression is not valid, it is kept at its current
9230 Apply erosion effect to the video.
9232 This filter replaces the pixel by the local(3x3) minimum.
9234 It accepts the following options:
9241 Limit the maximum change for each plane, default is 65535.
9242 If 0, plane will remain unchanged.
9245 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9248 Flags to local 3x3 coordinates maps like this:
9255 @section extractplanes
9257 Extract color channel components from input video stream into
9258 separate grayscale video streams.
9260 The filter accepts the following option:
9264 Set plane(s) to extract.
9266 Available values for planes are:
9277 Choosing planes not available in the input will result in an error.
9278 That means you cannot select @code{r}, @code{g}, @code{b} planes
9279 with @code{y}, @code{u}, @code{v} planes at same time.
9282 @subsection Examples
9286 Extract luma, u and v color channel component from input video frame
9287 into 3 grayscale outputs:
9289 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
9295 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9297 For each input image, the filter will compute the optimal mapping from
9298 the input to the output given the codebook length, that is the number
9299 of distinct output colors.
9301 This filter accepts the following options.
9304 @item codebook_length, l
9305 Set codebook length. The value must be a positive integer, and
9306 represents the number of distinct output colors. Default value is 256.
9309 Set the maximum number of iterations to apply for computing the optimal
9310 mapping. The higher the value the better the result and the higher the
9311 computation time. Default value is 1.
9314 Set a random seed, must be an integer included between 0 and
9315 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9316 will try to use a good random seed on a best effort basis.
9319 Set pal8 output pixel format. This option does not work with codebook
9320 length greater than 256.
9325 Measure graylevel entropy in histogram of color channels of video frames.
9327 It accepts the following parameters:
9331 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9333 @var{diff} mode measures entropy of histogram delta values, absolute differences
9334 between neighbour histogram values.
9339 Apply a fade-in/out effect to the input video.
9341 It accepts the following parameters:
9345 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9347 Default is @code{in}.
9349 @item start_frame, s
9350 Specify the number of the frame to start applying the fade
9351 effect at. Default is 0.
9354 The number of frames that the fade effect lasts. At the end of the
9355 fade-in effect, the output video will have the same intensity as the input video.
9356 At the end of the fade-out transition, the output video will be filled with the
9357 selected @option{color}.
9361 If set to 1, fade only alpha channel, if one exists on the input.
9364 @item start_time, st
9365 Specify the timestamp (in seconds) of the frame to start to apply the fade
9366 effect. If both start_frame and start_time are specified, the fade will start at
9367 whichever comes last. Default is 0.
9370 The number of seconds for which the fade effect has to last. At the end of the
9371 fade-in effect the output video will have the same intensity as the input video,
9372 at the end of the fade-out transition the output video will be filled with the
9373 selected @option{color}.
9374 If both duration and nb_frames are specified, duration is used. Default is 0
9375 (nb_frames is used by default).
9378 Specify the color of the fade. Default is "black".
9381 @subsection Examples
9385 Fade in the first 30 frames of video:
9390 The command above is equivalent to:
9396 Fade out the last 45 frames of a 200-frame video:
9399 fade=type=out:start_frame=155:nb_frames=45
9403 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9405 fade=in:0:25, fade=out:975:25
9409 Make the first 5 frames yellow, then fade in from frame 5-24:
9411 fade=in:5:20:color=yellow
9415 Fade in alpha over first 25 frames of video:
9417 fade=in:0:25:alpha=1
9421 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9423 fade=t=in:st=5.5:d=0.5
9429 Apply arbitrary expressions to samples in frequency domain
9433 Adjust the dc value (gain) of the luma plane of the image. The filter
9434 accepts an integer value in range @code{0} to @code{1000}. The default
9435 value is set to @code{0}.
9438 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9439 filter accepts an integer value in range @code{0} to @code{1000}. The
9440 default value is set to @code{0}.
9443 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9444 filter accepts an integer value in range @code{0} to @code{1000}. The
9445 default value is set to @code{0}.
9448 Set the frequency domain weight expression for the luma plane.
9451 Set the frequency domain weight expression for the 1st chroma plane.
9454 Set the frequency domain weight expression for the 2nd chroma plane.
9457 Set when the expressions are evaluated.
9459 It accepts the following values:
9462 Only evaluate expressions once during the filter initialization.
9465 Evaluate expressions for each incoming frame.
9468 Default value is @samp{init}.
9470 The filter accepts the following variables:
9473 The coordinates of the current sample.
9477 The width and height of the image.
9480 The number of input frame, starting from 0.
9483 @subsection Examples
9489 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9495 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9501 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9507 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9513 Denoise frames using 3D FFT (frequency domain filtering).
9515 The filter accepts the following options:
9519 Set the noise sigma constant. This sets denoising strength.
9520 Default value is 1. Allowed range is from 0 to 30.
9521 Using very high sigma with low overlap may give blocking artifacts.
9524 Set amount of denoising. By default all detected noise is reduced.
9525 Default value is 1. Allowed range is from 0 to 1.
9528 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9529 Actual size of block in pixels is 2 to power of @var{block}, so by default
9530 block size in pixels is 2^4 which is 16.
9533 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9536 Set number of previous frames to use for denoising. By default is set to 0.
9539 Set number of next frames to to use for denoising. By default is set to 0.
9542 Set planes which will be filtered, by default are all available filtered
9548 Extract a single field from an interlaced image using stride
9549 arithmetic to avoid wasting CPU time. The output frames are marked as
9552 The filter accepts the following options:
9556 Specify whether to extract the top (if the value is @code{0} or
9557 @code{top}) or the bottom field (if the value is @code{1} or
9563 Create new frames by copying the top and bottom fields from surrounding frames
9564 supplied as numbers by the hint file.
9568 Set file containing hints: absolute/relative frame numbers.
9570 There must be one line for each frame in a clip. Each line must contain two
9571 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9572 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9573 is current frame number for @code{absolute} mode or out of [-1, 1] range
9574 for @code{relative} mode. First number tells from which frame to pick up top
9575 field and second number tells from which frame to pick up bottom field.
9577 If optionally followed by @code{+} output frame will be marked as interlaced,
9578 else if followed by @code{-} output frame will be marked as progressive, else
9579 it will be marked same as input frame.
9580 If line starts with @code{#} or @code{;} that line is skipped.
9583 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9586 Example of first several lines of @code{hint} file for @code{relative} mode:
9589 1,0 - # second frame, use third's frame top field and second's frame bottom field
9590 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9607 Field matching filter for inverse telecine. It is meant to reconstruct the
9608 progressive frames from a telecined stream. The filter does not drop duplicated
9609 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9610 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9612 The separation of the field matching and the decimation is notably motivated by
9613 the possibility of inserting a de-interlacing filter fallback between the two.
9614 If the source has mixed telecined and real interlaced content,
9615 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9616 But these remaining combed frames will be marked as interlaced, and thus can be
9617 de-interlaced by a later filter such as @ref{yadif} before decimation.
9619 In addition to the various configuration options, @code{fieldmatch} can take an
9620 optional second stream, activated through the @option{ppsrc} option. If
9621 enabled, the frames reconstruction will be based on the fields and frames from
9622 this second stream. This allows the first input to be pre-processed in order to
9623 help the various algorithms of the filter, while keeping the output lossless
9624 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9625 or brightness/contrast adjustments can help.
9627 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9628 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9629 which @code{fieldmatch} is based on. While the semantic and usage are very
9630 close, some behaviour and options names can differ.
9632 The @ref{decimate} filter currently only works for constant frame rate input.
9633 If your input has mixed telecined (30fps) and progressive content with a lower
9634 framerate like 24fps use the following filterchain to produce the necessary cfr
9635 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9637 The filter accepts the following options:
9641 Specify the assumed field order of the input stream. Available values are:
9645 Auto detect parity (use FFmpeg's internal parity value).
9647 Assume bottom field first.
9649 Assume top field first.
9652 Note that it is sometimes recommended not to trust the parity announced by the
9655 Default value is @var{auto}.
9658 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9659 sense that it won't risk creating jerkiness due to duplicate frames when
9660 possible, but if there are bad edits or blended fields it will end up
9661 outputting combed frames when a good match might actually exist. On the other
9662 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9663 but will almost always find a good frame if there is one. The other values are
9664 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9665 jerkiness and creating duplicate frames versus finding good matches in sections
9666 with bad edits, orphaned fields, blended fields, etc.
9668 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9670 Available values are:
9674 2-way matching (p/c)
9676 2-way matching, and trying 3rd match if still combed (p/c + n)
9678 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9680 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9681 still combed (p/c + n + u/b)
9683 3-way matching (p/c/n)
9685 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9686 detected as combed (p/c/n + u/b)
9689 The parenthesis at the end indicate the matches that would be used for that
9690 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9693 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9696 Default value is @var{pc_n}.
9699 Mark the main input stream as a pre-processed input, and enable the secondary
9700 input stream as the clean source to pick the fields from. See the filter
9701 introduction for more details. It is similar to the @option{clip2} feature from
9704 Default value is @code{0} (disabled).
9707 Set the field to match from. It is recommended to set this to the same value as
9708 @option{order} unless you experience matching failures with that setting. In
9709 certain circumstances changing the field that is used to match from can have a
9710 large impact on matching performance. Available values are:
9714 Automatic (same value as @option{order}).
9716 Match from the bottom field.
9718 Match from the top field.
9721 Default value is @var{auto}.
9724 Set whether or not chroma is included during the match comparisons. In most
9725 cases it is recommended to leave this enabled. You should set this to @code{0}
9726 only if your clip has bad chroma problems such as heavy rainbowing or other
9727 artifacts. Setting this to @code{0} could also be used to speed things up at
9728 the cost of some accuracy.
9730 Default value is @code{1}.
9734 These define an exclusion band which excludes the lines between @option{y0} and
9735 @option{y1} from being included in the field matching decision. An exclusion
9736 band can be used to ignore subtitles, a logo, or other things that may
9737 interfere with the matching. @option{y0} sets the starting scan line and
9738 @option{y1} sets the ending line; all lines in between @option{y0} and
9739 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9740 @option{y0} and @option{y1} to the same value will disable the feature.
9741 @option{y0} and @option{y1} defaults to @code{0}.
9744 Set the scene change detection threshold as a percentage of maximum change on
9745 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9746 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9747 @option{scthresh} is @code{[0.0, 100.0]}.
9749 Default value is @code{12.0}.
9752 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9753 account the combed scores of matches when deciding what match to use as the
9754 final match. Available values are:
9758 No final matching based on combed scores.
9760 Combed scores are only used when a scene change is detected.
9762 Use combed scores all the time.
9765 Default is @var{sc}.
9768 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9769 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9770 Available values are:
9774 No forced calculation.
9776 Force p/c/n calculations.
9778 Force p/c/n/u/b calculations.
9781 Default value is @var{none}.
9784 This is the area combing threshold used for combed frame detection. This
9785 essentially controls how "strong" or "visible" combing must be to be detected.
9786 Larger values mean combing must be more visible and smaller values mean combing
9787 can be less visible or strong and still be detected. Valid settings are from
9788 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9789 be detected as combed). This is basically a pixel difference value. A good
9790 range is @code{[8, 12]}.
9792 Default value is @code{9}.
9795 Sets whether or not chroma is considered in the combed frame decision. Only
9796 disable this if your source has chroma problems (rainbowing, etc.) that are
9797 causing problems for the combed frame detection with chroma enabled. Actually,
9798 using @option{chroma}=@var{0} is usually more reliable, except for the case
9799 where there is chroma only combing in the source.
9801 Default value is @code{0}.
9805 Respectively set the x-axis and y-axis size of the window used during combed
9806 frame detection. This has to do with the size of the area in which
9807 @option{combpel} pixels are required to be detected as combed for a frame to be
9808 declared combed. See the @option{combpel} parameter description for more info.
9809 Possible values are any number that is a power of 2 starting at 4 and going up
9812 Default value is @code{16}.
9815 The number of combed pixels inside any of the @option{blocky} by
9816 @option{blockx} size blocks on the frame for the frame to be detected as
9817 combed. While @option{cthresh} controls how "visible" the combing must be, this
9818 setting controls "how much" combing there must be in any localized area (a
9819 window defined by the @option{blockx} and @option{blocky} settings) on the
9820 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9821 which point no frames will ever be detected as combed). This setting is known
9822 as @option{MI} in TFM/VFM vocabulary.
9824 Default value is @code{80}.
9827 @anchor{p/c/n/u/b meaning}
9828 @subsection p/c/n/u/b meaning
9830 @subsubsection p/c/n
9832 We assume the following telecined stream:
9835 Top fields: 1 2 2 3 4
9836 Bottom fields: 1 2 3 4 4
9839 The numbers correspond to the progressive frame the fields relate to. Here, the
9840 first two frames are progressive, the 3rd and 4th are combed, and so on.
9842 When @code{fieldmatch} is configured to run a matching from bottom
9843 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9848 B 1 2 3 4 4 <-- matching reference
9857 As a result of the field matching, we can see that some frames get duplicated.
9858 To perform a complete inverse telecine, you need to rely on a decimation filter
9859 after this operation. See for instance the @ref{decimate} filter.
9861 The same operation now matching from top fields (@option{field}=@var{top})
9866 T 1 2 2 3 4 <-- matching reference
9876 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9877 basically, they refer to the frame and field of the opposite parity:
9880 @item @var{p} matches the field of the opposite parity in the previous frame
9881 @item @var{c} matches the field of the opposite parity in the current frame
9882 @item @var{n} matches the field of the opposite parity in the next frame
9887 The @var{u} and @var{b} matching are a bit special in the sense that they match
9888 from the opposite parity flag. In the following examples, we assume that we are
9889 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9890 'x' is placed above and below each matched fields.
9892 With bottom matching (@option{field}=@var{bottom}):
9897 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9898 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9906 With top matching (@option{field}=@var{top}):
9911 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9912 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9920 @subsection Examples
9922 Simple IVTC of a top field first telecined stream:
9924 fieldmatch=order=tff:combmatch=none, decimate
9927 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9929 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9934 Transform the field order of the input video.
9936 It accepts the following parameters:
9941 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9942 for bottom field first.
9945 The default value is @samp{tff}.
9947 The transformation is done by shifting the picture content up or down
9948 by one line, and filling the remaining line with appropriate picture content.
9949 This method is consistent with most broadcast field order converters.
9951 If the input video is not flagged as being interlaced, or it is already
9952 flagged as being of the required output field order, then this filter does
9953 not alter the incoming video.
9955 It is very useful when converting to or from PAL DV material,
9956 which is bottom field first.
9960 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9963 @section fifo, afifo
9965 Buffer input images and send them when they are requested.
9967 It is mainly useful when auto-inserted by the libavfilter
9970 It does not take parameters.
9972 @section fillborders
9974 Fill borders of the input video, without changing video stream dimensions.
9975 Sometimes video can have garbage at the four edges and you may not want to
9976 crop video input to keep size multiple of some number.
9978 This filter accepts the following options:
9982 Number of pixels to fill from left border.
9985 Number of pixels to fill from right border.
9988 Number of pixels to fill from top border.
9991 Number of pixels to fill from bottom border.
9996 It accepts the following values:
9999 fill pixels using outermost pixels
10002 fill pixels using mirroring
10005 fill pixels with constant value
10008 Default is @var{smear}.
10011 Set color for pixels in fixed mode. Default is @var{black}.
10016 Find a rectangular object
10018 It accepts the following options:
10022 Filepath of the object image, needs to be in gray8.
10025 Detection threshold, default is 0.5.
10028 Number of mipmaps, default is 3.
10030 @item xmin, ymin, xmax, ymax
10031 Specifies the rectangle in which to search.
10034 @subsection Examples
10038 Generate a representative palette of a given video using @command{ffmpeg}:
10040 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
10044 @section cover_rect
10046 Cover a rectangular object
10048 It accepts the following options:
10052 Filepath of the optional cover image, needs to be in yuv420.
10057 It accepts the following values:
10060 cover it by the supplied image
10062 cover it by interpolating the surrounding pixels
10065 Default value is @var{blur}.
10068 @subsection Examples
10072 Generate a representative palette of a given video using @command{ffmpeg}:
10074 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
10080 Flood area with values of same pixel components with another values.
10082 It accepts the following options:
10085 Set pixel x coordinate.
10088 Set pixel y coordinate.
10091 Set source #0 component value.
10094 Set source #1 component value.
10097 Set source #2 component value.
10100 Set source #3 component value.
10103 Set destination #0 component value.
10106 Set destination #1 component value.
10109 Set destination #2 component value.
10112 Set destination #3 component value.
10118 Convert the input video to one of the specified pixel formats.
10119 Libavfilter will try to pick one that is suitable as input to
10122 It accepts the following parameters:
10126 A '|'-separated list of pixel format names, such as
10127 "pix_fmts=yuv420p|monow|rgb24".
10131 @subsection Examples
10135 Convert the input video to the @var{yuv420p} format
10137 format=pix_fmts=yuv420p
10140 Convert the input video to any of the formats in the list
10142 format=pix_fmts=yuv420p|yuv444p|yuv410p
10149 Convert the video to specified constant frame rate by duplicating or dropping
10150 frames as necessary.
10152 It accepts the following parameters:
10156 The desired output frame rate. The default is @code{25}.
10159 Assume the first PTS should be the given value, in seconds. This allows for
10160 padding/trimming at the start of stream. By default, no assumption is made
10161 about the first frame's expected PTS, so no padding or trimming is done.
10162 For example, this could be set to 0 to pad the beginning with duplicates of
10163 the first frame if a video stream starts after the audio stream or to trim any
10164 frames with a negative PTS.
10167 Timestamp (PTS) rounding method.
10169 Possible values are:
10176 round towards -infinity
10178 round towards +infinity
10182 The default is @code{near}.
10185 Action performed when reading the last frame.
10187 Possible values are:
10190 Use same timestamp rounding method as used for other frames.
10192 Pass through last frame if input duration has not been reached yet.
10194 The default is @code{round}.
10198 Alternatively, the options can be specified as a flat string:
10199 @var{fps}[:@var{start_time}[:@var{round}]].
10201 See also the @ref{setpts} filter.
10203 @subsection Examples
10207 A typical usage in order to set the fps to 25:
10213 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10215 fps=fps=film:round=near
10221 Pack two different video streams into a stereoscopic video, setting proper
10222 metadata on supported codecs. The two views should have the same size and
10223 framerate and processing will stop when the shorter video ends. Please note
10224 that you may conveniently adjust view properties with the @ref{scale} and
10227 It accepts the following parameters:
10231 The desired packing format. Supported values are:
10236 The views are next to each other (default).
10239 The views are on top of each other.
10242 The views are packed by line.
10245 The views are packed by column.
10248 The views are temporally interleaved.
10257 # Convert left and right views into a frame-sequential video
10258 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10260 # Convert views into a side-by-side video with the same output resolution as the input
10261 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
10266 Change the frame rate by interpolating new video output frames from the source
10269 This filter is not designed to function correctly with interlaced media. If
10270 you wish to change the frame rate of interlaced media then you are required
10271 to deinterlace before this filter and re-interlace after this filter.
10273 A description of the accepted options follows.
10277 Specify the output frames per second. This option can also be specified
10278 as a value alone. The default is @code{50}.
10281 Specify the start of a range where the output frame will be created as a
10282 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10283 the default is @code{15}.
10286 Specify the end of a range where the output frame will be created as a
10287 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10288 the default is @code{240}.
10291 Specify the level at which a scene change is detected as a value between
10292 0 and 100 to indicate a new scene; a low value reflects a low
10293 probability for the current frame to introduce a new scene, while a higher
10294 value means the current frame is more likely to be one.
10295 The default is @code{8.2}.
10298 Specify flags influencing the filter process.
10300 Available value for @var{flags} is:
10303 @item scene_change_detect, scd
10304 Enable scene change detection using the value of the option @var{scene}.
10305 This flag is enabled by default.
10311 Select one frame every N-th frame.
10313 This filter accepts the following option:
10316 Select frame after every @code{step} frames.
10317 Allowed values are positive integers higher than 0. Default value is @code{1}.
10320 @section freezedetect
10322 Detect frozen video.
10324 This filter logs a message and sets frame metadata when it detects that the
10325 input video has no significant change in content during a specified duration.
10326 Video freeze detection calculates the mean average absolute difference of all
10327 the components of video frames and compares it to a noise floor.
10329 The printed times and duration are expressed in seconds. The
10330 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10331 whose timestamp equals or exceeds the detection duration and it contains the
10332 timestamp of the first frame of the freeze. The
10333 @code{lavfi.freezedetect.freeze_duration} and
10334 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10337 The filter accepts the following options:
10341 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10342 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10346 Set freeze duration until notification (default is 2 seconds).
10352 Apply a frei0r effect to the input video.
10354 To enable the compilation of this filter, you need to install the frei0r
10355 header and configure FFmpeg with @code{--enable-frei0r}.
10357 It accepts the following parameters:
10362 The name of the frei0r effect to load. If the environment variable
10363 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10364 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10365 Otherwise, the standard frei0r paths are searched, in this order:
10366 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10367 @file{/usr/lib/frei0r-1/}.
10369 @item filter_params
10370 A '|'-separated list of parameters to pass to the frei0r effect.
10374 A frei0r effect parameter can be a boolean (its value is either
10375 "y" or "n"), a double, a color (specified as
10376 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10377 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10378 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10379 a position (specified as @var{X}/@var{Y}, where
10380 @var{X} and @var{Y} are floating point numbers) and/or a string.
10382 The number and types of parameters depend on the loaded effect. If an
10383 effect parameter is not specified, the default value is set.
10385 @subsection Examples
10389 Apply the distort0r effect, setting the first two double parameters:
10391 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10395 Apply the colordistance effect, taking a color as the first parameter:
10397 frei0r=colordistance:0.2/0.3/0.4
10398 frei0r=colordistance:violet
10399 frei0r=colordistance:0x112233
10403 Apply the perspective effect, specifying the top left and top right image
10406 frei0r=perspective:0.2/0.2|0.8/0.2
10410 For more information, see
10411 @url{http://frei0r.dyne.org}
10415 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10417 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10418 processing filter, one of them is performed once per block, not per pixel.
10419 This allows for much higher speed.
10421 The filter accepts the following options:
10425 Set quality. This option defines the number of levels for averaging. It accepts
10426 an integer in the range 4-5. Default value is @code{4}.
10429 Force a constant quantization parameter. It accepts an integer in range 0-63.
10430 If not set, the filter will use the QP from the video stream (if available).
10433 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10434 more details but also more artifacts, while higher values make the image smoother
10435 but also blurrier. Default value is @code{0} − PSNR optimal.
10437 @item use_bframe_qp
10438 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10439 option may cause flicker since the B-Frames have often larger QP. Default is
10440 @code{0} (not enabled).
10446 Apply Gaussian blur filter.
10448 The filter accepts the following options:
10452 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10455 Set number of steps for Gaussian approximation. Default is @code{1}.
10458 Set which planes to filter. By default all planes are filtered.
10461 Set vertical sigma, if negative it will be same as @code{sigma}.
10462 Default is @code{-1}.
10467 Apply generic equation to each pixel.
10469 The filter accepts the following options:
10472 @item lum_expr, lum
10473 Set the luminance expression.
10475 Set the chrominance blue expression.
10477 Set the chrominance red expression.
10478 @item alpha_expr, a
10479 Set the alpha expression.
10481 Set the red expression.
10482 @item green_expr, g
10483 Set the green expression.
10485 Set the blue expression.
10488 The colorspace is selected according to the specified options. If one
10489 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10490 options is specified, the filter will automatically select a YCbCr
10491 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10492 @option{blue_expr} options is specified, it will select an RGB
10495 If one of the chrominance expression is not defined, it falls back on the other
10496 one. If no alpha expression is specified it will evaluate to opaque value.
10497 If none of chrominance expressions are specified, they will evaluate
10498 to the luminance expression.
10500 The expressions can use the following variables and functions:
10504 The sequential number of the filtered frame, starting from @code{0}.
10508 The coordinates of the current sample.
10512 The width and height of the image.
10516 Width and height scale depending on the currently filtered plane. It is the
10517 ratio between the corresponding luma plane number of pixels and the current
10518 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10519 @code{0.5,0.5} for chroma planes.
10522 Time of the current frame, expressed in seconds.
10525 Return the value of the pixel at location (@var{x},@var{y}) of the current
10529 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10533 Return the value of the pixel at location (@var{x},@var{y}) of the
10534 blue-difference chroma plane. Return 0 if there is no such plane.
10537 Return the value of the pixel at location (@var{x},@var{y}) of the
10538 red-difference chroma plane. Return 0 if there is no such plane.
10543 Return the value of the pixel at location (@var{x},@var{y}) of the
10544 red/green/blue component. Return 0 if there is no such component.
10547 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10548 plane. Return 0 if there is no such plane.
10551 For functions, if @var{x} and @var{y} are outside the area, the value will be
10552 automatically clipped to the closer edge.
10554 @subsection Examples
10558 Flip the image horizontally:
10564 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10565 wavelength of 100 pixels:
10567 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10571 Generate a fancy enigmatic moving light:
10573 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
10577 Generate a quick emboss effect:
10579 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10583 Modify RGB components depending on pixel position:
10585 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10589 Create a radial gradient that is the same size as the input (also see
10590 the @ref{vignette} filter):
10592 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10598 Fix the banding artifacts that are sometimes introduced into nearly flat
10599 regions by truncation to 8-bit color depth.
10600 Interpolate the gradients that should go where the bands are, and
10603 It is designed for playback only. Do not use it prior to
10604 lossy compression, because compression tends to lose the dither and
10605 bring back the bands.
10607 It accepts the following parameters:
10612 The maximum amount by which the filter will change any one pixel. This is also
10613 the threshold for detecting nearly flat regions. Acceptable values range from
10614 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10618 The neighborhood to fit the gradient to. A larger radius makes for smoother
10619 gradients, but also prevents the filter from modifying the pixels near detailed
10620 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10621 values will be clipped to the valid range.
10625 Alternatively, the options can be specified as a flat string:
10626 @var{strength}[:@var{radius}]
10628 @subsection Examples
10632 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10638 Specify radius, omitting the strength (which will fall-back to the default
10646 @section graphmonitor, agraphmonitor
10647 Show various filtergraph stats.
10649 With this filter one can debug complete filtergraph.
10650 Especially issues with links filling with queued frames.
10652 The filter accepts the following options:
10656 Set video output size. Default is @var{hd720}.
10659 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10662 Set output mode, can be @var{fulll} or @var{compact}.
10663 In @var{compact} mode only filters with some queued frames have displayed stats.
10666 Set flags which enable which stats are shown in video.
10668 Available values for flags are:
10671 Display number of queued frames in each link.
10673 @item frame_count_in
10674 Display number of frames taken from filter.
10676 @item frame_count_out
10677 Display number of frames given out from filter.
10680 Display current filtered frame pts.
10683 Display current filtered frame time.
10686 Display time base for filter link.
10689 Display used format for filter link.
10692 Display video size or number of audio channels in case of audio used by filter link.
10695 Display video frame rate or sample rate in case of audio used by filter link.
10699 Set upper limit for video rate of output stream, Default value is @var{25}.
10700 This guarantee that output video frame rate will not be higher than this value.
10704 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10705 and corrects the scene colors accordingly.
10707 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10709 The filter accepts the following options:
10713 The order of differentiation to be applied on the scene. Must be chosen in the range
10714 [0,2] and default value is 1.
10717 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10718 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10719 max value instead of calculating Minkowski distance.
10722 The standard deviation of Gaussian blur to be applied on the scene. Must be
10723 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10724 can't be equal to 0 if @var{difford} is greater than 0.
10727 @subsection Examples
10733 greyedge=difford=1:minknorm=5:sigma=2
10739 greyedge=difford=1:minknorm=0:sigma=2
10747 Apply a Hald CLUT to a video stream.
10749 First input is the video stream to process, and second one is the Hald CLUT.
10750 The Hald CLUT input can be a simple picture or a complete video stream.
10752 The filter accepts the following options:
10756 Force termination when the shortest input terminates. Default is @code{0}.
10758 Continue applying the last CLUT after the end of the stream. A value of
10759 @code{0} disable the filter after the last frame of the CLUT is reached.
10760 Default is @code{1}.
10763 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10764 filters share the same internals).
10766 More information about the Hald CLUT can be found on Eskil Steenberg's website
10767 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10769 @subsection Workflow examples
10771 @subsubsection Hald CLUT video stream
10773 Generate an identity Hald CLUT stream altered with various effects:
10775 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
10778 Note: make sure you use a lossless codec.
10780 Then use it with @code{haldclut} to apply it on some random stream:
10782 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10785 The Hald CLUT will be applied to the 10 first seconds (duration of
10786 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10787 to the remaining frames of the @code{mandelbrot} stream.
10789 @subsubsection Hald CLUT with preview
10791 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10792 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10793 biggest possible square starting at the top left of the picture. The remaining
10794 padding pixels (bottom or right) will be ignored. This area can be used to add
10795 a preview of the Hald CLUT.
10797 Typically, the following generated Hald CLUT will be supported by the
10798 @code{haldclut} filter:
10801 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10802 pad=iw+320 [padded_clut];
10803 smptebars=s=320x256, split [a][b];
10804 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10805 [main][b] overlay=W-320" -frames:v 1 clut.png
10808 It contains the original and a preview of the effect of the CLUT: SMPTE color
10809 bars are displayed on the right-top, and below the same color bars processed by
10812 Then, the effect of this Hald CLUT can be visualized with:
10814 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10819 Flip the input video horizontally.
10821 For example, to horizontally flip the input video with @command{ffmpeg}:
10823 ffmpeg -i in.avi -vf "hflip" out.avi
10827 This filter applies a global color histogram equalization on a
10830 It can be used to correct video that has a compressed range of pixel
10831 intensities. The filter redistributes the pixel intensities to
10832 equalize their distribution across the intensity range. It may be
10833 viewed as an "automatically adjusting contrast filter". This filter is
10834 useful only for correcting degraded or poorly captured source
10837 The filter accepts the following options:
10841 Determine the amount of equalization to be applied. As the strength
10842 is reduced, the distribution of pixel intensities more-and-more
10843 approaches that of the input frame. The value must be a float number
10844 in the range [0,1] and defaults to 0.200.
10847 Set the maximum intensity that can generated and scale the output
10848 values appropriately. The strength should be set as desired and then
10849 the intensity can be limited if needed to avoid washing-out. The value
10850 must be a float number in the range [0,1] and defaults to 0.210.
10853 Set the antibanding level. If enabled the filter will randomly vary
10854 the luminance of output pixels by a small amount to avoid banding of
10855 the histogram. Possible values are @code{none}, @code{weak} or
10856 @code{strong}. It defaults to @code{none}.
10861 Compute and draw a color distribution histogram for the input video.
10863 The computed histogram is a representation of the color component
10864 distribution in an image.
10866 Standard histogram displays the color components distribution in an image.
10867 Displays color graph for each color component. Shows distribution of
10868 the Y, U, V, A or R, G, B components, depending on input format, in the
10869 current frame. Below each graph a color component scale meter is shown.
10871 The filter accepts the following options:
10875 Set height of level. Default value is @code{200}.
10876 Allowed range is [50, 2048].
10879 Set height of color scale. Default value is @code{12}.
10880 Allowed range is [0, 40].
10884 It accepts the following values:
10887 Per color component graphs are placed below each other.
10890 Per color component graphs are placed side by side.
10893 Presents information identical to that in the @code{parade}, except
10894 that the graphs representing color components are superimposed directly
10897 Default is @code{stack}.
10900 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10901 Default is @code{linear}.
10904 Set what color components to display.
10905 Default is @code{7}.
10908 Set foreground opacity. Default is @code{0.7}.
10911 Set background opacity. Default is @code{0.5}.
10914 @subsection Examples
10919 Calculate and draw histogram:
10921 ffplay -i input -vf histogram
10929 This is a high precision/quality 3d denoise filter. It aims to reduce
10930 image noise, producing smooth images and making still images really
10931 still. It should enhance compressibility.
10933 It accepts the following optional parameters:
10937 A non-negative floating point number which specifies spatial luma strength.
10938 It defaults to 4.0.
10940 @item chroma_spatial
10941 A non-negative floating point number which specifies spatial chroma strength.
10942 It defaults to 3.0*@var{luma_spatial}/4.0.
10945 A floating point number which specifies luma temporal strength. It defaults to
10946 6.0*@var{luma_spatial}/4.0.
10949 A floating point number which specifies chroma temporal strength. It defaults to
10950 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10953 @anchor{hwdownload}
10954 @section hwdownload
10956 Download hardware frames to system memory.
10958 The input must be in hardware frames, and the output a non-hardware format.
10959 Not all formats will be supported on the output - it may be necessary to insert
10960 an additional @option{format} filter immediately following in the graph to get
10961 the output in a supported format.
10965 Map hardware frames to system memory or to another device.
10967 This filter has several different modes of operation; which one is used depends
10968 on the input and output formats:
10971 Hardware frame input, normal frame output
10973 Map the input frames to system memory and pass them to the output. If the
10974 original hardware frame is later required (for example, after overlaying
10975 something else on part of it), the @option{hwmap} filter can be used again
10976 in the next mode to retrieve it.
10978 Normal frame input, hardware frame output
10980 If the input is actually a software-mapped hardware frame, then unmap it -
10981 that is, return the original hardware frame.
10983 Otherwise, a device must be provided. Create new hardware surfaces on that
10984 device for the output, then map them back to the software format at the input
10985 and give those frames to the preceding filter. This will then act like the
10986 @option{hwupload} filter, but may be able to avoid an additional copy when
10987 the input is already in a compatible format.
10989 Hardware frame input and output
10991 A device must be supplied for the output, either directly or with the
10992 @option{derive_device} option. The input and output devices must be of
10993 different types and compatible - the exact meaning of this is
10994 system-dependent, but typically it means that they must refer to the same
10995 underlying hardware context (for example, refer to the same graphics card).
10997 If the input frames were originally created on the output device, then unmap
10998 to retrieve the original frames.
11000 Otherwise, map the frames to the output device - create new hardware frames
11001 on the output corresponding to the frames on the input.
11004 The following additional parameters are accepted:
11008 Set the frame mapping mode. Some combination of:
11011 The mapped frame should be readable.
11013 The mapped frame should be writeable.
11015 The mapping will always overwrite the entire frame.
11017 This may improve performance in some cases, as the original contents of the
11018 frame need not be loaded.
11020 The mapping must not involve any copying.
11022 Indirect mappings to copies of frames are created in some cases where either
11023 direct mapping is not possible or it would have unexpected properties.
11024 Setting this flag ensures that the mapping is direct and will fail if that is
11027 Defaults to @var{read+write} if not specified.
11029 @item derive_device @var{type}
11030 Rather than using the device supplied at initialisation, instead derive a new
11031 device of type @var{type} from the device the input frames exist on.
11034 In a hardware to hardware mapping, map in reverse - create frames in the sink
11035 and map them back to the source. This may be necessary in some cases where
11036 a mapping in one direction is required but only the opposite direction is
11037 supported by the devices being used.
11039 This option is dangerous - it may break the preceding filter in undefined
11040 ways if there are any additional constraints on that filter's output.
11041 Do not use it without fully understanding the implications of its use.
11047 Upload system memory frames to hardware surfaces.
11049 The device to upload to must be supplied when the filter is initialised. If
11050 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
11053 @anchor{hwupload_cuda}
11054 @section hwupload_cuda
11056 Upload system memory frames to a CUDA device.
11058 It accepts the following optional parameters:
11062 The number of the CUDA device to use
11067 Apply a high-quality magnification filter designed for pixel art. This filter
11068 was originally created by Maxim Stepin.
11070 It accepts the following option:
11074 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
11075 @code{hq3x} and @code{4} for @code{hq4x}.
11076 Default is @code{3}.
11080 Stack input videos horizontally.
11082 All streams must be of same pixel format and of same height.
11084 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
11085 to create same output.
11087 The filter accept the following option:
11091 Set number of input streams. Default is 2.
11094 If set to 1, force the output to terminate when the shortest input
11095 terminates. Default value is 0.
11100 Modify the hue and/or the saturation of the input.
11102 It accepts the following parameters:
11106 Specify the hue angle as a number of degrees. It accepts an expression,
11107 and defaults to "0".
11110 Specify the saturation in the [-10,10] range. It accepts an expression and
11114 Specify the hue angle as a number of radians. It accepts an
11115 expression, and defaults to "0".
11118 Specify the brightness in the [-10,10] range. It accepts an expression and
11122 @option{h} and @option{H} are mutually exclusive, and can't be
11123 specified at the same time.
11125 The @option{b}, @option{h}, @option{H} and @option{s} option values are
11126 expressions containing the following constants:
11130 frame count of the input frame starting from 0
11133 presentation timestamp of the input frame expressed in time base units
11136 frame rate of the input video, NAN if the input frame rate is unknown
11139 timestamp expressed in seconds, NAN if the input timestamp is unknown
11142 time base of the input video
11145 @subsection Examples
11149 Set the hue to 90 degrees and the saturation to 1.0:
11155 Same command but expressing the hue in radians:
11161 Rotate hue and make the saturation swing between 0
11162 and 2 over a period of 1 second:
11164 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11168 Apply a 3 seconds saturation fade-in effect starting at 0:
11170 hue="s=min(t/3\,1)"
11173 The general fade-in expression can be written as:
11175 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11179 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11181 hue="s=max(0\, min(1\, (8-t)/3))"
11184 The general fade-out expression can be written as:
11186 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11191 @subsection Commands
11193 This filter supports the following commands:
11199 Modify the hue and/or the saturation and/or brightness of the input video.
11200 The command accepts the same syntax of the corresponding option.
11202 If the specified expression is not valid, it is kept at its current
11206 @section hysteresis
11208 Grow first stream into second stream by connecting components.
11209 This makes it possible to build more robust edge masks.
11211 This filter accepts the following options:
11215 Set which planes will be processed as bitmap, unprocessed planes will be
11216 copied from first stream.
11217 By default value 0xf, all planes will be processed.
11220 Set threshold which is used in filtering. If pixel component value is higher than
11221 this value filter algorithm for connecting components is activated.
11222 By default value is 0.
11227 Detect video interlacing type.
11229 This filter tries to detect if the input frames are interlaced, progressive,
11230 top or bottom field first. It will also try to detect fields that are
11231 repeated between adjacent frames (a sign of telecine).
11233 Single frame detection considers only immediately adjacent frames when classifying each frame.
11234 Multiple frame detection incorporates the classification history of previous frames.
11236 The filter will log these metadata values:
11239 @item single.current_frame
11240 Detected type of current frame using single-frame detection. One of:
11241 ``tff'' (top field first), ``bff'' (bottom field first),
11242 ``progressive'', or ``undetermined''
11245 Cumulative number of frames detected as top field first using single-frame detection.
11248 Cumulative number of frames detected as top field first using multiple-frame detection.
11251 Cumulative number of frames detected as bottom field first using single-frame detection.
11253 @item multiple.current_frame
11254 Detected type of current frame using multiple-frame detection. One of:
11255 ``tff'' (top field first), ``bff'' (bottom field first),
11256 ``progressive'', or ``undetermined''
11259 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11261 @item single.progressive
11262 Cumulative number of frames detected as progressive using single-frame detection.
11264 @item multiple.progressive
11265 Cumulative number of frames detected as progressive using multiple-frame detection.
11267 @item single.undetermined
11268 Cumulative number of frames that could not be classified using single-frame detection.
11270 @item multiple.undetermined
11271 Cumulative number of frames that could not be classified using multiple-frame detection.
11273 @item repeated.current_frame
11274 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11276 @item repeated.neither
11277 Cumulative number of frames with no repeated field.
11280 Cumulative number of frames with the top field repeated from the previous frame's top field.
11282 @item repeated.bottom
11283 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11286 The filter accepts the following options:
11290 Set interlacing threshold.
11292 Set progressive threshold.
11294 Threshold for repeated field detection.
11296 Number of frames after which a given frame's contribution to the
11297 statistics is halved (i.e., it contributes only 0.5 to its
11298 classification). The default of 0 means that all frames seen are given
11299 full weight of 1.0 forever.
11300 @item analyze_interlaced_flag
11301 When this is not 0 then idet will use the specified number of frames to determine
11302 if the interlaced flag is accurate, it will not count undetermined frames.
11303 If the flag is found to be accurate it will be used without any further
11304 computations, if it is found to be inaccurate it will be cleared without any
11305 further computations. This allows inserting the idet filter as a low computational
11306 method to clean up the interlaced flag
11311 Deinterleave or interleave fields.
11313 This filter allows one to process interlaced images fields without
11314 deinterlacing them. Deinterleaving splits the input frame into 2
11315 fields (so called half pictures). Odd lines are moved to the top
11316 half of the output image, even lines to the bottom half.
11317 You can process (filter) them independently and then re-interleave them.
11319 The filter accepts the following options:
11323 @item chroma_mode, c
11324 @item alpha_mode, a
11325 Available values for @var{luma_mode}, @var{chroma_mode} and
11326 @var{alpha_mode} are:
11332 @item deinterleave, d
11333 Deinterleave fields, placing one above the other.
11335 @item interleave, i
11336 Interleave fields. Reverse the effect of deinterleaving.
11338 Default value is @code{none}.
11340 @item luma_swap, ls
11341 @item chroma_swap, cs
11342 @item alpha_swap, as
11343 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11348 Apply inflate effect to the video.
11350 This filter replaces the pixel by the local(3x3) average by taking into account
11351 only values higher than the pixel.
11353 It accepts the following options:
11360 Limit the maximum change for each plane, default is 65535.
11361 If 0, plane will remain unchanged.
11366 Simple interlacing filter from progressive contents. This interleaves upper (or
11367 lower) lines from odd frames with lower (or upper) lines from even frames,
11368 halving the frame rate and preserving image height.
11371 Original Original New Frame
11372 Frame 'j' Frame 'j+1' (tff)
11373 ========== =========== ==================
11374 Line 0 --------------------> Frame 'j' Line 0
11375 Line 1 Line 1 ----> Frame 'j+1' Line 1
11376 Line 2 ---------------------> Frame 'j' Line 2
11377 Line 3 Line 3 ----> Frame 'j+1' Line 3
11379 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11382 It accepts the following optional parameters:
11386 This determines whether the interlaced frame is taken from the even
11387 (tff - default) or odd (bff) lines of the progressive frame.
11390 Vertical lowpass filter to avoid twitter interlacing and
11391 reduce moire patterns.
11395 Disable vertical lowpass filter
11398 Enable linear filter (default)
11401 Enable complex filter. This will slightly less reduce twitter and moire
11402 but better retain detail and subjective sharpness impression.
11409 Deinterlace input video by applying Donald Graft's adaptive kernel
11410 deinterling. Work on interlaced parts of a video to produce
11411 progressive frames.
11413 The description of the accepted parameters follows.
11417 Set the threshold which affects the filter's tolerance when
11418 determining if a pixel line must be processed. It must be an integer
11419 in the range [0,255] and defaults to 10. A value of 0 will result in
11420 applying the process on every pixels.
11423 Paint pixels exceeding the threshold value to white if set to 1.
11427 Set the fields order. Swap fields if set to 1, leave fields alone if
11431 Enable additional sharpening if set to 1. Default is 0.
11434 Enable twoway sharpening if set to 1. Default is 0.
11437 @subsection Examples
11441 Apply default values:
11443 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11447 Enable additional sharpening:
11453 Paint processed pixels in white:
11461 Slowly update darker pixels.
11463 This filter makes short flashes of light appear longer.
11464 This filter accepts the following options:
11468 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
11471 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
11474 @section lenscorrection
11476 Correct radial lens distortion
11478 This filter can be used to correct for radial distortion as can result from the use
11479 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11480 one can use tools available for example as part of opencv or simply trial-and-error.
11481 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11482 and extract the k1 and k2 coefficients from the resulting matrix.
11484 Note that effectively the same filter is available in the open-source tools Krita and
11485 Digikam from the KDE project.
11487 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11488 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11489 brightness distribution, so you may want to use both filters together in certain
11490 cases, though you will have to take care of ordering, i.e. whether vignetting should
11491 be applied before or after lens correction.
11493 @subsection Options
11495 The filter accepts the following options:
11499 Relative x-coordinate of the focal point of the image, and thereby the center of the
11500 distortion. This value has a range [0,1] and is expressed as fractions of the image
11501 width. Default is 0.5.
11503 Relative y-coordinate of the focal point of the image, and thereby the center of the
11504 distortion. This value has a range [0,1] and is expressed as fractions of the image
11505 height. Default is 0.5.
11507 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11508 no correction. Default is 0.
11510 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11511 0 means no correction. Default is 0.
11514 The formula that generates the correction is:
11516 @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)
11518 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11519 distances from the focal point in the source and target images, respectively.
11523 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11525 The @code{lensfun} filter requires the camera make, camera model, and lens model
11526 to apply the lens correction. The filter will load the lensfun database and
11527 query it to find the corresponding camera and lens entries in the database. As
11528 long as these entries can be found with the given options, the filter can
11529 perform corrections on frames. Note that incomplete strings will result in the
11530 filter choosing the best match with the given options, and the filter will
11531 output the chosen camera and lens models (logged with level "info"). You must
11532 provide the make, camera model, and lens model as they are required.
11534 The filter accepts the following options:
11538 The make of the camera (for example, "Canon"). This option is required.
11541 The model of the camera (for example, "Canon EOS 100D"). This option is
11545 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11546 option is required.
11549 The type of correction to apply. The following values are valid options:
11553 Enables fixing lens vignetting.
11556 Enables fixing lens geometry. This is the default.
11559 Enables fixing chromatic aberrations.
11562 Enables fixing lens vignetting and lens geometry.
11565 Enables fixing lens vignetting and chromatic aberrations.
11568 Enables fixing both lens geometry and chromatic aberrations.
11571 Enables all possible corrections.
11575 The focal length of the image/video (zoom; expected constant for video). For
11576 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11577 range should be chosen when using that lens. Default 18.
11580 The aperture of the image/video (expected constant for video). Note that
11581 aperture is only used for vignetting correction. Default 3.5.
11583 @item focus_distance
11584 The focus distance of the image/video (expected constant for video). Note that
11585 focus distance is only used for vignetting and only slightly affects the
11586 vignetting correction process. If unknown, leave it at the default value (which
11590 The scale factor which is applied after transformation. After correction the
11591 video is no longer necessarily rectangular. This parameter controls how much of
11592 the resulting image is visible. The value 0 means that a value will be chosen
11593 automatically such that there is little or no unmapped area in the output
11594 image. 1.0 means that no additional scaling is done. Lower values may result
11595 in more of the corrected image being visible, while higher values may avoid
11596 unmapped areas in the output.
11598 @item target_geometry
11599 The target geometry of the output image/video. The following values are valid
11603 @item rectilinear (default)
11606 @item equirectangular
11607 @item fisheye_orthographic
11608 @item fisheye_stereographic
11609 @item fisheye_equisolid
11610 @item fisheye_thoby
11613 Apply the reverse of image correction (instead of correcting distortion, apply
11616 @item interpolation
11617 The type of interpolation used when correcting distortion. The following values
11622 @item linear (default)
11627 @subsection Examples
11631 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11632 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11636 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
11640 Apply the same as before, but only for the first 5 seconds of video.
11643 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
11650 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11651 score between two input videos.
11653 The obtained VMAF score is printed through the logging system.
11655 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11656 After installing the library it can be enabled using:
11657 @code{./configure --enable-libvmaf --enable-version3}.
11658 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11660 The filter has following options:
11664 Set the model path which is to be used for SVM.
11665 Default value: @code{"vmaf_v0.6.1.pkl"}
11668 Set the file path to be used to store logs.
11671 Set the format of the log file (xml or json).
11673 @item enable_transform
11674 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11675 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11676 Default value: @code{false}
11679 Invokes the phone model which will generate VMAF scores higher than in the
11680 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11683 Enables computing psnr along with vmaf.
11686 Enables computing ssim along with vmaf.
11689 Enables computing ms_ssim along with vmaf.
11692 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11695 Set number of threads to be used when computing vmaf.
11698 Set interval for frame subsampling used when computing vmaf.
11700 @item enable_conf_interval
11701 Enables confidence interval.
11704 This filter also supports the @ref{framesync} options.
11706 On the below examples the input file @file{main.mpg} being processed is
11707 compared with the reference file @file{ref.mpg}.
11710 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11713 Example with options:
11715 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11720 Limits the pixel components values to the specified range [min, max].
11722 The filter accepts the following options:
11726 Lower bound. Defaults to the lowest allowed value for the input.
11729 Upper bound. Defaults to the highest allowed value for the input.
11732 Specify which planes will be processed. Defaults to all available.
11739 The filter accepts the following options:
11743 Set the number of loops. Setting this value to -1 will result in infinite loops.
11747 Set maximal size in number of frames. Default is 0.
11750 Set first frame of loop. Default is 0.
11753 @subsection Examples
11757 Loop single first frame infinitely:
11759 loop=loop=-1:size=1:start=0
11763 Loop single first frame 10 times:
11765 loop=loop=10:size=1:start=0
11769 Loop 10 first frames 5 times:
11771 loop=loop=5:size=10:start=0
11777 Apply a 1D LUT to an input video.
11779 The filter accepts the following options:
11783 Set the 1D LUT file name.
11785 Currently supported formats:
11794 Select interpolation mode.
11796 Available values are:
11800 Use values from the nearest defined point.
11802 Interpolate values using the linear interpolation.
11804 Interpolate values using the cosine interpolation.
11806 Interpolate values using the cubic interpolation.
11808 Interpolate values using the spline interpolation.
11815 Apply a 3D LUT to an input video.
11817 The filter accepts the following options:
11821 Set the 3D LUT file name.
11823 Currently supported formats:
11837 Select interpolation mode.
11839 Available values are:
11843 Use values from the nearest defined point.
11845 Interpolate values using the 8 points defining a cube.
11847 Interpolate values using a tetrahedron.
11851 This filter also supports the @ref{framesync} options.
11855 Turn certain luma values into transparency.
11857 The filter accepts the following options:
11861 Set the luma which will be used as base for transparency.
11862 Default value is @code{0}.
11865 Set the range of luma values to be keyed out.
11866 Default value is @code{0}.
11869 Set the range of softness. Default value is @code{0}.
11870 Use this to control gradual transition from zero to full transparency.
11873 @section lut, lutrgb, lutyuv
11875 Compute a look-up table for binding each pixel component input value
11876 to an output value, and apply it to the input video.
11878 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11879 to an RGB input video.
11881 These filters accept the following parameters:
11884 set first pixel component expression
11886 set second pixel component expression
11888 set third pixel component expression
11890 set fourth pixel component expression, corresponds to the alpha component
11893 set red component expression
11895 set green component expression
11897 set blue component expression
11899 alpha component expression
11902 set Y/luminance component expression
11904 set U/Cb component expression
11906 set V/Cr component expression
11909 Each of them specifies the expression to use for computing the lookup table for
11910 the corresponding pixel component values.
11912 The exact component associated to each of the @var{c*} options depends on the
11915 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11916 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11918 The expressions can contain the following constants and functions:
11923 The input width and height.
11926 The input value for the pixel component.
11929 The input value, clipped to the @var{minval}-@var{maxval} range.
11932 The maximum value for the pixel component.
11935 The minimum value for the pixel component.
11938 The negated value for the pixel component value, clipped to the
11939 @var{minval}-@var{maxval} range; it corresponds to the expression
11940 "maxval-clipval+minval".
11943 The computed value in @var{val}, clipped to the
11944 @var{minval}-@var{maxval} range.
11946 @item gammaval(gamma)
11947 The computed gamma correction value of the pixel component value,
11948 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11950 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11954 All expressions default to "val".
11956 @subsection Examples
11960 Negate input video:
11962 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11963 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11966 The above is the same as:
11968 lutrgb="r=negval:g=negval:b=negval"
11969 lutyuv="y=negval:u=negval:v=negval"
11979 Remove chroma components, turning the video into a graytone image:
11981 lutyuv="u=128:v=128"
11985 Apply a luma burning effect:
11991 Remove green and blue components:
11997 Set a constant alpha channel value on input:
11999 format=rgba,lutrgb=a="maxval-minval/2"
12003 Correct luminance gamma by a factor of 0.5:
12005 lutyuv=y=gammaval(0.5)
12009 Discard least significant bits of luma:
12011 lutyuv=y='bitand(val, 128+64+32)'
12015 Technicolor like effect:
12017 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
12021 @section lut2, tlut2
12023 The @code{lut2} filter takes two input streams and outputs one
12026 The @code{tlut2} (time lut2) filter takes two consecutive frames
12027 from one single stream.
12029 This filter accepts the following parameters:
12032 set first pixel component expression
12034 set second pixel component expression
12036 set third pixel component expression
12038 set fourth pixel component expression, corresponds to the alpha component
12041 set output bit depth, only available for @code{lut2} filter. By default is 0,
12042 which means bit depth is automatically picked from first input format.
12045 Each of them specifies the expression to use for computing the lookup table for
12046 the corresponding pixel component values.
12048 The exact component associated to each of the @var{c*} options depends on the
12051 The expressions can contain the following constants:
12056 The input width and height.
12059 The first input value for the pixel component.
12062 The second input value for the pixel component.
12065 The first input video bit depth.
12068 The second input video bit depth.
12071 All expressions default to "x".
12073 @subsection Examples
12077 Highlight differences between two RGB video streams:
12079 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)'
12083 Highlight differences between two YUV video streams:
12085 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)'
12089 Show max difference between two video streams:
12091 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)))'
12095 @section maskedclamp
12097 Clamp the first input stream with the second input and third input stream.
12099 Returns the value of first stream to be between second input
12100 stream - @code{undershoot} and third input stream + @code{overshoot}.
12102 This filter accepts the following options:
12105 Default value is @code{0}.
12108 Default value is @code{0}.
12111 Set which planes will be processed as bitmap, unprocessed planes will be
12112 copied from first stream.
12113 By default value 0xf, all planes will be processed.
12116 @section maskedmerge
12118 Merge the first input stream with the second input stream using per pixel
12119 weights in the third input stream.
12121 A value of 0 in the third stream pixel component means that pixel component
12122 from first stream is returned unchanged, while maximum value (eg. 255 for
12123 8-bit videos) means that pixel component from second stream is returned
12124 unchanged. Intermediate values define the amount of merging between both
12125 input stream's pixel components.
12127 This filter accepts the following options:
12130 Set which planes will be processed as bitmap, unprocessed planes will be
12131 copied from first stream.
12132 By default value 0xf, all planes will be processed.
12136 Create mask from input video.
12138 For example it is useful to create motion masks after @code{tblend} filter.
12140 This filter accepts the following options:
12144 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
12147 Set high threshold. Any pixel component higher than this value will be set to max value
12148 allowed for current pixel format.
12151 Set planes to filter, by default all available planes are filtered.
12154 Fill all frame pixels with this value.
12157 Set max average pixel value for frame. If sum of all pixel components is higher that this
12158 average, output frame will be completely filled with value set by @var{fill} option.
12159 Typically useful for scene changes when used in combination with @code{tblend} filter.
12164 Apply motion-compensation deinterlacing.
12166 It needs one field per frame as input and must thus be used together
12167 with yadif=1/3 or equivalent.
12169 This filter accepts the following options:
12172 Set the deinterlacing mode.
12174 It accepts one of the following values:
12179 use iterative motion estimation
12181 like @samp{slow}, but use multiple reference frames.
12183 Default value is @samp{fast}.
12186 Set the picture field parity assumed for the input video. It must be
12187 one of the following values:
12191 assume top field first
12193 assume bottom field first
12196 Default value is @samp{bff}.
12199 Set per-block quantization parameter (QP) used by the internal
12202 Higher values should result in a smoother motion vector field but less
12203 optimal individual vectors. Default value is 1.
12206 @section mergeplanes
12208 Merge color channel components from several video streams.
12210 The filter accepts up to 4 input streams, and merge selected input
12211 planes to the output video.
12213 This filter accepts the following options:
12216 Set input to output plane mapping. Default is @code{0}.
12218 The mappings is specified as a bitmap. It should be specified as a
12219 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12220 mapping for the first plane of the output stream. 'A' sets the number of
12221 the input stream to use (from 0 to 3), and 'a' the plane number of the
12222 corresponding input to use (from 0 to 3). The rest of the mappings is
12223 similar, 'Bb' describes the mapping for the output stream second
12224 plane, 'Cc' describes the mapping for the output stream third plane and
12225 'Dd' describes the mapping for the output stream fourth plane.
12228 Set output pixel format. Default is @code{yuva444p}.
12231 @subsection Examples
12235 Merge three gray video streams of same width and height into single video stream:
12237 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12241 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12243 [a0][a1]mergeplanes=0x00010210:yuva444p
12247 Swap Y and A plane in yuva444p stream:
12249 format=yuva444p,mergeplanes=0x03010200:yuva444p
12253 Swap U and V plane in yuv420p stream:
12255 format=yuv420p,mergeplanes=0x000201:yuv420p
12259 Cast a rgb24 clip to yuv444p:
12261 format=rgb24,mergeplanes=0x000102:yuv444p
12267 Estimate and export motion vectors using block matching algorithms.
12268 Motion vectors are stored in frame side data to be used by other filters.
12270 This filter accepts the following options:
12273 Specify the motion estimation method. Accepts one of the following values:
12277 Exhaustive search algorithm.
12279 Three step search algorithm.
12281 Two dimensional logarithmic search algorithm.
12283 New three step search algorithm.
12285 Four step search algorithm.
12287 Diamond search algorithm.
12289 Hexagon-based search algorithm.
12291 Enhanced predictive zonal search algorithm.
12293 Uneven multi-hexagon search algorithm.
12295 Default value is @samp{esa}.
12298 Macroblock size. Default @code{16}.
12301 Search parameter. Default @code{7}.
12304 @section midequalizer
12306 Apply Midway Image Equalization effect using two video streams.
12308 Midway Image Equalization adjusts a pair of images to have the same
12309 histogram, while maintaining their dynamics as much as possible. It's
12310 useful for e.g. matching exposures from a pair of stereo cameras.
12312 This filter has two inputs and one output, which must be of same pixel format, but
12313 may be of different sizes. The output of filter is first input adjusted with
12314 midway histogram of both inputs.
12316 This filter accepts the following option:
12320 Set which planes to process. Default is @code{15}, which is all available planes.
12323 @section minterpolate
12325 Convert the video to specified frame rate using motion interpolation.
12327 This filter accepts the following options:
12330 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}.
12333 Motion interpolation mode. Following values are accepted:
12336 Duplicate previous or next frame for interpolating new ones.
12338 Blend source frames. Interpolated frame is mean of previous and next frames.
12340 Motion compensated interpolation. Following options are effective when this mode is selected:
12344 Motion compensation mode. Following values are accepted:
12347 Overlapped block motion compensation.
12349 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12351 Default mode is @samp{obmc}.
12354 Motion estimation mode. Following values are accepted:
12357 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12359 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12361 Default mode is @samp{bilat}.
12364 The algorithm to be used for motion estimation. Following values are accepted:
12367 Exhaustive search algorithm.
12369 Three step search algorithm.
12371 Two dimensional logarithmic search algorithm.
12373 New three step search algorithm.
12375 Four step search algorithm.
12377 Diamond search algorithm.
12379 Hexagon-based search algorithm.
12381 Enhanced predictive zonal search algorithm.
12383 Uneven multi-hexagon search algorithm.
12385 Default algorithm is @samp{epzs}.
12388 Macroblock size. Default @code{16}.
12391 Motion estimation search parameter. Default @code{32}.
12394 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).
12399 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:
12402 Disable scene change detection.
12404 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12406 Default method is @samp{fdiff}.
12408 @item scd_threshold
12409 Scene change detection threshold. Default is @code{5.0}.
12414 Mix several video input streams into one video stream.
12416 A description of the accepted options follows.
12420 The number of inputs. If unspecified, it defaults to 2.
12423 Specify weight of each input video stream as sequence.
12424 Each weight is separated by space. If number of weights
12425 is smaller than number of @var{frames} last specified
12426 weight will be used for all remaining unset weights.
12429 Specify scale, if it is set it will be multiplied with sum
12430 of each weight multiplied with pixel values to give final destination
12431 pixel value. By default @var{scale} is auto scaled to sum of weights.
12434 Specify how end of stream is determined.
12437 The duration of the longest input. (default)
12440 The duration of the shortest input.
12443 The duration of the first input.
12447 @section mpdecimate
12449 Drop frames that do not differ greatly from the previous frame in
12450 order to reduce frame rate.
12452 The main use of this filter is for very-low-bitrate encoding
12453 (e.g. streaming over dialup modem), but it could in theory be used for
12454 fixing movies that were inverse-telecined incorrectly.
12456 A description of the accepted options follows.
12460 Set the maximum number of consecutive frames which can be dropped (if
12461 positive), or the minimum interval between dropped frames (if
12462 negative). If the value is 0, the frame is dropped disregarding the
12463 number of previous sequentially dropped frames.
12465 Default value is 0.
12470 Set the dropping threshold values.
12472 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12473 represent actual pixel value differences, so a threshold of 64
12474 corresponds to 1 unit of difference for each pixel, or the same spread
12475 out differently over the block.
12477 A frame is a candidate for dropping if no 8x8 blocks differ by more
12478 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12479 meaning the whole image) differ by more than a threshold of @option{lo}.
12481 Default value for @option{hi} is 64*12, default value for @option{lo} is
12482 64*5, and default value for @option{frac} is 0.33.
12488 Negate (invert) the input video.
12490 It accepts the following option:
12495 With value 1, it negates the alpha component, if present. Default value is 0.
12501 Denoise frames using Non-Local Means algorithm.
12503 Each pixel is adjusted by looking for other pixels with similar contexts. This
12504 context similarity is defined by comparing their surrounding patches of size
12505 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12508 Note that the research area defines centers for patches, which means some
12509 patches will be made of pixels outside that research area.
12511 The filter accepts the following options.
12515 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12518 Set patch size. Default is 7. Must be odd number in range [0, 99].
12521 Same as @option{p} but for chroma planes.
12523 The default value is @var{0} and means automatic.
12526 Set research size. Default is 15. Must be odd number in range [0, 99].
12529 Same as @option{r} but for chroma planes.
12531 The default value is @var{0} and means automatic.
12536 Deinterlace video using neural network edge directed interpolation.
12538 This filter accepts the following options:
12542 Mandatory option, without binary file filter can not work.
12543 Currently file can be found here:
12544 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12547 Set which frames to deinterlace, by default it is @code{all}.
12548 Can be @code{all} or @code{interlaced}.
12551 Set mode of operation.
12553 Can be one of the following:
12557 Use frame flags, both fields.
12559 Use frame flags, single field.
12561 Use top field only.
12563 Use bottom field only.
12565 Use both fields, top first.
12567 Use both fields, bottom first.
12571 Set which planes to process, by default filter process all frames.
12574 Set size of local neighborhood around each pixel, used by the predictor neural
12577 Can be one of the following:
12590 Set the number of neurons in predictor neural network.
12591 Can be one of the following:
12602 Controls the number of different neural network predictions that are blended
12603 together to compute the final output value. Can be @code{fast}, default or
12607 Set which set of weights to use in the predictor.
12608 Can be one of the following:
12612 weights trained to minimize absolute error
12614 weights trained to minimize squared error
12618 Controls whether or not the prescreener neural network is used to decide
12619 which pixels should be processed by the predictor neural network and which
12620 can be handled by simple cubic interpolation.
12621 The prescreener is trained to know whether cubic interpolation will be
12622 sufficient for a pixel or whether it should be predicted by the predictor nn.
12623 The computational complexity of the prescreener nn is much less than that of
12624 the predictor nn. Since most pixels can be handled by cubic interpolation,
12625 using the prescreener generally results in much faster processing.
12626 The prescreener is pretty accurate, so the difference between using it and not
12627 using it is almost always unnoticeable.
12629 Can be one of the following:
12637 Default is @code{new}.
12640 Set various debugging flags.
12645 Force libavfilter not to use any of the specified pixel formats for the
12646 input to the next filter.
12648 It accepts the following parameters:
12652 A '|'-separated list of pixel format names, such as
12653 pix_fmts=yuv420p|monow|rgb24".
12657 @subsection Examples
12661 Force libavfilter to use a format different from @var{yuv420p} for the
12662 input to the vflip filter:
12664 noformat=pix_fmts=yuv420p,vflip
12668 Convert the input video to any of the formats not contained in the list:
12670 noformat=yuv420p|yuv444p|yuv410p
12676 Add noise on video input frame.
12678 The filter accepts the following options:
12686 Set noise seed for specific pixel component or all pixel components in case
12687 of @var{all_seed}. Default value is @code{123457}.
12689 @item all_strength, alls
12690 @item c0_strength, c0s
12691 @item c1_strength, c1s
12692 @item c2_strength, c2s
12693 @item c3_strength, c3s
12694 Set noise strength for specific pixel component or all pixel components in case
12695 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12697 @item all_flags, allf
12698 @item c0_flags, c0f
12699 @item c1_flags, c1f
12700 @item c2_flags, c2f
12701 @item c3_flags, c3f
12702 Set pixel component flags or set flags for all components if @var{all_flags}.
12703 Available values for component flags are:
12706 averaged temporal noise (smoother)
12708 mix random noise with a (semi)regular pattern
12710 temporal noise (noise pattern changes between frames)
12712 uniform noise (gaussian otherwise)
12716 @subsection Examples
12718 Add temporal and uniform noise to input video:
12720 noise=alls=20:allf=t+u
12725 Normalize RGB video (aka histogram stretching, contrast stretching).
12726 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12728 For each channel of each frame, the filter computes the input range and maps
12729 it linearly to the user-specified output range. The output range defaults
12730 to the full dynamic range from pure black to pure white.
12732 Temporal smoothing can be used on the input range to reduce flickering (rapid
12733 changes in brightness) caused when small dark or bright objects enter or leave
12734 the scene. This is similar to the auto-exposure (automatic gain control) on a
12735 video camera, and, like a video camera, it may cause a period of over- or
12736 under-exposure of the video.
12738 The R,G,B channels can be normalized independently, which may cause some
12739 color shifting, or linked together as a single channel, which prevents
12740 color shifting. Linked normalization preserves hue. Independent normalization
12741 does not, so it can be used to remove some color casts. Independent and linked
12742 normalization can be combined in any ratio.
12744 The normalize filter accepts the following options:
12749 Colors which define the output range. The minimum input value is mapped to
12750 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12751 The defaults are black and white respectively. Specifying white for
12752 @var{blackpt} and black for @var{whitept} will give color-inverted,
12753 normalized video. Shades of grey can be used to reduce the dynamic range
12754 (contrast). Specifying saturated colors here can create some interesting
12758 The number of previous frames to use for temporal smoothing. The input range
12759 of each channel is smoothed using a rolling average over the current frame
12760 and the @var{smoothing} previous frames. The default is 0 (no temporal
12764 Controls the ratio of independent (color shifting) channel normalization to
12765 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12766 independent. Defaults to 1.0 (fully independent).
12769 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12770 expensive no-op. Defaults to 1.0 (full strength).
12774 @subsection Examples
12776 Stretch video contrast to use the full dynamic range, with no temporal
12777 smoothing; may flicker depending on the source content:
12779 normalize=blackpt=black:whitept=white:smoothing=0
12782 As above, but with 50 frames of temporal smoothing; flicker should be
12783 reduced, depending on the source content:
12785 normalize=blackpt=black:whitept=white:smoothing=50
12788 As above, but with hue-preserving linked channel normalization:
12790 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12793 As above, but with half strength:
12795 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12798 Map the darkest input color to red, the brightest input color to cyan:
12800 normalize=blackpt=red:whitept=cyan
12805 Pass the video source unchanged to the output.
12808 Optical Character Recognition
12810 This filter uses Tesseract for optical character recognition. To enable
12811 compilation of this filter, you need to configure FFmpeg with
12812 @code{--enable-libtesseract}.
12814 It accepts the following options:
12818 Set datapath to tesseract data. Default is to use whatever was
12819 set at installation.
12822 Set language, default is "eng".
12825 Set character whitelist.
12828 Set character blacklist.
12831 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12835 Apply a video transform using libopencv.
12837 To enable this filter, install the libopencv library and headers and
12838 configure FFmpeg with @code{--enable-libopencv}.
12840 It accepts the following parameters:
12845 The name of the libopencv filter to apply.
12847 @item filter_params
12848 The parameters to pass to the libopencv filter. If not specified, the default
12849 values are assumed.
12853 Refer to the official libopencv documentation for more precise
12855 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12857 Several libopencv filters are supported; see the following subsections.
12862 Dilate an image by using a specific structuring element.
12863 It corresponds to the libopencv function @code{cvDilate}.
12865 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12867 @var{struct_el} represents a structuring element, and has the syntax:
12868 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12870 @var{cols} and @var{rows} represent the number of columns and rows of
12871 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12872 point, and @var{shape} the shape for the structuring element. @var{shape}
12873 must be "rect", "cross", "ellipse", or "custom".
12875 If the value for @var{shape} is "custom", it must be followed by a
12876 string of the form "=@var{filename}". The file with name
12877 @var{filename} is assumed to represent a binary image, with each
12878 printable character corresponding to a bright pixel. When a custom
12879 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12880 or columns and rows of the read file are assumed instead.
12882 The default value for @var{struct_el} is "3x3+0x0/rect".
12884 @var{nb_iterations} specifies the number of times the transform is
12885 applied to the image, and defaults to 1.
12889 # Use the default values
12892 # Dilate using a structuring element with a 5x5 cross, iterating two times
12893 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12895 # Read the shape from the file diamond.shape, iterating two times.
12896 # The file diamond.shape may contain a pattern of characters like this
12902 # The specified columns and rows are ignored
12903 # but the anchor point coordinates are not
12904 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12909 Erode an image by using a specific structuring element.
12910 It corresponds to the libopencv function @code{cvErode}.
12912 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12913 with the same syntax and semantics as the @ref{dilate} filter.
12917 Smooth the input video.
12919 The filter takes the following parameters:
12920 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12922 @var{type} is the type of smooth filter to apply, and must be one of
12923 the following values: "blur", "blur_no_scale", "median", "gaussian",
12924 or "bilateral". The default value is "gaussian".
12926 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12927 depend on the smooth type. @var{param1} and
12928 @var{param2} accept integer positive values or 0. @var{param3} and
12929 @var{param4} accept floating point values.
12931 The default value for @var{param1} is 3. The default value for the
12932 other parameters is 0.
12934 These parameters correspond to the parameters assigned to the
12935 libopencv function @code{cvSmooth}.
12937 @section oscilloscope
12939 2D Video Oscilloscope.
12941 Useful to measure spatial impulse, step responses, chroma delays, etc.
12943 It accepts the following parameters:
12947 Set scope center x position.
12950 Set scope center y position.
12953 Set scope size, relative to frame diagonal.
12956 Set scope tilt/rotation.
12962 Set trace center x position.
12965 Set trace center y position.
12968 Set trace width, relative to width of frame.
12971 Set trace height, relative to height of frame.
12974 Set which components to trace. By default it traces first three components.
12977 Draw trace grid. By default is enabled.
12980 Draw some statistics. By default is enabled.
12983 Draw scope. By default is enabled.
12986 @subsection Examples
12990 Inspect full first row of video frame.
12992 oscilloscope=x=0.5:y=0:s=1
12996 Inspect full last row of video frame.
12998 oscilloscope=x=0.5:y=1:s=1
13002 Inspect full 5th line of video frame of height 1080.
13004 oscilloscope=x=0.5:y=5/1080:s=1
13008 Inspect full last column of video frame.
13010 oscilloscope=x=1:y=0.5:s=1:t=1
13018 Overlay one video on top of another.
13020 It takes two inputs and has one output. The first input is the "main"
13021 video on which the second input is overlaid.
13023 It accepts the following parameters:
13025 A description of the accepted options follows.
13030 Set the expression for the x and y coordinates of the overlaid video
13031 on the main video. Default value is "0" for both expressions. In case
13032 the expression is invalid, it is set to a huge value (meaning that the
13033 overlay will not be displayed within the output visible area).
13036 See @ref{framesync}.
13039 Set when the expressions for @option{x}, and @option{y} are evaluated.
13041 It accepts the following values:
13044 only evaluate expressions once during the filter initialization or
13045 when a command is processed
13048 evaluate expressions for each incoming frame
13051 Default value is @samp{frame}.
13054 See @ref{framesync}.
13057 Set the format for the output video.
13059 It accepts the following values:
13062 force YUV420 output
13065 force YUV422 output
13068 force YUV444 output
13071 force packed RGB output
13074 force planar RGB output
13077 automatically pick format
13080 Default value is @samp{yuv420}.
13083 See @ref{framesync}.
13086 Set format of alpha of the overlaid video, it can be @var{straight} or
13087 @var{premultiplied}. Default is @var{straight}.
13090 The @option{x}, and @option{y} expressions can contain the following
13096 The main input width and height.
13100 The overlay input width and height.
13104 The computed values for @var{x} and @var{y}. They are evaluated for
13109 horizontal and vertical chroma subsample values of the output
13110 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
13114 the number of input frame, starting from 0
13117 the position in the file of the input frame, NAN if unknown
13120 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
13124 This filter also supports the @ref{framesync} options.
13126 Note that the @var{n}, @var{pos}, @var{t} variables are available only
13127 when evaluation is done @emph{per frame}, and will evaluate to NAN
13128 when @option{eval} is set to @samp{init}.
13130 Be aware that frames are taken from each input video in timestamp
13131 order, hence, if their initial timestamps differ, it is a good idea
13132 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
13133 have them begin in the same zero timestamp, as the example for
13134 the @var{movie} filter does.
13136 You can chain together more overlays but you should test the
13137 efficiency of such approach.
13139 @subsection Commands
13141 This filter supports the following commands:
13145 Modify the x and y of the overlay input.
13146 The command accepts the same syntax of the corresponding option.
13148 If the specified expression is not valid, it is kept at its current
13152 @subsection Examples
13156 Draw the overlay at 10 pixels from the bottom right corner of the main
13159 overlay=main_w-overlay_w-10:main_h-overlay_h-10
13162 Using named options the example above becomes:
13164 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
13168 Insert a transparent PNG logo in the bottom left corner of the input,
13169 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
13171 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
13175 Insert 2 different transparent PNG logos (second logo on bottom
13176 right corner) using the @command{ffmpeg} tool:
13178 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
13182 Add a transparent color layer on top of the main video; @code{WxH}
13183 must specify the size of the main input to the overlay filter:
13185 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
13189 Play an original video and a filtered version (here with the deshake
13190 filter) side by side using the @command{ffplay} tool:
13192 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13195 The above command is the same as:
13197 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13201 Make a sliding overlay appearing from the left to the right top part of the
13202 screen starting since time 2:
13204 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13208 Compose output by putting two input videos side to side:
13210 ffmpeg -i left.avi -i right.avi -filter_complex "
13211 nullsrc=size=200x100 [background];
13212 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13213 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13214 [background][left] overlay=shortest=1 [background+left];
13215 [background+left][right] overlay=shortest=1:x=100 [left+right]
13220 Mask 10-20 seconds of a video by applying the delogo filter to a section
13222 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13223 -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]'
13228 Chain several overlays in cascade:
13230 nullsrc=s=200x200 [bg];
13231 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13232 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13233 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13234 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13235 [in3] null, [mid2] overlay=100:100 [out0]
13242 Apply Overcomplete Wavelet denoiser.
13244 The filter accepts the following options:
13250 Larger depth values will denoise lower frequency components more, but
13251 slow down filtering.
13253 Must be an int in the range 8-16, default is @code{8}.
13255 @item luma_strength, ls
13258 Must be a double value in the range 0-1000, default is @code{1.0}.
13260 @item chroma_strength, cs
13261 Set chroma strength.
13263 Must be a double value in the range 0-1000, default is @code{1.0}.
13269 Add paddings to the input image, and place the original input at the
13270 provided @var{x}, @var{y} coordinates.
13272 It accepts the following parameters:
13277 Specify an expression for the size of the output image with the
13278 paddings added. If the value for @var{width} or @var{height} is 0, the
13279 corresponding input size is used for the output.
13281 The @var{width} expression can reference the value set by the
13282 @var{height} expression, and vice versa.
13284 The default value of @var{width} and @var{height} is 0.
13288 Specify the offsets to place the input image at within the padded area,
13289 with respect to the top/left border of the output image.
13291 The @var{x} expression can reference the value set by the @var{y}
13292 expression, and vice versa.
13294 The default value of @var{x} and @var{y} is 0.
13296 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13297 so the input image is centered on the padded area.
13300 Specify the color of the padded area. For the syntax of this option,
13301 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13302 manual,ffmpeg-utils}.
13304 The default value of @var{color} is "black".
13307 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13309 It accepts the following values:
13313 Only evaluate expressions once during the filter initialization or when
13314 a command is processed.
13317 Evaluate expressions for each incoming frame.
13321 Default value is @samp{init}.
13324 Pad to aspect instead to a resolution.
13328 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13329 options are expressions containing the following constants:
13334 The input video width and height.
13338 These are the same as @var{in_w} and @var{in_h}.
13342 The output width and height (the size of the padded area), as
13343 specified by the @var{width} and @var{height} expressions.
13347 These are the same as @var{out_w} and @var{out_h}.
13351 The x and y offsets as specified by the @var{x} and @var{y}
13352 expressions, or NAN if not yet specified.
13355 same as @var{iw} / @var{ih}
13358 input sample aspect ratio
13361 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13365 The horizontal and vertical chroma subsample values. For example for the
13366 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13369 @subsection Examples
13373 Add paddings with the color "violet" to the input video. The output video
13374 size is 640x480, and the top-left corner of the input video is placed at
13377 pad=640:480:0:40:violet
13380 The example above is equivalent to the following command:
13382 pad=width=640:height=480:x=0:y=40:color=violet
13386 Pad the input to get an output with dimensions increased by 3/2,
13387 and put the input video at the center of the padded area:
13389 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13393 Pad the input to get a squared output with size equal to the maximum
13394 value between the input width and height, and put the input video at
13395 the center of the padded area:
13397 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13401 Pad the input to get a final w/h ratio of 16:9:
13403 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13407 In case of anamorphic video, in order to set the output display aspect
13408 correctly, it is necessary to use @var{sar} in the expression,
13409 according to the relation:
13411 (ih * X / ih) * sar = output_dar
13412 X = output_dar / sar
13415 Thus the previous example needs to be modified to:
13417 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13421 Double the output size and put the input video in the bottom-right
13422 corner of the output padded area:
13424 pad="2*iw:2*ih:ow-iw:oh-ih"
13428 @anchor{palettegen}
13429 @section palettegen
13431 Generate one palette for a whole video stream.
13433 It accepts the following options:
13437 Set the maximum number of colors to quantize in the palette.
13438 Note: the palette will still contain 256 colors; the unused palette entries
13441 @item reserve_transparent
13442 Create a palette of 255 colors maximum and reserve the last one for
13443 transparency. Reserving the transparency color is useful for GIF optimization.
13444 If not set, the maximum of colors in the palette will be 256. You probably want
13445 to disable this option for a standalone image.
13448 @item transparency_color
13449 Set the color that will be used as background for transparency.
13452 Set statistics mode.
13454 It accepts the following values:
13457 Compute full frame histograms.
13459 Compute histograms only for the part that differs from previous frame. This
13460 might be relevant to give more importance to the moving part of your input if
13461 the background is static.
13463 Compute new histogram for each frame.
13466 Default value is @var{full}.
13469 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13470 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13471 color quantization of the palette. This information is also visible at
13472 @var{info} logging level.
13474 @subsection Examples
13478 Generate a representative palette of a given video using @command{ffmpeg}:
13480 ffmpeg -i input.mkv -vf palettegen palette.png
13484 @section paletteuse
13486 Use a palette to downsample an input video stream.
13488 The filter takes two inputs: one video stream and a palette. The palette must
13489 be a 256 pixels image.
13491 It accepts the following options:
13495 Select dithering mode. Available algorithms are:
13498 Ordered 8x8 bayer dithering (deterministic)
13500 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13501 Note: this dithering is sometimes considered "wrong" and is included as a
13503 @item floyd_steinberg
13504 Floyd and Steingberg dithering (error diffusion)
13506 Frankie Sierra dithering v2 (error diffusion)
13508 Frankie Sierra dithering v2 "Lite" (error diffusion)
13511 Default is @var{sierra2_4a}.
13514 When @var{bayer} dithering is selected, this option defines the scale of the
13515 pattern (how much the crosshatch pattern is visible). A low value means more
13516 visible pattern for less banding, and higher value means less visible pattern
13517 at the cost of more banding.
13519 The option must be an integer value in the range [0,5]. Default is @var{2}.
13522 If set, define the zone to process
13526 Only the changing rectangle will be reprocessed. This is similar to GIF
13527 cropping/offsetting compression mechanism. This option can be useful for speed
13528 if only a part of the image is changing, and has use cases such as limiting the
13529 scope of the error diffusal @option{dither} to the rectangle that bounds the
13530 moving scene (it leads to more deterministic output if the scene doesn't change
13531 much, and as a result less moving noise and better GIF compression).
13534 Default is @var{none}.
13537 Take new palette for each output frame.
13539 @item alpha_threshold
13540 Sets the alpha threshold for transparency. Alpha values above this threshold
13541 will be treated as completely opaque, and values below this threshold will be
13542 treated as completely transparent.
13544 The option must be an integer value in the range [0,255]. Default is @var{128}.
13547 @subsection Examples
13551 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13552 using @command{ffmpeg}:
13554 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13558 @section perspective
13560 Correct perspective of video not recorded perpendicular to the screen.
13562 A description of the accepted parameters follows.
13573 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13574 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13575 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13576 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13577 then the corners of the source will be sent to the specified coordinates.
13579 The expressions can use the following variables:
13584 the width and height of video frame.
13588 Output frame count.
13591 @item interpolation
13592 Set interpolation for perspective correction.
13594 It accepts the following values:
13600 Default value is @samp{linear}.
13603 Set interpretation of coordinate options.
13605 It accepts the following values:
13609 Send point in the source specified by the given coordinates to
13610 the corners of the destination.
13612 @item 1, destination
13614 Send the corners of the source to the point in the destination specified
13615 by the given coordinates.
13617 Default value is @samp{source}.
13621 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13623 It accepts the following values:
13626 only evaluate expressions once during the filter initialization or
13627 when a command is processed
13630 evaluate expressions for each incoming frame
13633 Default value is @samp{init}.
13638 Delay interlaced video by one field time so that the field order changes.
13640 The intended use is to fix PAL movies that have been captured with the
13641 opposite field order to the film-to-video transfer.
13643 A description of the accepted parameters follows.
13649 It accepts the following values:
13652 Capture field order top-first, transfer bottom-first.
13653 Filter will delay the bottom field.
13656 Capture field order bottom-first, transfer top-first.
13657 Filter will delay the top field.
13660 Capture and transfer with the same field order. This mode only exists
13661 for the documentation of the other options to refer to, but if you
13662 actually select it, the filter will faithfully do nothing.
13665 Capture field order determined automatically by field flags, transfer
13667 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13668 basis using field flags. If no field information is available,
13669 then this works just like @samp{u}.
13672 Capture unknown or varying, transfer opposite.
13673 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13674 analyzing the images and selecting the alternative that produces best
13675 match between the fields.
13678 Capture top-first, transfer unknown or varying.
13679 Filter selects among @samp{t} and @samp{p} using image analysis.
13682 Capture bottom-first, transfer unknown or varying.
13683 Filter selects among @samp{b} and @samp{p} using image analysis.
13686 Capture determined by field flags, transfer unknown or varying.
13687 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13688 image analysis. If no field information is available, then this works just
13689 like @samp{U}. This is the default mode.
13692 Both capture and transfer unknown or varying.
13693 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13697 @section pixdesctest
13699 Pixel format descriptor test filter, mainly useful for internal
13700 testing. The output video should be equal to the input video.
13704 format=monow, pixdesctest
13707 can be used to test the monowhite pixel format descriptor definition.
13711 Display sample values of color channels. Mainly useful for checking color
13712 and levels. Minimum supported resolution is 640x480.
13714 The filters accept the following options:
13718 Set scope X position, relative offset on X axis.
13721 Set scope Y position, relative offset on Y axis.
13730 Set window opacity. This window also holds statistics about pixel area.
13733 Set window X position, relative offset on X axis.
13736 Set window Y position, relative offset on Y axis.
13741 Enable the specified chain of postprocessing subfilters using libpostproc. This
13742 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13743 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13744 Each subfilter and some options have a short and a long name that can be used
13745 interchangeably, i.e. dr/dering are the same.
13747 The filters accept the following options:
13751 Set postprocessing subfilters string.
13754 All subfilters share common options to determine their scope:
13758 Honor the quality commands for this subfilter.
13761 Do chrominance filtering, too (default).
13764 Do luminance filtering only (no chrominance).
13767 Do chrominance filtering only (no luminance).
13770 These options can be appended after the subfilter name, separated by a '|'.
13772 Available subfilters are:
13775 @item hb/hdeblock[|difference[|flatness]]
13776 Horizontal deblocking filter
13779 Difference factor where higher values mean more deblocking (default: @code{32}).
13781 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13784 @item vb/vdeblock[|difference[|flatness]]
13785 Vertical deblocking filter
13788 Difference factor where higher values mean more deblocking (default: @code{32}).
13790 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13793 @item ha/hadeblock[|difference[|flatness]]
13794 Accurate horizontal deblocking filter
13797 Difference factor where higher values mean more deblocking (default: @code{32}).
13799 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13802 @item va/vadeblock[|difference[|flatness]]
13803 Accurate vertical deblocking filter
13806 Difference factor where higher values mean more deblocking (default: @code{32}).
13808 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13812 The horizontal and vertical deblocking filters share the difference and
13813 flatness values so you cannot set different horizontal and vertical
13817 @item h1/x1hdeblock
13818 Experimental horizontal deblocking filter
13820 @item v1/x1vdeblock
13821 Experimental vertical deblocking filter
13826 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13829 larger -> stronger filtering
13831 larger -> stronger filtering
13833 larger -> stronger filtering
13836 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13839 Stretch luminance to @code{0-255}.
13842 @item lb/linblenddeint
13843 Linear blend deinterlacing filter that deinterlaces the given block by
13844 filtering all lines with a @code{(1 2 1)} filter.
13846 @item li/linipoldeint
13847 Linear interpolating deinterlacing filter that deinterlaces the given block by
13848 linearly interpolating every second line.
13850 @item ci/cubicipoldeint
13851 Cubic interpolating deinterlacing filter deinterlaces the given block by
13852 cubically interpolating every second line.
13854 @item md/mediandeint
13855 Median deinterlacing filter that deinterlaces the given block by applying a
13856 median filter to every second line.
13858 @item fd/ffmpegdeint
13859 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13860 second line with a @code{(-1 4 2 4 -1)} filter.
13863 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13864 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13866 @item fq/forceQuant[|quantizer]
13867 Overrides the quantizer table from the input with the constant quantizer you
13875 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13878 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13881 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13884 @subsection Examples
13888 Apply horizontal and vertical deblocking, deringing and automatic
13889 brightness/contrast:
13895 Apply default filters without brightness/contrast correction:
13901 Apply default filters and temporal denoiser:
13903 pp=default/tmpnoise|1|2|3
13907 Apply deblocking on luminance only, and switch vertical deblocking on or off
13908 automatically depending on available CPU time:
13915 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13916 similar to spp = 6 with 7 point DCT, where only the center sample is
13919 The filter accepts the following options:
13923 Force a constant quantization parameter. It accepts an integer in range
13924 0 to 63. If not set, the filter will use the QP from the video stream
13928 Set thresholding mode. Available modes are:
13932 Set hard thresholding.
13934 Set soft thresholding (better de-ringing effect, but likely blurrier).
13936 Set medium thresholding (good results, default).
13940 @section premultiply
13941 Apply alpha premultiply effect to input video stream using first plane
13942 of second stream as alpha.
13944 Both streams must have same dimensions and same pixel format.
13946 The filter accepts the following option:
13950 Set which planes will be processed, unprocessed planes will be copied.
13951 By default value 0xf, all planes will be processed.
13954 Do not require 2nd input for processing, instead use alpha plane from input stream.
13958 Apply prewitt operator to input video stream.
13960 The filter accepts the following option:
13964 Set which planes will be processed, unprocessed planes will be copied.
13965 By default value 0xf, all planes will be processed.
13968 Set value which will be multiplied with filtered result.
13971 Set value which will be added to filtered result.
13974 @anchor{program_opencl}
13975 @section program_opencl
13977 Filter video using an OpenCL program.
13982 OpenCL program source file.
13985 Kernel name in program.
13988 Number of inputs to the filter. Defaults to 1.
13991 Size of output frames. Defaults to the same as the first input.
13995 The program source file must contain a kernel function with the given name,
13996 which will be run once for each plane of the output. Each run on a plane
13997 gets enqueued as a separate 2D global NDRange with one work-item for each
13998 pixel to be generated. The global ID offset for each work-item is therefore
13999 the coordinates of a pixel in the destination image.
14001 The kernel function needs to take the following arguments:
14004 Destination image, @var{__write_only image2d_t}.
14006 This image will become the output; the kernel should write all of it.
14008 Frame index, @var{unsigned int}.
14010 This is a counter starting from zero and increasing by one for each frame.
14012 Source images, @var{__read_only image2d_t}.
14014 These are the most recent images on each input. The kernel may read from
14015 them to generate the output, but they can't be written to.
14022 Copy the input to the output (output must be the same size as the input).
14024 __kernel void copy(__write_only image2d_t destination,
14025 unsigned int index,
14026 __read_only image2d_t source)
14028 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
14030 int2 location = (int2)(get_global_id(0), get_global_id(1));
14032 float4 value = read_imagef(source, sampler, location);
14034 write_imagef(destination, location, value);
14039 Apply a simple transformation, rotating the input by an amount increasing
14040 with the index counter. Pixel values are linearly interpolated by the
14041 sampler, and the output need not have the same dimensions as the input.
14043 __kernel void rotate_image(__write_only image2d_t dst,
14044 unsigned int index,
14045 __read_only image2d_t src)
14047 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
14048 CLK_FILTER_LINEAR);
14050 float angle = (float)index / 100.0f;
14052 float2 dst_dim = convert_float2(get_image_dim(dst));
14053 float2 src_dim = convert_float2(get_image_dim(src));
14055 float2 dst_cen = dst_dim / 2.0f;
14056 float2 src_cen = src_dim / 2.0f;
14058 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
14060 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
14062 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
14063 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
14065 src_pos = src_pos * src_dim / dst_dim;
14067 float2 src_loc = src_pos + src_cen;
14069 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
14070 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
14071 write_imagef(dst, dst_loc, 0.5f);
14073 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
14078 Blend two inputs together, with the amount of each input used varying
14079 with the index counter.
14081 __kernel void blend_images(__write_only image2d_t dst,
14082 unsigned int index,
14083 __read_only image2d_t src1,
14084 __read_only image2d_t src2)
14086 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
14087 CLK_FILTER_LINEAR);
14089 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
14091 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
14092 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
14093 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
14095 float4 val1 = read_imagef(src1, sampler, src1_loc);
14096 float4 val2 = read_imagef(src2, sampler, src2_loc);
14098 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
14104 @section pseudocolor
14106 Alter frame colors in video with pseudocolors.
14108 This filter accept the following options:
14112 set pixel first component expression
14115 set pixel second component expression
14118 set pixel third component expression
14121 set pixel fourth component expression, corresponds to the alpha component
14124 set component to use as base for altering colors
14127 Each of them specifies the expression to use for computing the lookup table for
14128 the corresponding pixel component values.
14130 The expressions can contain the following constants and functions:
14135 The input width and height.
14138 The input value for the pixel component.
14140 @item ymin, umin, vmin, amin
14141 The minimum allowed component value.
14143 @item ymax, umax, vmax, amax
14144 The maximum allowed component value.
14147 All expressions default to "val".
14149 @subsection Examples
14153 Change too high luma values to gradient:
14155 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'"
14161 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
14162 Ratio) between two input videos.
14164 This filter takes in input two input videos, the first input is
14165 considered the "main" source and is passed unchanged to the
14166 output. The second input is used as a "reference" video for computing
14169 Both video inputs must have the same resolution and pixel format for
14170 this filter to work correctly. Also it assumes that both inputs
14171 have the same number of frames, which are compared one by one.
14173 The obtained average PSNR is printed through the logging system.
14175 The filter stores the accumulated MSE (mean squared error) of each
14176 frame, and at the end of the processing it is averaged across all frames
14177 equally, and the following formula is applied to obtain the PSNR:
14180 PSNR = 10*log10(MAX^2/MSE)
14183 Where MAX is the average of the maximum values of each component of the
14186 The description of the accepted parameters follows.
14189 @item stats_file, f
14190 If specified the filter will use the named file to save the PSNR of
14191 each individual frame. When filename equals "-" the data is sent to
14194 @item stats_version
14195 Specifies which version of the stats file format to use. Details of
14196 each format are written below.
14197 Default value is 1.
14199 @item stats_add_max
14200 Determines whether the max value is output to the stats log.
14201 Default value is 0.
14202 Requires stats_version >= 2. If this is set and stats_version < 2,
14203 the filter will return an error.
14206 This filter also supports the @ref{framesync} options.
14208 The file printed if @var{stats_file} is selected, contains a sequence of
14209 key/value pairs of the form @var{key}:@var{value} for each compared
14212 If a @var{stats_version} greater than 1 is specified, a header line precedes
14213 the list of per-frame-pair stats, with key value pairs following the frame
14214 format with the following parameters:
14217 @item psnr_log_version
14218 The version of the log file format. Will match @var{stats_version}.
14221 A comma separated list of the per-frame-pair parameters included in
14225 A description of each shown per-frame-pair parameter follows:
14229 sequential number of the input frame, starting from 1
14232 Mean Square Error pixel-by-pixel average difference of the compared
14233 frames, averaged over all the image components.
14235 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14236 Mean Square Error pixel-by-pixel average difference of the compared
14237 frames for the component specified by the suffix.
14239 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14240 Peak Signal to Noise ratio of the compared frames for the component
14241 specified by the suffix.
14243 @item max_avg, max_y, max_u, max_v
14244 Maximum allowed value for each channel, and average over all
14250 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14251 [main][ref] psnr="stats_file=stats.log" [out]
14254 On this example the input file being processed is compared with the
14255 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14256 is stored in @file{stats.log}.
14261 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14262 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14265 The pullup filter is designed to take advantage of future context in making
14266 its decisions. This filter is stateless in the sense that it does not lock
14267 onto a pattern to follow, but it instead looks forward to the following
14268 fields in order to identify matches and rebuild progressive frames.
14270 To produce content with an even framerate, insert the fps filter after
14271 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14272 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14274 The filter accepts the following options:
14281 These options set the amount of "junk" to ignore at the left, right, top, and
14282 bottom of the image, respectively. Left and right are in units of 8 pixels,
14283 while top and bottom are in units of 2 lines.
14284 The default is 8 pixels on each side.
14287 Set the strict breaks. Setting this option to 1 will reduce the chances of
14288 filter generating an occasional mismatched frame, but it may also cause an
14289 excessive number of frames to be dropped during high motion sequences.
14290 Conversely, setting it to -1 will make filter match fields more easily.
14291 This may help processing of video where there is slight blurring between
14292 the fields, but may also cause there to be interlaced frames in the output.
14293 Default value is @code{0}.
14296 Set the metric plane to use. It accepts the following values:
14302 Use chroma blue plane.
14305 Use chroma red plane.
14308 This option may be set to use chroma plane instead of the default luma plane
14309 for doing filter's computations. This may improve accuracy on very clean
14310 source material, but more likely will decrease accuracy, especially if there
14311 is chroma noise (rainbow effect) or any grayscale video.
14312 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14313 load and make pullup usable in realtime on slow machines.
14316 For best results (without duplicated frames in the output file) it is
14317 necessary to change the output frame rate. For example, to inverse
14318 telecine NTSC input:
14320 ffmpeg -i input -vf pullup -r 24000/1001 ...
14325 Change video quantization parameters (QP).
14327 The filter accepts the following option:
14331 Set expression for quantization parameter.
14334 The expression is evaluated through the eval API and can contain, among others,
14335 the following constants:
14339 1 if index is not 129, 0 otherwise.
14342 Sequential index starting from -129 to 128.
14345 @subsection Examples
14349 Some equation like:
14357 Flush video frames from internal cache of frames into a random order.
14358 No frame is discarded.
14359 Inspired by @ref{frei0r} nervous filter.
14363 Set size in number of frames of internal cache, in range from @code{2} to
14364 @code{512}. Default is @code{30}.
14367 Set seed for random number generator, must be an integer included between
14368 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14369 less than @code{0}, the filter will try to use a good random seed on a
14373 @section readeia608
14375 Read closed captioning (EIA-608) information from the top lines of a video frame.
14377 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14378 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14379 with EIA-608 data (starting from 0). A description of each metadata value follows:
14382 @item lavfi.readeia608.X.cc
14383 The two bytes stored as EIA-608 data (printed in hexadecimal).
14385 @item lavfi.readeia608.X.line
14386 The number of the line on which the EIA-608 data was identified and read.
14389 This filter accepts the following options:
14393 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14396 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14399 Set minimal acceptable amplitude change for sync codes detection.
14400 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14403 Set the ratio of width reserved for sync code detection.
14404 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14407 Set the max peaks height difference for sync code detection.
14408 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14411 Set max peaks period difference for sync code detection.
14412 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14415 Set the first two max start code bits differences.
14416 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14419 Set the minimum ratio of bits height compared to 3rd start code bit.
14420 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14423 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14426 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14429 Enable checking the parity bit. In the event of a parity error, the filter will output
14430 @code{0x00} for that character. Default is false.
14433 @subsection Examples
14437 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14439 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
14445 Read vertical interval timecode (VITC) information from the top lines of a
14448 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14449 timecode value, if a valid timecode has been detected. Further metadata key
14450 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14451 timecode data has been found or not.
14453 This filter accepts the following options:
14457 Set the maximum number of lines to scan for VITC data. If the value is set to
14458 @code{-1} the full video frame is scanned. Default is @code{45}.
14461 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14462 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14465 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14466 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14469 @subsection Examples
14473 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14474 draw @code{--:--:--:--} as a placeholder:
14476 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14482 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14484 Destination pixel at position (X, Y) will be picked from source (x, y) position
14485 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14486 value for pixel will be used for destination pixel.
14488 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14489 will have Xmap/Ymap video stream dimensions.
14490 Xmap and Ymap input video streams are 16bit depth, single channel.
14492 @section removegrain
14494 The removegrain filter is a spatial denoiser for progressive video.
14498 Set mode for the first plane.
14501 Set mode for the second plane.
14504 Set mode for the third plane.
14507 Set mode for the fourth plane.
14510 Range of mode is from 0 to 24. Description of each mode follows:
14514 Leave input plane unchanged. Default.
14517 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14520 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14523 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14526 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14527 This is equivalent to a median filter.
14530 Line-sensitive clipping giving the minimal change.
14533 Line-sensitive clipping, intermediate.
14536 Line-sensitive clipping, intermediate.
14539 Line-sensitive clipping, intermediate.
14542 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14545 Replaces the target pixel with the closest neighbour.
14548 [1 2 1] horizontal and vertical kernel blur.
14554 Bob mode, interpolates top field from the line where the neighbours
14555 pixels are the closest.
14558 Bob mode, interpolates bottom field from the line where the neighbours
14559 pixels are the closest.
14562 Bob mode, interpolates top field. Same as 13 but with a more complicated
14563 interpolation formula.
14566 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14567 interpolation formula.
14570 Clips the pixel with the minimum and maximum of respectively the maximum and
14571 minimum of each pair of opposite neighbour pixels.
14574 Line-sensitive clipping using opposite neighbours whose greatest distance from
14575 the current pixel is minimal.
14578 Replaces the pixel with the average of its 8 neighbours.
14581 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14584 Clips pixels using the averages of opposite neighbour.
14587 Same as mode 21 but simpler and faster.
14590 Small edge and halo removal, but reputed useless.
14596 @section removelogo
14598 Suppress a TV station logo, using an image file to determine which
14599 pixels comprise the logo. It works by filling in the pixels that
14600 comprise the logo with neighboring pixels.
14602 The filter accepts the following options:
14606 Set the filter bitmap file, which can be any image format supported by
14607 libavformat. The width and height of the image file must match those of the
14608 video stream being processed.
14611 Pixels in the provided bitmap image with a value of zero are not
14612 considered part of the logo, non-zero pixels are considered part of
14613 the logo. If you use white (255) for the logo and black (0) for the
14614 rest, you will be safe. For making the filter bitmap, it is
14615 recommended to take a screen capture of a black frame with the logo
14616 visible, and then using a threshold filter followed by the erode
14617 filter once or twice.
14619 If needed, little splotches can be fixed manually. Remember that if
14620 logo pixels are not covered, the filter quality will be much
14621 reduced. Marking too many pixels as part of the logo does not hurt as
14622 much, but it will increase the amount of blurring needed to cover over
14623 the image and will destroy more information than necessary, and extra
14624 pixels will slow things down on a large logo.
14626 @section repeatfields
14628 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14629 fields based on its value.
14633 Reverse a video clip.
14635 Warning: This filter requires memory to buffer the entire clip, so trimming
14638 @subsection Examples
14642 Take the first 5 seconds of a clip, and reverse it.
14649 Shift R/G/B/A pixels horizontally and/or vertically.
14651 The filter accepts the following options:
14654 Set amount to shift red horizontally.
14656 Set amount to shift red vertically.
14658 Set amount to shift green horizontally.
14660 Set amount to shift green vertically.
14662 Set amount to shift blue horizontally.
14664 Set amount to shift blue vertically.
14666 Set amount to shift alpha horizontally.
14668 Set amount to shift alpha vertically.
14670 Set edge mode, can be @var{smear}, default, or @var{warp}.
14674 Apply roberts cross operator to input video stream.
14676 The filter accepts the following option:
14680 Set which planes will be processed, unprocessed planes will be copied.
14681 By default value 0xf, all planes will be processed.
14684 Set value which will be multiplied with filtered result.
14687 Set value which will be added to filtered result.
14692 Rotate video by an arbitrary angle expressed in radians.
14694 The filter accepts the following options:
14696 A description of the optional parameters follows.
14699 Set an expression for the angle by which to rotate the input video
14700 clockwise, expressed as a number of radians. A negative value will
14701 result in a counter-clockwise rotation. By default it is set to "0".
14703 This expression is evaluated for each frame.
14706 Set the output width expression, default value is "iw".
14707 This expression is evaluated just once during configuration.
14710 Set the output height expression, default value is "ih".
14711 This expression is evaluated just once during configuration.
14714 Enable bilinear interpolation if set to 1, a value of 0 disables
14715 it. Default value is 1.
14718 Set the color used to fill the output area not covered by the rotated
14719 image. For the general syntax of this option, check the
14720 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14721 If the special value "none" is selected then no
14722 background is printed (useful for example if the background is never shown).
14724 Default value is "black".
14727 The expressions for the angle and the output size can contain the
14728 following constants and functions:
14732 sequential number of the input frame, starting from 0. It is always NAN
14733 before the first frame is filtered.
14736 time in seconds of the input frame, it is set to 0 when the filter is
14737 configured. It is always NAN before the first frame is filtered.
14741 horizontal and vertical chroma subsample values. For example for the
14742 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14746 the input video width and height
14750 the output width and height, that is the size of the padded area as
14751 specified by the @var{width} and @var{height} expressions
14755 the minimal width/height required for completely containing the input
14756 video rotated by @var{a} radians.
14758 These are only available when computing the @option{out_w} and
14759 @option{out_h} expressions.
14762 @subsection Examples
14766 Rotate the input by PI/6 radians clockwise:
14772 Rotate the input by PI/6 radians counter-clockwise:
14778 Rotate the input by 45 degrees clockwise:
14784 Apply a constant rotation with period T, starting from an angle of PI/3:
14786 rotate=PI/3+2*PI*t/T
14790 Make the input video rotation oscillating with a period of T
14791 seconds and an amplitude of A radians:
14793 rotate=A*sin(2*PI/T*t)
14797 Rotate the video, output size is chosen so that the whole rotating
14798 input video is always completely contained in the output:
14800 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14804 Rotate the video, reduce the output size so that no background is ever
14807 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14811 @subsection Commands
14813 The filter supports the following commands:
14817 Set the angle expression.
14818 The command accepts the same syntax of the corresponding option.
14820 If the specified expression is not valid, it is kept at its current
14826 Apply Shape Adaptive Blur.
14828 The filter accepts the following options:
14831 @item luma_radius, lr
14832 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14833 value is 1.0. A greater value will result in a more blurred image, and
14834 in slower processing.
14836 @item luma_pre_filter_radius, lpfr
14837 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14840 @item luma_strength, ls
14841 Set luma maximum difference between pixels to still be considered, must
14842 be a value in the 0.1-100.0 range, default value is 1.0.
14844 @item chroma_radius, cr
14845 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14846 greater value will result in a more blurred image, and in slower
14849 @item chroma_pre_filter_radius, cpfr
14850 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14852 @item chroma_strength, cs
14853 Set chroma maximum difference between pixels to still be considered,
14854 must be a value in the -0.9-100.0 range.
14857 Each chroma option value, if not explicitly specified, is set to the
14858 corresponding luma option value.
14863 Scale (resize) the input video, using the libswscale library.
14865 The scale filter forces the output display aspect ratio to be the same
14866 of the input, by changing the output sample aspect ratio.
14868 If the input image format is different from the format requested by
14869 the next filter, the scale filter will convert the input to the
14872 @subsection Options
14873 The filter accepts the following options, or any of the options
14874 supported by the libswscale scaler.
14876 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14877 the complete list of scaler options.
14882 Set the output video dimension expression. Default value is the input
14885 If the @var{width} or @var{w} value is 0, the input width is used for
14886 the output. If the @var{height} or @var{h} value is 0, the input height
14887 is used for the output.
14889 If one and only one of the values is -n with n >= 1, the scale filter
14890 will use a value that maintains the aspect ratio of the input image,
14891 calculated from the other specified dimension. After that it will,
14892 however, make sure that the calculated dimension is divisible by n and
14893 adjust the value if necessary.
14895 If both values are -n with n >= 1, the behavior will be identical to
14896 both values being set to 0 as previously detailed.
14898 See below for the list of accepted constants for use in the dimension
14902 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14906 Only evaluate expressions once during the filter initialization or when a command is processed.
14909 Evaluate expressions for each incoming frame.
14913 Default value is @samp{init}.
14917 Set the interlacing mode. It accepts the following values:
14921 Force interlaced aware scaling.
14924 Do not apply interlaced scaling.
14927 Select interlaced aware scaling depending on whether the source frames
14928 are flagged as interlaced or not.
14931 Default value is @samp{0}.
14934 Set libswscale scaling flags. See
14935 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14936 complete list of values. If not explicitly specified the filter applies
14940 @item param0, param1
14941 Set libswscale input parameters for scaling algorithms that need them. See
14942 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14943 complete documentation. If not explicitly specified the filter applies
14949 Set the video size. For the syntax of this option, check the
14950 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14952 @item in_color_matrix
14953 @item out_color_matrix
14954 Set in/output YCbCr color space type.
14956 This allows the autodetected value to be overridden as well as allows forcing
14957 a specific value used for the output and encoder.
14959 If not specified, the color space type depends on the pixel format.
14965 Choose automatically.
14968 Format conforming to International Telecommunication Union (ITU)
14969 Recommendation BT.709.
14972 Set color space conforming to the United States Federal Communications
14973 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14976 Set color space conforming to:
14980 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14983 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14986 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14991 Set color space conforming to SMPTE ST 240:1999.
14996 Set in/output YCbCr sample range.
14998 This allows the autodetected value to be overridden as well as allows forcing
14999 a specific value used for the output and encoder. If not specified, the
15000 range depends on the pixel format. Possible values:
15004 Choose automatically.
15007 Set full range (0-255 in case of 8-bit luma).
15009 @item mpeg/limited/tv
15010 Set "MPEG" range (16-235 in case of 8-bit luma).
15013 @item force_original_aspect_ratio
15014 Enable decreasing or increasing output video width or height if necessary to
15015 keep the original aspect ratio. Possible values:
15019 Scale the video as specified and disable this feature.
15022 The output video dimensions will automatically be decreased if needed.
15025 The output video dimensions will automatically be increased if needed.
15029 One useful instance of this option is that when you know a specific device's
15030 maximum allowed resolution, you can use this to limit the output video to
15031 that, while retaining the aspect ratio. For example, device A allows
15032 1280x720 playback, and your video is 1920x800. Using this option (set it to
15033 decrease) and specifying 1280x720 to the command line makes the output
15036 Please note that this is a different thing than specifying -1 for @option{w}
15037 or @option{h}, you still need to specify the output resolution for this option
15042 The values of the @option{w} and @option{h} options are expressions
15043 containing the following constants:
15048 The input width and height
15052 These are the same as @var{in_w} and @var{in_h}.
15056 The output (scaled) width and height
15060 These are the same as @var{out_w} and @var{out_h}
15063 The same as @var{iw} / @var{ih}
15066 input sample aspect ratio
15069 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
15073 horizontal and vertical input chroma subsample values. For example for the
15074 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15078 horizontal and vertical output chroma subsample values. For example for the
15079 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15082 @subsection Examples
15086 Scale the input video to a size of 200x100
15091 This is equivalent to:
15102 Specify a size abbreviation for the output size:
15107 which can also be written as:
15113 Scale the input to 2x:
15115 scale=w=2*iw:h=2*ih
15119 The above is the same as:
15121 scale=2*in_w:2*in_h
15125 Scale the input to 2x with forced interlaced scaling:
15127 scale=2*iw:2*ih:interl=1
15131 Scale the input to half size:
15133 scale=w=iw/2:h=ih/2
15137 Increase the width, and set the height to the same size:
15143 Seek Greek harmony:
15150 Increase the height, and set the width to 3/2 of the height:
15152 scale=w=3/2*oh:h=3/5*ih
15156 Increase the size, making the size a multiple of the chroma
15159 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
15163 Increase the width to a maximum of 500 pixels,
15164 keeping the same aspect ratio as the input:
15166 scale=w='min(500\, iw*3/2):h=-1'
15170 Make pixels square by combining scale and setsar:
15172 scale='trunc(ih*dar):ih',setsar=1/1
15176 Make pixels square by combining scale and setsar,
15177 making sure the resulting resolution is even (required by some codecs):
15179 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
15183 @subsection Commands
15185 This filter supports the following commands:
15189 Set the output video dimension expression.
15190 The command accepts the same syntax of the corresponding option.
15192 If the specified expression is not valid, it is kept at its current
15198 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15199 format conversion on CUDA video frames. Setting the output width and height
15200 works in the same way as for the @var{scale} filter.
15202 The following additional options are accepted:
15205 The pixel format of the output CUDA frames. If set to the string "same" (the
15206 default), the input format will be kept. Note that automatic format negotiation
15207 and conversion is not yet supported for hardware frames
15210 The interpolation algorithm used for resizing. One of the following:
15217 @item cubic2p_bspline
15218 2-parameter cubic (B=1, C=0)
15220 @item cubic2p_catmullrom
15221 2-parameter cubic (B=0, C=1/2)
15223 @item cubic2p_b05c03
15224 2-parameter cubic (B=1/2, C=3/10)
15236 Scale (resize) the input video, based on a reference video.
15238 See the scale filter for available options, scale2ref supports the same but
15239 uses the reference video instead of the main input as basis. scale2ref also
15240 supports the following additional constants for the @option{w} and
15241 @option{h} options:
15246 The main input video's width and height
15249 The same as @var{main_w} / @var{main_h}
15252 The main input video's sample aspect ratio
15254 @item main_dar, mdar
15255 The main input video's display aspect ratio. Calculated from
15256 @code{(main_w / main_h) * main_sar}.
15260 The main input video's horizontal and vertical chroma subsample values.
15261 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15265 @subsection Examples
15269 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15271 'scale2ref[b][a];[a][b]overlay'
15275 @anchor{selectivecolor}
15276 @section selectivecolor
15278 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15279 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15280 by the "purity" of the color (that is, how saturated it already is).
15282 This filter is similar to the Adobe Photoshop Selective Color tool.
15284 The filter accepts the following options:
15287 @item correction_method
15288 Select color correction method.
15290 Available values are:
15293 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15296 Specified adjustments are relative to the original component value.
15298 Default is @code{absolute}.
15300 Adjustments for red pixels (pixels where the red component is the maximum)
15302 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15304 Adjustments for green pixels (pixels where the green component is the maximum)
15306 Adjustments for cyan pixels (pixels where the red component is the minimum)
15308 Adjustments for blue pixels (pixels where the blue component is the maximum)
15310 Adjustments for magenta pixels (pixels where the green component is the minimum)
15312 Adjustments for white pixels (pixels where all components are greater than 128)
15314 Adjustments for all pixels except pure black and pure white
15316 Adjustments for black pixels (pixels where all components are lesser than 128)
15318 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15321 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15322 4 space separated floating point adjustment values in the [-1,1] range,
15323 respectively to adjust the amount of cyan, magenta, yellow and black for the
15324 pixels of its range.
15326 @subsection Examples
15330 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15331 increase magenta by 27% in blue areas:
15333 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15337 Use a Photoshop selective color preset:
15339 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15343 @anchor{separatefields}
15344 @section separatefields
15346 The @code{separatefields} takes a frame-based video input and splits
15347 each frame into its components fields, producing a new half height clip
15348 with twice the frame rate and twice the frame count.
15350 This filter use field-dominance information in frame to decide which
15351 of each pair of fields to place first in the output.
15352 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15354 @section setdar, setsar
15356 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15359 This is done by changing the specified Sample (aka Pixel) Aspect
15360 Ratio, according to the following equation:
15362 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15365 Keep in mind that the @code{setdar} filter does not modify the pixel
15366 dimensions of the video frame. Also, the display aspect ratio set by
15367 this filter may be changed by later filters in the filterchain,
15368 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15371 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15372 the filter output video.
15374 Note that as a consequence of the application of this filter, the
15375 output display aspect ratio will change according to the equation
15378 Keep in mind that the sample aspect ratio set by the @code{setsar}
15379 filter may be changed by later filters in the filterchain, e.g. if
15380 another "setsar" or a "setdar" filter is applied.
15382 It accepts the following parameters:
15385 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15386 Set the aspect ratio used by the filter.
15388 The parameter can be a floating point number string, an expression, or
15389 a string of the form @var{num}:@var{den}, where @var{num} and
15390 @var{den} are the numerator and denominator of the aspect ratio. If
15391 the parameter is not specified, it is assumed the value "0".
15392 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15396 Set the maximum integer value to use for expressing numerator and
15397 denominator when reducing the expressed aspect ratio to a rational.
15398 Default value is @code{100}.
15402 The parameter @var{sar} is an expression containing
15403 the following constants:
15407 These are approximated values for the mathematical constants e
15408 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15411 The input width and height.
15414 These are the same as @var{w} / @var{h}.
15417 The input sample aspect ratio.
15420 The input display aspect ratio. It is the same as
15421 (@var{w} / @var{h}) * @var{sar}.
15424 Horizontal and vertical chroma subsample values. For example, for the
15425 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15428 @subsection Examples
15433 To change the display aspect ratio to 16:9, specify one of the following:
15440 To change the sample aspect ratio to 10:11, specify:
15446 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15447 1000 in the aspect ratio reduction, use the command:
15449 setdar=ratio=16/9:max=1000
15457 Force field for the output video frame.
15459 The @code{setfield} filter marks the interlace type field for the
15460 output frames. It does not change the input frame, but only sets the
15461 corresponding property, which affects how the frame is treated by
15462 following filters (e.g. @code{fieldorder} or @code{yadif}).
15464 The filter accepts the following options:
15469 Available values are:
15473 Keep the same field property.
15476 Mark the frame as bottom-field-first.
15479 Mark the frame as top-field-first.
15482 Mark the frame as progressive.
15489 Force frame parameter for the output video frame.
15491 The @code{setparams} filter marks interlace and color range for the
15492 output frames. It does not change the input frame, but only sets the
15493 corresponding property, which affects how the frame is treated by
15498 Available values are:
15502 Keep the same field property (default).
15505 Mark the frame as bottom-field-first.
15508 Mark the frame as top-field-first.
15511 Mark the frame as progressive.
15515 Available values are:
15519 Keep the same color range property (default).
15521 @item unspecified, unknown
15522 Mark the frame as unspecified color range.
15524 @item limited, tv, mpeg
15525 Mark the frame as limited range.
15527 @item full, pc, jpeg
15528 Mark the frame as full range.
15531 @item color_primaries
15532 Set the color primaries.
15533 Available values are:
15537 Keep the same color primaries property (default).
15554 Set the color transfer.
15555 Available values are:
15559 Keep the same color trc property (default).
15581 Set the colorspace.
15582 Available values are:
15586 Keep the same colorspace property (default).
15599 @item chroma-derived-nc
15600 @item chroma-derived-c
15607 Show a line containing various information for each input video frame.
15608 The input video is not modified.
15610 This filter supports the following options:
15614 Calculate checksums of each plane. By default enabled.
15617 The shown line contains a sequence of key/value pairs of the form
15618 @var{key}:@var{value}.
15620 The following values are shown in the output:
15624 The (sequential) number of the input frame, starting from 0.
15627 The Presentation TimeStamp of the input frame, expressed as a number of
15628 time base units. The time base unit depends on the filter input pad.
15631 The Presentation TimeStamp of the input frame, expressed as a number of
15635 The position of the frame in the input stream, or -1 if this information is
15636 unavailable and/or meaningless (for example in case of synthetic video).
15639 The pixel format name.
15642 The sample aspect ratio of the input frame, expressed in the form
15643 @var{num}/@var{den}.
15646 The size of the input frame. For the syntax of this option, check the
15647 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15650 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15651 for bottom field first).
15654 This is 1 if the frame is a key frame, 0 otherwise.
15657 The picture type of the input frame ("I" for an I-frame, "P" for a
15658 P-frame, "B" for a B-frame, or "?" for an unknown type).
15659 Also refer to the documentation of the @code{AVPictureType} enum and of
15660 the @code{av_get_picture_type_char} function defined in
15661 @file{libavutil/avutil.h}.
15664 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15666 @item plane_checksum
15667 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15668 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15671 @section showpalette
15673 Displays the 256 colors palette of each frame. This filter is only relevant for
15674 @var{pal8} pixel format frames.
15676 It accepts the following option:
15680 Set the size of the box used to represent one palette color entry. Default is
15681 @code{30} (for a @code{30x30} pixel box).
15684 @section shuffleframes
15686 Reorder and/or duplicate and/or drop video frames.
15688 It accepts the following parameters:
15692 Set the destination indexes of input frames.
15693 This is space or '|' separated list of indexes that maps input frames to output
15694 frames. Number of indexes also sets maximal value that each index may have.
15695 '-1' index have special meaning and that is to drop frame.
15698 The first frame has the index 0. The default is to keep the input unchanged.
15700 @subsection Examples
15704 Swap second and third frame of every three frames of the input:
15706 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15710 Swap 10th and 1st frame of every ten frames of the input:
15712 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15716 @section shuffleplanes
15718 Reorder and/or duplicate video planes.
15720 It accepts the following parameters:
15725 The index of the input plane to be used as the first output plane.
15728 The index of the input plane to be used as the second output plane.
15731 The index of the input plane to be used as the third output plane.
15734 The index of the input plane to be used as the fourth output plane.
15738 The first plane has the index 0. The default is to keep the input unchanged.
15740 @subsection Examples
15744 Swap the second and third planes of the input:
15746 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15750 @anchor{signalstats}
15751 @section signalstats
15752 Evaluate various visual metrics that assist in determining issues associated
15753 with the digitization of analog video media.
15755 By default the filter will log these metadata values:
15759 Display the minimal Y value contained within the input frame. Expressed in
15763 Display the Y value at the 10% percentile within the input frame. Expressed in
15767 Display the average Y value within the input frame. Expressed in range of
15771 Display the Y value at the 90% percentile within the input frame. Expressed in
15775 Display the maximum Y value contained within the input frame. Expressed in
15779 Display the minimal U value contained within the input frame. Expressed in
15783 Display the U value at the 10% percentile within the input frame. Expressed in
15787 Display the average U value within the input frame. Expressed in range of
15791 Display the U value at the 90% percentile within the input frame. Expressed in
15795 Display the maximum U value contained within the input frame. Expressed in
15799 Display the minimal V value contained within the input frame. Expressed in
15803 Display the V value at the 10% percentile within the input frame. Expressed in
15807 Display the average V value within the input frame. Expressed in range of
15811 Display the V value at the 90% percentile within the input frame. Expressed in
15815 Display the maximum V value contained within the input frame. Expressed in
15819 Display the minimal saturation value contained within the input frame.
15820 Expressed in range of [0-~181.02].
15823 Display the saturation value at the 10% percentile within the input frame.
15824 Expressed in range of [0-~181.02].
15827 Display the average saturation value within the input frame. Expressed in range
15831 Display the saturation value at the 90% percentile within the input frame.
15832 Expressed in range of [0-~181.02].
15835 Display the maximum saturation value contained within the input frame.
15836 Expressed in range of [0-~181.02].
15839 Display the median value for hue within the input frame. Expressed in range of
15843 Display the average value for hue within the input frame. Expressed in range of
15847 Display the average of sample value difference between all values of the Y
15848 plane in the current frame and corresponding values of the previous input frame.
15849 Expressed in range of [0-255].
15852 Display the average of sample value difference between all values of the U
15853 plane in the current frame and corresponding values of the previous input frame.
15854 Expressed in range of [0-255].
15857 Display the average of sample value difference between all values of the V
15858 plane in the current frame and corresponding values of the previous input frame.
15859 Expressed in range of [0-255].
15862 Display bit depth of Y plane in current frame.
15863 Expressed in range of [0-16].
15866 Display bit depth of U plane in current frame.
15867 Expressed in range of [0-16].
15870 Display bit depth of V plane in current frame.
15871 Expressed in range of [0-16].
15874 The filter accepts the following options:
15880 @option{stat} specify an additional form of image analysis.
15881 @option{out} output video with the specified type of pixel highlighted.
15883 Both options accept the following values:
15887 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15888 unlike the neighboring pixels of the same field. Examples of temporal outliers
15889 include the results of video dropouts, head clogs, or tape tracking issues.
15892 Identify @var{vertical line repetition}. Vertical line repetition includes
15893 similar rows of pixels within a frame. In born-digital video vertical line
15894 repetition is common, but this pattern is uncommon in video digitized from an
15895 analog source. When it occurs in video that results from the digitization of an
15896 analog source it can indicate concealment from a dropout compensator.
15899 Identify pixels that fall outside of legal broadcast range.
15903 Set the highlight color for the @option{out} option. The default color is
15907 @subsection Examples
15911 Output data of various video metrics:
15913 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15917 Output specific data about the minimum and maximum values of the Y plane per frame:
15919 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15923 Playback video while highlighting pixels that are outside of broadcast range in red.
15925 ffplay example.mov -vf signalstats="out=brng:color=red"
15929 Playback video with signalstats metadata drawn over the frame.
15931 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15934 The contents of signalstat_drawtext.txt used in the command are:
15937 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15938 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15939 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15940 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15948 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15949 input. In this case the matching between the inputs can be calculated additionally.
15950 The filter always passes through the first input. The signature of each stream can
15951 be written into a file.
15953 It accepts the following options:
15957 Enable or disable the matching process.
15959 Available values are:
15963 Disable the calculation of a matching (default).
15965 Calculate the matching for the whole video and output whether the whole video
15966 matches or only parts.
15968 Calculate only until a matching is found or the video ends. Should be faster in
15973 Set the number of inputs. The option value must be a non negative integer.
15974 Default value is 1.
15977 Set the path to which the output is written. If there is more than one input,
15978 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15979 integer), that will be replaced with the input number. If no filename is
15980 specified, no output will be written. This is the default.
15983 Choose the output format.
15985 Available values are:
15989 Use the specified binary representation (default).
15991 Use the specified xml representation.
15995 Set threshold to detect one word as similar. The option value must be an integer
15996 greater than zero. The default value is 9000.
15999 Set threshold to detect all words as similar. The option value must be an integer
16000 greater than zero. The default value is 60000.
16003 Set threshold to detect frames as similar. The option value must be an integer
16004 greater than zero. The default value is 116.
16007 Set the minimum length of a sequence in frames to recognize it as matching
16008 sequence. The option value must be a non negative integer value.
16009 The default value is 0.
16012 Set the minimum relation, that matching frames to all frames must have.
16013 The option value must be a double value between 0 and 1. The default value is 0.5.
16016 @subsection Examples
16020 To calculate the signature of an input video and store it in signature.bin:
16022 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
16026 To detect whether two videos match and store the signatures in XML format in
16027 signature0.xml and signature1.xml:
16029 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 -
16037 Blur the input video without impacting the outlines.
16039 It accepts the following options:
16042 @item luma_radius, lr
16043 Set the luma radius. The option value must be a float number in
16044 the range [0.1,5.0] that specifies the variance of the gaussian filter
16045 used to blur the image (slower if larger). Default value is 1.0.
16047 @item luma_strength, ls
16048 Set the luma strength. The option value must be a float number
16049 in the range [-1.0,1.0] that configures the blurring. A value included
16050 in [0.0,1.0] will blur the image whereas a value included in
16051 [-1.0,0.0] will sharpen the image. Default value is 1.0.
16053 @item luma_threshold, lt
16054 Set the luma threshold used as a coefficient to determine
16055 whether a pixel should be blurred or not. The option value must be an
16056 integer in the range [-30,30]. A value of 0 will filter all the image,
16057 a value included in [0,30] will filter flat areas and a value included
16058 in [-30,0] will filter edges. Default value is 0.
16060 @item chroma_radius, cr
16061 Set the chroma radius. The option value must be a float number in
16062 the range [0.1,5.0] that specifies the variance of the gaussian filter
16063 used to blur the image (slower if larger). Default value is @option{luma_radius}.
16065 @item chroma_strength, cs
16066 Set the chroma strength. The option value must be a float number
16067 in the range [-1.0,1.0] that configures the blurring. A value included
16068 in [0.0,1.0] will blur the image whereas a value included in
16069 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
16071 @item chroma_threshold, ct
16072 Set the chroma threshold used as a coefficient to determine
16073 whether a pixel should be blurred or not. The option value must be an
16074 integer in the range [-30,30]. A value of 0 will filter all the image,
16075 a value included in [0,30] will filter flat areas and a value included
16076 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
16079 If a chroma option is not explicitly set, the corresponding luma value
16084 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
16086 This filter takes in input two input videos, the first input is
16087 considered the "main" source and is passed unchanged to the
16088 output. The second input is used as a "reference" video for computing
16091 Both video inputs must have the same resolution and pixel format for
16092 this filter to work correctly. Also it assumes that both inputs
16093 have the same number of frames, which are compared one by one.
16095 The filter stores the calculated SSIM of each frame.
16097 The description of the accepted parameters follows.
16100 @item stats_file, f
16101 If specified the filter will use the named file to save the SSIM of
16102 each individual frame. When filename equals "-" the data is sent to
16106 The file printed if @var{stats_file} is selected, contains a sequence of
16107 key/value pairs of the form @var{key}:@var{value} for each compared
16110 A description of each shown parameter follows:
16114 sequential number of the input frame, starting from 1
16116 @item Y, U, V, R, G, B
16117 SSIM of the compared frames for the component specified by the suffix.
16120 SSIM of the compared frames for the whole frame.
16123 Same as above but in dB representation.
16126 This filter also supports the @ref{framesync} options.
16130 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16131 [main][ref] ssim="stats_file=stats.log" [out]
16134 On this example the input file being processed is compared with the
16135 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
16136 is stored in @file{stats.log}.
16138 Another example with both psnr and ssim at same time:
16140 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
16145 Convert between different stereoscopic image formats.
16147 The filters accept the following options:
16151 Set stereoscopic image format of input.
16153 Available values for input image formats are:
16156 side by side parallel (left eye left, right eye right)
16159 side by side crosseye (right eye left, left eye right)
16162 side by side parallel with half width resolution
16163 (left eye left, right eye right)
16166 side by side crosseye with half width resolution
16167 (right eye left, left eye right)
16170 above-below (left eye above, right eye below)
16173 above-below (right eye above, left eye below)
16176 above-below with half height resolution
16177 (left eye above, right eye below)
16180 above-below with half height resolution
16181 (right eye above, left eye below)
16184 alternating frames (left eye first, right eye second)
16187 alternating frames (right eye first, left eye second)
16190 interleaved rows (left eye has top row, right eye starts on next row)
16193 interleaved rows (right eye has top row, left eye starts on next row)
16196 interleaved columns, left eye first
16199 interleaved columns, right eye first
16201 Default value is @samp{sbsl}.
16205 Set stereoscopic image format of output.
16209 side by side parallel (left eye left, right eye right)
16212 side by side crosseye (right eye left, left eye right)
16215 side by side parallel with half width resolution
16216 (left eye left, right eye right)
16219 side by side crosseye with half width resolution
16220 (right eye left, left eye right)
16223 above-below (left eye above, right eye below)
16226 above-below (right eye above, left eye below)
16229 above-below with half height resolution
16230 (left eye above, right eye below)
16233 above-below with half height resolution
16234 (right eye above, left eye below)
16237 alternating frames (left eye first, right eye second)
16240 alternating frames (right eye first, left eye second)
16243 interleaved rows (left eye has top row, right eye starts on next row)
16246 interleaved rows (right eye has top row, left eye starts on next row)
16249 anaglyph red/blue gray
16250 (red filter on left eye, blue filter on right eye)
16253 anaglyph red/green gray
16254 (red filter on left eye, green filter on right eye)
16257 anaglyph red/cyan gray
16258 (red filter on left eye, cyan filter on right eye)
16261 anaglyph red/cyan half colored
16262 (red filter on left eye, cyan filter on right eye)
16265 anaglyph red/cyan color
16266 (red filter on left eye, cyan filter on right eye)
16269 anaglyph red/cyan color optimized with the least squares projection of dubois
16270 (red filter on left eye, cyan filter on right eye)
16273 anaglyph green/magenta gray
16274 (green filter on left eye, magenta filter on right eye)
16277 anaglyph green/magenta half colored
16278 (green filter on left eye, magenta filter on right eye)
16281 anaglyph green/magenta colored
16282 (green filter on left eye, magenta filter on right eye)
16285 anaglyph green/magenta color optimized with the least squares projection of dubois
16286 (green filter on left eye, magenta filter on right eye)
16289 anaglyph yellow/blue gray
16290 (yellow filter on left eye, blue filter on right eye)
16293 anaglyph yellow/blue half colored
16294 (yellow filter on left eye, blue filter on right eye)
16297 anaglyph yellow/blue colored
16298 (yellow filter on left eye, blue filter on right eye)
16301 anaglyph yellow/blue color optimized with the least squares projection of dubois
16302 (yellow filter on left eye, blue filter on right eye)
16305 mono output (left eye only)
16308 mono output (right eye only)
16311 checkerboard, left eye first
16314 checkerboard, right eye first
16317 interleaved columns, left eye first
16320 interleaved columns, right eye first
16326 Default value is @samp{arcd}.
16329 @subsection Examples
16333 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16339 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16345 @section streamselect, astreamselect
16346 Select video or audio streams.
16348 The filter accepts the following options:
16352 Set number of inputs. Default is 2.
16355 Set input indexes to remap to outputs.
16358 @subsection Commands
16360 The @code{streamselect} and @code{astreamselect} filter supports the following
16365 Set input indexes to remap to outputs.
16368 @subsection Examples
16372 Select first 5 seconds 1st stream and rest of time 2nd stream:
16374 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16378 Same as above, but for audio:
16380 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16385 Apply sobel operator to input video stream.
16387 The filter accepts the following option:
16391 Set which planes will be processed, unprocessed planes will be copied.
16392 By default value 0xf, all planes will be processed.
16395 Set value which will be multiplied with filtered result.
16398 Set value which will be added to filtered result.
16404 Apply a simple postprocessing filter that compresses and decompresses the image
16405 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16406 and average the results.
16408 The filter accepts the following options:
16412 Set quality. This option defines the number of levels for averaging. It accepts
16413 an integer in the range 0-6. If set to @code{0}, the filter will have no
16414 effect. A value of @code{6} means the higher quality. For each increment of
16415 that value the speed drops by a factor of approximately 2. Default value is
16419 Force a constant quantization parameter. If not set, the filter will use the QP
16420 from the video stream (if available).
16423 Set thresholding mode. Available modes are:
16427 Set hard thresholding (default).
16429 Set soft thresholding (better de-ringing effect, but likely blurrier).
16432 @item use_bframe_qp
16433 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16434 option may cause flicker since the B-Frames have often larger QP. Default is
16435 @code{0} (not enabled).
16440 Scale the input by applying one of the super-resolution methods based on
16441 convolutional neural networks. Supported models:
16445 Super-Resolution Convolutional Neural Network model (SRCNN).
16446 See @url{https://arxiv.org/abs/1501.00092}.
16449 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16450 See @url{https://arxiv.org/abs/1609.05158}.
16453 Training scripts as well as scripts for model generation are provided in
16454 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16456 The filter accepts the following options:
16460 Specify which DNN backend to use for model loading and execution. This option accepts
16461 the following values:
16465 Native implementation of DNN loading and execution.
16468 TensorFlow backend. To enable this backend you
16469 need to install the TensorFlow for C library (see
16470 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16471 @code{--enable-libtensorflow}
16474 Default value is @samp{native}.
16477 Set path to model file specifying network architecture and its parameters.
16478 Note that different backends use different file formats. TensorFlow backend
16479 can load files for both formats, while native backend can load files for only
16483 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16484 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16485 input upscaled using bicubic upscaling with proper scale factor.
16491 Draw subtitles on top of input video using the libass library.
16493 To enable compilation of this filter you need to configure FFmpeg with
16494 @code{--enable-libass}. This filter also requires a build with libavcodec and
16495 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16496 Alpha) subtitles format.
16498 The filter accepts the following options:
16502 Set the filename of the subtitle file to read. It must be specified.
16504 @item original_size
16505 Specify the size of the original video, the video for which the ASS file
16506 was composed. For the syntax of this option, check the
16507 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16508 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16509 correctly scale the fonts if the aspect ratio has been changed.
16512 Set a directory path containing fonts that can be used by the filter.
16513 These fonts will be used in addition to whatever the font provider uses.
16516 Process alpha channel, by default alpha channel is untouched.
16519 Set subtitles input character encoding. @code{subtitles} filter only. Only
16520 useful if not UTF-8.
16522 @item stream_index, si
16523 Set subtitles stream index. @code{subtitles} filter only.
16526 Override default style or script info parameters of the subtitles. It accepts a
16527 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16530 If the first key is not specified, it is assumed that the first value
16531 specifies the @option{filename}.
16533 For example, to render the file @file{sub.srt} on top of the input
16534 video, use the command:
16539 which is equivalent to:
16541 subtitles=filename=sub.srt
16544 To render the default subtitles stream from file @file{video.mkv}, use:
16546 subtitles=video.mkv
16549 To render the second subtitles stream from that file, use:
16551 subtitles=video.mkv:si=1
16554 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16555 @code{DejaVu Serif}, use:
16557 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16560 @section super2xsai
16562 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16563 Interpolate) pixel art scaling algorithm.
16565 Useful for enlarging pixel art images without reducing sharpness.
16569 Swap two rectangular objects in video.
16571 This filter accepts the following options:
16581 Set 1st rect x coordinate.
16584 Set 1st rect y coordinate.
16587 Set 2nd rect x coordinate.
16590 Set 2nd rect y coordinate.
16592 All expressions are evaluated once for each frame.
16595 The all options are expressions containing the following constants:
16600 The input width and height.
16603 same as @var{w} / @var{h}
16606 input sample aspect ratio
16609 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16612 The number of the input frame, starting from 0.
16615 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16618 the position in the file of the input frame, NAN if unknown
16626 Apply telecine process to the video.
16628 This filter accepts the following options:
16637 The default value is @code{top}.
16641 A string of numbers representing the pulldown pattern you wish to apply.
16642 The default value is @code{23}.
16646 Some typical patterns:
16651 24p: 2332 (preferred)
16658 24p: 222222222223 ("Euro pulldown")
16665 Apply threshold effect to video stream.
16667 This filter needs four video streams to perform thresholding.
16668 First stream is stream we are filtering.
16669 Second stream is holding threshold values, third stream is holding min values,
16670 and last, fourth stream is holding max values.
16672 The filter accepts the following option:
16676 Set which planes will be processed, unprocessed planes will be copied.
16677 By default value 0xf, all planes will be processed.
16680 For example if first stream pixel's component value is less then threshold value
16681 of pixel component from 2nd threshold stream, third stream value will picked,
16682 otherwise fourth stream pixel component value will be picked.
16684 Using color source filter one can perform various types of thresholding:
16686 @subsection Examples
16690 Binary threshold, using gray color as threshold:
16692 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16696 Inverted binary threshold, using gray color as threshold:
16698 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16702 Truncate binary threshold, using gray color as threshold:
16704 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16708 Threshold to zero, using gray color as threshold:
16710 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16714 Inverted threshold to zero, using gray color as threshold:
16716 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16721 Select the most representative frame in a given sequence of consecutive frames.
16723 The filter accepts the following options:
16727 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16728 will pick one of them, and then handle the next batch of @var{n} frames until
16729 the end. Default is @code{100}.
16732 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16733 value will result in a higher memory usage, so a high value is not recommended.
16735 @subsection Examples
16739 Extract one picture each 50 frames:
16745 Complete example of a thumbnail creation with @command{ffmpeg}:
16747 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16753 Tile several successive frames together.
16755 The filter accepts the following options:
16760 Set the grid size (i.e. the number of lines and columns). For the syntax of
16761 this option, check the
16762 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16765 Set the maximum number of frames to render in the given area. It must be less
16766 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16767 the area will be used.
16770 Set the outer border margin in pixels.
16773 Set the inner border thickness (i.e. the number of pixels between frames). For
16774 more advanced padding options (such as having different values for the edges),
16775 refer to the pad video filter.
16778 Specify the color of the unused area. For the syntax of this option, check the
16779 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16780 The default value of @var{color} is "black".
16783 Set the number of frames to overlap when tiling several successive frames together.
16784 The value must be between @code{0} and @var{nb_frames - 1}.
16787 Set the number of frames to initially be empty before displaying first output frame.
16788 This controls how soon will one get first output frame.
16789 The value must be between @code{0} and @var{nb_frames - 1}.
16792 @subsection Examples
16796 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16798 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16800 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16801 duplicating each output frame to accommodate the originally detected frame
16805 Display @code{5} pictures in an area of @code{3x2} frames,
16806 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16807 mixed flat and named options:
16809 tile=3x2:nb_frames=5:padding=7:margin=2
16813 @section tinterlace
16815 Perform various types of temporal field interlacing.
16817 Frames are counted starting from 1, so the first input frame is
16820 The filter accepts the following options:
16825 Specify the mode of the interlacing. This option can also be specified
16826 as a value alone. See below for a list of values for this option.
16828 Available values are:
16832 Move odd frames into the upper field, even into the lower field,
16833 generating a double height frame at half frame rate.
16837 Frame 1 Frame 2 Frame 3 Frame 4
16839 11111 22222 33333 44444
16840 11111 22222 33333 44444
16841 11111 22222 33333 44444
16842 11111 22222 33333 44444
16856 Only output odd frames, even frames are dropped, generating a frame with
16857 unchanged height at half frame rate.
16862 Frame 1 Frame 2 Frame 3 Frame 4
16864 11111 22222 33333 44444
16865 11111 22222 33333 44444
16866 11111 22222 33333 44444
16867 11111 22222 33333 44444
16877 Only output even frames, odd frames are dropped, generating a frame with
16878 unchanged height at half frame rate.
16883 Frame 1 Frame 2 Frame 3 Frame 4
16885 11111 22222 33333 44444
16886 11111 22222 33333 44444
16887 11111 22222 33333 44444
16888 11111 22222 33333 44444
16898 Expand each frame to full height, but pad alternate lines with black,
16899 generating a frame with double height at the same input frame rate.
16904 Frame 1 Frame 2 Frame 3 Frame 4
16906 11111 22222 33333 44444
16907 11111 22222 33333 44444
16908 11111 22222 33333 44444
16909 11111 22222 33333 44444
16912 11111 ..... 33333 .....
16913 ..... 22222 ..... 44444
16914 11111 ..... 33333 .....
16915 ..... 22222 ..... 44444
16916 11111 ..... 33333 .....
16917 ..... 22222 ..... 44444
16918 11111 ..... 33333 .....
16919 ..... 22222 ..... 44444
16923 @item interleave_top, 4
16924 Interleave the upper field from odd frames with the lower field from
16925 even frames, generating a frame with unchanged height at half frame rate.
16930 Frame 1 Frame 2 Frame 3 Frame 4
16932 11111<- 22222 33333<- 44444
16933 11111 22222<- 33333 44444<-
16934 11111<- 22222 33333<- 44444
16935 11111 22222<- 33333 44444<-
16945 @item interleave_bottom, 5
16946 Interleave the lower field from odd frames with the upper field from
16947 even frames, generating a frame with unchanged height at half frame rate.
16952 Frame 1 Frame 2 Frame 3 Frame 4
16954 11111 22222<- 33333 44444<-
16955 11111<- 22222 33333<- 44444
16956 11111 22222<- 33333 44444<-
16957 11111<- 22222 33333<- 44444
16967 @item interlacex2, 6
16968 Double frame rate with unchanged height. Frames are inserted each
16969 containing the second temporal field from the previous input frame and
16970 the first temporal field from the next input frame. This mode relies on
16971 the top_field_first flag. Useful for interlaced video displays with no
16972 field synchronisation.
16977 Frame 1 Frame 2 Frame 3 Frame 4
16979 11111 22222 33333 44444
16980 11111 22222 33333 44444
16981 11111 22222 33333 44444
16982 11111 22222 33333 44444
16985 11111 22222 22222 33333 33333 44444 44444
16986 11111 11111 22222 22222 33333 33333 44444
16987 11111 22222 22222 33333 33333 44444 44444
16988 11111 11111 22222 22222 33333 33333 44444
16993 Move odd frames into the upper field, even into the lower field,
16994 generating a double height frame at same frame rate.
16999 Frame 1 Frame 2 Frame 3 Frame 4
17001 11111 22222 33333 44444
17002 11111 22222 33333 44444
17003 11111 22222 33333 44444
17004 11111 22222 33333 44444
17007 11111 33333 33333 55555
17008 22222 22222 44444 44444
17009 11111 33333 33333 55555
17010 22222 22222 44444 44444
17011 11111 33333 33333 55555
17012 22222 22222 44444 44444
17013 11111 33333 33333 55555
17014 22222 22222 44444 44444
17019 Numeric values are deprecated but are accepted for backward
17020 compatibility reasons.
17022 Default mode is @code{merge}.
17025 Specify flags influencing the filter process.
17027 Available value for @var{flags} is:
17030 @item low_pass_filter, vlfp
17031 Enable linear vertical low-pass filtering in the filter.
17032 Vertical low-pass filtering is required when creating an interlaced
17033 destination from a progressive source which contains high-frequency
17034 vertical detail. Filtering will reduce interlace 'twitter' and Moire
17037 @item complex_filter, cvlfp
17038 Enable complex vertical low-pass filtering.
17039 This will slightly less reduce interlace 'twitter' and Moire
17040 patterning but better retain detail and subjective sharpness impression.
17044 Vertical low-pass filtering can only be enabled for @option{mode}
17045 @var{interleave_top} and @var{interleave_bottom}.
17051 Mix successive video frames.
17053 A description of the accepted options follows.
17057 The number of successive frames to mix. If unspecified, it defaults to 3.
17060 Specify weight of each input video frame.
17061 Each weight is separated by space. If number of weights is smaller than
17062 number of @var{frames} last specified weight will be used for all remaining
17066 Specify scale, if it is set it will be multiplied with sum
17067 of each weight multiplied with pixel values to give final destination
17068 pixel value. By default @var{scale} is auto scaled to sum of weights.
17071 @subsection Examples
17075 Average 7 successive frames:
17077 tmix=frames=7:weights="1 1 1 1 1 1 1"
17081 Apply simple temporal convolution:
17083 tmix=frames=3:weights="-1 3 -1"
17087 Similar as above but only showing temporal differences:
17089 tmix=frames=3:weights="-1 2 -1":scale=1
17095 Tone map colors from different dynamic ranges.
17097 This filter expects data in single precision floating point, as it needs to
17098 operate on (and can output) out-of-range values. Another filter, such as
17099 @ref{zscale}, is needed to convert the resulting frame to a usable format.
17101 The tonemapping algorithms implemented only work on linear light, so input
17102 data should be linearized beforehand (and possibly correctly tagged).
17105 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
17108 @subsection Options
17109 The filter accepts the following options.
17113 Set the tone map algorithm to use.
17115 Possible values are:
17118 Do not apply any tone map, only desaturate overbright pixels.
17121 Hard-clip any out-of-range values. Use it for perfect color accuracy for
17122 in-range values, while distorting out-of-range values.
17125 Stretch the entire reference gamut to a linear multiple of the display.
17128 Fit a logarithmic transfer between the tone curves.
17131 Preserve overall image brightness with a simple curve, using nonlinear
17132 contrast, which results in flattening details and degrading color accuracy.
17135 Preserve both dark and bright details better than @var{reinhard}, at the cost
17136 of slightly darkening everything. Use it when detail preservation is more
17137 important than color and brightness accuracy.
17140 Smoothly map out-of-range values, while retaining contrast and colors for
17141 in-range material as much as possible. Use it when color accuracy is more
17142 important than detail preservation.
17148 Tune the tone mapping algorithm.
17150 This affects the following algorithms:
17156 Specifies the scale factor to use while stretching.
17160 Specifies the exponent of the function.
17164 Specify an extra linear coefficient to multiply into the signal before clipping.
17168 Specify the local contrast coefficient at the display peak.
17169 Default to 0.5, which means that in-gamut values will be about half as bright
17176 Specify the transition point from linear to mobius transform. Every value
17177 below this point is guaranteed to be mapped 1:1. The higher the value, the
17178 more accurate the result will be, at the cost of losing bright details.
17179 Default to 0.3, which due to the steep initial slope still preserves in-range
17180 colors fairly accurately.
17184 Apply desaturation for highlights that exceed this level of brightness. The
17185 higher the parameter, the more color information will be preserved. This
17186 setting helps prevent unnaturally blown-out colors for super-highlights, by
17187 (smoothly) turning into white instead. This makes images feel more natural,
17188 at the cost of reducing information about out-of-range colors.
17190 The default of 2.0 is somewhat conservative and will mostly just apply to
17191 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17193 This option works only if the input frame has a supported color tag.
17196 Override signal/nominal/reference peak with this value. Useful when the
17197 embedded peak information in display metadata is not reliable or when tone
17198 mapping from a lower range to a higher range.
17203 Temporarily pad video frames.
17205 The filter accepts the following options:
17209 Specify number of delay frames before input video stream.
17212 Specify number of padding frames after input video stream.
17213 Set to -1 to pad indefinitely.
17216 Set kind of frames added to beginning of stream.
17217 Can be either @var{add} or @var{clone}.
17218 With @var{add} frames of solid-color are added.
17219 With @var{clone} frames are clones of first frame.
17222 Set kind of frames added to end of stream.
17223 Can be either @var{add} or @var{clone}.
17224 With @var{add} frames of solid-color are added.
17225 With @var{clone} frames are clones of last frame.
17227 @item start_duration, stop_duration
17228 Specify the duration of the start/stop delay. See
17229 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17230 for the accepted syntax.
17231 These options override @var{start} and @var{stop}.
17234 Specify the color of the padded area. For the syntax of this option,
17235 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17236 manual,ffmpeg-utils}.
17238 The default value of @var{color} is "black".
17244 Transpose rows with columns in the input video and optionally flip it.
17246 It accepts the following parameters:
17251 Specify the transposition direction.
17253 Can assume the following values:
17255 @item 0, 4, cclock_flip
17256 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17264 Rotate by 90 degrees clockwise, that is:
17272 Rotate by 90 degrees counterclockwise, that is:
17279 @item 3, 7, clock_flip
17280 Rotate by 90 degrees clockwise and vertically flip, that is:
17288 For values between 4-7, the transposition is only done if the input
17289 video geometry is portrait and not landscape. These values are
17290 deprecated, the @code{passthrough} option should be used instead.
17292 Numerical values are deprecated, and should be dropped in favor of
17293 symbolic constants.
17296 Do not apply the transposition if the input geometry matches the one
17297 specified by the specified value. It accepts the following values:
17300 Always apply transposition.
17302 Preserve portrait geometry (when @var{height} >= @var{width}).
17304 Preserve landscape geometry (when @var{width} >= @var{height}).
17307 Default value is @code{none}.
17310 For example to rotate by 90 degrees clockwise and preserve portrait
17313 transpose=dir=1:passthrough=portrait
17316 The command above can also be specified as:
17318 transpose=1:portrait
17321 @section transpose_npp
17323 Transpose rows with columns in the input video and optionally flip it.
17324 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17326 It accepts the following parameters:
17331 Specify the transposition direction.
17333 Can assume the following values:
17336 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17339 Rotate by 90 degrees clockwise.
17342 Rotate by 90 degrees counterclockwise.
17345 Rotate by 90 degrees clockwise and vertically flip.
17349 Do not apply the transposition if the input geometry matches the one
17350 specified by the specified value. It accepts the following values:
17353 Always apply transposition. (default)
17355 Preserve portrait geometry (when @var{height} >= @var{width}).
17357 Preserve landscape geometry (when @var{width} >= @var{height}).
17363 Trim the input so that the output contains one continuous subpart of the input.
17365 It accepts the following parameters:
17368 Specify the time of the start of the kept section, i.e. the frame with the
17369 timestamp @var{start} will be the first frame in the output.
17372 Specify the time of the first frame that will be dropped, i.e. the frame
17373 immediately preceding the one with the timestamp @var{end} will be the last
17374 frame in the output.
17377 This is the same as @var{start}, except this option sets the start timestamp
17378 in timebase units instead of seconds.
17381 This is the same as @var{end}, except this option sets the end timestamp
17382 in timebase units instead of seconds.
17385 The maximum duration of the output in seconds.
17388 The number of the first frame that should be passed to the output.
17391 The number of the first frame that should be dropped.
17394 @option{start}, @option{end}, and @option{duration} are expressed as time
17395 duration specifications; see
17396 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17397 for the accepted syntax.
17399 Note that the first two sets of the start/end options and the @option{duration}
17400 option look at the frame timestamp, while the _frame variants simply count the
17401 frames that pass through the filter. Also note that this filter does not modify
17402 the timestamps. If you wish for the output timestamps to start at zero, insert a
17403 setpts filter after the trim filter.
17405 If multiple start or end options are set, this filter tries to be greedy and
17406 keep all the frames that match at least one of the specified constraints. To keep
17407 only the part that matches all the constraints at once, chain multiple trim
17410 The defaults are such that all the input is kept. So it is possible to set e.g.
17411 just the end values to keep everything before the specified time.
17416 Drop everything except the second minute of input:
17418 ffmpeg -i INPUT -vf trim=60:120
17422 Keep only the first second:
17424 ffmpeg -i INPUT -vf trim=duration=1
17429 @section unpremultiply
17430 Apply alpha unpremultiply effect to input video stream using first plane
17431 of second stream as alpha.
17433 Both streams must have same dimensions and same pixel format.
17435 The filter accepts the following option:
17439 Set which planes will be processed, unprocessed planes will be copied.
17440 By default value 0xf, all planes will be processed.
17442 If the format has 1 or 2 components, then luma is bit 0.
17443 If the format has 3 or 4 components:
17444 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17445 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17446 If present, the alpha channel is always the last bit.
17449 Do not require 2nd input for processing, instead use alpha plane from input stream.
17455 Sharpen or blur the input video.
17457 It accepts the following parameters:
17460 @item luma_msize_x, lx
17461 Set the luma matrix horizontal size. It must be an odd integer between
17462 3 and 23. The default value is 5.
17464 @item luma_msize_y, ly
17465 Set the luma matrix vertical size. It must be an odd integer between 3
17466 and 23. The default value is 5.
17468 @item luma_amount, la
17469 Set the luma effect strength. It must be a floating point number, reasonable
17470 values lay between -1.5 and 1.5.
17472 Negative values will blur the input video, while positive values will
17473 sharpen it, a value of zero will disable the effect.
17475 Default value is 1.0.
17477 @item chroma_msize_x, cx
17478 Set the chroma matrix horizontal size. It must be an odd integer
17479 between 3 and 23. The default value is 5.
17481 @item chroma_msize_y, cy
17482 Set the chroma matrix vertical size. It must be an odd integer
17483 between 3 and 23. The default value is 5.
17485 @item chroma_amount, ca
17486 Set the chroma effect strength. It must be a floating point number, reasonable
17487 values lay between -1.5 and 1.5.
17489 Negative values will blur the input video, while positive values will
17490 sharpen it, a value of zero will disable the effect.
17492 Default value is 0.0.
17496 All parameters are optional and default to the equivalent of the
17497 string '5:5:1.0:5:5:0.0'.
17499 @subsection Examples
17503 Apply strong luma sharpen effect:
17505 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17509 Apply a strong blur of both luma and chroma parameters:
17511 unsharp=7:7:-2:7:7:-2
17517 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17518 the image at several (or - in the case of @option{quality} level @code{8} - all)
17519 shifts and average the results.
17521 The way this differs from the behavior of spp is that uspp actually encodes &
17522 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17523 DCT similar to MJPEG.
17525 The filter accepts the following options:
17529 Set quality. This option defines the number of levels for averaging. It accepts
17530 an integer in the range 0-8. If set to @code{0}, the filter will have no
17531 effect. A value of @code{8} means the higher quality. For each increment of
17532 that value the speed drops by a factor of approximately 2. Default value is
17536 Force a constant quantization parameter. If not set, the filter will use the QP
17537 from the video stream (if available).
17540 @section vaguedenoiser
17542 Apply a wavelet based denoiser.
17544 It transforms each frame from the video input into the wavelet domain,
17545 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17546 the obtained coefficients. It does an inverse wavelet transform after.
17547 Due to wavelet properties, it should give a nice smoothed result, and
17548 reduced noise, without blurring picture features.
17550 This filter accepts the following options:
17554 The filtering strength. The higher, the more filtered the video will be.
17555 Hard thresholding can use a higher threshold than soft thresholding
17556 before the video looks overfiltered. Default value is 2.
17559 The filtering method the filter will use.
17561 It accepts the following values:
17564 All values under the threshold will be zeroed.
17567 All values under the threshold will be zeroed. All values above will be
17568 reduced by the threshold.
17571 Scales or nullifies coefficients - intermediary between (more) soft and
17572 (less) hard thresholding.
17575 Default is garrote.
17578 Number of times, the wavelet will decompose the picture. Picture can't
17579 be decomposed beyond a particular point (typically, 8 for a 640x480
17580 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17583 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17586 A list of the planes to process. By default all planes are processed.
17589 @section vectorscope
17591 Display 2 color component values in the two dimensional graph (which is called
17594 This filter accepts the following options:
17598 Set vectorscope mode.
17600 It accepts the following values:
17603 Gray values are displayed on graph, higher brightness means more pixels have
17604 same component color value on location in graph. This is the default mode.
17607 Gray values are displayed on graph. Surrounding pixels values which are not
17608 present in video frame are drawn in gradient of 2 color components which are
17609 set by option @code{x} and @code{y}. The 3rd color component is static.
17612 Actual color components values present in video frame are displayed on graph.
17615 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17616 on graph increases value of another color component, which is luminance by
17617 default values of @code{x} and @code{y}.
17620 Actual colors present in video frame are displayed on graph. If two different
17621 colors map to same position on graph then color with higher value of component
17622 not present in graph is picked.
17625 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17626 component picked from radial gradient.
17630 Set which color component will be represented on X-axis. Default is @code{1}.
17633 Set which color component will be represented on Y-axis. Default is @code{2}.
17636 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17637 of color component which represents frequency of (X, Y) location in graph.
17642 No envelope, this is default.
17645 Instant envelope, even darkest single pixel will be clearly highlighted.
17648 Hold maximum and minimum values presented in graph over time. This way you
17649 can still spot out of range values without constantly looking at vectorscope.
17652 Peak and instant envelope combined together.
17656 Set what kind of graticule to draw.
17664 Set graticule opacity.
17667 Set graticule flags.
17671 Draw graticule for white point.
17674 Draw graticule for black point.
17677 Draw color points short names.
17681 Set background opacity.
17683 @item lthreshold, l
17684 Set low threshold for color component not represented on X or Y axis.
17685 Values lower than this value will be ignored. Default is 0.
17686 Note this value is multiplied with actual max possible value one pixel component
17687 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17690 @item hthreshold, h
17691 Set high threshold for color component not represented on X or Y axis.
17692 Values higher than this value will be ignored. Default is 1.
17693 Note this value is multiplied with actual max possible value one pixel component
17694 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17695 is 0.9 * 255 = 230.
17697 @item colorspace, c
17698 Set what kind of colorspace to use when drawing graticule.
17707 @anchor{vidstabdetect}
17708 @section vidstabdetect
17710 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17711 @ref{vidstabtransform} for pass 2.
17713 This filter generates a file with relative translation and rotation
17714 transform information about subsequent frames, which is then used by
17715 the @ref{vidstabtransform} filter.
17717 To enable compilation of this filter you need to configure FFmpeg with
17718 @code{--enable-libvidstab}.
17720 This filter accepts the following options:
17724 Set the path to the file used to write the transforms information.
17725 Default value is @file{transforms.trf}.
17728 Set how shaky the video is and how quick the camera is. It accepts an
17729 integer in the range 1-10, a value of 1 means little shakiness, a
17730 value of 10 means strong shakiness. Default value is 5.
17733 Set the accuracy of the detection process. It must be a value in the
17734 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17735 accuracy. Default value is 15.
17738 Set stepsize of the search process. The region around minimum is
17739 scanned with 1 pixel resolution. Default value is 6.
17742 Set minimum contrast. Below this value a local measurement field is
17743 discarded. Must be a floating point value in the range 0-1. Default
17747 Set reference frame number for tripod mode.
17749 If enabled, the motion of the frames is compared to a reference frame
17750 in the filtered stream, identified by the specified number. The idea
17751 is to compensate all movements in a more-or-less static scene and keep
17752 the camera view absolutely still.
17754 If set to 0, it is disabled. The frames are counted starting from 1.
17757 Show fields and transforms in the resulting frames. It accepts an
17758 integer in the range 0-2. Default value is 0, which disables any
17762 @subsection Examples
17766 Use default values:
17772 Analyze strongly shaky movie and put the results in file
17773 @file{mytransforms.trf}:
17775 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17779 Visualize the result of internal transformations in the resulting
17782 vidstabdetect=show=1
17786 Analyze a video with medium shakiness using @command{ffmpeg}:
17788 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17792 @anchor{vidstabtransform}
17793 @section vidstabtransform
17795 Video stabilization/deshaking: pass 2 of 2,
17796 see @ref{vidstabdetect} for pass 1.
17798 Read a file with transform information for each frame and
17799 apply/compensate them. Together with the @ref{vidstabdetect}
17800 filter this can be used to deshake videos. See also
17801 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17802 the @ref{unsharp} filter, see below.
17804 To enable compilation of this filter you need to configure FFmpeg with
17805 @code{--enable-libvidstab}.
17807 @subsection Options
17811 Set path to the file used to read the transforms. Default value is
17812 @file{transforms.trf}.
17815 Set the number of frames (value*2 + 1) used for lowpass filtering the
17816 camera movements. Default value is 10.
17818 For example a number of 10 means that 21 frames are used (10 in the
17819 past and 10 in the future) to smoothen the motion in the video. A
17820 larger value leads to a smoother video, but limits the acceleration of
17821 the camera (pan/tilt movements). 0 is a special case where a static
17822 camera is simulated.
17825 Set the camera path optimization algorithm.
17827 Accepted values are:
17830 gaussian kernel low-pass filter on camera motion (default)
17832 averaging on transformations
17836 Set maximal number of pixels to translate frames. Default value is -1,
17840 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17841 value is -1, meaning no limit.
17844 Specify how to deal with borders that may be visible due to movement
17847 Available values are:
17850 keep image information from previous frame (default)
17852 fill the border black
17856 Invert transforms if set to 1. Default value is 0.
17859 Consider transforms as relative to previous frame if set to 1,
17860 absolute if set to 0. Default value is 0.
17863 Set percentage to zoom. A positive value will result in a zoom-in
17864 effect, a negative value in a zoom-out effect. Default value is 0 (no
17868 Set optimal zooming to avoid borders.
17870 Accepted values are:
17875 optimal static zoom value is determined (only very strong movements
17876 will lead to visible borders) (default)
17878 optimal adaptive zoom value is determined (no borders will be
17879 visible), see @option{zoomspeed}
17882 Note that the value given at zoom is added to the one calculated here.
17885 Set percent to zoom maximally each frame (enabled when
17886 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17890 Specify type of interpolation.
17892 Available values are:
17897 linear only horizontal
17899 linear in both directions (default)
17901 cubic in both directions (slow)
17905 Enable virtual tripod mode if set to 1, which is equivalent to
17906 @code{relative=0:smoothing=0}. Default value is 0.
17908 Use also @code{tripod} option of @ref{vidstabdetect}.
17911 Increase log verbosity if set to 1. Also the detected global motions
17912 are written to the temporary file @file{global_motions.trf}. Default
17916 @subsection Examples
17920 Use @command{ffmpeg} for a typical stabilization with default values:
17922 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17925 Note the use of the @ref{unsharp} filter which is always recommended.
17928 Zoom in a bit more and load transform data from a given file:
17930 vidstabtransform=zoom=5:input="mytransforms.trf"
17934 Smoothen the video even more:
17936 vidstabtransform=smoothing=30
17942 Flip the input video vertically.
17944 For example, to vertically flip a video with @command{ffmpeg}:
17946 ffmpeg -i in.avi -vf "vflip" out.avi
17951 Detect variable frame rate video.
17953 This filter tries to detect if the input is variable or constant frame rate.
17955 At end it will output number of frames detected as having variable delta pts,
17956 and ones with constant delta pts.
17957 If there was frames with variable delta, than it will also show min and max delta
17962 Boost or alter saturation.
17964 The filter accepts the following options:
17967 Set strength of boost if positive value or strength of alter if negative value.
17968 Default is 0. Allowed range is from -2 to 2.
17971 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17974 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17977 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17980 Set the red luma coefficient.
17983 Set the green luma coefficient.
17986 Set the blue luma coefficient.
17992 Make or reverse a natural vignetting effect.
17994 The filter accepts the following options:
17998 Set lens angle expression as a number of radians.
18000 The value is clipped in the @code{[0,PI/2]} range.
18002 Default value: @code{"PI/5"}
18006 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
18010 Set forward/backward mode.
18012 Available modes are:
18015 The larger the distance from the central point, the darker the image becomes.
18018 The larger the distance from the central point, the brighter the image becomes.
18019 This can be used to reverse a vignette effect, though there is no automatic
18020 detection to extract the lens @option{angle} and other settings (yet). It can
18021 also be used to create a burning effect.
18024 Default value is @samp{forward}.
18027 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
18029 It accepts the following values:
18032 Evaluate expressions only once during the filter initialization.
18035 Evaluate expressions for each incoming frame. This is way slower than the
18036 @samp{init} mode since it requires all the scalers to be re-computed, but it
18037 allows advanced dynamic expressions.
18040 Default value is @samp{init}.
18043 Set dithering to reduce the circular banding effects. Default is @code{1}
18047 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
18048 Setting this value to the SAR of the input will make a rectangular vignetting
18049 following the dimensions of the video.
18051 Default is @code{1/1}.
18054 @subsection Expressions
18056 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
18057 following parameters.
18062 input width and height
18065 the number of input frame, starting from 0
18068 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
18069 @var{TB} units, NAN if undefined
18072 frame rate of the input video, NAN if the input frame rate is unknown
18075 the PTS (Presentation TimeStamp) of the filtered video frame,
18076 expressed in seconds, NAN if undefined
18079 time base of the input video
18083 @subsection Examples
18087 Apply simple strong vignetting effect:
18093 Make a flickering vignetting:
18095 vignette='PI/4+random(1)*PI/50':eval=frame
18100 @section vmafmotion
18102 Obtain the average vmaf motion score of a video.
18103 It is one of the component filters of VMAF.
18105 The obtained average motion score is printed through the logging system.
18107 In the below example the input file @file{ref.mpg} is being processed and score
18111 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
18115 Stack input videos vertically.
18117 All streams must be of same pixel format and of same width.
18119 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
18120 to create same output.
18122 The filter accept the following option:
18126 Set number of input streams. Default is 2.
18129 If set to 1, force the output to terminate when the shortest input
18130 terminates. Default value is 0.
18135 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
18136 Deinterlacing Filter").
18138 Based on the process described by Martin Weston for BBC R&D, and
18139 implemented based on the de-interlace algorithm written by Jim
18140 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
18141 uses filter coefficients calculated by BBC R&D.
18143 There are two sets of filter coefficients, so called "simple":
18144 and "complex". Which set of filter coefficients is used can
18145 be set by passing an optional parameter:
18149 Set the interlacing filter coefficients. Accepts one of the following values:
18153 Simple filter coefficient set.
18155 More-complex filter coefficient set.
18157 Default value is @samp{complex}.
18160 Specify which frames to deinterlace. Accept one of the following values:
18164 Deinterlace all frames,
18166 Only deinterlace frames marked as interlaced.
18169 Default value is @samp{all}.
18173 Video waveform monitor.
18175 The waveform monitor plots color component intensity. By default luminance
18176 only. Each column of the waveform corresponds to a column of pixels in the
18179 It accepts the following options:
18183 Can be either @code{row}, or @code{column}. Default is @code{column}.
18184 In row mode, the graph on the left side represents color component value 0 and
18185 the right side represents value = 255. In column mode, the top side represents
18186 color component value = 0 and bottom side represents value = 255.
18189 Set intensity. Smaller values are useful to find out how many values of the same
18190 luminance are distributed across input rows/columns.
18191 Default value is @code{0.04}. Allowed range is [0, 1].
18194 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18195 In mirrored mode, higher values will be represented on the left
18196 side for @code{row} mode and at the top for @code{column} mode. Default is
18197 @code{1} (mirrored).
18201 It accepts the following values:
18204 Presents information identical to that in the @code{parade}, except
18205 that the graphs representing color components are superimposed directly
18208 This display mode makes it easier to spot relative differences or similarities
18209 in overlapping areas of the color components that are supposed to be identical,
18210 such as neutral whites, grays, or blacks.
18213 Display separate graph for the color components side by side in
18214 @code{row} mode or one below the other in @code{column} mode.
18217 Display separate graph for the color components side by side in
18218 @code{column} mode or one below the other in @code{row} mode.
18220 Using this display mode makes it easy to spot color casts in the highlights
18221 and shadows of an image, by comparing the contours of the top and the bottom
18222 graphs of each waveform. Since whites, grays, and blacks are characterized
18223 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18224 should display three waveforms of roughly equal width/height. If not, the
18225 correction is easy to perform by making level adjustments the three waveforms.
18227 Default is @code{stack}.
18229 @item components, c
18230 Set which color components to display. Default is 1, which means only luminance
18231 or red color component if input is in RGB colorspace. If is set for example to
18232 7 it will display all 3 (if) available color components.
18237 No envelope, this is default.
18240 Instant envelope, minimum and maximum values presented in graph will be easily
18241 visible even with small @code{step} value.
18244 Hold minimum and maximum values presented in graph across time. This way you
18245 can still spot out of range values without constantly looking at waveforms.
18248 Peak and instant envelope combined together.
18254 No filtering, this is default.
18257 Luma and chroma combined together.
18260 Similar as above, but shows difference between blue and red chroma.
18263 Similar as above, but use different colors.
18266 Displays only chroma.
18269 Displays actual color value on waveform.
18272 Similar as above, but with luma showing frequency of chroma values.
18276 Set which graticule to display.
18280 Do not display graticule.
18283 Display green graticule showing legal broadcast ranges.
18286 Display orange graticule showing legal broadcast ranges.
18290 Set graticule opacity.
18293 Set graticule flags.
18297 Draw numbers above lines. By default enabled.
18300 Draw dots instead of lines.
18304 Set scale used for displaying graticule.
18311 Default is digital.
18314 Set background opacity.
18317 @section weave, doubleweave
18319 The @code{weave} takes a field-based video input and join
18320 each two sequential fields into single frame, producing a new double
18321 height clip with half the frame rate and half the frame count.
18323 The @code{doubleweave} works same as @code{weave} but without
18324 halving frame rate and frame count.
18326 It accepts the following option:
18330 Set first field. Available values are:
18334 Set the frame as top-field-first.
18337 Set the frame as bottom-field-first.
18341 @subsection Examples
18345 Interlace video using @ref{select} and @ref{separatefields} filter:
18347 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18352 Apply the xBR high-quality magnification filter which is designed for pixel
18353 art. It follows a set of edge-detection rules, see
18354 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18356 It accepts the following option:
18360 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18361 @code{3xBR} and @code{4} for @code{4xBR}.
18362 Default is @code{3}.
18366 Stack video inputs into custom layout.
18368 All streams must be of same pixel format.
18370 The filter accept the following option:
18374 Set number of input streams. Default is 2.
18377 Specify layout of inputs.
18378 This option requires the desired layout configuration to be explicitly set by the user.
18379 This sets position of each video input in output. Each input
18380 is separated by '|'.
18381 The first number represents the column, and the second number represents the row.
18382 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18383 where X is video input from which to take width or height.
18384 Multiple values can be used when separated by '+'. In such
18385 case values are summed together.
18388 If set to 1, force the output to terminate when the shortest input
18389 terminates. Default value is 0.
18392 @subsection Examples
18396 Display 4 inputs into 2x2 grid,
18397 note that if inputs are of different sizes unused gaps might appear,
18398 as not all of output video is used.
18400 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18404 Display 4 inputs into 1x4 grid,
18405 note that if inputs are of different sizes unused gaps might appear,
18406 as not all of output video is used.
18408 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18412 Display 9 inputs into 3x3 grid,
18413 note that if inputs are of different sizes unused gaps might appear,
18414 as not all of output video is used.
18416 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
18423 Deinterlace the input video ("yadif" means "yet another deinterlacing
18426 It accepts the following parameters:
18432 The interlacing mode to adopt. It accepts one of the following values:
18435 @item 0, send_frame
18436 Output one frame for each frame.
18437 @item 1, send_field
18438 Output one frame for each field.
18439 @item 2, send_frame_nospatial
18440 Like @code{send_frame}, but it skips the spatial interlacing check.
18441 @item 3, send_field_nospatial
18442 Like @code{send_field}, but it skips the spatial interlacing check.
18445 The default value is @code{send_frame}.
18448 The picture field parity assumed for the input interlaced video. It accepts one
18449 of the following values:
18453 Assume the top field is first.
18455 Assume the bottom field is first.
18457 Enable automatic detection of field parity.
18460 The default value is @code{auto}.
18461 If the interlacing is unknown or the decoder does not export this information,
18462 top field first will be assumed.
18465 Specify which frames to deinterlace. Accept one of the following
18470 Deinterlace all frames.
18471 @item 1, interlaced
18472 Only deinterlace frames marked as interlaced.
18475 The default value is @code{all}.
18478 @section yadif_cuda
18480 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18481 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18484 It accepts the following parameters:
18490 The interlacing mode to adopt. It accepts one of the following values:
18493 @item 0, send_frame
18494 Output one frame for each frame.
18495 @item 1, send_field
18496 Output one frame for each field.
18497 @item 2, send_frame_nospatial
18498 Like @code{send_frame}, but it skips the spatial interlacing check.
18499 @item 3, send_field_nospatial
18500 Like @code{send_field}, but it skips the spatial interlacing check.
18503 The default value is @code{send_frame}.
18506 The picture field parity assumed for the input interlaced video. It accepts one
18507 of the following values:
18511 Assume the top field is first.
18513 Assume the bottom field is first.
18515 Enable automatic detection of field parity.
18518 The default value is @code{auto}.
18519 If the interlacing is unknown or the decoder does not export this information,
18520 top field first will be assumed.
18523 Specify which frames to deinterlace. Accept one of the following
18528 Deinterlace all frames.
18529 @item 1, interlaced
18530 Only deinterlace frames marked as interlaced.
18533 The default value is @code{all}.
18538 Apply Zoom & Pan effect.
18540 This filter accepts the following options:
18544 Set the zoom expression. Range is 1-10. Default is 1.
18548 Set the x and y expression. Default is 0.
18551 Set the duration expression in number of frames.
18552 This sets for how many number of frames effect will last for
18553 single input image.
18556 Set the output image size, default is 'hd720'.
18559 Set the output frame rate, default is '25'.
18562 Each expression can contain the following constants:
18581 Output frame count.
18585 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18586 for current input frame.
18590 'x' and 'y' of last output frame of previous input frame or 0 when there was
18591 not yet such frame (first input frame).
18594 Last calculated zoom from 'z' expression for current input frame.
18597 Last calculated zoom of last output frame of previous input frame.
18600 Number of output frames for current input frame. Calculated from 'd' expression
18601 for each input frame.
18604 number of output frames created for previous input frame
18607 Rational number: input width / input height
18610 sample aspect ratio
18613 display aspect ratio
18617 @subsection Examples
18621 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18623 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
18627 Zoom-in up to 1.5 and pan always at center of picture:
18629 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18633 Same as above but without pausing:
18635 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18641 Scale (resize) the input video, using the z.lib library:
18642 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18643 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18645 The zscale filter forces the output display aspect ratio to be the same
18646 as the input, by changing the output sample aspect ratio.
18648 If the input image format is different from the format requested by
18649 the next filter, the zscale filter will convert the input to the
18652 @subsection Options
18653 The filter accepts the following options.
18658 Set the output video dimension expression. Default value is the input
18661 If the @var{width} or @var{w} value is 0, the input width is used for
18662 the output. If the @var{height} or @var{h} value is 0, the input height
18663 is used for the output.
18665 If one and only one of the values is -n with n >= 1, the zscale filter
18666 will use a value that maintains the aspect ratio of the input image,
18667 calculated from the other specified dimension. After that it will,
18668 however, make sure that the calculated dimension is divisible by n and
18669 adjust the value if necessary.
18671 If both values are -n with n >= 1, the behavior will be identical to
18672 both values being set to 0 as previously detailed.
18674 See below for the list of accepted constants for use in the dimension
18678 Set the video size. For the syntax of this option, check the
18679 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18682 Set the dither type.
18684 Possible values are:
18689 @item error_diffusion
18695 Set the resize filter type.
18697 Possible values are:
18707 Default is bilinear.
18710 Set the color range.
18712 Possible values are:
18719 Default is same as input.
18722 Set the color primaries.
18724 Possible values are:
18734 Default is same as input.
18737 Set the transfer characteristics.
18739 Possible values are:
18753 Default is same as input.
18756 Set the colorspace matrix.
18758 Possible value are:
18769 Default is same as input.
18772 Set the input color range.
18774 Possible values are:
18781 Default is same as input.
18783 @item primariesin, pin
18784 Set the input color primaries.
18786 Possible values are:
18796 Default is same as input.
18798 @item transferin, tin
18799 Set the input transfer characteristics.
18801 Possible values are:
18812 Default is same as input.
18814 @item matrixin, min
18815 Set the input colorspace matrix.
18817 Possible value are:
18829 Set the output chroma location.
18831 Possible values are:
18842 @item chromalin, cin
18843 Set the input chroma location.
18845 Possible values are:
18857 Set the nominal peak luminance.
18860 The values of the @option{w} and @option{h} options are expressions
18861 containing the following constants:
18866 The input width and height
18870 These are the same as @var{in_w} and @var{in_h}.
18874 The output (scaled) width and height
18878 These are the same as @var{out_w} and @var{out_h}
18881 The same as @var{iw} / @var{ih}
18884 input sample aspect ratio
18887 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18891 horizontal and vertical input chroma subsample values. For example for the
18892 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18896 horizontal and vertical output chroma subsample values. For example for the
18897 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18903 @c man end VIDEO FILTERS
18905 @chapter OpenCL Video Filters
18906 @c man begin OPENCL VIDEO FILTERS
18908 Below is a description of the currently available OpenCL video filters.
18910 To enable compilation of these filters you need to configure FFmpeg with
18911 @code{--enable-opencl}.
18913 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18916 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18917 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18918 given device parameters.
18920 @item -filter_hw_device @var{name}
18921 Pass the hardware device called @var{name} to all filters in any filter graph.
18925 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18929 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18931 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18935 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.
18937 @section avgblur_opencl
18939 Apply average blur filter.
18941 The filter accepts the following options:
18945 Set horizontal radius size.
18946 Range is @code{[1, 1024]} and default value is @code{1}.
18949 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18952 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18955 @subsection Example
18959 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.
18961 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18965 @section boxblur_opencl
18967 Apply a boxblur algorithm to the input video.
18969 It accepts the following parameters:
18973 @item luma_radius, lr
18974 @item luma_power, lp
18975 @item chroma_radius, cr
18976 @item chroma_power, cp
18977 @item alpha_radius, ar
18978 @item alpha_power, ap
18982 A description of the accepted options follows.
18985 @item luma_radius, lr
18986 @item chroma_radius, cr
18987 @item alpha_radius, ar
18988 Set an expression for the box radius in pixels used for blurring the
18989 corresponding input plane.
18991 The radius value must be a non-negative number, and must not be
18992 greater than the value of the expression @code{min(w,h)/2} for the
18993 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18996 Default value for @option{luma_radius} is "2". If not specified,
18997 @option{chroma_radius} and @option{alpha_radius} default to the
18998 corresponding value set for @option{luma_radius}.
19000 The expressions can contain the following constants:
19004 The input width and height in pixels.
19008 The input chroma image width and height in pixels.
19012 The horizontal and vertical chroma subsample values. For example, for the
19013 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
19016 @item luma_power, lp
19017 @item chroma_power, cp
19018 @item alpha_power, ap
19019 Specify how many times the boxblur filter is applied to the
19020 corresponding plane.
19022 Default value for @option{luma_power} is 2. If not specified,
19023 @option{chroma_power} and @option{alpha_power} default to the
19024 corresponding value set for @option{luma_power}.
19026 A value of 0 will disable the effect.
19029 @subsection Examples
19031 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.
19035 Apply a boxblur filter with the luma, chroma, and alpha radius
19036 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.
19038 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
19039 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
19043 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.
19045 For the luma plane, a 2x2 box radius will be run once.
19047 For the chroma plane, a 4x4 box radius will be run 5 times.
19049 For the alpha plane, a 3x3 box radius will be run 7 times.
19051 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
19055 @section convolution_opencl
19057 Apply convolution of 3x3, 5x5, 7x7 matrix.
19059 The filter accepts the following options:
19066 Set matrix for each plane.
19067 Matrix is sequence of 9, 25 or 49 signed numbers.
19068 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
19074 Set multiplier for calculated value for each plane.
19075 If unset or 0, it will be sum of all matrix elements.
19076 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
19082 Set bias for each plane. This value is added to the result of the multiplication.
19083 Useful for making the overall image brighter or darker.
19084 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
19088 @subsection Examples
19094 -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
19100 -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
19104 Apply edge enhance:
19106 -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
19112 -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
19116 Apply laplacian edge detector which includes diagonals:
19118 -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
19124 -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
19128 @section dilation_opencl
19130 Apply dilation effect to the video.
19132 This filter replaces the pixel by the local(3x3) maximum.
19134 It accepts the following options:
19141 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19142 If @code{0}, plane will remain unchanged.
19145 Flag which specifies the pixel to refer to.
19146 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19148 Flags to local 3x3 coordinates region centered on @code{x}:
19157 @subsection Example
19161 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.
19163 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19167 @section erosion_opencl
19169 Apply erosion effect to the video.
19171 This filter replaces the pixel by the local(3x3) minimum.
19173 It accepts the following options:
19180 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19181 If @code{0}, plane will remain unchanged.
19184 Flag which specifies the pixel to refer to.
19185 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19187 Flags to local 3x3 coordinates region centered on @code{x}:
19196 @subsection Example
19200 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.
19202 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19206 @section colorkey_opencl
19207 RGB colorspace color keying.
19209 The filter accepts the following options:
19213 The color which will be replaced with transparency.
19216 Similarity percentage with the key color.
19218 0.01 matches only the exact key color, while 1.0 matches everything.
19223 0.0 makes pixels either fully transparent, or not transparent at all.
19225 Higher values result in semi-transparent pixels, with a higher transparency
19226 the more similar the pixels color is to the key color.
19229 @subsection Examples
19233 Make every semi-green pixel in the input transparent with some slight blending:
19235 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
19239 @section overlay_opencl
19241 Overlay one video on top of another.
19243 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19244 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19246 The filter accepts the following options:
19251 Set the x coordinate of the overlaid video on the main video.
19252 Default value is @code{0}.
19255 Set the x coordinate of the overlaid video on the main video.
19256 Default value is @code{0}.
19260 @subsection Examples
19264 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19266 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19269 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19271 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19276 @section prewitt_opencl
19278 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19280 The filter accepts the following option:
19284 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19287 Set value which will be multiplied with filtered result.
19288 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19291 Set value which will be added to filtered result.
19292 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19295 @subsection Example
19299 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19301 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19305 @section roberts_opencl
19306 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19308 The filter accepts the following option:
19312 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19315 Set value which will be multiplied with filtered result.
19316 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19319 Set value which will be added to filtered result.
19320 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19323 @subsection Example
19327 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19329 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19333 @section sobel_opencl
19335 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19337 The filter accepts the following option:
19341 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19344 Set value which will be multiplied with filtered result.
19345 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19348 Set value which will be added to filtered result.
19349 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19352 @subsection Example
19356 Apply sobel operator with scale set to 2 and delta set to 10
19358 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19362 @section tonemap_opencl
19364 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19366 It accepts the following parameters:
19370 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19373 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19376 Apply desaturation for highlights that exceed this level of brightness. The
19377 higher the parameter, the more color information will be preserved. This
19378 setting helps prevent unnaturally blown-out colors for super-highlights, by
19379 (smoothly) turning into white instead. This makes images feel more natural,
19380 at the cost of reducing information about out-of-range colors.
19382 The default value is 0.5, and the algorithm here is a little different from
19383 the cpu version tonemap currently. A setting of 0.0 disables this option.
19386 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19387 is used to detect whether the scene has changed or not. If the distance between
19388 the current frame average brightness and the current running average exceeds
19389 a threshold value, we would re-calculate scene average and peak brightness.
19390 The default value is 0.2.
19393 Specify the output pixel format.
19395 Currently supported formats are:
19402 Set the output color range.
19404 Possible values are:
19410 Default is same as input.
19413 Set the output color primaries.
19415 Possible values are:
19421 Default is same as input.
19424 Set the output transfer characteristics.
19426 Possible values are:
19435 Set the output colorspace matrix.
19437 Possible value are:
19443 Default is same as input.
19447 @subsection Example
19451 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19453 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19457 @section unsharp_opencl
19459 Sharpen or blur the input video.
19461 It accepts the following parameters:
19464 @item luma_msize_x, lx
19465 Set the luma matrix horizontal size.
19466 Range is @code{[1, 23]} and default value is @code{5}.
19468 @item luma_msize_y, ly
19469 Set the luma matrix vertical size.
19470 Range is @code{[1, 23]} and default value is @code{5}.
19472 @item luma_amount, la
19473 Set the luma effect strength.
19474 Range is @code{[-10, 10]} and default value is @code{1.0}.
19476 Negative values will blur the input video, while positive values will
19477 sharpen it, a value of zero will disable the effect.
19479 @item chroma_msize_x, cx
19480 Set the chroma matrix horizontal size.
19481 Range is @code{[1, 23]} and default value is @code{5}.
19483 @item chroma_msize_y, cy
19484 Set the chroma matrix vertical size.
19485 Range is @code{[1, 23]} and default value is @code{5}.
19487 @item chroma_amount, ca
19488 Set the chroma effect strength.
19489 Range is @code{[-10, 10]} and default value is @code{0.0}.
19491 Negative values will blur the input video, while positive values will
19492 sharpen it, a value of zero will disable the effect.
19496 All parameters are optional and default to the equivalent of the
19497 string '5:5:1.0:5:5:0.0'.
19499 @subsection Examples
19503 Apply strong luma sharpen effect:
19505 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19509 Apply a strong blur of both luma and chroma parameters:
19511 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19515 @c man end OPENCL VIDEO FILTERS
19517 @chapter Video Sources
19518 @c man begin VIDEO SOURCES
19520 Below is a description of the currently available video sources.
19524 Buffer video frames, and make them available to the filter chain.
19526 This source is mainly intended for a programmatic use, in particular
19527 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19529 It accepts the following parameters:
19534 Specify the size (width and height) of the buffered video frames. For the
19535 syntax of this option, check the
19536 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19539 The input video width.
19542 The input video height.
19545 A string representing the pixel format of the buffered video frames.
19546 It may be a number corresponding to a pixel format, or a pixel format
19550 Specify the timebase assumed by the timestamps of the buffered frames.
19553 Specify the frame rate expected for the video stream.
19555 @item pixel_aspect, sar
19556 The sample (pixel) aspect ratio of the input video.
19559 Specify the optional parameters to be used for the scale filter which
19560 is automatically inserted when an input change is detected in the
19561 input size or format.
19563 @item hw_frames_ctx
19564 When using a hardware pixel format, this should be a reference to an
19565 AVHWFramesContext describing input frames.
19570 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19573 will instruct the source to accept video frames with size 320x240 and
19574 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19575 square pixels (1:1 sample aspect ratio).
19576 Since the pixel format with name "yuv410p" corresponds to the number 6
19577 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19578 this example corresponds to:
19580 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19583 Alternatively, the options can be specified as a flat string, but this
19584 syntax is deprecated:
19586 @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}]
19590 Create a pattern generated by an elementary cellular automaton.
19592 The initial state of the cellular automaton can be defined through the
19593 @option{filename} and @option{pattern} options. If such options are
19594 not specified an initial state is created randomly.
19596 At each new frame a new row in the video is filled with the result of
19597 the cellular automaton next generation. The behavior when the whole
19598 frame is filled is defined by the @option{scroll} option.
19600 This source accepts the following options:
19604 Read the initial cellular automaton state, i.e. the starting row, from
19605 the specified file.
19606 In the file, each non-whitespace character is considered an alive
19607 cell, a newline will terminate the row, and further characters in the
19608 file will be ignored.
19611 Read the initial cellular automaton state, i.e. the starting row, from
19612 the specified string.
19614 Each non-whitespace character in the string is considered an alive
19615 cell, a newline will terminate the row, and further characters in the
19616 string will be ignored.
19619 Set the video rate, that is the number of frames generated per second.
19622 @item random_fill_ratio, ratio
19623 Set the random fill ratio for the initial cellular automaton row. It
19624 is a floating point number value ranging from 0 to 1, defaults to
19627 This option is ignored when a file or a pattern is specified.
19629 @item random_seed, seed
19630 Set the seed for filling randomly the initial row, must be an integer
19631 included between 0 and UINT32_MAX. If not specified, or if explicitly
19632 set to -1, the filter will try to use a good random seed on a best
19636 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19637 Default value is 110.
19640 Set the size of the output video. For the syntax of this option, check the
19641 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19643 If @option{filename} or @option{pattern} is specified, the size is set
19644 by default to the width of the specified initial state row, and the
19645 height is set to @var{width} * PHI.
19647 If @option{size} is set, it must contain the width of the specified
19648 pattern string, and the specified pattern will be centered in the
19651 If a filename or a pattern string is not specified, the size value
19652 defaults to "320x518" (used for a randomly generated initial state).
19655 If set to 1, scroll the output upward when all the rows in the output
19656 have been already filled. If set to 0, the new generated row will be
19657 written over the top row just after the bottom row is filled.
19660 @item start_full, full
19661 If set to 1, completely fill the output with generated rows before
19662 outputting the first frame.
19663 This is the default behavior, for disabling set the value to 0.
19666 If set to 1, stitch the left and right row edges together.
19667 This is the default behavior, for disabling set the value to 0.
19670 @subsection Examples
19674 Read the initial state from @file{pattern}, and specify an output of
19677 cellauto=f=pattern:s=200x400
19681 Generate a random initial row with a width of 200 cells, with a fill
19684 cellauto=ratio=2/3:s=200x200
19688 Create a pattern generated by rule 18 starting by a single alive cell
19689 centered on an initial row with width 100:
19691 cellauto=p=@@:s=100x400:full=0:rule=18
19695 Specify a more elaborated initial pattern:
19697 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19702 @anchor{coreimagesrc}
19703 @section coreimagesrc
19704 Video source generated on GPU using Apple's CoreImage API on OSX.
19706 This video source is a specialized version of the @ref{coreimage} video filter.
19707 Use a core image generator at the beginning of the applied filterchain to
19708 generate the content.
19710 The coreimagesrc video source accepts the following options:
19712 @item list_generators
19713 List all available generators along with all their respective options as well as
19714 possible minimum and maximum values along with the default values.
19716 list_generators=true
19720 Specify the size of the sourced video. For the syntax of this option, check the
19721 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19722 The default value is @code{320x240}.
19725 Specify the frame rate of the sourced video, as the number of frames
19726 generated per second. It has to be a string in the format
19727 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19728 number or a valid video frame rate abbreviation. The default value is
19732 Set the sample aspect ratio of the sourced video.
19735 Set the duration of the sourced video. See
19736 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19737 for the accepted syntax.
19739 If not specified, or the expressed duration is negative, the video is
19740 supposed to be generated forever.
19743 Additionally, all options of the @ref{coreimage} video filter are accepted.
19744 A complete filterchain can be used for further processing of the
19745 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19746 and examples for details.
19748 @subsection Examples
19753 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19754 given as complete and escaped command-line for Apple's standard bash shell:
19756 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19758 This example is equivalent to the QRCode example of @ref{coreimage} without the
19759 need for a nullsrc video source.
19763 @section mandelbrot
19765 Generate a Mandelbrot set fractal, and progressively zoom towards the
19766 point specified with @var{start_x} and @var{start_y}.
19768 This source accepts the following options:
19773 Set the terminal pts value. Default value is 400.
19776 Set the terminal scale value.
19777 Must be a floating point value. Default value is 0.3.
19780 Set the inner coloring mode, that is the algorithm used to draw the
19781 Mandelbrot fractal internal region.
19783 It shall assume one of the following values:
19788 Show time until convergence.
19790 Set color based on point closest to the origin of the iterations.
19795 Default value is @var{mincol}.
19798 Set the bailout value. Default value is 10.0.
19801 Set the maximum of iterations performed by the rendering
19802 algorithm. Default value is 7189.
19805 Set outer coloring mode.
19806 It shall assume one of following values:
19808 @item iteration_count
19809 Set iteration count mode.
19810 @item normalized_iteration_count
19811 set normalized iteration count mode.
19813 Default value is @var{normalized_iteration_count}.
19816 Set frame rate, expressed as number of frames per second. Default
19820 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19821 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19824 Set the initial scale value. Default value is 3.0.
19827 Set the initial x position. Must be a floating point value between
19828 -100 and 100. Default value is -0.743643887037158704752191506114774.
19831 Set the initial y position. Must be a floating point value between
19832 -100 and 100. Default value is -0.131825904205311970493132056385139.
19837 Generate various test patterns, as generated by the MPlayer test filter.
19839 The size of the generated video is fixed, and is 256x256.
19840 This source is useful in particular for testing encoding features.
19842 This source accepts the following options:
19847 Specify the frame rate of the sourced video, as the number of frames
19848 generated per second. It has to be a string in the format
19849 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19850 number or a valid video frame rate abbreviation. The default value is
19854 Set the duration of the sourced video. See
19855 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19856 for the accepted syntax.
19858 If not specified, or the expressed duration is negative, the video is
19859 supposed to be generated forever.
19863 Set the number or the name of the test to perform. Supported tests are:
19879 Default value is "all", which will cycle through the list of all tests.
19884 mptestsrc=t=dc_luma
19887 will generate a "dc_luma" test pattern.
19889 @section frei0r_src
19891 Provide a frei0r source.
19893 To enable compilation of this filter you need to install the frei0r
19894 header and configure FFmpeg with @code{--enable-frei0r}.
19896 This source accepts the following parameters:
19901 The size of the video to generate. For the syntax of this option, check the
19902 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19905 The framerate of the generated video. It may be a string of the form
19906 @var{num}/@var{den} or a frame rate abbreviation.
19909 The name to the frei0r source to load. For more information regarding frei0r and
19910 how to set the parameters, read the @ref{frei0r} section in the video filters
19913 @item filter_params
19914 A '|'-separated list of parameters to pass to the frei0r source.
19918 For example, to generate a frei0r partik0l source with size 200x200
19919 and frame rate 10 which is overlaid on the overlay filter main input:
19921 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19926 Generate a life pattern.
19928 This source is based on a generalization of John Conway's life game.
19930 The sourced input represents a life grid, each pixel represents a cell
19931 which can be in one of two possible states, alive or dead. Every cell
19932 interacts with its eight neighbours, which are the cells that are
19933 horizontally, vertically, or diagonally adjacent.
19935 At each interaction the grid evolves according to the adopted rule,
19936 which specifies the number of neighbor alive cells which will make a
19937 cell stay alive or born. The @option{rule} option allows one to specify
19940 This source accepts the following options:
19944 Set the file from which to read the initial grid state. In the file,
19945 each non-whitespace character is considered an alive cell, and newline
19946 is used to delimit the end of each row.
19948 If this option is not specified, the initial grid is generated
19952 Set the video rate, that is the number of frames generated per second.
19955 @item random_fill_ratio, ratio
19956 Set the random fill ratio for the initial random grid. It is a
19957 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19958 It is ignored when a file is specified.
19960 @item random_seed, seed
19961 Set the seed for filling the initial random grid, must be an integer
19962 included between 0 and UINT32_MAX. If not specified, or if explicitly
19963 set to -1, the filter will try to use a good random seed on a best
19969 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19970 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19971 @var{NS} specifies the number of alive neighbor cells which make a
19972 live cell stay alive, and @var{NB} the number of alive neighbor cells
19973 which make a dead cell to become alive (i.e. to "born").
19974 "s" and "b" can be used in place of "S" and "B", respectively.
19976 Alternatively a rule can be specified by an 18-bits integer. The 9
19977 high order bits are used to encode the next cell state if it is alive
19978 for each number of neighbor alive cells, the low order bits specify
19979 the rule for "borning" new cells. Higher order bits encode for an
19980 higher number of neighbor cells.
19981 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19982 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19984 Default value is "S23/B3", which is the original Conway's game of life
19985 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19986 cells, and will born a new cell if there are three alive cells around
19990 Set the size of the output video. For the syntax of this option, check the
19991 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19993 If @option{filename} is specified, the size is set by default to the
19994 same size of the input file. If @option{size} is set, it must contain
19995 the size specified in the input file, and the initial grid defined in
19996 that file is centered in the larger resulting area.
19998 If a filename is not specified, the size value defaults to "320x240"
19999 (used for a randomly generated initial grid).
20002 If set to 1, stitch the left and right grid edges together, and the
20003 top and bottom edges also. Defaults to 1.
20006 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
20007 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
20008 value from 0 to 255.
20011 Set the color of living (or new born) cells.
20014 Set the color of dead cells. If @option{mold} is set, this is the first color
20015 used to represent a dead cell.
20018 Set mold color, for definitely dead and moldy cells.
20020 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
20021 ffmpeg-utils manual,ffmpeg-utils}.
20024 @subsection Examples
20028 Read a grid from @file{pattern}, and center it on a grid of size
20031 life=f=pattern:s=300x300
20035 Generate a random grid of size 200x200, with a fill ratio of 2/3:
20037 life=ratio=2/3:s=200x200
20041 Specify a custom rule for evolving a randomly generated grid:
20047 Full example with slow death effect (mold) using @command{ffplay}:
20049 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
20056 @anchor{haldclutsrc}
20059 @anchor{pal100bars}
20060 @anchor{rgbtestsrc}
20062 @anchor{smptehdbars}
20065 @anchor{yuvtestsrc}
20066 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
20068 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
20070 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
20072 The @code{color} source provides an uniformly colored input.
20074 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
20075 @ref{haldclut} filter.
20077 The @code{nullsrc} source returns unprocessed video frames. It is
20078 mainly useful to be employed in analysis / debugging tools, or as the
20079 source for filters which ignore the input data.
20081 The @code{pal75bars} source generates a color bars pattern, based on
20082 EBU PAL recommendations with 75% color levels.
20084 The @code{pal100bars} source generates a color bars pattern, based on
20085 EBU PAL recommendations with 100% color levels.
20087 The @code{rgbtestsrc} source generates an RGB test pattern useful for
20088 detecting RGB vs BGR issues. You should see a red, green and blue
20089 stripe from top to bottom.
20091 The @code{smptebars} source generates a color bars pattern, based on
20092 the SMPTE Engineering Guideline EG 1-1990.
20094 The @code{smptehdbars} source generates a color bars pattern, based on
20095 the SMPTE RP 219-2002.
20097 The @code{testsrc} source generates a test video pattern, showing a
20098 color pattern, a scrolling gradient and a timestamp. This is mainly
20099 intended for testing purposes.
20101 The @code{testsrc2} source is similar to testsrc, but supports more
20102 pixel formats instead of just @code{rgb24}. This allows using it as an
20103 input for other tests without requiring a format conversion.
20105 The @code{yuvtestsrc} source generates an YUV test pattern. You should
20106 see a y, cb and cr stripe from top to bottom.
20108 The sources accept the following parameters:
20113 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
20114 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
20115 pixels to be used as identity matrix for 3D lookup tables. Each component is
20116 coded on a @code{1/(N*N)} scale.
20119 Specify the color of the source, only available in the @code{color}
20120 source. For the syntax of this option, check the
20121 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
20124 Specify the size of the sourced video. For the syntax of this option, check the
20125 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20126 The default value is @code{320x240}.
20128 This option is not available with the @code{allrgb}, @code{allyuv}, and
20129 @code{haldclutsrc} filters.
20132 Specify the frame rate of the sourced video, as the number of frames
20133 generated per second. It has to be a string in the format
20134 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
20135 number or a valid video frame rate abbreviation. The default value is
20139 Set the duration of the sourced video. See
20140 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20141 for the accepted syntax.
20143 If not specified, or the expressed duration is negative, the video is
20144 supposed to be generated forever.
20147 Set the sample aspect ratio of the sourced video.
20150 Specify the alpha (opacity) of the background, only available in the
20151 @code{testsrc2} source. The value must be between 0 (fully transparent) and
20152 255 (fully opaque, the default).
20155 Set the number of decimals to show in the timestamp, only available in the
20156 @code{testsrc} source.
20158 The displayed timestamp value will correspond to the original
20159 timestamp value multiplied by the power of 10 of the specified
20160 value. Default value is 0.
20163 @subsection Examples
20167 Generate a video with a duration of 5.3 seconds, with size
20168 176x144 and a frame rate of 10 frames per second:
20170 testsrc=duration=5.3:size=qcif:rate=10
20174 The following graph description will generate a red source
20175 with an opacity of 0.2, with size "qcif" and a frame rate of 10
20178 color=c=red@@0.2:s=qcif:r=10
20182 If the input content is to be ignored, @code{nullsrc} can be used. The
20183 following command generates noise in the luminance plane by employing
20184 the @code{geq} filter:
20186 nullsrc=s=256x256, geq=random(1)*255:128:128
20190 @subsection Commands
20192 The @code{color} source supports the following commands:
20196 Set the color of the created image. Accepts the same syntax of the
20197 corresponding @option{color} option.
20202 Generate video using an OpenCL program.
20207 OpenCL program source file.
20210 Kernel name in program.
20213 Size of frames to generate. This must be set.
20216 Pixel format to use for the generated frames. This must be set.
20219 Number of frames generated every second. Default value is '25'.
20223 For details of how the program loading works, see the @ref{program_opencl}
20230 Generate a colour ramp by setting pixel values from the position of the pixel
20231 in the output image. (Note that this will work with all pixel formats, but
20232 the generated output will not be the same.)
20234 __kernel void ramp(__write_only image2d_t dst,
20235 unsigned int index)
20237 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20240 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20242 write_imagef(dst, loc, val);
20247 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20249 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20250 unsigned int index)
20252 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20254 float4 value = 0.0f;
20255 int x = loc.x + index;
20256 int y = loc.y + index;
20257 while (x > 0 || y > 0) {
20258 if (x % 3 == 1 && y % 3 == 1) {
20266 write_imagef(dst, loc, value);
20272 @c man end VIDEO SOURCES
20274 @chapter Video Sinks
20275 @c man begin VIDEO SINKS
20277 Below is a description of the currently available video sinks.
20279 @section buffersink
20281 Buffer video frames, and make them available to the end of the filter
20284 This sink is mainly intended for programmatic use, in particular
20285 through the interface defined in @file{libavfilter/buffersink.h}
20286 or the options system.
20288 It accepts a pointer to an AVBufferSinkContext structure, which
20289 defines the incoming buffers' formats, to be passed as the opaque
20290 parameter to @code{avfilter_init_filter} for initialization.
20294 Null video sink: do absolutely nothing with the input video. It is
20295 mainly useful as a template and for use in analysis / debugging
20298 @c man end VIDEO SINKS
20300 @chapter Multimedia Filters
20301 @c man begin MULTIMEDIA FILTERS
20303 Below is a description of the currently available multimedia filters.
20307 Convert input audio to a video output, displaying the audio bit scope.
20309 The filter accepts the following options:
20313 Set frame rate, expressed as number of frames per second. Default
20317 Specify the video size for the output. For the syntax of this option, check the
20318 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20319 Default value is @code{1024x256}.
20322 Specify list of colors separated by space or by '|' which will be used to
20323 draw channels. Unrecognized or missing colors will be replaced
20327 @section ahistogram
20329 Convert input audio to a video output, displaying the volume histogram.
20331 The filter accepts the following options:
20335 Specify how histogram is calculated.
20337 It accepts the following values:
20340 Use single histogram for all channels.
20342 Use separate histogram for each channel.
20344 Default is @code{single}.
20347 Set frame rate, expressed as number of frames per second. Default
20351 Specify the video size for the output. For the syntax of this option, check the
20352 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20353 Default value is @code{hd720}.
20358 It accepts the following values:
20369 reverse logarithmic
20371 Default is @code{log}.
20374 Set amplitude scale.
20376 It accepts the following values:
20383 Default is @code{log}.
20386 Set how much frames to accumulate in histogram.
20387 Default is 1. Setting this to -1 accumulates all frames.
20390 Set histogram ratio of window height.
20393 Set sonogram sliding.
20395 It accepts the following values:
20398 replace old rows with new ones.
20400 scroll from top to bottom.
20402 Default is @code{replace}.
20405 @section aphasemeter
20407 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20408 representing mean phase of current audio frame. A video output can also be produced and is
20409 enabled by default. The audio is passed through as first output.
20411 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20412 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20413 and @code{1} means channels are in phase.
20415 The filter accepts the following options, all related to its video output:
20419 Set the output frame rate. Default value is @code{25}.
20422 Set the video size for the output. For the syntax of this option, check the
20423 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20424 Default value is @code{800x400}.
20429 Specify the red, green, blue contrast. Default values are @code{2},
20430 @code{7} and @code{1}.
20431 Allowed range is @code{[0, 255]}.
20434 Set color which will be used for drawing median phase. If color is
20435 @code{none} which is default, no median phase value will be drawn.
20438 Enable video output. Default is enabled.
20441 @section avectorscope
20443 Convert input audio to a video output, representing the audio vector
20446 The filter is used to measure the difference between channels of stereo
20447 audio stream. A monoaural signal, consisting of identical left and right
20448 signal, results in straight vertical line. Any stereo separation is visible
20449 as a deviation from this line, creating a Lissajous figure.
20450 If the straight (or deviation from it) but horizontal line appears this
20451 indicates that the left and right channels are out of phase.
20453 The filter accepts the following options:
20457 Set the vectorscope mode.
20459 Available values are:
20462 Lissajous rotated by 45 degrees.
20465 Same as above but not rotated.
20468 Shape resembling half of circle.
20471 Default value is @samp{lissajous}.
20474 Set the video size for the output. For the syntax of this option, check the
20475 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20476 Default value is @code{400x400}.
20479 Set the output frame rate. Default value is @code{25}.
20485 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20486 @code{160}, @code{80} and @code{255}.
20487 Allowed range is @code{[0, 255]}.
20493 Specify the red, green, blue and alpha fade. Default values are @code{15},
20494 @code{10}, @code{5} and @code{5}.
20495 Allowed range is @code{[0, 255]}.
20498 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20499 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20502 Set the vectorscope drawing mode.
20504 Available values are:
20507 Draw dot for each sample.
20510 Draw line between previous and current sample.
20513 Default value is @samp{dot}.
20516 Specify amplitude scale of audio samples.
20518 Available values are:
20534 Swap left channel axis with right channel axis.
20544 Mirror only x axis.
20547 Mirror only y axis.
20555 @subsection Examples
20559 Complete example using @command{ffplay}:
20561 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20562 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20566 @section bench, abench
20568 Benchmark part of a filtergraph.
20570 The filter accepts the following options:
20574 Start or stop a timer.
20576 Available values are:
20579 Get the current time, set it as frame metadata (using the key
20580 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20583 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20584 the input frame metadata to get the time difference. Time difference, average,
20585 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20586 @code{min}) are then printed. The timestamps are expressed in seconds.
20590 @subsection Examples
20594 Benchmark @ref{selectivecolor} filter:
20596 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20602 Concatenate audio and video streams, joining them together one after the
20605 The filter works on segments of synchronized video and audio streams. All
20606 segments must have the same number of streams of each type, and that will
20607 also be the number of streams at output.
20609 The filter accepts the following options:
20614 Set the number of segments. Default is 2.
20617 Set the number of output video streams, that is also the number of video
20618 streams in each segment. Default is 1.
20621 Set the number of output audio streams, that is also the number of audio
20622 streams in each segment. Default is 0.
20625 Activate unsafe mode: do not fail if segments have a different format.
20629 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20630 @var{a} audio outputs.
20632 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20633 segment, in the same order as the outputs, then the inputs for the second
20636 Related streams do not always have exactly the same duration, for various
20637 reasons including codec frame size or sloppy authoring. For that reason,
20638 related synchronized streams (e.g. a video and its audio track) should be
20639 concatenated at once. The concat filter will use the duration of the longest
20640 stream in each segment (except the last one), and if necessary pad shorter
20641 audio streams with silence.
20643 For this filter to work correctly, all segments must start at timestamp 0.
20645 All corresponding streams must have the same parameters in all segments; the
20646 filtering system will automatically select a common pixel format for video
20647 streams, and a common sample format, sample rate and channel layout for
20648 audio streams, but other settings, such as resolution, must be converted
20649 explicitly by the user.
20651 Different frame rates are acceptable but will result in variable frame rate
20652 at output; be sure to configure the output file to handle it.
20654 @subsection Examples
20658 Concatenate an opening, an episode and an ending, all in bilingual version
20659 (video in stream 0, audio in streams 1 and 2):
20661 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20662 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20663 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20664 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20668 Concatenate two parts, handling audio and video separately, using the
20669 (a)movie sources, and adjusting the resolution:
20671 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20672 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20673 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20675 Note that a desync will happen at the stitch if the audio and video streams
20676 do not have exactly the same duration in the first file.
20680 @subsection Commands
20682 This filter supports the following commands:
20685 Close the current segment and step to the next one
20688 @section drawgraph, adrawgraph
20690 Draw a graph using input video or audio metadata.
20692 It accepts the following parameters:
20696 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20699 Set 1st foreground color expression.
20702 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20705 Set 2nd foreground color expression.
20708 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20711 Set 3rd foreground color expression.
20714 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20717 Set 4th foreground color expression.
20720 Set minimal value of metadata value.
20723 Set maximal value of metadata value.
20726 Set graph background color. Default is white.
20731 Available values for mode is:
20738 Default is @code{line}.
20743 Available values for slide is:
20746 Draw new frame when right border is reached.
20749 Replace old columns with new ones.
20752 Scroll from right to left.
20755 Scroll from left to right.
20758 Draw single picture.
20761 Default is @code{frame}.
20764 Set size of graph video. For the syntax of this option, check the
20765 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20766 The default value is @code{900x256}.
20768 The foreground color expressions can use the following variables:
20771 Minimal value of metadata value.
20774 Maximal value of metadata value.
20777 Current metadata key value.
20780 The color is defined as 0xAABBGGRR.
20783 Example using metadata from @ref{signalstats} filter:
20785 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20788 Example using metadata from @ref{ebur128} filter:
20790 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20796 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
20797 level. By default, it logs a message at a frequency of 10Hz with the
20798 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20799 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20801 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
20802 sample format is double-precision floating point. The input stream will be converted to
20803 this specification, if needed. Users may need to insert aformat and/or aresample filters
20804 after this filter to obtain the original parameters.
20806 The filter also has a video output (see the @var{video} option) with a real
20807 time graph to observe the loudness evolution. The graphic contains the logged
20808 message mentioned above, so it is not printed anymore when this option is set,
20809 unless the verbose logging is set. The main graphing area contains the
20810 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20811 the momentary loudness (400 milliseconds), but can optionally be configured
20812 to instead display short-term loudness (see @var{gauge}).
20814 The green area marks a +/- 1LU target range around the target loudness
20815 (-23LUFS by default, unless modified through @var{target}).
20817 More information about the Loudness Recommendation EBU R128 on
20818 @url{http://tech.ebu.ch/loudness}.
20820 The filter accepts the following options:
20825 Activate the video output. The audio stream is passed unchanged whether this
20826 option is set or no. The video stream will be the first output stream if
20827 activated. Default is @code{0}.
20830 Set the video size. This option is for video only. For the syntax of this
20832 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20833 Default and minimum resolution is @code{640x480}.
20836 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20837 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20838 other integer value between this range is allowed.
20841 Set metadata injection. If set to @code{1}, the audio input will be segmented
20842 into 100ms output frames, each of them containing various loudness information
20843 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20845 Default is @code{0}.
20848 Force the frame logging level.
20850 Available values are:
20853 information logging level
20855 verbose logging level
20858 By default, the logging level is set to @var{info}. If the @option{video} or
20859 the @option{metadata} options are set, it switches to @var{verbose}.
20864 Available modes can be cumulated (the option is a @code{flag} type). Possible
20868 Disable any peak mode (default).
20870 Enable sample-peak mode.
20872 Simple peak mode looking for the higher sample value. It logs a message
20873 for sample-peak (identified by @code{SPK}).
20875 Enable true-peak mode.
20877 If enabled, the peak lookup is done on an over-sampled version of the input
20878 stream for better peak accuracy. It logs a message for true-peak.
20879 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20880 This mode requires a build with @code{libswresample}.
20884 Treat mono input files as "dual mono". If a mono file is intended for playback
20885 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20886 If set to @code{true}, this option will compensate for this effect.
20887 Multi-channel input files are not affected by this option.
20890 Set a specific pan law to be used for the measurement of dual mono files.
20891 This parameter is optional, and has a default value of -3.01dB.
20894 Set a specific target level (in LUFS) used as relative zero in the visualization.
20895 This parameter is optional and has a default value of -23LUFS as specified
20896 by EBU R128. However, material published online may prefer a level of -16LUFS
20897 (e.g. for use with podcasts or video platforms).
20900 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20901 @code{shortterm}. By default the momentary value will be used, but in certain
20902 scenarios it may be more useful to observe the short term value instead (e.g.
20906 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20907 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20908 video output, not the summary or continuous log output.
20911 @subsection Examples
20915 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20917 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20921 Run an analysis with @command{ffmpeg}:
20923 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20927 @section interleave, ainterleave
20929 Temporally interleave frames from several inputs.
20931 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20933 These filters read frames from several inputs and send the oldest
20934 queued frame to the output.
20936 Input streams must have well defined, monotonically increasing frame
20939 In order to submit one frame to output, these filters need to enqueue
20940 at least one frame for each input, so they cannot work in case one
20941 input is not yet terminated and will not receive incoming frames.
20943 For example consider the case when one input is a @code{select} filter
20944 which always drops input frames. The @code{interleave} filter will keep
20945 reading from that input, but it will never be able to send new frames
20946 to output until the input sends an end-of-stream signal.
20948 Also, depending on inputs synchronization, the filters will drop
20949 frames in case one input receives more frames than the other ones, and
20950 the queue is already filled.
20952 These filters accept the following options:
20956 Set the number of different inputs, it is 2 by default.
20959 @subsection Examples
20963 Interleave frames belonging to different streams using @command{ffmpeg}:
20965 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20969 Add flickering blur effect:
20971 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20975 @section metadata, ametadata
20977 Manipulate frame metadata.
20979 This filter accepts the following options:
20983 Set mode of operation of the filter.
20985 Can be one of the following:
20989 If both @code{value} and @code{key} is set, select frames
20990 which have such metadata. If only @code{key} is set, select
20991 every frame that has such key in metadata.
20994 Add new metadata @code{key} and @code{value}. If key is already available
20998 Modify value of already present key.
21001 If @code{value} is set, delete only keys that have such value.
21002 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
21006 Print key and its value if metadata was found. If @code{key} is not set print all
21007 metadata values available in frame.
21011 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
21014 Set metadata value which will be used. This option is mandatory for
21015 @code{modify} and @code{add} mode.
21018 Which function to use when comparing metadata value and @code{value}.
21020 Can be one of following:
21024 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
21027 Values are interpreted as strings, returns true if metadata value starts with
21028 the @code{value} option string.
21031 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
21034 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
21037 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
21040 Values are interpreted as floats, returns true if expression from option @code{expr}
21045 Set expression which is used when @code{function} is set to @code{expr}.
21046 The expression is evaluated through the eval API and can contain the following
21051 Float representation of @code{value} from metadata key.
21054 Float representation of @code{value} as supplied by user in @code{value} option.
21058 If specified in @code{print} mode, output is written to the named file. Instead of
21059 plain filename any writable url can be specified. Filename ``-'' is a shorthand
21060 for standard output. If @code{file} option is not set, output is written to the log
21061 with AV_LOG_INFO loglevel.
21065 @subsection Examples
21069 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
21072 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
21075 Print silencedetect output to file @file{metadata.txt}.
21077 silencedetect,ametadata=mode=print:file=metadata.txt
21080 Direct all metadata to a pipe with file descriptor 4.
21082 metadata=mode=print:file='pipe\:4'
21086 @section perms, aperms
21088 Set read/write permissions for the output frames.
21090 These filters are mainly aimed at developers to test direct path in the
21091 following filter in the filtergraph.
21093 The filters accept the following options:
21097 Select the permissions mode.
21099 It accepts the following values:
21102 Do nothing. This is the default.
21104 Set all the output frames read-only.
21106 Set all the output frames directly writable.
21108 Make the frame read-only if writable, and writable if read-only.
21110 Set each output frame read-only or writable randomly.
21114 Set the seed for the @var{random} mode, must be an integer included between
21115 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
21116 @code{-1}, the filter will try to use a good random seed on a best effort
21120 Note: in case of auto-inserted filter between the permission filter and the
21121 following one, the permission might not be received as expected in that
21122 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
21123 perms/aperms filter can avoid this problem.
21125 @section realtime, arealtime
21127 Slow down filtering to match real time approximately.
21129 These filters will pause the filtering for a variable amount of time to
21130 match the output rate with the input timestamps.
21131 They are similar to the @option{re} option to @code{ffmpeg}.
21133 They accept the following options:
21137 Time limit for the pauses. Any pause longer than that will be considered
21138 a timestamp discontinuity and reset the timer. Default is 2 seconds.
21142 @section select, aselect
21144 Select frames to pass in output.
21146 This filter accepts the following options:
21151 Set expression, which is evaluated for each input frame.
21153 If the expression is evaluated to zero, the frame is discarded.
21155 If the evaluation result is negative or NaN, the frame is sent to the
21156 first output; otherwise it is sent to the output with index
21157 @code{ceil(val)-1}, assuming that the input index starts from 0.
21159 For example a value of @code{1.2} corresponds to the output with index
21160 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
21163 Set the number of outputs. The output to which to send the selected
21164 frame is based on the result of the evaluation. Default value is 1.
21167 The expression can contain the following constants:
21171 The (sequential) number of the filtered frame, starting from 0.
21174 The (sequential) number of the selected frame, starting from 0.
21176 @item prev_selected_n
21177 The sequential number of the last selected frame. It's NAN if undefined.
21180 The timebase of the input timestamps.
21183 The PTS (Presentation TimeStamp) of the filtered video frame,
21184 expressed in @var{TB} units. It's NAN if undefined.
21187 The PTS of the filtered video frame,
21188 expressed in seconds. It's NAN if undefined.
21191 The PTS of the previously filtered video frame. It's NAN if undefined.
21193 @item prev_selected_pts
21194 The PTS of the last previously filtered video frame. It's NAN if undefined.
21196 @item prev_selected_t
21197 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
21200 The PTS of the first video frame in the video. It's NAN if undefined.
21203 The time of the first video frame in the video. It's NAN if undefined.
21205 @item pict_type @emph{(video only)}
21206 The type of the filtered frame. It can assume one of the following
21218 @item interlace_type @emph{(video only)}
21219 The frame interlace type. It can assume one of the following values:
21222 The frame is progressive (not interlaced).
21224 The frame is top-field-first.
21226 The frame is bottom-field-first.
21229 @item consumed_sample_n @emph{(audio only)}
21230 the number of selected samples before the current frame
21232 @item samples_n @emph{(audio only)}
21233 the number of samples in the current frame
21235 @item sample_rate @emph{(audio only)}
21236 the input sample rate
21239 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21242 the position in the file of the filtered frame, -1 if the information
21243 is not available (e.g. for synthetic video)
21245 @item scene @emph{(video only)}
21246 value between 0 and 1 to indicate a new scene; a low value reflects a low
21247 probability for the current frame to introduce a new scene, while a higher
21248 value means the current frame is more likely to be one (see the example below)
21250 @item concatdec_select
21251 The concat demuxer can select only part of a concat input file by setting an
21252 inpoint and an outpoint, but the output packets may not be entirely contained
21253 in the selected interval. By using this variable, it is possible to skip frames
21254 generated by the concat demuxer which are not exactly contained in the selected
21257 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21258 and the @var{lavf.concat.duration} packet metadata values which are also
21259 present in the decoded frames.
21261 The @var{concatdec_select} variable is -1 if the frame pts is at least
21262 start_time and either the duration metadata is missing or the frame pts is less
21263 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21266 That basically means that an input frame is selected if its pts is within the
21267 interval set by the concat demuxer.
21271 The default value of the select expression is "1".
21273 @subsection Examples
21277 Select all frames in input:
21282 The example above is the same as:
21294 Select only I-frames:
21296 select='eq(pict_type\,I)'
21300 Select one frame every 100:
21302 select='not(mod(n\,100))'
21306 Select only frames contained in the 10-20 time interval:
21308 select=between(t\,10\,20)
21312 Select only I-frames contained in the 10-20 time interval:
21314 select=between(t\,10\,20)*eq(pict_type\,I)
21318 Select frames with a minimum distance of 10 seconds:
21320 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21324 Use aselect to select only audio frames with samples number > 100:
21326 aselect='gt(samples_n\,100)'
21330 Create a mosaic of the first scenes:
21332 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21335 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21339 Send even and odd frames to separate outputs, and compose them:
21341 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21345 Select useful frames from an ffconcat file which is using inpoints and
21346 outpoints but where the source files are not intra frame only.
21348 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21352 @section sendcmd, asendcmd
21354 Send commands to filters in the filtergraph.
21356 These filters read commands to be sent to other filters in the
21359 @code{sendcmd} must be inserted between two video filters,
21360 @code{asendcmd} must be inserted between two audio filters, but apart
21361 from that they act the same way.
21363 The specification of commands can be provided in the filter arguments
21364 with the @var{commands} option, or in a file specified by the
21365 @var{filename} option.
21367 These filters accept the following options:
21370 Set the commands to be read and sent to the other filters.
21372 Set the filename of the commands to be read and sent to the other
21376 @subsection Commands syntax
21378 A commands description consists of a sequence of interval
21379 specifications, comprising a list of commands to be executed when a
21380 particular event related to that interval occurs. The occurring event
21381 is typically the current frame time entering or leaving a given time
21384 An interval is specified by the following syntax:
21386 @var{START}[-@var{END}] @var{COMMANDS};
21389 The time interval is specified by the @var{START} and @var{END} times.
21390 @var{END} is optional and defaults to the maximum time.
21392 The current frame time is considered within the specified interval if
21393 it is included in the interval [@var{START}, @var{END}), that is when
21394 the time is greater or equal to @var{START} and is lesser than
21397 @var{COMMANDS} consists of a sequence of one or more command
21398 specifications, separated by ",", relating to that interval. The
21399 syntax of a command specification is given by:
21401 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21404 @var{FLAGS} is optional and specifies the type of events relating to
21405 the time interval which enable sending the specified command, and must
21406 be a non-null sequence of identifier flags separated by "+" or "|" and
21407 enclosed between "[" and "]".
21409 The following flags are recognized:
21412 The command is sent when the current frame timestamp enters the
21413 specified interval. In other words, the command is sent when the
21414 previous frame timestamp was not in the given interval, and the
21418 The command is sent when the current frame timestamp leaves the
21419 specified interval. In other words, the command is sent when the
21420 previous frame timestamp was in the given interval, and the
21424 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21427 @var{TARGET} specifies the target of the command, usually the name of
21428 the filter class or a specific filter instance name.
21430 @var{COMMAND} specifies the name of the command for the target filter.
21432 @var{ARG} is optional and specifies the optional list of argument for
21433 the given @var{COMMAND}.
21435 Between one interval specification and another, whitespaces, or
21436 sequences of characters starting with @code{#} until the end of line,
21437 are ignored and can be used to annotate comments.
21439 A simplified BNF description of the commands specification syntax
21442 @var{COMMAND_FLAG} ::= "enter" | "leave"
21443 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21444 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21445 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21446 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21447 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21450 @subsection Examples
21454 Specify audio tempo change at second 4:
21456 asendcmd=c='4.0 atempo tempo 1.5',atempo
21460 Target a specific filter instance:
21462 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21466 Specify a list of drawtext and hue commands in a file.
21468 # show text in the interval 5-10
21469 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21470 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21472 # desaturate the image in the interval 15-20
21473 15.0-20.0 [enter] hue s 0,
21474 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21476 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21478 # apply an exponential saturation fade-out effect, starting from time 25
21479 25 [enter] hue s exp(25-t)
21482 A filtergraph allowing to read and process the above command list
21483 stored in a file @file{test.cmd}, can be specified with:
21485 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21490 @section setpts, asetpts
21492 Change the PTS (presentation timestamp) of the input frames.
21494 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21496 This filter accepts the following options:
21501 The expression which is evaluated for each frame to construct its timestamp.
21505 The expression is evaluated through the eval API and can contain the following
21509 @item FRAME_RATE, FR
21510 frame rate, only defined for constant frame-rate video
21513 The presentation timestamp in input
21516 The count of the input frame for video or the number of consumed samples,
21517 not including the current frame for audio, starting from 0.
21519 @item NB_CONSUMED_SAMPLES
21520 The number of consumed samples, not including the current frame (only
21523 @item NB_SAMPLES, S
21524 The number of samples in the current frame (only audio)
21526 @item SAMPLE_RATE, SR
21527 The audio sample rate.
21530 The PTS of the first frame.
21533 the time in seconds of the first frame
21536 State whether the current frame is interlaced.
21539 the time in seconds of the current frame
21542 original position in the file of the frame, or undefined if undefined
21543 for the current frame
21546 The previous input PTS.
21549 previous input time in seconds
21552 The previous output PTS.
21555 previous output time in seconds
21558 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21562 The wallclock (RTC) time at the start of the movie in microseconds.
21565 The timebase of the input timestamps.
21569 @subsection Examples
21573 Start counting PTS from zero
21575 setpts=PTS-STARTPTS
21579 Apply fast motion effect:
21585 Apply slow motion effect:
21591 Set fixed rate of 25 frames per second:
21597 Set fixed rate 25 fps with some jitter:
21599 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21603 Apply an offset of 10 seconds to the input PTS:
21609 Generate timestamps from a "live source" and rebase onto the current timebase:
21611 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21615 Generate timestamps by counting samples:
21624 Force color range for the output video frame.
21626 The @code{setrange} filter marks the color range property for the
21627 output frames. It does not change the input frame, but only sets the
21628 corresponding property, which affects how the frame is treated by
21631 The filter accepts the following options:
21636 Available values are:
21640 Keep the same color range property.
21642 @item unspecified, unknown
21643 Set the color range as unspecified.
21645 @item limited, tv, mpeg
21646 Set the color range as limited.
21648 @item full, pc, jpeg
21649 Set the color range as full.
21653 @section settb, asettb
21655 Set the timebase to use for the output frames timestamps.
21656 It is mainly useful for testing timebase configuration.
21658 It accepts the following parameters:
21663 The expression which is evaluated into the output timebase.
21667 The value for @option{tb} is an arithmetic expression representing a
21668 rational. The expression can contain the constants "AVTB" (the default
21669 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21670 audio only). Default value is "intb".
21672 @subsection Examples
21676 Set the timebase to 1/25:
21682 Set the timebase to 1/10:
21688 Set the timebase to 1001/1000:
21694 Set the timebase to 2*intb:
21700 Set the default timebase value:
21707 Convert input audio to a video output representing frequency spectrum
21708 logarithmically using Brown-Puckette constant Q transform algorithm with
21709 direct frequency domain coefficient calculation (but the transform itself
21710 is not really constant Q, instead the Q factor is actually variable/clamped),
21711 with musical tone scale, from E0 to D#10.
21713 The filter accepts the following options:
21717 Specify the video size for the output. It must be even. For the syntax of this option,
21718 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21719 Default value is @code{1920x1080}.
21722 Set the output frame rate. Default value is @code{25}.
21725 Set the bargraph height. It must be even. Default value is @code{-1} which
21726 computes the bargraph height automatically.
21729 Set the axis height. It must be even. Default value is @code{-1} which computes
21730 the axis height automatically.
21733 Set the sonogram height. It must be even. Default value is @code{-1} which
21734 computes the sonogram height automatically.
21737 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21738 instead. Default value is @code{1}.
21740 @item sono_v, volume
21741 Specify the sonogram volume expression. It can contain variables:
21744 the @var{bar_v} evaluated expression
21745 @item frequency, freq, f
21746 the frequency where it is evaluated
21747 @item timeclamp, tc
21748 the value of @var{timeclamp} option
21752 @item a_weighting(f)
21753 A-weighting of equal loudness
21754 @item b_weighting(f)
21755 B-weighting of equal loudness
21756 @item c_weighting(f)
21757 C-weighting of equal loudness.
21759 Default value is @code{16}.
21761 @item bar_v, volume2
21762 Specify the bargraph volume expression. It can contain variables:
21765 the @var{sono_v} evaluated expression
21766 @item frequency, freq, f
21767 the frequency where it is evaluated
21768 @item timeclamp, tc
21769 the value of @var{timeclamp} option
21773 @item a_weighting(f)
21774 A-weighting of equal loudness
21775 @item b_weighting(f)
21776 B-weighting of equal loudness
21777 @item c_weighting(f)
21778 C-weighting of equal loudness.
21780 Default value is @code{sono_v}.
21782 @item sono_g, gamma
21783 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21784 higher gamma makes the spectrum having more range. Default value is @code{3}.
21785 Acceptable range is @code{[1, 7]}.
21787 @item bar_g, gamma2
21788 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21792 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21793 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21795 @item timeclamp, tc
21796 Specify the transform timeclamp. At low frequency, there is trade-off between
21797 accuracy in time domain and frequency domain. If timeclamp is lower,
21798 event in time domain is represented more accurately (such as fast bass drum),
21799 otherwise event in frequency domain is represented more accurately
21800 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21803 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21804 limits future samples by applying asymmetric windowing in time domain, useful
21805 when low latency is required. Accepted range is @code{[0, 1]}.
21808 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21809 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21812 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21813 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21816 This option is deprecated and ignored.
21819 Specify the transform length in time domain. Use this option to control accuracy
21820 trade-off between time domain and frequency domain at every frequency sample.
21821 It can contain variables:
21823 @item frequency, freq, f
21824 the frequency where it is evaluated
21825 @item timeclamp, tc
21826 the value of @var{timeclamp} option.
21828 Default value is @code{384*tc/(384+tc*f)}.
21831 Specify the transform count for every video frame. Default value is @code{6}.
21832 Acceptable range is @code{[1, 30]}.
21835 Specify the transform count for every single pixel. Default value is @code{0},
21836 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21839 Specify font file for use with freetype to draw the axis. If not specified,
21840 use embedded font. Note that drawing with font file or embedded font is not
21841 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21845 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21846 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21849 Specify font color expression. This is arithmetic expression that should return
21850 integer value 0xRRGGBB. It can contain variables:
21852 @item frequency, freq, f
21853 the frequency where it is evaluated
21854 @item timeclamp, tc
21855 the value of @var{timeclamp} option
21860 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21861 @item r(x), g(x), b(x)
21862 red, green, and blue value of intensity x.
21864 Default value is @code{st(0, (midi(f)-59.5)/12);
21865 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21866 r(1-ld(1)) + b(ld(1))}.
21869 Specify image file to draw the axis. This option override @var{fontfile} and
21870 @var{fontcolor} option.
21873 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21874 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21875 Default value is @code{1}.
21878 Set colorspace. The accepted values are:
21881 Unspecified (default)
21890 BT.470BG or BT.601-6 625
21893 SMPTE-170M or BT.601-6 525
21899 BT.2020 with non-constant luminance
21904 Set spectrogram color scheme. This is list of floating point values with format
21905 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21906 The default is @code{1|0.5|0|0|0.5|1}.
21910 @subsection Examples
21914 Playing audio while showing the spectrum:
21916 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21920 Same as above, but with frame rate 30 fps:
21922 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21926 Playing at 1280x720:
21928 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21932 Disable sonogram display:
21938 A1 and its harmonics: A1, A2, (near)E3, A3:
21940 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),
21941 asplit[a][out1]; [a] showcqt [out0]'
21945 Same as above, but with more accuracy in frequency domain:
21947 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),
21948 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21954 bar_v=10:sono_v=bar_v*a_weighting(f)
21958 Custom gamma, now spectrum is linear to the amplitude.
21964 Custom tlength equation:
21966 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)))'
21970 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21972 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21976 Custom font using fontconfig:
21978 font='Courier New,Monospace,mono|bold'
21982 Custom frequency range with custom axis using image file:
21984 axisfile=myaxis.png:basefreq=40:endfreq=10000
21990 Convert input audio to video output representing the audio power spectrum.
21991 Audio amplitude is on Y-axis while frequency is on X-axis.
21993 The filter accepts the following options:
21997 Specify size of video. For the syntax of this option, check the
21998 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21999 Default is @code{1024x512}.
22003 This set how each frequency bin will be represented.
22005 It accepts the following values:
22011 Default is @code{bar}.
22014 Set amplitude scale.
22016 It accepts the following values:
22030 Default is @code{log}.
22033 Set frequency scale.
22035 It accepts the following values:
22044 Reverse logarithmic scale.
22046 Default is @code{lin}.
22051 It accepts the following values:
22067 Default is @code{w2048}
22070 Set windowing function.
22072 It accepts the following values:
22095 Default is @code{hanning}.
22098 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22099 which means optimal overlap for selected window function will be picked.
22102 Set time averaging. Setting this to 0 will display current maximal peaks.
22103 Default is @code{1}, which means time averaging is disabled.
22106 Specify list of colors separated by space or by '|' which will be used to
22107 draw channel frequencies. Unrecognized or missing colors will be replaced
22111 Set channel display mode.
22113 It accepts the following values:
22118 Default is @code{combined}.
22121 Set minimum amplitude used in @code{log} amplitude scaler.
22125 @anchor{showspectrum}
22126 @section showspectrum
22128 Convert input audio to a video output, representing the audio frequency
22131 The filter accepts the following options:
22135 Specify the video size for the output. For the syntax of this option, check the
22136 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22137 Default value is @code{640x512}.
22140 Specify how the spectrum should slide along the window.
22142 It accepts the following values:
22145 the samples start again on the left when they reach the right
22147 the samples scroll from right to left
22149 frames are only produced when the samples reach the right
22151 the samples scroll from left to right
22154 Default value is @code{replace}.
22157 Specify display mode.
22159 It accepts the following values:
22162 all channels are displayed in the same row
22164 all channels are displayed in separate rows
22167 Default value is @samp{combined}.
22170 Specify display color mode.
22172 It accepts the following values:
22175 each channel is displayed in a separate color
22177 each channel is displayed using the same color scheme
22179 each channel is displayed using the rainbow color scheme
22181 each channel is displayed using the moreland color scheme
22183 each channel is displayed using the nebulae color scheme
22185 each channel is displayed using the fire color scheme
22187 each channel is displayed using the fiery color scheme
22189 each channel is displayed using the fruit color scheme
22191 each channel is displayed using the cool color scheme
22193 each channel is displayed using the magma color scheme
22195 each channel is displayed using the green color scheme
22197 each channel is displayed using the viridis color scheme
22199 each channel is displayed using the plasma color scheme
22201 each channel is displayed using the cividis color scheme
22203 each channel is displayed using the terrain color scheme
22206 Default value is @samp{channel}.
22209 Specify scale used for calculating intensity color values.
22211 It accepts the following values:
22216 square root, default
22227 Default value is @samp{sqrt}.
22230 Specify frequency scale.
22232 It accepts the following values:
22240 Default value is @samp{lin}.
22243 Set saturation modifier for displayed colors. Negative values provide
22244 alternative color scheme. @code{0} is no saturation at all.
22245 Saturation must be in [-10.0, 10.0] range.
22246 Default value is @code{1}.
22249 Set window function.
22251 It accepts the following values:
22276 Default value is @code{hann}.
22279 Set orientation of time vs frequency axis. Can be @code{vertical} or
22280 @code{horizontal}. Default is @code{vertical}.
22283 Set ratio of overlap window. Default value is @code{0}.
22284 When value is @code{1} overlap is set to recommended size for specific
22285 window function currently used.
22288 Set scale gain for calculating intensity color values.
22289 Default value is @code{1}.
22292 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22295 Set color rotation, must be in [-1.0, 1.0] range.
22296 Default value is @code{0}.
22299 Set start frequency from which to display spectrogram. Default is @code{0}.
22302 Set stop frequency to which to display spectrogram. Default is @code{0}.
22305 Set upper frame rate limit. Default is @code{auto}, unlimited.
22308 Draw time and frequency axes and legends. Default is disabled.
22311 The usage is very similar to the showwaves filter; see the examples in that
22314 @subsection Examples
22318 Large window with logarithmic color scaling:
22320 showspectrum=s=1280x480:scale=log
22324 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22326 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22327 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22331 @section showspectrumpic
22333 Convert input audio to a single video frame, representing the audio frequency
22336 The filter accepts the following options:
22340 Specify the video size for the output. For the syntax of this option, check the
22341 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22342 Default value is @code{4096x2048}.
22345 Specify display mode.
22347 It accepts the following values:
22350 all channels are displayed in the same row
22352 all channels are displayed in separate rows
22354 Default value is @samp{combined}.
22357 Specify display color mode.
22359 It accepts the following values:
22362 each channel is displayed in a separate color
22364 each channel is displayed using the same color scheme
22366 each channel is displayed using the rainbow color scheme
22368 each channel is displayed using the moreland color scheme
22370 each channel is displayed using the nebulae color scheme
22372 each channel is displayed using the fire color scheme
22374 each channel is displayed using the fiery color scheme
22376 each channel is displayed using the fruit color scheme
22378 each channel is displayed using the cool color scheme
22380 each channel is displayed using the magma color scheme
22382 each channel is displayed using the green color scheme
22384 each channel is displayed using the viridis color scheme
22386 each channel is displayed using the plasma color scheme
22388 each channel is displayed using the cividis color scheme
22390 each channel is displayed using the terrain color scheme
22392 Default value is @samp{intensity}.
22395 Specify scale used for calculating intensity color values.
22397 It accepts the following values:
22402 square root, default
22412 Default value is @samp{log}.
22415 Specify frequency scale.
22417 It accepts the following values:
22425 Default value is @samp{lin}.
22428 Set saturation modifier for displayed colors. Negative values provide
22429 alternative color scheme. @code{0} is no saturation at all.
22430 Saturation must be in [-10.0, 10.0] range.
22431 Default value is @code{1}.
22434 Set window function.
22436 It accepts the following values:
22460 Default value is @code{hann}.
22463 Set orientation of time vs frequency axis. Can be @code{vertical} or
22464 @code{horizontal}. Default is @code{vertical}.
22467 Set scale gain for calculating intensity color values.
22468 Default value is @code{1}.
22471 Draw time and frequency axes and legends. Default is enabled.
22474 Set color rotation, must be in [-1.0, 1.0] range.
22475 Default value is @code{0}.
22478 Set start frequency from which to display spectrogram. Default is @code{0}.
22481 Set stop frequency to which to display spectrogram. Default is @code{0}.
22484 @subsection Examples
22488 Extract an audio spectrogram of a whole audio track
22489 in a 1024x1024 picture using @command{ffmpeg}:
22491 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22495 @section showvolume
22497 Convert input audio volume to a video output.
22499 The filter accepts the following options:
22506 Set border width, allowed range is [0, 5]. Default is 1.
22509 Set channel width, allowed range is [80, 8192]. Default is 400.
22512 Set channel height, allowed range is [1, 900]. Default is 20.
22515 Set fade, allowed range is [0, 1]. Default is 0.95.
22518 Set volume color expression.
22520 The expression can use the following variables:
22524 Current max volume of channel in dB.
22530 Current channel number, starting from 0.
22534 If set, displays channel names. Default is enabled.
22537 If set, displays volume values. Default is enabled.
22540 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22541 default is @code{h}.
22544 Set step size, allowed range is [0, 5]. Default is 0, which means
22548 Set background opacity, allowed range is [0, 1]. Default is 0.
22551 Set metering mode, can be peak: @code{p} or rms: @code{r},
22552 default is @code{p}.
22555 Set display scale, can be linear: @code{lin} or log: @code{log},
22556 default is @code{lin}.
22560 If set to > 0., display a line for the max level
22561 in the previous seconds.
22562 default is disabled: @code{0.}
22565 The color of the max line. Use when @code{dm} option is set to > 0.
22566 default is: @code{orange}
22571 Convert input audio to a video output, representing the samples waves.
22573 The filter accepts the following options:
22577 Specify the video size for the output. For the syntax of this option, check the
22578 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22579 Default value is @code{600x240}.
22584 Available values are:
22587 Draw a point for each sample.
22590 Draw a vertical line for each sample.
22593 Draw a point for each sample and a line between them.
22596 Draw a centered vertical line for each sample.
22599 Default value is @code{point}.
22602 Set the number of samples which are printed on the same column. A
22603 larger value will decrease the frame rate. Must be a positive
22604 integer. This option can be set only if the value for @var{rate}
22605 is not explicitly specified.
22608 Set the (approximate) output frame rate. This is done by setting the
22609 option @var{n}. Default value is "25".
22611 @item split_channels
22612 Set if channels should be drawn separately or overlap. Default value is 0.
22615 Set colors separated by '|' which are going to be used for drawing of each channel.
22618 Set amplitude scale.
22620 Available values are:
22638 Set the draw mode. This is mostly useful to set for high @var{n}.
22640 Available values are:
22643 Scale pixel values for each drawn sample.
22646 Draw every sample directly.
22649 Default value is @code{scale}.
22652 @subsection Examples
22656 Output the input file audio and the corresponding video representation
22659 amovie=a.mp3,asplit[out0],showwaves[out1]
22663 Create a synthetic signal and show it with showwaves, forcing a
22664 frame rate of 30 frames per second:
22666 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22670 @section showwavespic
22672 Convert input audio to a single video frame, representing the samples waves.
22674 The filter accepts the following options:
22678 Specify the video size for the output. For the syntax of this option, check the
22679 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22680 Default value is @code{600x240}.
22682 @item split_channels
22683 Set if channels should be drawn separately or overlap. Default value is 0.
22686 Set colors separated by '|' which are going to be used for drawing of each channel.
22689 Set amplitude scale.
22691 Available values are:
22711 Available values are:
22714 Scale pixel values for each drawn sample.
22717 Draw every sample directly.
22720 Default value is @code{scale}.
22723 @subsection Examples
22727 Extract a channel split representation of the wave form of a whole audio track
22728 in a 1024x800 picture using @command{ffmpeg}:
22730 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22734 @section sidedata, asidedata
22736 Delete frame side data, or select frames based on it.
22738 This filter accepts the following options:
22742 Set mode of operation of the filter.
22744 Can be one of the following:
22748 Select every frame with side data of @code{type}.
22751 Delete side data of @code{type}. If @code{type} is not set, delete all side
22757 Set side data type used with all modes. Must be set for @code{select} mode. For
22758 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22759 in @file{libavutil/frame.h}. For example, to choose
22760 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22764 @section spectrumsynth
22766 Sythesize audio from 2 input video spectrums, first input stream represents
22767 magnitude across time and second represents phase across time.
22768 The filter will transform from frequency domain as displayed in videos back
22769 to time domain as presented in audio output.
22771 This filter is primarily created for reversing processed @ref{showspectrum}
22772 filter outputs, but can synthesize sound from other spectrograms too.
22773 But in such case results are going to be poor if the phase data is not
22774 available, because in such cases phase data need to be recreated, usually
22775 it's just recreated from random noise.
22776 For best results use gray only output (@code{channel} color mode in
22777 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22778 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22779 @code{data} option. Inputs videos should generally use @code{fullframe}
22780 slide mode as that saves resources needed for decoding video.
22782 The filter accepts the following options:
22786 Specify sample rate of output audio, the sample rate of audio from which
22787 spectrum was generated may differ.
22790 Set number of channels represented in input video spectrums.
22793 Set scale which was used when generating magnitude input spectrum.
22794 Can be @code{lin} or @code{log}. Default is @code{log}.
22797 Set slide which was used when generating inputs spectrums.
22798 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22799 Default is @code{fullframe}.
22802 Set window function used for resynthesis.
22805 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22806 which means optimal overlap for selected window function will be picked.
22809 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22810 Default is @code{vertical}.
22813 @subsection Examples
22817 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22818 then resynthesize videos back to audio with spectrumsynth:
22820 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
22821 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
22822 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22826 @section split, asplit
22828 Split input into several identical outputs.
22830 @code{asplit} works with audio input, @code{split} with video.
22832 The filter accepts a single parameter which specifies the number of outputs. If
22833 unspecified, it defaults to 2.
22835 @subsection Examples
22839 Create two separate outputs from the same input:
22841 [in] split [out0][out1]
22845 To create 3 or more outputs, you need to specify the number of
22848 [in] asplit=3 [out0][out1][out2]
22852 Create two separate outputs from the same input, one cropped and
22855 [in] split [splitout1][splitout2];
22856 [splitout1] crop=100:100:0:0 [cropout];
22857 [splitout2] pad=200:200:100:100 [padout];
22861 Create 5 copies of the input audio with @command{ffmpeg}:
22863 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22869 Receive commands sent through a libzmq client, and forward them to
22870 filters in the filtergraph.
22872 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22873 must be inserted between two video filters, @code{azmq} between two
22874 audio filters. Both are capable to send messages to any filter type.
22876 To enable these filters you need to install the libzmq library and
22877 headers and configure FFmpeg with @code{--enable-libzmq}.
22879 For more information about libzmq see:
22880 @url{http://www.zeromq.org/}
22882 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22883 receives messages sent through a network interface defined by the
22884 @option{bind_address} (or the abbreviation "@option{b}") option.
22885 Default value of this option is @file{tcp://localhost:5555}. You may
22886 want to alter this value to your needs, but do not forget to escape any
22887 ':' signs (see @ref{filtergraph escaping}).
22889 The received message must be in the form:
22891 @var{TARGET} @var{COMMAND} [@var{ARG}]
22894 @var{TARGET} specifies the target of the command, usually the name of
22895 the filter class or a specific filter instance name. The default
22896 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22897 but you can override this by using the @samp{filter_name@@id} syntax
22898 (see @ref{Filtergraph syntax}).
22900 @var{COMMAND} specifies the name of the command for the target filter.
22902 @var{ARG} is optional and specifies the optional argument list for the
22903 given @var{COMMAND}.
22905 Upon reception, the message is processed and the corresponding command
22906 is injected into the filtergraph. Depending on the result, the filter
22907 will send a reply to the client, adopting the format:
22909 @var{ERROR_CODE} @var{ERROR_REASON}
22913 @var{MESSAGE} is optional.
22915 @subsection Examples
22917 Look at @file{tools/zmqsend} for an example of a zmq client which can
22918 be used to send commands processed by these filters.
22920 Consider the following filtergraph generated by @command{ffplay}.
22921 In this example the last overlay filter has an instance name. All other
22922 filters will have default instance names.
22925 ffplay -dumpgraph 1 -f lavfi "
22926 color=s=100x100:c=red [l];
22927 color=s=100x100:c=blue [r];
22928 nullsrc=s=200x100, zmq [bg];
22929 [bg][l] overlay [bg+l];
22930 [bg+l][r] overlay@@my=x=100 "
22933 To change the color of the left side of the video, the following
22934 command can be used:
22936 echo Parsed_color_0 c yellow | tools/zmqsend
22939 To change the right side:
22941 echo Parsed_color_1 c pink | tools/zmqsend
22944 To change the position of the right side:
22946 echo overlay@@my x 150 | tools/zmqsend
22950 @c man end MULTIMEDIA FILTERS
22952 @chapter Multimedia Sources
22953 @c man begin MULTIMEDIA SOURCES
22955 Below is a description of the currently available multimedia sources.
22959 This is the same as @ref{movie} source, except it selects an audio
22965 Read audio and/or video stream(s) from a movie container.
22967 It accepts the following parameters:
22971 The name of the resource to read (not necessarily a file; it can also be a
22972 device or a stream accessed through some protocol).
22974 @item format_name, f
22975 Specifies the format assumed for the movie to read, and can be either
22976 the name of a container or an input device. If not specified, the
22977 format is guessed from @var{movie_name} or by probing.
22979 @item seek_point, sp
22980 Specifies the seek point in seconds. The frames will be output
22981 starting from this seek point. The parameter is evaluated with
22982 @code{av_strtod}, so the numerical value may be suffixed by an IS
22983 postfix. The default value is "0".
22986 Specifies the streams to read. Several streams can be specified,
22987 separated by "+". The source will then have as many outputs, in the
22988 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22989 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22990 respectively the default (best suited) video and audio stream. Default
22991 is "dv", or "da" if the filter is called as "amovie".
22993 @item stream_index, si
22994 Specifies the index of the video stream to read. If the value is -1,
22995 the most suitable video stream will be automatically selected. The default
22996 value is "-1". Deprecated. If the filter is called "amovie", it will select
22997 audio instead of video.
23000 Specifies how many times to read the stream in sequence.
23001 If the value is 0, the stream will be looped infinitely.
23002 Default value is "1".
23004 Note that when the movie is looped the source timestamps are not
23005 changed, so it will generate non monotonically increasing timestamps.
23007 @item discontinuity
23008 Specifies the time difference between frames above which the point is
23009 considered a timestamp discontinuity which is removed by adjusting the later
23013 It allows overlaying a second video on top of the main input of
23014 a filtergraph, as shown in this graph:
23016 input -----------> deltapts0 --> overlay --> output
23019 movie --> scale--> deltapts1 -------+
23021 @subsection Examples
23025 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
23026 on top of the input labelled "in":
23028 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
23029 [in] setpts=PTS-STARTPTS [main];
23030 [main][over] overlay=16:16 [out]
23034 Read from a video4linux2 device, and overlay it on top of the input
23037 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
23038 [in] setpts=PTS-STARTPTS [main];
23039 [main][over] overlay=16:16 [out]
23043 Read the first video stream and the audio stream with id 0x81 from
23044 dvd.vob; the video is connected to the pad named "video" and the audio is
23045 connected to the pad named "audio":
23047 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
23051 @subsection Commands
23053 Both movie and amovie support the following commands:
23056 Perform seek using "av_seek_frame".
23057 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
23060 @var{stream_index}: If stream_index is -1, a default
23061 stream is selected, and @var{timestamp} is automatically converted
23062 from AV_TIME_BASE units to the stream specific time_base.
23064 @var{timestamp}: Timestamp in AVStream.time_base units
23065 or, if no stream is specified, in AV_TIME_BASE units.
23067 @var{flags}: Flags which select direction and seeking mode.
23071 Get movie duration in AV_TIME_BASE units.
23075 @c man end MULTIMEDIA SOURCES