1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}@@@var{id}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program optionally followed by "@@@var{id}".
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{FILTER_NAME} ::= @var{NAME}["@@"@var{NAME}]
216 @var{LINKLABEL} ::= "[" @var{NAME} "]"
217 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
218 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
219 @var{FILTER} ::= [@var{LINKLABELS}] @var{FILTER_NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
220 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
221 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
224 @anchor{filtergraph escaping}
225 @section Notes on filtergraph escaping
227 Filtergraph description composition entails several levels of
228 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
229 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
230 information about the employed escaping procedure.
232 A first level escaping affects the content of each filter option
233 value, which may contain the special character @code{:} used to
234 separate values, or one of the escaping characters @code{\'}.
236 A second level escaping affects the whole filter description, which
237 may contain the escaping characters @code{\'} or the special
238 characters @code{[],;} used by the filtergraph description.
240 Finally, when you specify a filtergraph on a shell commandline, you
241 need to perform a third level escaping for the shell special
242 characters contained within it.
244 For example, consider the following string to be embedded in
245 the @ref{drawtext} filter description @option{text} value:
247 this is a 'string': may contain one, or more, special characters
250 This string contains the @code{'} special escaping character, and the
251 @code{:} special character, so it needs to be escaped in this way:
253 text=this is a \'string\'\: may contain one, or more, special characters
256 A second level of escaping is required when embedding the filter
257 description in a filtergraph description, in order to escape all the
258 filtergraph special characters. Thus the example above becomes:
260 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
262 (note that in addition to the @code{\'} escaping special characters,
263 also @code{,} needs to be escaped).
265 Finally an additional level of escaping is needed when writing the
266 filtergraph description in a shell command, which depends on the
267 escaping rules of the adopted shell. For example, assuming that
268 @code{\} is special and needs to be escaped with another @code{\}, the
269 previous string will finally result in:
271 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
274 @chapter Timeline editing
276 Some filters support a generic @option{enable} option. For the filters
277 supporting timeline editing, this option can be set to an expression which is
278 evaluated before sending a frame to the filter. If the evaluation is non-zero,
279 the filter will be enabled, otherwise the frame will be sent unchanged to the
280 next filter in the filtergraph.
282 The expression accepts the following values:
285 timestamp expressed in seconds, NAN if the input timestamp is unknown
288 sequential number of the input frame, starting from 0
291 the position in the file of the input frame, NAN if unknown
295 width and height of the input frame if video
298 Additionally, these filters support an @option{enable} command that can be used
299 to re-define the expression.
301 Like any other filtering option, the @option{enable} option follows the same
304 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
305 minutes, and a @ref{curves} filter starting at 3 seconds:
307 smartblur = enable='between(t,10,3*60)',
308 curves = enable='gte(t,3)' : preset=cross_process
311 See @code{ffmpeg -filters} to view which filters have timeline support.
313 @c man end FILTERGRAPH DESCRIPTION
316 @chapter Options for filters with several inputs (framesync)
317 @c man begin OPTIONS FOR FILTERS WITH SEVERAL INPUTS
319 Some filters with several inputs support a common set of options.
320 These options can only be set by name, not with the short notation.
324 The action to take when EOF is encountered on the secondary input; it accepts
325 one of the following values:
329 Repeat the last frame (the default).
333 Pass the main input through.
337 If set to 1, force the output to terminate when the shortest input
338 terminates. Default value is 0.
341 If set to 1, force the filter to extend the last frame of secondary streams
342 until the end of the primary stream. A value of 0 disables this behavior.
346 @c man end OPTIONS FOR FILTERS WITH SEVERAL INPUTS
348 @chapter Audio Filters
349 @c man begin AUDIO FILTERS
351 When you configure your FFmpeg build, you can disable any of the
352 existing filters using @code{--disable-filters}.
353 The configure output will show the audio filters included in your
356 Below is a description of the currently available audio filters.
360 A compressor is mainly used to reduce the dynamic range of a signal.
361 Especially modern music is mostly compressed at a high ratio to
362 improve the overall loudness. It's done to get the highest attention
363 of a listener, "fatten" the sound and bring more "power" to the track.
364 If a signal is compressed too much it may sound dull or "dead"
365 afterwards or it may start to "pump" (which could be a powerful effect
366 but can also destroy a track completely).
367 The right compression is the key to reach a professional sound and is
368 the high art of mixing and mastering. Because of its complex settings
369 it may take a long time to get the right feeling for this kind of effect.
371 Compression is done by detecting the volume above a chosen level
372 @code{threshold} and dividing it by the factor set with @code{ratio}.
373 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
374 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
375 the signal would cause distortion of the waveform the reduction can be
376 levelled over the time. This is done by setting "Attack" and "Release".
377 @code{attack} determines how long the signal has to rise above the threshold
378 before any reduction will occur and @code{release} sets the time the signal
379 has to fall below the threshold to reduce the reduction again. Shorter signals
380 than the chosen attack time will be left untouched.
381 The overall reduction of the signal can be made up afterwards with the
382 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
383 raising the makeup to this level results in a signal twice as loud than the
384 source. To gain a softer entry in the compression the @code{knee} flattens the
385 hard edge at the threshold in the range of the chosen decibels.
387 The filter accepts the following options:
391 Set input gain. Default is 1. Range is between 0.015625 and 64.
394 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
395 Default is @code{downward}.
398 If a signal of stream rises above this level it will affect the gain
400 By default it is 0.125. Range is between 0.00097563 and 1.
403 Set a ratio by which the signal is reduced. 1:2 means that if the level
404 rose 4dB above the threshold, it will be only 2dB above after the reduction.
405 Default is 2. Range is between 1 and 20.
408 Amount of milliseconds the signal has to rise above the threshold before gain
409 reduction starts. Default is 20. Range is between 0.01 and 2000.
412 Amount of milliseconds the signal has to fall below the threshold before
413 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
416 Set the amount by how much signal will be amplified after processing.
417 Default is 1. Range is from 1 to 64.
420 Curve the sharp knee around the threshold to enter gain reduction more softly.
421 Default is 2.82843. Range is between 1 and 8.
424 Choose if the @code{average} level between all channels of input stream
425 or the louder(@code{maximum}) channel of input stream affects the
426 reduction. Default is @code{average}.
429 Should the exact signal be taken in case of @code{peak} or an RMS one in case
430 of @code{rms}. Default is @code{rms} which is mostly smoother.
433 How much to use compressed signal in output. Default is 1.
434 Range is between 0 and 1.
438 Simple audio dynamic range compression/expansion filter.
440 The filter accepts the following options:
444 Set contrast. Default is 33. Allowed range is between 0 and 100.
449 Copy the input audio source unchanged to the output. This is mainly useful for
454 Apply cross fade from one input audio stream to another input audio stream.
455 The cross fade is applied for specified duration near the end of first stream.
457 The filter accepts the following options:
461 Specify the number of samples for which the cross fade effect has to last.
462 At the end of the cross fade effect the first input audio will be completely
463 silent. Default is 44100.
466 Specify the duration of the cross fade effect. See
467 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
468 for the accepted syntax.
469 By default the duration is determined by @var{nb_samples}.
470 If set this option is used instead of @var{nb_samples}.
473 Should first stream end overlap with second stream start. Default is enabled.
476 Set curve for cross fade transition for first stream.
479 Set curve for cross fade transition for second stream.
481 For description of available curve types see @ref{afade} filter description.
488 Cross fade from one input to another:
490 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
494 Cross fade from one input to another but without overlapping:
496 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
501 Split audio stream into several bands.
503 This filter splits audio stream into two or more frequency ranges.
504 Summing all streams back will give flat output.
506 The filter accepts the following options:
510 Set split frequencies. Those must be positive and increasing.
513 Set filter order, can be @var{2nd}, @var{4th} or @var{8th}.
514 Default is @var{4th}.
519 Reduce audio bit resolution.
521 This filter is bit crusher with enhanced functionality. A bit crusher
522 is used to audibly reduce number of bits an audio signal is sampled
523 with. This doesn't change the bit depth at all, it just produces the
524 effect. Material reduced in bit depth sounds more harsh and "digital".
525 This filter is able to even round to continuous values instead of discrete
527 Additionally it has a D/C offset which results in different crushing of
528 the lower and the upper half of the signal.
529 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
531 Another feature of this filter is the logarithmic mode.
532 This setting switches from linear distances between bits to logarithmic ones.
533 The result is a much more "natural" sounding crusher which doesn't gate low
534 signals for example. The human ear has a logarithmic perception,
535 so this kind of crushing is much more pleasant.
536 Logarithmic crushing is also able to get anti-aliased.
538 The filter accepts the following options:
554 Can be linear: @code{lin} or logarithmic: @code{log}.
563 Set sample reduction.
566 Enable LFO. By default disabled.
577 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
581 Remove impulsive noise from input audio.
583 Samples detected as impulsive noise are replaced by interpolated samples using
584 autoregressive modelling.
588 Set window size, in milliseconds. Allowed range is from @code{10} to
589 @code{100}. Default value is @code{55} milliseconds.
590 This sets size of window which will be processed at once.
593 Set window overlap, in percentage of window size. Allowed range is from
594 @code{50} to @code{95}. Default value is @code{75} percent.
595 Setting this to a very high value increases impulsive noise removal but makes
596 whole process much slower.
599 Set autoregression order, in percentage of window size. Allowed range is from
600 @code{0} to @code{25}. Default value is @code{2} percent. This option also
601 controls quality of interpolated samples using neighbour good samples.
604 Set threshold value. Allowed range is from @code{1} to @code{100}.
605 Default value is @code{2}.
606 This controls the strength of impulsive noise which is going to be removed.
607 The lower value, the more samples will be detected as impulsive noise.
610 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
611 @code{10}. Default value is @code{2}.
612 If any two samples detected as noise are spaced less than this value then any
613 sample between those two samples will be also detected as noise.
618 It accepts the following values:
621 Select overlap-add method. Even not interpolated samples are slightly
622 changed with this method.
625 Select overlap-save method. Not interpolated samples remain unchanged.
628 Default value is @code{a}.
632 Remove clipped samples from input audio.
634 Samples detected as clipped are replaced by interpolated samples using
635 autoregressive modelling.
639 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
640 Default value is @code{55} milliseconds.
641 This sets size of window which will be processed at once.
644 Set window overlap, in percentage of window size. Allowed range is from @code{50}
645 to @code{95}. Default value is @code{75} percent.
648 Set autoregression order, in percentage of window size. Allowed range is from
649 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
650 quality of interpolated samples using neighbour good samples.
653 Set threshold value. Allowed range is from @code{1} to @code{100}.
654 Default value is @code{10}. Higher values make clip detection less aggressive.
657 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
658 Default value is @code{1000}. Higher values make clip detection less aggressive.
663 It accepts the following values:
666 Select overlap-add method. Even not interpolated samples are slightly changed
670 Select overlap-save method. Not interpolated samples remain unchanged.
673 Default value is @code{a}.
678 Delay one or more audio channels.
680 Samples in delayed channel are filled with silence.
682 The filter accepts the following option:
686 Set list of delays in milliseconds for each channel separated by '|'.
687 Unused delays will be silently ignored. If number of given delays is
688 smaller than number of channels all remaining channels will not be delayed.
689 If you want to delay exact number of samples, append 'S' to number.
690 If you want instead to delay in seconds, append 's' to number.
697 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
698 the second channel (and any other channels that may be present) unchanged.
704 Delay second channel by 500 samples, the third channel by 700 samples and leave
705 the first channel (and any other channels that may be present) unchanged.
711 @section aderivative, aintegral
713 Compute derivative/integral of audio stream.
715 Applying both filters one after another produces original audio.
719 Apply echoing to the input audio.
721 Echoes are reflected sound and can occur naturally amongst mountains
722 (and sometimes large buildings) when talking or shouting; digital echo
723 effects emulate this behaviour and are often used to help fill out the
724 sound of a single instrument or vocal. The time difference between the
725 original signal and the reflection is the @code{delay}, and the
726 loudness of the reflected signal is the @code{decay}.
727 Multiple echoes can have different delays and decays.
729 A description of the accepted parameters follows.
733 Set input gain of reflected signal. Default is @code{0.6}.
736 Set output gain of reflected signal. Default is @code{0.3}.
739 Set list of time intervals in milliseconds between original signal and reflections
740 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
741 Default is @code{1000}.
744 Set list of loudness of reflected signals separated by '|'.
745 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
746 Default is @code{0.5}.
753 Make it sound as if there are twice as many instruments as are actually playing:
755 aecho=0.8:0.88:60:0.4
759 If delay is very short, then it sound like a (metallic) robot playing music:
765 A longer delay will sound like an open air concert in the mountains:
767 aecho=0.8:0.9:1000:0.3
771 Same as above but with one more mountain:
773 aecho=0.8:0.9:1000|1800:0.3|0.25
778 Audio emphasis filter creates or restores material directly taken from LPs or
779 emphased CDs with different filter curves. E.g. to store music on vinyl the
780 signal has to be altered by a filter first to even out the disadvantages of
781 this recording medium.
782 Once the material is played back the inverse filter has to be applied to
783 restore the distortion of the frequency response.
785 The filter accepts the following options:
795 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
796 use @code{production} mode. Default is @code{reproduction} mode.
799 Set filter type. Selects medium. Can be one of the following:
811 select Compact Disc (CD).
817 select 50µs (FM-KF).
819 select 75µs (FM-KF).
825 Modify an audio signal according to the specified expressions.
827 This filter accepts one or more expressions (one for each channel),
828 which are evaluated and used to modify a corresponding audio signal.
830 It accepts the following parameters:
834 Set the '|'-separated expressions list for each separate channel. If
835 the number of input channels is greater than the number of
836 expressions, the last specified expression is used for the remaining
839 @item channel_layout, c
840 Set output channel layout. If not specified, the channel layout is
841 specified by the number of expressions. If set to @samp{same}, it will
842 use by default the same input channel layout.
845 Each expression in @var{exprs} can contain the following constants and functions:
849 channel number of the current expression
852 number of the evaluated sample, starting from 0
858 time of the evaluated sample expressed in seconds
861 @item nb_out_channels
862 input and output number of channels
865 the value of input channel with number @var{CH}
868 Note: this filter is slow. For faster processing you should use a
877 aeval=val(ch)/2:c=same
881 Invert phase of the second channel:
890 Apply fade-in/out effect to input audio.
892 A description of the accepted parameters follows.
896 Specify the effect type, can be either @code{in} for fade-in, or
897 @code{out} for a fade-out effect. Default is @code{in}.
899 @item start_sample, ss
900 Specify the number of the start sample for starting to apply the fade
901 effect. Default is 0.
904 Specify the number of samples for which the fade effect has to last. At
905 the end of the fade-in effect the output audio will have the same
906 volume as the input audio, at the end of the fade-out transition
907 the output audio will be silence. Default is 44100.
910 Specify the start time of the fade effect. Default is 0.
911 The value must be specified as a time duration; see
912 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
913 for the accepted syntax.
914 If set this option is used instead of @var{start_sample}.
917 Specify the duration of the fade effect. See
918 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
919 for the accepted syntax.
920 At the end of the fade-in effect the output audio will have the same
921 volume as the input audio, at the end of the fade-out transition
922 the output audio will be silence.
923 By default the duration is determined by @var{nb_samples}.
924 If set this option is used instead of @var{nb_samples}.
927 Set curve for fade transition.
929 It accepts the following values:
932 select triangular, linear slope (default)
934 select quarter of sine wave
936 select half of sine wave
938 select exponential sine wave
942 select inverted parabola
956 select inverted quarter of sine wave
958 select inverted half of sine wave
960 select double-exponential seat
962 select double-exponential sigmoid
964 select logistic sigmoid
974 Fade in first 15 seconds of audio:
980 Fade out last 25 seconds of a 900 seconds audio:
982 afade=t=out:st=875:d=25
987 Denoise audio samples with FFT.
989 A description of the accepted parameters follows.
993 Set the noise reduction in dB, allowed range is 0.01 to 97.
994 Default value is 12 dB.
997 Set the noise floor in dB, allowed range is -80 to -20.
998 Default value is -50 dB.
1003 It accepts the following values:
1012 Select shellac noise.
1015 Select custom noise, defined in @code{bn} option.
1017 Default value is white noise.
1021 Set custom band noise for every one of 15 bands.
1022 Bands are separated by ' ' or '|'.
1025 Set the residual floor in dB, allowed range is -80 to -20.
1026 Default value is -38 dB.
1029 Enable noise tracking. By default is disabled.
1030 With this enabled, noise floor is automatically adjusted.
1033 Enable residual tracking. By default is disabled.
1036 Set the output mode.
1038 It accepts the following values:
1041 Pass input unchanged.
1044 Pass noise filtered out.
1049 Default value is @var{o}.
1053 @subsection Commands
1055 This filter supports the following commands:
1057 @item sample_noise, sn
1058 Start or stop measuring noise profile.
1059 Syntax for the command is : "start" or "stop" string.
1060 After measuring noise profile is stopped it will be
1061 automatically applied in filtering.
1063 @item noise_reduction, nr
1064 Change noise reduction. Argument is single float number.
1065 Syntax for the command is : "@var{noise_reduction}"
1067 @item noise_floor, nf
1068 Change noise floor. Argument is single float number.
1069 Syntax for the command is : "@var{noise_floor}"
1071 @item output_mode, om
1072 Change output mode operation.
1073 Syntax for the command is : "i", "o" or "n" string.
1077 Apply arbitrary expressions to samples in frequency domain.
1081 Set frequency domain real expression for each separate channel separated
1082 by '|'. Default is "re".
1083 If the number of input channels is greater than the number of
1084 expressions, the last specified expression is used for the remaining
1088 Set frequency domain imaginary expression for each separate channel
1089 separated by '|'. Default is "im".
1091 Each expression in @var{real} and @var{imag} can contain the following
1092 constants and functions:
1099 current frequency bin number
1102 number of available bins
1105 channel number of the current expression
1114 current real part of frequency bin of current channel
1117 current imaginary part of frequency bin of current channel
1120 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1123 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1129 It accepts the following values:
1145 Default is @code{w4096}
1148 Set window function. Default is @code{hann}.
1151 Set window overlap. If set to 1, the recommended overlap for selected
1152 window function will be picked. Default is @code{0.75}.
1155 @subsection Examples
1159 Leave almost only low frequencies in audio:
1161 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1168 Apply an arbitrary Frequency Impulse Response filter.
1170 This filter is designed for applying long FIR filters,
1171 up to 60 seconds long.
1173 It can be used as component for digital crossover filters,
1174 room equalization, cross talk cancellation, wavefield synthesis,
1175 auralization, ambiophonics, ambisonics and spatialization.
1177 This filter uses second stream as FIR coefficients.
1178 If second stream holds single channel, it will be used
1179 for all input channels in first stream, otherwise
1180 number of channels in second stream must be same as
1181 number of channels in first stream.
1183 It accepts the following parameters:
1187 Set dry gain. This sets input gain.
1190 Set wet gain. This sets final output gain.
1193 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1196 Enable applying gain measured from power of IR.
1198 Set which approach to use for auto gain measurement.
1202 Do not apply any gain.
1205 select peak gain, very conservative approach. This is default value.
1208 select DC gain, limited application.
1211 select gain to noise approach, this is most popular one.
1215 Set gain to be applied to IR coefficients before filtering.
1216 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1219 Set format of IR stream. Can be @code{mono} or @code{input}.
1220 Default is @code{input}.
1223 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1224 Allowed range is 0.1 to 60 seconds.
1227 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1228 By default it is disabled.
1231 Set for which IR channel to display frequency response. By default is first channel
1232 displayed. This option is used only when @var{response} is enabled.
1235 Set video stream size. This option is used only when @var{response} is enabled.
1238 Set video stream frame rate. This option is used only when @var{response} is enabled.
1241 Set minimal partition size used for convolution. Default is @var{8192}.
1242 Allowed range is from @var{8} to @var{32768}.
1243 Lower values decreases latency at cost of higher CPU usage.
1246 Set maximal partition size used for convolution. Default is @var{8192}.
1247 Allowed range is from @var{8} to @var{32768}.
1248 Lower values may increase CPU usage.
1251 @subsection Examples
1255 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1257 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1264 Set output format constraints for the input audio. The framework will
1265 negotiate the most appropriate format to minimize conversions.
1267 It accepts the following parameters:
1271 A '|'-separated list of requested sample formats.
1274 A '|'-separated list of requested sample rates.
1276 @item channel_layouts
1277 A '|'-separated list of requested channel layouts.
1279 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1280 for the required syntax.
1283 If a parameter is omitted, all values are allowed.
1285 Force the output to either unsigned 8-bit or signed 16-bit stereo
1287 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1292 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1293 processing reduces disturbing noise between useful signals.
1295 Gating is done by detecting the volume below a chosen level @var{threshold}
1296 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1297 floor is set via @var{range}. Because an exact manipulation of the signal
1298 would cause distortion of the waveform the reduction can be levelled over
1299 time. This is done by setting @var{attack} and @var{release}.
1301 @var{attack} determines how long the signal has to fall below the threshold
1302 before any reduction will occur and @var{release} sets the time the signal
1303 has to rise above the threshold to reduce the reduction again.
1304 Shorter signals than the chosen attack time will be left untouched.
1308 Set input level before filtering.
1309 Default is 1. Allowed range is from 0.015625 to 64.
1312 Set the mode of operation. Can be @code{upward} or @code{downward}.
1313 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
1314 will be amplified, expanding dynamic range in upward direction.
1315 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
1318 Set the level of gain reduction when the signal is below the threshold.
1319 Default is 0.06125. Allowed range is from 0 to 1.
1320 Setting this to 0 disables reduction and then filter behaves like expander.
1323 If a signal rises above this level the gain reduction is released.
1324 Default is 0.125. Allowed range is from 0 to 1.
1327 Set a ratio by which the signal is reduced.
1328 Default is 2. Allowed range is from 1 to 9000.
1331 Amount of milliseconds the signal has to rise above the threshold before gain
1333 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1336 Amount of milliseconds the signal has to fall below the threshold before the
1337 reduction is increased again. Default is 250 milliseconds.
1338 Allowed range is from 0.01 to 9000.
1341 Set amount of amplification of signal after processing.
1342 Default is 1. Allowed range is from 1 to 64.
1345 Curve the sharp knee around the threshold to enter gain reduction more softly.
1346 Default is 2.828427125. Allowed range is from 1 to 8.
1349 Choose if exact signal should be taken for detection or an RMS like one.
1350 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1353 Choose if the average level between all channels or the louder channel affects
1355 Default is @code{average}. Can be @code{average} or @code{maximum}.
1360 Apply an arbitrary Infinite Impulse Response filter.
1362 It accepts the following parameters:
1366 Set numerator/zeros coefficients.
1369 Set denominator/poles coefficients.
1381 Set coefficients format.
1387 Z-plane zeros/poles, cartesian (default)
1389 Z-plane zeros/poles, polar radians
1391 Z-plane zeros/poles, polar degrees
1395 Set kind of processing.
1396 Can be @code{d} - direct or @code{s} - serial cascading. Default is @code{s}.
1399 Set filtering precision.
1403 double-precision floating-point (default)
1405 single-precision floating-point
1413 Show IR frequency response, magnitude and phase in additional video stream.
1414 By default it is disabled.
1417 Set for which IR channel to display frequency response. By default is first channel
1418 displayed. This option is used only when @var{response} is enabled.
1421 Set video stream size. This option is used only when @var{response} is enabled.
1424 Coefficients in @code{tf} format are separated by spaces and are in ascending
1427 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1428 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1431 Different coefficients and gains can be provided for every channel, in such case
1432 use '|' to separate coefficients or gains. Last provided coefficients will be
1433 used for all remaining channels.
1435 @subsection Examples
1439 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1441 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1445 Same as above but in @code{zp} format:
1447 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1453 The limiter prevents an input signal from rising over a desired threshold.
1454 This limiter uses lookahead technology to prevent your signal from distorting.
1455 It means that there is a small delay after the signal is processed. Keep in mind
1456 that the delay it produces is the attack time you set.
1458 The filter accepts the following options:
1462 Set input gain. Default is 1.
1465 Set output gain. Default is 1.
1468 Don't let signals above this level pass the limiter. Default is 1.
1471 The limiter will reach its attenuation level in this amount of time in
1472 milliseconds. Default is 5 milliseconds.
1475 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1476 Default is 50 milliseconds.
1479 When gain reduction is always needed ASC takes care of releasing to an
1480 average reduction level rather than reaching a reduction of 0 in the release
1484 Select how much the release time is affected by ASC, 0 means nearly no changes
1485 in release time while 1 produces higher release times.
1488 Auto level output signal. Default is enabled.
1489 This normalizes audio back to 0dB if enabled.
1492 Depending on picked setting it is recommended to upsample input 2x or 4x times
1493 with @ref{aresample} before applying this filter.
1497 Apply a two-pole all-pass filter with central frequency (in Hz)
1498 @var{frequency}, and filter-width @var{width}.
1499 An all-pass filter changes the audio's frequency to phase relationship
1500 without changing its frequency to amplitude relationship.
1502 The filter accepts the following options:
1506 Set frequency in Hz.
1509 Set method to specify band-width of filter.
1524 Specify the band-width of a filter in width_type units.
1527 Specify which channels to filter, by default all available are filtered.
1530 @subsection Commands
1532 This filter supports the following commands:
1535 Change allpass frequency.
1536 Syntax for the command is : "@var{frequency}"
1539 Change allpass width_type.
1540 Syntax for the command is : "@var{width_type}"
1543 Change allpass width.
1544 Syntax for the command is : "@var{width}"
1551 The filter accepts the following options:
1555 Set the number of loops. Setting this value to -1 will result in infinite loops.
1559 Set maximal number of samples. Default is 0.
1562 Set first sample of loop. Default is 0.
1568 Merge two or more audio streams into a single multi-channel stream.
1570 The filter accepts the following options:
1575 Set the number of inputs. Default is 2.
1579 If the channel layouts of the inputs are disjoint, and therefore compatible,
1580 the channel layout of the output will be set accordingly and the channels
1581 will be reordered as necessary. If the channel layouts of the inputs are not
1582 disjoint, the output will have all the channels of the first input then all
1583 the channels of the second input, in that order, and the channel layout of
1584 the output will be the default value corresponding to the total number of
1587 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1588 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1589 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1590 first input, b1 is the first channel of the second input).
1592 On the other hand, if both input are in stereo, the output channels will be
1593 in the default order: a1, a2, b1, b2, and the channel layout will be
1594 arbitrarily set to 4.0, which may or may not be the expected value.
1596 All inputs must have the same sample rate, and format.
1598 If inputs do not have the same duration, the output will stop with the
1601 @subsection Examples
1605 Merge two mono files into a stereo stream:
1607 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1611 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1613 ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
1619 Mixes multiple audio inputs into a single output.
1621 Note that this filter only supports float samples (the @var{amerge}
1622 and @var{pan} audio filters support many formats). If the @var{amix}
1623 input has integer samples then @ref{aresample} will be automatically
1624 inserted to perform the conversion to float samples.
1628 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1630 will mix 3 input audio streams to a single output with the same duration as the
1631 first input and a dropout transition time of 3 seconds.
1633 It accepts the following parameters:
1637 The number of inputs. If unspecified, it defaults to 2.
1640 How to determine the end-of-stream.
1644 The duration of the longest input. (default)
1647 The duration of the shortest input.
1650 The duration of the first input.
1654 @item dropout_transition
1655 The transition time, in seconds, for volume renormalization when an input
1656 stream ends. The default value is 2 seconds.
1659 Specify weight of each input audio stream as sequence.
1660 Each weight is separated by space. By default all inputs have same weight.
1665 Multiply first audio stream with second audio stream and store result
1666 in output audio stream. Multiplication is done by multiplying each
1667 sample from first stream with sample at same position from second stream.
1669 With this element-wise multiplication one can create amplitude fades and
1670 amplitude modulations.
1672 @section anequalizer
1674 High-order parametric multiband equalizer for each channel.
1676 It accepts the following parameters:
1680 This option string is in format:
1681 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1682 Each equalizer band is separated by '|'.
1686 Set channel number to which equalization will be applied.
1687 If input doesn't have that channel the entry is ignored.
1690 Set central frequency for band.
1691 If input doesn't have that frequency the entry is ignored.
1694 Set band width in hertz.
1697 Set band gain in dB.
1700 Set filter type for band, optional, can be:
1704 Butterworth, this is default.
1715 With this option activated frequency response of anequalizer is displayed
1719 Set video stream size. Only useful if curves option is activated.
1722 Set max gain that will be displayed. Only useful if curves option is activated.
1723 Setting this to a reasonable value makes it possible to display gain which is derived from
1724 neighbour bands which are too close to each other and thus produce higher gain
1725 when both are activated.
1728 Set frequency scale used to draw frequency response in video output.
1729 Can be linear or logarithmic. Default is logarithmic.
1732 Set color for each channel curve which is going to be displayed in video stream.
1733 This is list of color names separated by space or by '|'.
1734 Unrecognised or missing colors will be replaced by white color.
1737 @subsection Examples
1741 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1742 for first 2 channels using Chebyshev type 1 filter:
1744 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1748 @subsection Commands
1750 This filter supports the following commands:
1753 Alter existing filter parameters.
1754 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1756 @var{fN} is existing filter number, starting from 0, if no such filter is available
1758 @var{freq} set new frequency parameter.
1759 @var{width} set new width parameter in herz.
1760 @var{gain} set new gain parameter in dB.
1762 Full filter invocation with asendcmd may look like this:
1763 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1768 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1770 Each sample is adjusted by looking for other samples with similar contexts. This
1771 context similarity is defined by comparing their surrounding patches of size
1772 @option{p}. Patches are searched in an area of @option{r} around the sample.
1774 The filter accepts the following options.
1778 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
1781 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
1782 Default value is 2 milliseconds.
1785 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
1786 Default value is 6 milliseconds.
1789 Set the output mode.
1791 It accepts the following values:
1794 Pass input unchanged.
1797 Pass noise filtered out.
1802 Default value is @var{o}.
1808 Pass the audio source unchanged to the output.
1812 Pad the end of an audio stream with silence.
1814 This can be used together with @command{ffmpeg} @option{-shortest} to
1815 extend audio streams to the same length as the video stream.
1817 A description of the accepted options follows.
1821 Set silence packet size. Default value is 4096.
1824 Set the number of samples of silence to add to the end. After the
1825 value is reached, the stream is terminated. This option is mutually
1826 exclusive with @option{whole_len}.
1829 Set the minimum total number of samples in the output audio stream. If
1830 the value is longer than the input audio length, silence is added to
1831 the end, until the value is reached. This option is mutually exclusive
1832 with @option{pad_len}.
1835 Specify the duration of samples of silence to add. See
1836 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1837 for the accepted syntax. Used only if set to non-zero value.
1840 Specify the minimum total duration in the output audio stream. See
1841 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1842 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1843 the input audio length, silence is added to the end, until the value is reached.
1844 This option is mutually exclusive with @option{pad_dur}
1847 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1848 nor @option{whole_dur} option is set, the filter will add silence to the end of
1849 the input stream indefinitely.
1851 @subsection Examples
1855 Add 1024 samples of silence to the end of the input:
1861 Make sure the audio output will contain at least 10000 samples, pad
1862 the input with silence if required:
1864 apad=whole_len=10000
1868 Use @command{ffmpeg} to pad the audio input with silence, so that the
1869 video stream will always result the shortest and will be converted
1870 until the end in the output file when using the @option{shortest}
1873 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1878 Add a phasing effect to the input audio.
1880 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1881 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1883 A description of the accepted parameters follows.
1887 Set input gain. Default is 0.4.
1890 Set output gain. Default is 0.74
1893 Set delay in milliseconds. Default is 3.0.
1896 Set decay. Default is 0.4.
1899 Set modulation speed in Hz. Default is 0.5.
1902 Set modulation type. Default is triangular.
1904 It accepts the following values:
1913 Audio pulsator is something between an autopanner and a tremolo.
1914 But it can produce funny stereo effects as well. Pulsator changes the volume
1915 of the left and right channel based on a LFO (low frequency oscillator) with
1916 different waveforms and shifted phases.
1917 This filter have the ability to define an offset between left and right
1918 channel. An offset of 0 means that both LFO shapes match each other.
1919 The left and right channel are altered equally - a conventional tremolo.
1920 An offset of 50% means that the shape of the right channel is exactly shifted
1921 in phase (or moved backwards about half of the frequency) - pulsator acts as
1922 an autopanner. At 1 both curves match again. Every setting in between moves the
1923 phase shift gapless between all stages and produces some "bypassing" sounds with
1924 sine and triangle waveforms. The more you set the offset near 1 (starting from
1925 the 0.5) the faster the signal passes from the left to the right speaker.
1927 The filter accepts the following options:
1931 Set input gain. By default it is 1. Range is [0.015625 - 64].
1934 Set output gain. By default it is 1. Range is [0.015625 - 64].
1937 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1938 sawup or sawdown. Default is sine.
1941 Set modulation. Define how much of original signal is affected by the LFO.
1944 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1947 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1950 Set pulse width. Default is 1. Allowed range is [0 - 2].
1953 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1956 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1960 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1964 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1965 if timing is set to hz.
1971 Resample the input audio to the specified parameters, using the
1972 libswresample library. If none are specified then the filter will
1973 automatically convert between its input and output.
1975 This filter is also able to stretch/squeeze the audio data to make it match
1976 the timestamps or to inject silence / cut out audio to make it match the
1977 timestamps, do a combination of both or do neither.
1979 The filter accepts the syntax
1980 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1981 expresses a sample rate and @var{resampler_options} is a list of
1982 @var{key}=@var{value} pairs, separated by ":". See the
1983 @ref{Resampler Options,,"Resampler Options" section in the
1984 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1985 for the complete list of supported options.
1987 @subsection Examples
1991 Resample the input audio to 44100Hz:
1997 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1998 samples per second compensation:
2000 aresample=async=1000
2006 Reverse an audio clip.
2008 Warning: This filter requires memory to buffer the entire clip, so trimming
2011 @subsection Examples
2015 Take the first 5 seconds of a clip, and reverse it.
2017 atrim=end=5,areverse
2021 @section asetnsamples
2023 Set the number of samples per each output audio frame.
2025 The last output packet may contain a different number of samples, as
2026 the filter will flush all the remaining samples when the input audio
2029 The filter accepts the following options:
2033 @item nb_out_samples, n
2034 Set the number of frames per each output audio frame. The number is
2035 intended as the number of samples @emph{per each channel}.
2036 Default value is 1024.
2039 If set to 1, the filter will pad the last audio frame with zeroes, so
2040 that the last frame will contain the same number of samples as the
2041 previous ones. Default value is 1.
2044 For example, to set the number of per-frame samples to 1234 and
2045 disable padding for the last frame, use:
2047 asetnsamples=n=1234:p=0
2052 Set the sample rate without altering the PCM data.
2053 This will result in a change of speed and pitch.
2055 The filter accepts the following options:
2058 @item sample_rate, r
2059 Set the output sample rate. Default is 44100 Hz.
2064 Show a line containing various information for each input audio frame.
2065 The input audio is not modified.
2067 The shown line contains a sequence of key/value pairs of the form
2068 @var{key}:@var{value}.
2070 The following values are shown in the output:
2074 The (sequential) number of the input frame, starting from 0.
2077 The presentation timestamp of the input frame, in time base units; the time base
2078 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2081 The presentation timestamp of the input frame in seconds.
2084 position of the frame in the input stream, -1 if this information in
2085 unavailable and/or meaningless (for example in case of synthetic audio)
2094 The sample rate for the audio frame.
2097 The number of samples (per channel) in the frame.
2100 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2101 audio, the data is treated as if all the planes were concatenated.
2103 @item plane_checksums
2104 A list of Adler-32 checksums for each data plane.
2110 Display time domain statistical information about the audio channels.
2111 Statistics are calculated and displayed for each audio channel and,
2112 where applicable, an overall figure is also given.
2114 It accepts the following option:
2117 Short window length in seconds, used for peak and trough RMS measurement.
2118 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2122 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2123 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2126 Available keys for each channel are:
2162 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2163 this @code{lavfi.astats.Overall.Peak_count}.
2165 For description what each key means read below.
2168 Set number of frame after which stats are going to be recalculated.
2169 Default is disabled.
2171 @item measure_perchannel
2172 Select the entries which need to be measured per channel. The metadata keys can
2173 be used as flags, default is @option{all} which measures everything.
2174 @option{none} disables all per channel measurement.
2176 @item measure_overall
2177 Select the entries which need to be measured overall. The metadata keys can
2178 be used as flags, default is @option{all} which measures everything.
2179 @option{none} disables all overall measurement.
2183 A description of each shown parameter follows:
2187 Mean amplitude displacement from zero.
2190 Minimal sample level.
2193 Maximal sample level.
2195 @item Min difference
2196 Minimal difference between two consecutive samples.
2198 @item Max difference
2199 Maximal difference between two consecutive samples.
2201 @item Mean difference
2202 Mean difference between two consecutive samples.
2203 The average of each difference between two consecutive samples.
2205 @item RMS difference
2206 Root Mean Square difference between two consecutive samples.
2210 Standard peak and RMS level measured in dBFS.
2214 Peak and trough values for RMS level measured over a short window.
2217 Standard ratio of peak to RMS level (note: not in dB).
2220 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2221 (i.e. either @var{Min level} or @var{Max level}).
2224 Number of occasions (not the number of samples) that the signal attained either
2225 @var{Min level} or @var{Max level}.
2228 Overall bit depth of audio. Number of bits used for each sample.
2231 Measured dynamic range of audio in dB.
2233 @item Zero crossings
2234 Number of points where the waveform crosses the zero level axis.
2236 @item Zero crossings rate
2237 Rate of Zero crossings and number of audio samples.
2244 The filter accepts exactly one parameter, the audio tempo. If not
2245 specified then the filter will assume nominal 1.0 tempo. Tempo must
2246 be in the [0.5, 100.0] range.
2248 Note that tempo greater than 2 will skip some samples rather than
2249 blend them in. If for any reason this is a concern it is always
2250 possible to daisy-chain several instances of atempo to achieve the
2251 desired product tempo.
2253 @subsection Examples
2257 Slow down audio to 80% tempo:
2263 To speed up audio to 300% tempo:
2269 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2271 atempo=sqrt(3),atempo=sqrt(3)
2277 Trim the input so that the output contains one continuous subpart of the input.
2279 It accepts the following parameters:
2282 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2283 sample with the timestamp @var{start} will be the first sample in the output.
2286 Specify time of the first audio sample that will be dropped, i.e. the
2287 audio sample immediately preceding the one with the timestamp @var{end} will be
2288 the last sample in the output.
2291 Same as @var{start}, except this option sets the start timestamp in samples
2295 Same as @var{end}, except this option sets the end timestamp in samples instead
2299 The maximum duration of the output in seconds.
2302 The number of the first sample that should be output.
2305 The number of the first sample that should be dropped.
2308 @option{start}, @option{end}, and @option{duration} are expressed as time
2309 duration specifications; see
2310 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2312 Note that the first two sets of the start/end options and the @option{duration}
2313 option look at the frame timestamp, while the _sample options simply count the
2314 samples that pass through the filter. So start/end_pts and start/end_sample will
2315 give different results when the timestamps are wrong, inexact or do not start at
2316 zero. Also note that this filter does not modify the timestamps. If you wish
2317 to have the output timestamps start at zero, insert the asetpts filter after the
2320 If multiple start or end options are set, this filter tries to be greedy and
2321 keep all samples that match at least one of the specified constraints. To keep
2322 only the part that matches all the constraints at once, chain multiple atrim
2325 The defaults are such that all the input is kept. So it is possible to set e.g.
2326 just the end values to keep everything before the specified time.
2331 Drop everything except the second minute of input:
2333 ffmpeg -i INPUT -af atrim=60:120
2337 Keep only the first 1000 samples:
2339 ffmpeg -i INPUT -af atrim=end_sample=1000
2346 Apply a two-pole Butterworth band-pass filter with central
2347 frequency @var{frequency}, and (3dB-point) band-width width.
2348 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2349 instead of the default: constant 0dB peak gain.
2350 The filter roll off at 6dB per octave (20dB per decade).
2352 The filter accepts the following options:
2356 Set the filter's central frequency. Default is @code{3000}.
2359 Constant skirt gain if set to 1. Defaults to 0.
2362 Set method to specify band-width of filter.
2377 Specify the band-width of a filter in width_type units.
2380 Specify which channels to filter, by default all available are filtered.
2383 @subsection Commands
2385 This filter supports the following commands:
2388 Change bandpass frequency.
2389 Syntax for the command is : "@var{frequency}"
2392 Change bandpass width_type.
2393 Syntax for the command is : "@var{width_type}"
2396 Change bandpass width.
2397 Syntax for the command is : "@var{width}"
2402 Apply a two-pole Butterworth band-reject filter with central
2403 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2404 The filter roll off at 6dB per octave (20dB per decade).
2406 The filter accepts the following options:
2410 Set the filter's central frequency. Default is @code{3000}.
2413 Set method to specify band-width of filter.
2428 Specify the band-width of a filter in width_type units.
2431 Specify which channels to filter, by default all available are filtered.
2434 @subsection Commands
2436 This filter supports the following commands:
2439 Change bandreject frequency.
2440 Syntax for the command is : "@var{frequency}"
2443 Change bandreject width_type.
2444 Syntax for the command is : "@var{width_type}"
2447 Change bandreject width.
2448 Syntax for the command is : "@var{width}"
2451 @section bass, lowshelf
2453 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2454 shelving filter with a response similar to that of a standard
2455 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2457 The filter accepts the following options:
2461 Give the gain at 0 Hz. Its useful range is about -20
2462 (for a large cut) to +20 (for a large boost).
2463 Beware of clipping when using a positive gain.
2466 Set the filter's central frequency and so can be used
2467 to extend or reduce the frequency range to be boosted or cut.
2468 The default value is @code{100} Hz.
2471 Set method to specify band-width of filter.
2486 Determine how steep is the filter's shelf transition.
2489 Specify which channels to filter, by default all available are filtered.
2492 @subsection Commands
2494 This filter supports the following commands:
2497 Change bass frequency.
2498 Syntax for the command is : "@var{frequency}"
2501 Change bass width_type.
2502 Syntax for the command is : "@var{width_type}"
2506 Syntax for the command is : "@var{width}"
2510 Syntax for the command is : "@var{gain}"
2515 Apply a biquad IIR filter with the given coefficients.
2516 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2517 are the numerator and denominator coefficients respectively.
2518 and @var{channels}, @var{c} specify which channels to filter, by default all
2519 available are filtered.
2521 @subsection Commands
2523 This filter supports the following commands:
2531 Change biquad parameter.
2532 Syntax for the command is : "@var{value}"
2536 Bauer stereo to binaural transformation, which improves headphone listening of
2537 stereo audio records.
2539 To enable compilation of this filter you need to configure FFmpeg with
2540 @code{--enable-libbs2b}.
2542 It accepts the following parameters:
2546 Pre-defined crossfeed level.
2550 Default level (fcut=700, feed=50).
2553 Chu Moy circuit (fcut=700, feed=60).
2556 Jan Meier circuit (fcut=650, feed=95).
2561 Cut frequency (in Hz).
2570 Remap input channels to new locations.
2572 It accepts the following parameters:
2575 Map channels from input to output. The argument is a '|'-separated list of
2576 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2577 @var{in_channel} form. @var{in_channel} can be either the name of the input
2578 channel (e.g. FL for front left) or its index in the input channel layout.
2579 @var{out_channel} is the name of the output channel or its index in the output
2580 channel layout. If @var{out_channel} is not given then it is implicitly an
2581 index, starting with zero and increasing by one for each mapping.
2583 @item channel_layout
2584 The channel layout of the output stream.
2587 If no mapping is present, the filter will implicitly map input channels to
2588 output channels, preserving indices.
2590 @subsection Examples
2594 For example, assuming a 5.1+downmix input MOV file,
2596 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2598 will create an output WAV file tagged as stereo from the downmix channels of
2602 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2604 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2608 @section channelsplit
2610 Split each channel from an input audio stream into a separate output stream.
2612 It accepts the following parameters:
2614 @item channel_layout
2615 The channel layout of the input stream. The default is "stereo".
2617 A channel layout describing the channels to be extracted as separate output streams
2618 or "all" to extract each input channel as a separate stream. The default is "all".
2620 Choosing channels not present in channel layout in the input will result in an error.
2623 @subsection Examples
2627 For example, assuming a stereo input MP3 file,
2629 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2631 will create an output Matroska file with two audio streams, one containing only
2632 the left channel and the other the right channel.
2635 Split a 5.1 WAV file into per-channel files:
2637 ffmpeg -i in.wav -filter_complex
2638 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2639 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2640 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2645 Extract only LFE from a 5.1 WAV file:
2647 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2648 -map '[LFE]' lfe.wav
2653 Add a chorus effect to the audio.
2655 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2657 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2658 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2659 The modulation depth defines the range the modulated delay is played before or after
2660 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2661 sound tuned around the original one, like in a chorus where some vocals are slightly
2664 It accepts the following parameters:
2667 Set input gain. Default is 0.4.
2670 Set output gain. Default is 0.4.
2673 Set delays. A typical delay is around 40ms to 60ms.
2685 @subsection Examples
2691 chorus=0.7:0.9:55:0.4:0.25:2
2697 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2701 Fuller sounding chorus with three delays:
2703 chorus=0.5:0.9:50|60|40:0.4|0.32|0.3:0.25|0.4|0.3:2|2.3|1.3
2708 Compress or expand the audio's dynamic range.
2710 It accepts the following parameters:
2716 A list of times in seconds for each channel over which the instantaneous level
2717 of the input signal is averaged to determine its volume. @var{attacks} refers to
2718 increase of volume and @var{decays} refers to decrease of volume. For most
2719 situations, the attack time (response to the audio getting louder) should be
2720 shorter than the decay time, because the human ear is more sensitive to sudden
2721 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2722 a typical value for decay is 0.8 seconds.
2723 If specified number of attacks & decays is lower than number of channels, the last
2724 set attack/decay will be used for all remaining channels.
2727 A list of points for the transfer function, specified in dB relative to the
2728 maximum possible signal amplitude. Each key points list must be defined using
2729 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2730 @code{x0/y0 x1/y1 x2/y2 ....}
2732 The input values must be in strictly increasing order but the transfer function
2733 does not have to be monotonically rising. The point @code{0/0} is assumed but
2734 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2735 function are @code{-70/-70|-60/-20|1/0}.
2738 Set the curve radius in dB for all joints. It defaults to 0.01.
2741 Set the additional gain in dB to be applied at all points on the transfer
2742 function. This allows for easy adjustment of the overall gain.
2746 Set an initial volume, in dB, to be assumed for each channel when filtering
2747 starts. This permits the user to supply a nominal level initially, so that, for
2748 example, a very large gain is not applied to initial signal levels before the
2749 companding has begun to operate. A typical value for audio which is initially
2750 quiet is -90 dB. It defaults to 0.
2753 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2754 delayed before being fed to the volume adjuster. Specifying a delay
2755 approximately equal to the attack/decay times allows the filter to effectively
2756 operate in predictive rather than reactive mode. It defaults to 0.
2760 @subsection Examples
2764 Make music with both quiet and loud passages suitable for listening to in a
2767 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2770 Another example for audio with whisper and explosion parts:
2772 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2776 A noise gate for when the noise is at a lower level than the signal:
2778 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2782 Here is another noise gate, this time for when the noise is at a higher level
2783 than the signal (making it, in some ways, similar to squelch):
2785 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2789 2:1 compression starting at -6dB:
2791 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2795 2:1 compression starting at -9dB:
2797 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2801 2:1 compression starting at -12dB:
2803 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2807 2:1 compression starting at -18dB:
2809 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2813 3:1 compression starting at -15dB:
2815 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2821 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2827 compand=attacks=0:points=-80/-169|-54/-80|-49.5/-64.6|-41.1/-41.1|-25.8/-15|-10.8/-4.5|0/0|20/8.3
2831 Hard limiter at -6dB:
2833 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2837 Hard limiter at -12dB:
2839 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2843 Hard noise gate at -35 dB:
2845 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2851 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2855 @section compensationdelay
2857 Compensation Delay Line is a metric based delay to compensate differing
2858 positions of microphones or speakers.
2860 For example, you have recorded guitar with two microphones placed in
2861 different location. Because the front of sound wave has fixed speed in
2862 normal conditions, the phasing of microphones can vary and depends on
2863 their location and interposition. The best sound mix can be achieved when
2864 these microphones are in phase (synchronized). Note that distance of
2865 ~30 cm between microphones makes one microphone to capture signal in
2866 antiphase to another microphone. That makes the final mix sounding moody.
2867 This filter helps to solve phasing problems by adding different delays
2868 to each microphone track and make them synchronized.
2870 The best result can be reached when you take one track as base and
2871 synchronize other tracks one by one with it.
2872 Remember that synchronization/delay tolerance depends on sample rate, too.
2873 Higher sample rates will give more tolerance.
2875 It accepts the following parameters:
2879 Set millimeters distance. This is compensation distance for fine tuning.
2883 Set cm distance. This is compensation distance for tightening distance setup.
2887 Set meters distance. This is compensation distance for hard distance setup.
2891 Set dry amount. Amount of unprocessed (dry) signal.
2895 Set wet amount. Amount of processed (wet) signal.
2899 Set temperature degree in Celsius. This is the temperature of the environment.
2904 Apply headphone crossfeed filter.
2906 Crossfeed is the process of blending the left and right channels of stereo
2908 It is mainly used to reduce extreme stereo separation of low frequencies.
2910 The intent is to produce more speaker like sound to the listener.
2912 The filter accepts the following options:
2916 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2917 This sets gain of low shelf filter for side part of stereo image.
2918 Default is -6dB. Max allowed is -30db when strength is set to 1.
2921 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2922 This sets cut off frequency of low shelf filter. Default is cut off near
2923 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2926 Set input gain. Default is 0.9.
2929 Set output gain. Default is 1.
2932 @section crystalizer
2933 Simple algorithm to expand audio dynamic range.
2935 The filter accepts the following options:
2939 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2940 (unchanged sound) to 10.0 (maximum effect).
2943 Enable clipping. By default is enabled.
2947 Apply a DC shift to the audio.
2949 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2950 in the recording chain) from the audio. The effect of a DC offset is reduced
2951 headroom and hence volume. The @ref{astats} filter can be used to determine if
2952 a signal has a DC offset.
2956 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2960 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2961 used to prevent clipping.
2965 Measure audio dynamic range.
2967 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2968 is found in transition material. And anything less that 8 have very poor dynamics
2969 and is very compressed.
2971 The filter accepts the following options:
2975 Set window length in seconds used to split audio into segments of equal length.
2976 Default is 3 seconds.
2980 Dynamic Audio Normalizer.
2982 This filter applies a certain amount of gain to the input audio in order
2983 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2984 contrast to more "simple" normalization algorithms, the Dynamic Audio
2985 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2986 This allows for applying extra gain to the "quiet" sections of the audio
2987 while avoiding distortions or clipping the "loud" sections. In other words:
2988 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2989 sections, in the sense that the volume of each section is brought to the
2990 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2991 this goal *without* applying "dynamic range compressing". It will retain 100%
2992 of the dynamic range *within* each section of the audio file.
2996 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2997 Default is 500 milliseconds.
2998 The Dynamic Audio Normalizer processes the input audio in small chunks,
2999 referred to as frames. This is required, because a peak magnitude has no
3000 meaning for just a single sample value. Instead, we need to determine the
3001 peak magnitude for a contiguous sequence of sample values. While a "standard"
3002 normalizer would simply use the peak magnitude of the complete file, the
3003 Dynamic Audio Normalizer determines the peak magnitude individually for each
3004 frame. The length of a frame is specified in milliseconds. By default, the
3005 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
3006 been found to give good results with most files.
3007 Note that the exact frame length, in number of samples, will be determined
3008 automatically, based on the sampling rate of the individual input audio file.
3011 Set the Gaussian filter window size. In range from 3 to 301, must be odd
3012 number. Default is 31.
3013 Probably the most important parameter of the Dynamic Audio Normalizer is the
3014 @code{window size} of the Gaussian smoothing filter. The filter's window size
3015 is specified in frames, centered around the current frame. For the sake of
3016 simplicity, this must be an odd number. Consequently, the default value of 31
3017 takes into account the current frame, as well as the 15 preceding frames and
3018 the 15 subsequent frames. Using a larger window results in a stronger
3019 smoothing effect and thus in less gain variation, i.e. slower gain
3020 adaptation. Conversely, using a smaller window results in a weaker smoothing
3021 effect and thus in more gain variation, i.e. faster gain adaptation.
3022 In other words, the more you increase this value, the more the Dynamic Audio
3023 Normalizer will behave like a "traditional" normalization filter. On the
3024 contrary, the more you decrease this value, the more the Dynamic Audio
3025 Normalizer will behave like a dynamic range compressor.
3028 Set the target peak value. This specifies the highest permissible magnitude
3029 level for the normalized audio input. This filter will try to approach the
3030 target peak magnitude as closely as possible, but at the same time it also
3031 makes sure that the normalized signal will never exceed the peak magnitude.
3032 A frame's maximum local gain factor is imposed directly by the target peak
3033 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3034 It is not recommended to go above this value.
3037 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3038 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3039 factor for each input frame, i.e. the maximum gain factor that does not
3040 result in clipping or distortion. The maximum gain factor is determined by
3041 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3042 additionally bounds the frame's maximum gain factor by a predetermined
3043 (global) maximum gain factor. This is done in order to avoid excessive gain
3044 factors in "silent" or almost silent frames. By default, the maximum gain
3045 factor is 10.0, For most inputs the default value should be sufficient and
3046 it usually is not recommended to increase this value. Though, for input
3047 with an extremely low overall volume level, it may be necessary to allow even
3048 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3049 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3050 Instead, a "sigmoid" threshold function will be applied. This way, the
3051 gain factors will smoothly approach the threshold value, but never exceed that
3055 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3056 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3057 This means that the maximum local gain factor for each frame is defined
3058 (only) by the frame's highest magnitude sample. This way, the samples can
3059 be amplified as much as possible without exceeding the maximum signal
3060 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3061 Normalizer can also take into account the frame's root mean square,
3062 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3063 determine the power of a time-varying signal. It is therefore considered
3064 that the RMS is a better approximation of the "perceived loudness" than
3065 just looking at the signal's peak magnitude. Consequently, by adjusting all
3066 frames to a constant RMS value, a uniform "perceived loudness" can be
3067 established. If a target RMS value has been specified, a frame's local gain
3068 factor is defined as the factor that would result in exactly that RMS value.
3069 Note, however, that the maximum local gain factor is still restricted by the
3070 frame's highest magnitude sample, in order to prevent clipping.
3073 Enable channels coupling. By default is enabled.
3074 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3075 amount. This means the same gain factor will be applied to all channels, i.e.
3076 the maximum possible gain factor is determined by the "loudest" channel.
3077 However, in some recordings, it may happen that the volume of the different
3078 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3079 In this case, this option can be used to disable the channel coupling. This way,
3080 the gain factor will be determined independently for each channel, depending
3081 only on the individual channel's highest magnitude sample. This allows for
3082 harmonizing the volume of the different channels.
3085 Enable DC bias correction. By default is disabled.
3086 An audio signal (in the time domain) is a sequence of sample values.
3087 In the Dynamic Audio Normalizer these sample values are represented in the
3088 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3089 audio signal, or "waveform", should be centered around the zero point.
3090 That means if we calculate the mean value of all samples in a file, or in a
3091 single frame, then the result should be 0.0 or at least very close to that
3092 value. If, however, there is a significant deviation of the mean value from
3093 0.0, in either positive or negative direction, this is referred to as a
3094 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3095 Audio Normalizer provides optional DC bias correction.
3096 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3097 the mean value, or "DC correction" offset, of each input frame and subtract
3098 that value from all of the frame's sample values which ensures those samples
3099 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3100 boundaries, the DC correction offset values will be interpolated smoothly
3101 between neighbouring frames.
3104 Enable alternative boundary mode. By default is disabled.
3105 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3106 around each frame. This includes the preceding frames as well as the
3107 subsequent frames. However, for the "boundary" frames, located at the very
3108 beginning and at the very end of the audio file, not all neighbouring
3109 frames are available. In particular, for the first few frames in the audio
3110 file, the preceding frames are not known. And, similarly, for the last few
3111 frames in the audio file, the subsequent frames are not known. Thus, the
3112 question arises which gain factors should be assumed for the missing frames
3113 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3114 to deal with this situation. The default boundary mode assumes a gain factor
3115 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3116 "fade out" at the beginning and at the end of the input, respectively.
3119 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3120 By default, the Dynamic Audio Normalizer does not apply "traditional"
3121 compression. This means that signal peaks will not be pruned and thus the
3122 full dynamic range will be retained within each local neighbourhood. However,
3123 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3124 normalization algorithm with a more "traditional" compression.
3125 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3126 (thresholding) function. If (and only if) the compression feature is enabled,
3127 all input frames will be processed by a soft knee thresholding function prior
3128 to the actual normalization process. Put simply, the thresholding function is
3129 going to prune all samples whose magnitude exceeds a certain threshold value.
3130 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3131 value. Instead, the threshold value will be adjusted for each individual
3133 In general, smaller parameters result in stronger compression, and vice versa.
3134 Values below 3.0 are not recommended, because audible distortion may appear.
3139 Make audio easier to listen to on headphones.
3141 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3142 so that when listened to on headphones the stereo image is moved from
3143 inside your head (standard for headphones) to outside and in front of
3144 the listener (standard for speakers).
3150 Apply a two-pole peaking equalisation (EQ) filter. With this
3151 filter, the signal-level at and around a selected frequency can
3152 be increased or decreased, whilst (unlike bandpass and bandreject
3153 filters) that at all other frequencies is unchanged.
3155 In order to produce complex equalisation curves, this filter can
3156 be given several times, each with a different central frequency.
3158 The filter accepts the following options:
3162 Set the filter's central frequency in Hz.
3165 Set method to specify band-width of filter.
3180 Specify the band-width of a filter in width_type units.
3183 Set the required gain or attenuation in dB.
3184 Beware of clipping when using a positive gain.
3187 Specify which channels to filter, by default all available are filtered.
3190 @subsection Examples
3193 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3195 equalizer=f=1000:t=h:width=200:g=-10
3199 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3201 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3205 @subsection Commands
3207 This filter supports the following commands:
3210 Change equalizer frequency.
3211 Syntax for the command is : "@var{frequency}"
3214 Change equalizer width_type.
3215 Syntax for the command is : "@var{width_type}"
3218 Change equalizer width.
3219 Syntax for the command is : "@var{width}"
3222 Change equalizer gain.
3223 Syntax for the command is : "@var{gain}"
3226 @section extrastereo
3228 Linearly increases the difference between left and right channels which
3229 adds some sort of "live" effect to playback.
3231 The filter accepts the following options:
3235 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3236 (average of both channels), with 1.0 sound will be unchanged, with
3237 -1.0 left and right channels will be swapped.
3240 Enable clipping. By default is enabled.
3243 @section firequalizer
3244 Apply FIR Equalization using arbitrary frequency response.
3246 The filter accepts the following option:
3250 Set gain curve equation (in dB). The expression can contain variables:
3253 the evaluated frequency
3257 channel number, set to 0 when multichannels evaluation is disabled
3259 channel id, see libavutil/channel_layout.h, set to the first channel id when
3260 multichannels evaluation is disabled
3264 channel_layout, see libavutil/channel_layout.h
3269 @item gain_interpolate(f)
3270 interpolate gain on frequency f based on gain_entry
3271 @item cubic_interpolate(f)
3272 same as gain_interpolate, but smoother
3274 This option is also available as command. Default is @code{gain_interpolate(f)}.
3277 Set gain entry for gain_interpolate function. The expression can
3281 store gain entry at frequency f with value g
3283 This option is also available as command.
3286 Set filter delay in seconds. Higher value means more accurate.
3287 Default is @code{0.01}.
3290 Set filter accuracy in Hz. Lower value means more accurate.
3291 Default is @code{5}.
3294 Set window function. Acceptable values are:
3297 rectangular window, useful when gain curve is already smooth
3299 hann window (default)
3305 3-terms continuous 1st derivative nuttall window
3307 minimum 3-terms discontinuous nuttall window
3309 4-terms continuous 1st derivative nuttall window
3311 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3313 blackman-harris window
3319 If enabled, use fixed number of audio samples. This improves speed when
3320 filtering with large delay. Default is disabled.
3323 Enable multichannels evaluation on gain. Default is disabled.
3326 Enable zero phase mode by subtracting timestamp to compensate delay.
3327 Default is disabled.
3330 Set scale used by gain. Acceptable values are:
3333 linear frequency, linear gain
3335 linear frequency, logarithmic (in dB) gain (default)
3337 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3339 logarithmic frequency, logarithmic gain
3343 Set file for dumping, suitable for gnuplot.
3346 Set scale for dumpfile. Acceptable values are same with scale option.
3350 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3351 Default is disabled.
3354 Enable minimum phase impulse response. Default is disabled.
3357 @subsection Examples
3362 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3365 lowpass at 1000 Hz with gain_entry:
3367 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3370 custom equalization:
3372 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3375 higher delay with zero phase to compensate delay:
3377 firequalizer=delay=0.1:fixed=on:zero_phase=on
3380 lowpass on left channel, highpass on right channel:
3382 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3383 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3388 Apply a flanging effect to the audio.
3390 The filter accepts the following options:
3394 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3397 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3400 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3404 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3405 Default value is 71.
3408 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3411 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3412 Default value is @var{sinusoidal}.
3415 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3416 Default value is 25.
3419 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3420 Default is @var{linear}.
3424 Apply Haas effect to audio.
3426 Note that this makes most sense to apply on mono signals.
3427 With this filter applied to mono signals it give some directionality and
3428 stretches its stereo image.
3430 The filter accepts the following options:
3434 Set input level. By default is @var{1}, or 0dB
3437 Set output level. By default is @var{1}, or 0dB.
3440 Set gain applied to side part of signal. By default is @var{1}.
3443 Set kind of middle source. Can be one of the following:
3453 Pick middle part signal of stereo image.
3456 Pick side part signal of stereo image.
3460 Change middle phase. By default is disabled.
3463 Set left channel delay. By default is @var{2.05} milliseconds.
3466 Set left channel balance. By default is @var{-1}.
3469 Set left channel gain. By default is @var{1}.
3472 Change left phase. By default is disabled.
3475 Set right channel delay. By defaults is @var{2.12} milliseconds.
3478 Set right channel balance. By default is @var{1}.
3481 Set right channel gain. By default is @var{1}.
3484 Change right phase. By default is enabled.
3489 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3490 embedded HDCD codes is expanded into a 20-bit PCM stream.
3492 The filter supports the Peak Extend and Low-level Gain Adjustment features
3493 of HDCD, and detects the Transient Filter flag.
3496 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3499 When using the filter with wav, note the default encoding for wav is 16-bit,
3500 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3501 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3503 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3504 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3507 The filter accepts the following options:
3510 @item disable_autoconvert
3511 Disable any automatic format conversion or resampling in the filter graph.
3513 @item process_stereo
3514 Process the stereo channels together. If target_gain does not match between
3515 channels, consider it invalid and use the last valid target_gain.
3518 Set the code detect timer period in ms.
3521 Always extend peaks above -3dBFS even if PE isn't signaled.
3524 Replace audio with a solid tone and adjust the amplitude to signal some
3525 specific aspect of the decoding process. The output file can be loaded in
3526 an audio editor alongside the original to aid analysis.
3528 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3535 Gain adjustment level at each sample
3537 Samples where peak extend occurs
3539 Samples where the code detect timer is active
3541 Samples where the target gain does not match between channels
3547 Apply head-related transfer functions (HRTFs) to create virtual
3548 loudspeakers around the user for binaural listening via headphones.
3549 The HRIRs are provided via additional streams, for each channel
3550 one stereo input stream is needed.
3552 The filter accepts the following options:
3556 Set mapping of input streams for convolution.
3557 The argument is a '|'-separated list of channel names in order as they
3558 are given as additional stream inputs for filter.
3559 This also specify number of input streams. Number of input streams
3560 must be not less than number of channels in first stream plus one.
3563 Set gain applied to audio. Value is in dB. Default is 0.
3566 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3567 processing audio in time domain which is slow.
3568 @var{freq} is processing audio in frequency domain which is fast.
3569 Default is @var{freq}.
3572 Set custom gain for LFE channels. Value is in dB. Default is 0.
3575 Set size of frame in number of samples which will be processed at once.
3576 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3579 Set format of hrir stream.
3580 Default value is @var{stereo}. Alternative value is @var{multich}.
3581 If value is set to @var{stereo}, number of additional streams should
3582 be greater or equal to number of input channels in first input stream.
3583 Also each additional stream should have stereo number of channels.
3584 If value is set to @var{multich}, number of additional streams should
3585 be exactly one. Also number of input channels of additional stream
3586 should be equal or greater than twice number of channels of first input
3590 @subsection Examples
3594 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3595 each amovie filter use stereo file with IR coefficients as input.
3596 The files give coefficients for each position of virtual loudspeaker:
3599 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3604 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3605 but now in @var{multich} @var{hrir} format.
3607 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3614 Apply a high-pass filter with 3dB point frequency.
3615 The filter can be either single-pole, or double-pole (the default).
3616 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3618 The filter accepts the following options:
3622 Set frequency in Hz. Default is 3000.
3625 Set number of poles. Default is 2.
3628 Set method to specify band-width of filter.
3643 Specify the band-width of a filter in width_type units.
3644 Applies only to double-pole filter.
3645 The default is 0.707q and gives a Butterworth response.
3648 Specify which channels to filter, by default all available are filtered.
3651 @subsection Commands
3653 This filter supports the following commands:
3656 Change highpass frequency.
3657 Syntax for the command is : "@var{frequency}"
3660 Change highpass width_type.
3661 Syntax for the command is : "@var{width_type}"
3664 Change highpass width.
3665 Syntax for the command is : "@var{width}"
3670 Join multiple input streams into one multi-channel stream.
3672 It accepts the following parameters:
3676 The number of input streams. It defaults to 2.
3678 @item channel_layout
3679 The desired output channel layout. It defaults to stereo.
3682 Map channels from inputs to output. The argument is a '|'-separated list of
3683 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3684 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3685 can be either the name of the input channel (e.g. FL for front left) or its
3686 index in the specified input stream. @var{out_channel} is the name of the output
3690 The filter will attempt to guess the mappings when they are not specified
3691 explicitly. It does so by first trying to find an unused matching input channel
3692 and if that fails it picks the first unused input channel.
3694 Join 3 inputs (with properly set channel layouts):
3696 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3699 Build a 5.1 output from 6 single-channel streams:
3701 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3702 'join=inputs=6:channel_layout=5.1:map=0.0-FL|1.0-FR|2.0-FC|3.0-SL|4.0-SR|5.0-LFE'
3708 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3710 To enable compilation of this filter you need to configure FFmpeg with
3711 @code{--enable-ladspa}.
3715 Specifies the name of LADSPA plugin library to load. If the environment
3716 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3717 each one of the directories specified by the colon separated list in
3718 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3719 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3720 @file{/usr/lib/ladspa/}.
3723 Specifies the plugin within the library. Some libraries contain only
3724 one plugin, but others contain many of them. If this is not set filter
3725 will list all available plugins within the specified library.
3728 Set the '|' separated list of controls which are zero or more floating point
3729 values that determine the behavior of the loaded plugin (for example delay,
3731 Controls need to be defined using the following syntax:
3732 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3733 @var{valuei} is the value set on the @var{i}-th control.
3734 Alternatively they can be also defined using the following syntax:
3735 @var{value0}|@var{value1}|@var{value2}|..., where
3736 @var{valuei} is the value set on the @var{i}-th control.
3737 If @option{controls} is set to @code{help}, all available controls and
3738 their valid ranges are printed.
3740 @item sample_rate, s
3741 Specify the sample rate, default to 44100. Only used if plugin have
3745 Set the number of samples per channel per each output frame, default
3746 is 1024. Only used if plugin have zero inputs.
3749 Set the minimum duration of the sourced audio. See
3750 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3751 for the accepted syntax.
3752 Note that the resulting duration may be greater than the specified duration,
3753 as the generated audio is always cut at the end of a complete frame.
3754 If not specified, or the expressed duration is negative, the audio is
3755 supposed to be generated forever.
3756 Only used if plugin have zero inputs.
3760 @subsection Examples
3764 List all available plugins within amp (LADSPA example plugin) library:
3770 List all available controls and their valid ranges for @code{vcf_notch}
3771 plugin from @code{VCF} library:
3773 ladspa=f=vcf:p=vcf_notch:c=help
3777 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3780 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3784 Add reverberation to the audio using TAP-plugins
3785 (Tom's Audio Processing plugins):
3787 ladspa=file=tap_reverb:tap_reverb
3791 Generate white noise, with 0.2 amplitude:
3793 ladspa=file=cmt:noise_source_white:c=c0=.2
3797 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3798 @code{C* Audio Plugin Suite} (CAPS) library:
3800 ladspa=file=caps:Click:c=c1=20'
3804 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3806 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3810 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3811 @code{SWH Plugins} collection:
3813 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3817 Attenuate low frequencies using Multiband EQ from Steve Harris
3818 @code{SWH Plugins} collection:
3820 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3824 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3827 ladspa=caps:Narrower
3831 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3833 ladspa=caps:White:.2
3837 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3839 ladspa=caps:Fractal:c=c1=1
3843 Dynamic volume normalization using @code{VLevel} plugin:
3845 ladspa=vlevel-ladspa:vlevel_mono
3849 @subsection Commands
3851 This filter supports the following commands:
3854 Modify the @var{N}-th control value.
3856 If the specified value is not valid, it is ignored and prior one is kept.
3861 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3862 Support for both single pass (livestreams, files) and double pass (files) modes.
3863 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3864 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3865 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3867 The filter accepts the following options:
3871 Set integrated loudness target.
3872 Range is -70.0 - -5.0. Default value is -24.0.
3875 Set loudness range target.
3876 Range is 1.0 - 20.0. Default value is 7.0.
3879 Set maximum true peak.
3880 Range is -9.0 - +0.0. Default value is -2.0.
3882 @item measured_I, measured_i
3883 Measured IL of input file.
3884 Range is -99.0 - +0.0.
3886 @item measured_LRA, measured_lra
3887 Measured LRA of input file.
3888 Range is 0.0 - 99.0.
3890 @item measured_TP, measured_tp
3891 Measured true peak of input file.
3892 Range is -99.0 - +99.0.
3894 @item measured_thresh
3895 Measured threshold of input file.
3896 Range is -99.0 - +0.0.
3899 Set offset gain. Gain is applied before the true-peak limiter.
3900 Range is -99.0 - +99.0. Default is +0.0.
3903 Normalize linearly if possible.
3904 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3905 to be specified in order to use this mode.
3906 Options are true or false. Default is true.
3909 Treat mono input files as "dual-mono". If a mono file is intended for playback
3910 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3911 If set to @code{true}, this option will compensate for this effect.
3912 Multi-channel input files are not affected by this option.
3913 Options are true or false. Default is false.
3916 Set print format for stats. Options are summary, json, or none.
3917 Default value is none.
3922 Apply a low-pass filter with 3dB point frequency.
3923 The filter can be either single-pole or double-pole (the default).
3924 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3926 The filter accepts the following options:
3930 Set frequency in Hz. Default is 500.
3933 Set number of poles. Default is 2.
3936 Set method to specify band-width of filter.
3951 Specify the band-width of a filter in width_type units.
3952 Applies only to double-pole filter.
3953 The default is 0.707q and gives a Butterworth response.
3956 Specify which channels to filter, by default all available are filtered.
3959 @subsection Examples
3962 Lowpass only LFE channel, it LFE is not present it does nothing:
3968 @subsection Commands
3970 This filter supports the following commands:
3973 Change lowpass frequency.
3974 Syntax for the command is : "@var{frequency}"
3977 Change lowpass width_type.
3978 Syntax for the command is : "@var{width_type}"
3981 Change lowpass width.
3982 Syntax for the command is : "@var{width}"
3987 Load a LV2 (LADSPA Version 2) plugin.
3989 To enable compilation of this filter you need to configure FFmpeg with
3990 @code{--enable-lv2}.
3994 Specifies the plugin URI. You may need to escape ':'.
3997 Set the '|' separated list of controls which are zero or more floating point
3998 values that determine the behavior of the loaded plugin (for example delay,
4000 If @option{controls} is set to @code{help}, all available controls and
4001 their valid ranges are printed.
4003 @item sample_rate, s
4004 Specify the sample rate, default to 44100. Only used if plugin have
4008 Set the number of samples per channel per each output frame, default
4009 is 1024. Only used if plugin have zero inputs.
4012 Set the minimum duration of the sourced audio. See
4013 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4014 for the accepted syntax.
4015 Note that the resulting duration may be greater than the specified duration,
4016 as the generated audio is always cut at the end of a complete frame.
4017 If not specified, or the expressed duration is negative, the audio is
4018 supposed to be generated forever.
4019 Only used if plugin have zero inputs.
4022 @subsection Examples
4026 Apply bass enhancer plugin from Calf:
4028 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4032 Apply vinyl plugin from Calf:
4034 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4038 Apply bit crusher plugin from ArtyFX:
4040 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4045 Multiband Compress or expand the audio's dynamic range.
4047 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4048 This is akin to the crossover of a loudspeaker, and results in flat frequency
4049 response when absent compander action.
4051 It accepts the following parameters:
4055 This option syntax is:
4056 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4057 For explanation of each item refer to compand filter documentation.
4063 Mix channels with specific gain levels. The filter accepts the output
4064 channel layout followed by a set of channels definitions.
4066 This filter is also designed to efficiently remap the channels of an audio
4069 The filter accepts parameters of the form:
4070 "@var{l}|@var{outdef}|@var{outdef}|..."
4074 output channel layout or number of channels
4077 output channel specification, of the form:
4078 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4081 output channel to define, either a channel name (FL, FR, etc.) or a channel
4082 number (c0, c1, etc.)
4085 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4088 input channel to use, see out_name for details; it is not possible to mix
4089 named and numbered input channels
4092 If the `=' in a channel specification is replaced by `<', then the gains for
4093 that specification will be renormalized so that the total is 1, thus
4094 avoiding clipping noise.
4096 @subsection Mixing examples
4098 For example, if you want to down-mix from stereo to mono, but with a bigger
4099 factor for the left channel:
4101 pan=1c|c0=0.9*c0+0.1*c1
4104 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4105 7-channels surround:
4107 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4110 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4111 that should be preferred (see "-ac" option) unless you have very specific
4114 @subsection Remapping examples
4116 The channel remapping will be effective if, and only if:
4119 @item gain coefficients are zeroes or ones,
4120 @item only one input per channel output,
4123 If all these conditions are satisfied, the filter will notify the user ("Pure
4124 channel mapping detected"), and use an optimized and lossless method to do the
4127 For example, if you have a 5.1 source and want a stereo audio stream by
4128 dropping the extra channels:
4130 pan="stereo| c0=FL | c1=FR"
4133 Given the same source, you can also switch front left and front right channels
4134 and keep the input channel layout:
4136 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4139 If the input is a stereo audio stream, you can mute the front left channel (and
4140 still keep the stereo channel layout) with:
4145 Still with a stereo audio stream input, you can copy the right channel in both
4146 front left and right:
4148 pan="stereo| c0=FR | c1=FR"
4153 ReplayGain scanner filter. This filter takes an audio stream as an input and
4154 outputs it unchanged.
4155 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4159 Convert the audio sample format, sample rate and channel layout. It is
4160 not meant to be used directly.
4163 Apply time-stretching and pitch-shifting with librubberband.
4165 To enable compilation of this filter, you need to configure FFmpeg with
4166 @code{--enable-librubberband}.
4168 The filter accepts the following options:
4172 Set tempo scale factor.
4175 Set pitch scale factor.
4178 Set transients detector.
4179 Possible values are:
4188 Possible values are:
4197 Possible values are:
4204 Set processing window size.
4205 Possible values are:
4214 Possible values are:
4221 Enable formant preservation when shift pitching.
4222 Possible values are:
4230 Possible values are:
4239 Possible values are:
4246 @section sidechaincompress
4248 This filter acts like normal compressor but has the ability to compress
4249 detected signal using second input signal.
4250 It needs two input streams and returns one output stream.
4251 First input stream will be processed depending on second stream signal.
4252 The filtered signal then can be filtered with other filters in later stages of
4253 processing. See @ref{pan} and @ref{amerge} filter.
4255 The filter accepts the following options:
4259 Set input gain. Default is 1. Range is between 0.015625 and 64.
4262 Set mode of compressor operation. Can be @code{upward} or @code{downward}.
4263 Default is @code{downward}.
4266 If a signal of second stream raises above this level it will affect the gain
4267 reduction of first stream.
4268 By default is 0.125. Range is between 0.00097563 and 1.
4271 Set a ratio about which the signal is reduced. 1:2 means that if the level
4272 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4273 Default is 2. Range is between 1 and 20.
4276 Amount of milliseconds the signal has to rise above the threshold before gain
4277 reduction starts. Default is 20. Range is between 0.01 and 2000.
4280 Amount of milliseconds the signal has to fall below the threshold before
4281 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4284 Set the amount by how much signal will be amplified after processing.
4285 Default is 1. Range is from 1 to 64.
4288 Curve the sharp knee around the threshold to enter gain reduction more softly.
4289 Default is 2.82843. Range is between 1 and 8.
4292 Choose if the @code{average} level between all channels of side-chain stream
4293 or the louder(@code{maximum}) channel of side-chain stream affects the
4294 reduction. Default is @code{average}.
4297 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4298 of @code{rms}. Default is @code{rms} which is mainly smoother.
4301 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4304 How much to use compressed signal in output. Default is 1.
4305 Range is between 0 and 1.
4308 @subsection Examples
4312 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4313 depending on the signal of 2nd input and later compressed signal to be
4314 merged with 2nd input:
4316 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4320 @section sidechaingate
4322 A sidechain gate acts like a normal (wideband) gate but has the ability to
4323 filter the detected signal before sending it to the gain reduction stage.
4324 Normally a gate uses the full range signal to detect a level above the
4326 For example: If you cut all lower frequencies from your sidechain signal
4327 the gate will decrease the volume of your track only if not enough highs
4328 appear. With this technique you are able to reduce the resonation of a
4329 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4331 It needs two input streams and returns one output stream.
4332 First input stream will be processed depending on second stream signal.
4334 The filter accepts the following options:
4338 Set input level before filtering.
4339 Default is 1. Allowed range is from 0.015625 to 64.
4342 Set the mode of operation. Can be @code{upward} or @code{downward}.
4343 Default is @code{downward}. If set to @code{upward} mode, higher parts of signal
4344 will be amplified, expanding dynamic range in upward direction.
4345 Otherwise, in case of @code{downward} lower parts of signal will be reduced.
4348 Set the level of gain reduction when the signal is below the threshold.
4349 Default is 0.06125. Allowed range is from 0 to 1.
4350 Setting this to 0 disables reduction and then filter behaves like expander.
4353 If a signal rises above this level the gain reduction is released.
4354 Default is 0.125. Allowed range is from 0 to 1.
4357 Set a ratio about which the signal is reduced.
4358 Default is 2. Allowed range is from 1 to 9000.
4361 Amount of milliseconds the signal has to rise above the threshold before gain
4363 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4366 Amount of milliseconds the signal has to fall below the threshold before the
4367 reduction is increased again. Default is 250 milliseconds.
4368 Allowed range is from 0.01 to 9000.
4371 Set amount of amplification of signal after processing.
4372 Default is 1. Allowed range is from 1 to 64.
4375 Curve the sharp knee around the threshold to enter gain reduction more softly.
4376 Default is 2.828427125. Allowed range is from 1 to 8.
4379 Choose if exact signal should be taken for detection or an RMS like one.
4380 Default is rms. Can be peak or rms.
4383 Choose if the average level between all channels or the louder channel affects
4385 Default is average. Can be average or maximum.
4388 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4391 @section silencedetect
4393 Detect silence in an audio stream.
4395 This filter logs a message when it detects that the input audio volume is less
4396 or equal to a noise tolerance value for a duration greater or equal to the
4397 minimum detected noise duration.
4399 The printed times and duration are expressed in seconds.
4401 The filter accepts the following options:
4405 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4406 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4409 Set silence duration until notification (default is 2 seconds).
4412 Process each channel separately, instead of combined. By default is disabled.
4415 @subsection Examples
4419 Detect 5 seconds of silence with -50dB noise tolerance:
4421 silencedetect=n=-50dB:d=5
4425 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4426 tolerance in @file{silence.mp3}:
4428 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4432 @section silenceremove
4434 Remove silence from the beginning, middle or end of the audio.
4436 The filter accepts the following options:
4440 This value is used to indicate if audio should be trimmed at beginning of
4441 the audio. A value of zero indicates no silence should be trimmed from the
4442 beginning. When specifying a non-zero value, it trims audio up until it
4443 finds non-silence. Normally, when trimming silence from beginning of audio
4444 the @var{start_periods} will be @code{1} but it can be increased to higher
4445 values to trim all audio up to specific count of non-silence periods.
4446 Default value is @code{0}.
4448 @item start_duration
4449 Specify the amount of time that non-silence must be detected before it stops
4450 trimming audio. By increasing the duration, bursts of noises can be treated
4451 as silence and trimmed off. Default value is @code{0}.
4453 @item start_threshold
4454 This indicates what sample value should be treated as silence. For digital
4455 audio, a value of @code{0} may be fine but for audio recorded from analog,
4456 you may wish to increase the value to account for background noise.
4457 Can be specified in dB (in case "dB" is appended to the specified value)
4458 or amplitude ratio. Default value is @code{0}.
4461 Specify max duration of silence at beginning that will be kept after
4462 trimming. Default is 0, which is equal to trimming all samples detected
4466 Specify mode of detection of silence end in start of multi-channel audio.
4467 Can be @var{any} or @var{all}. Default is @var{any}.
4468 With @var{any}, any sample that is detected as non-silence will cause
4469 stopped trimming of silence.
4470 With @var{all}, only if all channels are detected as non-silence will cause
4471 stopped trimming of silence.
4474 Set the count for trimming silence from the end of audio.
4475 To remove silence from the middle of a file, specify a @var{stop_periods}
4476 that is negative. This value is then treated as a positive value and is
4477 used to indicate the effect should restart processing as specified by
4478 @var{start_periods}, making it suitable for removing periods of silence
4479 in the middle of the audio.
4480 Default value is @code{0}.
4483 Specify a duration of silence that must exist before audio is not copied any
4484 more. By specifying a higher duration, silence that is wanted can be left in
4486 Default value is @code{0}.
4488 @item stop_threshold
4489 This is the same as @option{start_threshold} but for trimming silence from
4491 Can be specified in dB (in case "dB" is appended to the specified value)
4492 or amplitude ratio. Default value is @code{0}.
4495 Specify max duration of silence at end that will be kept after
4496 trimming. Default is 0, which is equal to trimming all samples detected
4500 Specify mode of detection of silence start in end of multi-channel audio.
4501 Can be @var{any} or @var{all}. Default is @var{any}.
4502 With @var{any}, any sample that is detected as non-silence will cause
4503 stopped trimming of silence.
4504 With @var{all}, only if all channels are detected as non-silence will cause
4505 stopped trimming of silence.
4508 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4509 and works better with digital silence which is exactly 0.
4510 Default value is @code{rms}.
4513 Set duration in number of seconds used to calculate size of window in number
4514 of samples for detecting silence.
4515 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4518 @subsection Examples
4522 The following example shows how this filter can be used to start a recording
4523 that does not contain the delay at the start which usually occurs between
4524 pressing the record button and the start of the performance:
4526 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4530 Trim all silence encountered from beginning to end where there is more than 1
4531 second of silence in audio:
4533 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4539 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4540 loudspeakers around the user for binaural listening via headphones (audio
4541 formats up to 9 channels supported).
4542 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4543 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4544 Austrian Academy of Sciences.
4546 To enable compilation of this filter you need to configure FFmpeg with
4547 @code{--enable-libmysofa}.
4549 The filter accepts the following options:
4553 Set the SOFA file used for rendering.
4556 Set gain applied to audio. Value is in dB. Default is 0.
4559 Set rotation of virtual loudspeakers in deg. Default is 0.
4562 Set elevation of virtual speakers in deg. Default is 0.
4565 Set distance in meters between loudspeakers and the listener with near-field
4566 HRTFs. Default is 1.
4569 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4570 processing audio in time domain which is slow.
4571 @var{freq} is processing audio in frequency domain which is fast.
4572 Default is @var{freq}.
4575 Set custom positions of virtual loudspeakers. Syntax for this option is:
4576 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4577 Each virtual loudspeaker is described with short channel name following with
4578 azimuth and elevation in degrees.
4579 Each virtual loudspeaker description is separated by '|'.
4580 For example to override front left and front right channel positions use:
4581 'speakers=FL 45 15|FR 345 15'.
4582 Descriptions with unrecognised channel names are ignored.
4585 Set custom gain for LFE channels. Value is in dB. Default is 0.
4588 Set custom frame size in number of samples. Default is 1024.
4589 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4590 is set to @var{freq}.
4593 Should all IRs be normalized upon importing SOFA file.
4594 By default is enabled.
4597 Should nearest IRs be interpolated with neighbor IRs if exact position
4598 does not match. By default is disabled.
4601 Minphase all IRs upon loading of SOFA file. By default is disabled.
4604 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4607 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4610 @subsection Examples
4614 Using ClubFritz6 sofa file:
4616 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4620 Using ClubFritz12 sofa file and bigger radius with small rotation:
4622 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4626 Similar as above but with custom speaker positions for front left, front right, back left and back right
4627 and also with custom gain:
4629 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4633 @section stereotools
4635 This filter has some handy utilities to manage stereo signals, for converting
4636 M/S stereo recordings to L/R signal while having control over the parameters
4637 or spreading the stereo image of master track.
4639 The filter accepts the following options:
4643 Set input level before filtering for both channels. Defaults is 1.
4644 Allowed range is from 0.015625 to 64.
4647 Set output level after filtering for both channels. Defaults is 1.
4648 Allowed range is from 0.015625 to 64.
4651 Set input balance between both channels. Default is 0.
4652 Allowed range is from -1 to 1.
4655 Set output balance between both channels. Default is 0.
4656 Allowed range is from -1 to 1.
4659 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4660 clipping. Disabled by default.
4663 Mute the left channel. Disabled by default.
4666 Mute the right channel. Disabled by default.
4669 Change the phase of the left channel. Disabled by default.
4672 Change the phase of the right channel. Disabled by default.
4675 Set stereo mode. Available values are:
4679 Left/Right to Left/Right, this is default.
4682 Left/Right to Mid/Side.
4685 Mid/Side to Left/Right.
4688 Left/Right to Left/Left.
4691 Left/Right to Right/Right.
4694 Left/Right to Left + Right.
4697 Left/Right to Right/Left.
4700 Mid/Side to Left/Left.
4703 Mid/Side to Right/Right.
4707 Set level of side signal. Default is 1.
4708 Allowed range is from 0.015625 to 64.
4711 Set balance of side signal. Default is 0.
4712 Allowed range is from -1 to 1.
4715 Set level of the middle signal. Default is 1.
4716 Allowed range is from 0.015625 to 64.
4719 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4722 Set stereo base between mono and inversed channels. Default is 0.
4723 Allowed range is from -1 to 1.
4726 Set delay in milliseconds how much to delay left from right channel and
4727 vice versa. Default is 0. Allowed range is from -20 to 20.
4730 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4733 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4735 @item bmode_in, bmode_out
4736 Set balance mode for balance_in/balance_out option.
4738 Can be one of the following:
4742 Classic balance mode. Attenuate one channel at time.
4743 Gain is raised up to 1.
4746 Similar as classic mode above but gain is raised up to 2.
4749 Equal power distribution, from -6dB to +6dB range.
4753 @subsection Examples
4757 Apply karaoke like effect:
4759 stereotools=mlev=0.015625
4763 Convert M/S signal to L/R:
4765 "stereotools=mode=ms>lr"
4769 @section stereowiden
4771 This filter enhance the stereo effect by suppressing signal common to both
4772 channels and by delaying the signal of left into right and vice versa,
4773 thereby widening the stereo effect.
4775 The filter accepts the following options:
4779 Time in milliseconds of the delay of left signal into right and vice versa.
4780 Default is 20 milliseconds.
4783 Amount of gain in delayed signal into right and vice versa. Gives a delay
4784 effect of left signal in right output and vice versa which gives widening
4785 effect. Default is 0.3.
4788 Cross feed of left into right with inverted phase. This helps in suppressing
4789 the mono. If the value is 1 it will cancel all the signal common to both
4790 channels. Default is 0.3.
4793 Set level of input signal of original channel. Default is 0.8.
4796 @section superequalizer
4797 Apply 18 band equalizer.
4799 The filter accepts the following options:
4806 Set 131Hz band gain.
4808 Set 185Hz band gain.
4810 Set 262Hz band gain.
4812 Set 370Hz band gain.
4814 Set 523Hz band gain.
4816 Set 740Hz band gain.
4818 Set 1047Hz band gain.
4820 Set 1480Hz band gain.
4822 Set 2093Hz band gain.
4824 Set 2960Hz band gain.
4826 Set 4186Hz band gain.
4828 Set 5920Hz band gain.
4830 Set 8372Hz band gain.
4832 Set 11840Hz band gain.
4834 Set 16744Hz band gain.
4836 Set 20000Hz band gain.
4840 Apply audio surround upmix filter.
4842 This filter allows to produce multichannel output from audio stream.
4844 The filter accepts the following options:
4848 Set output channel layout. By default, this is @var{5.1}.
4850 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4851 for the required syntax.
4854 Set input channel layout. By default, this is @var{stereo}.
4856 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4857 for the required syntax.
4860 Set input volume level. By default, this is @var{1}.
4863 Set output volume level. By default, this is @var{1}.
4866 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4869 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4872 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4875 Set LFE mode, can be @var{add} or @var{sub}. Default is @var{add}.
4876 In @var{add} mode, LFE channel is created from input audio and added to output.
4877 In @var{sub} mode, LFE channel is created from input audio and added to output but
4878 also all non-LFE output channels are subtracted with output LFE channel.
4881 Set front center input volume. By default, this is @var{1}.
4884 Set front center output volume. By default, this is @var{1}.
4887 Set LFE input volume. By default, this is @var{1}.
4890 Set LFE output volume. By default, this is @var{1}.
4893 Set spread usage of stereo image across X axis for all channels.
4896 Set spread usage of stereo image across Y axis for all channels.
4898 @item fcx, flx, frx, blx, brx, slx, srx, bcx
4899 Set spread usage of stereo image across X axis for each channel.
4901 @item fcy, fly, fry, bly, bry, sly, sry, bcy
4902 Set spread usage of stereo image across Y axis for each channel.
4905 Set window size. Allowed range is from @var{1024} to @var{65536}. Default size is @var{4096}.
4908 Set window function.
4910 It accepts the following values:
4933 Default is @code{hann}.
4936 Set window overlap. If set to 1, the recommended overlap for selected
4937 window function will be picked. Default is @code{0.5}.
4940 @section treble, highshelf
4942 Boost or cut treble (upper) frequencies of the audio using a two-pole
4943 shelving filter with a response similar to that of a standard
4944 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4946 The filter accepts the following options:
4950 Give the gain at whichever is the lower of ~22 kHz and the
4951 Nyquist frequency. Its useful range is about -20 (for a large cut)
4952 to +20 (for a large boost). Beware of clipping when using a positive gain.
4955 Set the filter's central frequency and so can be used
4956 to extend or reduce the frequency range to be boosted or cut.
4957 The default value is @code{3000} Hz.
4960 Set method to specify band-width of filter.
4975 Determine how steep is the filter's shelf transition.
4978 Specify which channels to filter, by default all available are filtered.
4981 @subsection Commands
4983 This filter supports the following commands:
4986 Change treble frequency.
4987 Syntax for the command is : "@var{frequency}"
4990 Change treble width_type.
4991 Syntax for the command is : "@var{width_type}"
4994 Change treble width.
4995 Syntax for the command is : "@var{width}"
4999 Syntax for the command is : "@var{gain}"
5004 Sinusoidal amplitude modulation.
5006 The filter accepts the following options:
5010 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
5011 (20 Hz or lower) will result in a tremolo effect.
5012 This filter may also be used as a ring modulator by specifying
5013 a modulation frequency higher than 20 Hz.
5014 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5017 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5018 Default value is 0.5.
5023 Sinusoidal phase modulation.
5025 The filter accepts the following options:
5029 Modulation frequency in Hertz.
5030 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
5033 Depth of modulation as a percentage. Range is 0.0 - 1.0.
5034 Default value is 0.5.
5039 Adjust the input audio volume.
5041 It accepts the following parameters:
5045 Set audio volume expression.
5047 Output values are clipped to the maximum value.
5049 The output audio volume is given by the relation:
5051 @var{output_volume} = @var{volume} * @var{input_volume}
5054 The default value for @var{volume} is "1.0".
5057 This parameter represents the mathematical precision.
5059 It determines which input sample formats will be allowed, which affects the
5060 precision of the volume scaling.
5064 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
5066 32-bit floating-point; this limits input sample format to FLT. (default)
5068 64-bit floating-point; this limits input sample format to DBL.
5072 Choose the behaviour on encountering ReplayGain side data in input frames.
5076 Remove ReplayGain side data, ignoring its contents (the default).
5079 Ignore ReplayGain side data, but leave it in the frame.
5082 Prefer the track gain, if present.
5085 Prefer the album gain, if present.
5088 @item replaygain_preamp
5089 Pre-amplification gain in dB to apply to the selected replaygain gain.
5091 Default value for @var{replaygain_preamp} is 0.0.
5094 Set when the volume expression is evaluated.
5096 It accepts the following values:
5099 only evaluate expression once during the filter initialization, or
5100 when the @samp{volume} command is sent
5103 evaluate expression for each incoming frame
5106 Default value is @samp{once}.
5109 The volume expression can contain the following parameters.
5113 frame number (starting at zero)
5116 @item nb_consumed_samples
5117 number of samples consumed by the filter
5119 number of samples in the current frame
5121 original frame position in the file
5127 PTS at start of stream
5129 time at start of stream
5135 last set volume value
5138 Note that when @option{eval} is set to @samp{once} only the
5139 @var{sample_rate} and @var{tb} variables are available, all other
5140 variables will evaluate to NAN.
5142 @subsection Commands
5144 This filter supports the following commands:
5147 Modify the volume expression.
5148 The command accepts the same syntax of the corresponding option.
5150 If the specified expression is not valid, it is kept at its current
5152 @item replaygain_noclip
5153 Prevent clipping by limiting the gain applied.
5155 Default value for @var{replaygain_noclip} is 1.
5159 @subsection Examples
5163 Halve the input audio volume:
5167 volume=volume=-6.0206dB
5170 In all the above example the named key for @option{volume} can be
5171 omitted, for example like in:
5177 Increase input audio power by 6 decibels using fixed-point precision:
5179 volume=volume=6dB:precision=fixed
5183 Fade volume after time 10 with an annihilation period of 5 seconds:
5185 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5189 @section volumedetect
5191 Detect the volume of the input video.
5193 The filter has no parameters. The input is not modified. Statistics about
5194 the volume will be printed in the log when the input stream end is reached.
5196 In particular it will show the mean volume (root mean square), maximum
5197 volume (on a per-sample basis), and the beginning of a histogram of the
5198 registered volume values (from the maximum value to a cumulated 1/1000 of
5201 All volumes are in decibels relative to the maximum PCM value.
5203 @subsection Examples
5205 Here is an excerpt of the output:
5207 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5208 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5209 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5210 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5211 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5212 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5213 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5214 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5215 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5221 The mean square energy is approximately -27 dB, or 10^-2.7.
5223 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5225 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5228 In other words, raising the volume by +4 dB does not cause any clipping,
5229 raising it by +5 dB causes clipping for 6 samples, etc.
5231 @c man end AUDIO FILTERS
5233 @chapter Audio Sources
5234 @c man begin AUDIO SOURCES
5236 Below is a description of the currently available audio sources.
5240 Buffer audio frames, and make them available to the filter chain.
5242 This source is mainly intended for a programmatic use, in particular
5243 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5245 It accepts the following parameters:
5249 The timebase which will be used for timestamps of submitted frames. It must be
5250 either a floating-point number or in @var{numerator}/@var{denominator} form.
5253 The sample rate of the incoming audio buffers.
5256 The sample format of the incoming audio buffers.
5257 Either a sample format name or its corresponding integer representation from
5258 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5260 @item channel_layout
5261 The channel layout of the incoming audio buffers.
5262 Either a channel layout name from channel_layout_map in
5263 @file{libavutil/channel_layout.c} or its corresponding integer representation
5264 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5267 The number of channels of the incoming audio buffers.
5268 If both @var{channels} and @var{channel_layout} are specified, then they
5273 @subsection Examples
5276 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5279 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5280 Since the sample format with name "s16p" corresponds to the number
5281 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5284 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5289 Generate an audio signal specified by an expression.
5291 This source accepts in input one or more expressions (one for each
5292 channel), which are evaluated and used to generate a corresponding
5295 This source accepts the following options:
5299 Set the '|'-separated expressions list for each separate channel. In case the
5300 @option{channel_layout} option is not specified, the selected channel layout
5301 depends on the number of provided expressions. Otherwise the last
5302 specified expression is applied to the remaining output channels.
5304 @item channel_layout, c
5305 Set the channel layout. The number of channels in the specified layout
5306 must be equal to the number of specified expressions.
5309 Set the minimum duration of the sourced audio. See
5310 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5311 for the accepted syntax.
5312 Note that the resulting duration may be greater than the specified
5313 duration, as the generated audio is always cut at the end of a
5316 If not specified, or the expressed duration is negative, the audio is
5317 supposed to be generated forever.
5320 Set the number of samples per channel per each output frame,
5323 @item sample_rate, s
5324 Specify the sample rate, default to 44100.
5327 Each expression in @var{exprs} can contain the following constants:
5331 number of the evaluated sample, starting from 0
5334 time of the evaluated sample expressed in seconds, starting from 0
5341 @subsection Examples
5351 Generate a sin signal with frequency of 440 Hz, set sample rate to
5354 aevalsrc="sin(440*2*PI*t):s=8000"
5358 Generate a two channels signal, specify the channel layout (Front
5359 Center + Back Center) explicitly:
5361 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5365 Generate white noise:
5367 aevalsrc="-2+random(0)"
5371 Generate an amplitude modulated signal:
5373 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5377 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5379 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5386 The null audio source, return unprocessed audio frames. It is mainly useful
5387 as a template and to be employed in analysis / debugging tools, or as
5388 the source for filters which ignore the input data (for example the sox
5391 This source accepts the following options:
5395 @item channel_layout, cl
5397 Specifies the channel layout, and can be either an integer or a string
5398 representing a channel layout. The default value of @var{channel_layout}
5401 Check the channel_layout_map definition in
5402 @file{libavutil/channel_layout.c} for the mapping between strings and
5403 channel layout values.
5405 @item sample_rate, r
5406 Specifies the sample rate, and defaults to 44100.
5409 Set the number of samples per requested frames.
5413 @subsection Examples
5417 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5419 anullsrc=r=48000:cl=4
5423 Do the same operation with a more obvious syntax:
5425 anullsrc=r=48000:cl=mono
5429 All the parameters need to be explicitly defined.
5433 Synthesize a voice utterance using the libflite library.
5435 To enable compilation of this filter you need to configure FFmpeg with
5436 @code{--enable-libflite}.
5438 Note that versions of the flite library prior to 2.0 are not thread-safe.
5440 The filter accepts the following options:
5445 If set to 1, list the names of the available voices and exit
5446 immediately. Default value is 0.
5449 Set the maximum number of samples per frame. Default value is 512.
5452 Set the filename containing the text to speak.
5455 Set the text to speak.
5458 Set the voice to use for the speech synthesis. Default value is
5459 @code{kal}. See also the @var{list_voices} option.
5462 @subsection Examples
5466 Read from file @file{speech.txt}, and synthesize the text using the
5467 standard flite voice:
5469 flite=textfile=speech.txt
5473 Read the specified text selecting the @code{slt} voice:
5475 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5479 Input text to ffmpeg:
5481 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5485 Make @file{ffplay} speak the specified text, using @code{flite} and
5486 the @code{lavfi} device:
5488 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5492 For more information about libflite, check:
5493 @url{http://www.festvox.org/flite/}
5497 Generate a noise audio signal.
5499 The filter accepts the following options:
5502 @item sample_rate, r
5503 Specify the sample rate. Default value is 48000 Hz.
5506 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5510 Specify the duration of the generated audio stream. Not specifying this option
5511 results in noise with an infinite length.
5513 @item color, colour, c
5514 Specify the color of noise. Available noise colors are white, pink, brown,
5515 blue and violet. Default color is white.
5518 Specify a value used to seed the PRNG.
5521 Set the number of samples per each output frame, default is 1024.
5524 @subsection Examples
5529 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5531 anoisesrc=d=60:c=pink:r=44100:a=0.5
5537 Generate odd-tap Hilbert transform FIR coefficients.
5539 The resulting stream can be used with @ref{afir} filter for phase-shifting
5540 the signal by 90 degrees.
5542 This is used in many matrix coding schemes and for analytic signal generation.
5543 The process is often written as a multiplication by i (or j), the imaginary unit.
5545 The filter accepts the following options:
5549 @item sample_rate, s
5550 Set sample rate, default is 44100.
5553 Set length of FIR filter, default is 22051.
5556 Set number of samples per each frame.
5559 Set window function to be used when generating FIR coefficients.
5564 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5566 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5568 The filter accepts the following options:
5571 @item sample_rate, r
5572 Set sample rate, default is 44100.
5575 Set number of samples per each frame. Default is 1024.
5578 Set high-pass frequency. Default is 0.
5581 Set low-pass frequency. Default is 0.
5582 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5583 is higher than 0 then filter will create band-pass filter coefficients,
5584 otherwise band-reject filter coefficients.
5587 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5590 Set Kaiser window beta.
5593 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5596 Enable rounding, by default is disabled.
5599 Set number of taps for high-pass filter.
5602 Set number of taps for low-pass filter.
5607 Generate an audio signal made of a sine wave with amplitude 1/8.
5609 The audio signal is bit-exact.
5611 The filter accepts the following options:
5616 Set the carrier frequency. Default is 440 Hz.
5618 @item beep_factor, b
5619 Enable a periodic beep every second with frequency @var{beep_factor} times
5620 the carrier frequency. Default is 0, meaning the beep is disabled.
5622 @item sample_rate, r
5623 Specify the sample rate, default is 44100.
5626 Specify the duration of the generated audio stream.
5628 @item samples_per_frame
5629 Set the number of samples per output frame.
5631 The expression can contain the following constants:
5635 The (sequential) number of the output audio frame, starting from 0.
5638 The PTS (Presentation TimeStamp) of the output audio frame,
5639 expressed in @var{TB} units.
5642 The PTS of the output audio frame, expressed in seconds.
5645 The timebase of the output audio frames.
5648 Default is @code{1024}.
5651 @subsection Examples
5656 Generate a simple 440 Hz sine wave:
5662 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5666 sine=frequency=220:beep_factor=4:duration=5
5670 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5673 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5677 @c man end AUDIO SOURCES
5679 @chapter Audio Sinks
5680 @c man begin AUDIO SINKS
5682 Below is a description of the currently available audio sinks.
5684 @section abuffersink
5686 Buffer audio frames, and make them available to the end of filter chain.
5688 This sink is mainly intended for programmatic use, in particular
5689 through the interface defined in @file{libavfilter/buffersink.h}
5690 or the options system.
5692 It accepts a pointer to an AVABufferSinkContext structure, which
5693 defines the incoming buffers' formats, to be passed as the opaque
5694 parameter to @code{avfilter_init_filter} for initialization.
5697 Null audio sink; do absolutely nothing with the input audio. It is
5698 mainly useful as a template and for use in analysis / debugging
5701 @c man end AUDIO SINKS
5703 @chapter Video Filters
5704 @c man begin VIDEO FILTERS
5706 When you configure your FFmpeg build, you can disable any of the
5707 existing filters using @code{--disable-filters}.
5708 The configure output will show the video filters included in your
5711 Below is a description of the currently available video filters.
5713 @section alphaextract
5715 Extract the alpha component from the input as a grayscale video. This
5716 is especially useful with the @var{alphamerge} filter.
5720 Add or replace the alpha component of the primary input with the
5721 grayscale value of a second input. This is intended for use with
5722 @var{alphaextract} to allow the transmission or storage of frame
5723 sequences that have alpha in a format that doesn't support an alpha
5726 For example, to reconstruct full frames from a normal YUV-encoded video
5727 and a separate video created with @var{alphaextract}, you might use:
5729 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5732 Since this filter is designed for reconstruction, it operates on frame
5733 sequences without considering timestamps, and terminates when either
5734 input reaches end of stream. This will cause problems if your encoding
5735 pipeline drops frames. If you're trying to apply an image as an
5736 overlay to a video stream, consider the @var{overlay} filter instead.
5740 Amplify differences between current pixel and pixels of adjacent frames in
5741 same pixel location.
5743 This filter accepts the following options:
5747 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5748 For example radius of 3 will instruct filter to calculate average of 7 frames.
5751 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5754 Set threshold for difference amplification. Any difference greater or equal to
5755 this value will not alter source pixel. Default is 10.
5756 Allowed range is from 0 to 65535.
5759 Set tolerance for difference amplification. Any difference lower to
5760 this value will not alter source pixel. Default is 0.
5761 Allowed range is from 0 to 65535.
5764 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5765 This option controls maximum possible value that will decrease source pixel value.
5768 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5769 This option controls maximum possible value that will increase source pixel value.
5772 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5777 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5778 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5779 Substation Alpha) subtitles files.
5781 This filter accepts the following option in addition to the common options from
5782 the @ref{subtitles} filter:
5786 Set the shaping engine
5788 Available values are:
5791 The default libass shaping engine, which is the best available.
5793 Fast, font-agnostic shaper that can do only substitutions
5795 Slower shaper using OpenType for substitutions and positioning
5798 The default is @code{auto}.
5802 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5804 The filter accepts the following options:
5808 Set threshold A for 1st plane. Default is 0.02.
5809 Valid range is 0 to 0.3.
5812 Set threshold B for 1st plane. Default is 0.04.
5813 Valid range is 0 to 5.
5816 Set threshold A for 2nd plane. Default is 0.02.
5817 Valid range is 0 to 0.3.
5820 Set threshold B for 2nd plane. Default is 0.04.
5821 Valid range is 0 to 5.
5824 Set threshold A for 3rd plane. Default is 0.02.
5825 Valid range is 0 to 0.3.
5828 Set threshold B for 3rd plane. Default is 0.04.
5829 Valid range is 0 to 5.
5831 Threshold A is designed to react on abrupt changes in the input signal and
5832 threshold B is designed to react on continuous changes in the input signal.
5835 Set number of frames filter will use for averaging. Default is 9. Must be odd
5836 number in range [5, 129].
5839 Set what planes of frame filter will use for averaging. Default is all.
5844 Apply average blur filter.
5846 The filter accepts the following options:
5850 Set horizontal radius size.
5853 Set which planes to filter. By default all planes are filtered.
5856 Set vertical radius size, if zero it will be same as @code{sizeX}.
5857 Default is @code{0}.
5862 Compute the bounding box for the non-black pixels in the input frame
5865 This filter computes the bounding box containing all the pixels with a
5866 luminance value greater than the minimum allowed value.
5867 The parameters describing the bounding box are printed on the filter
5870 The filter accepts the following option:
5874 Set the minimal luminance value. Default is @code{16}.
5877 @section bitplanenoise
5879 Show and measure bit plane noise.
5881 The filter accepts the following options:
5885 Set which plane to analyze. Default is @code{1}.
5888 Filter out noisy pixels from @code{bitplane} set above.
5889 Default is disabled.
5892 @section blackdetect
5894 Detect video intervals that are (almost) completely black. Can be
5895 useful to detect chapter transitions, commercials, or invalid
5896 recordings. Output lines contains the time for the start, end and
5897 duration of the detected black interval expressed in seconds.
5899 In order to display the output lines, you need to set the loglevel at
5900 least to the AV_LOG_INFO value.
5902 The filter accepts the following options:
5905 @item black_min_duration, d
5906 Set the minimum detected black duration expressed in seconds. It must
5907 be a non-negative floating point number.
5909 Default value is 2.0.
5911 @item picture_black_ratio_th, pic_th
5912 Set the threshold for considering a picture "black".
5913 Express the minimum value for the ratio:
5915 @var{nb_black_pixels} / @var{nb_pixels}
5918 for which a picture is considered black.
5919 Default value is 0.98.
5921 @item pixel_black_th, pix_th
5922 Set the threshold for considering a pixel "black".
5924 The threshold expresses the maximum pixel luminance value for which a
5925 pixel is considered "black". The provided value is scaled according to
5926 the following equation:
5928 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5931 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5932 the input video format, the range is [0-255] for YUV full-range
5933 formats and [16-235] for YUV non full-range formats.
5935 Default value is 0.10.
5938 The following example sets the maximum pixel threshold to the minimum
5939 value, and detects only black intervals of 2 or more seconds:
5941 blackdetect=d=2:pix_th=0.00
5946 Detect frames that are (almost) completely black. Can be useful to
5947 detect chapter transitions or commercials. Output lines consist of
5948 the frame number of the detected frame, the percentage of blackness,
5949 the position in the file if known or -1 and the timestamp in seconds.
5951 In order to display the output lines, you need to set the loglevel at
5952 least to the AV_LOG_INFO value.
5954 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5955 The value represents the percentage of pixels in the picture that
5956 are below the threshold value.
5958 It accepts the following parameters:
5963 The percentage of the pixels that have to be below the threshold; it defaults to
5966 @item threshold, thresh
5967 The threshold below which a pixel value is considered black; it defaults to
5972 @section blend, tblend
5974 Blend two video frames into each other.
5976 The @code{blend} filter takes two input streams and outputs one
5977 stream, the first input is the "top" layer and second input is
5978 "bottom" layer. By default, the output terminates when the longest input terminates.
5980 The @code{tblend} (time blend) filter takes two consecutive frames
5981 from one single stream, and outputs the result obtained by blending
5982 the new frame on top of the old frame.
5984 A description of the accepted options follows.
5992 Set blend mode for specific pixel component or all pixel components in case
5993 of @var{all_mode}. Default value is @code{normal}.
5995 Available values for component modes are:
6037 Set blend opacity for specific pixel component or all pixel components in case
6038 of @var{all_opacity}. Only used in combination with pixel component blend modes.
6045 Set blend expression for specific pixel component or all pixel components in case
6046 of @var{all_expr}. Note that related mode options will be ignored if those are set.
6048 The expressions can use the following variables:
6052 The sequential number of the filtered frame, starting from @code{0}.
6056 the coordinates of the current sample
6060 the width and height of currently filtered plane
6064 Width and height scale for the plane being filtered. It is the
6065 ratio between the dimensions of the current plane to the luma plane,
6066 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
6067 the luma plane and @code{0.5,0.5} for the chroma planes.
6070 Time of the current frame, expressed in seconds.
6073 Value of pixel component at current location for first video frame (top layer).
6076 Value of pixel component at current location for second video frame (bottom layer).
6080 The @code{blend} filter also supports the @ref{framesync} options.
6082 @subsection Examples
6086 Apply transition from bottom layer to top layer in first 10 seconds:
6088 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6092 Apply linear horizontal transition from top layer to bottom layer:
6094 blend=all_expr='A*(X/W)+B*(1-X/W)'
6098 Apply 1x1 checkerboard effect:
6100 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6104 Apply uncover left effect:
6106 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6110 Apply uncover down effect:
6112 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6116 Apply uncover up-left effect:
6118 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6122 Split diagonally video and shows top and bottom layer on each side:
6124 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6128 Display differences between the current and the previous frame:
6130 tblend=all_mode=grainextract
6136 Denoise frames using Block-Matching 3D algorithm.
6138 The filter accepts the following options.
6142 Set denoising strength. Default value is 1.
6143 Allowed range is from 0 to 999.9.
6144 The denoising algorithm is very sensitive to sigma, so adjust it
6145 according to the source.
6148 Set local patch size. This sets dimensions in 2D.
6151 Set sliding step for processing blocks. Default value is 4.
6152 Allowed range is from 1 to 64.
6153 Smaller values allows processing more reference blocks and is slower.
6156 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6157 When set to 1, no block matching is done. Larger values allows more blocks
6159 Allowed range is from 1 to 256.
6162 Set radius for search block matching. Default is 9.
6163 Allowed range is from 1 to INT32_MAX.
6166 Set step between two search locations for block matching. Default is 1.
6167 Allowed range is from 1 to 64. Smaller is slower.
6170 Set threshold of mean square error for block matching. Valid range is 0 to
6174 Set thresholding parameter for hard thresholding in 3D transformed domain.
6175 Larger values results in stronger hard-thresholding filtering in frequency
6179 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6180 Default is @code{basic}.
6183 If enabled, filter will use 2nd stream for block matching.
6184 Default is disabled for @code{basic} value of @var{estim} option,
6185 and always enabled if value of @var{estim} is @code{final}.
6188 Set planes to filter. Default is all available except alpha.
6191 @subsection Examples
6195 Basic filtering with bm3d:
6197 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6201 Same as above, but filtering only luma:
6203 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6207 Same as above, but with both estimation modes:
6209 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
6213 Same as above, but prefilter with @ref{nlmeans} filter instead:
6215 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
6221 Apply a boxblur algorithm to the input video.
6223 It accepts the following parameters:
6227 @item luma_radius, lr
6228 @item luma_power, lp
6229 @item chroma_radius, cr
6230 @item chroma_power, cp
6231 @item alpha_radius, ar
6232 @item alpha_power, ap
6236 A description of the accepted options follows.
6239 @item luma_radius, lr
6240 @item chroma_radius, cr
6241 @item alpha_radius, ar
6242 Set an expression for the box radius in pixels used for blurring the
6243 corresponding input plane.
6245 The radius value must be a non-negative number, and must not be
6246 greater than the value of the expression @code{min(w,h)/2} for the
6247 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6250 Default value for @option{luma_radius} is "2". If not specified,
6251 @option{chroma_radius} and @option{alpha_radius} default to the
6252 corresponding value set for @option{luma_radius}.
6254 The expressions can contain the following constants:
6258 The input width and height in pixels.
6262 The input chroma image width and height in pixels.
6266 The horizontal and vertical chroma subsample values. For example, for the
6267 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6270 @item luma_power, lp
6271 @item chroma_power, cp
6272 @item alpha_power, ap
6273 Specify how many times the boxblur filter is applied to the
6274 corresponding plane.
6276 Default value for @option{luma_power} is 2. If not specified,
6277 @option{chroma_power} and @option{alpha_power} default to the
6278 corresponding value set for @option{luma_power}.
6280 A value of 0 will disable the effect.
6283 @subsection Examples
6287 Apply a boxblur filter with the luma, chroma, and alpha radii
6290 boxblur=luma_radius=2:luma_power=1
6295 Set the luma radius to 2, and alpha and chroma radius to 0:
6297 boxblur=2:1:cr=0:ar=0
6301 Set the luma and chroma radii to a fraction of the video dimension:
6303 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6309 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6310 Deinterlacing Filter").
6312 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6313 interpolation algorithms.
6314 It accepts the following parameters:
6318 The interlacing mode to adopt. It accepts one of the following values:
6322 Output one frame for each frame.
6324 Output one frame for each field.
6327 The default value is @code{send_field}.
6330 The picture field parity assumed for the input interlaced video. It accepts one
6331 of the following values:
6335 Assume the top field is first.
6337 Assume the bottom field is first.
6339 Enable automatic detection of field parity.
6342 The default value is @code{auto}.
6343 If the interlacing is unknown or the decoder does not export this information,
6344 top field first will be assumed.
6347 Specify which frames to deinterlace. Accept one of the following
6352 Deinterlace all frames.
6354 Only deinterlace frames marked as interlaced.
6357 The default value is @code{all}.
6361 Remove all color information for all colors except for certain one.
6363 The filter accepts the following options:
6367 The color which will not be replaced with neutral chroma.
6370 Similarity percentage with the above color.
6371 0.01 matches only the exact key color, while 1.0 matches everything.
6374 Signals that the color passed is already in YUV instead of RGB.
6376 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6377 This can be used to pass exact YUV values as hexadecimal numbers.
6381 YUV colorspace color/chroma keying.
6383 The filter accepts the following options:
6387 The color which will be replaced with transparency.
6390 Similarity percentage with the key color.
6392 0.01 matches only the exact key color, while 1.0 matches everything.
6397 0.0 makes pixels either fully transparent, or not transparent at all.
6399 Higher values result in semi-transparent pixels, with a higher transparency
6400 the more similar the pixels color is to the key color.
6403 Signals that the color passed is already in YUV instead of RGB.
6405 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6406 This can be used to pass exact YUV values as hexadecimal numbers.
6409 @subsection Examples
6413 Make every green pixel in the input image transparent:
6415 ffmpeg -i input.png -vf chromakey=green out.png
6419 Overlay a greenscreen-video on top of a static black background.
6421 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
6425 @section chromashift
6426 Shift chroma pixels horizontally and/or vertically.
6428 The filter accepts the following options:
6431 Set amount to shift chroma-blue horizontally.
6433 Set amount to shift chroma-blue vertically.
6435 Set amount to shift chroma-red horizontally.
6437 Set amount to shift chroma-red vertically.
6439 Set edge mode, can be @var{smear}, default, or @var{warp}.
6444 Display CIE color diagram with pixels overlaid onto it.
6446 The filter accepts the following options:
6461 @item uhdtv, rec2020
6474 Set what gamuts to draw.
6476 See @code{system} option for available values.
6479 Set ciescope size, by default set to 512.
6482 Set intensity used to map input pixel values to CIE diagram.
6485 Set contrast used to draw tongue colors that are out of active color system gamut.
6488 Correct gamma displayed on scope, by default enabled.
6491 Show white point on CIE diagram, by default disabled.
6494 Set input gamma. Used only with XYZ input color space.
6499 Visualize information exported by some codecs.
6501 Some codecs can export information through frames using side-data or other
6502 means. For example, some MPEG based codecs export motion vectors through the
6503 @var{export_mvs} flag in the codec @option{flags2} option.
6505 The filter accepts the following option:
6509 Set motion vectors to visualize.
6511 Available flags for @var{mv} are:
6515 forward predicted MVs of P-frames
6517 forward predicted MVs of B-frames
6519 backward predicted MVs of B-frames
6523 Display quantization parameters using the chroma planes.
6526 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6528 Available flags for @var{mv_type} are:
6532 forward predicted MVs
6534 backward predicted MVs
6537 @item frame_type, ft
6538 Set frame type to visualize motion vectors of.
6540 Available flags for @var{frame_type} are:
6544 intra-coded frames (I-frames)
6546 predicted frames (P-frames)
6548 bi-directionally predicted frames (B-frames)
6552 @subsection Examples
6556 Visualize forward predicted MVs of all frames using @command{ffplay}:
6558 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6562 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6564 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6568 @section colorbalance
6569 Modify intensity of primary colors (red, green and blue) of input frames.
6571 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6572 regions for the red-cyan, green-magenta or blue-yellow balance.
6574 A positive adjustment value shifts the balance towards the primary color, a negative
6575 value towards the complementary color.
6577 The filter accepts the following options:
6583 Adjust red, green and blue shadows (darkest pixels).
6588 Adjust red, green and blue midtones (medium pixels).
6593 Adjust red, green and blue highlights (brightest pixels).
6595 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6598 @subsection Examples
6602 Add red color cast to shadows:
6609 RGB colorspace color keying.
6611 The filter accepts the following options:
6615 The color which will be replaced with transparency.
6618 Similarity percentage with the key color.
6620 0.01 matches only the exact key color, while 1.0 matches everything.
6625 0.0 makes pixels either fully transparent, or not transparent at all.
6627 Higher values result in semi-transparent pixels, with a higher transparency
6628 the more similar the pixels color is to the key color.
6631 @subsection Examples
6635 Make every green pixel in the input image transparent:
6637 ffmpeg -i input.png -vf colorkey=green out.png
6641 Overlay a greenscreen-video on top of a static background image.
6643 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
6647 @section colorlevels
6649 Adjust video input frames using levels.
6651 The filter accepts the following options:
6658 Adjust red, green, blue and alpha input black point.
6659 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6665 Adjust red, green, blue and alpha input white point.
6666 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6668 Input levels are used to lighten highlights (bright tones), darken shadows
6669 (dark tones), change the balance of bright and dark tones.
6675 Adjust red, green, blue and alpha output black point.
6676 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6682 Adjust red, green, blue and alpha output white point.
6683 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6685 Output levels allows manual selection of a constrained output level range.
6688 @subsection Examples
6692 Make video output darker:
6694 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6700 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6704 Make video output lighter:
6706 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6710 Increase brightness:
6712 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6716 @section colorchannelmixer
6718 Adjust video input frames by re-mixing color channels.
6720 This filter modifies a color channel by adding the values associated to
6721 the other channels of the same pixels. For example if the value to
6722 modify is red, the output value will be:
6724 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6727 The filter accepts the following options:
6734 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6735 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6741 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6742 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6748 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6749 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6755 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6756 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6758 Allowed ranges for options are @code{[-2.0, 2.0]}.
6761 @subsection Examples
6765 Convert source to grayscale:
6767 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6770 Simulate sepia tones:
6772 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6776 @section colormatrix
6778 Convert color matrix.
6780 The filter accepts the following options:
6785 Specify the source and destination color matrix. Both values must be
6788 The accepted values are:
6816 For example to convert from BT.601 to SMPTE-240M, use the command:
6818 colormatrix=bt601:smpte240m
6823 Convert colorspace, transfer characteristics or color primaries.
6824 Input video needs to have an even size.
6826 The filter accepts the following options:
6831 Specify all color properties at once.
6833 The accepted values are:
6863 Specify output colorspace.
6865 The accepted values are:
6874 BT.470BG or BT.601-6 625
6877 SMPTE-170M or BT.601-6 525
6886 BT.2020 with non-constant luminance
6892 Specify output transfer characteristics.
6894 The accepted values are:
6906 Constant gamma of 2.2
6909 Constant gamma of 2.8
6912 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6930 BT.2020 for 10-bits content
6933 BT.2020 for 12-bits content
6939 Specify output color primaries.
6941 The accepted values are:
6950 BT.470BG or BT.601-6 625
6953 SMPTE-170M or BT.601-6 525
6977 Specify output color range.
6979 The accepted values are:
6982 TV (restricted) range
6985 MPEG (restricted) range
6996 Specify output color format.
6998 The accepted values are:
7001 YUV 4:2:0 planar 8-bits
7004 YUV 4:2:0 planar 10-bits
7007 YUV 4:2:0 planar 12-bits
7010 YUV 4:2:2 planar 8-bits
7013 YUV 4:2:2 planar 10-bits
7016 YUV 4:2:2 planar 12-bits
7019 YUV 4:4:4 planar 8-bits
7022 YUV 4:4:4 planar 10-bits
7025 YUV 4:4:4 planar 12-bits
7030 Do a fast conversion, which skips gamma/primary correction. This will take
7031 significantly less CPU, but will be mathematically incorrect. To get output
7032 compatible with that produced by the colormatrix filter, use fast=1.
7035 Specify dithering mode.
7037 The accepted values are:
7043 Floyd-Steinberg dithering
7047 Whitepoint adaptation mode.
7049 The accepted values are:
7052 Bradford whitepoint adaptation
7055 von Kries whitepoint adaptation
7058 identity whitepoint adaptation (i.e. no whitepoint adaptation)
7062 Override all input properties at once. Same accepted values as @ref{all}.
7065 Override input colorspace. Same accepted values as @ref{space}.
7068 Override input color primaries. Same accepted values as @ref{primaries}.
7071 Override input transfer characteristics. Same accepted values as @ref{trc}.
7074 Override input color range. Same accepted values as @ref{range}.
7078 The filter converts the transfer characteristics, color space and color
7079 primaries to the specified user values. The output value, if not specified,
7080 is set to a default value based on the "all" property. If that property is
7081 also not specified, the filter will log an error. The output color range and
7082 format default to the same value as the input color range and format. The
7083 input transfer characteristics, color space, color primaries and color range
7084 should be set on the input data. If any of these are missing, the filter will
7085 log an error and no conversion will take place.
7087 For example to convert the input to SMPTE-240M, use the command:
7089 colorspace=smpte240m
7092 @section convolution
7094 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7096 The filter accepts the following options:
7103 Set matrix for each plane.
7104 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7105 and from 1 to 49 odd number of signed integers in @var{row} mode.
7111 Set multiplier for calculated value for each plane.
7112 If unset or 0, it will be sum of all matrix elements.
7118 Set bias for each plane. This value is added to the result of the multiplication.
7119 Useful for making the overall image brighter or darker. Default is 0.0.
7125 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7126 Default is @var{square}.
7129 @subsection Examples
7135 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"
7141 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"
7147 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"
7153 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"
7157 Apply laplacian edge detector which includes diagonals:
7159 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"
7165 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"
7171 Apply 2D convolution of video stream in frequency domain using second stream
7174 The filter accepts the following options:
7178 Set which planes to process.
7181 Set which impulse video frames will be processed, can be @var{first}
7182 or @var{all}. Default is @var{all}.
7185 The @code{convolve} filter also supports the @ref{framesync} options.
7189 Copy the input video source unchanged to the output. This is mainly useful for
7194 Video filtering on GPU using Apple's CoreImage API on OSX.
7196 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7197 processed by video hardware. However, software-based OpenGL implementations
7198 exist which means there is no guarantee for hardware processing. It depends on
7201 There are many filters and image generators provided by Apple that come with a
7202 large variety of options. The filter has to be referenced by its name along
7205 The coreimage filter accepts the following options:
7208 List all available filters and generators along with all their respective
7209 options as well as possible minimum and maximum values along with the default
7216 Specify all filters by their respective name and options.
7217 Use @var{list_filters} to determine all valid filter names and options.
7218 Numerical options are specified by a float value and are automatically clamped
7219 to their respective value range. Vector and color options have to be specified
7220 by a list of space separated float values. Character escaping has to be done.
7221 A special option name @code{default} is available to use default options for a
7224 It is required to specify either @code{default} or at least one of the filter options.
7225 All omitted options are used with their default values.
7226 The syntax of the filter string is as follows:
7228 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7232 Specify a rectangle where the output of the filter chain is copied into the
7233 input image. It is given by a list of space separated float values:
7235 output_rect=x\ y\ width\ height
7237 If not given, the output rectangle equals the dimensions of the input image.
7238 The output rectangle is automatically cropped at the borders of the input
7239 image. Negative values are valid for each component.
7241 output_rect=25\ 25\ 100\ 100
7245 Several filters can be chained for successive processing without GPU-HOST
7246 transfers allowing for fast processing of complex filter chains.
7247 Currently, only filters with zero (generators) or exactly one (filters) input
7248 image and one output image are supported. Also, transition filters are not yet
7251 Some filters generate output images with additional padding depending on the
7252 respective filter kernel. The padding is automatically removed to ensure the
7253 filter output has the same size as the input image.
7255 For image generators, the size of the output image is determined by the
7256 previous output image of the filter chain or the input image of the whole
7257 filterchain, respectively. The generators do not use the pixel information of
7258 this image to generate their output. However, the generated output is
7259 blended onto this image, resulting in partial or complete coverage of the
7262 The @ref{coreimagesrc} video source can be used for generating input images
7263 which are directly fed into the filter chain. By using it, providing input
7264 images by another video source or an input video is not required.
7266 @subsection Examples
7271 List all filters available:
7273 coreimage=list_filters=true
7277 Use the CIBoxBlur filter with default options to blur an image:
7279 coreimage=filter=CIBoxBlur@@default
7283 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7284 its center at 100x100 and a radius of 50 pixels:
7286 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7290 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7291 given as complete and escaped command-line for Apple's standard bash shell:
7293 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7299 Crop the input video to given dimensions.
7301 It accepts the following parameters:
7305 The width of the output video. It defaults to @code{iw}.
7306 This expression is evaluated only once during the filter
7307 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7310 The height of the output video. It defaults to @code{ih}.
7311 This expression is evaluated only once during the filter
7312 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7315 The horizontal position, in the input video, of the left edge of the output
7316 video. It defaults to @code{(in_w-out_w)/2}.
7317 This expression is evaluated per-frame.
7320 The vertical position, in the input video, of the top edge of the output video.
7321 It defaults to @code{(in_h-out_h)/2}.
7322 This expression is evaluated per-frame.
7325 If set to 1 will force the output display aspect ratio
7326 to be the same of the input, by changing the output sample aspect
7327 ratio. It defaults to 0.
7330 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7331 width/height/x/y as specified and will not be rounded to nearest smaller value.
7335 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7336 expressions containing the following constants:
7341 The computed values for @var{x} and @var{y}. They are evaluated for
7346 The input width and height.
7350 These are the same as @var{in_w} and @var{in_h}.
7354 The output (cropped) width and height.
7358 These are the same as @var{out_w} and @var{out_h}.
7361 same as @var{iw} / @var{ih}
7364 input sample aspect ratio
7367 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7371 horizontal and vertical chroma subsample values. For example for the
7372 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7375 The number of the input frame, starting from 0.
7378 the position in the file of the input frame, NAN if unknown
7381 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7385 The expression for @var{out_w} may depend on the value of @var{out_h},
7386 and the expression for @var{out_h} may depend on @var{out_w}, but they
7387 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7388 evaluated after @var{out_w} and @var{out_h}.
7390 The @var{x} and @var{y} parameters specify the expressions for the
7391 position of the top-left corner of the output (non-cropped) area. They
7392 are evaluated for each frame. If the evaluated value is not valid, it
7393 is approximated to the nearest valid value.
7395 The expression for @var{x} may depend on @var{y}, and the expression
7396 for @var{y} may depend on @var{x}.
7398 @subsection Examples
7402 Crop area with size 100x100 at position (12,34).
7407 Using named options, the example above becomes:
7409 crop=w=100:h=100:x=12:y=34
7413 Crop the central input area with size 100x100:
7419 Crop the central input area with size 2/3 of the input video:
7421 crop=2/3*in_w:2/3*in_h
7425 Crop the input video central square:
7432 Delimit the rectangle with the top-left corner placed at position
7433 100:100 and the right-bottom corner corresponding to the right-bottom
7434 corner of the input image.
7436 crop=in_w-100:in_h-100:100:100
7440 Crop 10 pixels from the left and right borders, and 20 pixels from
7441 the top and bottom borders
7443 crop=in_w-2*10:in_h-2*20
7447 Keep only the bottom right quarter of the input image:
7449 crop=in_w/2:in_h/2:in_w/2:in_h/2
7453 Crop height for getting Greek harmony:
7455 crop=in_w:1/PHI*in_w
7459 Apply trembling effect:
7461 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)
7465 Apply erratic camera effect depending on timestamp:
7467 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)"
7471 Set x depending on the value of y:
7473 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7477 @subsection Commands
7479 This filter supports the following commands:
7485 Set width/height of the output video and the horizontal/vertical position
7487 The command accepts the same syntax of the corresponding option.
7489 If the specified expression is not valid, it is kept at its current
7495 Auto-detect the crop size.
7497 It calculates the necessary cropping parameters and prints the
7498 recommended parameters via the logging system. The detected dimensions
7499 correspond to the non-black area of the input video.
7501 It accepts the following parameters:
7506 Set higher black value threshold, which can be optionally specified
7507 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7508 value greater to the set value is considered non-black. It defaults to 24.
7509 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7510 on the bitdepth of the pixel format.
7513 The value which the width/height should be divisible by. It defaults to
7514 16. The offset is automatically adjusted to center the video. Use 2 to
7515 get only even dimensions (needed for 4:2:2 video). 16 is best when
7516 encoding to most video codecs.
7518 @item reset_count, reset
7519 Set the counter that determines after how many frames cropdetect will
7520 reset the previously detected largest video area and start over to
7521 detect the current optimal crop area. Default value is 0.
7523 This can be useful when channel logos distort the video area. 0
7524 indicates 'never reset', and returns the largest area encountered during
7531 Delay video filtering until a given wallclock timestamp. The filter first
7532 passes on @option{preroll} amount of frames, then it buffers at most
7533 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7534 it forwards the buffered frames and also any subsequent frames coming in its
7537 The filter can be used synchronize the output of multiple ffmpeg processes for
7538 realtime output devices like decklink. By putting the delay in the filtering
7539 chain and pre-buffering frames the process can pass on data to output almost
7540 immediately after the target wallclock timestamp is reached.
7542 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7548 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7551 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7554 The maximum duration of content to buffer before waiting for the cue expressed
7555 in seconds. Default is 0.
7562 Apply color adjustments using curves.
7564 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7565 component (red, green and blue) has its values defined by @var{N} key points
7566 tied from each other using a smooth curve. The x-axis represents the pixel
7567 values from the input frame, and the y-axis the new pixel values to be set for
7570 By default, a component curve is defined by the two points @var{(0;0)} and
7571 @var{(1;1)}. This creates a straight line where each original pixel value is
7572 "adjusted" to its own value, which means no change to the image.
7574 The filter allows you to redefine these two points and add some more. A new
7575 curve (using a natural cubic spline interpolation) will be define to pass
7576 smoothly through all these new coordinates. The new defined points needs to be
7577 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7578 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7579 the vector spaces, the values will be clipped accordingly.
7581 The filter accepts the following options:
7585 Select one of the available color presets. This option can be used in addition
7586 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7587 options takes priority on the preset values.
7588 Available presets are:
7591 @item color_negative
7594 @item increase_contrast
7596 @item linear_contrast
7597 @item medium_contrast
7599 @item strong_contrast
7602 Default is @code{none}.
7604 Set the master key points. These points will define a second pass mapping. It
7605 is sometimes called a "luminance" or "value" mapping. It can be used with
7606 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7607 post-processing LUT.
7609 Set the key points for the red component.
7611 Set the key points for the green component.
7613 Set the key points for the blue component.
7615 Set the key points for all components (not including master).
7616 Can be used in addition to the other key points component
7617 options. In this case, the unset component(s) will fallback on this
7618 @option{all} setting.
7620 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7622 Save Gnuplot script of the curves in specified file.
7625 To avoid some filtergraph syntax conflicts, each key points list need to be
7626 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7628 @subsection Examples
7632 Increase slightly the middle level of blue:
7634 curves=blue='0/0 0.5/0.58 1/1'
7640 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'
7642 Here we obtain the following coordinates for each components:
7645 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7647 @code{(0;0) (0.50;0.48) (1;1)}
7649 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7653 The previous example can also be achieved with the associated built-in preset:
7655 curves=preset=vintage
7665 Use a Photoshop preset and redefine the points of the green component:
7667 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7671 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7672 and @command{gnuplot}:
7674 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7675 gnuplot -p /tmp/curves.plt
7681 Video data analysis filter.
7683 This filter shows hexadecimal pixel values of part of video.
7685 The filter accepts the following options:
7689 Set output video size.
7692 Set x offset from where to pick pixels.
7695 Set y offset from where to pick pixels.
7698 Set scope mode, can be one of the following:
7701 Draw hexadecimal pixel values with white color on black background.
7704 Draw hexadecimal pixel values with input video pixel color on black
7708 Draw hexadecimal pixel values on color background picked from input video,
7709 the text color is picked in such way so its always visible.
7713 Draw rows and columns numbers on left and top of video.
7716 Set background opacity.
7721 Denoise frames using 2D DCT (frequency domain filtering).
7723 This filter is not designed for real time.
7725 The filter accepts the following options:
7729 Set the noise sigma constant.
7731 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7732 coefficient (absolute value) below this threshold with be dropped.
7734 If you need a more advanced filtering, see @option{expr}.
7736 Default is @code{0}.
7739 Set number overlapping pixels for each block. Since the filter can be slow, you
7740 may want to reduce this value, at the cost of a less effective filter and the
7741 risk of various artefacts.
7743 If the overlapping value doesn't permit processing the whole input width or
7744 height, a warning will be displayed and according borders won't be denoised.
7746 Default value is @var{blocksize}-1, which is the best possible setting.
7749 Set the coefficient factor expression.
7751 For each coefficient of a DCT block, this expression will be evaluated as a
7752 multiplier value for the coefficient.
7754 If this is option is set, the @option{sigma} option will be ignored.
7756 The absolute value of the coefficient can be accessed through the @var{c}
7760 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7761 @var{blocksize}, which is the width and height of the processed blocks.
7763 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7764 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7765 on the speed processing. Also, a larger block size does not necessarily means a
7769 @subsection Examples
7771 Apply a denoise with a @option{sigma} of @code{4.5}:
7776 The same operation can be achieved using the expression system:
7778 dctdnoiz=e='gte(c, 4.5*3)'
7781 Violent denoise using a block size of @code{16x16}:
7788 Remove banding artifacts from input video.
7789 It works by replacing banded pixels with average value of referenced pixels.
7791 The filter accepts the following options:
7798 Set banding detection threshold for each plane. Default is 0.02.
7799 Valid range is 0.00003 to 0.5.
7800 If difference between current pixel and reference pixel is less than threshold,
7801 it will be considered as banded.
7804 Banding detection range in pixels. Default is 16. If positive, random number
7805 in range 0 to set value will be used. If negative, exact absolute value
7807 The range defines square of four pixels around current pixel.
7810 Set direction in radians from which four pixel will be compared. If positive,
7811 random direction from 0 to set direction will be picked. If negative, exact of
7812 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7813 will pick only pixels on same row and -PI/2 will pick only pixels on same
7817 If enabled, current pixel is compared with average value of all four
7818 surrounding pixels. The default is enabled. If disabled current pixel is
7819 compared with all four surrounding pixels. The pixel is considered banded
7820 if only all four differences with surrounding pixels are less than threshold.
7823 If enabled, current pixel is changed if and only if all pixel components are banded,
7824 e.g. banding detection threshold is triggered for all color components.
7825 The default is disabled.
7830 Remove blocking artifacts from input video.
7832 The filter accepts the following options:
7836 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7837 This controls what kind of deblocking is applied.
7840 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7846 Set blocking detection thresholds. Allowed range is 0 to 1.
7847 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7848 Using higher threshold gives more deblocking strength.
7849 Setting @var{alpha} controls threshold detection at exact edge of block.
7850 Remaining options controls threshold detection near the edge. Each one for
7851 below/above or left/right. Setting any of those to @var{0} disables
7855 Set planes to filter. Default is to filter all available planes.
7858 @subsection Examples
7862 Deblock using weak filter and block size of 4 pixels.
7864 deblock=filter=weak:block=4
7868 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7869 deblocking more edges.
7871 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7875 Similar as above, but filter only first plane.
7877 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7881 Similar as above, but filter only second and third plane.
7883 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7890 Drop duplicated frames at regular intervals.
7892 The filter accepts the following options:
7896 Set the number of frames from which one will be dropped. Setting this to
7897 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7898 Default is @code{5}.
7901 Set the threshold for duplicate detection. If the difference metric for a frame
7902 is less than or equal to this value, then it is declared as duplicate. Default
7906 Set scene change threshold. Default is @code{15}.
7910 Set the size of the x and y-axis blocks used during metric calculations.
7911 Larger blocks give better noise suppression, but also give worse detection of
7912 small movements. Must be a power of two. Default is @code{32}.
7915 Mark main input as a pre-processed input and activate clean source input
7916 stream. This allows the input to be pre-processed with various filters to help
7917 the metrics calculation while keeping the frame selection lossless. When set to
7918 @code{1}, the first stream is for the pre-processed input, and the second
7919 stream is the clean source from where the kept frames are chosen. Default is
7923 Set whether or not chroma is considered in the metric calculations. Default is
7929 Apply 2D deconvolution of video stream in frequency domain using second stream
7932 The filter accepts the following options:
7936 Set which planes to process.
7939 Set which impulse video frames will be processed, can be @var{first}
7940 or @var{all}. Default is @var{all}.
7943 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7944 and height are not same and not power of 2 or if stream prior to convolving
7948 The @code{deconvolve} filter also supports the @ref{framesync} options.
7952 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
7954 It accepts the following options:
7958 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
7959 @var{rainbows} for cross-color reduction.
7962 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
7965 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
7968 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
7971 Set temporal chroma threshold. Lower values increases reduction of cross-color.
7976 Apply deflate effect to the video.
7978 This filter replaces the pixel by the local(3x3) average by taking into account
7979 only values lower than the pixel.
7981 It accepts the following options:
7988 Limit the maximum change for each plane, default is 65535.
7989 If 0, plane will remain unchanged.
7994 Remove temporal frame luminance variations.
7996 It accepts the following options:
8000 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
8003 Set averaging mode to smooth temporal luminance variations.
8005 Available values are:
8030 Do not actually modify frame. Useful when one only wants metadata.
8035 Remove judder produced by partially interlaced telecined content.
8037 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
8038 source was partially telecined content then the output of @code{pullup,dejudder}
8039 will have a variable frame rate. May change the recorded frame rate of the
8040 container. Aside from that change, this filter will not affect constant frame
8043 The option available in this filter is:
8047 Specify the length of the window over which the judder repeats.
8049 Accepts any integer greater than 1. Useful values are:
8053 If the original was telecined from 24 to 30 fps (Film to NTSC).
8056 If the original was telecined from 25 to 30 fps (PAL to NTSC).
8059 If a mixture of the two.
8062 The default is @samp{4}.
8067 Suppress a TV station logo by a simple interpolation of the surrounding
8068 pixels. Just set a rectangle covering the logo and watch it disappear
8069 (and sometimes something even uglier appear - your mileage may vary).
8071 It accepts the following parameters:
8076 Specify the top left corner coordinates of the logo. They must be
8081 Specify the width and height of the logo to clear. They must be
8085 Specify the thickness of the fuzzy edge of the rectangle (added to
8086 @var{w} and @var{h}). The default value is 1. This option is
8087 deprecated, setting higher values should no longer be necessary and
8091 When set to 1, a green rectangle is drawn on the screen to simplify
8092 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8093 The default value is 0.
8095 The rectangle is drawn on the outermost pixels which will be (partly)
8096 replaced with interpolated values. The values of the next pixels
8097 immediately outside this rectangle in each direction will be used to
8098 compute the interpolated pixel values inside the rectangle.
8102 @subsection Examples
8106 Set a rectangle covering the area with top left corner coordinates 0,0
8107 and size 100x77, and a band of size 10:
8109 delogo=x=0:y=0:w=100:h=77:band=10
8116 Attempt to fix small changes in horizontal and/or vertical shift. This
8117 filter helps remove camera shake from hand-holding a camera, bumping a
8118 tripod, moving on a vehicle, etc.
8120 The filter accepts the following options:
8128 Specify a rectangular area where to limit the search for motion
8130 If desired the search for motion vectors can be limited to a
8131 rectangular area of the frame defined by its top left corner, width
8132 and height. These parameters have the same meaning as the drawbox
8133 filter which can be used to visualise the position of the bounding
8136 This is useful when simultaneous movement of subjects within the frame
8137 might be confused for camera motion by the motion vector search.
8139 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8140 then the full frame is used. This allows later options to be set
8141 without specifying the bounding box for the motion vector search.
8143 Default - search the whole frame.
8147 Specify the maximum extent of movement in x and y directions in the
8148 range 0-64 pixels. Default 16.
8151 Specify how to generate pixels to fill blanks at the edge of the
8152 frame. Available values are:
8155 Fill zeroes at blank locations
8157 Original image at blank locations
8159 Extruded edge value at blank locations
8161 Mirrored edge at blank locations
8163 Default value is @samp{mirror}.
8166 Specify the blocksize to use for motion search. Range 4-128 pixels,
8170 Specify the contrast threshold for blocks. Only blocks with more than
8171 the specified contrast (difference between darkest and lightest
8172 pixels) will be considered. Range 1-255, default 125.
8175 Specify the search strategy. Available values are:
8178 Set exhaustive search
8180 Set less exhaustive search.
8182 Default value is @samp{exhaustive}.
8185 If set then a detailed log of the motion search is written to the
8192 Remove unwanted contamination of foreground colors, caused by reflected color of
8193 greenscreen or bluescreen.
8195 This filter accepts the following options:
8199 Set what type of despill to use.
8202 Set how spillmap will be generated.
8205 Set how much to get rid of still remaining spill.
8208 Controls amount of red in spill area.
8211 Controls amount of green in spill area.
8212 Should be -1 for greenscreen.
8215 Controls amount of blue in spill area.
8216 Should be -1 for bluescreen.
8219 Controls brightness of spill area, preserving colors.
8222 Modify alpha from generated spillmap.
8227 Apply an exact inverse of the telecine operation. It requires a predefined
8228 pattern specified using the pattern option which must be the same as that passed
8229 to the telecine filter.
8231 This filter accepts the following options:
8240 The default value is @code{top}.
8244 A string of numbers representing the pulldown pattern you wish to apply.
8245 The default value is @code{23}.
8248 A number representing position of the first frame with respect to the telecine
8249 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8254 Apply dilation effect to the video.
8256 This filter replaces the pixel by the local(3x3) maximum.
8258 It accepts the following options:
8265 Limit the maximum change for each plane, default is 65535.
8266 If 0, plane will remain unchanged.
8269 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8272 Flags to local 3x3 coordinates maps like this:
8281 Displace pixels as indicated by second and third input stream.
8283 It takes three input streams and outputs one stream, the first input is the
8284 source, and second and third input are displacement maps.
8286 The second input specifies how much to displace pixels along the
8287 x-axis, while the third input specifies how much to displace pixels
8289 If one of displacement map streams terminates, last frame from that
8290 displacement map will be used.
8292 Note that once generated, displacements maps can be reused over and over again.
8294 A description of the accepted options follows.
8298 Set displace behavior for pixels that are out of range.
8300 Available values are:
8303 Missing pixels are replaced by black pixels.
8306 Adjacent pixels will spread out to replace missing pixels.
8309 Out of range pixels are wrapped so they point to pixels of other side.
8312 Out of range pixels will be replaced with mirrored pixels.
8314 Default is @samp{smear}.
8318 @subsection Examples
8322 Add ripple effect to rgb input of video size hd720:
8324 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
8328 Add wave effect to rgb input of video size hd720:
8330 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
8336 Draw a colored box on the input image.
8338 It accepts the following parameters:
8343 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8347 The expressions which specify the width and height of the box; if 0 they are interpreted as
8348 the input width and height. It defaults to 0.
8351 Specify the color of the box to write. For the general syntax of this option,
8352 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8353 value @code{invert} is used, the box edge color is the same as the
8354 video with inverted luma.
8357 The expression which sets the thickness of the box edge.
8358 A value of @code{fill} will create a filled box. Default value is @code{3}.
8360 See below for the list of accepted constants.
8363 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8364 will overwrite the video's color and alpha pixels.
8365 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8368 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8369 following constants:
8373 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8377 horizontal and vertical chroma subsample values. For example for the
8378 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8382 The input width and height.
8385 The input sample aspect ratio.
8389 The x and y offset coordinates where the box is drawn.
8393 The width and height of the drawn box.
8396 The thickness of the drawn box.
8398 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8399 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8403 @subsection Examples
8407 Draw a black box around the edge of the input image:
8413 Draw a box with color red and an opacity of 50%:
8415 drawbox=10:20:200:60:red@@0.5
8418 The previous example can be specified as:
8420 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8424 Fill the box with pink color:
8426 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8430 Draw a 2-pixel red 2.40:1 mask:
8432 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
8438 Draw a grid on the input image.
8440 It accepts the following parameters:
8445 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8449 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8450 input width and height, respectively, minus @code{thickness}, so image gets
8451 framed. Default to 0.
8454 Specify the color of the grid. For the general syntax of this option,
8455 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8456 value @code{invert} is used, the grid color is the same as the
8457 video with inverted luma.
8460 The expression which sets the thickness of the grid line. Default value is @code{1}.
8462 See below for the list of accepted constants.
8465 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8466 will overwrite the video's color and alpha pixels.
8467 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8470 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8471 following constants:
8475 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8479 horizontal and vertical chroma subsample values. For example for the
8480 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8484 The input grid cell width and height.
8487 The input sample aspect ratio.
8491 The x and y coordinates of some point of grid intersection (meant to configure offset).
8495 The width and height of the drawn cell.
8498 The thickness of the drawn cell.
8500 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8501 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8505 @subsection Examples
8509 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8511 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8515 Draw a white 3x3 grid with an opacity of 50%:
8517 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8524 Draw a text string or text from a specified file on top of a video, using the
8525 libfreetype library.
8527 To enable compilation of this filter, you need to configure FFmpeg with
8528 @code{--enable-libfreetype}.
8529 To enable default font fallback and the @var{font} option you need to
8530 configure FFmpeg with @code{--enable-libfontconfig}.
8531 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8532 @code{--enable-libfribidi}.
8536 It accepts the following parameters:
8541 Used to draw a box around text using the background color.
8542 The value must be either 1 (enable) or 0 (disable).
8543 The default value of @var{box} is 0.
8546 Set the width of the border to be drawn around the box using @var{boxcolor}.
8547 The default value of @var{boxborderw} is 0.
8550 The color to be used for drawing box around text. For the syntax of this
8551 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8553 The default value of @var{boxcolor} is "white".
8556 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8557 The default value of @var{line_spacing} is 0.
8560 Set the width of the border to be drawn around the text using @var{bordercolor}.
8561 The default value of @var{borderw} is 0.
8564 Set the color to be used for drawing border around text. For the syntax of this
8565 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8567 The default value of @var{bordercolor} is "black".
8570 Select how the @var{text} is expanded. Can be either @code{none},
8571 @code{strftime} (deprecated) or
8572 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8576 Set a start time for the count. Value is in microseconds. Only applied
8577 in the deprecated strftime expansion mode. To emulate in normal expansion
8578 mode use the @code{pts} function, supplying the start time (in seconds)
8579 as the second argument.
8582 If true, check and fix text coords to avoid clipping.
8585 The color to be used for drawing fonts. For the syntax of this option, check
8586 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8588 The default value of @var{fontcolor} is "black".
8590 @item fontcolor_expr
8591 String which is expanded the same way as @var{text} to obtain dynamic
8592 @var{fontcolor} value. By default this option has empty value and is not
8593 processed. When this option is set, it overrides @var{fontcolor} option.
8596 The font family to be used for drawing text. By default Sans.
8599 The font file to be used for drawing text. The path must be included.
8600 This parameter is mandatory if the fontconfig support is disabled.
8603 Draw the text applying alpha blending. The value can
8604 be a number between 0.0 and 1.0.
8605 The expression accepts the same variables @var{x, y} as well.
8606 The default value is 1.
8607 Please see @var{fontcolor_expr}.
8610 The font size to be used for drawing text.
8611 The default value of @var{fontsize} is 16.
8614 If set to 1, attempt to shape the text (for example, reverse the order of
8615 right-to-left text and join Arabic characters) before drawing it.
8616 Otherwise, just draw the text exactly as given.
8617 By default 1 (if supported).
8620 The flags to be used for loading the fonts.
8622 The flags map the corresponding flags supported by libfreetype, and are
8623 a combination of the following values:
8630 @item vertical_layout
8631 @item force_autohint
8634 @item ignore_global_advance_width
8636 @item ignore_transform
8642 Default value is "default".
8644 For more information consult the documentation for the FT_LOAD_*
8648 The color to be used for drawing a shadow behind the drawn text. For the
8649 syntax of this option, check the @ref{color syntax,,"Color" section in the
8650 ffmpeg-utils manual,ffmpeg-utils}.
8652 The default value of @var{shadowcolor} is "black".
8656 The x and y offsets for the text shadow position with respect to the
8657 position of the text. They can be either positive or negative
8658 values. The default value for both is "0".
8661 The starting frame number for the n/frame_num variable. The default value
8665 The size in number of spaces to use for rendering the tab.
8669 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8670 format. It can be used with or without text parameter. @var{timecode_rate}
8671 option must be specified.
8673 @item timecode_rate, rate, r
8674 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8675 integer. Minimum value is "1".
8676 Drop-frame timecode is supported for frame rates 30 & 60.
8679 If set to 1, the output of the timecode option will wrap around at 24 hours.
8680 Default is 0 (disabled).
8683 The text string to be drawn. The text must be a sequence of UTF-8
8685 This parameter is mandatory if no file is specified with the parameter
8689 A text file containing text to be drawn. The text must be a sequence
8690 of UTF-8 encoded characters.
8692 This parameter is mandatory if no text string is specified with the
8693 parameter @var{text}.
8695 If both @var{text} and @var{textfile} are specified, an error is thrown.
8698 If set to 1, the @var{textfile} will be reloaded before each frame.
8699 Be sure to update it atomically, or it may be read partially, or even fail.
8703 The expressions which specify the offsets where text will be drawn
8704 within the video frame. They are relative to the top/left border of the
8707 The default value of @var{x} and @var{y} is "0".
8709 See below for the list of accepted constants and functions.
8712 The parameters for @var{x} and @var{y} are expressions containing the
8713 following constants and functions:
8717 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8721 horizontal and vertical chroma subsample values. For example for the
8722 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8725 the height of each text line
8733 @item max_glyph_a, ascent
8734 the maximum distance from the baseline to the highest/upper grid
8735 coordinate used to place a glyph outline point, for all the rendered
8737 It is a positive value, due to the grid's orientation with the Y axis
8740 @item max_glyph_d, descent
8741 the maximum distance from the baseline to the lowest grid coordinate
8742 used to place a glyph outline point, for all the rendered glyphs.
8743 This is a negative value, due to the grid's orientation, with the Y axis
8747 maximum glyph height, that is the maximum height for all the glyphs
8748 contained in the rendered text, it is equivalent to @var{ascent} -
8752 maximum glyph width, that is the maximum width for all the glyphs
8753 contained in the rendered text
8756 the number of input frame, starting from 0
8758 @item rand(min, max)
8759 return a random number included between @var{min} and @var{max}
8762 The input sample aspect ratio.
8765 timestamp expressed in seconds, NAN if the input timestamp is unknown
8768 the height of the rendered text
8771 the width of the rendered text
8775 the x and y offset coordinates where the text is drawn.
8777 These parameters allow the @var{x} and @var{y} expressions to refer
8778 each other, so you can for example specify @code{y=x/dar}.
8781 @anchor{drawtext_expansion}
8782 @subsection Text expansion
8784 If @option{expansion} is set to @code{strftime},
8785 the filter recognizes strftime() sequences in the provided text and
8786 expands them accordingly. Check the documentation of strftime(). This
8787 feature is deprecated.
8789 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8791 If @option{expansion} is set to @code{normal} (which is the default),
8792 the following expansion mechanism is used.
8794 The backslash character @samp{\}, followed by any character, always expands to
8795 the second character.
8797 Sequences of the form @code{%@{...@}} are expanded. The text between the
8798 braces is a function name, possibly followed by arguments separated by ':'.
8799 If the arguments contain special characters or delimiters (':' or '@}'),
8800 they should be escaped.
8802 Note that they probably must also be escaped as the value for the
8803 @option{text} option in the filter argument string and as the filter
8804 argument in the filtergraph description, and possibly also for the shell,
8805 that makes up to four levels of escaping; using a text file avoids these
8808 The following functions are available:
8813 The expression evaluation result.
8815 It must take one argument specifying the expression to be evaluated,
8816 which accepts the same constants and functions as the @var{x} and
8817 @var{y} values. Note that not all constants should be used, for
8818 example the text size is not known when evaluating the expression, so
8819 the constants @var{text_w} and @var{text_h} will have an undefined
8822 @item expr_int_format, eif
8823 Evaluate the expression's value and output as formatted integer.
8825 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8826 The second argument specifies the output format. Allowed values are @samp{x},
8827 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8828 @code{printf} function.
8829 The third parameter is optional and sets the number of positions taken by the output.
8830 It can be used to add padding with zeros from the left.
8833 The time at which the filter is running, expressed in UTC.
8834 It can accept an argument: a strftime() format string.
8837 The time at which the filter is running, expressed in the local time zone.
8838 It can accept an argument: a strftime() format string.
8841 Frame metadata. Takes one or two arguments.
8843 The first argument is mandatory and specifies the metadata key.
8845 The second argument is optional and specifies a default value, used when the
8846 metadata key is not found or empty.
8849 The frame number, starting from 0.
8852 A 1 character description of the current picture type.
8855 The timestamp of the current frame.
8856 It can take up to three arguments.
8858 The first argument is the format of the timestamp; it defaults to @code{flt}
8859 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8860 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8861 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8862 @code{localtime} stands for the timestamp of the frame formatted as
8863 local time zone time.
8865 The second argument is an offset added to the timestamp.
8867 If the format is set to @code{hms}, a third argument @code{24HH} may be
8868 supplied to present the hour part of the formatted timestamp in 24h format
8871 If the format is set to @code{localtime} or @code{gmtime},
8872 a third argument may be supplied: a strftime() format string.
8873 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8876 @subsection Examples
8880 Draw "Test Text" with font FreeSerif, using the default values for the
8881 optional parameters.
8884 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8888 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8889 and y=50 (counting from the top-left corner of the screen), text is
8890 yellow with a red box around it. Both the text and the box have an
8894 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8895 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8898 Note that the double quotes are not necessary if spaces are not used
8899 within the parameter list.
8902 Show the text at the center of the video frame:
8904 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8908 Show the text at a random position, switching to a new position every 30 seconds:
8910 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)"
8914 Show a text line sliding from right to left in the last row of the video
8915 frame. The file @file{LONG_LINE} is assumed to contain a single line
8918 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8922 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8924 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8928 Draw a single green letter "g", at the center of the input video.
8929 The glyph baseline is placed at half screen height.
8931 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8935 Show text for 1 second every 3 seconds:
8937 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8941 Use fontconfig to set the font. Note that the colons need to be escaped.
8943 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8947 Print the date of a real-time encoding (see strftime(3)):
8949 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8953 Show text fading in and out (appearing/disappearing):
8956 DS=1.0 # display start
8957 DE=10.0 # display end
8958 FID=1.5 # fade in duration
8959 FOD=5 # fade out duration
8960 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 @}"
8964 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8965 and the @option{fontsize} value are included in the @option{y} offset.
8967 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8968 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8973 For more information about libfreetype, check:
8974 @url{http://www.freetype.org/}.
8976 For more information about fontconfig, check:
8977 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8979 For more information about libfribidi, check:
8980 @url{http://fribidi.org/}.
8984 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8986 The filter accepts the following options:
8991 Set low and high threshold values used by the Canny thresholding
8994 The high threshold selects the "strong" edge pixels, which are then
8995 connected through 8-connectivity with the "weak" edge pixels selected
8996 by the low threshold.
8998 @var{low} and @var{high} threshold values must be chosen in the range
8999 [0,1], and @var{low} should be lesser or equal to @var{high}.
9001 Default value for @var{low} is @code{20/255}, and default value for @var{high}
9005 Define the drawing mode.
9009 Draw white/gray wires on black background.
9012 Mix the colors to create a paint/cartoon effect.
9015 Apply Canny edge detector on all selected planes.
9017 Default value is @var{wires}.
9020 Select planes for filtering. By default all available planes are filtered.
9023 @subsection Examples
9027 Standard edge detection with custom values for the hysteresis thresholding:
9029 edgedetect=low=0.1:high=0.4
9033 Painting effect without thresholding:
9035 edgedetect=mode=colormix:high=0
9040 Set brightness, contrast, saturation and approximate gamma adjustment.
9042 The filter accepts the following options:
9046 Set the contrast expression. The value must be a float value in range
9047 @code{-2.0} to @code{2.0}. The default value is "1".
9050 Set the brightness expression. The value must be a float value in
9051 range @code{-1.0} to @code{1.0}. The default value is "0".
9054 Set the saturation expression. The value must be a float in
9055 range @code{0.0} to @code{3.0}. The default value is "1".
9058 Set the gamma expression. The value must be a float in range
9059 @code{0.1} to @code{10.0}. The default value is "1".
9062 Set the gamma expression for red. The value must be a float in
9063 range @code{0.1} to @code{10.0}. The default value is "1".
9066 Set the gamma expression for green. The value must be a float in range
9067 @code{0.1} to @code{10.0}. The default value is "1".
9070 Set the gamma expression for blue. The value must be a float in range
9071 @code{0.1} to @code{10.0}. The default value is "1".
9074 Set the gamma weight expression. It can be used to reduce the effect
9075 of a high gamma value on bright image areas, e.g. keep them from
9076 getting overamplified and just plain white. The value must be a float
9077 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
9078 gamma correction all the way down while @code{1.0} leaves it at its
9079 full strength. Default is "1".
9082 Set when the expressions for brightness, contrast, saturation and
9083 gamma expressions are evaluated.
9085 It accepts the following values:
9088 only evaluate expressions once during the filter initialization or
9089 when a command is processed
9092 evaluate expressions for each incoming frame
9095 Default value is @samp{init}.
9098 The expressions accept the following parameters:
9101 frame count of the input frame starting from 0
9104 byte position of the corresponding packet in the input file, NAN if
9108 frame rate of the input video, NAN if the input frame rate is unknown
9111 timestamp expressed in seconds, NAN if the input timestamp is unknown
9114 @subsection Commands
9115 The filter supports the following commands:
9119 Set the contrast expression.
9122 Set the brightness expression.
9125 Set the saturation expression.
9128 Set the gamma expression.
9131 Set the gamma_r expression.
9134 Set gamma_g expression.
9137 Set gamma_b expression.
9140 Set gamma_weight expression.
9142 The command accepts the same syntax of the corresponding option.
9144 If the specified expression is not valid, it is kept at its current
9151 Apply erosion effect to the video.
9153 This filter replaces the pixel by the local(3x3) minimum.
9155 It accepts the following options:
9162 Limit the maximum change for each plane, default is 65535.
9163 If 0, plane will remain unchanged.
9166 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9169 Flags to local 3x3 coordinates maps like this:
9176 @section extractplanes
9178 Extract color channel components from input video stream into
9179 separate grayscale video streams.
9181 The filter accepts the following option:
9185 Set plane(s) to extract.
9187 Available values for planes are:
9198 Choosing planes not available in the input will result in an error.
9199 That means you cannot select @code{r}, @code{g}, @code{b} planes
9200 with @code{y}, @code{u}, @code{v} planes at same time.
9203 @subsection Examples
9207 Extract luma, u and v color channel component from input video frame
9208 into 3 grayscale outputs:
9210 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
9216 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9218 For each input image, the filter will compute the optimal mapping from
9219 the input to the output given the codebook length, that is the number
9220 of distinct output colors.
9222 This filter accepts the following options.
9225 @item codebook_length, l
9226 Set codebook length. The value must be a positive integer, and
9227 represents the number of distinct output colors. Default value is 256.
9230 Set the maximum number of iterations to apply for computing the optimal
9231 mapping. The higher the value the better the result and the higher the
9232 computation time. Default value is 1.
9235 Set a random seed, must be an integer included between 0 and
9236 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9237 will try to use a good random seed on a best effort basis.
9240 Set pal8 output pixel format. This option does not work with codebook
9241 length greater than 256.
9246 Measure graylevel entropy in histogram of color channels of video frames.
9248 It accepts the following parameters:
9252 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9254 @var{diff} mode measures entropy of histogram delta values, absolute differences
9255 between neighbour histogram values.
9260 Apply a fade-in/out effect to the input video.
9262 It accepts the following parameters:
9266 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9268 Default is @code{in}.
9270 @item start_frame, s
9271 Specify the number of the frame to start applying the fade
9272 effect at. Default is 0.
9275 The number of frames that the fade effect lasts. At the end of the
9276 fade-in effect, the output video will have the same intensity as the input video.
9277 At the end of the fade-out transition, the output video will be filled with the
9278 selected @option{color}.
9282 If set to 1, fade only alpha channel, if one exists on the input.
9285 @item start_time, st
9286 Specify the timestamp (in seconds) of the frame to start to apply the fade
9287 effect. If both start_frame and start_time are specified, the fade will start at
9288 whichever comes last. Default is 0.
9291 The number of seconds for which the fade effect has to last. At the end of the
9292 fade-in effect the output video will have the same intensity as the input video,
9293 at the end of the fade-out transition the output video will be filled with the
9294 selected @option{color}.
9295 If both duration and nb_frames are specified, duration is used. Default is 0
9296 (nb_frames is used by default).
9299 Specify the color of the fade. Default is "black".
9302 @subsection Examples
9306 Fade in the first 30 frames of video:
9311 The command above is equivalent to:
9317 Fade out the last 45 frames of a 200-frame video:
9320 fade=type=out:start_frame=155:nb_frames=45
9324 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9326 fade=in:0:25, fade=out:975:25
9330 Make the first 5 frames yellow, then fade in from frame 5-24:
9332 fade=in:5:20:color=yellow
9336 Fade in alpha over first 25 frames of video:
9338 fade=in:0:25:alpha=1
9342 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9344 fade=t=in:st=5.5:d=0.5
9350 Apply arbitrary expressions to samples in frequency domain
9354 Adjust the dc value (gain) of the luma plane of the image. The filter
9355 accepts an integer value in range @code{0} to @code{1000}. The default
9356 value is set to @code{0}.
9359 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9360 filter accepts an integer value in range @code{0} to @code{1000}. The
9361 default value is set to @code{0}.
9364 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9365 filter accepts an integer value in range @code{0} to @code{1000}. The
9366 default value is set to @code{0}.
9369 Set the frequency domain weight expression for the luma plane.
9372 Set the frequency domain weight expression for the 1st chroma plane.
9375 Set the frequency domain weight expression for the 2nd chroma plane.
9378 Set when the expressions are evaluated.
9380 It accepts the following values:
9383 Only evaluate expressions once during the filter initialization.
9386 Evaluate expressions for each incoming frame.
9389 Default value is @samp{init}.
9391 The filter accepts the following variables:
9394 The coordinates of the current sample.
9398 The width and height of the image.
9401 The number of input frame, starting from 0.
9404 @subsection Examples
9410 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9416 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9422 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9428 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9434 Denoise frames using 3D FFT (frequency domain filtering).
9436 The filter accepts the following options:
9440 Set the noise sigma constant. This sets denoising strength.
9441 Default value is 1. Allowed range is from 0 to 30.
9442 Using very high sigma with low overlap may give blocking artifacts.
9445 Set amount of denoising. By default all detected noise is reduced.
9446 Default value is 1. Allowed range is from 0 to 1.
9449 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9450 Actual size of block in pixels is 2 to power of @var{block}, so by default
9451 block size in pixels is 2^4 which is 16.
9454 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9457 Set number of previous frames to use for denoising. By default is set to 0.
9460 Set number of next frames to to use for denoising. By default is set to 0.
9463 Set planes which will be filtered, by default are all available filtered
9469 Extract a single field from an interlaced image using stride
9470 arithmetic to avoid wasting CPU time. The output frames are marked as
9473 The filter accepts the following options:
9477 Specify whether to extract the top (if the value is @code{0} or
9478 @code{top}) or the bottom field (if the value is @code{1} or
9484 Create new frames by copying the top and bottom fields from surrounding frames
9485 supplied as numbers by the hint file.
9489 Set file containing hints: absolute/relative frame numbers.
9491 There must be one line for each frame in a clip. Each line must contain two
9492 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9493 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9494 is current frame number for @code{absolute} mode or out of [-1, 1] range
9495 for @code{relative} mode. First number tells from which frame to pick up top
9496 field and second number tells from which frame to pick up bottom field.
9498 If optionally followed by @code{+} output frame will be marked as interlaced,
9499 else if followed by @code{-} output frame will be marked as progressive, else
9500 it will be marked same as input frame.
9501 If line starts with @code{#} or @code{;} that line is skipped.
9504 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9507 Example of first several lines of @code{hint} file for @code{relative} mode:
9510 1,0 - # second frame, use third's frame top field and second's frame bottom field
9511 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9528 Field matching filter for inverse telecine. It is meant to reconstruct the
9529 progressive frames from a telecined stream. The filter does not drop duplicated
9530 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9531 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9533 The separation of the field matching and the decimation is notably motivated by
9534 the possibility of inserting a de-interlacing filter fallback between the two.
9535 If the source has mixed telecined and real interlaced content,
9536 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9537 But these remaining combed frames will be marked as interlaced, and thus can be
9538 de-interlaced by a later filter such as @ref{yadif} before decimation.
9540 In addition to the various configuration options, @code{fieldmatch} can take an
9541 optional second stream, activated through the @option{ppsrc} option. If
9542 enabled, the frames reconstruction will be based on the fields and frames from
9543 this second stream. This allows the first input to be pre-processed in order to
9544 help the various algorithms of the filter, while keeping the output lossless
9545 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9546 or brightness/contrast adjustments can help.
9548 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9549 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9550 which @code{fieldmatch} is based on. While the semantic and usage are very
9551 close, some behaviour and options names can differ.
9553 The @ref{decimate} filter currently only works for constant frame rate input.
9554 If your input has mixed telecined (30fps) and progressive content with a lower
9555 framerate like 24fps use the following filterchain to produce the necessary cfr
9556 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9558 The filter accepts the following options:
9562 Specify the assumed field order of the input stream. Available values are:
9566 Auto detect parity (use FFmpeg's internal parity value).
9568 Assume bottom field first.
9570 Assume top field first.
9573 Note that it is sometimes recommended not to trust the parity announced by the
9576 Default value is @var{auto}.
9579 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9580 sense that it won't risk creating jerkiness due to duplicate frames when
9581 possible, but if there are bad edits or blended fields it will end up
9582 outputting combed frames when a good match might actually exist. On the other
9583 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9584 but will almost always find a good frame if there is one. The other values are
9585 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9586 jerkiness and creating duplicate frames versus finding good matches in sections
9587 with bad edits, orphaned fields, blended fields, etc.
9589 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9591 Available values are:
9595 2-way matching (p/c)
9597 2-way matching, and trying 3rd match if still combed (p/c + n)
9599 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9601 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9602 still combed (p/c + n + u/b)
9604 3-way matching (p/c/n)
9606 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9607 detected as combed (p/c/n + u/b)
9610 The parenthesis at the end indicate the matches that would be used for that
9611 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9614 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9617 Default value is @var{pc_n}.
9620 Mark the main input stream as a pre-processed input, and enable the secondary
9621 input stream as the clean source to pick the fields from. See the filter
9622 introduction for more details. It is similar to the @option{clip2} feature from
9625 Default value is @code{0} (disabled).
9628 Set the field to match from. It is recommended to set this to the same value as
9629 @option{order} unless you experience matching failures with that setting. In
9630 certain circumstances changing the field that is used to match from can have a
9631 large impact on matching performance. Available values are:
9635 Automatic (same value as @option{order}).
9637 Match from the bottom field.
9639 Match from the top field.
9642 Default value is @var{auto}.
9645 Set whether or not chroma is included during the match comparisons. In most
9646 cases it is recommended to leave this enabled. You should set this to @code{0}
9647 only if your clip has bad chroma problems such as heavy rainbowing or other
9648 artifacts. Setting this to @code{0} could also be used to speed things up at
9649 the cost of some accuracy.
9651 Default value is @code{1}.
9655 These define an exclusion band which excludes the lines between @option{y0} and
9656 @option{y1} from being included in the field matching decision. An exclusion
9657 band can be used to ignore subtitles, a logo, or other things that may
9658 interfere with the matching. @option{y0} sets the starting scan line and
9659 @option{y1} sets the ending line; all lines in between @option{y0} and
9660 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9661 @option{y0} and @option{y1} to the same value will disable the feature.
9662 @option{y0} and @option{y1} defaults to @code{0}.
9665 Set the scene change detection threshold as a percentage of maximum change on
9666 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9667 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9668 @option{scthresh} is @code{[0.0, 100.0]}.
9670 Default value is @code{12.0}.
9673 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9674 account the combed scores of matches when deciding what match to use as the
9675 final match. Available values are:
9679 No final matching based on combed scores.
9681 Combed scores are only used when a scene change is detected.
9683 Use combed scores all the time.
9686 Default is @var{sc}.
9689 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9690 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9691 Available values are:
9695 No forced calculation.
9697 Force p/c/n calculations.
9699 Force p/c/n/u/b calculations.
9702 Default value is @var{none}.
9705 This is the area combing threshold used for combed frame detection. This
9706 essentially controls how "strong" or "visible" combing must be to be detected.
9707 Larger values mean combing must be more visible and smaller values mean combing
9708 can be less visible or strong and still be detected. Valid settings are from
9709 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9710 be detected as combed). This is basically a pixel difference value. A good
9711 range is @code{[8, 12]}.
9713 Default value is @code{9}.
9716 Sets whether or not chroma is considered in the combed frame decision. Only
9717 disable this if your source has chroma problems (rainbowing, etc.) that are
9718 causing problems for the combed frame detection with chroma enabled. Actually,
9719 using @option{chroma}=@var{0} is usually more reliable, except for the case
9720 where there is chroma only combing in the source.
9722 Default value is @code{0}.
9726 Respectively set the x-axis and y-axis size of the window used during combed
9727 frame detection. This has to do with the size of the area in which
9728 @option{combpel} pixels are required to be detected as combed for a frame to be
9729 declared combed. See the @option{combpel} parameter description for more info.
9730 Possible values are any number that is a power of 2 starting at 4 and going up
9733 Default value is @code{16}.
9736 The number of combed pixels inside any of the @option{blocky} by
9737 @option{blockx} size blocks on the frame for the frame to be detected as
9738 combed. While @option{cthresh} controls how "visible" the combing must be, this
9739 setting controls "how much" combing there must be in any localized area (a
9740 window defined by the @option{blockx} and @option{blocky} settings) on the
9741 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9742 which point no frames will ever be detected as combed). This setting is known
9743 as @option{MI} in TFM/VFM vocabulary.
9745 Default value is @code{80}.
9748 @anchor{p/c/n/u/b meaning}
9749 @subsection p/c/n/u/b meaning
9751 @subsubsection p/c/n
9753 We assume the following telecined stream:
9756 Top fields: 1 2 2 3 4
9757 Bottom fields: 1 2 3 4 4
9760 The numbers correspond to the progressive frame the fields relate to. Here, the
9761 first two frames are progressive, the 3rd and 4th are combed, and so on.
9763 When @code{fieldmatch} is configured to run a matching from bottom
9764 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9769 B 1 2 3 4 4 <-- matching reference
9778 As a result of the field matching, we can see that some frames get duplicated.
9779 To perform a complete inverse telecine, you need to rely on a decimation filter
9780 after this operation. See for instance the @ref{decimate} filter.
9782 The same operation now matching from top fields (@option{field}=@var{top})
9787 T 1 2 2 3 4 <-- matching reference
9797 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9798 basically, they refer to the frame and field of the opposite parity:
9801 @item @var{p} matches the field of the opposite parity in the previous frame
9802 @item @var{c} matches the field of the opposite parity in the current frame
9803 @item @var{n} matches the field of the opposite parity in the next frame
9808 The @var{u} and @var{b} matching are a bit special in the sense that they match
9809 from the opposite parity flag. In the following examples, we assume that we are
9810 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9811 'x' is placed above and below each matched fields.
9813 With bottom matching (@option{field}=@var{bottom}):
9818 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9819 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9827 With top matching (@option{field}=@var{top}):
9832 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9833 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9841 @subsection Examples
9843 Simple IVTC of a top field first telecined stream:
9845 fieldmatch=order=tff:combmatch=none, decimate
9848 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9850 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9855 Transform the field order of the input video.
9857 It accepts the following parameters:
9862 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9863 for bottom field first.
9866 The default value is @samp{tff}.
9868 The transformation is done by shifting the picture content up or down
9869 by one line, and filling the remaining line with appropriate picture content.
9870 This method is consistent with most broadcast field order converters.
9872 If the input video is not flagged as being interlaced, or it is already
9873 flagged as being of the required output field order, then this filter does
9874 not alter the incoming video.
9876 It is very useful when converting to or from PAL DV material,
9877 which is bottom field first.
9881 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9884 @section fifo, afifo
9886 Buffer input images and send them when they are requested.
9888 It is mainly useful when auto-inserted by the libavfilter
9891 It does not take parameters.
9893 @section fillborders
9895 Fill borders of the input video, without changing video stream dimensions.
9896 Sometimes video can have garbage at the four edges and you may not want to
9897 crop video input to keep size multiple of some number.
9899 This filter accepts the following options:
9903 Number of pixels to fill from left border.
9906 Number of pixels to fill from right border.
9909 Number of pixels to fill from top border.
9912 Number of pixels to fill from bottom border.
9917 It accepts the following values:
9920 fill pixels using outermost pixels
9923 fill pixels using mirroring
9926 fill pixels with constant value
9929 Default is @var{smear}.
9932 Set color for pixels in fixed mode. Default is @var{black}.
9937 Find a rectangular object
9939 It accepts the following options:
9943 Filepath of the object image, needs to be in gray8.
9946 Detection threshold, default is 0.5.
9949 Number of mipmaps, default is 3.
9951 @item xmin, ymin, xmax, ymax
9952 Specifies the rectangle in which to search.
9955 @subsection Examples
9959 Generate a representative palette of a given video using @command{ffmpeg}:
9961 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9967 Cover a rectangular object
9969 It accepts the following options:
9973 Filepath of the optional cover image, needs to be in yuv420.
9978 It accepts the following values:
9981 cover it by the supplied image
9983 cover it by interpolating the surrounding pixels
9986 Default value is @var{blur}.
9989 @subsection Examples
9993 Generate a representative palette of a given video using @command{ffmpeg}:
9995 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
10001 Flood area with values of same pixel components with another values.
10003 It accepts the following options:
10006 Set pixel x coordinate.
10009 Set pixel y coordinate.
10012 Set source #0 component value.
10015 Set source #1 component value.
10018 Set source #2 component value.
10021 Set source #3 component value.
10024 Set destination #0 component value.
10027 Set destination #1 component value.
10030 Set destination #2 component value.
10033 Set destination #3 component value.
10039 Convert the input video to one of the specified pixel formats.
10040 Libavfilter will try to pick one that is suitable as input to
10043 It accepts the following parameters:
10047 A '|'-separated list of pixel format names, such as
10048 "pix_fmts=yuv420p|monow|rgb24".
10052 @subsection Examples
10056 Convert the input video to the @var{yuv420p} format
10058 format=pix_fmts=yuv420p
10061 Convert the input video to any of the formats in the list
10063 format=pix_fmts=yuv420p|yuv444p|yuv410p
10070 Convert the video to specified constant frame rate by duplicating or dropping
10071 frames as necessary.
10073 It accepts the following parameters:
10077 The desired output frame rate. The default is @code{25}.
10080 Assume the first PTS should be the given value, in seconds. This allows for
10081 padding/trimming at the start of stream. By default, no assumption is made
10082 about the first frame's expected PTS, so no padding or trimming is done.
10083 For example, this could be set to 0 to pad the beginning with duplicates of
10084 the first frame if a video stream starts after the audio stream or to trim any
10085 frames with a negative PTS.
10088 Timestamp (PTS) rounding method.
10090 Possible values are:
10097 round towards -infinity
10099 round towards +infinity
10103 The default is @code{near}.
10106 Action performed when reading the last frame.
10108 Possible values are:
10111 Use same timestamp rounding method as used for other frames.
10113 Pass through last frame if input duration has not been reached yet.
10115 The default is @code{round}.
10119 Alternatively, the options can be specified as a flat string:
10120 @var{fps}[:@var{start_time}[:@var{round}]].
10122 See also the @ref{setpts} filter.
10124 @subsection Examples
10128 A typical usage in order to set the fps to 25:
10134 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10136 fps=fps=film:round=near
10142 Pack two different video streams into a stereoscopic video, setting proper
10143 metadata on supported codecs. The two views should have the same size and
10144 framerate and processing will stop when the shorter video ends. Please note
10145 that you may conveniently adjust view properties with the @ref{scale} and
10148 It accepts the following parameters:
10152 The desired packing format. Supported values are:
10157 The views are next to each other (default).
10160 The views are on top of each other.
10163 The views are packed by line.
10166 The views are packed by column.
10169 The views are temporally interleaved.
10178 # Convert left and right views into a frame-sequential video
10179 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10181 # Convert views into a side-by-side video with the same output resolution as the input
10182 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
10187 Change the frame rate by interpolating new video output frames from the source
10190 This filter is not designed to function correctly with interlaced media. If
10191 you wish to change the frame rate of interlaced media then you are required
10192 to deinterlace before this filter and re-interlace after this filter.
10194 A description of the accepted options follows.
10198 Specify the output frames per second. This option can also be specified
10199 as a value alone. The default is @code{50}.
10202 Specify the start of a range where the output frame will be created as a
10203 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10204 the default is @code{15}.
10207 Specify the end of a range where the output frame will be created as a
10208 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10209 the default is @code{240}.
10212 Specify the level at which a scene change is detected as a value between
10213 0 and 100 to indicate a new scene; a low value reflects a low
10214 probability for the current frame to introduce a new scene, while a higher
10215 value means the current frame is more likely to be one.
10216 The default is @code{8.2}.
10219 Specify flags influencing the filter process.
10221 Available value for @var{flags} is:
10224 @item scene_change_detect, scd
10225 Enable scene change detection using the value of the option @var{scene}.
10226 This flag is enabled by default.
10232 Select one frame every N-th frame.
10234 This filter accepts the following option:
10237 Select frame after every @code{step} frames.
10238 Allowed values are positive integers higher than 0. Default value is @code{1}.
10241 @section freezedetect
10243 Detect frozen video.
10245 This filter logs a message and sets frame metadata when it detects that the
10246 input video has no significant change in content during a specified duration.
10247 Video freeze detection calculates the mean average absolute difference of all
10248 the components of video frames and compares it to a noise floor.
10250 The printed times and duration are expressed in seconds. The
10251 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10252 whose timestamp equals or exceeds the detection duration and it contains the
10253 timestamp of the first frame of the freeze. The
10254 @code{lavfi.freezedetect.freeze_duration} and
10255 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10258 The filter accepts the following options:
10262 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10263 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10267 Set freeze duration until notification (default is 2 seconds).
10273 Apply a frei0r effect to the input video.
10275 To enable the compilation of this filter, you need to install the frei0r
10276 header and configure FFmpeg with @code{--enable-frei0r}.
10278 It accepts the following parameters:
10283 The name of the frei0r effect to load. If the environment variable
10284 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10285 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10286 Otherwise, the standard frei0r paths are searched, in this order:
10287 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10288 @file{/usr/lib/frei0r-1/}.
10290 @item filter_params
10291 A '|'-separated list of parameters to pass to the frei0r effect.
10295 A frei0r effect parameter can be a boolean (its value is either
10296 "y" or "n"), a double, a color (specified as
10297 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10298 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10299 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10300 a position (specified as @var{X}/@var{Y}, where
10301 @var{X} and @var{Y} are floating point numbers) and/or a string.
10303 The number and types of parameters depend on the loaded effect. If an
10304 effect parameter is not specified, the default value is set.
10306 @subsection Examples
10310 Apply the distort0r effect, setting the first two double parameters:
10312 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10316 Apply the colordistance effect, taking a color as the first parameter:
10318 frei0r=colordistance:0.2/0.3/0.4
10319 frei0r=colordistance:violet
10320 frei0r=colordistance:0x112233
10324 Apply the perspective effect, specifying the top left and top right image
10327 frei0r=perspective:0.2/0.2|0.8/0.2
10331 For more information, see
10332 @url{http://frei0r.dyne.org}
10336 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10338 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10339 processing filter, one of them is performed once per block, not per pixel.
10340 This allows for much higher speed.
10342 The filter accepts the following options:
10346 Set quality. This option defines the number of levels for averaging. It accepts
10347 an integer in the range 4-5. Default value is @code{4}.
10350 Force a constant quantization parameter. It accepts an integer in range 0-63.
10351 If not set, the filter will use the QP from the video stream (if available).
10354 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10355 more details but also more artifacts, while higher values make the image smoother
10356 but also blurrier. Default value is @code{0} − PSNR optimal.
10358 @item use_bframe_qp
10359 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10360 option may cause flicker since the B-Frames have often larger QP. Default is
10361 @code{0} (not enabled).
10367 Apply Gaussian blur filter.
10369 The filter accepts the following options:
10373 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10376 Set number of steps for Gaussian approximation. Default is @code{1}.
10379 Set which planes to filter. By default all planes are filtered.
10382 Set vertical sigma, if negative it will be same as @code{sigma}.
10383 Default is @code{-1}.
10388 Apply generic equation to each pixel.
10390 The filter accepts the following options:
10393 @item lum_expr, lum
10394 Set the luminance expression.
10396 Set the chrominance blue expression.
10398 Set the chrominance red expression.
10399 @item alpha_expr, a
10400 Set the alpha expression.
10402 Set the red expression.
10403 @item green_expr, g
10404 Set the green expression.
10406 Set the blue expression.
10409 The colorspace is selected according to the specified options. If one
10410 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10411 options is specified, the filter will automatically select a YCbCr
10412 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10413 @option{blue_expr} options is specified, it will select an RGB
10416 If one of the chrominance expression is not defined, it falls back on the other
10417 one. If no alpha expression is specified it will evaluate to opaque value.
10418 If none of chrominance expressions are specified, they will evaluate
10419 to the luminance expression.
10421 The expressions can use the following variables and functions:
10425 The sequential number of the filtered frame, starting from @code{0}.
10429 The coordinates of the current sample.
10433 The width and height of the image.
10437 Width and height scale depending on the currently filtered plane. It is the
10438 ratio between the corresponding luma plane number of pixels and the current
10439 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10440 @code{0.5,0.5} for chroma planes.
10443 Time of the current frame, expressed in seconds.
10446 Return the value of the pixel at location (@var{x},@var{y}) of the current
10450 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10454 Return the value of the pixel at location (@var{x},@var{y}) of the
10455 blue-difference chroma plane. Return 0 if there is no such plane.
10458 Return the value of the pixel at location (@var{x},@var{y}) of the
10459 red-difference chroma plane. Return 0 if there is no such plane.
10464 Return the value of the pixel at location (@var{x},@var{y}) of the
10465 red/green/blue component. Return 0 if there is no such component.
10468 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10469 plane. Return 0 if there is no such plane.
10472 For functions, if @var{x} and @var{y} are outside the area, the value will be
10473 automatically clipped to the closer edge.
10475 @subsection Examples
10479 Flip the image horizontally:
10485 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10486 wavelength of 100 pixels:
10488 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10492 Generate a fancy enigmatic moving light:
10494 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
10498 Generate a quick emboss effect:
10500 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10504 Modify RGB components depending on pixel position:
10506 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10510 Create a radial gradient that is the same size as the input (also see
10511 the @ref{vignette} filter):
10513 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10519 Fix the banding artifacts that are sometimes introduced into nearly flat
10520 regions by truncation to 8-bit color depth.
10521 Interpolate the gradients that should go where the bands are, and
10524 It is designed for playback only. Do not use it prior to
10525 lossy compression, because compression tends to lose the dither and
10526 bring back the bands.
10528 It accepts the following parameters:
10533 The maximum amount by which the filter will change any one pixel. This is also
10534 the threshold for detecting nearly flat regions. Acceptable values range from
10535 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10539 The neighborhood to fit the gradient to. A larger radius makes for smoother
10540 gradients, but also prevents the filter from modifying the pixels near detailed
10541 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10542 values will be clipped to the valid range.
10546 Alternatively, the options can be specified as a flat string:
10547 @var{strength}[:@var{radius}]
10549 @subsection Examples
10553 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10559 Specify radius, omitting the strength (which will fall-back to the default
10567 @section graphmonitor, agraphmonitor
10568 Show various filtergraph stats.
10570 With this filter one can debug complete filtergraph.
10571 Especially issues with links filling with queued frames.
10573 The filter accepts the following options:
10577 Set video output size. Default is @var{hd720}.
10580 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10583 Set output mode, can be @var{fulll} or @var{compact}.
10584 In @var{compact} mode only filters with some queued frames have displayed stats.
10587 Set flags which enable which stats are shown in video.
10589 Available values for flags are:
10592 Display number of queued frames in each link.
10594 @item frame_count_in
10595 Display number of frames taken from filter.
10597 @item frame_count_out
10598 Display number of frames given out from filter.
10601 Display current filtered frame pts.
10604 Display current filtered frame time.
10607 Display time base for filter link.
10610 Display used format for filter link.
10613 Display video size or number of audio channels in case of audio used by filter link.
10616 Display video frame rate or sample rate in case of audio used by filter link.
10620 Set upper limit for video rate of output stream, Default value is @var{25}.
10621 This guarantee that output video frame rate will not be higher than this value.
10625 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10626 and corrects the scene colors accordingly.
10628 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10630 The filter accepts the following options:
10634 The order of differentiation to be applied on the scene. Must be chosen in the range
10635 [0,2] and default value is 1.
10638 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10639 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10640 max value instead of calculating Minkowski distance.
10643 The standard deviation of Gaussian blur to be applied on the scene. Must be
10644 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10645 can't be equal to 0 if @var{difford} is greater than 0.
10648 @subsection Examples
10654 greyedge=difford=1:minknorm=5:sigma=2
10660 greyedge=difford=1:minknorm=0:sigma=2
10668 Apply a Hald CLUT to a video stream.
10670 First input is the video stream to process, and second one is the Hald CLUT.
10671 The Hald CLUT input can be a simple picture or a complete video stream.
10673 The filter accepts the following options:
10677 Force termination when the shortest input terminates. Default is @code{0}.
10679 Continue applying the last CLUT after the end of the stream. A value of
10680 @code{0} disable the filter after the last frame of the CLUT is reached.
10681 Default is @code{1}.
10684 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10685 filters share the same internals).
10687 More information about the Hald CLUT can be found on Eskil Steenberg's website
10688 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10690 @subsection Workflow examples
10692 @subsubsection Hald CLUT video stream
10694 Generate an identity Hald CLUT stream altered with various effects:
10696 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
10699 Note: make sure you use a lossless codec.
10701 Then use it with @code{haldclut} to apply it on some random stream:
10703 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10706 The Hald CLUT will be applied to the 10 first seconds (duration of
10707 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10708 to the remaining frames of the @code{mandelbrot} stream.
10710 @subsubsection Hald CLUT with preview
10712 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10713 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10714 biggest possible square starting at the top left of the picture. The remaining
10715 padding pixels (bottom or right) will be ignored. This area can be used to add
10716 a preview of the Hald CLUT.
10718 Typically, the following generated Hald CLUT will be supported by the
10719 @code{haldclut} filter:
10722 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10723 pad=iw+320 [padded_clut];
10724 smptebars=s=320x256, split [a][b];
10725 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10726 [main][b] overlay=W-320" -frames:v 1 clut.png
10729 It contains the original and a preview of the effect of the CLUT: SMPTE color
10730 bars are displayed on the right-top, and below the same color bars processed by
10733 Then, the effect of this Hald CLUT can be visualized with:
10735 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10740 Flip the input video horizontally.
10742 For example, to horizontally flip the input video with @command{ffmpeg}:
10744 ffmpeg -i in.avi -vf "hflip" out.avi
10748 This filter applies a global color histogram equalization on a
10751 It can be used to correct video that has a compressed range of pixel
10752 intensities. The filter redistributes the pixel intensities to
10753 equalize their distribution across the intensity range. It may be
10754 viewed as an "automatically adjusting contrast filter". This filter is
10755 useful only for correcting degraded or poorly captured source
10758 The filter accepts the following options:
10762 Determine the amount of equalization to be applied. As the strength
10763 is reduced, the distribution of pixel intensities more-and-more
10764 approaches that of the input frame. The value must be a float number
10765 in the range [0,1] and defaults to 0.200.
10768 Set the maximum intensity that can generated and scale the output
10769 values appropriately. The strength should be set as desired and then
10770 the intensity can be limited if needed to avoid washing-out. The value
10771 must be a float number in the range [0,1] and defaults to 0.210.
10774 Set the antibanding level. If enabled the filter will randomly vary
10775 the luminance of output pixels by a small amount to avoid banding of
10776 the histogram. Possible values are @code{none}, @code{weak} or
10777 @code{strong}. It defaults to @code{none}.
10782 Compute and draw a color distribution histogram for the input video.
10784 The computed histogram is a representation of the color component
10785 distribution in an image.
10787 Standard histogram displays the color components distribution in an image.
10788 Displays color graph for each color component. Shows distribution of
10789 the Y, U, V, A or R, G, B components, depending on input format, in the
10790 current frame. Below each graph a color component scale meter is shown.
10792 The filter accepts the following options:
10796 Set height of level. Default value is @code{200}.
10797 Allowed range is [50, 2048].
10800 Set height of color scale. Default value is @code{12}.
10801 Allowed range is [0, 40].
10805 It accepts the following values:
10808 Per color component graphs are placed below each other.
10811 Per color component graphs are placed side by side.
10814 Presents information identical to that in the @code{parade}, except
10815 that the graphs representing color components are superimposed directly
10818 Default is @code{stack}.
10821 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10822 Default is @code{linear}.
10825 Set what color components to display.
10826 Default is @code{7}.
10829 Set foreground opacity. Default is @code{0.7}.
10832 Set background opacity. Default is @code{0.5}.
10835 @subsection Examples
10840 Calculate and draw histogram:
10842 ffplay -i input -vf histogram
10850 This is a high precision/quality 3d denoise filter. It aims to reduce
10851 image noise, producing smooth images and making still images really
10852 still. It should enhance compressibility.
10854 It accepts the following optional parameters:
10858 A non-negative floating point number which specifies spatial luma strength.
10859 It defaults to 4.0.
10861 @item chroma_spatial
10862 A non-negative floating point number which specifies spatial chroma strength.
10863 It defaults to 3.0*@var{luma_spatial}/4.0.
10866 A floating point number which specifies luma temporal strength. It defaults to
10867 6.0*@var{luma_spatial}/4.0.
10870 A floating point number which specifies chroma temporal strength. It defaults to
10871 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10874 @anchor{hwdownload}
10875 @section hwdownload
10877 Download hardware frames to system memory.
10879 The input must be in hardware frames, and the output a non-hardware format.
10880 Not all formats will be supported on the output - it may be necessary to insert
10881 an additional @option{format} filter immediately following in the graph to get
10882 the output in a supported format.
10886 Map hardware frames to system memory or to another device.
10888 This filter has several different modes of operation; which one is used depends
10889 on the input and output formats:
10892 Hardware frame input, normal frame output
10894 Map the input frames to system memory and pass them to the output. If the
10895 original hardware frame is later required (for example, after overlaying
10896 something else on part of it), the @option{hwmap} filter can be used again
10897 in the next mode to retrieve it.
10899 Normal frame input, hardware frame output
10901 If the input is actually a software-mapped hardware frame, then unmap it -
10902 that is, return the original hardware frame.
10904 Otherwise, a device must be provided. Create new hardware surfaces on that
10905 device for the output, then map them back to the software format at the input
10906 and give those frames to the preceding filter. This will then act like the
10907 @option{hwupload} filter, but may be able to avoid an additional copy when
10908 the input is already in a compatible format.
10910 Hardware frame input and output
10912 A device must be supplied for the output, either directly or with the
10913 @option{derive_device} option. The input and output devices must be of
10914 different types and compatible - the exact meaning of this is
10915 system-dependent, but typically it means that they must refer to the same
10916 underlying hardware context (for example, refer to the same graphics card).
10918 If the input frames were originally created on the output device, then unmap
10919 to retrieve the original frames.
10921 Otherwise, map the frames to the output device - create new hardware frames
10922 on the output corresponding to the frames on the input.
10925 The following additional parameters are accepted:
10929 Set the frame mapping mode. Some combination of:
10932 The mapped frame should be readable.
10934 The mapped frame should be writeable.
10936 The mapping will always overwrite the entire frame.
10938 This may improve performance in some cases, as the original contents of the
10939 frame need not be loaded.
10941 The mapping must not involve any copying.
10943 Indirect mappings to copies of frames are created in some cases where either
10944 direct mapping is not possible or it would have unexpected properties.
10945 Setting this flag ensures that the mapping is direct and will fail if that is
10948 Defaults to @var{read+write} if not specified.
10950 @item derive_device @var{type}
10951 Rather than using the device supplied at initialisation, instead derive a new
10952 device of type @var{type} from the device the input frames exist on.
10955 In a hardware to hardware mapping, map in reverse - create frames in the sink
10956 and map them back to the source. This may be necessary in some cases where
10957 a mapping in one direction is required but only the opposite direction is
10958 supported by the devices being used.
10960 This option is dangerous - it may break the preceding filter in undefined
10961 ways if there are any additional constraints on that filter's output.
10962 Do not use it without fully understanding the implications of its use.
10968 Upload system memory frames to hardware surfaces.
10970 The device to upload to must be supplied when the filter is initialised. If
10971 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10974 @anchor{hwupload_cuda}
10975 @section hwupload_cuda
10977 Upload system memory frames to a CUDA device.
10979 It accepts the following optional parameters:
10983 The number of the CUDA device to use
10988 Apply a high-quality magnification filter designed for pixel art. This filter
10989 was originally created by Maxim Stepin.
10991 It accepts the following option:
10995 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10996 @code{hq3x} and @code{4} for @code{hq4x}.
10997 Default is @code{3}.
11001 Stack input videos horizontally.
11003 All streams must be of same pixel format and of same height.
11005 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
11006 to create same output.
11008 The filter accept the following option:
11012 Set number of input streams. Default is 2.
11015 If set to 1, force the output to terminate when the shortest input
11016 terminates. Default value is 0.
11021 Modify the hue and/or the saturation of the input.
11023 It accepts the following parameters:
11027 Specify the hue angle as a number of degrees. It accepts an expression,
11028 and defaults to "0".
11031 Specify the saturation in the [-10,10] range. It accepts an expression and
11035 Specify the hue angle as a number of radians. It accepts an
11036 expression, and defaults to "0".
11039 Specify the brightness in the [-10,10] range. It accepts an expression and
11043 @option{h} and @option{H} are mutually exclusive, and can't be
11044 specified at the same time.
11046 The @option{b}, @option{h}, @option{H} and @option{s} option values are
11047 expressions containing the following constants:
11051 frame count of the input frame starting from 0
11054 presentation timestamp of the input frame expressed in time base units
11057 frame rate of the input video, NAN if the input frame rate is unknown
11060 timestamp expressed in seconds, NAN if the input timestamp is unknown
11063 time base of the input video
11066 @subsection Examples
11070 Set the hue to 90 degrees and the saturation to 1.0:
11076 Same command but expressing the hue in radians:
11082 Rotate hue and make the saturation swing between 0
11083 and 2 over a period of 1 second:
11085 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11089 Apply a 3 seconds saturation fade-in effect starting at 0:
11091 hue="s=min(t/3\,1)"
11094 The general fade-in expression can be written as:
11096 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11100 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11102 hue="s=max(0\, min(1\, (8-t)/3))"
11105 The general fade-out expression can be written as:
11107 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11112 @subsection Commands
11114 This filter supports the following commands:
11120 Modify the hue and/or the saturation and/or brightness of the input video.
11121 The command accepts the same syntax of the corresponding option.
11123 If the specified expression is not valid, it is kept at its current
11127 @section hysteresis
11129 Grow first stream into second stream by connecting components.
11130 This makes it possible to build more robust edge masks.
11132 This filter accepts the following options:
11136 Set which planes will be processed as bitmap, unprocessed planes will be
11137 copied from first stream.
11138 By default value 0xf, all planes will be processed.
11141 Set threshold which is used in filtering. If pixel component value is higher than
11142 this value filter algorithm for connecting components is activated.
11143 By default value is 0.
11148 Detect video interlacing type.
11150 This filter tries to detect if the input frames are interlaced, progressive,
11151 top or bottom field first. It will also try to detect fields that are
11152 repeated between adjacent frames (a sign of telecine).
11154 Single frame detection considers only immediately adjacent frames when classifying each frame.
11155 Multiple frame detection incorporates the classification history of previous frames.
11157 The filter will log these metadata values:
11160 @item single.current_frame
11161 Detected type of current frame using single-frame detection. One of:
11162 ``tff'' (top field first), ``bff'' (bottom field first),
11163 ``progressive'', or ``undetermined''
11166 Cumulative number of frames detected as top field first using single-frame detection.
11169 Cumulative number of frames detected as top field first using multiple-frame detection.
11172 Cumulative number of frames detected as bottom field first using single-frame detection.
11174 @item multiple.current_frame
11175 Detected type of current frame using multiple-frame detection. One of:
11176 ``tff'' (top field first), ``bff'' (bottom field first),
11177 ``progressive'', or ``undetermined''
11180 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11182 @item single.progressive
11183 Cumulative number of frames detected as progressive using single-frame detection.
11185 @item multiple.progressive
11186 Cumulative number of frames detected as progressive using multiple-frame detection.
11188 @item single.undetermined
11189 Cumulative number of frames that could not be classified using single-frame detection.
11191 @item multiple.undetermined
11192 Cumulative number of frames that could not be classified using multiple-frame detection.
11194 @item repeated.current_frame
11195 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11197 @item repeated.neither
11198 Cumulative number of frames with no repeated field.
11201 Cumulative number of frames with the top field repeated from the previous frame's top field.
11203 @item repeated.bottom
11204 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11207 The filter accepts the following options:
11211 Set interlacing threshold.
11213 Set progressive threshold.
11215 Threshold for repeated field detection.
11217 Number of frames after which a given frame's contribution to the
11218 statistics is halved (i.e., it contributes only 0.5 to its
11219 classification). The default of 0 means that all frames seen are given
11220 full weight of 1.0 forever.
11221 @item analyze_interlaced_flag
11222 When this is not 0 then idet will use the specified number of frames to determine
11223 if the interlaced flag is accurate, it will not count undetermined frames.
11224 If the flag is found to be accurate it will be used without any further
11225 computations, if it is found to be inaccurate it will be cleared without any
11226 further computations. This allows inserting the idet filter as a low computational
11227 method to clean up the interlaced flag
11232 Deinterleave or interleave fields.
11234 This filter allows one to process interlaced images fields without
11235 deinterlacing them. Deinterleaving splits the input frame into 2
11236 fields (so called half pictures). Odd lines are moved to the top
11237 half of the output image, even lines to the bottom half.
11238 You can process (filter) them independently and then re-interleave them.
11240 The filter accepts the following options:
11244 @item chroma_mode, c
11245 @item alpha_mode, a
11246 Available values for @var{luma_mode}, @var{chroma_mode} and
11247 @var{alpha_mode} are:
11253 @item deinterleave, d
11254 Deinterleave fields, placing one above the other.
11256 @item interleave, i
11257 Interleave fields. Reverse the effect of deinterleaving.
11259 Default value is @code{none}.
11261 @item luma_swap, ls
11262 @item chroma_swap, cs
11263 @item alpha_swap, as
11264 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11269 Apply inflate effect to the video.
11271 This filter replaces the pixel by the local(3x3) average by taking into account
11272 only values higher than the pixel.
11274 It accepts the following options:
11281 Limit the maximum change for each plane, default is 65535.
11282 If 0, plane will remain unchanged.
11287 Simple interlacing filter from progressive contents. This interleaves upper (or
11288 lower) lines from odd frames with lower (or upper) lines from even frames,
11289 halving the frame rate and preserving image height.
11292 Original Original New Frame
11293 Frame 'j' Frame 'j+1' (tff)
11294 ========== =========== ==================
11295 Line 0 --------------------> Frame 'j' Line 0
11296 Line 1 Line 1 ----> Frame 'j+1' Line 1
11297 Line 2 ---------------------> Frame 'j' Line 2
11298 Line 3 Line 3 ----> Frame 'j+1' Line 3
11300 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11303 It accepts the following optional parameters:
11307 This determines whether the interlaced frame is taken from the even
11308 (tff - default) or odd (bff) lines of the progressive frame.
11311 Vertical lowpass filter to avoid twitter interlacing and
11312 reduce moire patterns.
11316 Disable vertical lowpass filter
11319 Enable linear filter (default)
11322 Enable complex filter. This will slightly less reduce twitter and moire
11323 but better retain detail and subjective sharpness impression.
11330 Deinterlace input video by applying Donald Graft's adaptive kernel
11331 deinterling. Work on interlaced parts of a video to produce
11332 progressive frames.
11334 The description of the accepted parameters follows.
11338 Set the threshold which affects the filter's tolerance when
11339 determining if a pixel line must be processed. It must be an integer
11340 in the range [0,255] and defaults to 10. A value of 0 will result in
11341 applying the process on every pixels.
11344 Paint pixels exceeding the threshold value to white if set to 1.
11348 Set the fields order. Swap fields if set to 1, leave fields alone if
11352 Enable additional sharpening if set to 1. Default is 0.
11355 Enable twoway sharpening if set to 1. Default is 0.
11358 @subsection Examples
11362 Apply default values:
11364 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11368 Enable additional sharpening:
11374 Paint processed pixels in white:
11382 Slowly update darker pixels.
11384 This filter makes short flashes of light appear longer.
11385 This filter accepts the following options:
11389 Set factor for decaying. Default is .95. Allowed range is from 0 to 1.
11392 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
11395 @section lenscorrection
11397 Correct radial lens distortion
11399 This filter can be used to correct for radial distortion as can result from the use
11400 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11401 one can use tools available for example as part of opencv or simply trial-and-error.
11402 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11403 and extract the k1 and k2 coefficients from the resulting matrix.
11405 Note that effectively the same filter is available in the open-source tools Krita and
11406 Digikam from the KDE project.
11408 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11409 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11410 brightness distribution, so you may want to use both filters together in certain
11411 cases, though you will have to take care of ordering, i.e. whether vignetting should
11412 be applied before or after lens correction.
11414 @subsection Options
11416 The filter accepts the following options:
11420 Relative x-coordinate of the focal point of the image, and thereby the center of the
11421 distortion. This value has a range [0,1] and is expressed as fractions of the image
11422 width. Default is 0.5.
11424 Relative y-coordinate of the focal point of the image, and thereby the center of the
11425 distortion. This value has a range [0,1] and is expressed as fractions of the image
11426 height. Default is 0.5.
11428 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11429 no correction. Default is 0.
11431 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11432 0 means no correction. Default is 0.
11435 The formula that generates the correction is:
11437 @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)
11439 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11440 distances from the focal point in the source and target images, respectively.
11444 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11446 The @code{lensfun} filter requires the camera make, camera model, and lens model
11447 to apply the lens correction. The filter will load the lensfun database and
11448 query it to find the corresponding camera and lens entries in the database. As
11449 long as these entries can be found with the given options, the filter can
11450 perform corrections on frames. Note that incomplete strings will result in the
11451 filter choosing the best match with the given options, and the filter will
11452 output the chosen camera and lens models (logged with level "info"). You must
11453 provide the make, camera model, and lens model as they are required.
11455 The filter accepts the following options:
11459 The make of the camera (for example, "Canon"). This option is required.
11462 The model of the camera (for example, "Canon EOS 100D"). This option is
11466 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11467 option is required.
11470 The type of correction to apply. The following values are valid options:
11474 Enables fixing lens vignetting.
11477 Enables fixing lens geometry. This is the default.
11480 Enables fixing chromatic aberrations.
11483 Enables fixing lens vignetting and lens geometry.
11486 Enables fixing lens vignetting and chromatic aberrations.
11489 Enables fixing both lens geometry and chromatic aberrations.
11492 Enables all possible corrections.
11496 The focal length of the image/video (zoom; expected constant for video). For
11497 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11498 range should be chosen when using that lens. Default 18.
11501 The aperture of the image/video (expected constant for video). Note that
11502 aperture is only used for vignetting correction. Default 3.5.
11504 @item focus_distance
11505 The focus distance of the image/video (expected constant for video). Note that
11506 focus distance is only used for vignetting and only slightly affects the
11507 vignetting correction process. If unknown, leave it at the default value (which
11511 The scale factor which is applied after transformation. After correction the
11512 video is no longer necessarily rectangular. This parameter controls how much of
11513 the resulting image is visible. The value 0 means that a value will be chosen
11514 automatically such that there is little or no unmapped area in the output
11515 image. 1.0 means that no additional scaling is done. Lower values may result
11516 in more of the corrected image being visible, while higher values may avoid
11517 unmapped areas in the output.
11519 @item target_geometry
11520 The target geometry of the output image/video. The following values are valid
11524 @item rectilinear (default)
11527 @item equirectangular
11528 @item fisheye_orthographic
11529 @item fisheye_stereographic
11530 @item fisheye_equisolid
11531 @item fisheye_thoby
11534 Apply the reverse of image correction (instead of correcting distortion, apply
11537 @item interpolation
11538 The type of interpolation used when correcting distortion. The following values
11543 @item linear (default)
11548 @subsection Examples
11552 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11553 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11557 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
11561 Apply the same as before, but only for the first 5 seconds of video.
11564 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
11571 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11572 score between two input videos.
11574 The obtained VMAF score is printed through the logging system.
11576 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11577 After installing the library it can be enabled using:
11578 @code{./configure --enable-libvmaf --enable-version3}.
11579 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11581 The filter has following options:
11585 Set the model path which is to be used for SVM.
11586 Default value: @code{"vmaf_v0.6.1.pkl"}
11589 Set the file path to be used to store logs.
11592 Set the format of the log file (xml or json).
11594 @item enable_transform
11595 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11596 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11597 Default value: @code{false}
11600 Invokes the phone model which will generate VMAF scores higher than in the
11601 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11604 Enables computing psnr along with vmaf.
11607 Enables computing ssim along with vmaf.
11610 Enables computing ms_ssim along with vmaf.
11613 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11616 Set number of threads to be used when computing vmaf.
11619 Set interval for frame subsampling used when computing vmaf.
11621 @item enable_conf_interval
11622 Enables confidence interval.
11625 This filter also supports the @ref{framesync} options.
11627 On the below examples the input file @file{main.mpg} being processed is
11628 compared with the reference file @file{ref.mpg}.
11631 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11634 Example with options:
11636 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11641 Limits the pixel components values to the specified range [min, max].
11643 The filter accepts the following options:
11647 Lower bound. Defaults to the lowest allowed value for the input.
11650 Upper bound. Defaults to the highest allowed value for the input.
11653 Specify which planes will be processed. Defaults to all available.
11660 The filter accepts the following options:
11664 Set the number of loops. Setting this value to -1 will result in infinite loops.
11668 Set maximal size in number of frames. Default is 0.
11671 Set first frame of loop. Default is 0.
11674 @subsection Examples
11678 Loop single first frame infinitely:
11680 loop=loop=-1:size=1:start=0
11684 Loop single first frame 10 times:
11686 loop=loop=10:size=1:start=0
11690 Loop 10 first frames 5 times:
11692 loop=loop=5:size=10:start=0
11698 Apply a 1D LUT to an input video.
11700 The filter accepts the following options:
11704 Set the 1D LUT file name.
11706 Currently supported formats:
11715 Select interpolation mode.
11717 Available values are:
11721 Use values from the nearest defined point.
11723 Interpolate values using the linear interpolation.
11725 Interpolate values using the cosine interpolation.
11727 Interpolate values using the cubic interpolation.
11729 Interpolate values using the spline interpolation.
11736 Apply a 3D LUT to an input video.
11738 The filter accepts the following options:
11742 Set the 3D LUT file name.
11744 Currently supported formats:
11758 Select interpolation mode.
11760 Available values are:
11764 Use values from the nearest defined point.
11766 Interpolate values using the 8 points defining a cube.
11768 Interpolate values using a tetrahedron.
11772 This filter also supports the @ref{framesync} options.
11776 Turn certain luma values into transparency.
11778 The filter accepts the following options:
11782 Set the luma which will be used as base for transparency.
11783 Default value is @code{0}.
11786 Set the range of luma values to be keyed out.
11787 Default value is @code{0}.
11790 Set the range of softness. Default value is @code{0}.
11791 Use this to control gradual transition from zero to full transparency.
11794 @section lut, lutrgb, lutyuv
11796 Compute a look-up table for binding each pixel component input value
11797 to an output value, and apply it to the input video.
11799 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11800 to an RGB input video.
11802 These filters accept the following parameters:
11805 set first pixel component expression
11807 set second pixel component expression
11809 set third pixel component expression
11811 set fourth pixel component expression, corresponds to the alpha component
11814 set red component expression
11816 set green component expression
11818 set blue component expression
11820 alpha component expression
11823 set Y/luminance component expression
11825 set U/Cb component expression
11827 set V/Cr component expression
11830 Each of them specifies the expression to use for computing the lookup table for
11831 the corresponding pixel component values.
11833 The exact component associated to each of the @var{c*} options depends on the
11836 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11837 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11839 The expressions can contain the following constants and functions:
11844 The input width and height.
11847 The input value for the pixel component.
11850 The input value, clipped to the @var{minval}-@var{maxval} range.
11853 The maximum value for the pixel component.
11856 The minimum value for the pixel component.
11859 The negated value for the pixel component value, clipped to the
11860 @var{minval}-@var{maxval} range; it corresponds to the expression
11861 "maxval-clipval+minval".
11864 The computed value in @var{val}, clipped to the
11865 @var{minval}-@var{maxval} range.
11867 @item gammaval(gamma)
11868 The computed gamma correction value of the pixel component value,
11869 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11871 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11875 All expressions default to "val".
11877 @subsection Examples
11881 Negate input video:
11883 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11884 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11887 The above is the same as:
11889 lutrgb="r=negval:g=negval:b=negval"
11890 lutyuv="y=negval:u=negval:v=negval"
11900 Remove chroma components, turning the video into a graytone image:
11902 lutyuv="u=128:v=128"
11906 Apply a luma burning effect:
11912 Remove green and blue components:
11918 Set a constant alpha channel value on input:
11920 format=rgba,lutrgb=a="maxval-minval/2"
11924 Correct luminance gamma by a factor of 0.5:
11926 lutyuv=y=gammaval(0.5)
11930 Discard least significant bits of luma:
11932 lutyuv=y='bitand(val, 128+64+32)'
11936 Technicolor like effect:
11938 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11942 @section lut2, tlut2
11944 The @code{lut2} filter takes two input streams and outputs one
11947 The @code{tlut2} (time lut2) filter takes two consecutive frames
11948 from one single stream.
11950 This filter accepts the following parameters:
11953 set first pixel component expression
11955 set second pixel component expression
11957 set third pixel component expression
11959 set fourth pixel component expression, corresponds to the alpha component
11962 set output bit depth, only available for @code{lut2} filter. By default is 0,
11963 which means bit depth is automatically picked from first input format.
11966 Each of them specifies the expression to use for computing the lookup table for
11967 the corresponding pixel component values.
11969 The exact component associated to each of the @var{c*} options depends on the
11972 The expressions can contain the following constants:
11977 The input width and height.
11980 The first input value for the pixel component.
11983 The second input value for the pixel component.
11986 The first input video bit depth.
11989 The second input video bit depth.
11992 All expressions default to "x".
11994 @subsection Examples
11998 Highlight differences between two RGB video streams:
12000 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)'
12004 Highlight differences between two YUV video streams:
12006 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)'
12010 Show max difference between two video streams:
12012 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)))'
12016 @section maskedclamp
12018 Clamp the first input stream with the second input and third input stream.
12020 Returns the value of first stream to be between second input
12021 stream - @code{undershoot} and third input stream + @code{overshoot}.
12023 This filter accepts the following options:
12026 Default value is @code{0}.
12029 Default value is @code{0}.
12032 Set which planes will be processed as bitmap, unprocessed planes will be
12033 copied from first stream.
12034 By default value 0xf, all planes will be processed.
12037 @section maskedmerge
12039 Merge the first input stream with the second input stream using per pixel
12040 weights in the third input stream.
12042 A value of 0 in the third stream pixel component means that pixel component
12043 from first stream is returned unchanged, while maximum value (eg. 255 for
12044 8-bit videos) means that pixel component from second stream is returned
12045 unchanged. Intermediate values define the amount of merging between both
12046 input stream's pixel components.
12048 This filter accepts the following options:
12051 Set which planes will be processed as bitmap, unprocessed planes will be
12052 copied from first stream.
12053 By default value 0xf, all planes will be processed.
12057 Create mask from input video.
12059 For example it is useful to create motion masks after @code{tblend} filter.
12061 This filter accepts the following options:
12065 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
12068 Set high threshold. Any pixel component higher than this value will be set to max value
12069 allowed for current pixel format.
12072 Set planes to filter, by default all available planes are filtered.
12075 Fill all frame pixels with this value.
12078 Set max average pixel value for frame. If sum of all pixel components is higher that this
12079 average, output frame will be completely filled with value set by @var{fill} option.
12080 Typically useful for scene changes when used in combination with @code{tblend} filter.
12085 Apply motion-compensation deinterlacing.
12087 It needs one field per frame as input and must thus be used together
12088 with yadif=1/3 or equivalent.
12090 This filter accepts the following options:
12093 Set the deinterlacing mode.
12095 It accepts one of the following values:
12100 use iterative motion estimation
12102 like @samp{slow}, but use multiple reference frames.
12104 Default value is @samp{fast}.
12107 Set the picture field parity assumed for the input video. It must be
12108 one of the following values:
12112 assume top field first
12114 assume bottom field first
12117 Default value is @samp{bff}.
12120 Set per-block quantization parameter (QP) used by the internal
12123 Higher values should result in a smoother motion vector field but less
12124 optimal individual vectors. Default value is 1.
12127 @section mergeplanes
12129 Merge color channel components from several video streams.
12131 The filter accepts up to 4 input streams, and merge selected input
12132 planes to the output video.
12134 This filter accepts the following options:
12137 Set input to output plane mapping. Default is @code{0}.
12139 The mappings is specified as a bitmap. It should be specified as a
12140 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12141 mapping for the first plane of the output stream. 'A' sets the number of
12142 the input stream to use (from 0 to 3), and 'a' the plane number of the
12143 corresponding input to use (from 0 to 3). The rest of the mappings is
12144 similar, 'Bb' describes the mapping for the output stream second
12145 plane, 'Cc' describes the mapping for the output stream third plane and
12146 'Dd' describes the mapping for the output stream fourth plane.
12149 Set output pixel format. Default is @code{yuva444p}.
12152 @subsection Examples
12156 Merge three gray video streams of same width and height into single video stream:
12158 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12162 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12164 [a0][a1]mergeplanes=0x00010210:yuva444p
12168 Swap Y and A plane in yuva444p stream:
12170 format=yuva444p,mergeplanes=0x03010200:yuva444p
12174 Swap U and V plane in yuv420p stream:
12176 format=yuv420p,mergeplanes=0x000201:yuv420p
12180 Cast a rgb24 clip to yuv444p:
12182 format=rgb24,mergeplanes=0x000102:yuv444p
12188 Estimate and export motion vectors using block matching algorithms.
12189 Motion vectors are stored in frame side data to be used by other filters.
12191 This filter accepts the following options:
12194 Specify the motion estimation method. Accepts one of the following values:
12198 Exhaustive search algorithm.
12200 Three step search algorithm.
12202 Two dimensional logarithmic search algorithm.
12204 New three step search algorithm.
12206 Four step search algorithm.
12208 Diamond search algorithm.
12210 Hexagon-based search algorithm.
12212 Enhanced predictive zonal search algorithm.
12214 Uneven multi-hexagon search algorithm.
12216 Default value is @samp{esa}.
12219 Macroblock size. Default @code{16}.
12222 Search parameter. Default @code{7}.
12225 @section midequalizer
12227 Apply Midway Image Equalization effect using two video streams.
12229 Midway Image Equalization adjusts a pair of images to have the same
12230 histogram, while maintaining their dynamics as much as possible. It's
12231 useful for e.g. matching exposures from a pair of stereo cameras.
12233 This filter has two inputs and one output, which must be of same pixel format, but
12234 may be of different sizes. The output of filter is first input adjusted with
12235 midway histogram of both inputs.
12237 This filter accepts the following option:
12241 Set which planes to process. Default is @code{15}, which is all available planes.
12244 @section minterpolate
12246 Convert the video to specified frame rate using motion interpolation.
12248 This filter accepts the following options:
12251 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}.
12254 Motion interpolation mode. Following values are accepted:
12257 Duplicate previous or next frame for interpolating new ones.
12259 Blend source frames. Interpolated frame is mean of previous and next frames.
12261 Motion compensated interpolation. Following options are effective when this mode is selected:
12265 Motion compensation mode. Following values are accepted:
12268 Overlapped block motion compensation.
12270 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12272 Default mode is @samp{obmc}.
12275 Motion estimation mode. Following values are accepted:
12278 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12280 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12282 Default mode is @samp{bilat}.
12285 The algorithm to be used for motion estimation. Following values are accepted:
12288 Exhaustive search algorithm.
12290 Three step search algorithm.
12292 Two dimensional logarithmic search algorithm.
12294 New three step search algorithm.
12296 Four step search algorithm.
12298 Diamond search algorithm.
12300 Hexagon-based search algorithm.
12302 Enhanced predictive zonal search algorithm.
12304 Uneven multi-hexagon search algorithm.
12306 Default algorithm is @samp{epzs}.
12309 Macroblock size. Default @code{16}.
12312 Motion estimation search parameter. Default @code{32}.
12315 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).
12320 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:
12323 Disable scene change detection.
12325 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12327 Default method is @samp{fdiff}.
12329 @item scd_threshold
12330 Scene change detection threshold. Default is @code{5.0}.
12335 Mix several video input streams into one video stream.
12337 A description of the accepted options follows.
12341 The number of inputs. If unspecified, it defaults to 2.
12344 Specify weight of each input video stream as sequence.
12345 Each weight is separated by space. If number of weights
12346 is smaller than number of @var{frames} last specified
12347 weight will be used for all remaining unset weights.
12350 Specify scale, if it is set it will be multiplied with sum
12351 of each weight multiplied with pixel values to give final destination
12352 pixel value. By default @var{scale} is auto scaled to sum of weights.
12355 Specify how end of stream is determined.
12358 The duration of the longest input. (default)
12361 The duration of the shortest input.
12364 The duration of the first input.
12368 @section mpdecimate
12370 Drop frames that do not differ greatly from the previous frame in
12371 order to reduce frame rate.
12373 The main use of this filter is for very-low-bitrate encoding
12374 (e.g. streaming over dialup modem), but it could in theory be used for
12375 fixing movies that were inverse-telecined incorrectly.
12377 A description of the accepted options follows.
12381 Set the maximum number of consecutive frames which can be dropped (if
12382 positive), or the minimum interval between dropped frames (if
12383 negative). If the value is 0, the frame is dropped disregarding the
12384 number of previous sequentially dropped frames.
12386 Default value is 0.
12391 Set the dropping threshold values.
12393 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12394 represent actual pixel value differences, so a threshold of 64
12395 corresponds to 1 unit of difference for each pixel, or the same spread
12396 out differently over the block.
12398 A frame is a candidate for dropping if no 8x8 blocks differ by more
12399 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12400 meaning the whole image) differ by more than a threshold of @option{lo}.
12402 Default value for @option{hi} is 64*12, default value for @option{lo} is
12403 64*5, and default value for @option{frac} is 0.33.
12409 Negate (invert) the input video.
12411 It accepts the following option:
12416 With value 1, it negates the alpha component, if present. Default value is 0.
12422 Denoise frames using Non-Local Means algorithm.
12424 Each pixel is adjusted by looking for other pixels with similar contexts. This
12425 context similarity is defined by comparing their surrounding patches of size
12426 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12429 Note that the research area defines centers for patches, which means some
12430 patches will be made of pixels outside that research area.
12432 The filter accepts the following options.
12436 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12439 Set patch size. Default is 7. Must be odd number in range [0, 99].
12442 Same as @option{p} but for chroma planes.
12444 The default value is @var{0} and means automatic.
12447 Set research size. Default is 15. Must be odd number in range [0, 99].
12450 Same as @option{r} but for chroma planes.
12452 The default value is @var{0} and means automatic.
12457 Deinterlace video using neural network edge directed interpolation.
12459 This filter accepts the following options:
12463 Mandatory option, without binary file filter can not work.
12464 Currently file can be found here:
12465 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12468 Set which frames to deinterlace, by default it is @code{all}.
12469 Can be @code{all} or @code{interlaced}.
12472 Set mode of operation.
12474 Can be one of the following:
12478 Use frame flags, both fields.
12480 Use frame flags, single field.
12482 Use top field only.
12484 Use bottom field only.
12486 Use both fields, top first.
12488 Use both fields, bottom first.
12492 Set which planes to process, by default filter process all frames.
12495 Set size of local neighborhood around each pixel, used by the predictor neural
12498 Can be one of the following:
12511 Set the number of neurons in predictor neural network.
12512 Can be one of the following:
12523 Controls the number of different neural network predictions that are blended
12524 together to compute the final output value. Can be @code{fast}, default or
12528 Set which set of weights to use in the predictor.
12529 Can be one of the following:
12533 weights trained to minimize absolute error
12535 weights trained to minimize squared error
12539 Controls whether or not the prescreener neural network is used to decide
12540 which pixels should be processed by the predictor neural network and which
12541 can be handled by simple cubic interpolation.
12542 The prescreener is trained to know whether cubic interpolation will be
12543 sufficient for a pixel or whether it should be predicted by the predictor nn.
12544 The computational complexity of the prescreener nn is much less than that of
12545 the predictor nn. Since most pixels can be handled by cubic interpolation,
12546 using the prescreener generally results in much faster processing.
12547 The prescreener is pretty accurate, so the difference between using it and not
12548 using it is almost always unnoticeable.
12550 Can be one of the following:
12558 Default is @code{new}.
12561 Set various debugging flags.
12566 Force libavfilter not to use any of the specified pixel formats for the
12567 input to the next filter.
12569 It accepts the following parameters:
12573 A '|'-separated list of pixel format names, such as
12574 pix_fmts=yuv420p|monow|rgb24".
12578 @subsection Examples
12582 Force libavfilter to use a format different from @var{yuv420p} for the
12583 input to the vflip filter:
12585 noformat=pix_fmts=yuv420p,vflip
12589 Convert the input video to any of the formats not contained in the list:
12591 noformat=yuv420p|yuv444p|yuv410p
12597 Add noise on video input frame.
12599 The filter accepts the following options:
12607 Set noise seed for specific pixel component or all pixel components in case
12608 of @var{all_seed}. Default value is @code{123457}.
12610 @item all_strength, alls
12611 @item c0_strength, c0s
12612 @item c1_strength, c1s
12613 @item c2_strength, c2s
12614 @item c3_strength, c3s
12615 Set noise strength for specific pixel component or all pixel components in case
12616 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12618 @item all_flags, allf
12619 @item c0_flags, c0f
12620 @item c1_flags, c1f
12621 @item c2_flags, c2f
12622 @item c3_flags, c3f
12623 Set pixel component flags or set flags for all components if @var{all_flags}.
12624 Available values for component flags are:
12627 averaged temporal noise (smoother)
12629 mix random noise with a (semi)regular pattern
12631 temporal noise (noise pattern changes between frames)
12633 uniform noise (gaussian otherwise)
12637 @subsection Examples
12639 Add temporal and uniform noise to input video:
12641 noise=alls=20:allf=t+u
12646 Normalize RGB video (aka histogram stretching, contrast stretching).
12647 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12649 For each channel of each frame, the filter computes the input range and maps
12650 it linearly to the user-specified output range. The output range defaults
12651 to the full dynamic range from pure black to pure white.
12653 Temporal smoothing can be used on the input range to reduce flickering (rapid
12654 changes in brightness) caused when small dark or bright objects enter or leave
12655 the scene. This is similar to the auto-exposure (automatic gain control) on a
12656 video camera, and, like a video camera, it may cause a period of over- or
12657 under-exposure of the video.
12659 The R,G,B channels can be normalized independently, which may cause some
12660 color shifting, or linked together as a single channel, which prevents
12661 color shifting. Linked normalization preserves hue. Independent normalization
12662 does not, so it can be used to remove some color casts. Independent and linked
12663 normalization can be combined in any ratio.
12665 The normalize filter accepts the following options:
12670 Colors which define the output range. The minimum input value is mapped to
12671 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12672 The defaults are black and white respectively. Specifying white for
12673 @var{blackpt} and black for @var{whitept} will give color-inverted,
12674 normalized video. Shades of grey can be used to reduce the dynamic range
12675 (contrast). Specifying saturated colors here can create some interesting
12679 The number of previous frames to use for temporal smoothing. The input range
12680 of each channel is smoothed using a rolling average over the current frame
12681 and the @var{smoothing} previous frames. The default is 0 (no temporal
12685 Controls the ratio of independent (color shifting) channel normalization to
12686 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12687 independent. Defaults to 1.0 (fully independent).
12690 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12691 expensive no-op. Defaults to 1.0 (full strength).
12695 @subsection Examples
12697 Stretch video contrast to use the full dynamic range, with no temporal
12698 smoothing; may flicker depending on the source content:
12700 normalize=blackpt=black:whitept=white:smoothing=0
12703 As above, but with 50 frames of temporal smoothing; flicker should be
12704 reduced, depending on the source content:
12706 normalize=blackpt=black:whitept=white:smoothing=50
12709 As above, but with hue-preserving linked channel normalization:
12711 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12714 As above, but with half strength:
12716 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12719 Map the darkest input color to red, the brightest input color to cyan:
12721 normalize=blackpt=red:whitept=cyan
12726 Pass the video source unchanged to the output.
12729 Optical Character Recognition
12731 This filter uses Tesseract for optical character recognition. To enable
12732 compilation of this filter, you need to configure FFmpeg with
12733 @code{--enable-libtesseract}.
12735 It accepts the following options:
12739 Set datapath to tesseract data. Default is to use whatever was
12740 set at installation.
12743 Set language, default is "eng".
12746 Set character whitelist.
12749 Set character blacklist.
12752 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12756 Apply a video transform using libopencv.
12758 To enable this filter, install the libopencv library and headers and
12759 configure FFmpeg with @code{--enable-libopencv}.
12761 It accepts the following parameters:
12766 The name of the libopencv filter to apply.
12768 @item filter_params
12769 The parameters to pass to the libopencv filter. If not specified, the default
12770 values are assumed.
12774 Refer to the official libopencv documentation for more precise
12776 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12778 Several libopencv filters are supported; see the following subsections.
12783 Dilate an image by using a specific structuring element.
12784 It corresponds to the libopencv function @code{cvDilate}.
12786 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12788 @var{struct_el} represents a structuring element, and has the syntax:
12789 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12791 @var{cols} and @var{rows} represent the number of columns and rows of
12792 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12793 point, and @var{shape} the shape for the structuring element. @var{shape}
12794 must be "rect", "cross", "ellipse", or "custom".
12796 If the value for @var{shape} is "custom", it must be followed by a
12797 string of the form "=@var{filename}". The file with name
12798 @var{filename} is assumed to represent a binary image, with each
12799 printable character corresponding to a bright pixel. When a custom
12800 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12801 or columns and rows of the read file are assumed instead.
12803 The default value for @var{struct_el} is "3x3+0x0/rect".
12805 @var{nb_iterations} specifies the number of times the transform is
12806 applied to the image, and defaults to 1.
12810 # Use the default values
12813 # Dilate using a structuring element with a 5x5 cross, iterating two times
12814 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12816 # Read the shape from the file diamond.shape, iterating two times.
12817 # The file diamond.shape may contain a pattern of characters like this
12823 # The specified columns and rows are ignored
12824 # but the anchor point coordinates are not
12825 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12830 Erode an image by using a specific structuring element.
12831 It corresponds to the libopencv function @code{cvErode}.
12833 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12834 with the same syntax and semantics as the @ref{dilate} filter.
12838 Smooth the input video.
12840 The filter takes the following parameters:
12841 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12843 @var{type} is the type of smooth filter to apply, and must be one of
12844 the following values: "blur", "blur_no_scale", "median", "gaussian",
12845 or "bilateral". The default value is "gaussian".
12847 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12848 depend on the smooth type. @var{param1} and
12849 @var{param2} accept integer positive values or 0. @var{param3} and
12850 @var{param4} accept floating point values.
12852 The default value for @var{param1} is 3. The default value for the
12853 other parameters is 0.
12855 These parameters correspond to the parameters assigned to the
12856 libopencv function @code{cvSmooth}.
12858 @section oscilloscope
12860 2D Video Oscilloscope.
12862 Useful to measure spatial impulse, step responses, chroma delays, etc.
12864 It accepts the following parameters:
12868 Set scope center x position.
12871 Set scope center y position.
12874 Set scope size, relative to frame diagonal.
12877 Set scope tilt/rotation.
12883 Set trace center x position.
12886 Set trace center y position.
12889 Set trace width, relative to width of frame.
12892 Set trace height, relative to height of frame.
12895 Set which components to trace. By default it traces first three components.
12898 Draw trace grid. By default is enabled.
12901 Draw some statistics. By default is enabled.
12904 Draw scope. By default is enabled.
12907 @subsection Examples
12911 Inspect full first row of video frame.
12913 oscilloscope=x=0.5:y=0:s=1
12917 Inspect full last row of video frame.
12919 oscilloscope=x=0.5:y=1:s=1
12923 Inspect full 5th line of video frame of height 1080.
12925 oscilloscope=x=0.5:y=5/1080:s=1
12929 Inspect full last column of video frame.
12931 oscilloscope=x=1:y=0.5:s=1:t=1
12939 Overlay one video on top of another.
12941 It takes two inputs and has one output. The first input is the "main"
12942 video on which the second input is overlaid.
12944 It accepts the following parameters:
12946 A description of the accepted options follows.
12951 Set the expression for the x and y coordinates of the overlaid video
12952 on the main video. Default value is "0" for both expressions. In case
12953 the expression is invalid, it is set to a huge value (meaning that the
12954 overlay will not be displayed within the output visible area).
12957 See @ref{framesync}.
12960 Set when the expressions for @option{x}, and @option{y} are evaluated.
12962 It accepts the following values:
12965 only evaluate expressions once during the filter initialization or
12966 when a command is processed
12969 evaluate expressions for each incoming frame
12972 Default value is @samp{frame}.
12975 See @ref{framesync}.
12978 Set the format for the output video.
12980 It accepts the following values:
12983 force YUV420 output
12986 force YUV422 output
12989 force YUV444 output
12992 force packed RGB output
12995 force planar RGB output
12998 automatically pick format
13001 Default value is @samp{yuv420}.
13004 See @ref{framesync}.
13007 Set format of alpha of the overlaid video, it can be @var{straight} or
13008 @var{premultiplied}. Default is @var{straight}.
13011 The @option{x}, and @option{y} expressions can contain the following
13017 The main input width and height.
13021 The overlay input width and height.
13025 The computed values for @var{x} and @var{y}. They are evaluated for
13030 horizontal and vertical chroma subsample values of the output
13031 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
13035 the number of input frame, starting from 0
13038 the position in the file of the input frame, NAN if unknown
13041 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
13045 This filter also supports the @ref{framesync} options.
13047 Note that the @var{n}, @var{pos}, @var{t} variables are available only
13048 when evaluation is done @emph{per frame}, and will evaluate to NAN
13049 when @option{eval} is set to @samp{init}.
13051 Be aware that frames are taken from each input video in timestamp
13052 order, hence, if their initial timestamps differ, it is a good idea
13053 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
13054 have them begin in the same zero timestamp, as the example for
13055 the @var{movie} filter does.
13057 You can chain together more overlays but you should test the
13058 efficiency of such approach.
13060 @subsection Commands
13062 This filter supports the following commands:
13066 Modify the x and y of the overlay input.
13067 The command accepts the same syntax of the corresponding option.
13069 If the specified expression is not valid, it is kept at its current
13073 @subsection Examples
13077 Draw the overlay at 10 pixels from the bottom right corner of the main
13080 overlay=main_w-overlay_w-10:main_h-overlay_h-10
13083 Using named options the example above becomes:
13085 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
13089 Insert a transparent PNG logo in the bottom left corner of the input,
13090 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
13092 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
13096 Insert 2 different transparent PNG logos (second logo on bottom
13097 right corner) using the @command{ffmpeg} tool:
13099 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
13103 Add a transparent color layer on top of the main video; @code{WxH}
13104 must specify the size of the main input to the overlay filter:
13106 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
13110 Play an original video and a filtered version (here with the deshake
13111 filter) side by side using the @command{ffplay} tool:
13113 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13116 The above command is the same as:
13118 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13122 Make a sliding overlay appearing from the left to the right top part of the
13123 screen starting since time 2:
13125 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13129 Compose output by putting two input videos side to side:
13131 ffmpeg -i left.avi -i right.avi -filter_complex "
13132 nullsrc=size=200x100 [background];
13133 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13134 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13135 [background][left] overlay=shortest=1 [background+left];
13136 [background+left][right] overlay=shortest=1:x=100 [left+right]
13141 Mask 10-20 seconds of a video by applying the delogo filter to a section
13143 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13144 -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]'
13149 Chain several overlays in cascade:
13151 nullsrc=s=200x200 [bg];
13152 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13153 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13154 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13155 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13156 [in3] null, [mid2] overlay=100:100 [out0]
13163 Apply Overcomplete Wavelet denoiser.
13165 The filter accepts the following options:
13171 Larger depth values will denoise lower frequency components more, but
13172 slow down filtering.
13174 Must be an int in the range 8-16, default is @code{8}.
13176 @item luma_strength, ls
13179 Must be a double value in the range 0-1000, default is @code{1.0}.
13181 @item chroma_strength, cs
13182 Set chroma strength.
13184 Must be a double value in the range 0-1000, default is @code{1.0}.
13190 Add paddings to the input image, and place the original input at the
13191 provided @var{x}, @var{y} coordinates.
13193 It accepts the following parameters:
13198 Specify an expression for the size of the output image with the
13199 paddings added. If the value for @var{width} or @var{height} is 0, the
13200 corresponding input size is used for the output.
13202 The @var{width} expression can reference the value set by the
13203 @var{height} expression, and vice versa.
13205 The default value of @var{width} and @var{height} is 0.
13209 Specify the offsets to place the input image at within the padded area,
13210 with respect to the top/left border of the output image.
13212 The @var{x} expression can reference the value set by the @var{y}
13213 expression, and vice versa.
13215 The default value of @var{x} and @var{y} is 0.
13217 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13218 so the input image is centered on the padded area.
13221 Specify the color of the padded area. For the syntax of this option,
13222 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13223 manual,ffmpeg-utils}.
13225 The default value of @var{color} is "black".
13228 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13230 It accepts the following values:
13234 Only evaluate expressions once during the filter initialization or when
13235 a command is processed.
13238 Evaluate expressions for each incoming frame.
13242 Default value is @samp{init}.
13245 Pad to aspect instead to a resolution.
13249 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13250 options are expressions containing the following constants:
13255 The input video width and height.
13259 These are the same as @var{in_w} and @var{in_h}.
13263 The output width and height (the size of the padded area), as
13264 specified by the @var{width} and @var{height} expressions.
13268 These are the same as @var{out_w} and @var{out_h}.
13272 The x and y offsets as specified by the @var{x} and @var{y}
13273 expressions, or NAN if not yet specified.
13276 same as @var{iw} / @var{ih}
13279 input sample aspect ratio
13282 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13286 The horizontal and vertical chroma subsample values. For example for the
13287 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13290 @subsection Examples
13294 Add paddings with the color "violet" to the input video. The output video
13295 size is 640x480, and the top-left corner of the input video is placed at
13298 pad=640:480:0:40:violet
13301 The example above is equivalent to the following command:
13303 pad=width=640:height=480:x=0:y=40:color=violet
13307 Pad the input to get an output with dimensions increased by 3/2,
13308 and put the input video at the center of the padded area:
13310 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13314 Pad the input to get a squared output with size equal to the maximum
13315 value between the input width and height, and put the input video at
13316 the center of the padded area:
13318 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13322 Pad the input to get a final w/h ratio of 16:9:
13324 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13328 In case of anamorphic video, in order to set the output display aspect
13329 correctly, it is necessary to use @var{sar} in the expression,
13330 according to the relation:
13332 (ih * X / ih) * sar = output_dar
13333 X = output_dar / sar
13336 Thus the previous example needs to be modified to:
13338 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13342 Double the output size and put the input video in the bottom-right
13343 corner of the output padded area:
13345 pad="2*iw:2*ih:ow-iw:oh-ih"
13349 @anchor{palettegen}
13350 @section palettegen
13352 Generate one palette for a whole video stream.
13354 It accepts the following options:
13358 Set the maximum number of colors to quantize in the palette.
13359 Note: the palette will still contain 256 colors; the unused palette entries
13362 @item reserve_transparent
13363 Create a palette of 255 colors maximum and reserve the last one for
13364 transparency. Reserving the transparency color is useful for GIF optimization.
13365 If not set, the maximum of colors in the palette will be 256. You probably want
13366 to disable this option for a standalone image.
13369 @item transparency_color
13370 Set the color that will be used as background for transparency.
13373 Set statistics mode.
13375 It accepts the following values:
13378 Compute full frame histograms.
13380 Compute histograms only for the part that differs from previous frame. This
13381 might be relevant to give more importance to the moving part of your input if
13382 the background is static.
13384 Compute new histogram for each frame.
13387 Default value is @var{full}.
13390 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13391 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13392 color quantization of the palette. This information is also visible at
13393 @var{info} logging level.
13395 @subsection Examples
13399 Generate a representative palette of a given video using @command{ffmpeg}:
13401 ffmpeg -i input.mkv -vf palettegen palette.png
13405 @section paletteuse
13407 Use a palette to downsample an input video stream.
13409 The filter takes two inputs: one video stream and a palette. The palette must
13410 be a 256 pixels image.
13412 It accepts the following options:
13416 Select dithering mode. Available algorithms are:
13419 Ordered 8x8 bayer dithering (deterministic)
13421 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13422 Note: this dithering is sometimes considered "wrong" and is included as a
13424 @item floyd_steinberg
13425 Floyd and Steingberg dithering (error diffusion)
13427 Frankie Sierra dithering v2 (error diffusion)
13429 Frankie Sierra dithering v2 "Lite" (error diffusion)
13432 Default is @var{sierra2_4a}.
13435 When @var{bayer} dithering is selected, this option defines the scale of the
13436 pattern (how much the crosshatch pattern is visible). A low value means more
13437 visible pattern for less banding, and higher value means less visible pattern
13438 at the cost of more banding.
13440 The option must be an integer value in the range [0,5]. Default is @var{2}.
13443 If set, define the zone to process
13447 Only the changing rectangle will be reprocessed. This is similar to GIF
13448 cropping/offsetting compression mechanism. This option can be useful for speed
13449 if only a part of the image is changing, and has use cases such as limiting the
13450 scope of the error diffusal @option{dither} to the rectangle that bounds the
13451 moving scene (it leads to more deterministic output if the scene doesn't change
13452 much, and as a result less moving noise and better GIF compression).
13455 Default is @var{none}.
13458 Take new palette for each output frame.
13460 @item alpha_threshold
13461 Sets the alpha threshold for transparency. Alpha values above this threshold
13462 will be treated as completely opaque, and values below this threshold will be
13463 treated as completely transparent.
13465 The option must be an integer value in the range [0,255]. Default is @var{128}.
13468 @subsection Examples
13472 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13473 using @command{ffmpeg}:
13475 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13479 @section perspective
13481 Correct perspective of video not recorded perpendicular to the screen.
13483 A description of the accepted parameters follows.
13494 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13495 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13496 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13497 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13498 then the corners of the source will be sent to the specified coordinates.
13500 The expressions can use the following variables:
13505 the width and height of video frame.
13509 Output frame count.
13512 @item interpolation
13513 Set interpolation for perspective correction.
13515 It accepts the following values:
13521 Default value is @samp{linear}.
13524 Set interpretation of coordinate options.
13526 It accepts the following values:
13530 Send point in the source specified by the given coordinates to
13531 the corners of the destination.
13533 @item 1, destination
13535 Send the corners of the source to the point in the destination specified
13536 by the given coordinates.
13538 Default value is @samp{source}.
13542 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13544 It accepts the following values:
13547 only evaluate expressions once during the filter initialization or
13548 when a command is processed
13551 evaluate expressions for each incoming frame
13554 Default value is @samp{init}.
13559 Delay interlaced video by one field time so that the field order changes.
13561 The intended use is to fix PAL movies that have been captured with the
13562 opposite field order to the film-to-video transfer.
13564 A description of the accepted parameters follows.
13570 It accepts the following values:
13573 Capture field order top-first, transfer bottom-first.
13574 Filter will delay the bottom field.
13577 Capture field order bottom-first, transfer top-first.
13578 Filter will delay the top field.
13581 Capture and transfer with the same field order. This mode only exists
13582 for the documentation of the other options to refer to, but if you
13583 actually select it, the filter will faithfully do nothing.
13586 Capture field order determined automatically by field flags, transfer
13588 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13589 basis using field flags. If no field information is available,
13590 then this works just like @samp{u}.
13593 Capture unknown or varying, transfer opposite.
13594 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13595 analyzing the images and selecting the alternative that produces best
13596 match between the fields.
13599 Capture top-first, transfer unknown or varying.
13600 Filter selects among @samp{t} and @samp{p} using image analysis.
13603 Capture bottom-first, transfer unknown or varying.
13604 Filter selects among @samp{b} and @samp{p} using image analysis.
13607 Capture determined by field flags, transfer unknown or varying.
13608 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13609 image analysis. If no field information is available, then this works just
13610 like @samp{U}. This is the default mode.
13613 Both capture and transfer unknown or varying.
13614 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13618 @section pixdesctest
13620 Pixel format descriptor test filter, mainly useful for internal
13621 testing. The output video should be equal to the input video.
13625 format=monow, pixdesctest
13628 can be used to test the monowhite pixel format descriptor definition.
13632 Display sample values of color channels. Mainly useful for checking color
13633 and levels. Minimum supported resolution is 640x480.
13635 The filters accept the following options:
13639 Set scope X position, relative offset on X axis.
13642 Set scope Y position, relative offset on Y axis.
13651 Set window opacity. This window also holds statistics about pixel area.
13654 Set window X position, relative offset on X axis.
13657 Set window Y position, relative offset on Y axis.
13662 Enable the specified chain of postprocessing subfilters using libpostproc. This
13663 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13664 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13665 Each subfilter and some options have a short and a long name that can be used
13666 interchangeably, i.e. dr/dering are the same.
13668 The filters accept the following options:
13672 Set postprocessing subfilters string.
13675 All subfilters share common options to determine their scope:
13679 Honor the quality commands for this subfilter.
13682 Do chrominance filtering, too (default).
13685 Do luminance filtering only (no chrominance).
13688 Do chrominance filtering only (no luminance).
13691 These options can be appended after the subfilter name, separated by a '|'.
13693 Available subfilters are:
13696 @item hb/hdeblock[|difference[|flatness]]
13697 Horizontal deblocking filter
13700 Difference factor where higher values mean more deblocking (default: @code{32}).
13702 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13705 @item vb/vdeblock[|difference[|flatness]]
13706 Vertical deblocking filter
13709 Difference factor where higher values mean more deblocking (default: @code{32}).
13711 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13714 @item ha/hadeblock[|difference[|flatness]]
13715 Accurate horizontal deblocking filter
13718 Difference factor where higher values mean more deblocking (default: @code{32}).
13720 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13723 @item va/vadeblock[|difference[|flatness]]
13724 Accurate vertical deblocking filter
13727 Difference factor where higher values mean more deblocking (default: @code{32}).
13729 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13733 The horizontal and vertical deblocking filters share the difference and
13734 flatness values so you cannot set different horizontal and vertical
13738 @item h1/x1hdeblock
13739 Experimental horizontal deblocking filter
13741 @item v1/x1vdeblock
13742 Experimental vertical deblocking filter
13747 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13750 larger -> stronger filtering
13752 larger -> stronger filtering
13754 larger -> stronger filtering
13757 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13760 Stretch luminance to @code{0-255}.
13763 @item lb/linblenddeint
13764 Linear blend deinterlacing filter that deinterlaces the given block by
13765 filtering all lines with a @code{(1 2 1)} filter.
13767 @item li/linipoldeint
13768 Linear interpolating deinterlacing filter that deinterlaces the given block by
13769 linearly interpolating every second line.
13771 @item ci/cubicipoldeint
13772 Cubic interpolating deinterlacing filter deinterlaces the given block by
13773 cubically interpolating every second line.
13775 @item md/mediandeint
13776 Median deinterlacing filter that deinterlaces the given block by applying a
13777 median filter to every second line.
13779 @item fd/ffmpegdeint
13780 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13781 second line with a @code{(-1 4 2 4 -1)} filter.
13784 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13785 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13787 @item fq/forceQuant[|quantizer]
13788 Overrides the quantizer table from the input with the constant quantizer you
13796 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13799 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13802 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13805 @subsection Examples
13809 Apply horizontal and vertical deblocking, deringing and automatic
13810 brightness/contrast:
13816 Apply default filters without brightness/contrast correction:
13822 Apply default filters and temporal denoiser:
13824 pp=default/tmpnoise|1|2|3
13828 Apply deblocking on luminance only, and switch vertical deblocking on or off
13829 automatically depending on available CPU time:
13836 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13837 similar to spp = 6 with 7 point DCT, where only the center sample is
13840 The filter accepts the following options:
13844 Force a constant quantization parameter. It accepts an integer in range
13845 0 to 63. If not set, the filter will use the QP from the video stream
13849 Set thresholding mode. Available modes are:
13853 Set hard thresholding.
13855 Set soft thresholding (better de-ringing effect, but likely blurrier).
13857 Set medium thresholding (good results, default).
13861 @section premultiply
13862 Apply alpha premultiply effect to input video stream using first plane
13863 of second stream as alpha.
13865 Both streams must have same dimensions and same pixel format.
13867 The filter accepts the following option:
13871 Set which planes will be processed, unprocessed planes will be copied.
13872 By default value 0xf, all planes will be processed.
13875 Do not require 2nd input for processing, instead use alpha plane from input stream.
13879 Apply prewitt operator to input video stream.
13881 The filter accepts the following option:
13885 Set which planes will be processed, unprocessed planes will be copied.
13886 By default value 0xf, all planes will be processed.
13889 Set value which will be multiplied with filtered result.
13892 Set value which will be added to filtered result.
13895 @anchor{program_opencl}
13896 @section program_opencl
13898 Filter video using an OpenCL program.
13903 OpenCL program source file.
13906 Kernel name in program.
13909 Number of inputs to the filter. Defaults to 1.
13912 Size of output frames. Defaults to the same as the first input.
13916 The program source file must contain a kernel function with the given name,
13917 which will be run once for each plane of the output. Each run on a plane
13918 gets enqueued as a separate 2D global NDRange with one work-item for each
13919 pixel to be generated. The global ID offset for each work-item is therefore
13920 the coordinates of a pixel in the destination image.
13922 The kernel function needs to take the following arguments:
13925 Destination image, @var{__write_only image2d_t}.
13927 This image will become the output; the kernel should write all of it.
13929 Frame index, @var{unsigned int}.
13931 This is a counter starting from zero and increasing by one for each frame.
13933 Source images, @var{__read_only image2d_t}.
13935 These are the most recent images on each input. The kernel may read from
13936 them to generate the output, but they can't be written to.
13943 Copy the input to the output (output must be the same size as the input).
13945 __kernel void copy(__write_only image2d_t destination,
13946 unsigned int index,
13947 __read_only image2d_t source)
13949 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13951 int2 location = (int2)(get_global_id(0), get_global_id(1));
13953 float4 value = read_imagef(source, sampler, location);
13955 write_imagef(destination, location, value);
13960 Apply a simple transformation, rotating the input by an amount increasing
13961 with the index counter. Pixel values are linearly interpolated by the
13962 sampler, and the output need not have the same dimensions as the input.
13964 __kernel void rotate_image(__write_only image2d_t dst,
13965 unsigned int index,
13966 __read_only image2d_t src)
13968 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13969 CLK_FILTER_LINEAR);
13971 float angle = (float)index / 100.0f;
13973 float2 dst_dim = convert_float2(get_image_dim(dst));
13974 float2 src_dim = convert_float2(get_image_dim(src));
13976 float2 dst_cen = dst_dim / 2.0f;
13977 float2 src_cen = src_dim / 2.0f;
13979 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13981 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13983 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13984 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13986 src_pos = src_pos * src_dim / dst_dim;
13988 float2 src_loc = src_pos + src_cen;
13990 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13991 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13992 write_imagef(dst, dst_loc, 0.5f);
13994 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13999 Blend two inputs together, with the amount of each input used varying
14000 with the index counter.
14002 __kernel void blend_images(__write_only image2d_t dst,
14003 unsigned int index,
14004 __read_only image2d_t src1,
14005 __read_only image2d_t src2)
14007 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
14008 CLK_FILTER_LINEAR);
14010 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
14012 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
14013 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
14014 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
14016 float4 val1 = read_imagef(src1, sampler, src1_loc);
14017 float4 val2 = read_imagef(src2, sampler, src2_loc);
14019 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
14025 @section pseudocolor
14027 Alter frame colors in video with pseudocolors.
14029 This filter accept the following options:
14033 set pixel first component expression
14036 set pixel second component expression
14039 set pixel third component expression
14042 set pixel fourth component expression, corresponds to the alpha component
14045 set component to use as base for altering colors
14048 Each of them specifies the expression to use for computing the lookup table for
14049 the corresponding pixel component values.
14051 The expressions can contain the following constants and functions:
14056 The input width and height.
14059 The input value for the pixel component.
14061 @item ymin, umin, vmin, amin
14062 The minimum allowed component value.
14064 @item ymax, umax, vmax, amax
14065 The maximum allowed component value.
14068 All expressions default to "val".
14070 @subsection Examples
14074 Change too high luma values to gradient:
14076 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'"
14082 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
14083 Ratio) between two input videos.
14085 This filter takes in input two input videos, the first input is
14086 considered the "main" source and is passed unchanged to the
14087 output. The second input is used as a "reference" video for computing
14090 Both video inputs must have the same resolution and pixel format for
14091 this filter to work correctly. Also it assumes that both inputs
14092 have the same number of frames, which are compared one by one.
14094 The obtained average PSNR is printed through the logging system.
14096 The filter stores the accumulated MSE (mean squared error) of each
14097 frame, and at the end of the processing it is averaged across all frames
14098 equally, and the following formula is applied to obtain the PSNR:
14101 PSNR = 10*log10(MAX^2/MSE)
14104 Where MAX is the average of the maximum values of each component of the
14107 The description of the accepted parameters follows.
14110 @item stats_file, f
14111 If specified the filter will use the named file to save the PSNR of
14112 each individual frame. When filename equals "-" the data is sent to
14115 @item stats_version
14116 Specifies which version of the stats file format to use. Details of
14117 each format are written below.
14118 Default value is 1.
14120 @item stats_add_max
14121 Determines whether the max value is output to the stats log.
14122 Default value is 0.
14123 Requires stats_version >= 2. If this is set and stats_version < 2,
14124 the filter will return an error.
14127 This filter also supports the @ref{framesync} options.
14129 The file printed if @var{stats_file} is selected, contains a sequence of
14130 key/value pairs of the form @var{key}:@var{value} for each compared
14133 If a @var{stats_version} greater than 1 is specified, a header line precedes
14134 the list of per-frame-pair stats, with key value pairs following the frame
14135 format with the following parameters:
14138 @item psnr_log_version
14139 The version of the log file format. Will match @var{stats_version}.
14142 A comma separated list of the per-frame-pair parameters included in
14146 A description of each shown per-frame-pair parameter follows:
14150 sequential number of the input frame, starting from 1
14153 Mean Square Error pixel-by-pixel average difference of the compared
14154 frames, averaged over all the image components.
14156 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14157 Mean Square Error pixel-by-pixel average difference of the compared
14158 frames for the component specified by the suffix.
14160 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14161 Peak Signal to Noise ratio of the compared frames for the component
14162 specified by the suffix.
14164 @item max_avg, max_y, max_u, max_v
14165 Maximum allowed value for each channel, and average over all
14171 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14172 [main][ref] psnr="stats_file=stats.log" [out]
14175 On this example the input file being processed is compared with the
14176 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14177 is stored in @file{stats.log}.
14182 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14183 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14186 The pullup filter is designed to take advantage of future context in making
14187 its decisions. This filter is stateless in the sense that it does not lock
14188 onto a pattern to follow, but it instead looks forward to the following
14189 fields in order to identify matches and rebuild progressive frames.
14191 To produce content with an even framerate, insert the fps filter after
14192 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14193 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14195 The filter accepts the following options:
14202 These options set the amount of "junk" to ignore at the left, right, top, and
14203 bottom of the image, respectively. Left and right are in units of 8 pixels,
14204 while top and bottom are in units of 2 lines.
14205 The default is 8 pixels on each side.
14208 Set the strict breaks. Setting this option to 1 will reduce the chances of
14209 filter generating an occasional mismatched frame, but it may also cause an
14210 excessive number of frames to be dropped during high motion sequences.
14211 Conversely, setting it to -1 will make filter match fields more easily.
14212 This may help processing of video where there is slight blurring between
14213 the fields, but may also cause there to be interlaced frames in the output.
14214 Default value is @code{0}.
14217 Set the metric plane to use. It accepts the following values:
14223 Use chroma blue plane.
14226 Use chroma red plane.
14229 This option may be set to use chroma plane instead of the default luma plane
14230 for doing filter's computations. This may improve accuracy on very clean
14231 source material, but more likely will decrease accuracy, especially if there
14232 is chroma noise (rainbow effect) or any grayscale video.
14233 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14234 load and make pullup usable in realtime on slow machines.
14237 For best results (without duplicated frames in the output file) it is
14238 necessary to change the output frame rate. For example, to inverse
14239 telecine NTSC input:
14241 ffmpeg -i input -vf pullup -r 24000/1001 ...
14246 Change video quantization parameters (QP).
14248 The filter accepts the following option:
14252 Set expression for quantization parameter.
14255 The expression is evaluated through the eval API and can contain, among others,
14256 the following constants:
14260 1 if index is not 129, 0 otherwise.
14263 Sequential index starting from -129 to 128.
14266 @subsection Examples
14270 Some equation like:
14278 Flush video frames from internal cache of frames into a random order.
14279 No frame is discarded.
14280 Inspired by @ref{frei0r} nervous filter.
14284 Set size in number of frames of internal cache, in range from @code{2} to
14285 @code{512}. Default is @code{30}.
14288 Set seed for random number generator, must be an integer included between
14289 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14290 less than @code{0}, the filter will try to use a good random seed on a
14294 @section readeia608
14296 Read closed captioning (EIA-608) information from the top lines of a video frame.
14298 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14299 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14300 with EIA-608 data (starting from 0). A description of each metadata value follows:
14303 @item lavfi.readeia608.X.cc
14304 The two bytes stored as EIA-608 data (printed in hexadecimal).
14306 @item lavfi.readeia608.X.line
14307 The number of the line on which the EIA-608 data was identified and read.
14310 This filter accepts the following options:
14314 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14317 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14320 Set minimal acceptable amplitude change for sync codes detection.
14321 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14324 Set the ratio of width reserved for sync code detection.
14325 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14328 Set the max peaks height difference for sync code detection.
14329 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14332 Set max peaks period difference for sync code detection.
14333 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14336 Set the first two max start code bits differences.
14337 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14340 Set the minimum ratio of bits height compared to 3rd start code bit.
14341 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14344 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14347 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14350 Enable checking the parity bit. In the event of a parity error, the filter will output
14351 @code{0x00} for that character. Default is false.
14354 @subsection Examples
14358 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14360 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
14366 Read vertical interval timecode (VITC) information from the top lines of a
14369 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14370 timecode value, if a valid timecode has been detected. Further metadata key
14371 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14372 timecode data has been found or not.
14374 This filter accepts the following options:
14378 Set the maximum number of lines to scan for VITC data. If the value is set to
14379 @code{-1} the full video frame is scanned. Default is @code{45}.
14382 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14383 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14386 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14387 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14390 @subsection Examples
14394 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14395 draw @code{--:--:--:--} as a placeholder:
14397 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14403 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14405 Destination pixel at position (X, Y) will be picked from source (x, y) position
14406 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14407 value for pixel will be used for destination pixel.
14409 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14410 will have Xmap/Ymap video stream dimensions.
14411 Xmap and Ymap input video streams are 16bit depth, single channel.
14413 @section removegrain
14415 The removegrain filter is a spatial denoiser for progressive video.
14419 Set mode for the first plane.
14422 Set mode for the second plane.
14425 Set mode for the third plane.
14428 Set mode for the fourth plane.
14431 Range of mode is from 0 to 24. Description of each mode follows:
14435 Leave input plane unchanged. Default.
14438 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14441 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14444 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14447 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14448 This is equivalent to a median filter.
14451 Line-sensitive clipping giving the minimal change.
14454 Line-sensitive clipping, intermediate.
14457 Line-sensitive clipping, intermediate.
14460 Line-sensitive clipping, intermediate.
14463 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14466 Replaces the target pixel with the closest neighbour.
14469 [1 2 1] horizontal and vertical kernel blur.
14475 Bob mode, interpolates top field from the line where the neighbours
14476 pixels are the closest.
14479 Bob mode, interpolates bottom field from the line where the neighbours
14480 pixels are the closest.
14483 Bob mode, interpolates top field. Same as 13 but with a more complicated
14484 interpolation formula.
14487 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14488 interpolation formula.
14491 Clips the pixel with the minimum and maximum of respectively the maximum and
14492 minimum of each pair of opposite neighbour pixels.
14495 Line-sensitive clipping using opposite neighbours whose greatest distance from
14496 the current pixel is minimal.
14499 Replaces the pixel with the average of its 8 neighbours.
14502 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14505 Clips pixels using the averages of opposite neighbour.
14508 Same as mode 21 but simpler and faster.
14511 Small edge and halo removal, but reputed useless.
14517 @section removelogo
14519 Suppress a TV station logo, using an image file to determine which
14520 pixels comprise the logo. It works by filling in the pixels that
14521 comprise the logo with neighboring pixels.
14523 The filter accepts the following options:
14527 Set the filter bitmap file, which can be any image format supported by
14528 libavformat. The width and height of the image file must match those of the
14529 video stream being processed.
14532 Pixels in the provided bitmap image with a value of zero are not
14533 considered part of the logo, non-zero pixels are considered part of
14534 the logo. If you use white (255) for the logo and black (0) for the
14535 rest, you will be safe. For making the filter bitmap, it is
14536 recommended to take a screen capture of a black frame with the logo
14537 visible, and then using a threshold filter followed by the erode
14538 filter once or twice.
14540 If needed, little splotches can be fixed manually. Remember that if
14541 logo pixels are not covered, the filter quality will be much
14542 reduced. Marking too many pixels as part of the logo does not hurt as
14543 much, but it will increase the amount of blurring needed to cover over
14544 the image and will destroy more information than necessary, and extra
14545 pixels will slow things down on a large logo.
14547 @section repeatfields
14549 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14550 fields based on its value.
14554 Reverse a video clip.
14556 Warning: This filter requires memory to buffer the entire clip, so trimming
14559 @subsection Examples
14563 Take the first 5 seconds of a clip, and reverse it.
14570 Shift R/G/B/A pixels horizontally and/or vertically.
14572 The filter accepts the following options:
14575 Set amount to shift red horizontally.
14577 Set amount to shift red vertically.
14579 Set amount to shift green horizontally.
14581 Set amount to shift green vertically.
14583 Set amount to shift blue horizontally.
14585 Set amount to shift blue vertically.
14587 Set amount to shift alpha horizontally.
14589 Set amount to shift alpha vertically.
14591 Set edge mode, can be @var{smear}, default, or @var{warp}.
14595 Apply roberts cross operator to input video stream.
14597 The filter accepts the following option:
14601 Set which planes will be processed, unprocessed planes will be copied.
14602 By default value 0xf, all planes will be processed.
14605 Set value which will be multiplied with filtered result.
14608 Set value which will be added to filtered result.
14613 Rotate video by an arbitrary angle expressed in radians.
14615 The filter accepts the following options:
14617 A description of the optional parameters follows.
14620 Set an expression for the angle by which to rotate the input video
14621 clockwise, expressed as a number of radians. A negative value will
14622 result in a counter-clockwise rotation. By default it is set to "0".
14624 This expression is evaluated for each frame.
14627 Set the output width expression, default value is "iw".
14628 This expression is evaluated just once during configuration.
14631 Set the output height expression, default value is "ih".
14632 This expression is evaluated just once during configuration.
14635 Enable bilinear interpolation if set to 1, a value of 0 disables
14636 it. Default value is 1.
14639 Set the color used to fill the output area not covered by the rotated
14640 image. For the general syntax of this option, check the
14641 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14642 If the special value "none" is selected then no
14643 background is printed (useful for example if the background is never shown).
14645 Default value is "black".
14648 The expressions for the angle and the output size can contain the
14649 following constants and functions:
14653 sequential number of the input frame, starting from 0. It is always NAN
14654 before the first frame is filtered.
14657 time in seconds of the input frame, it is set to 0 when the filter is
14658 configured. It is always NAN before the first frame is filtered.
14662 horizontal and vertical chroma subsample values. For example for the
14663 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14667 the input video width and height
14671 the output width and height, that is the size of the padded area as
14672 specified by the @var{width} and @var{height} expressions
14676 the minimal width/height required for completely containing the input
14677 video rotated by @var{a} radians.
14679 These are only available when computing the @option{out_w} and
14680 @option{out_h} expressions.
14683 @subsection Examples
14687 Rotate the input by PI/6 radians clockwise:
14693 Rotate the input by PI/6 radians counter-clockwise:
14699 Rotate the input by 45 degrees clockwise:
14705 Apply a constant rotation with period T, starting from an angle of PI/3:
14707 rotate=PI/3+2*PI*t/T
14711 Make the input video rotation oscillating with a period of T
14712 seconds and an amplitude of A radians:
14714 rotate=A*sin(2*PI/T*t)
14718 Rotate the video, output size is chosen so that the whole rotating
14719 input video is always completely contained in the output:
14721 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14725 Rotate the video, reduce the output size so that no background is ever
14728 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14732 @subsection Commands
14734 The filter supports the following commands:
14738 Set the angle expression.
14739 The command accepts the same syntax of the corresponding option.
14741 If the specified expression is not valid, it is kept at its current
14747 Apply Shape Adaptive Blur.
14749 The filter accepts the following options:
14752 @item luma_radius, lr
14753 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14754 value is 1.0. A greater value will result in a more blurred image, and
14755 in slower processing.
14757 @item luma_pre_filter_radius, lpfr
14758 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14761 @item luma_strength, ls
14762 Set luma maximum difference between pixels to still be considered, must
14763 be a value in the 0.1-100.0 range, default value is 1.0.
14765 @item chroma_radius, cr
14766 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14767 greater value will result in a more blurred image, and in slower
14770 @item chroma_pre_filter_radius, cpfr
14771 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14773 @item chroma_strength, cs
14774 Set chroma maximum difference between pixels to still be considered,
14775 must be a value in the -0.9-100.0 range.
14778 Each chroma option value, if not explicitly specified, is set to the
14779 corresponding luma option value.
14784 Scale (resize) the input video, using the libswscale library.
14786 The scale filter forces the output display aspect ratio to be the same
14787 of the input, by changing the output sample aspect ratio.
14789 If the input image format is different from the format requested by
14790 the next filter, the scale filter will convert the input to the
14793 @subsection Options
14794 The filter accepts the following options, or any of the options
14795 supported by the libswscale scaler.
14797 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14798 the complete list of scaler options.
14803 Set the output video dimension expression. Default value is the input
14806 If the @var{width} or @var{w} value is 0, the input width is used for
14807 the output. If the @var{height} or @var{h} value is 0, the input height
14808 is used for the output.
14810 If one and only one of the values is -n with n >= 1, the scale filter
14811 will use a value that maintains the aspect ratio of the input image,
14812 calculated from the other specified dimension. After that it will,
14813 however, make sure that the calculated dimension is divisible by n and
14814 adjust the value if necessary.
14816 If both values are -n with n >= 1, the behavior will be identical to
14817 both values being set to 0 as previously detailed.
14819 See below for the list of accepted constants for use in the dimension
14823 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14827 Only evaluate expressions once during the filter initialization or when a command is processed.
14830 Evaluate expressions for each incoming frame.
14834 Default value is @samp{init}.
14838 Set the interlacing mode. It accepts the following values:
14842 Force interlaced aware scaling.
14845 Do not apply interlaced scaling.
14848 Select interlaced aware scaling depending on whether the source frames
14849 are flagged as interlaced or not.
14852 Default value is @samp{0}.
14855 Set libswscale scaling flags. See
14856 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14857 complete list of values. If not explicitly specified the filter applies
14861 @item param0, param1
14862 Set libswscale input parameters for scaling algorithms that need them. See
14863 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14864 complete documentation. If not explicitly specified the filter applies
14870 Set the video size. For the syntax of this option, check the
14871 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14873 @item in_color_matrix
14874 @item out_color_matrix
14875 Set in/output YCbCr color space type.
14877 This allows the autodetected value to be overridden as well as allows forcing
14878 a specific value used for the output and encoder.
14880 If not specified, the color space type depends on the pixel format.
14886 Choose automatically.
14889 Format conforming to International Telecommunication Union (ITU)
14890 Recommendation BT.709.
14893 Set color space conforming to the United States Federal Communications
14894 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14897 Set color space conforming to:
14901 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14904 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14907 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14912 Set color space conforming to SMPTE ST 240:1999.
14917 Set in/output YCbCr sample range.
14919 This allows the autodetected value to be overridden as well as allows forcing
14920 a specific value used for the output and encoder. If not specified, the
14921 range depends on the pixel format. Possible values:
14925 Choose automatically.
14928 Set full range (0-255 in case of 8-bit luma).
14930 @item mpeg/limited/tv
14931 Set "MPEG" range (16-235 in case of 8-bit luma).
14934 @item force_original_aspect_ratio
14935 Enable decreasing or increasing output video width or height if necessary to
14936 keep the original aspect ratio. Possible values:
14940 Scale the video as specified and disable this feature.
14943 The output video dimensions will automatically be decreased if needed.
14946 The output video dimensions will automatically be increased if needed.
14950 One useful instance of this option is that when you know a specific device's
14951 maximum allowed resolution, you can use this to limit the output video to
14952 that, while retaining the aspect ratio. For example, device A allows
14953 1280x720 playback, and your video is 1920x800. Using this option (set it to
14954 decrease) and specifying 1280x720 to the command line makes the output
14957 Please note that this is a different thing than specifying -1 for @option{w}
14958 or @option{h}, you still need to specify the output resolution for this option
14963 The values of the @option{w} and @option{h} options are expressions
14964 containing the following constants:
14969 The input width and height
14973 These are the same as @var{in_w} and @var{in_h}.
14977 The output (scaled) width and height
14981 These are the same as @var{out_w} and @var{out_h}
14984 The same as @var{iw} / @var{ih}
14987 input sample aspect ratio
14990 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14994 horizontal and vertical input chroma subsample values. For example for the
14995 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14999 horizontal and vertical output chroma subsample values. For example for the
15000 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15003 @subsection Examples
15007 Scale the input video to a size of 200x100
15012 This is equivalent to:
15023 Specify a size abbreviation for the output size:
15028 which can also be written as:
15034 Scale the input to 2x:
15036 scale=w=2*iw:h=2*ih
15040 The above is the same as:
15042 scale=2*in_w:2*in_h
15046 Scale the input to 2x with forced interlaced scaling:
15048 scale=2*iw:2*ih:interl=1
15052 Scale the input to half size:
15054 scale=w=iw/2:h=ih/2
15058 Increase the width, and set the height to the same size:
15064 Seek Greek harmony:
15071 Increase the height, and set the width to 3/2 of the height:
15073 scale=w=3/2*oh:h=3/5*ih
15077 Increase the size, making the size a multiple of the chroma
15080 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
15084 Increase the width to a maximum of 500 pixels,
15085 keeping the same aspect ratio as the input:
15087 scale=w='min(500\, iw*3/2):h=-1'
15091 Make pixels square by combining scale and setsar:
15093 scale='trunc(ih*dar):ih',setsar=1/1
15097 Make pixels square by combining scale and setsar,
15098 making sure the resulting resolution is even (required by some codecs):
15100 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
15104 @subsection Commands
15106 This filter supports the following commands:
15110 Set the output video dimension expression.
15111 The command accepts the same syntax of the corresponding option.
15113 If the specified expression is not valid, it is kept at its current
15119 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15120 format conversion on CUDA video frames. Setting the output width and height
15121 works in the same way as for the @var{scale} filter.
15123 The following additional options are accepted:
15126 The pixel format of the output CUDA frames. If set to the string "same" (the
15127 default), the input format will be kept. Note that automatic format negotiation
15128 and conversion is not yet supported for hardware frames
15131 The interpolation algorithm used for resizing. One of the following:
15138 @item cubic2p_bspline
15139 2-parameter cubic (B=1, C=0)
15141 @item cubic2p_catmullrom
15142 2-parameter cubic (B=0, C=1/2)
15144 @item cubic2p_b05c03
15145 2-parameter cubic (B=1/2, C=3/10)
15157 Scale (resize) the input video, based on a reference video.
15159 See the scale filter for available options, scale2ref supports the same but
15160 uses the reference video instead of the main input as basis. scale2ref also
15161 supports the following additional constants for the @option{w} and
15162 @option{h} options:
15167 The main input video's width and height
15170 The same as @var{main_w} / @var{main_h}
15173 The main input video's sample aspect ratio
15175 @item main_dar, mdar
15176 The main input video's display aspect ratio. Calculated from
15177 @code{(main_w / main_h) * main_sar}.
15181 The main input video's horizontal and vertical chroma subsample values.
15182 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15186 @subsection Examples
15190 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15192 'scale2ref[b][a];[a][b]overlay'
15196 @anchor{selectivecolor}
15197 @section selectivecolor
15199 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15200 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15201 by the "purity" of the color (that is, how saturated it already is).
15203 This filter is similar to the Adobe Photoshop Selective Color tool.
15205 The filter accepts the following options:
15208 @item correction_method
15209 Select color correction method.
15211 Available values are:
15214 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15217 Specified adjustments are relative to the original component value.
15219 Default is @code{absolute}.
15221 Adjustments for red pixels (pixels where the red component is the maximum)
15223 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15225 Adjustments for green pixels (pixels where the green component is the maximum)
15227 Adjustments for cyan pixels (pixels where the red component is the minimum)
15229 Adjustments for blue pixels (pixels where the blue component is the maximum)
15231 Adjustments for magenta pixels (pixels where the green component is the minimum)
15233 Adjustments for white pixels (pixels where all components are greater than 128)
15235 Adjustments for all pixels except pure black and pure white
15237 Adjustments for black pixels (pixels where all components are lesser than 128)
15239 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15242 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15243 4 space separated floating point adjustment values in the [-1,1] range,
15244 respectively to adjust the amount of cyan, magenta, yellow and black for the
15245 pixels of its range.
15247 @subsection Examples
15251 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15252 increase magenta by 27% in blue areas:
15254 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15258 Use a Photoshop selective color preset:
15260 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15264 @anchor{separatefields}
15265 @section separatefields
15267 The @code{separatefields} takes a frame-based video input and splits
15268 each frame into its components fields, producing a new half height clip
15269 with twice the frame rate and twice the frame count.
15271 This filter use field-dominance information in frame to decide which
15272 of each pair of fields to place first in the output.
15273 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15275 @section setdar, setsar
15277 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15280 This is done by changing the specified Sample (aka Pixel) Aspect
15281 Ratio, according to the following equation:
15283 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15286 Keep in mind that the @code{setdar} filter does not modify the pixel
15287 dimensions of the video frame. Also, the display aspect ratio set by
15288 this filter may be changed by later filters in the filterchain,
15289 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15292 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15293 the filter output video.
15295 Note that as a consequence of the application of this filter, the
15296 output display aspect ratio will change according to the equation
15299 Keep in mind that the sample aspect ratio set by the @code{setsar}
15300 filter may be changed by later filters in the filterchain, e.g. if
15301 another "setsar" or a "setdar" filter is applied.
15303 It accepts the following parameters:
15306 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15307 Set the aspect ratio used by the filter.
15309 The parameter can be a floating point number string, an expression, or
15310 a string of the form @var{num}:@var{den}, where @var{num} and
15311 @var{den} are the numerator and denominator of the aspect ratio. If
15312 the parameter is not specified, it is assumed the value "0".
15313 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15317 Set the maximum integer value to use for expressing numerator and
15318 denominator when reducing the expressed aspect ratio to a rational.
15319 Default value is @code{100}.
15323 The parameter @var{sar} is an expression containing
15324 the following constants:
15328 These are approximated values for the mathematical constants e
15329 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15332 The input width and height.
15335 These are the same as @var{w} / @var{h}.
15338 The input sample aspect ratio.
15341 The input display aspect ratio. It is the same as
15342 (@var{w} / @var{h}) * @var{sar}.
15345 Horizontal and vertical chroma subsample values. For example, for the
15346 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15349 @subsection Examples
15354 To change the display aspect ratio to 16:9, specify one of the following:
15361 To change the sample aspect ratio to 10:11, specify:
15367 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15368 1000 in the aspect ratio reduction, use the command:
15370 setdar=ratio=16/9:max=1000
15378 Force field for the output video frame.
15380 The @code{setfield} filter marks the interlace type field for the
15381 output frames. It does not change the input frame, but only sets the
15382 corresponding property, which affects how the frame is treated by
15383 following filters (e.g. @code{fieldorder} or @code{yadif}).
15385 The filter accepts the following options:
15390 Available values are:
15394 Keep the same field property.
15397 Mark the frame as bottom-field-first.
15400 Mark the frame as top-field-first.
15403 Mark the frame as progressive.
15410 Force frame parameter for the output video frame.
15412 The @code{setparams} filter marks interlace and color range for the
15413 output frames. It does not change the input frame, but only sets the
15414 corresponding property, which affects how the frame is treated by
15419 Available values are:
15423 Keep the same field property (default).
15426 Mark the frame as bottom-field-first.
15429 Mark the frame as top-field-first.
15432 Mark the frame as progressive.
15436 Available values are:
15440 Keep the same color range property (default).
15442 @item unspecified, unknown
15443 Mark the frame as unspecified color range.
15445 @item limited, tv, mpeg
15446 Mark the frame as limited range.
15448 @item full, pc, jpeg
15449 Mark the frame as full range.
15452 @item color_primaries
15453 Set the color primaries.
15454 Available values are:
15458 Keep the same color primaries property (default).
15475 Set the color transfer.
15476 Available values are:
15480 Keep the same color trc property (default).
15502 Set the colorspace.
15503 Available values are:
15507 Keep the same colorspace property (default).
15520 @item chroma-derived-nc
15521 @item chroma-derived-c
15528 Show a line containing various information for each input video frame.
15529 The input video is not modified.
15531 This filter supports the following options:
15535 Calculate checksums of each plane. By default enabled.
15538 The shown line contains a sequence of key/value pairs of the form
15539 @var{key}:@var{value}.
15541 The following values are shown in the output:
15545 The (sequential) number of the input frame, starting from 0.
15548 The Presentation TimeStamp of the input frame, expressed as a number of
15549 time base units. The time base unit depends on the filter input pad.
15552 The Presentation TimeStamp of the input frame, expressed as a number of
15556 The position of the frame in the input stream, or -1 if this information is
15557 unavailable and/or meaningless (for example in case of synthetic video).
15560 The pixel format name.
15563 The sample aspect ratio of the input frame, expressed in the form
15564 @var{num}/@var{den}.
15567 The size of the input frame. For the syntax of this option, check the
15568 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15571 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15572 for bottom field first).
15575 This is 1 if the frame is a key frame, 0 otherwise.
15578 The picture type of the input frame ("I" for an I-frame, "P" for a
15579 P-frame, "B" for a B-frame, or "?" for an unknown type).
15580 Also refer to the documentation of the @code{AVPictureType} enum and of
15581 the @code{av_get_picture_type_char} function defined in
15582 @file{libavutil/avutil.h}.
15585 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15587 @item plane_checksum
15588 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15589 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15592 @section showpalette
15594 Displays the 256 colors palette of each frame. This filter is only relevant for
15595 @var{pal8} pixel format frames.
15597 It accepts the following option:
15601 Set the size of the box used to represent one palette color entry. Default is
15602 @code{30} (for a @code{30x30} pixel box).
15605 @section shuffleframes
15607 Reorder and/or duplicate and/or drop video frames.
15609 It accepts the following parameters:
15613 Set the destination indexes of input frames.
15614 This is space or '|' separated list of indexes that maps input frames to output
15615 frames. Number of indexes also sets maximal value that each index may have.
15616 '-1' index have special meaning and that is to drop frame.
15619 The first frame has the index 0. The default is to keep the input unchanged.
15621 @subsection Examples
15625 Swap second and third frame of every three frames of the input:
15627 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15631 Swap 10th and 1st frame of every ten frames of the input:
15633 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15637 @section shuffleplanes
15639 Reorder and/or duplicate video planes.
15641 It accepts the following parameters:
15646 The index of the input plane to be used as the first output plane.
15649 The index of the input plane to be used as the second output plane.
15652 The index of the input plane to be used as the third output plane.
15655 The index of the input plane to be used as the fourth output plane.
15659 The first plane has the index 0. The default is to keep the input unchanged.
15661 @subsection Examples
15665 Swap the second and third planes of the input:
15667 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15671 @anchor{signalstats}
15672 @section signalstats
15673 Evaluate various visual metrics that assist in determining issues associated
15674 with the digitization of analog video media.
15676 By default the filter will log these metadata values:
15680 Display the minimal Y value contained within the input frame. Expressed in
15684 Display the Y value at the 10% percentile within the input frame. Expressed in
15688 Display the average Y value within the input frame. Expressed in range of
15692 Display the Y value at the 90% percentile within the input frame. Expressed in
15696 Display the maximum Y value contained within the input frame. Expressed in
15700 Display the minimal U value contained within the input frame. Expressed in
15704 Display the U value at the 10% percentile within the input frame. Expressed in
15708 Display the average U value within the input frame. Expressed in range of
15712 Display the U value at the 90% percentile within the input frame. Expressed in
15716 Display the maximum U value contained within the input frame. Expressed in
15720 Display the minimal V value contained within the input frame. Expressed in
15724 Display the V value at the 10% percentile within the input frame. Expressed in
15728 Display the average V value within the input frame. Expressed in range of
15732 Display the V value at the 90% percentile within the input frame. Expressed in
15736 Display the maximum V value contained within the input frame. Expressed in
15740 Display the minimal saturation value contained within the input frame.
15741 Expressed in range of [0-~181.02].
15744 Display the saturation value at the 10% percentile within the input frame.
15745 Expressed in range of [0-~181.02].
15748 Display the average saturation value within the input frame. Expressed in range
15752 Display the saturation value at the 90% percentile within the input frame.
15753 Expressed in range of [0-~181.02].
15756 Display the maximum saturation value contained within the input frame.
15757 Expressed in range of [0-~181.02].
15760 Display the median value for hue within the input frame. Expressed in range of
15764 Display the average value for hue within the input frame. Expressed in range of
15768 Display the average of sample value difference between all values of the Y
15769 plane in the current frame and corresponding values of the previous input frame.
15770 Expressed in range of [0-255].
15773 Display the average of sample value difference between all values of the U
15774 plane in the current frame and corresponding values of the previous input frame.
15775 Expressed in range of [0-255].
15778 Display the average of sample value difference between all values of the V
15779 plane in the current frame and corresponding values of the previous input frame.
15780 Expressed in range of [0-255].
15783 Display bit depth of Y plane in current frame.
15784 Expressed in range of [0-16].
15787 Display bit depth of U plane in current frame.
15788 Expressed in range of [0-16].
15791 Display bit depth of V plane in current frame.
15792 Expressed in range of [0-16].
15795 The filter accepts the following options:
15801 @option{stat} specify an additional form of image analysis.
15802 @option{out} output video with the specified type of pixel highlighted.
15804 Both options accept the following values:
15808 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15809 unlike the neighboring pixels of the same field. Examples of temporal outliers
15810 include the results of video dropouts, head clogs, or tape tracking issues.
15813 Identify @var{vertical line repetition}. Vertical line repetition includes
15814 similar rows of pixels within a frame. In born-digital video vertical line
15815 repetition is common, but this pattern is uncommon in video digitized from an
15816 analog source. When it occurs in video that results from the digitization of an
15817 analog source it can indicate concealment from a dropout compensator.
15820 Identify pixels that fall outside of legal broadcast range.
15824 Set the highlight color for the @option{out} option. The default color is
15828 @subsection Examples
15832 Output data of various video metrics:
15834 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15838 Output specific data about the minimum and maximum values of the Y plane per frame:
15840 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15844 Playback video while highlighting pixels that are outside of broadcast range in red.
15846 ffplay example.mov -vf signalstats="out=brng:color=red"
15850 Playback video with signalstats metadata drawn over the frame.
15852 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15855 The contents of signalstat_drawtext.txt used in the command are:
15858 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15859 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15860 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15861 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15869 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15870 input. In this case the matching between the inputs can be calculated additionally.
15871 The filter always passes through the first input. The signature of each stream can
15872 be written into a file.
15874 It accepts the following options:
15878 Enable or disable the matching process.
15880 Available values are:
15884 Disable the calculation of a matching (default).
15886 Calculate the matching for the whole video and output whether the whole video
15887 matches or only parts.
15889 Calculate only until a matching is found or the video ends. Should be faster in
15894 Set the number of inputs. The option value must be a non negative integer.
15895 Default value is 1.
15898 Set the path to which the output is written. If there is more than one input,
15899 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15900 integer), that will be replaced with the input number. If no filename is
15901 specified, no output will be written. This is the default.
15904 Choose the output format.
15906 Available values are:
15910 Use the specified binary representation (default).
15912 Use the specified xml representation.
15916 Set threshold to detect one word as similar. The option value must be an integer
15917 greater than zero. The default value is 9000.
15920 Set threshold to detect all words as similar. The option value must be an integer
15921 greater than zero. The default value is 60000.
15924 Set threshold to detect frames as similar. The option value must be an integer
15925 greater than zero. The default value is 116.
15928 Set the minimum length of a sequence in frames to recognize it as matching
15929 sequence. The option value must be a non negative integer value.
15930 The default value is 0.
15933 Set the minimum relation, that matching frames to all frames must have.
15934 The option value must be a double value between 0 and 1. The default value is 0.5.
15937 @subsection Examples
15941 To calculate the signature of an input video and store it in signature.bin:
15943 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15947 To detect whether two videos match and store the signatures in XML format in
15948 signature0.xml and signature1.xml:
15950 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 -
15958 Blur the input video without impacting the outlines.
15960 It accepts the following options:
15963 @item luma_radius, lr
15964 Set the luma radius. The option value must be a float number in
15965 the range [0.1,5.0] that specifies the variance of the gaussian filter
15966 used to blur the image (slower if larger). Default value is 1.0.
15968 @item luma_strength, ls
15969 Set the luma strength. The option value must be a float number
15970 in the range [-1.0,1.0] that configures the blurring. A value included
15971 in [0.0,1.0] will blur the image whereas a value included in
15972 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15974 @item luma_threshold, lt
15975 Set the luma threshold used as a coefficient to determine
15976 whether a pixel should be blurred or not. The option value must be an
15977 integer in the range [-30,30]. A value of 0 will filter all the image,
15978 a value included in [0,30] will filter flat areas and a value included
15979 in [-30,0] will filter edges. Default value is 0.
15981 @item chroma_radius, cr
15982 Set the chroma radius. The option value must be a float number in
15983 the range [0.1,5.0] that specifies the variance of the gaussian filter
15984 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15986 @item chroma_strength, cs
15987 Set the chroma strength. The option value must be a float number
15988 in the range [-1.0,1.0] that configures the blurring. A value included
15989 in [0.0,1.0] will blur the image whereas a value included in
15990 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15992 @item chroma_threshold, ct
15993 Set the chroma threshold used as a coefficient to determine
15994 whether a pixel should be blurred or not. The option value must be an
15995 integer in the range [-30,30]. A value of 0 will filter all the image,
15996 a value included in [0,30] will filter flat areas and a value included
15997 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
16000 If a chroma option is not explicitly set, the corresponding luma value
16005 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
16007 This filter takes in input two input videos, the first input is
16008 considered the "main" source and is passed unchanged to the
16009 output. The second input is used as a "reference" video for computing
16012 Both video inputs must have the same resolution and pixel format for
16013 this filter to work correctly. Also it assumes that both inputs
16014 have the same number of frames, which are compared one by one.
16016 The filter stores the calculated SSIM of each frame.
16018 The description of the accepted parameters follows.
16021 @item stats_file, f
16022 If specified the filter will use the named file to save the SSIM of
16023 each individual frame. When filename equals "-" the data is sent to
16027 The file printed if @var{stats_file} is selected, contains a sequence of
16028 key/value pairs of the form @var{key}:@var{value} for each compared
16031 A description of each shown parameter follows:
16035 sequential number of the input frame, starting from 1
16037 @item Y, U, V, R, G, B
16038 SSIM of the compared frames for the component specified by the suffix.
16041 SSIM of the compared frames for the whole frame.
16044 Same as above but in dB representation.
16047 This filter also supports the @ref{framesync} options.
16051 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
16052 [main][ref] ssim="stats_file=stats.log" [out]
16055 On this example the input file being processed is compared with the
16056 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
16057 is stored in @file{stats.log}.
16059 Another example with both psnr and ssim at same time:
16061 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
16066 Convert between different stereoscopic image formats.
16068 The filters accept the following options:
16072 Set stereoscopic image format of input.
16074 Available values for input image formats are:
16077 side by side parallel (left eye left, right eye right)
16080 side by side crosseye (right eye left, left eye right)
16083 side by side parallel with half width resolution
16084 (left eye left, right eye right)
16087 side by side crosseye with half width resolution
16088 (right eye left, left eye right)
16091 above-below (left eye above, right eye below)
16094 above-below (right eye above, left eye below)
16097 above-below with half height resolution
16098 (left eye above, right eye below)
16101 above-below with half height resolution
16102 (right eye above, left eye below)
16105 alternating frames (left eye first, right eye second)
16108 alternating frames (right eye first, left eye second)
16111 interleaved rows (left eye has top row, right eye starts on next row)
16114 interleaved rows (right eye has top row, left eye starts on next row)
16117 interleaved columns, left eye first
16120 interleaved columns, right eye first
16122 Default value is @samp{sbsl}.
16126 Set stereoscopic image format of output.
16130 side by side parallel (left eye left, right eye right)
16133 side by side crosseye (right eye left, left eye right)
16136 side by side parallel with half width resolution
16137 (left eye left, right eye right)
16140 side by side crosseye with half width resolution
16141 (right eye left, left eye right)
16144 above-below (left eye above, right eye below)
16147 above-below (right eye above, left eye below)
16150 above-below with half height resolution
16151 (left eye above, right eye below)
16154 above-below with half height resolution
16155 (right eye above, left eye below)
16158 alternating frames (left eye first, right eye second)
16161 alternating frames (right eye first, left eye second)
16164 interleaved rows (left eye has top row, right eye starts on next row)
16167 interleaved rows (right eye has top row, left eye starts on next row)
16170 anaglyph red/blue gray
16171 (red filter on left eye, blue filter on right eye)
16174 anaglyph red/green gray
16175 (red filter on left eye, green filter on right eye)
16178 anaglyph red/cyan gray
16179 (red filter on left eye, cyan filter on right eye)
16182 anaglyph red/cyan half colored
16183 (red filter on left eye, cyan filter on right eye)
16186 anaglyph red/cyan color
16187 (red filter on left eye, cyan filter on right eye)
16190 anaglyph red/cyan color optimized with the least squares projection of dubois
16191 (red filter on left eye, cyan filter on right eye)
16194 anaglyph green/magenta gray
16195 (green filter on left eye, magenta filter on right eye)
16198 anaglyph green/magenta half colored
16199 (green filter on left eye, magenta filter on right eye)
16202 anaglyph green/magenta colored
16203 (green filter on left eye, magenta filter on right eye)
16206 anaglyph green/magenta color optimized with the least squares projection of dubois
16207 (green filter on left eye, magenta filter on right eye)
16210 anaglyph yellow/blue gray
16211 (yellow filter on left eye, blue filter on right eye)
16214 anaglyph yellow/blue half colored
16215 (yellow filter on left eye, blue filter on right eye)
16218 anaglyph yellow/blue colored
16219 (yellow filter on left eye, blue filter on right eye)
16222 anaglyph yellow/blue color optimized with the least squares projection of dubois
16223 (yellow filter on left eye, blue filter on right eye)
16226 mono output (left eye only)
16229 mono output (right eye only)
16232 checkerboard, left eye first
16235 checkerboard, right eye first
16238 interleaved columns, left eye first
16241 interleaved columns, right eye first
16247 Default value is @samp{arcd}.
16250 @subsection Examples
16254 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16260 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16266 @section streamselect, astreamselect
16267 Select video or audio streams.
16269 The filter accepts the following options:
16273 Set number of inputs. Default is 2.
16276 Set input indexes to remap to outputs.
16279 @subsection Commands
16281 The @code{streamselect} and @code{astreamselect} filter supports the following
16286 Set input indexes to remap to outputs.
16289 @subsection Examples
16293 Select first 5 seconds 1st stream and rest of time 2nd stream:
16295 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16299 Same as above, but for audio:
16301 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16306 Apply sobel operator to input video stream.
16308 The filter accepts the following option:
16312 Set which planes will be processed, unprocessed planes will be copied.
16313 By default value 0xf, all planes will be processed.
16316 Set value which will be multiplied with filtered result.
16319 Set value which will be added to filtered result.
16325 Apply a simple postprocessing filter that compresses and decompresses the image
16326 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16327 and average the results.
16329 The filter accepts the following options:
16333 Set quality. This option defines the number of levels for averaging. It accepts
16334 an integer in the range 0-6. If set to @code{0}, the filter will have no
16335 effect. A value of @code{6} means the higher quality. For each increment of
16336 that value the speed drops by a factor of approximately 2. Default value is
16340 Force a constant quantization parameter. If not set, the filter will use the QP
16341 from the video stream (if available).
16344 Set thresholding mode. Available modes are:
16348 Set hard thresholding (default).
16350 Set soft thresholding (better de-ringing effect, but likely blurrier).
16353 @item use_bframe_qp
16354 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16355 option may cause flicker since the B-Frames have often larger QP. Default is
16356 @code{0} (not enabled).
16361 Scale the input by applying one of the super-resolution methods based on
16362 convolutional neural networks. Supported models:
16366 Super-Resolution Convolutional Neural Network model (SRCNN).
16367 See @url{https://arxiv.org/abs/1501.00092}.
16370 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16371 See @url{https://arxiv.org/abs/1609.05158}.
16374 Training scripts as well as scripts for model generation are provided in
16375 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16377 The filter accepts the following options:
16381 Specify which DNN backend to use for model loading and execution. This option accepts
16382 the following values:
16386 Native implementation of DNN loading and execution.
16389 TensorFlow backend. To enable this backend you
16390 need to install the TensorFlow for C library (see
16391 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16392 @code{--enable-libtensorflow}
16395 Default value is @samp{native}.
16398 Set path to model file specifying network architecture and its parameters.
16399 Note that different backends use different file formats. TensorFlow backend
16400 can load files for both formats, while native backend can load files for only
16404 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16405 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16406 input upscaled using bicubic upscaling with proper scale factor.
16412 Draw subtitles on top of input video using the libass library.
16414 To enable compilation of this filter you need to configure FFmpeg with
16415 @code{--enable-libass}. This filter also requires a build with libavcodec and
16416 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16417 Alpha) subtitles format.
16419 The filter accepts the following options:
16423 Set the filename of the subtitle file to read. It must be specified.
16425 @item original_size
16426 Specify the size of the original video, the video for which the ASS file
16427 was composed. For the syntax of this option, check the
16428 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16429 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16430 correctly scale the fonts if the aspect ratio has been changed.
16433 Set a directory path containing fonts that can be used by the filter.
16434 These fonts will be used in addition to whatever the font provider uses.
16437 Process alpha channel, by default alpha channel is untouched.
16440 Set subtitles input character encoding. @code{subtitles} filter only. Only
16441 useful if not UTF-8.
16443 @item stream_index, si
16444 Set subtitles stream index. @code{subtitles} filter only.
16447 Override default style or script info parameters of the subtitles. It accepts a
16448 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16451 If the first key is not specified, it is assumed that the first value
16452 specifies the @option{filename}.
16454 For example, to render the file @file{sub.srt} on top of the input
16455 video, use the command:
16460 which is equivalent to:
16462 subtitles=filename=sub.srt
16465 To render the default subtitles stream from file @file{video.mkv}, use:
16467 subtitles=video.mkv
16470 To render the second subtitles stream from that file, use:
16472 subtitles=video.mkv:si=1
16475 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16476 @code{DejaVu Serif}, use:
16478 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16481 @section super2xsai
16483 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16484 Interpolate) pixel art scaling algorithm.
16486 Useful for enlarging pixel art images without reducing sharpness.
16490 Swap two rectangular objects in video.
16492 This filter accepts the following options:
16502 Set 1st rect x coordinate.
16505 Set 1st rect y coordinate.
16508 Set 2nd rect x coordinate.
16511 Set 2nd rect y coordinate.
16513 All expressions are evaluated once for each frame.
16516 The all options are expressions containing the following constants:
16521 The input width and height.
16524 same as @var{w} / @var{h}
16527 input sample aspect ratio
16530 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16533 The number of the input frame, starting from 0.
16536 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16539 the position in the file of the input frame, NAN if unknown
16547 Apply telecine process to the video.
16549 This filter accepts the following options:
16558 The default value is @code{top}.
16562 A string of numbers representing the pulldown pattern you wish to apply.
16563 The default value is @code{23}.
16567 Some typical patterns:
16572 24p: 2332 (preferred)
16579 24p: 222222222223 ("Euro pulldown")
16586 Apply threshold effect to video stream.
16588 This filter needs four video streams to perform thresholding.
16589 First stream is stream we are filtering.
16590 Second stream is holding threshold values, third stream is holding min values,
16591 and last, fourth stream is holding max values.
16593 The filter accepts the following option:
16597 Set which planes will be processed, unprocessed planes will be copied.
16598 By default value 0xf, all planes will be processed.
16601 For example if first stream pixel's component value is less then threshold value
16602 of pixel component from 2nd threshold stream, third stream value will picked,
16603 otherwise fourth stream pixel component value will be picked.
16605 Using color source filter one can perform various types of thresholding:
16607 @subsection Examples
16611 Binary threshold, using gray color as threshold:
16613 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16617 Inverted binary threshold, using gray color as threshold:
16619 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16623 Truncate binary threshold, using gray color as threshold:
16625 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16629 Threshold to zero, using gray color as threshold:
16631 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16635 Inverted threshold to zero, using gray color as threshold:
16637 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16642 Select the most representative frame in a given sequence of consecutive frames.
16644 The filter accepts the following options:
16648 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16649 will pick one of them, and then handle the next batch of @var{n} frames until
16650 the end. Default is @code{100}.
16653 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16654 value will result in a higher memory usage, so a high value is not recommended.
16656 @subsection Examples
16660 Extract one picture each 50 frames:
16666 Complete example of a thumbnail creation with @command{ffmpeg}:
16668 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16674 Tile several successive frames together.
16676 The filter accepts the following options:
16681 Set the grid size (i.e. the number of lines and columns). For the syntax of
16682 this option, check the
16683 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16686 Set the maximum number of frames to render in the given area. It must be less
16687 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16688 the area will be used.
16691 Set the outer border margin in pixels.
16694 Set the inner border thickness (i.e. the number of pixels between frames). For
16695 more advanced padding options (such as having different values for the edges),
16696 refer to the pad video filter.
16699 Specify the color of the unused area. For the syntax of this option, check the
16700 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16701 The default value of @var{color} is "black".
16704 Set the number of frames to overlap when tiling several successive frames together.
16705 The value must be between @code{0} and @var{nb_frames - 1}.
16708 Set the number of frames to initially be empty before displaying first output frame.
16709 This controls how soon will one get first output frame.
16710 The value must be between @code{0} and @var{nb_frames - 1}.
16713 @subsection Examples
16717 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16719 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16721 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16722 duplicating each output frame to accommodate the originally detected frame
16726 Display @code{5} pictures in an area of @code{3x2} frames,
16727 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16728 mixed flat and named options:
16730 tile=3x2:nb_frames=5:padding=7:margin=2
16734 @section tinterlace
16736 Perform various types of temporal field interlacing.
16738 Frames are counted starting from 1, so the first input frame is
16741 The filter accepts the following options:
16746 Specify the mode of the interlacing. This option can also be specified
16747 as a value alone. See below for a list of values for this option.
16749 Available values are:
16753 Move odd frames into the upper field, even into the lower field,
16754 generating a double height frame at half frame rate.
16758 Frame 1 Frame 2 Frame 3 Frame 4
16760 11111 22222 33333 44444
16761 11111 22222 33333 44444
16762 11111 22222 33333 44444
16763 11111 22222 33333 44444
16777 Only output odd frames, even frames are dropped, generating a frame with
16778 unchanged height at half frame rate.
16783 Frame 1 Frame 2 Frame 3 Frame 4
16785 11111 22222 33333 44444
16786 11111 22222 33333 44444
16787 11111 22222 33333 44444
16788 11111 22222 33333 44444
16798 Only output even frames, odd frames are dropped, generating a frame with
16799 unchanged height at half frame rate.
16804 Frame 1 Frame 2 Frame 3 Frame 4
16806 11111 22222 33333 44444
16807 11111 22222 33333 44444
16808 11111 22222 33333 44444
16809 11111 22222 33333 44444
16819 Expand each frame to full height, but pad alternate lines with black,
16820 generating a frame with double height at the same input frame rate.
16825 Frame 1 Frame 2 Frame 3 Frame 4
16827 11111 22222 33333 44444
16828 11111 22222 33333 44444
16829 11111 22222 33333 44444
16830 11111 22222 33333 44444
16833 11111 ..... 33333 .....
16834 ..... 22222 ..... 44444
16835 11111 ..... 33333 .....
16836 ..... 22222 ..... 44444
16837 11111 ..... 33333 .....
16838 ..... 22222 ..... 44444
16839 11111 ..... 33333 .....
16840 ..... 22222 ..... 44444
16844 @item interleave_top, 4
16845 Interleave the upper field from odd frames with the lower field from
16846 even frames, generating a frame with unchanged height at half frame rate.
16851 Frame 1 Frame 2 Frame 3 Frame 4
16853 11111<- 22222 33333<- 44444
16854 11111 22222<- 33333 44444<-
16855 11111<- 22222 33333<- 44444
16856 11111 22222<- 33333 44444<-
16866 @item interleave_bottom, 5
16867 Interleave the lower field from odd frames with the upper field from
16868 even frames, generating a frame with unchanged height at half frame rate.
16873 Frame 1 Frame 2 Frame 3 Frame 4
16875 11111 22222<- 33333 44444<-
16876 11111<- 22222 33333<- 44444
16877 11111 22222<- 33333 44444<-
16878 11111<- 22222 33333<- 44444
16888 @item interlacex2, 6
16889 Double frame rate with unchanged height. Frames are inserted each
16890 containing the second temporal field from the previous input frame and
16891 the first temporal field from the next input frame. This mode relies on
16892 the top_field_first flag. Useful for interlaced video displays with no
16893 field synchronisation.
16898 Frame 1 Frame 2 Frame 3 Frame 4
16900 11111 22222 33333 44444
16901 11111 22222 33333 44444
16902 11111 22222 33333 44444
16903 11111 22222 33333 44444
16906 11111 22222 22222 33333 33333 44444 44444
16907 11111 11111 22222 22222 33333 33333 44444
16908 11111 22222 22222 33333 33333 44444 44444
16909 11111 11111 22222 22222 33333 33333 44444
16914 Move odd frames into the upper field, even into the lower field,
16915 generating a double height frame at same frame rate.
16920 Frame 1 Frame 2 Frame 3 Frame 4
16922 11111 22222 33333 44444
16923 11111 22222 33333 44444
16924 11111 22222 33333 44444
16925 11111 22222 33333 44444
16928 11111 33333 33333 55555
16929 22222 22222 44444 44444
16930 11111 33333 33333 55555
16931 22222 22222 44444 44444
16932 11111 33333 33333 55555
16933 22222 22222 44444 44444
16934 11111 33333 33333 55555
16935 22222 22222 44444 44444
16940 Numeric values are deprecated but are accepted for backward
16941 compatibility reasons.
16943 Default mode is @code{merge}.
16946 Specify flags influencing the filter process.
16948 Available value for @var{flags} is:
16951 @item low_pass_filter, vlfp
16952 Enable linear vertical low-pass filtering in the filter.
16953 Vertical low-pass filtering is required when creating an interlaced
16954 destination from a progressive source which contains high-frequency
16955 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16958 @item complex_filter, cvlfp
16959 Enable complex vertical low-pass filtering.
16960 This will slightly less reduce interlace 'twitter' and Moire
16961 patterning but better retain detail and subjective sharpness impression.
16965 Vertical low-pass filtering can only be enabled for @option{mode}
16966 @var{interleave_top} and @var{interleave_bottom}.
16972 Mix successive video frames.
16974 A description of the accepted options follows.
16978 The number of successive frames to mix. If unspecified, it defaults to 3.
16981 Specify weight of each input video frame.
16982 Each weight is separated by space. If number of weights is smaller than
16983 number of @var{frames} last specified weight will be used for all remaining
16987 Specify scale, if it is set it will be multiplied with sum
16988 of each weight multiplied with pixel values to give final destination
16989 pixel value. By default @var{scale} is auto scaled to sum of weights.
16992 @subsection Examples
16996 Average 7 successive frames:
16998 tmix=frames=7:weights="1 1 1 1 1 1 1"
17002 Apply simple temporal convolution:
17004 tmix=frames=3:weights="-1 3 -1"
17008 Similar as above but only showing temporal differences:
17010 tmix=frames=3:weights="-1 2 -1":scale=1
17016 Tone map colors from different dynamic ranges.
17018 This filter expects data in single precision floating point, as it needs to
17019 operate on (and can output) out-of-range values. Another filter, such as
17020 @ref{zscale}, is needed to convert the resulting frame to a usable format.
17022 The tonemapping algorithms implemented only work on linear light, so input
17023 data should be linearized beforehand (and possibly correctly tagged).
17026 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
17029 @subsection Options
17030 The filter accepts the following options.
17034 Set the tone map algorithm to use.
17036 Possible values are:
17039 Do not apply any tone map, only desaturate overbright pixels.
17042 Hard-clip any out-of-range values. Use it for perfect color accuracy for
17043 in-range values, while distorting out-of-range values.
17046 Stretch the entire reference gamut to a linear multiple of the display.
17049 Fit a logarithmic transfer between the tone curves.
17052 Preserve overall image brightness with a simple curve, using nonlinear
17053 contrast, which results in flattening details and degrading color accuracy.
17056 Preserve both dark and bright details better than @var{reinhard}, at the cost
17057 of slightly darkening everything. Use it when detail preservation is more
17058 important than color and brightness accuracy.
17061 Smoothly map out-of-range values, while retaining contrast and colors for
17062 in-range material as much as possible. Use it when color accuracy is more
17063 important than detail preservation.
17069 Tune the tone mapping algorithm.
17071 This affects the following algorithms:
17077 Specifies the scale factor to use while stretching.
17081 Specifies the exponent of the function.
17085 Specify an extra linear coefficient to multiply into the signal before clipping.
17089 Specify the local contrast coefficient at the display peak.
17090 Default to 0.5, which means that in-gamut values will be about half as bright
17097 Specify the transition point from linear to mobius transform. Every value
17098 below this point is guaranteed to be mapped 1:1. The higher the value, the
17099 more accurate the result will be, at the cost of losing bright details.
17100 Default to 0.3, which due to the steep initial slope still preserves in-range
17101 colors fairly accurately.
17105 Apply desaturation for highlights that exceed this level of brightness. The
17106 higher the parameter, the more color information will be preserved. This
17107 setting helps prevent unnaturally blown-out colors for super-highlights, by
17108 (smoothly) turning into white instead. This makes images feel more natural,
17109 at the cost of reducing information about out-of-range colors.
17111 The default of 2.0 is somewhat conservative and will mostly just apply to
17112 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17114 This option works only if the input frame has a supported color tag.
17117 Override signal/nominal/reference peak with this value. Useful when the
17118 embedded peak information in display metadata is not reliable or when tone
17119 mapping from a lower range to a higher range.
17124 Temporarily pad video frames.
17126 The filter accepts the following options:
17130 Specify number of delay frames before input video stream.
17133 Specify number of padding frames after input video stream.
17134 Set to -1 to pad indefinitely.
17137 Set kind of frames added to beginning of stream.
17138 Can be either @var{add} or @var{clone}.
17139 With @var{add} frames of solid-color are added.
17140 With @var{clone} frames are clones of first frame.
17143 Set kind of frames added to end of stream.
17144 Can be either @var{add} or @var{clone}.
17145 With @var{add} frames of solid-color are added.
17146 With @var{clone} frames are clones of last frame.
17148 @item start_duration, stop_duration
17149 Specify the duration of the start/stop delay. See
17150 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17151 for the accepted syntax.
17152 These options override @var{start} and @var{stop}.
17155 Specify the color of the padded area. For the syntax of this option,
17156 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17157 manual,ffmpeg-utils}.
17159 The default value of @var{color} is "black".
17165 Transpose rows with columns in the input video and optionally flip it.
17167 It accepts the following parameters:
17172 Specify the transposition direction.
17174 Can assume the following values:
17176 @item 0, 4, cclock_flip
17177 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17185 Rotate by 90 degrees clockwise, that is:
17193 Rotate by 90 degrees counterclockwise, that is:
17200 @item 3, 7, clock_flip
17201 Rotate by 90 degrees clockwise and vertically flip, that is:
17209 For values between 4-7, the transposition is only done if the input
17210 video geometry is portrait and not landscape. These values are
17211 deprecated, the @code{passthrough} option should be used instead.
17213 Numerical values are deprecated, and should be dropped in favor of
17214 symbolic constants.
17217 Do not apply the transposition if the input geometry matches the one
17218 specified by the specified value. It accepts the following values:
17221 Always apply transposition.
17223 Preserve portrait geometry (when @var{height} >= @var{width}).
17225 Preserve landscape geometry (when @var{width} >= @var{height}).
17228 Default value is @code{none}.
17231 For example to rotate by 90 degrees clockwise and preserve portrait
17234 transpose=dir=1:passthrough=portrait
17237 The command above can also be specified as:
17239 transpose=1:portrait
17242 @section transpose_npp
17244 Transpose rows with columns in the input video and optionally flip it.
17245 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17247 It accepts the following parameters:
17252 Specify the transposition direction.
17254 Can assume the following values:
17257 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17260 Rotate by 90 degrees clockwise.
17263 Rotate by 90 degrees counterclockwise.
17266 Rotate by 90 degrees clockwise and vertically flip.
17270 Do not apply the transposition if the input geometry matches the one
17271 specified by the specified value. It accepts the following values:
17274 Always apply transposition. (default)
17276 Preserve portrait geometry (when @var{height} >= @var{width}).
17278 Preserve landscape geometry (when @var{width} >= @var{height}).
17284 Trim the input so that the output contains one continuous subpart of the input.
17286 It accepts the following parameters:
17289 Specify the time of the start of the kept section, i.e. the frame with the
17290 timestamp @var{start} will be the first frame in the output.
17293 Specify the time of the first frame that will be dropped, i.e. the frame
17294 immediately preceding the one with the timestamp @var{end} will be the last
17295 frame in the output.
17298 This is the same as @var{start}, except this option sets the start timestamp
17299 in timebase units instead of seconds.
17302 This is the same as @var{end}, except this option sets the end timestamp
17303 in timebase units instead of seconds.
17306 The maximum duration of the output in seconds.
17309 The number of the first frame that should be passed to the output.
17312 The number of the first frame that should be dropped.
17315 @option{start}, @option{end}, and @option{duration} are expressed as time
17316 duration specifications; see
17317 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17318 for the accepted syntax.
17320 Note that the first two sets of the start/end options and the @option{duration}
17321 option look at the frame timestamp, while the _frame variants simply count the
17322 frames that pass through the filter. Also note that this filter does not modify
17323 the timestamps. If you wish for the output timestamps to start at zero, insert a
17324 setpts filter after the trim filter.
17326 If multiple start or end options are set, this filter tries to be greedy and
17327 keep all the frames that match at least one of the specified constraints. To keep
17328 only the part that matches all the constraints at once, chain multiple trim
17331 The defaults are such that all the input is kept. So it is possible to set e.g.
17332 just the end values to keep everything before the specified time.
17337 Drop everything except the second minute of input:
17339 ffmpeg -i INPUT -vf trim=60:120
17343 Keep only the first second:
17345 ffmpeg -i INPUT -vf trim=duration=1
17350 @section unpremultiply
17351 Apply alpha unpremultiply effect to input video stream using first plane
17352 of second stream as alpha.
17354 Both streams must have same dimensions and same pixel format.
17356 The filter accepts the following option:
17360 Set which planes will be processed, unprocessed planes will be copied.
17361 By default value 0xf, all planes will be processed.
17363 If the format has 1 or 2 components, then luma is bit 0.
17364 If the format has 3 or 4 components:
17365 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17366 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17367 If present, the alpha channel is always the last bit.
17370 Do not require 2nd input for processing, instead use alpha plane from input stream.
17376 Sharpen or blur the input video.
17378 It accepts the following parameters:
17381 @item luma_msize_x, lx
17382 Set the luma matrix horizontal size. It must be an odd integer between
17383 3 and 23. The default value is 5.
17385 @item luma_msize_y, ly
17386 Set the luma matrix vertical size. It must be an odd integer between 3
17387 and 23. The default value is 5.
17389 @item luma_amount, la
17390 Set the luma effect strength. It must be a floating point number, reasonable
17391 values lay between -1.5 and 1.5.
17393 Negative values will blur the input video, while positive values will
17394 sharpen it, a value of zero will disable the effect.
17396 Default value is 1.0.
17398 @item chroma_msize_x, cx
17399 Set the chroma matrix horizontal size. It must be an odd integer
17400 between 3 and 23. The default value is 5.
17402 @item chroma_msize_y, cy
17403 Set the chroma matrix vertical size. It must be an odd integer
17404 between 3 and 23. The default value is 5.
17406 @item chroma_amount, ca
17407 Set the chroma effect strength. It must be a floating point number, reasonable
17408 values lay between -1.5 and 1.5.
17410 Negative values will blur the input video, while positive values will
17411 sharpen it, a value of zero will disable the effect.
17413 Default value is 0.0.
17417 All parameters are optional and default to the equivalent of the
17418 string '5:5:1.0:5:5:0.0'.
17420 @subsection Examples
17424 Apply strong luma sharpen effect:
17426 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17430 Apply a strong blur of both luma and chroma parameters:
17432 unsharp=7:7:-2:7:7:-2
17438 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17439 the image at several (or - in the case of @option{quality} level @code{8} - all)
17440 shifts and average the results.
17442 The way this differs from the behavior of spp is that uspp actually encodes &
17443 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17444 DCT similar to MJPEG.
17446 The filter accepts the following options:
17450 Set quality. This option defines the number of levels for averaging. It accepts
17451 an integer in the range 0-8. If set to @code{0}, the filter will have no
17452 effect. A value of @code{8} means the higher quality. For each increment of
17453 that value the speed drops by a factor of approximately 2. Default value is
17457 Force a constant quantization parameter. If not set, the filter will use the QP
17458 from the video stream (if available).
17461 @section vaguedenoiser
17463 Apply a wavelet based denoiser.
17465 It transforms each frame from the video input into the wavelet domain,
17466 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17467 the obtained coefficients. It does an inverse wavelet transform after.
17468 Due to wavelet properties, it should give a nice smoothed result, and
17469 reduced noise, without blurring picture features.
17471 This filter accepts the following options:
17475 The filtering strength. The higher, the more filtered the video will be.
17476 Hard thresholding can use a higher threshold than soft thresholding
17477 before the video looks overfiltered. Default value is 2.
17480 The filtering method the filter will use.
17482 It accepts the following values:
17485 All values under the threshold will be zeroed.
17488 All values under the threshold will be zeroed. All values above will be
17489 reduced by the threshold.
17492 Scales or nullifies coefficients - intermediary between (more) soft and
17493 (less) hard thresholding.
17496 Default is garrote.
17499 Number of times, the wavelet will decompose the picture. Picture can't
17500 be decomposed beyond a particular point (typically, 8 for a 640x480
17501 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17504 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17507 A list of the planes to process. By default all planes are processed.
17510 @section vectorscope
17512 Display 2 color component values in the two dimensional graph (which is called
17515 This filter accepts the following options:
17519 Set vectorscope mode.
17521 It accepts the following values:
17524 Gray values are displayed on graph, higher brightness means more pixels have
17525 same component color value on location in graph. This is the default mode.
17528 Gray values are displayed on graph. Surrounding pixels values which are not
17529 present in video frame are drawn in gradient of 2 color components which are
17530 set by option @code{x} and @code{y}. The 3rd color component is static.
17533 Actual color components values present in video frame are displayed on graph.
17536 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17537 on graph increases value of another color component, which is luminance by
17538 default values of @code{x} and @code{y}.
17541 Actual colors present in video frame are displayed on graph. If two different
17542 colors map to same position on graph then color with higher value of component
17543 not present in graph is picked.
17546 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17547 component picked from radial gradient.
17551 Set which color component will be represented on X-axis. Default is @code{1}.
17554 Set which color component will be represented on Y-axis. Default is @code{2}.
17557 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17558 of color component which represents frequency of (X, Y) location in graph.
17563 No envelope, this is default.
17566 Instant envelope, even darkest single pixel will be clearly highlighted.
17569 Hold maximum and minimum values presented in graph over time. This way you
17570 can still spot out of range values without constantly looking at vectorscope.
17573 Peak and instant envelope combined together.
17577 Set what kind of graticule to draw.
17585 Set graticule opacity.
17588 Set graticule flags.
17592 Draw graticule for white point.
17595 Draw graticule for black point.
17598 Draw color points short names.
17602 Set background opacity.
17604 @item lthreshold, l
17605 Set low threshold for color component not represented on X or Y axis.
17606 Values lower than this value will be ignored. Default is 0.
17607 Note this value is multiplied with actual max possible value one pixel component
17608 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17611 @item hthreshold, h
17612 Set high threshold for color component not represented on X or Y axis.
17613 Values higher than this value will be ignored. Default is 1.
17614 Note this value is multiplied with actual max possible value one pixel component
17615 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17616 is 0.9 * 255 = 230.
17618 @item colorspace, c
17619 Set what kind of colorspace to use when drawing graticule.
17628 @anchor{vidstabdetect}
17629 @section vidstabdetect
17631 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17632 @ref{vidstabtransform} for pass 2.
17634 This filter generates a file with relative translation and rotation
17635 transform information about subsequent frames, which is then used by
17636 the @ref{vidstabtransform} filter.
17638 To enable compilation of this filter you need to configure FFmpeg with
17639 @code{--enable-libvidstab}.
17641 This filter accepts the following options:
17645 Set the path to the file used to write the transforms information.
17646 Default value is @file{transforms.trf}.
17649 Set how shaky the video is and how quick the camera is. It accepts an
17650 integer in the range 1-10, a value of 1 means little shakiness, a
17651 value of 10 means strong shakiness. Default value is 5.
17654 Set the accuracy of the detection process. It must be a value in the
17655 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17656 accuracy. Default value is 15.
17659 Set stepsize of the search process. The region around minimum is
17660 scanned with 1 pixel resolution. Default value is 6.
17663 Set minimum contrast. Below this value a local measurement field is
17664 discarded. Must be a floating point value in the range 0-1. Default
17668 Set reference frame number for tripod mode.
17670 If enabled, the motion of the frames is compared to a reference frame
17671 in the filtered stream, identified by the specified number. The idea
17672 is to compensate all movements in a more-or-less static scene and keep
17673 the camera view absolutely still.
17675 If set to 0, it is disabled. The frames are counted starting from 1.
17678 Show fields and transforms in the resulting frames. It accepts an
17679 integer in the range 0-2. Default value is 0, which disables any
17683 @subsection Examples
17687 Use default values:
17693 Analyze strongly shaky movie and put the results in file
17694 @file{mytransforms.trf}:
17696 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17700 Visualize the result of internal transformations in the resulting
17703 vidstabdetect=show=1
17707 Analyze a video with medium shakiness using @command{ffmpeg}:
17709 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17713 @anchor{vidstabtransform}
17714 @section vidstabtransform
17716 Video stabilization/deshaking: pass 2 of 2,
17717 see @ref{vidstabdetect} for pass 1.
17719 Read a file with transform information for each frame and
17720 apply/compensate them. Together with the @ref{vidstabdetect}
17721 filter this can be used to deshake videos. See also
17722 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17723 the @ref{unsharp} filter, see below.
17725 To enable compilation of this filter you need to configure FFmpeg with
17726 @code{--enable-libvidstab}.
17728 @subsection Options
17732 Set path to the file used to read the transforms. Default value is
17733 @file{transforms.trf}.
17736 Set the number of frames (value*2 + 1) used for lowpass filtering the
17737 camera movements. Default value is 10.
17739 For example a number of 10 means that 21 frames are used (10 in the
17740 past and 10 in the future) to smoothen the motion in the video. A
17741 larger value leads to a smoother video, but limits the acceleration of
17742 the camera (pan/tilt movements). 0 is a special case where a static
17743 camera is simulated.
17746 Set the camera path optimization algorithm.
17748 Accepted values are:
17751 gaussian kernel low-pass filter on camera motion (default)
17753 averaging on transformations
17757 Set maximal number of pixels to translate frames. Default value is -1,
17761 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17762 value is -1, meaning no limit.
17765 Specify how to deal with borders that may be visible due to movement
17768 Available values are:
17771 keep image information from previous frame (default)
17773 fill the border black
17777 Invert transforms if set to 1. Default value is 0.
17780 Consider transforms as relative to previous frame if set to 1,
17781 absolute if set to 0. Default value is 0.
17784 Set percentage to zoom. A positive value will result in a zoom-in
17785 effect, a negative value in a zoom-out effect. Default value is 0 (no
17789 Set optimal zooming to avoid borders.
17791 Accepted values are:
17796 optimal static zoom value is determined (only very strong movements
17797 will lead to visible borders) (default)
17799 optimal adaptive zoom value is determined (no borders will be
17800 visible), see @option{zoomspeed}
17803 Note that the value given at zoom is added to the one calculated here.
17806 Set percent to zoom maximally each frame (enabled when
17807 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17811 Specify type of interpolation.
17813 Available values are:
17818 linear only horizontal
17820 linear in both directions (default)
17822 cubic in both directions (slow)
17826 Enable virtual tripod mode if set to 1, which is equivalent to
17827 @code{relative=0:smoothing=0}. Default value is 0.
17829 Use also @code{tripod} option of @ref{vidstabdetect}.
17832 Increase log verbosity if set to 1. Also the detected global motions
17833 are written to the temporary file @file{global_motions.trf}. Default
17837 @subsection Examples
17841 Use @command{ffmpeg} for a typical stabilization with default values:
17843 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17846 Note the use of the @ref{unsharp} filter which is always recommended.
17849 Zoom in a bit more and load transform data from a given file:
17851 vidstabtransform=zoom=5:input="mytransforms.trf"
17855 Smoothen the video even more:
17857 vidstabtransform=smoothing=30
17863 Flip the input video vertically.
17865 For example, to vertically flip a video with @command{ffmpeg}:
17867 ffmpeg -i in.avi -vf "vflip" out.avi
17872 Detect variable frame rate video.
17874 This filter tries to detect if the input is variable or constant frame rate.
17876 At end it will output number of frames detected as having variable delta pts,
17877 and ones with constant delta pts.
17878 If there was frames with variable delta, than it will also show min and max delta
17883 Boost or alter saturation.
17885 The filter accepts the following options:
17888 Set strength of boost if positive value or strength of alter if negative value.
17889 Default is 0. Allowed range is from -2 to 2.
17892 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17895 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17898 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17901 Set the red luma coefficient.
17904 Set the green luma coefficient.
17907 Set the blue luma coefficient.
17913 Make or reverse a natural vignetting effect.
17915 The filter accepts the following options:
17919 Set lens angle expression as a number of radians.
17921 The value is clipped in the @code{[0,PI/2]} range.
17923 Default value: @code{"PI/5"}
17927 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17931 Set forward/backward mode.
17933 Available modes are:
17936 The larger the distance from the central point, the darker the image becomes.
17939 The larger the distance from the central point, the brighter the image becomes.
17940 This can be used to reverse a vignette effect, though there is no automatic
17941 detection to extract the lens @option{angle} and other settings (yet). It can
17942 also be used to create a burning effect.
17945 Default value is @samp{forward}.
17948 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17950 It accepts the following values:
17953 Evaluate expressions only once during the filter initialization.
17956 Evaluate expressions for each incoming frame. This is way slower than the
17957 @samp{init} mode since it requires all the scalers to be re-computed, but it
17958 allows advanced dynamic expressions.
17961 Default value is @samp{init}.
17964 Set dithering to reduce the circular banding effects. Default is @code{1}
17968 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17969 Setting this value to the SAR of the input will make a rectangular vignetting
17970 following the dimensions of the video.
17972 Default is @code{1/1}.
17975 @subsection Expressions
17977 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17978 following parameters.
17983 input width and height
17986 the number of input frame, starting from 0
17989 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17990 @var{TB} units, NAN if undefined
17993 frame rate of the input video, NAN if the input frame rate is unknown
17996 the PTS (Presentation TimeStamp) of the filtered video frame,
17997 expressed in seconds, NAN if undefined
18000 time base of the input video
18004 @subsection Examples
18008 Apply simple strong vignetting effect:
18014 Make a flickering vignetting:
18016 vignette='PI/4+random(1)*PI/50':eval=frame
18021 @section vmafmotion
18023 Obtain the average vmaf motion score of a video.
18024 It is one of the component filters of VMAF.
18026 The obtained average motion score is printed through the logging system.
18028 In the below example the input file @file{ref.mpg} is being processed and score
18032 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
18036 Stack input videos vertically.
18038 All streams must be of same pixel format and of same width.
18040 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
18041 to create same output.
18043 The filter accept the following option:
18047 Set number of input streams. Default is 2.
18050 If set to 1, force the output to terminate when the shortest input
18051 terminates. Default value is 0.
18056 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
18057 Deinterlacing Filter").
18059 Based on the process described by Martin Weston for BBC R&D, and
18060 implemented based on the de-interlace algorithm written by Jim
18061 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
18062 uses filter coefficients calculated by BBC R&D.
18064 There are two sets of filter coefficients, so called "simple":
18065 and "complex". Which set of filter coefficients is used can
18066 be set by passing an optional parameter:
18070 Set the interlacing filter coefficients. Accepts one of the following values:
18074 Simple filter coefficient set.
18076 More-complex filter coefficient set.
18078 Default value is @samp{complex}.
18081 Specify which frames to deinterlace. Accept one of the following values:
18085 Deinterlace all frames,
18087 Only deinterlace frames marked as interlaced.
18090 Default value is @samp{all}.
18094 Video waveform monitor.
18096 The waveform monitor plots color component intensity. By default luminance
18097 only. Each column of the waveform corresponds to a column of pixels in the
18100 It accepts the following options:
18104 Can be either @code{row}, or @code{column}. Default is @code{column}.
18105 In row mode, the graph on the left side represents color component value 0 and
18106 the right side represents value = 255. In column mode, the top side represents
18107 color component value = 0 and bottom side represents value = 255.
18110 Set intensity. Smaller values are useful to find out how many values of the same
18111 luminance are distributed across input rows/columns.
18112 Default value is @code{0.04}. Allowed range is [0, 1].
18115 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18116 In mirrored mode, higher values will be represented on the left
18117 side for @code{row} mode and at the top for @code{column} mode. Default is
18118 @code{1} (mirrored).
18122 It accepts the following values:
18125 Presents information identical to that in the @code{parade}, except
18126 that the graphs representing color components are superimposed directly
18129 This display mode makes it easier to spot relative differences or similarities
18130 in overlapping areas of the color components that are supposed to be identical,
18131 such as neutral whites, grays, or blacks.
18134 Display separate graph for the color components side by side in
18135 @code{row} mode or one below the other in @code{column} mode.
18138 Display separate graph for the color components side by side in
18139 @code{column} mode or one below the other in @code{row} mode.
18141 Using this display mode makes it easy to spot color casts in the highlights
18142 and shadows of an image, by comparing the contours of the top and the bottom
18143 graphs of each waveform. Since whites, grays, and blacks are characterized
18144 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18145 should display three waveforms of roughly equal width/height. If not, the
18146 correction is easy to perform by making level adjustments the three waveforms.
18148 Default is @code{stack}.
18150 @item components, c
18151 Set which color components to display. Default is 1, which means only luminance
18152 or red color component if input is in RGB colorspace. If is set for example to
18153 7 it will display all 3 (if) available color components.
18158 No envelope, this is default.
18161 Instant envelope, minimum and maximum values presented in graph will be easily
18162 visible even with small @code{step} value.
18165 Hold minimum and maximum values presented in graph across time. This way you
18166 can still spot out of range values without constantly looking at waveforms.
18169 Peak and instant envelope combined together.
18175 No filtering, this is default.
18178 Luma and chroma combined together.
18181 Similar as above, but shows difference between blue and red chroma.
18184 Similar as above, but use different colors.
18187 Displays only chroma.
18190 Displays actual color value on waveform.
18193 Similar as above, but with luma showing frequency of chroma values.
18197 Set which graticule to display.
18201 Do not display graticule.
18204 Display green graticule showing legal broadcast ranges.
18207 Display orange graticule showing legal broadcast ranges.
18211 Set graticule opacity.
18214 Set graticule flags.
18218 Draw numbers above lines. By default enabled.
18221 Draw dots instead of lines.
18225 Set scale used for displaying graticule.
18232 Default is digital.
18235 Set background opacity.
18238 @section weave, doubleweave
18240 The @code{weave} takes a field-based video input and join
18241 each two sequential fields into single frame, producing a new double
18242 height clip with half the frame rate and half the frame count.
18244 The @code{doubleweave} works same as @code{weave} but without
18245 halving frame rate and frame count.
18247 It accepts the following option:
18251 Set first field. Available values are:
18255 Set the frame as top-field-first.
18258 Set the frame as bottom-field-first.
18262 @subsection Examples
18266 Interlace video using @ref{select} and @ref{separatefields} filter:
18268 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18273 Apply the xBR high-quality magnification filter which is designed for pixel
18274 art. It follows a set of edge-detection rules, see
18275 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18277 It accepts the following option:
18281 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18282 @code{3xBR} and @code{4} for @code{4xBR}.
18283 Default is @code{3}.
18287 Stack video inputs into custom layout.
18289 All streams must be of same pixel format.
18291 The filter accept the following option:
18295 Set number of input streams. Default is 2.
18298 Specify layout of inputs.
18299 This option requires the desired layout configuration to be explicitly set by the user.
18300 This sets position of each video input in output. Each input
18301 is separated by '|'.
18302 The first number represents the column, and the second number represents the row.
18303 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18304 where X is video input from which to take width or height.
18305 Multiple values can be used when separated by '+'. In such
18306 case values are summed together.
18309 If set to 1, force the output to terminate when the shortest input
18310 terminates. Default value is 0.
18313 @subsection Examples
18317 Display 4 inputs into 2x2 grid,
18318 note that if inputs are of different sizes unused gaps might appear,
18319 as not all of output video is used.
18321 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18325 Display 4 inputs into 1x4 grid,
18326 note that if inputs are of different sizes unused gaps might appear,
18327 as not all of output video is used.
18329 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18333 Display 9 inputs into 3x3 grid,
18334 note that if inputs are of different sizes unused gaps might appear,
18335 as not all of output video is used.
18337 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
18344 Deinterlace the input video ("yadif" means "yet another deinterlacing
18347 It accepts the following parameters:
18353 The interlacing mode to adopt. It accepts one of the following values:
18356 @item 0, send_frame
18357 Output one frame for each frame.
18358 @item 1, send_field
18359 Output one frame for each field.
18360 @item 2, send_frame_nospatial
18361 Like @code{send_frame}, but it skips the spatial interlacing check.
18362 @item 3, send_field_nospatial
18363 Like @code{send_field}, but it skips the spatial interlacing check.
18366 The default value is @code{send_frame}.
18369 The picture field parity assumed for the input interlaced video. It accepts one
18370 of the following values:
18374 Assume the top field is first.
18376 Assume the bottom field is first.
18378 Enable automatic detection of field parity.
18381 The default value is @code{auto}.
18382 If the interlacing is unknown or the decoder does not export this information,
18383 top field first will be assumed.
18386 Specify which frames to deinterlace. Accept one of the following
18391 Deinterlace all frames.
18392 @item 1, interlaced
18393 Only deinterlace frames marked as interlaced.
18396 The default value is @code{all}.
18399 @section yadif_cuda
18401 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18402 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18405 It accepts the following parameters:
18411 The interlacing mode to adopt. It accepts one of the following values:
18414 @item 0, send_frame
18415 Output one frame for each frame.
18416 @item 1, send_field
18417 Output one frame for each field.
18418 @item 2, send_frame_nospatial
18419 Like @code{send_frame}, but it skips the spatial interlacing check.
18420 @item 3, send_field_nospatial
18421 Like @code{send_field}, but it skips the spatial interlacing check.
18424 The default value is @code{send_frame}.
18427 The picture field parity assumed for the input interlaced video. It accepts one
18428 of the following values:
18432 Assume the top field is first.
18434 Assume the bottom field is first.
18436 Enable automatic detection of field parity.
18439 The default value is @code{auto}.
18440 If the interlacing is unknown or the decoder does not export this information,
18441 top field first will be assumed.
18444 Specify which frames to deinterlace. Accept one of the following
18449 Deinterlace all frames.
18450 @item 1, interlaced
18451 Only deinterlace frames marked as interlaced.
18454 The default value is @code{all}.
18459 Apply Zoom & Pan effect.
18461 This filter accepts the following options:
18465 Set the zoom expression. Range is 1-10. Default is 1.
18469 Set the x and y expression. Default is 0.
18472 Set the duration expression in number of frames.
18473 This sets for how many number of frames effect will last for
18474 single input image.
18477 Set the output image size, default is 'hd720'.
18480 Set the output frame rate, default is '25'.
18483 Each expression can contain the following constants:
18502 Output frame count.
18506 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18507 for current input frame.
18511 'x' and 'y' of last output frame of previous input frame or 0 when there was
18512 not yet such frame (first input frame).
18515 Last calculated zoom from 'z' expression for current input frame.
18518 Last calculated zoom of last output frame of previous input frame.
18521 Number of output frames for current input frame. Calculated from 'd' expression
18522 for each input frame.
18525 number of output frames created for previous input frame
18528 Rational number: input width / input height
18531 sample aspect ratio
18534 display aspect ratio
18538 @subsection Examples
18542 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18544 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
18548 Zoom-in up to 1.5 and pan always at center of picture:
18550 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18554 Same as above but without pausing:
18556 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18562 Scale (resize) the input video, using the z.lib library:
18563 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18564 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18566 The zscale filter forces the output display aspect ratio to be the same
18567 as the input, by changing the output sample aspect ratio.
18569 If the input image format is different from the format requested by
18570 the next filter, the zscale filter will convert the input to the
18573 @subsection Options
18574 The filter accepts the following options.
18579 Set the output video dimension expression. Default value is the input
18582 If the @var{width} or @var{w} value is 0, the input width is used for
18583 the output. If the @var{height} or @var{h} value is 0, the input height
18584 is used for the output.
18586 If one and only one of the values is -n with n >= 1, the zscale filter
18587 will use a value that maintains the aspect ratio of the input image,
18588 calculated from the other specified dimension. After that it will,
18589 however, make sure that the calculated dimension is divisible by n and
18590 adjust the value if necessary.
18592 If both values are -n with n >= 1, the behavior will be identical to
18593 both values being set to 0 as previously detailed.
18595 See below for the list of accepted constants for use in the dimension
18599 Set the video size. For the syntax of this option, check the
18600 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18603 Set the dither type.
18605 Possible values are:
18610 @item error_diffusion
18616 Set the resize filter type.
18618 Possible values are:
18628 Default is bilinear.
18631 Set the color range.
18633 Possible values are:
18640 Default is same as input.
18643 Set the color primaries.
18645 Possible values are:
18655 Default is same as input.
18658 Set the transfer characteristics.
18660 Possible values are:
18674 Default is same as input.
18677 Set the colorspace matrix.
18679 Possible value are:
18690 Default is same as input.
18693 Set the input color range.
18695 Possible values are:
18702 Default is same as input.
18704 @item primariesin, pin
18705 Set the input color primaries.
18707 Possible values are:
18717 Default is same as input.
18719 @item transferin, tin
18720 Set the input transfer characteristics.
18722 Possible values are:
18733 Default is same as input.
18735 @item matrixin, min
18736 Set the input colorspace matrix.
18738 Possible value are:
18750 Set the output chroma location.
18752 Possible values are:
18763 @item chromalin, cin
18764 Set the input chroma location.
18766 Possible values are:
18778 Set the nominal peak luminance.
18781 The values of the @option{w} and @option{h} options are expressions
18782 containing the following constants:
18787 The input width and height
18791 These are the same as @var{in_w} and @var{in_h}.
18795 The output (scaled) width and height
18799 These are the same as @var{out_w} and @var{out_h}
18802 The same as @var{iw} / @var{ih}
18805 input sample aspect ratio
18808 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18812 horizontal and vertical input chroma subsample values. For example for the
18813 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18817 horizontal and vertical output chroma subsample values. For example for the
18818 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18824 @c man end VIDEO FILTERS
18826 @chapter OpenCL Video Filters
18827 @c man begin OPENCL VIDEO FILTERS
18829 Below is a description of the currently available OpenCL video filters.
18831 To enable compilation of these filters you need to configure FFmpeg with
18832 @code{--enable-opencl}.
18834 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18837 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18838 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18839 given device parameters.
18841 @item -filter_hw_device @var{name}
18842 Pass the hardware device called @var{name} to all filters in any filter graph.
18846 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18850 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18852 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18856 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.
18858 @section avgblur_opencl
18860 Apply average blur filter.
18862 The filter accepts the following options:
18866 Set horizontal radius size.
18867 Range is @code{[1, 1024]} and default value is @code{1}.
18870 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18873 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18876 @subsection Example
18880 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.
18882 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18886 @section boxblur_opencl
18888 Apply a boxblur algorithm to the input video.
18890 It accepts the following parameters:
18894 @item luma_radius, lr
18895 @item luma_power, lp
18896 @item chroma_radius, cr
18897 @item chroma_power, cp
18898 @item alpha_radius, ar
18899 @item alpha_power, ap
18903 A description of the accepted options follows.
18906 @item luma_radius, lr
18907 @item chroma_radius, cr
18908 @item alpha_radius, ar
18909 Set an expression for the box radius in pixels used for blurring the
18910 corresponding input plane.
18912 The radius value must be a non-negative number, and must not be
18913 greater than the value of the expression @code{min(w,h)/2} for the
18914 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18917 Default value for @option{luma_radius} is "2". If not specified,
18918 @option{chroma_radius} and @option{alpha_radius} default to the
18919 corresponding value set for @option{luma_radius}.
18921 The expressions can contain the following constants:
18925 The input width and height in pixels.
18929 The input chroma image width and height in pixels.
18933 The horizontal and vertical chroma subsample values. For example, for the
18934 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
18937 @item luma_power, lp
18938 @item chroma_power, cp
18939 @item alpha_power, ap
18940 Specify how many times the boxblur filter is applied to the
18941 corresponding plane.
18943 Default value for @option{luma_power} is 2. If not specified,
18944 @option{chroma_power} and @option{alpha_power} default to the
18945 corresponding value set for @option{luma_power}.
18947 A value of 0 will disable the effect.
18950 @subsection Examples
18952 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.
18956 Apply a boxblur filter with the luma, chroma, and alpha radius
18957 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.
18959 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
18960 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
18964 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.
18966 For the luma plane, a 2x2 box radius will be run once.
18968 For the chroma plane, a 4x4 box radius will be run 5 times.
18970 For the alpha plane, a 3x3 box radius will be run 7 times.
18972 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
18976 @section convolution_opencl
18978 Apply convolution of 3x3, 5x5, 7x7 matrix.
18980 The filter accepts the following options:
18987 Set matrix for each plane.
18988 Matrix is sequence of 9, 25 or 49 signed numbers.
18989 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
18995 Set multiplier for calculated value for each plane.
18996 If unset or 0, it will be sum of all matrix elements.
18997 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
19003 Set bias for each plane. This value is added to the result of the multiplication.
19004 Useful for making the overall image brighter or darker.
19005 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
19009 @subsection Examples
19015 -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
19021 -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
19025 Apply edge enhance:
19027 -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
19033 -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
19037 Apply laplacian edge detector which includes diagonals:
19039 -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
19045 -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
19049 @section dilation_opencl
19051 Apply dilation effect to the video.
19053 This filter replaces the pixel by the local(3x3) maximum.
19055 It accepts the following options:
19062 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19063 If @code{0}, plane will remain unchanged.
19066 Flag which specifies the pixel to refer to.
19067 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19069 Flags to local 3x3 coordinates region centered on @code{x}:
19078 @subsection Example
19082 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.
19084 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19088 @section erosion_opencl
19090 Apply erosion effect to the video.
19092 This filter replaces the pixel by the local(3x3) minimum.
19094 It accepts the following options:
19101 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
19102 If @code{0}, plane will remain unchanged.
19105 Flag which specifies the pixel to refer to.
19106 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
19108 Flags to local 3x3 coordinates region centered on @code{x}:
19117 @subsection Example
19121 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.
19123 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19127 @section colorkey_opencl
19128 RGB colorspace color keying.
19130 The filter accepts the following options:
19134 The color which will be replaced with transparency.
19137 Similarity percentage with the key color.
19139 0.01 matches only the exact key color, while 1.0 matches everything.
19144 0.0 makes pixels either fully transparent, or not transparent at all.
19146 Higher values result in semi-transparent pixels, with a higher transparency
19147 the more similar the pixels color is to the key color.
19150 @subsection Examples
19154 Make every semi-green pixel in the input transparent with some slight blending:
19156 -i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
19160 @section overlay_opencl
19162 Overlay one video on top of another.
19164 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19165 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19167 The filter accepts the following options:
19172 Set the x coordinate of the overlaid video on the main video.
19173 Default value is @code{0}.
19176 Set the x coordinate of the overlaid video on the main video.
19177 Default value is @code{0}.
19181 @subsection Examples
19185 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19187 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19190 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19192 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19197 @section prewitt_opencl
19199 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19201 The filter accepts the following option:
19205 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19208 Set value which will be multiplied with filtered result.
19209 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19212 Set value which will be added to filtered result.
19213 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19216 @subsection Example
19220 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19222 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19226 @section roberts_opencl
19227 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19229 The filter accepts the following option:
19233 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19236 Set value which will be multiplied with filtered result.
19237 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19240 Set value which will be added to filtered result.
19241 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19244 @subsection Example
19248 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19250 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19254 @section sobel_opencl
19256 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19258 The filter accepts the following option:
19262 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19265 Set value which will be multiplied with filtered result.
19266 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19269 Set value which will be added to filtered result.
19270 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19273 @subsection Example
19277 Apply sobel operator with scale set to 2 and delta set to 10
19279 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19283 @section tonemap_opencl
19285 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19287 It accepts the following parameters:
19291 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19294 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19297 Apply desaturation for highlights that exceed this level of brightness. The
19298 higher the parameter, the more color information will be preserved. This
19299 setting helps prevent unnaturally blown-out colors for super-highlights, by
19300 (smoothly) turning into white instead. This makes images feel more natural,
19301 at the cost of reducing information about out-of-range colors.
19303 The default value is 0.5, and the algorithm here is a little different from
19304 the cpu version tonemap currently. A setting of 0.0 disables this option.
19307 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19308 is used to detect whether the scene has changed or not. If the distance between
19309 the current frame average brightness and the current running average exceeds
19310 a threshold value, we would re-calculate scene average and peak brightness.
19311 The default value is 0.2.
19314 Specify the output pixel format.
19316 Currently supported formats are:
19323 Set the output color range.
19325 Possible values are:
19331 Default is same as input.
19334 Set the output color primaries.
19336 Possible values are:
19342 Default is same as input.
19345 Set the output transfer characteristics.
19347 Possible values are:
19356 Set the output colorspace matrix.
19358 Possible value are:
19364 Default is same as input.
19368 @subsection Example
19372 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19374 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19378 @section unsharp_opencl
19380 Sharpen or blur the input video.
19382 It accepts the following parameters:
19385 @item luma_msize_x, lx
19386 Set the luma matrix horizontal size.
19387 Range is @code{[1, 23]} and default value is @code{5}.
19389 @item luma_msize_y, ly
19390 Set the luma matrix vertical size.
19391 Range is @code{[1, 23]} and default value is @code{5}.
19393 @item luma_amount, la
19394 Set the luma effect strength.
19395 Range is @code{[-10, 10]} and default value is @code{1.0}.
19397 Negative values will blur the input video, while positive values will
19398 sharpen it, a value of zero will disable the effect.
19400 @item chroma_msize_x, cx
19401 Set the chroma matrix horizontal size.
19402 Range is @code{[1, 23]} and default value is @code{5}.
19404 @item chroma_msize_y, cy
19405 Set the chroma matrix vertical size.
19406 Range is @code{[1, 23]} and default value is @code{5}.
19408 @item chroma_amount, ca
19409 Set the chroma effect strength.
19410 Range is @code{[-10, 10]} and default value is @code{0.0}.
19412 Negative values will blur the input video, while positive values will
19413 sharpen it, a value of zero will disable the effect.
19417 All parameters are optional and default to the equivalent of the
19418 string '5:5:1.0:5:5:0.0'.
19420 @subsection Examples
19424 Apply strong luma sharpen effect:
19426 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19430 Apply a strong blur of both luma and chroma parameters:
19432 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19436 @c man end OPENCL VIDEO FILTERS
19438 @chapter Video Sources
19439 @c man begin VIDEO SOURCES
19441 Below is a description of the currently available video sources.
19445 Buffer video frames, and make them available to the filter chain.
19447 This source is mainly intended for a programmatic use, in particular
19448 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19450 It accepts the following parameters:
19455 Specify the size (width and height) of the buffered video frames. For the
19456 syntax of this option, check the
19457 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19460 The input video width.
19463 The input video height.
19466 A string representing the pixel format of the buffered video frames.
19467 It may be a number corresponding to a pixel format, or a pixel format
19471 Specify the timebase assumed by the timestamps of the buffered frames.
19474 Specify the frame rate expected for the video stream.
19476 @item pixel_aspect, sar
19477 The sample (pixel) aspect ratio of the input video.
19480 Specify the optional parameters to be used for the scale filter which
19481 is automatically inserted when an input change is detected in the
19482 input size or format.
19484 @item hw_frames_ctx
19485 When using a hardware pixel format, this should be a reference to an
19486 AVHWFramesContext describing input frames.
19491 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19494 will instruct the source to accept video frames with size 320x240 and
19495 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19496 square pixels (1:1 sample aspect ratio).
19497 Since the pixel format with name "yuv410p" corresponds to the number 6
19498 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19499 this example corresponds to:
19501 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19504 Alternatively, the options can be specified as a flat string, but this
19505 syntax is deprecated:
19507 @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}]
19511 Create a pattern generated by an elementary cellular automaton.
19513 The initial state of the cellular automaton can be defined through the
19514 @option{filename} and @option{pattern} options. If such options are
19515 not specified an initial state is created randomly.
19517 At each new frame a new row in the video is filled with the result of
19518 the cellular automaton next generation. The behavior when the whole
19519 frame is filled is defined by the @option{scroll} option.
19521 This source accepts the following options:
19525 Read the initial cellular automaton state, i.e. the starting row, from
19526 the specified file.
19527 In the file, each non-whitespace character is considered an alive
19528 cell, a newline will terminate the row, and further characters in the
19529 file will be ignored.
19532 Read the initial cellular automaton state, i.e. the starting row, from
19533 the specified string.
19535 Each non-whitespace character in the string is considered an alive
19536 cell, a newline will terminate the row, and further characters in the
19537 string will be ignored.
19540 Set the video rate, that is the number of frames generated per second.
19543 @item random_fill_ratio, ratio
19544 Set the random fill ratio for the initial cellular automaton row. It
19545 is a floating point number value ranging from 0 to 1, defaults to
19548 This option is ignored when a file or a pattern is specified.
19550 @item random_seed, seed
19551 Set the seed for filling randomly the initial row, must be an integer
19552 included between 0 and UINT32_MAX. If not specified, or if explicitly
19553 set to -1, the filter will try to use a good random seed on a best
19557 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19558 Default value is 110.
19561 Set the size of the output video. For the syntax of this option, check the
19562 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19564 If @option{filename} or @option{pattern} is specified, the size is set
19565 by default to the width of the specified initial state row, and the
19566 height is set to @var{width} * PHI.
19568 If @option{size} is set, it must contain the width of the specified
19569 pattern string, and the specified pattern will be centered in the
19572 If a filename or a pattern string is not specified, the size value
19573 defaults to "320x518" (used for a randomly generated initial state).
19576 If set to 1, scroll the output upward when all the rows in the output
19577 have been already filled. If set to 0, the new generated row will be
19578 written over the top row just after the bottom row is filled.
19581 @item start_full, full
19582 If set to 1, completely fill the output with generated rows before
19583 outputting the first frame.
19584 This is the default behavior, for disabling set the value to 0.
19587 If set to 1, stitch the left and right row edges together.
19588 This is the default behavior, for disabling set the value to 0.
19591 @subsection Examples
19595 Read the initial state from @file{pattern}, and specify an output of
19598 cellauto=f=pattern:s=200x400
19602 Generate a random initial row with a width of 200 cells, with a fill
19605 cellauto=ratio=2/3:s=200x200
19609 Create a pattern generated by rule 18 starting by a single alive cell
19610 centered on an initial row with width 100:
19612 cellauto=p=@@:s=100x400:full=0:rule=18
19616 Specify a more elaborated initial pattern:
19618 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19623 @anchor{coreimagesrc}
19624 @section coreimagesrc
19625 Video source generated on GPU using Apple's CoreImage API on OSX.
19627 This video source is a specialized version of the @ref{coreimage} video filter.
19628 Use a core image generator at the beginning of the applied filterchain to
19629 generate the content.
19631 The coreimagesrc video source accepts the following options:
19633 @item list_generators
19634 List all available generators along with all their respective options as well as
19635 possible minimum and maximum values along with the default values.
19637 list_generators=true
19641 Specify the size of the sourced video. For the syntax of this option, check the
19642 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19643 The default value is @code{320x240}.
19646 Specify the frame rate of the sourced video, as the number of frames
19647 generated per second. It has to be a string in the format
19648 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19649 number or a valid video frame rate abbreviation. The default value is
19653 Set the sample aspect ratio of the sourced video.
19656 Set the duration of the sourced video. See
19657 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19658 for the accepted syntax.
19660 If not specified, or the expressed duration is negative, the video is
19661 supposed to be generated forever.
19664 Additionally, all options of the @ref{coreimage} video filter are accepted.
19665 A complete filterchain can be used for further processing of the
19666 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19667 and examples for details.
19669 @subsection Examples
19674 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19675 given as complete and escaped command-line for Apple's standard bash shell:
19677 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19679 This example is equivalent to the QRCode example of @ref{coreimage} without the
19680 need for a nullsrc video source.
19684 @section mandelbrot
19686 Generate a Mandelbrot set fractal, and progressively zoom towards the
19687 point specified with @var{start_x} and @var{start_y}.
19689 This source accepts the following options:
19694 Set the terminal pts value. Default value is 400.
19697 Set the terminal scale value.
19698 Must be a floating point value. Default value is 0.3.
19701 Set the inner coloring mode, that is the algorithm used to draw the
19702 Mandelbrot fractal internal region.
19704 It shall assume one of the following values:
19709 Show time until convergence.
19711 Set color based on point closest to the origin of the iterations.
19716 Default value is @var{mincol}.
19719 Set the bailout value. Default value is 10.0.
19722 Set the maximum of iterations performed by the rendering
19723 algorithm. Default value is 7189.
19726 Set outer coloring mode.
19727 It shall assume one of following values:
19729 @item iteration_count
19730 Set iteration count mode.
19731 @item normalized_iteration_count
19732 set normalized iteration count mode.
19734 Default value is @var{normalized_iteration_count}.
19737 Set frame rate, expressed as number of frames per second. Default
19741 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19742 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19745 Set the initial scale value. Default value is 3.0.
19748 Set the initial x position. Must be a floating point value between
19749 -100 and 100. Default value is -0.743643887037158704752191506114774.
19752 Set the initial y position. Must be a floating point value between
19753 -100 and 100. Default value is -0.131825904205311970493132056385139.
19758 Generate various test patterns, as generated by the MPlayer test filter.
19760 The size of the generated video is fixed, and is 256x256.
19761 This source is useful in particular for testing encoding features.
19763 This source accepts the following options:
19768 Specify the frame rate of the sourced video, as the number of frames
19769 generated per second. It has to be a string in the format
19770 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19771 number or a valid video frame rate abbreviation. The default value is
19775 Set the duration of the sourced video. See
19776 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19777 for the accepted syntax.
19779 If not specified, or the expressed duration is negative, the video is
19780 supposed to be generated forever.
19784 Set the number or the name of the test to perform. Supported tests are:
19800 Default value is "all", which will cycle through the list of all tests.
19805 mptestsrc=t=dc_luma
19808 will generate a "dc_luma" test pattern.
19810 @section frei0r_src
19812 Provide a frei0r source.
19814 To enable compilation of this filter you need to install the frei0r
19815 header and configure FFmpeg with @code{--enable-frei0r}.
19817 This source accepts the following parameters:
19822 The size of the video to generate. For the syntax of this option, check the
19823 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19826 The framerate of the generated video. It may be a string of the form
19827 @var{num}/@var{den} or a frame rate abbreviation.
19830 The name to the frei0r source to load. For more information regarding frei0r and
19831 how to set the parameters, read the @ref{frei0r} section in the video filters
19834 @item filter_params
19835 A '|'-separated list of parameters to pass to the frei0r source.
19839 For example, to generate a frei0r partik0l source with size 200x200
19840 and frame rate 10 which is overlaid on the overlay filter main input:
19842 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19847 Generate a life pattern.
19849 This source is based on a generalization of John Conway's life game.
19851 The sourced input represents a life grid, each pixel represents a cell
19852 which can be in one of two possible states, alive or dead. Every cell
19853 interacts with its eight neighbours, which are the cells that are
19854 horizontally, vertically, or diagonally adjacent.
19856 At each interaction the grid evolves according to the adopted rule,
19857 which specifies the number of neighbor alive cells which will make a
19858 cell stay alive or born. The @option{rule} option allows one to specify
19861 This source accepts the following options:
19865 Set the file from which to read the initial grid state. In the file,
19866 each non-whitespace character is considered an alive cell, and newline
19867 is used to delimit the end of each row.
19869 If this option is not specified, the initial grid is generated
19873 Set the video rate, that is the number of frames generated per second.
19876 @item random_fill_ratio, ratio
19877 Set the random fill ratio for the initial random grid. It is a
19878 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19879 It is ignored when a file is specified.
19881 @item random_seed, seed
19882 Set the seed for filling the initial random grid, must be an integer
19883 included between 0 and UINT32_MAX. If not specified, or if explicitly
19884 set to -1, the filter will try to use a good random seed on a best
19890 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19891 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19892 @var{NS} specifies the number of alive neighbor cells which make a
19893 live cell stay alive, and @var{NB} the number of alive neighbor cells
19894 which make a dead cell to become alive (i.e. to "born").
19895 "s" and "b" can be used in place of "S" and "B", respectively.
19897 Alternatively a rule can be specified by an 18-bits integer. The 9
19898 high order bits are used to encode the next cell state if it is alive
19899 for each number of neighbor alive cells, the low order bits specify
19900 the rule for "borning" new cells. Higher order bits encode for an
19901 higher number of neighbor cells.
19902 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19903 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19905 Default value is "S23/B3", which is the original Conway's game of life
19906 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19907 cells, and will born a new cell if there are three alive cells around
19911 Set the size of the output video. For the syntax of this option, check the
19912 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19914 If @option{filename} is specified, the size is set by default to the
19915 same size of the input file. If @option{size} is set, it must contain
19916 the size specified in the input file, and the initial grid defined in
19917 that file is centered in the larger resulting area.
19919 If a filename is not specified, the size value defaults to "320x240"
19920 (used for a randomly generated initial grid).
19923 If set to 1, stitch the left and right grid edges together, and the
19924 top and bottom edges also. Defaults to 1.
19927 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
19928 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
19929 value from 0 to 255.
19932 Set the color of living (or new born) cells.
19935 Set the color of dead cells. If @option{mold} is set, this is the first color
19936 used to represent a dead cell.
19939 Set mold color, for definitely dead and moldy cells.
19941 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
19942 ffmpeg-utils manual,ffmpeg-utils}.
19945 @subsection Examples
19949 Read a grid from @file{pattern}, and center it on a grid of size
19952 life=f=pattern:s=300x300
19956 Generate a random grid of size 200x200, with a fill ratio of 2/3:
19958 life=ratio=2/3:s=200x200
19962 Specify a custom rule for evolving a randomly generated grid:
19968 Full example with slow death effect (mold) using @command{ffplay}:
19970 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
19977 @anchor{haldclutsrc}
19980 @anchor{pal100bars}
19981 @anchor{rgbtestsrc}
19983 @anchor{smptehdbars}
19986 @anchor{yuvtestsrc}
19987 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
19989 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
19991 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
19993 The @code{color} source provides an uniformly colored input.
19995 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
19996 @ref{haldclut} filter.
19998 The @code{nullsrc} source returns unprocessed video frames. It is
19999 mainly useful to be employed in analysis / debugging tools, or as the
20000 source for filters which ignore the input data.
20002 The @code{pal75bars} source generates a color bars pattern, based on
20003 EBU PAL recommendations with 75% color levels.
20005 The @code{pal100bars} source generates a color bars pattern, based on
20006 EBU PAL recommendations with 100% color levels.
20008 The @code{rgbtestsrc} source generates an RGB test pattern useful for
20009 detecting RGB vs BGR issues. You should see a red, green and blue
20010 stripe from top to bottom.
20012 The @code{smptebars} source generates a color bars pattern, based on
20013 the SMPTE Engineering Guideline EG 1-1990.
20015 The @code{smptehdbars} source generates a color bars pattern, based on
20016 the SMPTE RP 219-2002.
20018 The @code{testsrc} source generates a test video pattern, showing a
20019 color pattern, a scrolling gradient and a timestamp. This is mainly
20020 intended for testing purposes.
20022 The @code{testsrc2} source is similar to testsrc, but supports more
20023 pixel formats instead of just @code{rgb24}. This allows using it as an
20024 input for other tests without requiring a format conversion.
20026 The @code{yuvtestsrc} source generates an YUV test pattern. You should
20027 see a y, cb and cr stripe from top to bottom.
20029 The sources accept the following parameters:
20034 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
20035 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
20036 pixels to be used as identity matrix for 3D lookup tables. Each component is
20037 coded on a @code{1/(N*N)} scale.
20040 Specify the color of the source, only available in the @code{color}
20041 source. For the syntax of this option, check the
20042 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
20045 Specify the size of the sourced video. For the syntax of this option, check the
20046 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20047 The default value is @code{320x240}.
20049 This option is not available with the @code{allrgb}, @code{allyuv}, and
20050 @code{haldclutsrc} filters.
20053 Specify the frame rate of the sourced video, as the number of frames
20054 generated per second. It has to be a string in the format
20055 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
20056 number or a valid video frame rate abbreviation. The default value is
20060 Set the duration of the sourced video. See
20061 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
20062 for the accepted syntax.
20064 If not specified, or the expressed duration is negative, the video is
20065 supposed to be generated forever.
20068 Set the sample aspect ratio of the sourced video.
20071 Specify the alpha (opacity) of the background, only available in the
20072 @code{testsrc2} source. The value must be between 0 (fully transparent) and
20073 255 (fully opaque, the default).
20076 Set the number of decimals to show in the timestamp, only available in the
20077 @code{testsrc} source.
20079 The displayed timestamp value will correspond to the original
20080 timestamp value multiplied by the power of 10 of the specified
20081 value. Default value is 0.
20084 @subsection Examples
20088 Generate a video with a duration of 5.3 seconds, with size
20089 176x144 and a frame rate of 10 frames per second:
20091 testsrc=duration=5.3:size=qcif:rate=10
20095 The following graph description will generate a red source
20096 with an opacity of 0.2, with size "qcif" and a frame rate of 10
20099 color=c=red@@0.2:s=qcif:r=10
20103 If the input content is to be ignored, @code{nullsrc} can be used. The
20104 following command generates noise in the luminance plane by employing
20105 the @code{geq} filter:
20107 nullsrc=s=256x256, geq=random(1)*255:128:128
20111 @subsection Commands
20113 The @code{color} source supports the following commands:
20117 Set the color of the created image. Accepts the same syntax of the
20118 corresponding @option{color} option.
20123 Generate video using an OpenCL program.
20128 OpenCL program source file.
20131 Kernel name in program.
20134 Size of frames to generate. This must be set.
20137 Pixel format to use for the generated frames. This must be set.
20140 Number of frames generated every second. Default value is '25'.
20144 For details of how the program loading works, see the @ref{program_opencl}
20151 Generate a colour ramp by setting pixel values from the position of the pixel
20152 in the output image. (Note that this will work with all pixel formats, but
20153 the generated output will not be the same.)
20155 __kernel void ramp(__write_only image2d_t dst,
20156 unsigned int index)
20158 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20161 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20163 write_imagef(dst, loc, val);
20168 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20170 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20171 unsigned int index)
20173 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20175 float4 value = 0.0f;
20176 int x = loc.x + index;
20177 int y = loc.y + index;
20178 while (x > 0 || y > 0) {
20179 if (x % 3 == 1 && y % 3 == 1) {
20187 write_imagef(dst, loc, value);
20193 @c man end VIDEO SOURCES
20195 @chapter Video Sinks
20196 @c man begin VIDEO SINKS
20198 Below is a description of the currently available video sinks.
20200 @section buffersink
20202 Buffer video frames, and make them available to the end of the filter
20205 This sink is mainly intended for programmatic use, in particular
20206 through the interface defined in @file{libavfilter/buffersink.h}
20207 or the options system.
20209 It accepts a pointer to an AVBufferSinkContext structure, which
20210 defines the incoming buffers' formats, to be passed as the opaque
20211 parameter to @code{avfilter_init_filter} for initialization.
20215 Null video sink: do absolutely nothing with the input video. It is
20216 mainly useful as a template and for use in analysis / debugging
20219 @c man end VIDEO SINKS
20221 @chapter Multimedia Filters
20222 @c man begin MULTIMEDIA FILTERS
20224 Below is a description of the currently available multimedia filters.
20228 Convert input audio to a video output, displaying the audio bit scope.
20230 The filter accepts the following options:
20234 Set frame rate, expressed as number of frames per second. Default
20238 Specify the video size for the output. For the syntax of this option, check the
20239 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20240 Default value is @code{1024x256}.
20243 Specify list of colors separated by space or by '|' which will be used to
20244 draw channels. Unrecognized or missing colors will be replaced
20248 @section ahistogram
20250 Convert input audio to a video output, displaying the volume histogram.
20252 The filter accepts the following options:
20256 Specify how histogram is calculated.
20258 It accepts the following values:
20261 Use single histogram for all channels.
20263 Use separate histogram for each channel.
20265 Default is @code{single}.
20268 Set frame rate, expressed as number of frames per second. Default
20272 Specify the video size for the output. For the syntax of this option, check the
20273 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20274 Default value is @code{hd720}.
20279 It accepts the following values:
20290 reverse logarithmic
20292 Default is @code{log}.
20295 Set amplitude scale.
20297 It accepts the following values:
20304 Default is @code{log}.
20307 Set how much frames to accumulate in histogram.
20308 Default is 1. Setting this to -1 accumulates all frames.
20311 Set histogram ratio of window height.
20314 Set sonogram sliding.
20316 It accepts the following values:
20319 replace old rows with new ones.
20321 scroll from top to bottom.
20323 Default is @code{replace}.
20326 @section aphasemeter
20328 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20329 representing mean phase of current audio frame. A video output can also be produced and is
20330 enabled by default. The audio is passed through as first output.
20332 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20333 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20334 and @code{1} means channels are in phase.
20336 The filter accepts the following options, all related to its video output:
20340 Set the output frame rate. Default value is @code{25}.
20343 Set the video size for the output. For the syntax of this option, check the
20344 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20345 Default value is @code{800x400}.
20350 Specify the red, green, blue contrast. Default values are @code{2},
20351 @code{7} and @code{1}.
20352 Allowed range is @code{[0, 255]}.
20355 Set color which will be used for drawing median phase. If color is
20356 @code{none} which is default, no median phase value will be drawn.
20359 Enable video output. Default is enabled.
20362 @section avectorscope
20364 Convert input audio to a video output, representing the audio vector
20367 The filter is used to measure the difference between channels of stereo
20368 audio stream. A monoaural signal, consisting of identical left and right
20369 signal, results in straight vertical line. Any stereo separation is visible
20370 as a deviation from this line, creating a Lissajous figure.
20371 If the straight (or deviation from it) but horizontal line appears this
20372 indicates that the left and right channels are out of phase.
20374 The filter accepts the following options:
20378 Set the vectorscope mode.
20380 Available values are:
20383 Lissajous rotated by 45 degrees.
20386 Same as above but not rotated.
20389 Shape resembling half of circle.
20392 Default value is @samp{lissajous}.
20395 Set the video size for the output. For the syntax of this option, check the
20396 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20397 Default value is @code{400x400}.
20400 Set the output frame rate. Default value is @code{25}.
20406 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20407 @code{160}, @code{80} and @code{255}.
20408 Allowed range is @code{[0, 255]}.
20414 Specify the red, green, blue and alpha fade. Default values are @code{15},
20415 @code{10}, @code{5} and @code{5}.
20416 Allowed range is @code{[0, 255]}.
20419 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20420 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20423 Set the vectorscope drawing mode.
20425 Available values are:
20428 Draw dot for each sample.
20431 Draw line between previous and current sample.
20434 Default value is @samp{dot}.
20437 Specify amplitude scale of audio samples.
20439 Available values are:
20455 Swap left channel axis with right channel axis.
20465 Mirror only x axis.
20468 Mirror only y axis.
20476 @subsection Examples
20480 Complete example using @command{ffplay}:
20482 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20483 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20487 @section bench, abench
20489 Benchmark part of a filtergraph.
20491 The filter accepts the following options:
20495 Start or stop a timer.
20497 Available values are:
20500 Get the current time, set it as frame metadata (using the key
20501 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20504 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20505 the input frame metadata to get the time difference. Time difference, average,
20506 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20507 @code{min}) are then printed. The timestamps are expressed in seconds.
20511 @subsection Examples
20515 Benchmark @ref{selectivecolor} filter:
20517 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20523 Concatenate audio and video streams, joining them together one after the
20526 The filter works on segments of synchronized video and audio streams. All
20527 segments must have the same number of streams of each type, and that will
20528 also be the number of streams at output.
20530 The filter accepts the following options:
20535 Set the number of segments. Default is 2.
20538 Set the number of output video streams, that is also the number of video
20539 streams in each segment. Default is 1.
20542 Set the number of output audio streams, that is also the number of audio
20543 streams in each segment. Default is 0.
20546 Activate unsafe mode: do not fail if segments have a different format.
20550 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20551 @var{a} audio outputs.
20553 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20554 segment, in the same order as the outputs, then the inputs for the second
20557 Related streams do not always have exactly the same duration, for various
20558 reasons including codec frame size or sloppy authoring. For that reason,
20559 related synchronized streams (e.g. a video and its audio track) should be
20560 concatenated at once. The concat filter will use the duration of the longest
20561 stream in each segment (except the last one), and if necessary pad shorter
20562 audio streams with silence.
20564 For this filter to work correctly, all segments must start at timestamp 0.
20566 All corresponding streams must have the same parameters in all segments; the
20567 filtering system will automatically select a common pixel format for video
20568 streams, and a common sample format, sample rate and channel layout for
20569 audio streams, but other settings, such as resolution, must be converted
20570 explicitly by the user.
20572 Different frame rates are acceptable but will result in variable frame rate
20573 at output; be sure to configure the output file to handle it.
20575 @subsection Examples
20579 Concatenate an opening, an episode and an ending, all in bilingual version
20580 (video in stream 0, audio in streams 1 and 2):
20582 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20583 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20584 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20585 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20589 Concatenate two parts, handling audio and video separately, using the
20590 (a)movie sources, and adjusting the resolution:
20592 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20593 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20594 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20596 Note that a desync will happen at the stitch if the audio and video streams
20597 do not have exactly the same duration in the first file.
20601 @subsection Commands
20603 This filter supports the following commands:
20606 Close the current segment and step to the next one
20609 @section drawgraph, adrawgraph
20611 Draw a graph using input video or audio metadata.
20613 It accepts the following parameters:
20617 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20620 Set 1st foreground color expression.
20623 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20626 Set 2nd foreground color expression.
20629 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20632 Set 3rd foreground color expression.
20635 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20638 Set 4th foreground color expression.
20641 Set minimal value of metadata value.
20644 Set maximal value of metadata value.
20647 Set graph background color. Default is white.
20652 Available values for mode is:
20659 Default is @code{line}.
20664 Available values for slide is:
20667 Draw new frame when right border is reached.
20670 Replace old columns with new ones.
20673 Scroll from right to left.
20676 Scroll from left to right.
20679 Draw single picture.
20682 Default is @code{frame}.
20685 Set size of graph video. For the syntax of this option, check the
20686 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20687 The default value is @code{900x256}.
20689 The foreground color expressions can use the following variables:
20692 Minimal value of metadata value.
20695 Maximal value of metadata value.
20698 Current metadata key value.
20701 The color is defined as 0xAABBGGRR.
20704 Example using metadata from @ref{signalstats} filter:
20706 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20709 Example using metadata from @ref{ebur128} filter:
20711 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20717 EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness
20718 level. By default, it logs a message at a frequency of 10Hz with the
20719 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20720 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20722 The filter can only analyze streams which have a sampling rate of 48000 Hz and whose
20723 sample format is double-precision floating point. The input stream will be converted to
20724 this specification, if needed. Users may need to insert aformat and/or aresample filters
20725 after this filter to obtain the original parameters.
20727 The filter also has a video output (see the @var{video} option) with a real
20728 time graph to observe the loudness evolution. The graphic contains the logged
20729 message mentioned above, so it is not printed anymore when this option is set,
20730 unless the verbose logging is set. The main graphing area contains the
20731 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20732 the momentary loudness (400 milliseconds), but can optionally be configured
20733 to instead display short-term loudness (see @var{gauge}).
20735 The green area marks a +/- 1LU target range around the target loudness
20736 (-23LUFS by default, unless modified through @var{target}).
20738 More information about the Loudness Recommendation EBU R128 on
20739 @url{http://tech.ebu.ch/loudness}.
20741 The filter accepts the following options:
20746 Activate the video output. The audio stream is passed unchanged whether this
20747 option is set or no. The video stream will be the first output stream if
20748 activated. Default is @code{0}.
20751 Set the video size. This option is for video only. For the syntax of this
20753 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20754 Default and minimum resolution is @code{640x480}.
20757 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20758 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20759 other integer value between this range is allowed.
20762 Set metadata injection. If set to @code{1}, the audio input will be segmented
20763 into 100ms output frames, each of them containing various loudness information
20764 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20766 Default is @code{0}.
20769 Force the frame logging level.
20771 Available values are:
20774 information logging level
20776 verbose logging level
20779 By default, the logging level is set to @var{info}. If the @option{video} or
20780 the @option{metadata} options are set, it switches to @var{verbose}.
20785 Available modes can be cumulated (the option is a @code{flag} type). Possible
20789 Disable any peak mode (default).
20791 Enable sample-peak mode.
20793 Simple peak mode looking for the higher sample value. It logs a message
20794 for sample-peak (identified by @code{SPK}).
20796 Enable true-peak mode.
20798 If enabled, the peak lookup is done on an over-sampled version of the input
20799 stream for better peak accuracy. It logs a message for true-peak.
20800 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20801 This mode requires a build with @code{libswresample}.
20805 Treat mono input files as "dual mono". If a mono file is intended for playback
20806 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20807 If set to @code{true}, this option will compensate for this effect.
20808 Multi-channel input files are not affected by this option.
20811 Set a specific pan law to be used for the measurement of dual mono files.
20812 This parameter is optional, and has a default value of -3.01dB.
20815 Set a specific target level (in LUFS) used as relative zero in the visualization.
20816 This parameter is optional and has a default value of -23LUFS as specified
20817 by EBU R128. However, material published online may prefer a level of -16LUFS
20818 (e.g. for use with podcasts or video platforms).
20821 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20822 @code{shortterm}. By default the momentary value will be used, but in certain
20823 scenarios it may be more useful to observe the short term value instead (e.g.
20827 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20828 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20829 video output, not the summary or continuous log output.
20832 @subsection Examples
20836 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20838 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20842 Run an analysis with @command{ffmpeg}:
20844 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20848 @section interleave, ainterleave
20850 Temporally interleave frames from several inputs.
20852 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20854 These filters read frames from several inputs and send the oldest
20855 queued frame to the output.
20857 Input streams must have well defined, monotonically increasing frame
20860 In order to submit one frame to output, these filters need to enqueue
20861 at least one frame for each input, so they cannot work in case one
20862 input is not yet terminated and will not receive incoming frames.
20864 For example consider the case when one input is a @code{select} filter
20865 which always drops input frames. The @code{interleave} filter will keep
20866 reading from that input, but it will never be able to send new frames
20867 to output until the input sends an end-of-stream signal.
20869 Also, depending on inputs synchronization, the filters will drop
20870 frames in case one input receives more frames than the other ones, and
20871 the queue is already filled.
20873 These filters accept the following options:
20877 Set the number of different inputs, it is 2 by default.
20880 @subsection Examples
20884 Interleave frames belonging to different streams using @command{ffmpeg}:
20886 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20890 Add flickering blur effect:
20892 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20896 @section metadata, ametadata
20898 Manipulate frame metadata.
20900 This filter accepts the following options:
20904 Set mode of operation of the filter.
20906 Can be one of the following:
20910 If both @code{value} and @code{key} is set, select frames
20911 which have such metadata. If only @code{key} is set, select
20912 every frame that has such key in metadata.
20915 Add new metadata @code{key} and @code{value}. If key is already available
20919 Modify value of already present key.
20922 If @code{value} is set, delete only keys that have such value.
20923 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
20927 Print key and its value if metadata was found. If @code{key} is not set print all
20928 metadata values available in frame.
20932 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
20935 Set metadata value which will be used. This option is mandatory for
20936 @code{modify} and @code{add} mode.
20939 Which function to use when comparing metadata value and @code{value}.
20941 Can be one of following:
20945 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
20948 Values are interpreted as strings, returns true if metadata value starts with
20949 the @code{value} option string.
20952 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
20955 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
20958 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
20961 Values are interpreted as floats, returns true if expression from option @code{expr}
20966 Set expression which is used when @code{function} is set to @code{expr}.
20967 The expression is evaluated through the eval API and can contain the following
20972 Float representation of @code{value} from metadata key.
20975 Float representation of @code{value} as supplied by user in @code{value} option.
20979 If specified in @code{print} mode, output is written to the named file. Instead of
20980 plain filename any writable url can be specified. Filename ``-'' is a shorthand
20981 for standard output. If @code{file} option is not set, output is written to the log
20982 with AV_LOG_INFO loglevel.
20986 @subsection Examples
20990 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
20993 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
20996 Print silencedetect output to file @file{metadata.txt}.
20998 silencedetect,ametadata=mode=print:file=metadata.txt
21001 Direct all metadata to a pipe with file descriptor 4.
21003 metadata=mode=print:file='pipe\:4'
21007 @section perms, aperms
21009 Set read/write permissions for the output frames.
21011 These filters are mainly aimed at developers to test direct path in the
21012 following filter in the filtergraph.
21014 The filters accept the following options:
21018 Select the permissions mode.
21020 It accepts the following values:
21023 Do nothing. This is the default.
21025 Set all the output frames read-only.
21027 Set all the output frames directly writable.
21029 Make the frame read-only if writable, and writable if read-only.
21031 Set each output frame read-only or writable randomly.
21035 Set the seed for the @var{random} mode, must be an integer included between
21036 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
21037 @code{-1}, the filter will try to use a good random seed on a best effort
21041 Note: in case of auto-inserted filter between the permission filter and the
21042 following one, the permission might not be received as expected in that
21043 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
21044 perms/aperms filter can avoid this problem.
21046 @section realtime, arealtime
21048 Slow down filtering to match real time approximately.
21050 These filters will pause the filtering for a variable amount of time to
21051 match the output rate with the input timestamps.
21052 They are similar to the @option{re} option to @code{ffmpeg}.
21054 They accept the following options:
21058 Time limit for the pauses. Any pause longer than that will be considered
21059 a timestamp discontinuity and reset the timer. Default is 2 seconds.
21063 @section select, aselect
21065 Select frames to pass in output.
21067 This filter accepts the following options:
21072 Set expression, which is evaluated for each input frame.
21074 If the expression is evaluated to zero, the frame is discarded.
21076 If the evaluation result is negative or NaN, the frame is sent to the
21077 first output; otherwise it is sent to the output with index
21078 @code{ceil(val)-1}, assuming that the input index starts from 0.
21080 For example a value of @code{1.2} corresponds to the output with index
21081 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
21084 Set the number of outputs. The output to which to send the selected
21085 frame is based on the result of the evaluation. Default value is 1.
21088 The expression can contain the following constants:
21092 The (sequential) number of the filtered frame, starting from 0.
21095 The (sequential) number of the selected frame, starting from 0.
21097 @item prev_selected_n
21098 The sequential number of the last selected frame. It's NAN if undefined.
21101 The timebase of the input timestamps.
21104 The PTS (Presentation TimeStamp) of the filtered video frame,
21105 expressed in @var{TB} units. It's NAN if undefined.
21108 The PTS of the filtered video frame,
21109 expressed in seconds. It's NAN if undefined.
21112 The PTS of the previously filtered video frame. It's NAN if undefined.
21114 @item prev_selected_pts
21115 The PTS of the last previously filtered video frame. It's NAN if undefined.
21117 @item prev_selected_t
21118 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
21121 The PTS of the first video frame in the video. It's NAN if undefined.
21124 The time of the first video frame in the video. It's NAN if undefined.
21126 @item pict_type @emph{(video only)}
21127 The type of the filtered frame. It can assume one of the following
21139 @item interlace_type @emph{(video only)}
21140 The frame interlace type. It can assume one of the following values:
21143 The frame is progressive (not interlaced).
21145 The frame is top-field-first.
21147 The frame is bottom-field-first.
21150 @item consumed_sample_n @emph{(audio only)}
21151 the number of selected samples before the current frame
21153 @item samples_n @emph{(audio only)}
21154 the number of samples in the current frame
21156 @item sample_rate @emph{(audio only)}
21157 the input sample rate
21160 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21163 the position in the file of the filtered frame, -1 if the information
21164 is not available (e.g. for synthetic video)
21166 @item scene @emph{(video only)}
21167 value between 0 and 1 to indicate a new scene; a low value reflects a low
21168 probability for the current frame to introduce a new scene, while a higher
21169 value means the current frame is more likely to be one (see the example below)
21171 @item concatdec_select
21172 The concat demuxer can select only part of a concat input file by setting an
21173 inpoint and an outpoint, but the output packets may not be entirely contained
21174 in the selected interval. By using this variable, it is possible to skip frames
21175 generated by the concat demuxer which are not exactly contained in the selected
21178 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21179 and the @var{lavf.concat.duration} packet metadata values which are also
21180 present in the decoded frames.
21182 The @var{concatdec_select} variable is -1 if the frame pts is at least
21183 start_time and either the duration metadata is missing or the frame pts is less
21184 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21187 That basically means that an input frame is selected if its pts is within the
21188 interval set by the concat demuxer.
21192 The default value of the select expression is "1".
21194 @subsection Examples
21198 Select all frames in input:
21203 The example above is the same as:
21215 Select only I-frames:
21217 select='eq(pict_type\,I)'
21221 Select one frame every 100:
21223 select='not(mod(n\,100))'
21227 Select only frames contained in the 10-20 time interval:
21229 select=between(t\,10\,20)
21233 Select only I-frames contained in the 10-20 time interval:
21235 select=between(t\,10\,20)*eq(pict_type\,I)
21239 Select frames with a minimum distance of 10 seconds:
21241 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21245 Use aselect to select only audio frames with samples number > 100:
21247 aselect='gt(samples_n\,100)'
21251 Create a mosaic of the first scenes:
21253 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21256 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21260 Send even and odd frames to separate outputs, and compose them:
21262 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21266 Select useful frames from an ffconcat file which is using inpoints and
21267 outpoints but where the source files are not intra frame only.
21269 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21273 @section sendcmd, asendcmd
21275 Send commands to filters in the filtergraph.
21277 These filters read commands to be sent to other filters in the
21280 @code{sendcmd} must be inserted between two video filters,
21281 @code{asendcmd} must be inserted between two audio filters, but apart
21282 from that they act the same way.
21284 The specification of commands can be provided in the filter arguments
21285 with the @var{commands} option, or in a file specified by the
21286 @var{filename} option.
21288 These filters accept the following options:
21291 Set the commands to be read and sent to the other filters.
21293 Set the filename of the commands to be read and sent to the other
21297 @subsection Commands syntax
21299 A commands description consists of a sequence of interval
21300 specifications, comprising a list of commands to be executed when a
21301 particular event related to that interval occurs. The occurring event
21302 is typically the current frame time entering or leaving a given time
21305 An interval is specified by the following syntax:
21307 @var{START}[-@var{END}] @var{COMMANDS};
21310 The time interval is specified by the @var{START} and @var{END} times.
21311 @var{END} is optional and defaults to the maximum time.
21313 The current frame time is considered within the specified interval if
21314 it is included in the interval [@var{START}, @var{END}), that is when
21315 the time is greater or equal to @var{START} and is lesser than
21318 @var{COMMANDS} consists of a sequence of one or more command
21319 specifications, separated by ",", relating to that interval. The
21320 syntax of a command specification is given by:
21322 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21325 @var{FLAGS} is optional and specifies the type of events relating to
21326 the time interval which enable sending the specified command, and must
21327 be a non-null sequence of identifier flags separated by "+" or "|" and
21328 enclosed between "[" and "]".
21330 The following flags are recognized:
21333 The command is sent when the current frame timestamp enters the
21334 specified interval. In other words, the command is sent when the
21335 previous frame timestamp was not in the given interval, and the
21339 The command is sent when the current frame timestamp leaves the
21340 specified interval. In other words, the command is sent when the
21341 previous frame timestamp was in the given interval, and the
21345 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21348 @var{TARGET} specifies the target of the command, usually the name of
21349 the filter class or a specific filter instance name.
21351 @var{COMMAND} specifies the name of the command for the target filter.
21353 @var{ARG} is optional and specifies the optional list of argument for
21354 the given @var{COMMAND}.
21356 Between one interval specification and another, whitespaces, or
21357 sequences of characters starting with @code{#} until the end of line,
21358 are ignored and can be used to annotate comments.
21360 A simplified BNF description of the commands specification syntax
21363 @var{COMMAND_FLAG} ::= "enter" | "leave"
21364 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21365 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21366 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21367 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21368 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21371 @subsection Examples
21375 Specify audio tempo change at second 4:
21377 asendcmd=c='4.0 atempo tempo 1.5',atempo
21381 Target a specific filter instance:
21383 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21387 Specify a list of drawtext and hue commands in a file.
21389 # show text in the interval 5-10
21390 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21391 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21393 # desaturate the image in the interval 15-20
21394 15.0-20.0 [enter] hue s 0,
21395 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21397 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21399 # apply an exponential saturation fade-out effect, starting from time 25
21400 25 [enter] hue s exp(25-t)
21403 A filtergraph allowing to read and process the above command list
21404 stored in a file @file{test.cmd}, can be specified with:
21406 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21411 @section setpts, asetpts
21413 Change the PTS (presentation timestamp) of the input frames.
21415 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21417 This filter accepts the following options:
21422 The expression which is evaluated for each frame to construct its timestamp.
21426 The expression is evaluated through the eval API and can contain the following
21430 @item FRAME_RATE, FR
21431 frame rate, only defined for constant frame-rate video
21434 The presentation timestamp in input
21437 The count of the input frame for video or the number of consumed samples,
21438 not including the current frame for audio, starting from 0.
21440 @item NB_CONSUMED_SAMPLES
21441 The number of consumed samples, not including the current frame (only
21444 @item NB_SAMPLES, S
21445 The number of samples in the current frame (only audio)
21447 @item SAMPLE_RATE, SR
21448 The audio sample rate.
21451 The PTS of the first frame.
21454 the time in seconds of the first frame
21457 State whether the current frame is interlaced.
21460 the time in seconds of the current frame
21463 original position in the file of the frame, or undefined if undefined
21464 for the current frame
21467 The previous input PTS.
21470 previous input time in seconds
21473 The previous output PTS.
21476 previous output time in seconds
21479 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21483 The wallclock (RTC) time at the start of the movie in microseconds.
21486 The timebase of the input timestamps.
21490 @subsection Examples
21494 Start counting PTS from zero
21496 setpts=PTS-STARTPTS
21500 Apply fast motion effect:
21506 Apply slow motion effect:
21512 Set fixed rate of 25 frames per second:
21518 Set fixed rate 25 fps with some jitter:
21520 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21524 Apply an offset of 10 seconds to the input PTS:
21530 Generate timestamps from a "live source" and rebase onto the current timebase:
21532 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21536 Generate timestamps by counting samples:
21545 Force color range for the output video frame.
21547 The @code{setrange} filter marks the color range property for the
21548 output frames. It does not change the input frame, but only sets the
21549 corresponding property, which affects how the frame is treated by
21552 The filter accepts the following options:
21557 Available values are:
21561 Keep the same color range property.
21563 @item unspecified, unknown
21564 Set the color range as unspecified.
21566 @item limited, tv, mpeg
21567 Set the color range as limited.
21569 @item full, pc, jpeg
21570 Set the color range as full.
21574 @section settb, asettb
21576 Set the timebase to use for the output frames timestamps.
21577 It is mainly useful for testing timebase configuration.
21579 It accepts the following parameters:
21584 The expression which is evaluated into the output timebase.
21588 The value for @option{tb} is an arithmetic expression representing a
21589 rational. The expression can contain the constants "AVTB" (the default
21590 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21591 audio only). Default value is "intb".
21593 @subsection Examples
21597 Set the timebase to 1/25:
21603 Set the timebase to 1/10:
21609 Set the timebase to 1001/1000:
21615 Set the timebase to 2*intb:
21621 Set the default timebase value:
21628 Convert input audio to a video output representing frequency spectrum
21629 logarithmically using Brown-Puckette constant Q transform algorithm with
21630 direct frequency domain coefficient calculation (but the transform itself
21631 is not really constant Q, instead the Q factor is actually variable/clamped),
21632 with musical tone scale, from E0 to D#10.
21634 The filter accepts the following options:
21638 Specify the video size for the output. It must be even. For the syntax of this option,
21639 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21640 Default value is @code{1920x1080}.
21643 Set the output frame rate. Default value is @code{25}.
21646 Set the bargraph height. It must be even. Default value is @code{-1} which
21647 computes the bargraph height automatically.
21650 Set the axis height. It must be even. Default value is @code{-1} which computes
21651 the axis height automatically.
21654 Set the sonogram height. It must be even. Default value is @code{-1} which
21655 computes the sonogram height automatically.
21658 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21659 instead. Default value is @code{1}.
21661 @item sono_v, volume
21662 Specify the sonogram volume expression. It can contain variables:
21665 the @var{bar_v} evaluated expression
21666 @item frequency, freq, f
21667 the frequency where it is evaluated
21668 @item timeclamp, tc
21669 the value of @var{timeclamp} option
21673 @item a_weighting(f)
21674 A-weighting of equal loudness
21675 @item b_weighting(f)
21676 B-weighting of equal loudness
21677 @item c_weighting(f)
21678 C-weighting of equal loudness.
21680 Default value is @code{16}.
21682 @item bar_v, volume2
21683 Specify the bargraph volume expression. It can contain variables:
21686 the @var{sono_v} evaluated expression
21687 @item frequency, freq, f
21688 the frequency where it is evaluated
21689 @item timeclamp, tc
21690 the value of @var{timeclamp} option
21694 @item a_weighting(f)
21695 A-weighting of equal loudness
21696 @item b_weighting(f)
21697 B-weighting of equal loudness
21698 @item c_weighting(f)
21699 C-weighting of equal loudness.
21701 Default value is @code{sono_v}.
21703 @item sono_g, gamma
21704 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21705 higher gamma makes the spectrum having more range. Default value is @code{3}.
21706 Acceptable range is @code{[1, 7]}.
21708 @item bar_g, gamma2
21709 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21713 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21714 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21716 @item timeclamp, tc
21717 Specify the transform timeclamp. At low frequency, there is trade-off between
21718 accuracy in time domain and frequency domain. If timeclamp is lower,
21719 event in time domain is represented more accurately (such as fast bass drum),
21720 otherwise event in frequency domain is represented more accurately
21721 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21724 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21725 limits future samples by applying asymmetric windowing in time domain, useful
21726 when low latency is required. Accepted range is @code{[0, 1]}.
21729 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21730 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21733 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21734 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21737 This option is deprecated and ignored.
21740 Specify the transform length in time domain. Use this option to control accuracy
21741 trade-off between time domain and frequency domain at every frequency sample.
21742 It can contain variables:
21744 @item frequency, freq, f
21745 the frequency where it is evaluated
21746 @item timeclamp, tc
21747 the value of @var{timeclamp} option.
21749 Default value is @code{384*tc/(384+tc*f)}.
21752 Specify the transform count for every video frame. Default value is @code{6}.
21753 Acceptable range is @code{[1, 30]}.
21756 Specify the transform count for every single pixel. Default value is @code{0},
21757 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21760 Specify font file for use with freetype to draw the axis. If not specified,
21761 use embedded font. Note that drawing with font file or embedded font is not
21762 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21766 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21767 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21770 Specify font color expression. This is arithmetic expression that should return
21771 integer value 0xRRGGBB. It can contain variables:
21773 @item frequency, freq, f
21774 the frequency where it is evaluated
21775 @item timeclamp, tc
21776 the value of @var{timeclamp} option
21781 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21782 @item r(x), g(x), b(x)
21783 red, green, and blue value of intensity x.
21785 Default value is @code{st(0, (midi(f)-59.5)/12);
21786 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21787 r(1-ld(1)) + b(ld(1))}.
21790 Specify image file to draw the axis. This option override @var{fontfile} and
21791 @var{fontcolor} option.
21794 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21795 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21796 Default value is @code{1}.
21799 Set colorspace. The accepted values are:
21802 Unspecified (default)
21811 BT.470BG or BT.601-6 625
21814 SMPTE-170M or BT.601-6 525
21820 BT.2020 with non-constant luminance
21825 Set spectrogram color scheme. This is list of floating point values with format
21826 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21827 The default is @code{1|0.5|0|0|0.5|1}.
21831 @subsection Examples
21835 Playing audio while showing the spectrum:
21837 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21841 Same as above, but with frame rate 30 fps:
21843 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21847 Playing at 1280x720:
21849 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21853 Disable sonogram display:
21859 A1 and its harmonics: A1, A2, (near)E3, A3:
21861 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),
21862 asplit[a][out1]; [a] showcqt [out0]'
21866 Same as above, but with more accuracy in frequency domain:
21868 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),
21869 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21875 bar_v=10:sono_v=bar_v*a_weighting(f)
21879 Custom gamma, now spectrum is linear to the amplitude.
21885 Custom tlength equation:
21887 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)))'
21891 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21893 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21897 Custom font using fontconfig:
21899 font='Courier New,Monospace,mono|bold'
21903 Custom frequency range with custom axis using image file:
21905 axisfile=myaxis.png:basefreq=40:endfreq=10000
21911 Convert input audio to video output representing the audio power spectrum.
21912 Audio amplitude is on Y-axis while frequency is on X-axis.
21914 The filter accepts the following options:
21918 Specify size of video. For the syntax of this option, check the
21919 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21920 Default is @code{1024x512}.
21924 This set how each frequency bin will be represented.
21926 It accepts the following values:
21932 Default is @code{bar}.
21935 Set amplitude scale.
21937 It accepts the following values:
21951 Default is @code{log}.
21954 Set frequency scale.
21956 It accepts the following values:
21965 Reverse logarithmic scale.
21967 Default is @code{lin}.
21972 It accepts the following values:
21988 Default is @code{w2048}
21991 Set windowing function.
21993 It accepts the following values:
22016 Default is @code{hanning}.
22019 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22020 which means optimal overlap for selected window function will be picked.
22023 Set time averaging. Setting this to 0 will display current maximal peaks.
22024 Default is @code{1}, which means time averaging is disabled.
22027 Specify list of colors separated by space or by '|' which will be used to
22028 draw channel frequencies. Unrecognized or missing colors will be replaced
22032 Set channel display mode.
22034 It accepts the following values:
22039 Default is @code{combined}.
22042 Set minimum amplitude used in @code{log} amplitude scaler.
22046 @anchor{showspectrum}
22047 @section showspectrum
22049 Convert input audio to a video output, representing the audio frequency
22052 The filter accepts the following options:
22056 Specify the video size for the output. For the syntax of this option, check the
22057 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22058 Default value is @code{640x512}.
22061 Specify how the spectrum should slide along the window.
22063 It accepts the following values:
22066 the samples start again on the left when they reach the right
22068 the samples scroll from right to left
22070 frames are only produced when the samples reach the right
22072 the samples scroll from left to right
22075 Default value is @code{replace}.
22078 Specify display mode.
22080 It accepts the following values:
22083 all channels are displayed in the same row
22085 all channels are displayed in separate rows
22088 Default value is @samp{combined}.
22091 Specify display color mode.
22093 It accepts the following values:
22096 each channel is displayed in a separate color
22098 each channel is displayed using the same color scheme
22100 each channel is displayed using the rainbow color scheme
22102 each channel is displayed using the moreland color scheme
22104 each channel is displayed using the nebulae color scheme
22106 each channel is displayed using the fire color scheme
22108 each channel is displayed using the fiery color scheme
22110 each channel is displayed using the fruit color scheme
22112 each channel is displayed using the cool color scheme
22114 each channel is displayed using the magma color scheme
22116 each channel is displayed using the green color scheme
22118 each channel is displayed using the viridis color scheme
22120 each channel is displayed using the plasma color scheme
22122 each channel is displayed using the cividis color scheme
22124 each channel is displayed using the terrain color scheme
22127 Default value is @samp{channel}.
22130 Specify scale used for calculating intensity color values.
22132 It accepts the following values:
22137 square root, default
22148 Default value is @samp{sqrt}.
22151 Set saturation modifier for displayed colors. Negative values provide
22152 alternative color scheme. @code{0} is no saturation at all.
22153 Saturation must be in [-10.0, 10.0] range.
22154 Default value is @code{1}.
22157 Set window function.
22159 It accepts the following values:
22184 Default value is @code{hann}.
22187 Set orientation of time vs frequency axis. Can be @code{vertical} or
22188 @code{horizontal}. Default is @code{vertical}.
22191 Set ratio of overlap window. Default value is @code{0}.
22192 When value is @code{1} overlap is set to recommended size for specific
22193 window function currently used.
22196 Set scale gain for calculating intensity color values.
22197 Default value is @code{1}.
22200 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22203 Set color rotation, must be in [-1.0, 1.0] range.
22204 Default value is @code{0}.
22207 Set start frequency from which to display spectrogram. Default is @code{0}.
22210 Set stop frequency to which to display spectrogram. Default is @code{0}.
22213 Set upper frame rate limit. Default is @code{auto}, unlimited.
22216 Draw time and frequency axes and legends. Default is disabled.
22219 The usage is very similar to the showwaves filter; see the examples in that
22222 @subsection Examples
22226 Large window with logarithmic color scaling:
22228 showspectrum=s=1280x480:scale=log
22232 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22234 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22235 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22239 @section showspectrumpic
22241 Convert input audio to a single video frame, representing the audio frequency
22244 The filter accepts the following options:
22248 Specify the video size for the output. For the syntax of this option, check the
22249 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22250 Default value is @code{4096x2048}.
22253 Specify display mode.
22255 It accepts the following values:
22258 all channels are displayed in the same row
22260 all channels are displayed in separate rows
22262 Default value is @samp{combined}.
22265 Specify display color mode.
22267 It accepts the following values:
22270 each channel is displayed in a separate color
22272 each channel is displayed using the same color scheme
22274 each channel is displayed using the rainbow color scheme
22276 each channel is displayed using the moreland color scheme
22278 each channel is displayed using the nebulae color scheme
22280 each channel is displayed using the fire color scheme
22282 each channel is displayed using the fiery color scheme
22284 each channel is displayed using the fruit color scheme
22286 each channel is displayed using the cool color scheme
22288 each channel is displayed using the magma color scheme
22290 each channel is displayed using the green color scheme
22292 each channel is displayed using the viridis color scheme
22294 each channel is displayed using the plasma color scheme
22296 each channel is displayed using the cividis color scheme
22298 each channel is displayed using the terrain color scheme
22300 Default value is @samp{intensity}.
22303 Specify scale used for calculating intensity color values.
22305 It accepts the following values:
22310 square root, default
22320 Default value is @samp{log}.
22323 Set saturation modifier for displayed colors. Negative values provide
22324 alternative color scheme. @code{0} is no saturation at all.
22325 Saturation must be in [-10.0, 10.0] range.
22326 Default value is @code{1}.
22329 Set window function.
22331 It accepts the following values:
22355 Default value is @code{hann}.
22358 Set orientation of time vs frequency axis. Can be @code{vertical} or
22359 @code{horizontal}. Default is @code{vertical}.
22362 Set scale gain for calculating intensity color values.
22363 Default value is @code{1}.
22366 Draw time and frequency axes and legends. Default is enabled.
22369 Set color rotation, must be in [-1.0, 1.0] range.
22370 Default value is @code{0}.
22373 Set start frequency from which to display spectrogram. Default is @code{0}.
22376 Set stop frequency to which to display spectrogram. Default is @code{0}.
22379 @subsection Examples
22383 Extract an audio spectrogram of a whole audio track
22384 in a 1024x1024 picture using @command{ffmpeg}:
22386 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22390 @section showvolume
22392 Convert input audio volume to a video output.
22394 The filter accepts the following options:
22401 Set border width, allowed range is [0, 5]. Default is 1.
22404 Set channel width, allowed range is [80, 8192]. Default is 400.
22407 Set channel height, allowed range is [1, 900]. Default is 20.
22410 Set fade, allowed range is [0, 1]. Default is 0.95.
22413 Set volume color expression.
22415 The expression can use the following variables:
22419 Current max volume of channel in dB.
22425 Current channel number, starting from 0.
22429 If set, displays channel names. Default is enabled.
22432 If set, displays volume values. Default is enabled.
22435 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22436 default is @code{h}.
22439 Set step size, allowed range is [0, 5]. Default is 0, which means
22443 Set background opacity, allowed range is [0, 1]. Default is 0.
22446 Set metering mode, can be peak: @code{p} or rms: @code{r},
22447 default is @code{p}.
22450 Set display scale, can be linear: @code{lin} or log: @code{log},
22451 default is @code{lin}.
22455 If set to > 0., display a line for the max level
22456 in the previous seconds.
22457 default is disabled: @code{0.}
22460 The color of the max line. Use when @code{dm} option is set to > 0.
22461 default is: @code{orange}
22466 Convert input audio to a video output, representing the samples waves.
22468 The filter accepts the following options:
22472 Specify the video size for the output. For the syntax of this option, check the
22473 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22474 Default value is @code{600x240}.
22479 Available values are:
22482 Draw a point for each sample.
22485 Draw a vertical line for each sample.
22488 Draw a point for each sample and a line between them.
22491 Draw a centered vertical line for each sample.
22494 Default value is @code{point}.
22497 Set the number of samples which are printed on the same column. A
22498 larger value will decrease the frame rate. Must be a positive
22499 integer. This option can be set only if the value for @var{rate}
22500 is not explicitly specified.
22503 Set the (approximate) output frame rate. This is done by setting the
22504 option @var{n}. Default value is "25".
22506 @item split_channels
22507 Set if channels should be drawn separately or overlap. Default value is 0.
22510 Set colors separated by '|' which are going to be used for drawing of each channel.
22513 Set amplitude scale.
22515 Available values are:
22533 Set the draw mode. This is mostly useful to set for high @var{n}.
22535 Available values are:
22538 Scale pixel values for each drawn sample.
22541 Draw every sample directly.
22544 Default value is @code{scale}.
22547 @subsection Examples
22551 Output the input file audio and the corresponding video representation
22554 amovie=a.mp3,asplit[out0],showwaves[out1]
22558 Create a synthetic signal and show it with showwaves, forcing a
22559 frame rate of 30 frames per second:
22561 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22565 @section showwavespic
22567 Convert input audio to a single video frame, representing the samples waves.
22569 The filter accepts the following options:
22573 Specify the video size for the output. For the syntax of this option, check the
22574 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22575 Default value is @code{600x240}.
22577 @item split_channels
22578 Set if channels should be drawn separately or overlap. Default value is 0.
22581 Set colors separated by '|' which are going to be used for drawing of each channel.
22584 Set amplitude scale.
22586 Available values are:
22604 @subsection Examples
22608 Extract a channel split representation of the wave form of a whole audio track
22609 in a 1024x800 picture using @command{ffmpeg}:
22611 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22615 @section sidedata, asidedata
22617 Delete frame side data, or select frames based on it.
22619 This filter accepts the following options:
22623 Set mode of operation of the filter.
22625 Can be one of the following:
22629 Select every frame with side data of @code{type}.
22632 Delete side data of @code{type}. If @code{type} is not set, delete all side
22638 Set side data type used with all modes. Must be set for @code{select} mode. For
22639 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22640 in @file{libavutil/frame.h}. For example, to choose
22641 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22645 @section spectrumsynth
22647 Sythesize audio from 2 input video spectrums, first input stream represents
22648 magnitude across time and second represents phase across time.
22649 The filter will transform from frequency domain as displayed in videos back
22650 to time domain as presented in audio output.
22652 This filter is primarily created for reversing processed @ref{showspectrum}
22653 filter outputs, but can synthesize sound from other spectrograms too.
22654 But in such case results are going to be poor if the phase data is not
22655 available, because in such cases phase data need to be recreated, usually
22656 it's just recreated from random noise.
22657 For best results use gray only output (@code{channel} color mode in
22658 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22659 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22660 @code{data} option. Inputs videos should generally use @code{fullframe}
22661 slide mode as that saves resources needed for decoding video.
22663 The filter accepts the following options:
22667 Specify sample rate of output audio, the sample rate of audio from which
22668 spectrum was generated may differ.
22671 Set number of channels represented in input video spectrums.
22674 Set scale which was used when generating magnitude input spectrum.
22675 Can be @code{lin} or @code{log}. Default is @code{log}.
22678 Set slide which was used when generating inputs spectrums.
22679 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22680 Default is @code{fullframe}.
22683 Set window function used for resynthesis.
22686 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22687 which means optimal overlap for selected window function will be picked.
22690 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22691 Default is @code{vertical}.
22694 @subsection Examples
22698 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22699 then resynthesize videos back to audio with spectrumsynth:
22701 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
22702 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
22703 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22707 @section split, asplit
22709 Split input into several identical outputs.
22711 @code{asplit} works with audio input, @code{split} with video.
22713 The filter accepts a single parameter which specifies the number of outputs. If
22714 unspecified, it defaults to 2.
22716 @subsection Examples
22720 Create two separate outputs from the same input:
22722 [in] split [out0][out1]
22726 To create 3 or more outputs, you need to specify the number of
22729 [in] asplit=3 [out0][out1][out2]
22733 Create two separate outputs from the same input, one cropped and
22736 [in] split [splitout1][splitout2];
22737 [splitout1] crop=100:100:0:0 [cropout];
22738 [splitout2] pad=200:200:100:100 [padout];
22742 Create 5 copies of the input audio with @command{ffmpeg}:
22744 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22750 Receive commands sent through a libzmq client, and forward them to
22751 filters in the filtergraph.
22753 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22754 must be inserted between two video filters, @code{azmq} between two
22755 audio filters. Both are capable to send messages to any filter type.
22757 To enable these filters you need to install the libzmq library and
22758 headers and configure FFmpeg with @code{--enable-libzmq}.
22760 For more information about libzmq see:
22761 @url{http://www.zeromq.org/}
22763 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22764 receives messages sent through a network interface defined by the
22765 @option{bind_address} (or the abbreviation "@option{b}") option.
22766 Default value of this option is @file{tcp://localhost:5555}. You may
22767 want to alter this value to your needs, but do not forget to escape any
22768 ':' signs (see @ref{filtergraph escaping}).
22770 The received message must be in the form:
22772 @var{TARGET} @var{COMMAND} [@var{ARG}]
22775 @var{TARGET} specifies the target of the command, usually the name of
22776 the filter class or a specific filter instance name. The default
22777 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22778 but you can override this by using the @samp{filter_name@@id} syntax
22779 (see @ref{Filtergraph syntax}).
22781 @var{COMMAND} specifies the name of the command for the target filter.
22783 @var{ARG} is optional and specifies the optional argument list for the
22784 given @var{COMMAND}.
22786 Upon reception, the message is processed and the corresponding command
22787 is injected into the filtergraph. Depending on the result, the filter
22788 will send a reply to the client, adopting the format:
22790 @var{ERROR_CODE} @var{ERROR_REASON}
22794 @var{MESSAGE} is optional.
22796 @subsection Examples
22798 Look at @file{tools/zmqsend} for an example of a zmq client which can
22799 be used to send commands processed by these filters.
22801 Consider the following filtergraph generated by @command{ffplay}.
22802 In this example the last overlay filter has an instance name. All other
22803 filters will have default instance names.
22806 ffplay -dumpgraph 1 -f lavfi "
22807 color=s=100x100:c=red [l];
22808 color=s=100x100:c=blue [r];
22809 nullsrc=s=200x100, zmq [bg];
22810 [bg][l] overlay [bg+l];
22811 [bg+l][r] overlay@@my=x=100 "
22814 To change the color of the left side of the video, the following
22815 command can be used:
22817 echo Parsed_color_0 c yellow | tools/zmqsend
22820 To change the right side:
22822 echo Parsed_color_1 c pink | tools/zmqsend
22825 To change the position of the right side:
22827 echo overlay@@my x 150 | tools/zmqsend
22831 @c man end MULTIMEDIA FILTERS
22833 @chapter Multimedia Sources
22834 @c man begin MULTIMEDIA SOURCES
22836 Below is a description of the currently available multimedia sources.
22840 This is the same as @ref{movie} source, except it selects an audio
22846 Read audio and/or video stream(s) from a movie container.
22848 It accepts the following parameters:
22852 The name of the resource to read (not necessarily a file; it can also be a
22853 device or a stream accessed through some protocol).
22855 @item format_name, f
22856 Specifies the format assumed for the movie to read, and can be either
22857 the name of a container or an input device. If not specified, the
22858 format is guessed from @var{movie_name} or by probing.
22860 @item seek_point, sp
22861 Specifies the seek point in seconds. The frames will be output
22862 starting from this seek point. The parameter is evaluated with
22863 @code{av_strtod}, so the numerical value may be suffixed by an IS
22864 postfix. The default value is "0".
22867 Specifies the streams to read. Several streams can be specified,
22868 separated by "+". The source will then have as many outputs, in the
22869 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22870 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22871 respectively the default (best suited) video and audio stream. Default
22872 is "dv", or "da" if the filter is called as "amovie".
22874 @item stream_index, si
22875 Specifies the index of the video stream to read. If the value is -1,
22876 the most suitable video stream will be automatically selected. The default
22877 value is "-1". Deprecated. If the filter is called "amovie", it will select
22878 audio instead of video.
22881 Specifies how many times to read the stream in sequence.
22882 If the value is 0, the stream will be looped infinitely.
22883 Default value is "1".
22885 Note that when the movie is looped the source timestamps are not
22886 changed, so it will generate non monotonically increasing timestamps.
22888 @item discontinuity
22889 Specifies the time difference between frames above which the point is
22890 considered a timestamp discontinuity which is removed by adjusting the later
22894 It allows overlaying a second video on top of the main input of
22895 a filtergraph, as shown in this graph:
22897 input -----------> deltapts0 --> overlay --> output
22900 movie --> scale--> deltapts1 -------+
22902 @subsection Examples
22906 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
22907 on top of the input labelled "in":
22909 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
22910 [in] setpts=PTS-STARTPTS [main];
22911 [main][over] overlay=16:16 [out]
22915 Read from a video4linux2 device, and overlay it on top of the input
22918 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
22919 [in] setpts=PTS-STARTPTS [main];
22920 [main][over] overlay=16:16 [out]
22924 Read the first video stream and the audio stream with id 0x81 from
22925 dvd.vob; the video is connected to the pad named "video" and the audio is
22926 connected to the pad named "audio":
22928 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
22932 @subsection Commands
22934 Both movie and amovie support the following commands:
22937 Perform seek using "av_seek_frame".
22938 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
22941 @var{stream_index}: If stream_index is -1, a default
22942 stream is selected, and @var{timestamp} is automatically converted
22943 from AV_TIME_BASE units to the stream specific time_base.
22945 @var{timestamp}: Timestamp in AVStream.time_base units
22946 or, if no stream is specified, in AV_TIME_BASE units.
22948 @var{flags}: Flags which select direction and seeking mode.
22952 Get movie duration in AV_TIME_BASE units.
22956 @c man end MULTIMEDIA SOURCES