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 If a signal of stream rises above this level it will affect the gain
396 By default it is 0.125. Range is between 0.00097563 and 1.
399 Set a ratio by which the signal is reduced. 1:2 means that if the level
400 rose 4dB above the threshold, it will be only 2dB above after the reduction.
401 Default is 2. Range is between 1 and 20.
404 Amount of milliseconds the signal has to rise above the threshold before gain
405 reduction starts. Default is 20. Range is between 0.01 and 2000.
408 Amount of milliseconds the signal has to fall below the threshold before
409 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
412 Set the amount by how much signal will be amplified after processing.
413 Default is 1. Range is from 1 to 64.
416 Curve the sharp knee around the threshold to enter gain reduction more softly.
417 Default is 2.82843. Range is between 1 and 8.
420 Choose if the @code{average} level between all channels of input stream
421 or the louder(@code{maximum}) channel of input stream affects the
422 reduction. Default is @code{average}.
425 Should the exact signal be taken in case of @code{peak} or an RMS one in case
426 of @code{rms}. Default is @code{rms} which is mostly smoother.
429 How much to use compressed signal in output. Default is 1.
430 Range is between 0 and 1.
434 Simple audio dynamic range commpression/expansion filter.
436 The filter accepts the following options:
440 Set contrast. Default is 33. Allowed range is between 0 and 100.
445 Copy the input audio source unchanged to the output. This is mainly useful for
450 Apply cross fade from one input audio stream to another input audio stream.
451 The cross fade is applied for specified duration near the end of first stream.
453 The filter accepts the following options:
457 Specify the number of samples for which the cross fade effect has to last.
458 At the end of the cross fade effect the first input audio will be completely
459 silent. Default is 44100.
462 Specify the duration of the cross fade effect. See
463 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
464 for the accepted syntax.
465 By default the duration is determined by @var{nb_samples}.
466 If set this option is used instead of @var{nb_samples}.
469 Should first stream end overlap with second stream start. Default is enabled.
472 Set curve for cross fade transition for first stream.
475 Set curve for cross fade transition for second stream.
477 For description of available curve types see @ref{afade} filter description.
484 Cross fade from one input to another:
486 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
490 Cross fade from one input to another but without overlapping:
492 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
497 Split audio stream into several bands.
499 This filter splits audio stream into two or more frequency ranges.
500 Summing all streams back will give flat output.
502 The filter accepts the following options:
506 Set split frequencies. Those must be positive and increasing.
509 Set filter order, can be @var{2nd}, @var{4th} or @var{8th}.
510 Default is @var{4th}.
515 Reduce audio bit resolution.
517 This filter is bit crusher with enhanced functionality. A bit crusher
518 is used to audibly reduce number of bits an audio signal is sampled
519 with. This doesn't change the bit depth at all, it just produces the
520 effect. Material reduced in bit depth sounds more harsh and "digital".
521 This filter is able to even round to continuous values instead of discrete
523 Additionally it has a D/C offset which results in different crushing of
524 the lower and the upper half of the signal.
525 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
527 Another feature of this filter is the logarithmic mode.
528 This setting switches from linear distances between bits to logarithmic ones.
529 The result is a much more "natural" sounding crusher which doesn't gate low
530 signals for example. The human ear has a logarithmic perception,
531 so this kind of crushing is much more pleasant.
532 Logarithmic crushing is also able to get anti-aliased.
534 The filter accepts the following options:
550 Can be linear: @code{lin} or logarithmic: @code{log}.
559 Set sample reduction.
562 Enable LFO. By default disabled.
573 Delay audio filtering until a given wallclock timestamp. See the @ref{cue}
577 Remove impulsive noise from input audio.
579 Samples detected as impulsive noise are replaced by interpolated samples using
580 autoregressive modelling.
584 Set window size, in milliseconds. Allowed range is from @code{10} to
585 @code{100}. Default value is @code{55} milliseconds.
586 This sets size of window which will be processed at once.
589 Set window overlap, in percentage of window size. Allowed range is from
590 @code{50} to @code{95}. Default value is @code{75} percent.
591 Setting this to a very high value increases impulsive noise removal but makes
592 whole process much slower.
595 Set autoregression order, in percentage of window size. Allowed range is from
596 @code{0} to @code{25}. Default value is @code{2} percent. This option also
597 controls quality of interpolated samples using neighbour good samples.
600 Set threshold value. Allowed range is from @code{1} to @code{100}.
601 Default value is @code{2}.
602 This controls the strength of impulsive noise which is going to be removed.
603 The lower value, the more samples will be detected as impulsive noise.
606 Set burst fusion, in percentage of window size. Allowed range is @code{0} to
607 @code{10}. Default value is @code{2}.
608 If any two samples deteced as noise are spaced less than this value then any
609 sample inbetween those two samples will be also detected as noise.
614 It accepts the following values:
617 Select overlap-add method. Even not interpolated samples are slightly
618 changed with this method.
621 Select overlap-save method. Not interpolated samples remain unchanged.
624 Default value is @code{a}.
628 Remove clipped samples from input audio.
630 Samples detected as clipped are replaced by interpolated samples using
631 autoregressive modelling.
635 Set window size, in milliseconds. Allowed range is from @code{10} to @code{100}.
636 Default value is @code{55} milliseconds.
637 This sets size of window which will be processed at once.
640 Set window overlap, in percentage of window size. Allowed range is from @code{50}
641 to @code{95}. Default value is @code{75} percent.
644 Set autoregression order, in percentage of window size. Allowed range is from
645 @code{0} to @code{25}. Default value is @code{8} percent. This option also controls
646 quality of interpolated samples using neighbour good samples.
649 Set threshold value. Allowed range is from @code{1} to @code{100}.
650 Default value is @code{10}. Higher values make clip detection less aggressive.
653 Set size of histogram used to detect clips. Allowed range is from @code{100} to @code{9999}.
654 Default value is @code{1000}. Higher values make clip detection less aggressive.
659 It accepts the following values:
662 Select overlap-add method. Even not interpolated samples are slightly changed
666 Select overlap-save method. Not interpolated samples remain unchanged.
669 Default value is @code{a}.
674 Delay one or more audio channels.
676 Samples in delayed channel are filled with silence.
678 The filter accepts the following option:
682 Set list of delays in milliseconds for each channel separated by '|'.
683 Unused delays will be silently ignored. If number of given delays is
684 smaller than number of channels all remaining channels will not be delayed.
685 If you want to delay exact number of samples, append 'S' to number.
686 If you want instead to delay in seconds, append 's' to number.
693 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
694 the second channel (and any other channels that may be present) unchanged.
700 Delay second channel by 500 samples, the third channel by 700 samples and leave
701 the first channel (and any other channels that may be present) unchanged.
707 @section aderivative, aintegral
709 Compute derivative/integral of audio stream.
711 Applying both filters one after another produces original audio.
715 Apply echoing to the input audio.
717 Echoes are reflected sound and can occur naturally amongst mountains
718 (and sometimes large buildings) when talking or shouting; digital echo
719 effects emulate this behaviour and are often used to help fill out the
720 sound of a single instrument or vocal. The time difference between the
721 original signal and the reflection is the @code{delay}, and the
722 loudness of the reflected signal is the @code{decay}.
723 Multiple echoes can have different delays and decays.
725 A description of the accepted parameters follows.
729 Set input gain of reflected signal. Default is @code{0.6}.
732 Set output gain of reflected signal. Default is @code{0.3}.
735 Set list of time intervals in milliseconds between original signal and reflections
736 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
737 Default is @code{1000}.
740 Set list of loudness of reflected signals separated by '|'.
741 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
742 Default is @code{0.5}.
749 Make it sound as if there are twice as many instruments as are actually playing:
751 aecho=0.8:0.88:60:0.4
755 If delay is very short, then it sound like a (metallic) robot playing music:
761 A longer delay will sound like an open air concert in the mountains:
763 aecho=0.8:0.9:1000:0.3
767 Same as above but with one more mountain:
769 aecho=0.8:0.9:1000|1800:0.3|0.25
774 Audio emphasis filter creates or restores material directly taken from LPs or
775 emphased CDs with different filter curves. E.g. to store music on vinyl the
776 signal has to be altered by a filter first to even out the disadvantages of
777 this recording medium.
778 Once the material is played back the inverse filter has to be applied to
779 restore the distortion of the frequency response.
781 The filter accepts the following options:
791 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
792 use @code{production} mode. Default is @code{reproduction} mode.
795 Set filter type. Selects medium. Can be one of the following:
807 select Compact Disc (CD).
813 select 50µs (FM-KF).
815 select 75µs (FM-KF).
821 Modify an audio signal according to the specified expressions.
823 This filter accepts one or more expressions (one for each channel),
824 which are evaluated and used to modify a corresponding audio signal.
826 It accepts the following parameters:
830 Set the '|'-separated expressions list for each separate channel. If
831 the number of input channels is greater than the number of
832 expressions, the last specified expression is used for the remaining
835 @item channel_layout, c
836 Set output channel layout. If not specified, the channel layout is
837 specified by the number of expressions. If set to @samp{same}, it will
838 use by default the same input channel layout.
841 Each expression in @var{exprs} can contain the following constants and functions:
845 channel number of the current expression
848 number of the evaluated sample, starting from 0
854 time of the evaluated sample expressed in seconds
857 @item nb_out_channels
858 input and output number of channels
861 the value of input channel with number @var{CH}
864 Note: this filter is slow. For faster processing you should use a
873 aeval=val(ch)/2:c=same
877 Invert phase of the second channel:
886 Apply fade-in/out effect to input audio.
888 A description of the accepted parameters follows.
892 Specify the effect type, can be either @code{in} for fade-in, or
893 @code{out} for a fade-out effect. Default is @code{in}.
895 @item start_sample, ss
896 Specify the number of the start sample for starting to apply the fade
897 effect. Default is 0.
900 Specify the number of samples for which the fade effect has to last. At
901 the end of the fade-in effect the output audio will have the same
902 volume as the input audio, at the end of the fade-out transition
903 the output audio will be silence. Default is 44100.
906 Specify the start time of the fade effect. Default is 0.
907 The value must be specified as a time duration; see
908 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
909 for the accepted syntax.
910 If set this option is used instead of @var{start_sample}.
913 Specify the duration of the fade effect. See
914 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
915 for the accepted syntax.
916 At the end of the fade-in effect the output audio will have the same
917 volume as the input audio, at the end of the fade-out transition
918 the output audio will be silence.
919 By default the duration is determined by @var{nb_samples}.
920 If set this option is used instead of @var{nb_samples}.
923 Set curve for fade transition.
925 It accepts the following values:
928 select triangular, linear slope (default)
930 select quarter of sine wave
932 select half of sine wave
934 select exponential sine wave
938 select inverted parabola
952 select inverted quarter of sine wave
954 select inverted half of sine wave
956 select double-exponential seat
958 select double-exponential sigmoid
960 select logistic sigmoid
968 Fade in first 15 seconds of audio:
974 Fade out last 25 seconds of a 900 seconds audio:
976 afade=t=out:st=875:d=25
981 Denoise audio samples with FFT.
983 A description of the accepted parameters follows.
987 Set the noise reduction in dB, allowed range is 0.01 to 97.
988 Default value is 12 dB.
991 Set the noise floor in dB, allowed range is -80 to -20.
992 Default value is -50 dB.
997 It accepts the following values:
1006 Select shellac noise.
1009 Select custom noise, defined in @code{bn} option.
1011 Default value is white noise.
1015 Set custom band noise for every one of 15 bands.
1016 Bands are separated by ' ' or '|'.
1019 Set the residual floor in dB, allowed range is -80 to -20.
1020 Default value is -38 dB.
1023 Enable noise tracking. By default is disabled.
1024 With this enabled, noise floor is automatically adjusted.
1027 Enable residual tracking. By default is disabled.
1030 Set the output mode.
1032 It accepts the following values:
1035 Pass input unchanged.
1038 Pass noise filtered out.
1043 Default value is @var{o}.
1047 @subsection Commands
1049 This filter supports the following commands:
1051 @item sample_noise, sn
1052 Start or stop measuring noise profile.
1053 Syntax for the command is : "start" or "stop" string.
1054 After measuring noise profile is stopped it will be
1055 automatically applied in filtering.
1057 @item noise_reduction, nr
1058 Change noise reduction. Argument is single float number.
1059 Syntax for the command is : "@var{noise_reduction}"
1061 @item noise_floor, nf
1062 Change noise floor. Argument is single float number.
1063 Syntax for the command is : "@var{noise_floor}"
1065 @item output_mode, om
1066 Change output mode operation.
1067 Syntax for the command is : "i", "o" or "n" string.
1071 Apply arbitrary expressions to samples in frequency domain.
1075 Set frequency domain real expression for each separate channel separated
1076 by '|'. Default is "re".
1077 If the number of input channels is greater than the number of
1078 expressions, the last specified expression is used for the remaining
1082 Set frequency domain imaginary expression for each separate channel
1083 separated by '|'. Default is "im".
1085 Each expression in @var{real} and @var{imag} can contain the following
1086 constants and functions:
1093 current frequency bin number
1096 number of available bins
1099 channel number of the current expression
1108 current real part of frequency bin of current channel
1111 current imaginary part of frequency bin of current channel
1114 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1117 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1123 It accepts the following values:
1139 Default is @code{w4096}
1142 Set window function. Default is @code{hann}.
1145 Set window overlap. If set to 1, the recommended overlap for selected
1146 window function will be picked. Default is @code{0.75}.
1149 @subsection Examples
1153 Leave almost only low frequencies in audio:
1155 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1162 Apply an arbitrary Frequency Impulse Response filter.
1164 This filter is designed for applying long FIR filters,
1165 up to 60 seconds long.
1167 It can be used as component for digital crossover filters,
1168 room equalization, cross talk cancellation, wavefield synthesis,
1169 auralization, ambiophonics, ambisonics and spatialization.
1171 This filter uses second stream as FIR coefficients.
1172 If second stream holds single channel, it will be used
1173 for all input channels in first stream, otherwise
1174 number of channels in second stream must be same as
1175 number of channels in first stream.
1177 It accepts the following parameters:
1181 Set dry gain. This sets input gain.
1184 Set wet gain. This sets final output gain.
1187 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1190 Enable applying gain measured from power of IR.
1192 Set which approach to use for auto gain measurement.
1196 Do not apply any gain.
1199 select peak gain, very conservative approach. This is default value.
1202 select DC gain, limited application.
1205 select gain to noise approach, this is most popular one.
1209 Set gain to be applied to IR coefficients before filtering.
1210 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1213 Set format of IR stream. Can be @code{mono} or @code{input}.
1214 Default is @code{input}.
1217 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1218 Allowed range is 0.1 to 60 seconds.
1221 Show IR frequency reponse, magnitude(magenta) and phase(green) and group delay(yellow) in additional video stream.
1222 By default it is disabled.
1225 Set for which IR channel to display frequency response. By default is first channel
1226 displayed. This option is used only when @var{response} is enabled.
1229 Set video stream size. This option is used only when @var{response} is enabled.
1232 Set video stream frame rate. This option is used only when @var{response} is enabled.
1235 Set minimal partition size used for convolution. Default is @var{8192}.
1236 Allowed range is from @var{8} to @var{32768}.
1237 Lower values decreases latency at cost of higher CPU usage.
1240 Set maximal partition size used for convolution. Default is @var{8192}.
1241 Allowed range is from @var{8} to @var{32768}.
1242 Lower values may increase CPU usage.
1245 @subsection Examples
1249 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1251 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1258 Set output format constraints for the input audio. The framework will
1259 negotiate the most appropriate format to minimize conversions.
1261 It accepts the following parameters:
1265 A '|'-separated list of requested sample formats.
1268 A '|'-separated list of requested sample rates.
1270 @item channel_layouts
1271 A '|'-separated list of requested channel layouts.
1273 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1274 for the required syntax.
1277 If a parameter is omitted, all values are allowed.
1279 Force the output to either unsigned 8-bit or signed 16-bit stereo
1281 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1286 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1287 processing reduces disturbing noise between useful signals.
1289 Gating is done by detecting the volume below a chosen level @var{threshold}
1290 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1291 floor is set via @var{range}. Because an exact manipulation of the signal
1292 would cause distortion of the waveform the reduction can be levelled over
1293 time. This is done by setting @var{attack} and @var{release}.
1295 @var{attack} determines how long the signal has to fall below the threshold
1296 before any reduction will occur and @var{release} sets the time the signal
1297 has to rise above the threshold to reduce the reduction again.
1298 Shorter signals than the chosen attack time will be left untouched.
1302 Set input level before filtering.
1303 Default is 1. Allowed range is from 0.015625 to 64.
1306 Set the level of gain reduction when the signal is below the threshold.
1307 Default is 0.06125. Allowed range is from 0 to 1.
1310 If a signal rises above this level the gain reduction is released.
1311 Default is 0.125. Allowed range is from 0 to 1.
1314 Set a ratio by which the signal is reduced.
1315 Default is 2. Allowed range is from 1 to 9000.
1318 Amount of milliseconds the signal has to rise above the threshold before gain
1320 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1323 Amount of milliseconds the signal has to fall below the threshold before the
1324 reduction is increased again. Default is 250 milliseconds.
1325 Allowed range is from 0.01 to 9000.
1328 Set amount of amplification of signal after processing.
1329 Default is 1. Allowed range is from 1 to 64.
1332 Curve the sharp knee around the threshold to enter gain reduction more softly.
1333 Default is 2.828427125. Allowed range is from 1 to 8.
1336 Choose if exact signal should be taken for detection or an RMS like one.
1337 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1340 Choose if the average level between all channels or the louder channel affects
1342 Default is @code{average}. Can be @code{average} or @code{maximum}.
1347 Apply an arbitrary Infinite Impulse Response filter.
1349 It accepts the following parameters:
1353 Set numerator/zeros coefficients.
1356 Set denominator/poles coefficients.
1368 Set coefficients format.
1374 Z-plane zeros/poles, cartesian (default)
1376 Z-plane zeros/poles, polar radians
1378 Z-plane zeros/poles, polar degrees
1382 Set kind of processing.
1383 Can be @code{d} - direct or @code{s} - serial cascading. Defauls is @code{s}.
1386 Set filtering precision.
1390 double-precision floating-point (default)
1392 single-precision floating-point
1400 Show IR frequency reponse, magnitude and phase in additional video stream.
1401 By default it is disabled.
1404 Set for which IR channel to display frequency response. By default is first channel
1405 displayed. This option is used only when @var{response} is enabled.
1408 Set video stream size. This option is used only when @var{response} is enabled.
1411 Coefficients in @code{tf} format are separated by spaces and are in ascending
1414 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1415 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1418 Different coefficients and gains can be provided for every channel, in such case
1419 use '|' to separate coefficients or gains. Last provided coefficients will be
1420 used for all remaining channels.
1422 @subsection Examples
1426 Apply 2 pole elliptic notch at arround 5000Hz for 48000 Hz sample rate:
1428 aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
1432 Same as above but in @code{zp} format:
1434 aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
1440 The limiter prevents an input signal from rising over a desired threshold.
1441 This limiter uses lookahead technology to prevent your signal from distorting.
1442 It means that there is a small delay after the signal is processed. Keep in mind
1443 that the delay it produces is the attack time you set.
1445 The filter accepts the following options:
1449 Set input gain. Default is 1.
1452 Set output gain. Default is 1.
1455 Don't let signals above this level pass the limiter. Default is 1.
1458 The limiter will reach its attenuation level in this amount of time in
1459 milliseconds. Default is 5 milliseconds.
1462 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1463 Default is 50 milliseconds.
1466 When gain reduction is always needed ASC takes care of releasing to an
1467 average reduction level rather than reaching a reduction of 0 in the release
1471 Select how much the release time is affected by ASC, 0 means nearly no changes
1472 in release time while 1 produces higher release times.
1475 Auto level output signal. Default is enabled.
1476 This normalizes audio back to 0dB if enabled.
1479 Depending on picked setting it is recommended to upsample input 2x or 4x times
1480 with @ref{aresample} before applying this filter.
1484 Apply a two-pole all-pass filter with central frequency (in Hz)
1485 @var{frequency}, and filter-width @var{width}.
1486 An all-pass filter changes the audio's frequency to phase relationship
1487 without changing its frequency to amplitude relationship.
1489 The filter accepts the following options:
1493 Set frequency in Hz.
1496 Set method to specify band-width of filter.
1511 Specify the band-width of a filter in width_type units.
1514 Specify which channels to filter, by default all available are filtered.
1517 @subsection Commands
1519 This filter supports the following commands:
1522 Change allpass frequency.
1523 Syntax for the command is : "@var{frequency}"
1526 Change allpass width_type.
1527 Syntax for the command is : "@var{width_type}"
1530 Change allpass width.
1531 Syntax for the command is : "@var{width}"
1538 The filter accepts the following options:
1542 Set the number of loops. Setting this value to -1 will result in infinite loops.
1546 Set maximal number of samples. Default is 0.
1549 Set first sample of loop. Default is 0.
1555 Merge two or more audio streams into a single multi-channel stream.
1557 The filter accepts the following options:
1562 Set the number of inputs. Default is 2.
1566 If the channel layouts of the inputs are disjoint, and therefore compatible,
1567 the channel layout of the output will be set accordingly and the channels
1568 will be reordered as necessary. If the channel layouts of the inputs are not
1569 disjoint, the output will have all the channels of the first input then all
1570 the channels of the second input, in that order, and the channel layout of
1571 the output will be the default value corresponding to the total number of
1574 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1575 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1576 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1577 first input, b1 is the first channel of the second input).
1579 On the other hand, if both input are in stereo, the output channels will be
1580 in the default order: a1, a2, b1, b2, and the channel layout will be
1581 arbitrarily set to 4.0, which may or may not be the expected value.
1583 All inputs must have the same sample rate, and format.
1585 If inputs do not have the same duration, the output will stop with the
1588 @subsection Examples
1592 Merge two mono files into a stereo stream:
1594 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1598 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1600 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
1606 Mixes multiple audio inputs into a single output.
1608 Note that this filter only supports float samples (the @var{amerge}
1609 and @var{pan} audio filters support many formats). If the @var{amix}
1610 input has integer samples then @ref{aresample} will be automatically
1611 inserted to perform the conversion to float samples.
1615 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1617 will mix 3 input audio streams to a single output with the same duration as the
1618 first input and a dropout transition time of 3 seconds.
1620 It accepts the following parameters:
1624 The number of inputs. If unspecified, it defaults to 2.
1627 How to determine the end-of-stream.
1631 The duration of the longest input. (default)
1634 The duration of the shortest input.
1637 The duration of the first input.
1641 @item dropout_transition
1642 The transition time, in seconds, for volume renormalization when an input
1643 stream ends. The default value is 2 seconds.
1646 Specify weight of each input audio stream as sequence.
1647 Each weight is separated by space. By default all inputs have same weight.
1652 Multiply first audio stream with second audio stream and store result
1653 in output audio stream. Multiplication is done by multiplying each
1654 sample from first stream with sample at same position from second stream.
1656 With this element-wise multiplication one can create amplitude fades and
1657 amplitude modulations.
1659 @section anequalizer
1661 High-order parametric multiband equalizer for each channel.
1663 It accepts the following parameters:
1667 This option string is in format:
1668 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1669 Each equalizer band is separated by '|'.
1673 Set channel number to which equalization will be applied.
1674 If input doesn't have that channel the entry is ignored.
1677 Set central frequency for band.
1678 If input doesn't have that frequency the entry is ignored.
1681 Set band width in hertz.
1684 Set band gain in dB.
1687 Set filter type for band, optional, can be:
1691 Butterworth, this is default.
1702 With this option activated frequency response of anequalizer is displayed
1706 Set video stream size. Only useful if curves option is activated.
1709 Set max gain that will be displayed. Only useful if curves option is activated.
1710 Setting this to a reasonable value makes it possible to display gain which is derived from
1711 neighbour bands which are too close to each other and thus produce higher gain
1712 when both are activated.
1715 Set frequency scale used to draw frequency response in video output.
1716 Can be linear or logarithmic. Default is logarithmic.
1719 Set color for each channel curve which is going to be displayed in video stream.
1720 This is list of color names separated by space or by '|'.
1721 Unrecognised or missing colors will be replaced by white color.
1724 @subsection Examples
1728 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1729 for first 2 channels using Chebyshev type 1 filter:
1731 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1735 @subsection Commands
1737 This filter supports the following commands:
1740 Alter existing filter parameters.
1741 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1743 @var{fN} is existing filter number, starting from 0, if no such filter is available
1745 @var{freq} set new frequency parameter.
1746 @var{width} set new width parameter in herz.
1747 @var{gain} set new gain parameter in dB.
1749 Full filter invocation with asendcmd may look like this:
1750 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1755 Pass the audio source unchanged to the output.
1759 Pad the end of an audio stream with silence.
1761 This can be used together with @command{ffmpeg} @option{-shortest} to
1762 extend audio streams to the same length as the video stream.
1764 A description of the accepted options follows.
1768 Set silence packet size. Default value is 4096.
1771 Set the number of samples of silence to add to the end. After the
1772 value is reached, the stream is terminated. This option is mutually
1773 exclusive with @option{whole_len}.
1776 Set the minimum total number of samples in the output audio stream. If
1777 the value is longer than the input audio length, silence is added to
1778 the end, until the value is reached. This option is mutually exclusive
1779 with @option{pad_len}.
1782 Specify the duration of samples of silence to add. See
1783 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1784 for the accepted syntax. Used only if set to non-zero value.
1787 Specify the minimum total duration in the output audio stream. See
1788 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1789 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1790 the input audio length, silence is added to the end, until the value is reached.
1791 This option is mutually exclusive with @option{pad_dur}
1794 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1795 nor @option{whole_dur} option is set, the filter will add silence to the end of
1796 the input stream indefinitely.
1798 @subsection Examples
1802 Add 1024 samples of silence to the end of the input:
1808 Make sure the audio output will contain at least 10000 samples, pad
1809 the input with silence if required:
1811 apad=whole_len=10000
1815 Use @command{ffmpeg} to pad the audio input with silence, so that the
1816 video stream will always result the shortest and will be converted
1817 until the end in the output file when using the @option{shortest}
1820 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1825 Add a phasing effect to the input audio.
1827 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1828 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1830 A description of the accepted parameters follows.
1834 Set input gain. Default is 0.4.
1837 Set output gain. Default is 0.74
1840 Set delay in milliseconds. Default is 3.0.
1843 Set decay. Default is 0.4.
1846 Set modulation speed in Hz. Default is 0.5.
1849 Set modulation type. Default is triangular.
1851 It accepts the following values:
1860 Audio pulsator is something between an autopanner and a tremolo.
1861 But it can produce funny stereo effects as well. Pulsator changes the volume
1862 of the left and right channel based on a LFO (low frequency oscillator) with
1863 different waveforms and shifted phases.
1864 This filter have the ability to define an offset between left and right
1865 channel. An offset of 0 means that both LFO shapes match each other.
1866 The left and right channel are altered equally - a conventional tremolo.
1867 An offset of 50% means that the shape of the right channel is exactly shifted
1868 in phase (or moved backwards about half of the frequency) - pulsator acts as
1869 an autopanner. At 1 both curves match again. Every setting in between moves the
1870 phase shift gapless between all stages and produces some "bypassing" sounds with
1871 sine and triangle waveforms. The more you set the offset near 1 (starting from
1872 the 0.5) the faster the signal passes from the left to the right speaker.
1874 The filter accepts the following options:
1878 Set input gain. By default it is 1. Range is [0.015625 - 64].
1881 Set output gain. By default it is 1. Range is [0.015625 - 64].
1884 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1885 sawup or sawdown. Default is sine.
1888 Set modulation. Define how much of original signal is affected by the LFO.
1891 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1894 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1897 Set pulse width. Default is 1. Allowed range is [0 - 2].
1900 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1903 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1907 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1911 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1912 if timing is set to hz.
1918 Resample the input audio to the specified parameters, using the
1919 libswresample library. If none are specified then the filter will
1920 automatically convert between its input and output.
1922 This filter is also able to stretch/squeeze the audio data to make it match
1923 the timestamps or to inject silence / cut out audio to make it match the
1924 timestamps, do a combination of both or do neither.
1926 The filter accepts the syntax
1927 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1928 expresses a sample rate and @var{resampler_options} is a list of
1929 @var{key}=@var{value} pairs, separated by ":". See the
1930 @ref{Resampler Options,,"Resampler Options" section in the
1931 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1932 for the complete list of supported options.
1934 @subsection Examples
1938 Resample the input audio to 44100Hz:
1944 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1945 samples per second compensation:
1947 aresample=async=1000
1953 Reverse an audio clip.
1955 Warning: This filter requires memory to buffer the entire clip, so trimming
1958 @subsection Examples
1962 Take the first 5 seconds of a clip, and reverse it.
1964 atrim=end=5,areverse
1968 @section asetnsamples
1970 Set the number of samples per each output audio frame.
1972 The last output packet may contain a different number of samples, as
1973 the filter will flush all the remaining samples when the input audio
1976 The filter accepts the following options:
1980 @item nb_out_samples, n
1981 Set the number of frames per each output audio frame. The number is
1982 intended as the number of samples @emph{per each channel}.
1983 Default value is 1024.
1986 If set to 1, the filter will pad the last audio frame with zeroes, so
1987 that the last frame will contain the same number of samples as the
1988 previous ones. Default value is 1.
1991 For example, to set the number of per-frame samples to 1234 and
1992 disable padding for the last frame, use:
1994 asetnsamples=n=1234:p=0
1999 Set the sample rate without altering the PCM data.
2000 This will result in a change of speed and pitch.
2002 The filter accepts the following options:
2005 @item sample_rate, r
2006 Set the output sample rate. Default is 44100 Hz.
2011 Show a line containing various information for each input audio frame.
2012 The input audio is not modified.
2014 The shown line contains a sequence of key/value pairs of the form
2015 @var{key}:@var{value}.
2017 The following values are shown in the output:
2021 The (sequential) number of the input frame, starting from 0.
2024 The presentation timestamp of the input frame, in time base units; the time base
2025 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2028 The presentation timestamp of the input frame in seconds.
2031 position of the frame in the input stream, -1 if this information in
2032 unavailable and/or meaningless (for example in case of synthetic audio)
2041 The sample rate for the audio frame.
2044 The number of samples (per channel) in the frame.
2047 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2048 audio, the data is treated as if all the planes were concatenated.
2050 @item plane_checksums
2051 A list of Adler-32 checksums for each data plane.
2057 Display time domain statistical information about the audio channels.
2058 Statistics are calculated and displayed for each audio channel and,
2059 where applicable, an overall figure is also given.
2061 It accepts the following option:
2064 Short window length in seconds, used for peak and trough RMS measurement.
2065 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2069 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2070 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2073 Available keys for each channel are:
2109 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2110 this @code{lavfi.astats.Overall.Peak_count}.
2112 For description what each key means read below.
2115 Set number of frame after which stats are going to be recalculated.
2116 Default is disabled.
2119 A description of each shown parameter follows:
2123 Mean amplitude displacement from zero.
2126 Minimal sample level.
2129 Maximal sample level.
2131 @item Min difference
2132 Minimal difference between two consecutive samples.
2134 @item Max difference
2135 Maximal difference between two consecutive samples.
2137 @item Mean difference
2138 Mean difference between two consecutive samples.
2139 The average of each difference between two consecutive samples.
2141 @item RMS difference
2142 Root Mean Square difference between two consecutive samples.
2146 Standard peak and RMS level measured in dBFS.
2150 Peak and trough values for RMS level measured over a short window.
2153 Standard ratio of peak to RMS level (note: not in dB).
2156 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2157 (i.e. either @var{Min level} or @var{Max level}).
2160 Number of occasions (not the number of samples) that the signal attained either
2161 @var{Min level} or @var{Max level}.
2164 Overall bit depth of audio. Number of bits used for each sample.
2167 Measured dynamic range of audio in dB.
2169 @item Zero crossings
2170 Number of points where the waveform crosses the zero level axis.
2172 @item Zero crossings rate
2173 Rate of Zero crossings and number of audio samples.
2180 The filter accepts exactly one parameter, the audio tempo. If not
2181 specified then the filter will assume nominal 1.0 tempo. Tempo must
2182 be in the [0.5, 100.0] range.
2184 Note that tempo greater than 2 will skip some samples rather than
2185 blend them in. If for any reason this is a concern it is always
2186 possible to daisy-chain several instances of atempo to achieve the
2187 desired product tempo.
2189 @subsection Examples
2193 Slow down audio to 80% tempo:
2199 To speed up audio to 300% tempo:
2205 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2207 atempo=sqrt(3),atempo=sqrt(3)
2213 Trim the input so that the output contains one continuous subpart of the input.
2215 It accepts the following parameters:
2218 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2219 sample with the timestamp @var{start} will be the first sample in the output.
2222 Specify time of the first audio sample that will be dropped, i.e. the
2223 audio sample immediately preceding the one with the timestamp @var{end} will be
2224 the last sample in the output.
2227 Same as @var{start}, except this option sets the start timestamp in samples
2231 Same as @var{end}, except this option sets the end timestamp in samples instead
2235 The maximum duration of the output in seconds.
2238 The number of the first sample that should be output.
2241 The number of the first sample that should be dropped.
2244 @option{start}, @option{end}, and @option{duration} are expressed as time
2245 duration specifications; see
2246 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2248 Note that the first two sets of the start/end options and the @option{duration}
2249 option look at the frame timestamp, while the _sample options simply count the
2250 samples that pass through the filter. So start/end_pts and start/end_sample will
2251 give different results when the timestamps are wrong, inexact or do not start at
2252 zero. Also note that this filter does not modify the timestamps. If you wish
2253 to have the output timestamps start at zero, insert the asetpts filter after the
2256 If multiple start or end options are set, this filter tries to be greedy and
2257 keep all samples that match at least one of the specified constraints. To keep
2258 only the part that matches all the constraints at once, chain multiple atrim
2261 The defaults are such that all the input is kept. So it is possible to set e.g.
2262 just the end values to keep everything before the specified time.
2267 Drop everything except the second minute of input:
2269 ffmpeg -i INPUT -af atrim=60:120
2273 Keep only the first 1000 samples:
2275 ffmpeg -i INPUT -af atrim=end_sample=1000
2282 Apply a two-pole Butterworth band-pass filter with central
2283 frequency @var{frequency}, and (3dB-point) band-width width.
2284 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2285 instead of the default: constant 0dB peak gain.
2286 The filter roll off at 6dB per octave (20dB per decade).
2288 The filter accepts the following options:
2292 Set the filter's central frequency. Default is @code{3000}.
2295 Constant skirt gain if set to 1. Defaults to 0.
2298 Set method to specify band-width of filter.
2313 Specify the band-width of a filter in width_type units.
2316 Specify which channels to filter, by default all available are filtered.
2319 @subsection Commands
2321 This filter supports the following commands:
2324 Change bandpass frequency.
2325 Syntax for the command is : "@var{frequency}"
2328 Change bandpass width_type.
2329 Syntax for the command is : "@var{width_type}"
2332 Change bandpass width.
2333 Syntax for the command is : "@var{width}"
2338 Apply a two-pole Butterworth band-reject filter with central
2339 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2340 The filter roll off at 6dB per octave (20dB per decade).
2342 The filter accepts the following options:
2346 Set the filter's central frequency. Default is @code{3000}.
2349 Set method to specify band-width of filter.
2364 Specify the band-width of a filter in width_type units.
2367 Specify which channels to filter, by default all available are filtered.
2370 @subsection Commands
2372 This filter supports the following commands:
2375 Change bandreject frequency.
2376 Syntax for the command is : "@var{frequency}"
2379 Change bandreject width_type.
2380 Syntax for the command is : "@var{width_type}"
2383 Change bandreject width.
2384 Syntax for the command is : "@var{width}"
2387 @section bass, lowshelf
2389 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2390 shelving filter with a response similar to that of a standard
2391 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2393 The filter accepts the following options:
2397 Give the gain at 0 Hz. Its useful range is about -20
2398 (for a large cut) to +20 (for a large boost).
2399 Beware of clipping when using a positive gain.
2402 Set the filter's central frequency and so can be used
2403 to extend or reduce the frequency range to be boosted or cut.
2404 The default value is @code{100} Hz.
2407 Set method to specify band-width of filter.
2422 Determine how steep is the filter's shelf transition.
2425 Specify which channels to filter, by default all available are filtered.
2428 @subsection Commands
2430 This filter supports the following commands:
2433 Change bass frequency.
2434 Syntax for the command is : "@var{frequency}"
2437 Change bass width_type.
2438 Syntax for the command is : "@var{width_type}"
2442 Syntax for the command is : "@var{width}"
2446 Syntax for the command is : "@var{gain}"
2451 Apply a biquad IIR filter with the given coefficients.
2452 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2453 are the numerator and denominator coefficients respectively.
2454 and @var{channels}, @var{c} specify which channels to filter, by default all
2455 available are filtered.
2457 @subsection Commands
2459 This filter supports the following commands:
2467 Change biquad parameter.
2468 Syntax for the command is : "@var{value}"
2472 Bauer stereo to binaural transformation, which improves headphone listening of
2473 stereo audio records.
2475 To enable compilation of this filter you need to configure FFmpeg with
2476 @code{--enable-libbs2b}.
2478 It accepts the following parameters:
2482 Pre-defined crossfeed level.
2486 Default level (fcut=700, feed=50).
2489 Chu Moy circuit (fcut=700, feed=60).
2492 Jan Meier circuit (fcut=650, feed=95).
2497 Cut frequency (in Hz).
2506 Remap input channels to new locations.
2508 It accepts the following parameters:
2511 Map channels from input to output. The argument is a '|'-separated list of
2512 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2513 @var{in_channel} form. @var{in_channel} can be either the name of the input
2514 channel (e.g. FL for front left) or its index in the input channel layout.
2515 @var{out_channel} is the name of the output channel or its index in the output
2516 channel layout. If @var{out_channel} is not given then it is implicitly an
2517 index, starting with zero and increasing by one for each mapping.
2519 @item channel_layout
2520 The channel layout of the output stream.
2523 If no mapping is present, the filter will implicitly map input channels to
2524 output channels, preserving indices.
2526 @subsection Examples
2530 For example, assuming a 5.1+downmix input MOV file,
2532 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2534 will create an output WAV file tagged as stereo from the downmix channels of
2538 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2540 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2544 @section channelsplit
2546 Split each channel from an input audio stream into a separate output stream.
2548 It accepts the following parameters:
2550 @item channel_layout
2551 The channel layout of the input stream. The default is "stereo".
2553 A channel layout describing the channels to be extracted as separate output streams
2554 or "all" to extract each input channel as a separate stream. The default is "all".
2556 Choosing channels not present in channel layout in the input will result in an error.
2559 @subsection Examples
2563 For example, assuming a stereo input MP3 file,
2565 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2567 will create an output Matroska file with two audio streams, one containing only
2568 the left channel and the other the right channel.
2571 Split a 5.1 WAV file into per-channel files:
2573 ffmpeg -i in.wav -filter_complex
2574 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2575 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2576 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2581 Extract only LFE from a 5.1 WAV file:
2583 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2584 -map '[LFE]' lfe.wav
2589 Add a chorus effect to the audio.
2591 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2593 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2594 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2595 The modulation depth defines the range the modulated delay is played before or after
2596 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2597 sound tuned around the original one, like in a chorus where some vocals are slightly
2600 It accepts the following parameters:
2603 Set input gain. Default is 0.4.
2606 Set output gain. Default is 0.4.
2609 Set delays. A typical delay is around 40ms to 60ms.
2621 @subsection Examples
2627 chorus=0.7:0.9:55:0.4:0.25:2
2633 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2637 Fuller sounding chorus with three delays:
2639 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
2644 Compress or expand the audio's dynamic range.
2646 It accepts the following parameters:
2652 A list of times in seconds for each channel over which the instantaneous level
2653 of the input signal is averaged to determine its volume. @var{attacks} refers to
2654 increase of volume and @var{decays} refers to decrease of volume. For most
2655 situations, the attack time (response to the audio getting louder) should be
2656 shorter than the decay time, because the human ear is more sensitive to sudden
2657 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2658 a typical value for decay is 0.8 seconds.
2659 If specified number of attacks & decays is lower than number of channels, the last
2660 set attack/decay will be used for all remaining channels.
2663 A list of points for the transfer function, specified in dB relative to the
2664 maximum possible signal amplitude. Each key points list must be defined using
2665 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2666 @code{x0/y0 x1/y1 x2/y2 ....}
2668 The input values must be in strictly increasing order but the transfer function
2669 does not have to be monotonically rising. The point @code{0/0} is assumed but
2670 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2671 function are @code{-70/-70|-60/-20|1/0}.
2674 Set the curve radius in dB for all joints. It defaults to 0.01.
2677 Set the additional gain in dB to be applied at all points on the transfer
2678 function. This allows for easy adjustment of the overall gain.
2682 Set an initial volume, in dB, to be assumed for each channel when filtering
2683 starts. This permits the user to supply a nominal level initially, so that, for
2684 example, a very large gain is not applied to initial signal levels before the
2685 companding has begun to operate. A typical value for audio which is initially
2686 quiet is -90 dB. It defaults to 0.
2689 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2690 delayed before being fed to the volume adjuster. Specifying a delay
2691 approximately equal to the attack/decay times allows the filter to effectively
2692 operate in predictive rather than reactive mode. It defaults to 0.
2696 @subsection Examples
2700 Make music with both quiet and loud passages suitable for listening to in a
2703 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2706 Another example for audio with whisper and explosion parts:
2708 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2712 A noise gate for when the noise is at a lower level than the signal:
2714 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2718 Here is another noise gate, this time for when the noise is at a higher level
2719 than the signal (making it, in some ways, similar to squelch):
2721 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2725 2:1 compression starting at -6dB:
2727 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2731 2:1 compression starting at -9dB:
2733 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2737 2:1 compression starting at -12dB:
2739 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2743 2:1 compression starting at -18dB:
2745 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2749 3:1 compression starting at -15dB:
2751 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2757 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2763 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
2767 Hard limiter at -6dB:
2769 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2773 Hard limiter at -12dB:
2775 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2779 Hard noise gate at -35 dB:
2781 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2787 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2791 @section compensationdelay
2793 Compensation Delay Line is a metric based delay to compensate differing
2794 positions of microphones or speakers.
2796 For example, you have recorded guitar with two microphones placed in
2797 different location. Because the front of sound wave has fixed speed in
2798 normal conditions, the phasing of microphones can vary and depends on
2799 their location and interposition. The best sound mix can be achieved when
2800 these microphones are in phase (synchronized). Note that distance of
2801 ~30 cm between microphones makes one microphone to capture signal in
2802 antiphase to another microphone. That makes the final mix sounding moody.
2803 This filter helps to solve phasing problems by adding different delays
2804 to each microphone track and make them synchronized.
2806 The best result can be reached when you take one track as base and
2807 synchronize other tracks one by one with it.
2808 Remember that synchronization/delay tolerance depends on sample rate, too.
2809 Higher sample rates will give more tolerance.
2811 It accepts the following parameters:
2815 Set millimeters distance. This is compensation distance for fine tuning.
2819 Set cm distance. This is compensation distance for tightening distance setup.
2823 Set meters distance. This is compensation distance for hard distance setup.
2827 Set dry amount. Amount of unprocessed (dry) signal.
2831 Set wet amount. Amount of processed (wet) signal.
2835 Set temperature degree in Celsius. This is the temperature of the environment.
2840 Apply headphone crossfeed filter.
2842 Crossfeed is the process of blending the left and right channels of stereo
2844 It is mainly used to reduce extreme stereo separation of low frequencies.
2846 The intent is to produce more speaker like sound to the listener.
2848 The filter accepts the following options:
2852 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2853 This sets gain of low shelf filter for side part of stereo image.
2854 Default is -6dB. Max allowed is -30db when strength is set to 1.
2857 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2858 This sets cut off frequency of low shelf filter. Default is cut off near
2859 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2862 Set input gain. Default is 0.9.
2865 Set output gain. Default is 1.
2868 @section crystalizer
2869 Simple algorithm to expand audio dynamic range.
2871 The filter accepts the following options:
2875 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2876 (unchanged sound) to 10.0 (maximum effect).
2879 Enable clipping. By default is enabled.
2883 Apply a DC shift to the audio.
2885 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2886 in the recording chain) from the audio. The effect of a DC offset is reduced
2887 headroom and hence volume. The @ref{astats} filter can be used to determine if
2888 a signal has a DC offset.
2892 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2896 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2897 used to prevent clipping.
2901 Measure audio dynamic range.
2903 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2904 is found in transition material. And anything less that 8 have very poor dynamics
2905 and is very compressed.
2907 The filter accepts the following options:
2911 Set window length in seconds used to split audio into segments of equal length.
2912 Default is 3 seconds.
2916 Dynamic Audio Normalizer.
2918 This filter applies a certain amount of gain to the input audio in order
2919 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2920 contrast to more "simple" normalization algorithms, the Dynamic Audio
2921 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2922 This allows for applying extra gain to the "quiet" sections of the audio
2923 while avoiding distortions or clipping the "loud" sections. In other words:
2924 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2925 sections, in the sense that the volume of each section is brought to the
2926 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2927 this goal *without* applying "dynamic range compressing". It will retain 100%
2928 of the dynamic range *within* each section of the audio file.
2932 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2933 Default is 500 milliseconds.
2934 The Dynamic Audio Normalizer processes the input audio in small chunks,
2935 referred to as frames. This is required, because a peak magnitude has no
2936 meaning for just a single sample value. Instead, we need to determine the
2937 peak magnitude for a contiguous sequence of sample values. While a "standard"
2938 normalizer would simply use the peak magnitude of the complete file, the
2939 Dynamic Audio Normalizer determines the peak magnitude individually for each
2940 frame. The length of a frame is specified in milliseconds. By default, the
2941 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2942 been found to give good results with most files.
2943 Note that the exact frame length, in number of samples, will be determined
2944 automatically, based on the sampling rate of the individual input audio file.
2947 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2948 number. Default is 31.
2949 Probably the most important parameter of the Dynamic Audio Normalizer is the
2950 @code{window size} of the Gaussian smoothing filter. The filter's window size
2951 is specified in frames, centered around the current frame. For the sake of
2952 simplicity, this must be an odd number. Consequently, the default value of 31
2953 takes into account the current frame, as well as the 15 preceding frames and
2954 the 15 subsequent frames. Using a larger window results in a stronger
2955 smoothing effect and thus in less gain variation, i.e. slower gain
2956 adaptation. Conversely, using a smaller window results in a weaker smoothing
2957 effect and thus in more gain variation, i.e. faster gain adaptation.
2958 In other words, the more you increase this value, the more the Dynamic Audio
2959 Normalizer will behave like a "traditional" normalization filter. On the
2960 contrary, the more you decrease this value, the more the Dynamic Audio
2961 Normalizer will behave like a dynamic range compressor.
2964 Set the target peak value. This specifies the highest permissible magnitude
2965 level for the normalized audio input. This filter will try to approach the
2966 target peak magnitude as closely as possible, but at the same time it also
2967 makes sure that the normalized signal will never exceed the peak magnitude.
2968 A frame's maximum local gain factor is imposed directly by the target peak
2969 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2970 It is not recommended to go above this value.
2973 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2974 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2975 factor for each input frame, i.e. the maximum gain factor that does not
2976 result in clipping or distortion. The maximum gain factor is determined by
2977 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2978 additionally bounds the frame's maximum gain factor by a predetermined
2979 (global) maximum gain factor. This is done in order to avoid excessive gain
2980 factors in "silent" or almost silent frames. By default, the maximum gain
2981 factor is 10.0, For most inputs the default value should be sufficient and
2982 it usually is not recommended to increase this value. Though, for input
2983 with an extremely low overall volume level, it may be necessary to allow even
2984 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2985 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2986 Instead, a "sigmoid" threshold function will be applied. This way, the
2987 gain factors will smoothly approach the threshold value, but never exceed that
2991 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2992 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2993 This means that the maximum local gain factor for each frame is defined
2994 (only) by the frame's highest magnitude sample. This way, the samples can
2995 be amplified as much as possible without exceeding the maximum signal
2996 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2997 Normalizer can also take into account the frame's root mean square,
2998 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2999 determine the power of a time-varying signal. It is therefore considered
3000 that the RMS is a better approximation of the "perceived loudness" than
3001 just looking at the signal's peak magnitude. Consequently, by adjusting all
3002 frames to a constant RMS value, a uniform "perceived loudness" can be
3003 established. If a target RMS value has been specified, a frame's local gain
3004 factor is defined as the factor that would result in exactly that RMS value.
3005 Note, however, that the maximum local gain factor is still restricted by the
3006 frame's highest magnitude sample, in order to prevent clipping.
3009 Enable channels coupling. By default is enabled.
3010 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3011 amount. This means the same gain factor will be applied to all channels, i.e.
3012 the maximum possible gain factor is determined by the "loudest" channel.
3013 However, in some recordings, it may happen that the volume of the different
3014 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3015 In this case, this option can be used to disable the channel coupling. This way,
3016 the gain factor will be determined independently for each channel, depending
3017 only on the individual channel's highest magnitude sample. This allows for
3018 harmonizing the volume of the different channels.
3021 Enable DC bias correction. By default is disabled.
3022 An audio signal (in the time domain) is a sequence of sample values.
3023 In the Dynamic Audio Normalizer these sample values are represented in the
3024 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3025 audio signal, or "waveform", should be centered around the zero point.
3026 That means if we calculate the mean value of all samples in a file, or in a
3027 single frame, then the result should be 0.0 or at least very close to that
3028 value. If, however, there is a significant deviation of the mean value from
3029 0.0, in either positive or negative direction, this is referred to as a
3030 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3031 Audio Normalizer provides optional DC bias correction.
3032 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3033 the mean value, or "DC correction" offset, of each input frame and subtract
3034 that value from all of the frame's sample values which ensures those samples
3035 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3036 boundaries, the DC correction offset values will be interpolated smoothly
3037 between neighbouring frames.
3040 Enable alternative boundary mode. By default is disabled.
3041 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3042 around each frame. This includes the preceding frames as well as the
3043 subsequent frames. However, for the "boundary" frames, located at the very
3044 beginning and at the very end of the audio file, not all neighbouring
3045 frames are available. In particular, for the first few frames in the audio
3046 file, the preceding frames are not known. And, similarly, for the last few
3047 frames in the audio file, the subsequent frames are not known. Thus, the
3048 question arises which gain factors should be assumed for the missing frames
3049 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3050 to deal with this situation. The default boundary mode assumes a gain factor
3051 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3052 "fade out" at the beginning and at the end of the input, respectively.
3055 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3056 By default, the Dynamic Audio Normalizer does not apply "traditional"
3057 compression. This means that signal peaks will not be pruned and thus the
3058 full dynamic range will be retained within each local neighbourhood. However,
3059 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3060 normalization algorithm with a more "traditional" compression.
3061 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3062 (thresholding) function. If (and only if) the compression feature is enabled,
3063 all input frames will be processed by a soft knee thresholding function prior
3064 to the actual normalization process. Put simply, the thresholding function is
3065 going to prune all samples whose magnitude exceeds a certain threshold value.
3066 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3067 value. Instead, the threshold value will be adjusted for each individual
3069 In general, smaller parameters result in stronger compression, and vice versa.
3070 Values below 3.0 are not recommended, because audible distortion may appear.
3075 Make audio easier to listen to on headphones.
3077 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3078 so that when listened to on headphones the stereo image is moved from
3079 inside your head (standard for headphones) to outside and in front of
3080 the listener (standard for speakers).
3086 Apply a two-pole peaking equalisation (EQ) filter. With this
3087 filter, the signal-level at and around a selected frequency can
3088 be increased or decreased, whilst (unlike bandpass and bandreject
3089 filters) that at all other frequencies is unchanged.
3091 In order to produce complex equalisation curves, this filter can
3092 be given several times, each with a different central frequency.
3094 The filter accepts the following options:
3098 Set the filter's central frequency in Hz.
3101 Set method to specify band-width of filter.
3116 Specify the band-width of a filter in width_type units.
3119 Set the required gain or attenuation in dB.
3120 Beware of clipping when using a positive gain.
3123 Specify which channels to filter, by default all available are filtered.
3126 @subsection Examples
3129 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3131 equalizer=f=1000:t=h:width=200:g=-10
3135 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3137 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3141 @subsection Commands
3143 This filter supports the following commands:
3146 Change equalizer frequency.
3147 Syntax for the command is : "@var{frequency}"
3150 Change equalizer width_type.
3151 Syntax for the command is : "@var{width_type}"
3154 Change equalizer width.
3155 Syntax for the command is : "@var{width}"
3158 Change equalizer gain.
3159 Syntax for the command is : "@var{gain}"
3162 @section extrastereo
3164 Linearly increases the difference between left and right channels which
3165 adds some sort of "live" effect to playback.
3167 The filter accepts the following options:
3171 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3172 (average of both channels), with 1.0 sound will be unchanged, with
3173 -1.0 left and right channels will be swapped.
3176 Enable clipping. By default is enabled.
3179 @section firequalizer
3180 Apply FIR Equalization using arbitrary frequency response.
3182 The filter accepts the following option:
3186 Set gain curve equation (in dB). The expression can contain variables:
3189 the evaluated frequency
3193 channel number, set to 0 when multichannels evaluation is disabled
3195 channel id, see libavutil/channel_layout.h, set to the first channel id when
3196 multichannels evaluation is disabled
3200 channel_layout, see libavutil/channel_layout.h
3205 @item gain_interpolate(f)
3206 interpolate gain on frequency f based on gain_entry
3207 @item cubic_interpolate(f)
3208 same as gain_interpolate, but smoother
3210 This option is also available as command. Default is @code{gain_interpolate(f)}.
3213 Set gain entry for gain_interpolate function. The expression can
3217 store gain entry at frequency f with value g
3219 This option is also available as command.
3222 Set filter delay in seconds. Higher value means more accurate.
3223 Default is @code{0.01}.
3226 Set filter accuracy in Hz. Lower value means more accurate.
3227 Default is @code{5}.
3230 Set window function. Acceptable values are:
3233 rectangular window, useful when gain curve is already smooth
3235 hann window (default)
3241 3-terms continuous 1st derivative nuttall window
3243 minimum 3-terms discontinuous nuttall window
3245 4-terms continuous 1st derivative nuttall window
3247 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3249 blackman-harris window
3255 If enabled, use fixed number of audio samples. This improves speed when
3256 filtering with large delay. Default is disabled.
3259 Enable multichannels evaluation on gain. Default is disabled.
3262 Enable zero phase mode by subtracting timestamp to compensate delay.
3263 Default is disabled.
3266 Set scale used by gain. Acceptable values are:
3269 linear frequency, linear gain
3271 linear frequency, logarithmic (in dB) gain (default)
3273 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3275 logarithmic frequency, logarithmic gain
3279 Set file for dumping, suitable for gnuplot.
3282 Set scale for dumpfile. Acceptable values are same with scale option.
3286 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3287 Default is disabled.
3290 Enable minimum phase impulse response. Default is disabled.
3293 @subsection Examples
3298 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3301 lowpass at 1000 Hz with gain_entry:
3303 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3306 custom equalization:
3308 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3311 higher delay with zero phase to compensate delay:
3313 firequalizer=delay=0.1:fixed=on:zero_phase=on
3316 lowpass on left channel, highpass on right channel:
3318 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3319 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3324 Apply a flanging effect to the audio.
3326 The filter accepts the following options:
3330 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3333 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3336 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3340 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3341 Default value is 71.
3344 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3347 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3348 Default value is @var{sinusoidal}.
3351 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3352 Default value is 25.
3355 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3356 Default is @var{linear}.
3360 Apply Haas effect to audio.
3362 Note that this makes most sense to apply on mono signals.
3363 With this filter applied to mono signals it give some directionality and
3364 stretches its stereo image.
3366 The filter accepts the following options:
3370 Set input level. By default is @var{1}, or 0dB
3373 Set output level. By default is @var{1}, or 0dB.
3376 Set gain applied to side part of signal. By default is @var{1}.
3379 Set kind of middle source. Can be one of the following:
3389 Pick middle part signal of stereo image.
3392 Pick side part signal of stereo image.
3396 Change middle phase. By default is disabled.
3399 Set left channel delay. By default is @var{2.05} milliseconds.
3402 Set left channel balance. By default is @var{-1}.
3405 Set left channel gain. By default is @var{1}.
3408 Change left phase. By default is disabled.
3411 Set right channel delay. By defaults is @var{2.12} milliseconds.
3414 Set right channel balance. By default is @var{1}.
3417 Set right channel gain. By default is @var{1}.
3420 Change right phase. By default is enabled.
3425 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3426 embedded HDCD codes is expanded into a 20-bit PCM stream.
3428 The filter supports the Peak Extend and Low-level Gain Adjustment features
3429 of HDCD, and detects the Transient Filter flag.
3432 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3435 When using the filter with wav, note the default encoding for wav is 16-bit,
3436 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3437 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3439 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3440 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3443 The filter accepts the following options:
3446 @item disable_autoconvert
3447 Disable any automatic format conversion or resampling in the filter graph.
3449 @item process_stereo
3450 Process the stereo channels together. If target_gain does not match between
3451 channels, consider it invalid and use the last valid target_gain.
3454 Set the code detect timer period in ms.
3457 Always extend peaks above -3dBFS even if PE isn't signaled.
3460 Replace audio with a solid tone and adjust the amplitude to signal some
3461 specific aspect of the decoding process. The output file can be loaded in
3462 an audio editor alongside the original to aid analysis.
3464 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3471 Gain adjustment level at each sample
3473 Samples where peak extend occurs
3475 Samples where the code detect timer is active
3477 Samples where the target gain does not match between channels
3483 Apply head-related transfer functions (HRTFs) to create virtual
3484 loudspeakers around the user for binaural listening via headphones.
3485 The HRIRs are provided via additional streams, for each channel
3486 one stereo input stream is needed.
3488 The filter accepts the following options:
3492 Set mapping of input streams for convolution.
3493 The argument is a '|'-separated list of channel names in order as they
3494 are given as additional stream inputs for filter.
3495 This also specify number of input streams. Number of input streams
3496 must be not less than number of channels in first stream plus one.
3499 Set gain applied to audio. Value is in dB. Default is 0.
3502 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3503 processing audio in time domain which is slow.
3504 @var{freq} is processing audio in frequency domain which is fast.
3505 Default is @var{freq}.
3508 Set custom gain for LFE channels. Value is in dB. Default is 0.
3511 Set size of frame in number of samples which will be processed at once.
3512 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3515 Set format of hrir stream.
3516 Default value is @var{stereo}. Alternative value is @var{multich}.
3517 If value is set to @var{stereo}, number of additional streams should
3518 be greater or equal to number of input channels in first input stream.
3519 Also each additional stream should have stereo number of channels.
3520 If value is set to @var{multich}, number of additional streams should
3521 be exactly one. Also number of input channels of additional stream
3522 should be equal or greater than twice number of channels of first input
3526 @subsection Examples
3530 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3531 each amovie filter use stereo file with IR coefficients as input.
3532 The files give coefficients for each position of virtual loudspeaker:
3534 ffmpeg -i input.wav -lavfi-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],[a:0][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3539 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3540 but now in @var{multich} @var{hrir} format.
3542 ffmpeg -i input.wav -lavfi-complex "amovie=minp.wav[hrirs],[a:0][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3549 Apply a high-pass filter with 3dB point frequency.
3550 The filter can be either single-pole, or double-pole (the default).
3551 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3553 The filter accepts the following options:
3557 Set frequency in Hz. Default is 3000.
3560 Set number of poles. Default is 2.
3563 Set method to specify band-width of filter.
3578 Specify the band-width of a filter in width_type units.
3579 Applies only to double-pole filter.
3580 The default is 0.707q and gives a Butterworth response.
3583 Specify which channels to filter, by default all available are filtered.
3586 @subsection Commands
3588 This filter supports the following commands:
3591 Change highpass frequency.
3592 Syntax for the command is : "@var{frequency}"
3595 Change highpass width_type.
3596 Syntax for the command is : "@var{width_type}"
3599 Change highpass width.
3600 Syntax for the command is : "@var{width}"
3605 Join multiple input streams into one multi-channel stream.
3607 It accepts the following parameters:
3611 The number of input streams. It defaults to 2.
3613 @item channel_layout
3614 The desired output channel layout. It defaults to stereo.
3617 Map channels from inputs to output. The argument is a '|'-separated list of
3618 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3619 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3620 can be either the name of the input channel (e.g. FL for front left) or its
3621 index in the specified input stream. @var{out_channel} is the name of the output
3625 The filter will attempt to guess the mappings when they are not specified
3626 explicitly. It does so by first trying to find an unused matching input channel
3627 and if that fails it picks the first unused input channel.
3629 Join 3 inputs (with properly set channel layouts):
3631 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3634 Build a 5.1 output from 6 single-channel streams:
3636 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3637 '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'
3643 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3645 To enable compilation of this filter you need to configure FFmpeg with
3646 @code{--enable-ladspa}.
3650 Specifies the name of LADSPA plugin library to load. If the environment
3651 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3652 each one of the directories specified by the colon separated list in
3653 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3654 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3655 @file{/usr/lib/ladspa/}.
3658 Specifies the plugin within the library. Some libraries contain only
3659 one plugin, but others contain many of them. If this is not set filter
3660 will list all available plugins within the specified library.
3663 Set the '|' separated list of controls which are zero or more floating point
3664 values that determine the behavior of the loaded plugin (for example delay,
3666 Controls need to be defined using the following syntax:
3667 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3668 @var{valuei} is the value set on the @var{i}-th control.
3669 Alternatively they can be also defined using the following syntax:
3670 @var{value0}|@var{value1}|@var{value2}|..., where
3671 @var{valuei} is the value set on the @var{i}-th control.
3672 If @option{controls} is set to @code{help}, all available controls and
3673 their valid ranges are printed.
3675 @item sample_rate, s
3676 Specify the sample rate, default to 44100. Only used if plugin have
3680 Set the number of samples per channel per each output frame, default
3681 is 1024. Only used if plugin have zero inputs.
3684 Set the minimum duration of the sourced audio. See
3685 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3686 for the accepted syntax.
3687 Note that the resulting duration may be greater than the specified duration,
3688 as the generated audio is always cut at the end of a complete frame.
3689 If not specified, or the expressed duration is negative, the audio is
3690 supposed to be generated forever.
3691 Only used if plugin have zero inputs.
3695 @subsection Examples
3699 List all available plugins within amp (LADSPA example plugin) library:
3705 List all available controls and their valid ranges for @code{vcf_notch}
3706 plugin from @code{VCF} library:
3708 ladspa=f=vcf:p=vcf_notch:c=help
3712 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3715 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3719 Add reverberation to the audio using TAP-plugins
3720 (Tom's Audio Processing plugins):
3722 ladspa=file=tap_reverb:tap_reverb
3726 Generate white noise, with 0.2 amplitude:
3728 ladspa=file=cmt:noise_source_white:c=c0=.2
3732 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3733 @code{C* Audio Plugin Suite} (CAPS) library:
3735 ladspa=file=caps:Click:c=c1=20'
3739 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3741 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3745 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3746 @code{SWH Plugins} collection:
3748 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3752 Attenuate low frequencies using Multiband EQ from Steve Harris
3753 @code{SWH Plugins} collection:
3755 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3759 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3762 ladspa=caps:Narrower
3766 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3768 ladspa=caps:White:.2
3772 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3774 ladspa=caps:Fractal:c=c1=1
3778 Dynamic volume normalization using @code{VLevel} plugin:
3780 ladspa=vlevel-ladspa:vlevel_mono
3784 @subsection Commands
3786 This filter supports the following commands:
3789 Modify the @var{N}-th control value.
3791 If the specified value is not valid, it is ignored and prior one is kept.
3796 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3797 Support for both single pass (livestreams, files) and double pass (files) modes.
3798 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3799 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3800 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3802 The filter accepts the following options:
3806 Set integrated loudness target.
3807 Range is -70.0 - -5.0. Default value is -24.0.
3810 Set loudness range target.
3811 Range is 1.0 - 20.0. Default value is 7.0.
3814 Set maximum true peak.
3815 Range is -9.0 - +0.0. Default value is -2.0.
3817 @item measured_I, measured_i
3818 Measured IL of input file.
3819 Range is -99.0 - +0.0.
3821 @item measured_LRA, measured_lra
3822 Measured LRA of input file.
3823 Range is 0.0 - 99.0.
3825 @item measured_TP, measured_tp
3826 Measured true peak of input file.
3827 Range is -99.0 - +99.0.
3829 @item measured_thresh
3830 Measured threshold of input file.
3831 Range is -99.0 - +0.0.
3834 Set offset gain. Gain is applied before the true-peak limiter.
3835 Range is -99.0 - +99.0. Default is +0.0.
3838 Normalize linearly if possible.
3839 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3840 to be specified in order to use this mode.
3841 Options are true or false. Default is true.
3844 Treat mono input files as "dual-mono". If a mono file is intended for playback
3845 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3846 If set to @code{true}, this option will compensate for this effect.
3847 Multi-channel input files are not affected by this option.
3848 Options are true or false. Default is false.
3851 Set print format for stats. Options are summary, json, or none.
3852 Default value is none.
3857 Apply a low-pass filter with 3dB point frequency.
3858 The filter can be either single-pole or double-pole (the default).
3859 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3861 The filter accepts the following options:
3865 Set frequency in Hz. Default is 500.
3868 Set number of poles. Default is 2.
3871 Set method to specify band-width of filter.
3886 Specify the band-width of a filter in width_type units.
3887 Applies only to double-pole filter.
3888 The default is 0.707q and gives a Butterworth response.
3891 Specify which channels to filter, by default all available are filtered.
3894 @subsection Examples
3897 Lowpass only LFE channel, it LFE is not present it does nothing:
3903 @subsection Commands
3905 This filter supports the following commands:
3908 Change lowpass frequency.
3909 Syntax for the command is : "@var{frequency}"
3912 Change lowpass width_type.
3913 Syntax for the command is : "@var{width_type}"
3916 Change lowpass width.
3917 Syntax for the command is : "@var{width}"
3922 Load a LV2 (LADSPA Version 2) plugin.
3924 To enable compilation of this filter you need to configure FFmpeg with
3925 @code{--enable-lv2}.
3929 Specifies the plugin URI. You may need to escape ':'.
3932 Set the '|' separated list of controls which are zero or more floating point
3933 values that determine the behavior of the loaded plugin (for example delay,
3935 If @option{controls} is set to @code{help}, all available controls and
3936 their valid ranges are printed.
3938 @item sample_rate, s
3939 Specify the sample rate, default to 44100. Only used if plugin have
3943 Set the number of samples per channel per each output frame, default
3944 is 1024. Only used if plugin have zero inputs.
3947 Set the minimum duration of the sourced audio. See
3948 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3949 for the accepted syntax.
3950 Note that the resulting duration may be greater than the specified duration,
3951 as the generated audio is always cut at the end of a complete frame.
3952 If not specified, or the expressed duration is negative, the audio is
3953 supposed to be generated forever.
3954 Only used if plugin have zero inputs.
3957 @subsection Examples
3961 Apply bass enhancer plugin from Calf:
3963 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
3967 Apply vinyl plugin from Calf:
3969 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
3973 Apply bit crusher plugin from ArtyFX:
3975 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
3980 Multiband Compress or expand the audio's dynamic range.
3982 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
3983 This is akin to the crossover of a loudspeaker, and results in flat frequency
3984 response when absent compander action.
3986 It accepts the following parameters:
3990 This option syntax is:
3991 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
3992 For explanation of each item refer to compand filter documentation.
3998 Mix channels with specific gain levels. The filter accepts the output
3999 channel layout followed by a set of channels definitions.
4001 This filter is also designed to efficiently remap the channels of an audio
4004 The filter accepts parameters of the form:
4005 "@var{l}|@var{outdef}|@var{outdef}|..."
4009 output channel layout or number of channels
4012 output channel specification, of the form:
4013 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4016 output channel to define, either a channel name (FL, FR, etc.) or a channel
4017 number (c0, c1, etc.)
4020 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4023 input channel to use, see out_name for details; it is not possible to mix
4024 named and numbered input channels
4027 If the `=' in a channel specification is replaced by `<', then the gains for
4028 that specification will be renormalized so that the total is 1, thus
4029 avoiding clipping noise.
4031 @subsection Mixing examples
4033 For example, if you want to down-mix from stereo to mono, but with a bigger
4034 factor for the left channel:
4036 pan=1c|c0=0.9*c0+0.1*c1
4039 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4040 7-channels surround:
4042 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4045 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4046 that should be preferred (see "-ac" option) unless you have very specific
4049 @subsection Remapping examples
4051 The channel remapping will be effective if, and only if:
4054 @item gain coefficients are zeroes or ones,
4055 @item only one input per channel output,
4058 If all these conditions are satisfied, the filter will notify the user ("Pure
4059 channel mapping detected"), and use an optimized and lossless method to do the
4062 For example, if you have a 5.1 source and want a stereo audio stream by
4063 dropping the extra channels:
4065 pan="stereo| c0=FL | c1=FR"
4068 Given the same source, you can also switch front left and front right channels
4069 and keep the input channel layout:
4071 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4074 If the input is a stereo audio stream, you can mute the front left channel (and
4075 still keep the stereo channel layout) with:
4080 Still with a stereo audio stream input, you can copy the right channel in both
4081 front left and right:
4083 pan="stereo| c0=FR | c1=FR"
4088 ReplayGain scanner filter. This filter takes an audio stream as an input and
4089 outputs it unchanged.
4090 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4094 Convert the audio sample format, sample rate and channel layout. It is
4095 not meant to be used directly.
4098 Apply time-stretching and pitch-shifting with librubberband.
4100 To enable compilation of this filter, you need to configure FFmpeg with
4101 @code{--enable-librubberband}.
4103 The filter accepts the following options:
4107 Set tempo scale factor.
4110 Set pitch scale factor.
4113 Set transients detector.
4114 Possible values are:
4123 Possible values are:
4132 Possible values are:
4139 Set processing window size.
4140 Possible values are:
4149 Possible values are:
4156 Enable formant preservation when shift pitching.
4157 Possible values are:
4165 Possible values are:
4174 Possible values are:
4181 @section sidechaincompress
4183 This filter acts like normal compressor but has the ability to compress
4184 detected signal using second input signal.
4185 It needs two input streams and returns one output stream.
4186 First input stream will be processed depending on second stream signal.
4187 The filtered signal then can be filtered with other filters in later stages of
4188 processing. See @ref{pan} and @ref{amerge} filter.
4190 The filter accepts the following options:
4194 Set input gain. Default is 1. Range is between 0.015625 and 64.
4197 If a signal of second stream raises above this level it will affect the gain
4198 reduction of first stream.
4199 By default is 0.125. Range is between 0.00097563 and 1.
4202 Set a ratio about which the signal is reduced. 1:2 means that if the level
4203 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4204 Default is 2. Range is between 1 and 20.
4207 Amount of milliseconds the signal has to rise above the threshold before gain
4208 reduction starts. Default is 20. Range is between 0.01 and 2000.
4211 Amount of milliseconds the signal has to fall below the threshold before
4212 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4215 Set the amount by how much signal will be amplified after processing.
4216 Default is 1. Range is from 1 to 64.
4219 Curve the sharp knee around the threshold to enter gain reduction more softly.
4220 Default is 2.82843. Range is between 1 and 8.
4223 Choose if the @code{average} level between all channels of side-chain stream
4224 or the louder(@code{maximum}) channel of side-chain stream affects the
4225 reduction. Default is @code{average}.
4228 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4229 of @code{rms}. Default is @code{rms} which is mainly smoother.
4232 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4235 How much to use compressed signal in output. Default is 1.
4236 Range is between 0 and 1.
4239 @subsection Examples
4243 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4244 depending on the signal of 2nd input and later compressed signal to be
4245 merged with 2nd input:
4247 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4251 @section sidechaingate
4253 A sidechain gate acts like a normal (wideband) gate but has the ability to
4254 filter the detected signal before sending it to the gain reduction stage.
4255 Normally a gate uses the full range signal to detect a level above the
4257 For example: If you cut all lower frequencies from your sidechain signal
4258 the gate will decrease the volume of your track only if not enough highs
4259 appear. With this technique you are able to reduce the resonation of a
4260 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4262 It needs two input streams and returns one output stream.
4263 First input stream will be processed depending on second stream signal.
4265 The filter accepts the following options:
4269 Set input level before filtering.
4270 Default is 1. Allowed range is from 0.015625 to 64.
4273 Set the level of gain reduction when the signal is below the threshold.
4274 Default is 0.06125. Allowed range is from 0 to 1.
4277 If a signal rises above this level the gain reduction is released.
4278 Default is 0.125. Allowed range is from 0 to 1.
4281 Set a ratio about which the signal is reduced.
4282 Default is 2. Allowed range is from 1 to 9000.
4285 Amount of milliseconds the signal has to rise above the threshold before gain
4287 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4290 Amount of milliseconds the signal has to fall below the threshold before the
4291 reduction is increased again. Default is 250 milliseconds.
4292 Allowed range is from 0.01 to 9000.
4295 Set amount of amplification of signal after processing.
4296 Default is 1. Allowed range is from 1 to 64.
4299 Curve the sharp knee around the threshold to enter gain reduction more softly.
4300 Default is 2.828427125. Allowed range is from 1 to 8.
4303 Choose if exact signal should be taken for detection or an RMS like one.
4304 Default is rms. Can be peak or rms.
4307 Choose if the average level between all channels or the louder channel affects
4309 Default is average. Can be average or maximum.
4312 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4315 @section silencedetect
4317 Detect silence in an audio stream.
4319 This filter logs a message when it detects that the input audio volume is less
4320 or equal to a noise tolerance value for a duration greater or equal to the
4321 minimum detected noise duration.
4323 The printed times and duration are expressed in seconds.
4325 The filter accepts the following options:
4329 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4330 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4333 Set silence duration until notification (default is 2 seconds).
4336 Process each channel separately, instead of combined. By default is disabled.
4339 @subsection Examples
4343 Detect 5 seconds of silence with -50dB noise tolerance:
4345 silencedetect=n=-50dB:d=5
4349 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4350 tolerance in @file{silence.mp3}:
4352 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4356 @section silenceremove
4358 Remove silence from the beginning, middle or end of the audio.
4360 The filter accepts the following options:
4364 This value is used to indicate if audio should be trimmed at beginning of
4365 the audio. A value of zero indicates no silence should be trimmed from the
4366 beginning. When specifying a non-zero value, it trims audio up until it
4367 finds non-silence. Normally, when trimming silence from beginning of audio
4368 the @var{start_periods} will be @code{1} but it can be increased to higher
4369 values to trim all audio up to specific count of non-silence periods.
4370 Default value is @code{0}.
4372 @item start_duration
4373 Specify the amount of time that non-silence must be detected before it stops
4374 trimming audio. By increasing the duration, bursts of noises can be treated
4375 as silence and trimmed off. Default value is @code{0}.
4377 @item start_threshold
4378 This indicates what sample value should be treated as silence. For digital
4379 audio, a value of @code{0} may be fine but for audio recorded from analog,
4380 you may wish to increase the value to account for background noise.
4381 Can be specified in dB (in case "dB" is appended to the specified value)
4382 or amplitude ratio. Default value is @code{0}.
4385 Specify max duration of silence at beginning that will be kept after
4386 trimming. Default is 0, which is equal to trimming all samples detected
4390 Specify mode of detection of silence end in start of multi-channel audio.
4391 Can be @var{any} or @var{all}. Default is @var{any}.
4392 With @var{any}, any sample that is detected as non-silence will cause
4393 stopped trimming of silence.
4394 With @var{all}, only if all channels are detected as non-silence will cause
4395 stopped trimming of silence.
4398 Set the count for trimming silence from the end of audio.
4399 To remove silence from the middle of a file, specify a @var{stop_periods}
4400 that is negative. This value is then treated as a positive value and is
4401 used to indicate the effect should restart processing as specified by
4402 @var{start_periods}, making it suitable for removing periods of silence
4403 in the middle of the audio.
4404 Default value is @code{0}.
4407 Specify a duration of silence that must exist before audio is not copied any
4408 more. By specifying a higher duration, silence that is wanted can be left in
4410 Default value is @code{0}.
4412 @item stop_threshold
4413 This is the same as @option{start_threshold} but for trimming silence from
4415 Can be specified in dB (in case "dB" is appended to the specified value)
4416 or amplitude ratio. Default value is @code{0}.
4419 Specify max duration of silence at end that will be kept after
4420 trimming. Default is 0, which is equal to trimming all samples detected
4424 Specify mode of detection of silence start in end of multi-channel audio.
4425 Can be @var{any} or @var{all}. Default is @var{any}.
4426 With @var{any}, any sample that is detected as non-silence will cause
4427 stopped trimming of silence.
4428 With @var{all}, only if all channels are detected as non-silence will cause
4429 stopped trimming of silence.
4432 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4433 and works better with digital silence which is exactly 0.
4434 Default value is @code{rms}.
4437 Set duration in number of seconds used to calculate size of window in number
4438 of samples for detecting silence.
4439 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4442 @subsection Examples
4446 The following example shows how this filter can be used to start a recording
4447 that does not contain the delay at the start which usually occurs between
4448 pressing the record button and the start of the performance:
4450 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4454 Trim all silence encountered from beginning to end where there is more than 1
4455 second of silence in audio:
4457 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4463 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4464 loudspeakers around the user for binaural listening via headphones (audio
4465 formats up to 9 channels supported).
4466 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4467 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4468 Austrian Academy of Sciences.
4470 To enable compilation of this filter you need to configure FFmpeg with
4471 @code{--enable-libmysofa}.
4473 The filter accepts the following options:
4477 Set the SOFA file used for rendering.
4480 Set gain applied to audio. Value is in dB. Default is 0.
4483 Set rotation of virtual loudspeakers in deg. Default is 0.
4486 Set elevation of virtual speakers in deg. Default is 0.
4489 Set distance in meters between loudspeakers and the listener with near-field
4490 HRTFs. Default is 1.
4493 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4494 processing audio in time domain which is slow.
4495 @var{freq} is processing audio in frequency domain which is fast.
4496 Default is @var{freq}.
4499 Set custom positions of virtual loudspeakers. Syntax for this option is:
4500 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4501 Each virtual loudspeaker is described with short channel name following with
4502 azimuth and elevation in degrees.
4503 Each virtual loudspeaker description is separated by '|'.
4504 For example to override front left and front right channel positions use:
4505 'speakers=FL 45 15|FR 345 15'.
4506 Descriptions with unrecognised channel names are ignored.
4509 Set custom gain for LFE channels. Value is in dB. Default is 0.
4512 Set custom frame size in number of samples. Default is 1024.
4513 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4514 is set to @var{freq}.
4517 Should all IRs be normalized upon importing SOFA file.
4518 By default is enabled.
4521 Should nearest IRs be interpolated with neighbor IRs if exact position
4522 does not match. By default is disabled.
4525 Minphase all IRs upon loading of SOFA file. By default is disabled.
4528 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4531 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4534 @subsection Examples
4538 Using ClubFritz6 sofa file:
4540 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4544 Using ClubFritz12 sofa file and bigger radius with small rotation:
4546 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4550 Similar as above but with custom speaker positions for front left, front right, back left and back right
4551 and also with custom gain:
4553 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4557 @section stereotools
4559 This filter has some handy utilities to manage stereo signals, for converting
4560 M/S stereo recordings to L/R signal while having control over the parameters
4561 or spreading the stereo image of master track.
4563 The filter accepts the following options:
4567 Set input level before filtering for both channels. Defaults is 1.
4568 Allowed range is from 0.015625 to 64.
4571 Set output level after filtering for both channels. Defaults is 1.
4572 Allowed range is from 0.015625 to 64.
4575 Set input balance between both channels. Default is 0.
4576 Allowed range is from -1 to 1.
4579 Set output balance between both channels. Default is 0.
4580 Allowed range is from -1 to 1.
4583 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4584 clipping. Disabled by default.
4587 Mute the left channel. Disabled by default.
4590 Mute the right channel. Disabled by default.
4593 Change the phase of the left channel. Disabled by default.
4596 Change the phase of the right channel. Disabled by default.
4599 Set stereo mode. Available values are:
4603 Left/Right to Left/Right, this is default.
4606 Left/Right to Mid/Side.
4609 Mid/Side to Left/Right.
4612 Left/Right to Left/Left.
4615 Left/Right to Right/Right.
4618 Left/Right to Left + Right.
4621 Left/Right to Right/Left.
4624 Mid/Side to Left/Left.
4627 Mid/Side to Right/Right.
4631 Set level of side signal. Default is 1.
4632 Allowed range is from 0.015625 to 64.
4635 Set balance of side signal. Default is 0.
4636 Allowed range is from -1 to 1.
4639 Set level of the middle signal. Default is 1.
4640 Allowed range is from 0.015625 to 64.
4643 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4646 Set stereo base between mono and inversed channels. Default is 0.
4647 Allowed range is from -1 to 1.
4650 Set delay in milliseconds how much to delay left from right channel and
4651 vice versa. Default is 0. Allowed range is from -20 to 20.
4654 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4657 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4659 @item bmode_in, bmode_out
4660 Set balance mode for balance_in/balance_out option.
4662 Can be one of the following:
4666 Classic balance mode. Attenuate one channel at time.
4667 Gain is raised up to 1.
4670 Similar as classic mode above but gain is raised up to 2.
4673 Equal power distribution, from -6dB to +6dB range.
4677 @subsection Examples
4681 Apply karaoke like effect:
4683 stereotools=mlev=0.015625
4687 Convert M/S signal to L/R:
4689 "stereotools=mode=ms>lr"
4693 @section stereowiden
4695 This filter enhance the stereo effect by suppressing signal common to both
4696 channels and by delaying the signal of left into right and vice versa,
4697 thereby widening the stereo effect.
4699 The filter accepts the following options:
4703 Time in milliseconds of the delay of left signal into right and vice versa.
4704 Default is 20 milliseconds.
4707 Amount of gain in delayed signal into right and vice versa. Gives a delay
4708 effect of left signal in right output and vice versa which gives widening
4709 effect. Default is 0.3.
4712 Cross feed of left into right with inverted phase. This helps in suppressing
4713 the mono. If the value is 1 it will cancel all the signal common to both
4714 channels. Default is 0.3.
4717 Set level of input signal of original channel. Default is 0.8.
4720 @section superequalizer
4721 Apply 18 band equalizer.
4723 The filter accepts the following options:
4730 Set 131Hz band gain.
4732 Set 185Hz band gain.
4734 Set 262Hz band gain.
4736 Set 370Hz band gain.
4738 Set 523Hz band gain.
4740 Set 740Hz band gain.
4742 Set 1047Hz band gain.
4744 Set 1480Hz band gain.
4746 Set 2093Hz band gain.
4748 Set 2960Hz band gain.
4750 Set 4186Hz band gain.
4752 Set 5920Hz band gain.
4754 Set 8372Hz band gain.
4756 Set 11840Hz band gain.
4758 Set 16744Hz band gain.
4760 Set 20000Hz band gain.
4764 Apply audio surround upmix filter.
4766 This filter allows to produce multichannel output from audio stream.
4768 The filter accepts the following options:
4772 Set output channel layout. By default, this is @var{5.1}.
4774 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4775 for the required syntax.
4778 Set input channel layout. By default, this is @var{stereo}.
4780 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4781 for the required syntax.
4784 Set input volume level. By default, this is @var{1}.
4787 Set output volume level. By default, this is @var{1}.
4790 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4793 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4796 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4799 Set front center input volume. By default, this is @var{1}.
4802 Set front center output volume. By default, this is @var{1}.
4805 Set LFE input volume. By default, this is @var{1}.
4808 Set LFE output volume. By default, this is @var{1}.
4811 @section treble, highshelf
4813 Boost or cut treble (upper) frequencies of the audio using a two-pole
4814 shelving filter with a response similar to that of a standard
4815 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4817 The filter accepts the following options:
4821 Give the gain at whichever is the lower of ~22 kHz and the
4822 Nyquist frequency. Its useful range is about -20 (for a large cut)
4823 to +20 (for a large boost). Beware of clipping when using a positive gain.
4826 Set the filter's central frequency and so can be used
4827 to extend or reduce the frequency range to be boosted or cut.
4828 The default value is @code{3000} Hz.
4831 Set method to specify band-width of filter.
4846 Determine how steep is the filter's shelf transition.
4849 Specify which channels to filter, by default all available are filtered.
4852 @subsection Commands
4854 This filter supports the following commands:
4857 Change treble frequency.
4858 Syntax for the command is : "@var{frequency}"
4861 Change treble width_type.
4862 Syntax for the command is : "@var{width_type}"
4865 Change treble width.
4866 Syntax for the command is : "@var{width}"
4870 Syntax for the command is : "@var{gain}"
4875 Sinusoidal amplitude modulation.
4877 The filter accepts the following options:
4881 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4882 (20 Hz or lower) will result in a tremolo effect.
4883 This filter may also be used as a ring modulator by specifying
4884 a modulation frequency higher than 20 Hz.
4885 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4888 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4889 Default value is 0.5.
4894 Sinusoidal phase modulation.
4896 The filter accepts the following options:
4900 Modulation frequency in Hertz.
4901 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4904 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4905 Default value is 0.5.
4910 Adjust the input audio volume.
4912 It accepts the following parameters:
4916 Set audio volume expression.
4918 Output values are clipped to the maximum value.
4920 The output audio volume is given by the relation:
4922 @var{output_volume} = @var{volume} * @var{input_volume}
4925 The default value for @var{volume} is "1.0".
4928 This parameter represents the mathematical precision.
4930 It determines which input sample formats will be allowed, which affects the
4931 precision of the volume scaling.
4935 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4937 32-bit floating-point; this limits input sample format to FLT. (default)
4939 64-bit floating-point; this limits input sample format to DBL.
4943 Choose the behaviour on encountering ReplayGain side data in input frames.
4947 Remove ReplayGain side data, ignoring its contents (the default).
4950 Ignore ReplayGain side data, but leave it in the frame.
4953 Prefer the track gain, if present.
4956 Prefer the album gain, if present.
4959 @item replaygain_preamp
4960 Pre-amplification gain in dB to apply to the selected replaygain gain.
4962 Default value for @var{replaygain_preamp} is 0.0.
4965 Set when the volume expression is evaluated.
4967 It accepts the following values:
4970 only evaluate expression once during the filter initialization, or
4971 when the @samp{volume} command is sent
4974 evaluate expression for each incoming frame
4977 Default value is @samp{once}.
4980 The volume expression can contain the following parameters.
4984 frame number (starting at zero)
4987 @item nb_consumed_samples
4988 number of samples consumed by the filter
4990 number of samples in the current frame
4992 original frame position in the file
4998 PTS at start of stream
5000 time at start of stream
5006 last set volume value
5009 Note that when @option{eval} is set to @samp{once} only the
5010 @var{sample_rate} and @var{tb} variables are available, all other
5011 variables will evaluate to NAN.
5013 @subsection Commands
5015 This filter supports the following commands:
5018 Modify the volume expression.
5019 The command accepts the same syntax of the corresponding option.
5021 If the specified expression is not valid, it is kept at its current
5023 @item replaygain_noclip
5024 Prevent clipping by limiting the gain applied.
5026 Default value for @var{replaygain_noclip} is 1.
5030 @subsection Examples
5034 Halve the input audio volume:
5038 volume=volume=-6.0206dB
5041 In all the above example the named key for @option{volume} can be
5042 omitted, for example like in:
5048 Increase input audio power by 6 decibels using fixed-point precision:
5050 volume=volume=6dB:precision=fixed
5054 Fade volume after time 10 with an annihilation period of 5 seconds:
5056 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5060 @section volumedetect
5062 Detect the volume of the input video.
5064 The filter has no parameters. The input is not modified. Statistics about
5065 the volume will be printed in the log when the input stream end is reached.
5067 In particular it will show the mean volume (root mean square), maximum
5068 volume (on a per-sample basis), and the beginning of a histogram of the
5069 registered volume values (from the maximum value to a cumulated 1/1000 of
5072 All volumes are in decibels relative to the maximum PCM value.
5074 @subsection Examples
5076 Here is an excerpt of the output:
5078 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5079 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5080 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5081 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5082 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5083 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5084 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5085 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5086 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5092 The mean square energy is approximately -27 dB, or 10^-2.7.
5094 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5096 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5099 In other words, raising the volume by +4 dB does not cause any clipping,
5100 raising it by +5 dB causes clipping for 6 samples, etc.
5102 @c man end AUDIO FILTERS
5104 @chapter Audio Sources
5105 @c man begin AUDIO SOURCES
5107 Below is a description of the currently available audio sources.
5111 Buffer audio frames, and make them available to the filter chain.
5113 This source is mainly intended for a programmatic use, in particular
5114 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5116 It accepts the following parameters:
5120 The timebase which will be used for timestamps of submitted frames. It must be
5121 either a floating-point number or in @var{numerator}/@var{denominator} form.
5124 The sample rate of the incoming audio buffers.
5127 The sample format of the incoming audio buffers.
5128 Either a sample format name or its corresponding integer representation from
5129 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5131 @item channel_layout
5132 The channel layout of the incoming audio buffers.
5133 Either a channel layout name from channel_layout_map in
5134 @file{libavutil/channel_layout.c} or its corresponding integer representation
5135 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5138 The number of channels of the incoming audio buffers.
5139 If both @var{channels} and @var{channel_layout} are specified, then they
5144 @subsection Examples
5147 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5150 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5151 Since the sample format with name "s16p" corresponds to the number
5152 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5155 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5160 Generate an audio signal specified by an expression.
5162 This source accepts in input one or more expressions (one for each
5163 channel), which are evaluated and used to generate a corresponding
5166 This source accepts the following options:
5170 Set the '|'-separated expressions list for each separate channel. In case the
5171 @option{channel_layout} option is not specified, the selected channel layout
5172 depends on the number of provided expressions. Otherwise the last
5173 specified expression is applied to the remaining output channels.
5175 @item channel_layout, c
5176 Set the channel layout. The number of channels in the specified layout
5177 must be equal to the number of specified expressions.
5180 Set the minimum duration of the sourced audio. See
5181 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5182 for the accepted syntax.
5183 Note that the resulting duration may be greater than the specified
5184 duration, as the generated audio is always cut at the end of a
5187 If not specified, or the expressed duration is negative, the audio is
5188 supposed to be generated forever.
5191 Set the number of samples per channel per each output frame,
5194 @item sample_rate, s
5195 Specify the sample rate, default to 44100.
5198 Each expression in @var{exprs} can contain the following constants:
5202 number of the evaluated sample, starting from 0
5205 time of the evaluated sample expressed in seconds, starting from 0
5212 @subsection Examples
5222 Generate a sin signal with frequency of 440 Hz, set sample rate to
5225 aevalsrc="sin(440*2*PI*t):s=8000"
5229 Generate a two channels signal, specify the channel layout (Front
5230 Center + Back Center) explicitly:
5232 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5236 Generate white noise:
5238 aevalsrc="-2+random(0)"
5242 Generate an amplitude modulated signal:
5244 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5248 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5250 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5257 The null audio source, return unprocessed audio frames. It is mainly useful
5258 as a template and to be employed in analysis / debugging tools, or as
5259 the source for filters which ignore the input data (for example the sox
5262 This source accepts the following options:
5266 @item channel_layout, cl
5268 Specifies the channel layout, and can be either an integer or a string
5269 representing a channel layout. The default value of @var{channel_layout}
5272 Check the channel_layout_map definition in
5273 @file{libavutil/channel_layout.c} for the mapping between strings and
5274 channel layout values.
5276 @item sample_rate, r
5277 Specifies the sample rate, and defaults to 44100.
5280 Set the number of samples per requested frames.
5284 @subsection Examples
5288 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5290 anullsrc=r=48000:cl=4
5294 Do the same operation with a more obvious syntax:
5296 anullsrc=r=48000:cl=mono
5300 All the parameters need to be explicitly defined.
5304 Synthesize a voice utterance using the libflite library.
5306 To enable compilation of this filter you need to configure FFmpeg with
5307 @code{--enable-libflite}.
5309 Note that versions of the flite library prior to 2.0 are not thread-safe.
5311 The filter accepts the following options:
5316 If set to 1, list the names of the available voices and exit
5317 immediately. Default value is 0.
5320 Set the maximum number of samples per frame. Default value is 512.
5323 Set the filename containing the text to speak.
5326 Set the text to speak.
5329 Set the voice to use for the speech synthesis. Default value is
5330 @code{kal}. See also the @var{list_voices} option.
5333 @subsection Examples
5337 Read from file @file{speech.txt}, and synthesize the text using the
5338 standard flite voice:
5340 flite=textfile=speech.txt
5344 Read the specified text selecting the @code{slt} voice:
5346 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5350 Input text to ffmpeg:
5352 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5356 Make @file{ffplay} speak the specified text, using @code{flite} and
5357 the @code{lavfi} device:
5359 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5363 For more information about libflite, check:
5364 @url{http://www.festvox.org/flite/}
5368 Generate a noise audio signal.
5370 The filter accepts the following options:
5373 @item sample_rate, r
5374 Specify the sample rate. Default value is 48000 Hz.
5377 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5381 Specify the duration of the generated audio stream. Not specifying this option
5382 results in noise with an infinite length.
5384 @item color, colour, c
5385 Specify the color of noise. Available noise colors are white, pink, brown,
5386 blue and violet. Default color is white.
5389 Specify a value used to seed the PRNG.
5392 Set the number of samples per each output frame, default is 1024.
5395 @subsection Examples
5400 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5402 anoisesrc=d=60:c=pink:r=44100:a=0.5
5408 Generate odd-tap Hilbert transform FIR coefficients.
5410 The resulting stream can be used with @ref{afir} filter for phase-shifting
5411 the signal by 90 degrees.
5413 This is used in many matrix coding schemes and for analytic signal generation.
5414 The process is often written as a multiplication by i (or j), the imaginary unit.
5416 The filter accepts the following options:
5420 @item sample_rate, s
5421 Set sample rate, default is 44100.
5424 Set length of FIR filter, default is 22051.
5427 Set number of samples per each frame.
5430 Set window function to be used when generating FIR coefficients.
5435 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5437 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5439 The filter accepts the following options:
5442 @item sample_rate, r
5443 Set sample rate, default is 44100.
5446 Set number of samples per each frame. Default is 1024.
5449 Set high-pass frequency. Default is 0.
5452 Set low-pass frequency. Default is 0.
5453 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5454 is higher than 0 then filter will create band-pass filter coefficients,
5455 otherwise band-reject filter coefficients.
5458 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5461 Set Kaiser window beta.
5464 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5467 Enable rounding, by default is disabled.
5470 Set number of taps for high-pass filter.
5473 Set number of taps for low-pass filter.
5478 Generate an audio signal made of a sine wave with amplitude 1/8.
5480 The audio signal is bit-exact.
5482 The filter accepts the following options:
5487 Set the carrier frequency. Default is 440 Hz.
5489 @item beep_factor, b
5490 Enable a periodic beep every second with frequency @var{beep_factor} times
5491 the carrier frequency. Default is 0, meaning the beep is disabled.
5493 @item sample_rate, r
5494 Specify the sample rate, default is 44100.
5497 Specify the duration of the generated audio stream.
5499 @item samples_per_frame
5500 Set the number of samples per output frame.
5502 The expression can contain the following constants:
5506 The (sequential) number of the output audio frame, starting from 0.
5509 The PTS (Presentation TimeStamp) of the output audio frame,
5510 expressed in @var{TB} units.
5513 The PTS of the output audio frame, expressed in seconds.
5516 The timebase of the output audio frames.
5519 Default is @code{1024}.
5522 @subsection Examples
5527 Generate a simple 440 Hz sine wave:
5533 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5537 sine=frequency=220:beep_factor=4:duration=5
5541 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5544 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5548 @c man end AUDIO SOURCES
5550 @chapter Audio Sinks
5551 @c man begin AUDIO SINKS
5553 Below is a description of the currently available audio sinks.
5555 @section abuffersink
5557 Buffer audio frames, and make them available to the end of filter chain.
5559 This sink is mainly intended for programmatic use, in particular
5560 through the interface defined in @file{libavfilter/buffersink.h}
5561 or the options system.
5563 It accepts a pointer to an AVABufferSinkContext structure, which
5564 defines the incoming buffers' formats, to be passed as the opaque
5565 parameter to @code{avfilter_init_filter} for initialization.
5568 Null audio sink; do absolutely nothing with the input audio. It is
5569 mainly useful as a template and for use in analysis / debugging
5572 @c man end AUDIO SINKS
5574 @chapter Video Filters
5575 @c man begin VIDEO FILTERS
5577 When you configure your FFmpeg build, you can disable any of the
5578 existing filters using @code{--disable-filters}.
5579 The configure output will show the video filters included in your
5582 Below is a description of the currently available video filters.
5584 @section alphaextract
5586 Extract the alpha component from the input as a grayscale video. This
5587 is especially useful with the @var{alphamerge} filter.
5591 Add or replace the alpha component of the primary input with the
5592 grayscale value of a second input. This is intended for use with
5593 @var{alphaextract} to allow the transmission or storage of frame
5594 sequences that have alpha in a format that doesn't support an alpha
5597 For example, to reconstruct full frames from a normal YUV-encoded video
5598 and a separate video created with @var{alphaextract}, you might use:
5600 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5603 Since this filter is designed for reconstruction, it operates on frame
5604 sequences without considering timestamps, and terminates when either
5605 input reaches end of stream. This will cause problems if your encoding
5606 pipeline drops frames. If you're trying to apply an image as an
5607 overlay to a video stream, consider the @var{overlay} filter instead.
5611 Amplify differences between current pixel and pixels of adjacent frames in
5612 same pixel location.
5614 This filter accepts the following options:
5618 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5619 For example radius of 3 will instruct filter to calculate average of 7 frames.
5622 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5625 Set threshold for difference amplification. Any differrence greater or equal to
5626 this value will not alter source pixel. Default is 10.
5627 Allowed range is from 0 to 65535.
5630 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5631 This option controls maximum possible value that will decrease source pixel value.
5634 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5635 This option controls maximum possible value that will increase source pixel value.
5638 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5643 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5644 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5645 Substation Alpha) subtitles files.
5647 This filter accepts the following option in addition to the common options from
5648 the @ref{subtitles} filter:
5652 Set the shaping engine
5654 Available values are:
5657 The default libass shaping engine, which is the best available.
5659 Fast, font-agnostic shaper that can do only substitutions
5661 Slower shaper using OpenType for substitutions and positioning
5664 The default is @code{auto}.
5668 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5670 The filter accepts the following options:
5674 Set threshold A for 1st plane. Default is 0.02.
5675 Valid range is 0 to 0.3.
5678 Set threshold B for 1st plane. Default is 0.04.
5679 Valid range is 0 to 5.
5682 Set threshold A for 2nd plane. Default is 0.02.
5683 Valid range is 0 to 0.3.
5686 Set threshold B for 2nd plane. Default is 0.04.
5687 Valid range is 0 to 5.
5690 Set threshold A for 3rd plane. Default is 0.02.
5691 Valid range is 0 to 0.3.
5694 Set threshold B for 3rd plane. Default is 0.04.
5695 Valid range is 0 to 5.
5697 Threshold A is designed to react on abrupt changes in the input signal and
5698 threshold B is designed to react on continuous changes in the input signal.
5701 Set number of frames filter will use for averaging. Default is 9. Must be odd
5702 number in range [5, 129].
5705 Set what planes of frame filter will use for averaging. Default is all.
5710 Apply average blur filter.
5712 The filter accepts the following options:
5716 Set horizontal radius size.
5719 Set which planes to filter. By default all planes are filtered.
5722 Set vertical radius size, if zero it will be same as @code{sizeX}.
5723 Default is @code{0}.
5728 Compute the bounding box for the non-black pixels in the input frame
5731 This filter computes the bounding box containing all the pixels with a
5732 luminance value greater than the minimum allowed value.
5733 The parameters describing the bounding box are printed on the filter
5736 The filter accepts the following option:
5740 Set the minimal luminance value. Default is @code{16}.
5743 @section bitplanenoise
5745 Show and measure bit plane noise.
5747 The filter accepts the following options:
5751 Set which plane to analyze. Default is @code{1}.
5754 Filter out noisy pixels from @code{bitplane} set above.
5755 Default is disabled.
5758 @section blackdetect
5760 Detect video intervals that are (almost) completely black. Can be
5761 useful to detect chapter transitions, commercials, or invalid
5762 recordings. Output lines contains the time for the start, end and
5763 duration of the detected black interval expressed in seconds.
5765 In order to display the output lines, you need to set the loglevel at
5766 least to the AV_LOG_INFO value.
5768 The filter accepts the following options:
5771 @item black_min_duration, d
5772 Set the minimum detected black duration expressed in seconds. It must
5773 be a non-negative floating point number.
5775 Default value is 2.0.
5777 @item picture_black_ratio_th, pic_th
5778 Set the threshold for considering a picture "black".
5779 Express the minimum value for the ratio:
5781 @var{nb_black_pixels} / @var{nb_pixels}
5784 for which a picture is considered black.
5785 Default value is 0.98.
5787 @item pixel_black_th, pix_th
5788 Set the threshold for considering a pixel "black".
5790 The threshold expresses the maximum pixel luminance value for which a
5791 pixel is considered "black". The provided value is scaled according to
5792 the following equation:
5794 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5797 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5798 the input video format, the range is [0-255] for YUV full-range
5799 formats and [16-235] for YUV non full-range formats.
5801 Default value is 0.10.
5804 The following example sets the maximum pixel threshold to the minimum
5805 value, and detects only black intervals of 2 or more seconds:
5807 blackdetect=d=2:pix_th=0.00
5812 Detect frames that are (almost) completely black. Can be useful to
5813 detect chapter transitions or commercials. Output lines consist of
5814 the frame number of the detected frame, the percentage of blackness,
5815 the position in the file if known or -1 and the timestamp in seconds.
5817 In order to display the output lines, you need to set the loglevel at
5818 least to the AV_LOG_INFO value.
5820 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5821 The value represents the percentage of pixels in the picture that
5822 are below the threshold value.
5824 It accepts the following parameters:
5829 The percentage of the pixels that have to be below the threshold; it defaults to
5832 @item threshold, thresh
5833 The threshold below which a pixel value is considered black; it defaults to
5838 @section blend, tblend
5840 Blend two video frames into each other.
5842 The @code{blend} filter takes two input streams and outputs one
5843 stream, the first input is the "top" layer and second input is
5844 "bottom" layer. By default, the output terminates when the longest input terminates.
5846 The @code{tblend} (time blend) filter takes two consecutive frames
5847 from one single stream, and outputs the result obtained by blending
5848 the new frame on top of the old frame.
5850 A description of the accepted options follows.
5858 Set blend mode for specific pixel component or all pixel components in case
5859 of @var{all_mode}. Default value is @code{normal}.
5861 Available values for component modes are:
5903 Set blend opacity for specific pixel component or all pixel components in case
5904 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5911 Set blend expression for specific pixel component or all pixel components in case
5912 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5914 The expressions can use the following variables:
5918 The sequential number of the filtered frame, starting from @code{0}.
5922 the coordinates of the current sample
5926 the width and height of currently filtered plane
5930 Width and height scale for the plane being filtered. It is the
5931 ratio between the dimensions of the current plane to the luma plane,
5932 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
5933 the luma plane and @code{0.5,0.5} for the chroma planes.
5936 Time of the current frame, expressed in seconds.
5939 Value of pixel component at current location for first video frame (top layer).
5942 Value of pixel component at current location for second video frame (bottom layer).
5946 The @code{blend} filter also supports the @ref{framesync} options.
5948 @subsection Examples
5952 Apply transition from bottom layer to top layer in first 10 seconds:
5954 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
5958 Apply linear horizontal transition from top layer to bottom layer:
5960 blend=all_expr='A*(X/W)+B*(1-X/W)'
5964 Apply 1x1 checkerboard effect:
5966 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
5970 Apply uncover left effect:
5972 blend=all_expr='if(gte(N*SW+X,W),A,B)'
5976 Apply uncover down effect:
5978 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
5982 Apply uncover up-left effect:
5984 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
5988 Split diagonally video and shows top and bottom layer on each side:
5990 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
5994 Display differences between the current and the previous frame:
5996 tblend=all_mode=grainextract
6002 Denoise frames using Block-Matching 3D algorithm.
6004 The filter accepts the following options.
6008 Set denoising strength. Default value is 1.
6009 Allowed range is from 0 to 999.9.
6010 The denoising algorith is very sensitive to sigma, so adjust it
6011 according to the source.
6014 Set local patch size. This sets dimensions in 2D.
6017 Set sliding step for processing blocks. Default value is 4.
6018 Allowed range is from 1 to 64.
6019 Smaller values allows processing more reference blocks and is slower.
6022 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6023 When set to 1, no block matching is done. Larger values allows more blocks
6025 Allowed range is from 1 to 256.
6028 Set radius for search block matching. Default is 9.
6029 Allowed range is from 1 to INT32_MAX.
6032 Set step between two search locations for block matching. Default is 1.
6033 Allowed range is from 1 to 64. Smaller is slower.
6036 Set threshold of mean square error for block matching. Valid range is 0 to
6040 Set thresholding parameter for hard thresholding in 3D transformed domain.
6041 Larger values results in stronger hard-thresholding filtering in frequency
6045 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6046 Default is @code{basic}.
6049 If enabled, filter will use 2nd stream for block matching.
6050 Default is disabled for @code{basic} value of @var{estim} option,
6051 and always enabled if value of @var{estim} is @code{final}.
6054 Set planes to filter. Default is all available except alpha.
6057 @subsection Examples
6061 Basic filtering with bm3d:
6063 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6067 Same as above, but filtering only luma:
6069 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6073 Same as above, but with both estimation modes:
6075 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
6079 Same as above, but prefilter with @ref{nlmeans} filter instead:
6081 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
6087 Apply a boxblur algorithm to the input video.
6089 It accepts the following parameters:
6093 @item luma_radius, lr
6094 @item luma_power, lp
6095 @item chroma_radius, cr
6096 @item chroma_power, cp
6097 @item alpha_radius, ar
6098 @item alpha_power, ap
6102 A description of the accepted options follows.
6105 @item luma_radius, lr
6106 @item chroma_radius, cr
6107 @item alpha_radius, ar
6108 Set an expression for the box radius in pixels used for blurring the
6109 corresponding input plane.
6111 The radius value must be a non-negative number, and must not be
6112 greater than the value of the expression @code{min(w,h)/2} for the
6113 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6116 Default value for @option{luma_radius} is "2". If not specified,
6117 @option{chroma_radius} and @option{alpha_radius} default to the
6118 corresponding value set for @option{luma_radius}.
6120 The expressions can contain the following constants:
6124 The input width and height in pixels.
6128 The input chroma image width and height in pixels.
6132 The horizontal and vertical chroma subsample values. For example, for the
6133 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6136 @item luma_power, lp
6137 @item chroma_power, cp
6138 @item alpha_power, ap
6139 Specify how many times the boxblur filter is applied to the
6140 corresponding plane.
6142 Default value for @option{luma_power} is 2. If not specified,
6143 @option{chroma_power} and @option{alpha_power} default to the
6144 corresponding value set for @option{luma_power}.
6146 A value of 0 will disable the effect.
6149 @subsection Examples
6153 Apply a boxblur filter with the luma, chroma, and alpha radii
6156 boxblur=luma_radius=2:luma_power=1
6161 Set the luma radius to 2, and alpha and chroma radius to 0:
6163 boxblur=2:1:cr=0:ar=0
6167 Set the luma and chroma radii to a fraction of the video dimension:
6169 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6175 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6176 Deinterlacing Filter").
6178 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6179 interpolation algorithms.
6180 It accepts the following parameters:
6184 The interlacing mode to adopt. It accepts one of the following values:
6188 Output one frame for each frame.
6190 Output one frame for each field.
6193 The default value is @code{send_field}.
6196 The picture field parity assumed for the input interlaced video. It accepts one
6197 of the following values:
6201 Assume the top field is first.
6203 Assume the bottom field is first.
6205 Enable automatic detection of field parity.
6208 The default value is @code{auto}.
6209 If the interlacing is unknown or the decoder does not export this information,
6210 top field first will be assumed.
6213 Specify which frames to deinterlace. Accept one of the following
6218 Deinterlace all frames.
6220 Only deinterlace frames marked as interlaced.
6223 The default value is @code{all}.
6227 Remove all color information for all colors except for certain one.
6229 The filter accepts the following options:
6233 The color which will not be replaced with neutral chroma.
6236 Similarity percentage with the above color.
6237 0.01 matches only the exact key color, while 1.0 matches everything.
6240 Signals that the color passed is already in YUV instead of RGB.
6242 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6243 This can be used to pass exact YUV values as hexadecimal numbers.
6247 YUV colorspace color/chroma keying.
6249 The filter accepts the following options:
6253 The color which will be replaced with transparency.
6256 Similarity percentage with the key color.
6258 0.01 matches only the exact key color, while 1.0 matches everything.
6263 0.0 makes pixels either fully transparent, or not transparent at all.
6265 Higher values result in semi-transparent pixels, with a higher transparency
6266 the more similar the pixels color is to the key color.
6269 Signals that the color passed is already in YUV instead of RGB.
6271 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6272 This can be used to pass exact YUV values as hexadecimal numbers.
6275 @subsection Examples
6279 Make every green pixel in the input image transparent:
6281 ffmpeg -i input.png -vf chromakey=green out.png
6285 Overlay a greenscreen-video on top of a static black background.
6287 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
6291 @section chromashift
6292 Shift chroma pixels horizontally and/or vertically.
6294 The filter accepts the following options:
6297 Set amount to shift chroma-blue horizontally.
6299 Set amount to shift chroma-blue vertically.
6301 Set amount to shift chroma-red horizontally.
6303 Set amount to shift chroma-red vertically.
6305 Set edge mode, can be @var{smear}, default, or @var{warp}.
6310 Display CIE color diagram with pixels overlaid onto it.
6312 The filter accepts the following options:
6327 @item uhdtv, rec2020
6340 Set what gamuts to draw.
6342 See @code{system} option for available values.
6345 Set ciescope size, by default set to 512.
6348 Set intensity used to map input pixel values to CIE diagram.
6351 Set contrast used to draw tongue colors that are out of active color system gamut.
6354 Correct gamma displayed on scope, by default enabled.
6357 Show white point on CIE diagram, by default disabled.
6360 Set input gamma. Used only with XYZ input color space.
6365 Visualize information exported by some codecs.
6367 Some codecs can export information through frames using side-data or other
6368 means. For example, some MPEG based codecs export motion vectors through the
6369 @var{export_mvs} flag in the codec @option{flags2} option.
6371 The filter accepts the following option:
6375 Set motion vectors to visualize.
6377 Available flags for @var{mv} are:
6381 forward predicted MVs of P-frames
6383 forward predicted MVs of B-frames
6385 backward predicted MVs of B-frames
6389 Display quantization parameters using the chroma planes.
6392 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6394 Available flags for @var{mv_type} are:
6398 forward predicted MVs
6400 backward predicted MVs
6403 @item frame_type, ft
6404 Set frame type to visualize motion vectors of.
6406 Available flags for @var{frame_type} are:
6410 intra-coded frames (I-frames)
6412 predicted frames (P-frames)
6414 bi-directionally predicted frames (B-frames)
6418 @subsection Examples
6422 Visualize forward predicted MVs of all frames using @command{ffplay}:
6424 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6428 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6430 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6434 @section colorbalance
6435 Modify intensity of primary colors (red, green and blue) of input frames.
6437 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6438 regions for the red-cyan, green-magenta or blue-yellow balance.
6440 A positive adjustment value shifts the balance towards the primary color, a negative
6441 value towards the complementary color.
6443 The filter accepts the following options:
6449 Adjust red, green and blue shadows (darkest pixels).
6454 Adjust red, green and blue midtones (medium pixels).
6459 Adjust red, green and blue highlights (brightest pixels).
6461 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6464 @subsection Examples
6468 Add red color cast to shadows:
6475 RGB colorspace color keying.
6477 The filter accepts the following options:
6481 The color which will be replaced with transparency.
6484 Similarity percentage with the key color.
6486 0.01 matches only the exact key color, while 1.0 matches everything.
6491 0.0 makes pixels either fully transparent, or not transparent at all.
6493 Higher values result in semi-transparent pixels, with a higher transparency
6494 the more similar the pixels color is to the key color.
6497 @subsection Examples
6501 Make every green pixel in the input image transparent:
6503 ffmpeg -i input.png -vf colorkey=green out.png
6507 Overlay a greenscreen-video on top of a static background image.
6509 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
6513 @section colorlevels
6515 Adjust video input frames using levels.
6517 The filter accepts the following options:
6524 Adjust red, green, blue and alpha input black point.
6525 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6531 Adjust red, green, blue and alpha input white point.
6532 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6534 Input levels are used to lighten highlights (bright tones), darken shadows
6535 (dark tones), change the balance of bright and dark tones.
6541 Adjust red, green, blue and alpha output black point.
6542 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6548 Adjust red, green, blue and alpha output white point.
6549 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6551 Output levels allows manual selection of a constrained output level range.
6554 @subsection Examples
6558 Make video output darker:
6560 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6566 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6570 Make video output lighter:
6572 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6576 Increase brightness:
6578 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6582 @section colorchannelmixer
6584 Adjust video input frames by re-mixing color channels.
6586 This filter modifies a color channel by adding the values associated to
6587 the other channels of the same pixels. For example if the value to
6588 modify is red, the output value will be:
6590 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6593 The filter accepts the following options:
6600 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6601 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6607 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6608 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6614 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6615 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6621 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6622 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6624 Allowed ranges for options are @code{[-2.0, 2.0]}.
6627 @subsection Examples
6631 Convert source to grayscale:
6633 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6636 Simulate sepia tones:
6638 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6642 @section colormatrix
6644 Convert color matrix.
6646 The filter accepts the following options:
6651 Specify the source and destination color matrix. Both values must be
6654 The accepted values are:
6682 For example to convert from BT.601 to SMPTE-240M, use the command:
6684 colormatrix=bt601:smpte240m
6689 Convert colorspace, transfer characteristics or color primaries.
6690 Input video needs to have an even size.
6692 The filter accepts the following options:
6697 Specify all color properties at once.
6699 The accepted values are:
6729 Specify output colorspace.
6731 The accepted values are:
6740 BT.470BG or BT.601-6 625
6743 SMPTE-170M or BT.601-6 525
6752 BT.2020 with non-constant luminance
6758 Specify output transfer characteristics.
6760 The accepted values are:
6772 Constant gamma of 2.2
6775 Constant gamma of 2.8
6778 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6796 BT.2020 for 10-bits content
6799 BT.2020 for 12-bits content
6805 Specify output color primaries.
6807 The accepted values are:
6816 BT.470BG or BT.601-6 625
6819 SMPTE-170M or BT.601-6 525
6843 Specify output color range.
6845 The accepted values are:
6848 TV (restricted) range
6851 MPEG (restricted) range
6862 Specify output color format.
6864 The accepted values are:
6867 YUV 4:2:0 planar 8-bits
6870 YUV 4:2:0 planar 10-bits
6873 YUV 4:2:0 planar 12-bits
6876 YUV 4:2:2 planar 8-bits
6879 YUV 4:2:2 planar 10-bits
6882 YUV 4:2:2 planar 12-bits
6885 YUV 4:4:4 planar 8-bits
6888 YUV 4:4:4 planar 10-bits
6891 YUV 4:4:4 planar 12-bits
6896 Do a fast conversion, which skips gamma/primary correction. This will take
6897 significantly less CPU, but will be mathematically incorrect. To get output
6898 compatible with that produced by the colormatrix filter, use fast=1.
6901 Specify dithering mode.
6903 The accepted values are:
6909 Floyd-Steinberg dithering
6913 Whitepoint adaptation mode.
6915 The accepted values are:
6918 Bradford whitepoint adaptation
6921 von Kries whitepoint adaptation
6924 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6928 Override all input properties at once. Same accepted values as @ref{all}.
6931 Override input colorspace. Same accepted values as @ref{space}.
6934 Override input color primaries. Same accepted values as @ref{primaries}.
6937 Override input transfer characteristics. Same accepted values as @ref{trc}.
6940 Override input color range. Same accepted values as @ref{range}.
6944 The filter converts the transfer characteristics, color space and color
6945 primaries to the specified user values. The output value, if not specified,
6946 is set to a default value based on the "all" property. If that property is
6947 also not specified, the filter will log an error. The output color range and
6948 format default to the same value as the input color range and format. The
6949 input transfer characteristics, color space, color primaries and color range
6950 should be set on the input data. If any of these are missing, the filter will
6951 log an error and no conversion will take place.
6953 For example to convert the input to SMPTE-240M, use the command:
6955 colorspace=smpte240m
6958 @section convolution
6960 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
6962 The filter accepts the following options:
6969 Set matrix for each plane.
6970 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
6971 and from 1 to 49 odd number of signed integers in @var{row} mode.
6977 Set multiplier for calculated value for each plane.
6978 If unset or 0, it will be sum of all matrix elements.
6984 Set bias for each plane. This value is added to the result of the multiplication.
6985 Useful for making the overall image brighter or darker. Default is 0.0.
6991 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
6992 Default is @var{square}.
6995 @subsection Examples
7001 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"
7007 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"
7013 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"
7019 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"
7023 Apply laplacian edge detector which includes diagonals:
7025 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"
7031 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"
7037 Apply 2D convolution of video stream in frequency domain using second stream
7040 The filter accepts the following options:
7044 Set which planes to process.
7047 Set which impulse video frames will be processed, can be @var{first}
7048 or @var{all}. Default is @var{all}.
7051 The @code{convolve} filter also supports the @ref{framesync} options.
7055 Copy the input video source unchanged to the output. This is mainly useful for
7060 Video filtering on GPU using Apple's CoreImage API on OSX.
7062 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7063 processed by video hardware. However, software-based OpenGL implementations
7064 exist which means there is no guarantee for hardware processing. It depends on
7067 There are many filters and image generators provided by Apple that come with a
7068 large variety of options. The filter has to be referenced by its name along
7071 The coreimage filter accepts the following options:
7074 List all available filters and generators along with all their respective
7075 options as well as possible minimum and maximum values along with the default
7082 Specify all filters by their respective name and options.
7083 Use @var{list_filters} to determine all valid filter names and options.
7084 Numerical options are specified by a float value and are automatically clamped
7085 to their respective value range. Vector and color options have to be specified
7086 by a list of space separated float values. Character escaping has to be done.
7087 A special option name @code{default} is available to use default options for a
7090 It is required to specify either @code{default} or at least one of the filter options.
7091 All omitted options are used with their default values.
7092 The syntax of the filter string is as follows:
7094 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7098 Specify a rectangle where the output of the filter chain is copied into the
7099 input image. It is given by a list of space separated float values:
7101 output_rect=x\ y\ width\ height
7103 If not given, the output rectangle equals the dimensions of the input image.
7104 The output rectangle is automatically cropped at the borders of the input
7105 image. Negative values are valid for each component.
7107 output_rect=25\ 25\ 100\ 100
7111 Several filters can be chained for successive processing without GPU-HOST
7112 transfers allowing for fast processing of complex filter chains.
7113 Currently, only filters with zero (generators) or exactly one (filters) input
7114 image and one output image are supported. Also, transition filters are not yet
7117 Some filters generate output images with additional padding depending on the
7118 respective filter kernel. The padding is automatically removed to ensure the
7119 filter output has the same size as the input image.
7121 For image generators, the size of the output image is determined by the
7122 previous output image of the filter chain or the input image of the whole
7123 filterchain, respectively. The generators do not use the pixel information of
7124 this image to generate their output. However, the generated output is
7125 blended onto this image, resulting in partial or complete coverage of the
7128 The @ref{coreimagesrc} video source can be used for generating input images
7129 which are directly fed into the filter chain. By using it, providing input
7130 images by another video source or an input video is not required.
7132 @subsection Examples
7137 List all filters available:
7139 coreimage=list_filters=true
7143 Use the CIBoxBlur filter with default options to blur an image:
7145 coreimage=filter=CIBoxBlur@@default
7149 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7150 its center at 100x100 and a radius of 50 pixels:
7152 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7156 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7157 given as complete and escaped command-line for Apple's standard bash shell:
7159 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7165 Crop the input video to given dimensions.
7167 It accepts the following parameters:
7171 The width of the output video. It defaults to @code{iw}.
7172 This expression is evaluated only once during the filter
7173 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7176 The height of the output video. It defaults to @code{ih}.
7177 This expression is evaluated only once during the filter
7178 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7181 The horizontal position, in the input video, of the left edge of the output
7182 video. It defaults to @code{(in_w-out_w)/2}.
7183 This expression is evaluated per-frame.
7186 The vertical position, in the input video, of the top edge of the output video.
7187 It defaults to @code{(in_h-out_h)/2}.
7188 This expression is evaluated per-frame.
7191 If set to 1 will force the output display aspect ratio
7192 to be the same of the input, by changing the output sample aspect
7193 ratio. It defaults to 0.
7196 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7197 width/height/x/y as specified and will not be rounded to nearest smaller value.
7201 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7202 expressions containing the following constants:
7207 The computed values for @var{x} and @var{y}. They are evaluated for
7212 The input width and height.
7216 These are the same as @var{in_w} and @var{in_h}.
7220 The output (cropped) width and height.
7224 These are the same as @var{out_w} and @var{out_h}.
7227 same as @var{iw} / @var{ih}
7230 input sample aspect ratio
7233 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7237 horizontal and vertical chroma subsample values. For example for the
7238 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7241 The number of the input frame, starting from 0.
7244 the position in the file of the input frame, NAN if unknown
7247 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7251 The expression for @var{out_w} may depend on the value of @var{out_h},
7252 and the expression for @var{out_h} may depend on @var{out_w}, but they
7253 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7254 evaluated after @var{out_w} and @var{out_h}.
7256 The @var{x} and @var{y} parameters specify the expressions for the
7257 position of the top-left corner of the output (non-cropped) area. They
7258 are evaluated for each frame. If the evaluated value is not valid, it
7259 is approximated to the nearest valid value.
7261 The expression for @var{x} may depend on @var{y}, and the expression
7262 for @var{y} may depend on @var{x}.
7264 @subsection Examples
7268 Crop area with size 100x100 at position (12,34).
7273 Using named options, the example above becomes:
7275 crop=w=100:h=100:x=12:y=34
7279 Crop the central input area with size 100x100:
7285 Crop the central input area with size 2/3 of the input video:
7287 crop=2/3*in_w:2/3*in_h
7291 Crop the input video central square:
7298 Delimit the rectangle with the top-left corner placed at position
7299 100:100 and the right-bottom corner corresponding to the right-bottom
7300 corner of the input image.
7302 crop=in_w-100:in_h-100:100:100
7306 Crop 10 pixels from the left and right borders, and 20 pixels from
7307 the top and bottom borders
7309 crop=in_w-2*10:in_h-2*20
7313 Keep only the bottom right quarter of the input image:
7315 crop=in_w/2:in_h/2:in_w/2:in_h/2
7319 Crop height for getting Greek harmony:
7321 crop=in_w:1/PHI*in_w
7325 Apply trembling effect:
7327 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)
7331 Apply erratic camera effect depending on timestamp:
7333 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)"
7337 Set x depending on the value of y:
7339 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7343 @subsection Commands
7345 This filter supports the following commands:
7351 Set width/height of the output video and the horizontal/vertical position
7353 The command accepts the same syntax of the corresponding option.
7355 If the specified expression is not valid, it is kept at its current
7361 Auto-detect the crop size.
7363 It calculates the necessary cropping parameters and prints the
7364 recommended parameters via the logging system. The detected dimensions
7365 correspond to the non-black area of the input video.
7367 It accepts the following parameters:
7372 Set higher black value threshold, which can be optionally specified
7373 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7374 value greater to the set value is considered non-black. It defaults to 24.
7375 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7376 on the bitdepth of the pixel format.
7379 The value which the width/height should be divisible by. It defaults to
7380 16. The offset is automatically adjusted to center the video. Use 2 to
7381 get only even dimensions (needed for 4:2:2 video). 16 is best when
7382 encoding to most video codecs.
7384 @item reset_count, reset
7385 Set the counter that determines after how many frames cropdetect will
7386 reset the previously detected largest video area and start over to
7387 detect the current optimal crop area. Default value is 0.
7389 This can be useful when channel logos distort the video area. 0
7390 indicates 'never reset', and returns the largest area encountered during
7397 Delay video filtering until a given wallclock timestamp. The filter first
7398 passes on @option{preroll} amount of frames, then it buffers at most
7399 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7400 it forwards the buffered frames and also any subsequent frames coming in its
7403 The filter can be used synchronize the output of multiple ffmpeg processes for
7404 realtime output devices like decklink. By putting the delay in the filtering
7405 chain and pre-buffering frames the process can pass on data to output almost
7406 immediately after the target wallclock timestamp is reached.
7408 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7414 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7417 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7420 The maximum duration of content to buffer before waiting for the cue expressed
7421 in seconds. Default is 0.
7428 Apply color adjustments using curves.
7430 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7431 component (red, green and blue) has its values defined by @var{N} key points
7432 tied from each other using a smooth curve. The x-axis represents the pixel
7433 values from the input frame, and the y-axis the new pixel values to be set for
7436 By default, a component curve is defined by the two points @var{(0;0)} and
7437 @var{(1;1)}. This creates a straight line where each original pixel value is
7438 "adjusted" to its own value, which means no change to the image.
7440 The filter allows you to redefine these two points and add some more. A new
7441 curve (using a natural cubic spline interpolation) will be define to pass
7442 smoothly through all these new coordinates. The new defined points needs to be
7443 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7444 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7445 the vector spaces, the values will be clipped accordingly.
7447 The filter accepts the following options:
7451 Select one of the available color presets. This option can be used in addition
7452 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7453 options takes priority on the preset values.
7454 Available presets are:
7457 @item color_negative
7460 @item increase_contrast
7462 @item linear_contrast
7463 @item medium_contrast
7465 @item strong_contrast
7468 Default is @code{none}.
7470 Set the master key points. These points will define a second pass mapping. It
7471 is sometimes called a "luminance" or "value" mapping. It can be used with
7472 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7473 post-processing LUT.
7475 Set the key points for the red component.
7477 Set the key points for the green component.
7479 Set the key points for the blue component.
7481 Set the key points for all components (not including master).
7482 Can be used in addition to the other key points component
7483 options. In this case, the unset component(s) will fallback on this
7484 @option{all} setting.
7486 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7488 Save Gnuplot script of the curves in specified file.
7491 To avoid some filtergraph syntax conflicts, each key points list need to be
7492 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7494 @subsection Examples
7498 Increase slightly the middle level of blue:
7500 curves=blue='0/0 0.5/0.58 1/1'
7506 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'
7508 Here we obtain the following coordinates for each components:
7511 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7513 @code{(0;0) (0.50;0.48) (1;1)}
7515 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7519 The previous example can also be achieved with the associated built-in preset:
7521 curves=preset=vintage
7531 Use a Photoshop preset and redefine the points of the green component:
7533 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7537 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7538 and @command{gnuplot}:
7540 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7541 gnuplot -p /tmp/curves.plt
7547 Video data analysis filter.
7549 This filter shows hexadecimal pixel values of part of video.
7551 The filter accepts the following options:
7555 Set output video size.
7558 Set x offset from where to pick pixels.
7561 Set y offset from where to pick pixels.
7564 Set scope mode, can be one of the following:
7567 Draw hexadecimal pixel values with white color on black background.
7570 Draw hexadecimal pixel values with input video pixel color on black
7574 Draw hexadecimal pixel values on color background picked from input video,
7575 the text color is picked in such way so its always visible.
7579 Draw rows and columns numbers on left and top of video.
7582 Set background opacity.
7587 Denoise frames using 2D DCT (frequency domain filtering).
7589 This filter is not designed for real time.
7591 The filter accepts the following options:
7595 Set the noise sigma constant.
7597 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7598 coefficient (absolute value) below this threshold with be dropped.
7600 If you need a more advanced filtering, see @option{expr}.
7602 Default is @code{0}.
7605 Set number overlapping pixels for each block. Since the filter can be slow, you
7606 may want to reduce this value, at the cost of a less effective filter and the
7607 risk of various artefacts.
7609 If the overlapping value doesn't permit processing the whole input width or
7610 height, a warning will be displayed and according borders won't be denoised.
7612 Default value is @var{blocksize}-1, which is the best possible setting.
7615 Set the coefficient factor expression.
7617 For each coefficient of a DCT block, this expression will be evaluated as a
7618 multiplier value for the coefficient.
7620 If this is option is set, the @option{sigma} option will be ignored.
7622 The absolute value of the coefficient can be accessed through the @var{c}
7626 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7627 @var{blocksize}, which is the width and height of the processed blocks.
7629 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7630 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7631 on the speed processing. Also, a larger block size does not necessarily means a
7635 @subsection Examples
7637 Apply a denoise with a @option{sigma} of @code{4.5}:
7642 The same operation can be achieved using the expression system:
7644 dctdnoiz=e='gte(c, 4.5*3)'
7647 Violent denoise using a block size of @code{16x16}:
7654 Remove banding artifacts from input video.
7655 It works by replacing banded pixels with average value of referenced pixels.
7657 The filter accepts the following options:
7664 Set banding detection threshold for each plane. Default is 0.02.
7665 Valid range is 0.00003 to 0.5.
7666 If difference between current pixel and reference pixel is less than threshold,
7667 it will be considered as banded.
7670 Banding detection range in pixels. Default is 16. If positive, random number
7671 in range 0 to set value will be used. If negative, exact absolute value
7673 The range defines square of four pixels around current pixel.
7676 Set direction in radians from which four pixel will be compared. If positive,
7677 random direction from 0 to set direction will be picked. If negative, exact of
7678 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7679 will pick only pixels on same row and -PI/2 will pick only pixels on same
7683 If enabled, current pixel is compared with average value of all four
7684 surrounding pixels. The default is enabled. If disabled current pixel is
7685 compared with all four surrounding pixels. The pixel is considered banded
7686 if only all four differences with surrounding pixels are less than threshold.
7689 If enabled, current pixel is changed if and only if all pixel components are banded,
7690 e.g. banding detection threshold is triggered for all color components.
7691 The default is disabled.
7696 Remove blocking artifacts from input video.
7698 The filter accepts the following options:
7702 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7703 This controls what kind of deblocking is applied.
7706 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7712 Set blocking detection thresholds. Allowed range is 0 to 1.
7713 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7714 Using higher threshold gives more deblocking strength.
7715 Setting @var{alpha} controls threshold detection at exact edge of block.
7716 Remaining options controls threshold detection near the edge. Each one for
7717 below/above or left/right. Setting any of those to @var{0} disables
7721 Set planes to filter. Default is to filter all available planes.
7724 @subsection Examples
7728 Deblock using weak filter and block size of 4 pixels.
7730 deblock=filter=weak:block=4
7734 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7735 deblocking more edges.
7737 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7741 Similar as above, but filter only first plane.
7743 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7747 Similar as above, but filter only second and third plane.
7749 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7756 Drop duplicated frames at regular intervals.
7758 The filter accepts the following options:
7762 Set the number of frames from which one will be dropped. Setting this to
7763 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7764 Default is @code{5}.
7767 Set the threshold for duplicate detection. If the difference metric for a frame
7768 is less than or equal to this value, then it is declared as duplicate. Default
7772 Set scene change threshold. Default is @code{15}.
7776 Set the size of the x and y-axis blocks used during metric calculations.
7777 Larger blocks give better noise suppression, but also give worse detection of
7778 small movements. Must be a power of two. Default is @code{32}.
7781 Mark main input as a pre-processed input and activate clean source input
7782 stream. This allows the input to be pre-processed with various filters to help
7783 the metrics calculation while keeping the frame selection lossless. When set to
7784 @code{1}, the first stream is for the pre-processed input, and the second
7785 stream is the clean source from where the kept frames are chosen. Default is
7789 Set whether or not chroma is considered in the metric calculations. Default is
7795 Apply 2D deconvolution of video stream in frequency domain using second stream
7798 The filter accepts the following options:
7802 Set which planes to process.
7805 Set which impulse video frames will be processed, can be @var{first}
7806 or @var{all}. Default is @var{all}.
7809 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7810 and height are not same and not power of 2 or if stream prior to convolving
7814 The @code{deconvolve} filter also supports the @ref{framesync} options.
7818 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
7820 It accepts the following options:
7824 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
7825 @var{rainbows} for cross-color reduction.
7828 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
7831 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
7834 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
7837 Set temporal chroma threshold. Lower values increases reduction of cross-color.
7842 Apply deflate effect to the video.
7844 This filter replaces the pixel by the local(3x3) average by taking into account
7845 only values lower than the pixel.
7847 It accepts the following options:
7854 Limit the maximum change for each plane, default is 65535.
7855 If 0, plane will remain unchanged.
7860 Remove temporal frame luminance variations.
7862 It accepts the following options:
7866 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7869 Set averaging mode to smooth temporal luminance variations.
7871 Available values are:
7896 Do not actually modify frame. Useful when one only wants metadata.
7901 Remove judder produced by partially interlaced telecined content.
7903 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7904 source was partially telecined content then the output of @code{pullup,dejudder}
7905 will have a variable frame rate. May change the recorded frame rate of the
7906 container. Aside from that change, this filter will not affect constant frame
7909 The option available in this filter is:
7913 Specify the length of the window over which the judder repeats.
7915 Accepts any integer greater than 1. Useful values are:
7919 If the original was telecined from 24 to 30 fps (Film to NTSC).
7922 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7925 If a mixture of the two.
7928 The default is @samp{4}.
7933 Suppress a TV station logo by a simple interpolation of the surrounding
7934 pixels. Just set a rectangle covering the logo and watch it disappear
7935 (and sometimes something even uglier appear - your mileage may vary).
7937 It accepts the following parameters:
7942 Specify the top left corner coordinates of the logo. They must be
7947 Specify the width and height of the logo to clear. They must be
7951 Specify the thickness of the fuzzy edge of the rectangle (added to
7952 @var{w} and @var{h}). The default value is 1. This option is
7953 deprecated, setting higher values should no longer be necessary and
7957 When set to 1, a green rectangle is drawn on the screen to simplify
7958 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
7959 The default value is 0.
7961 The rectangle is drawn on the outermost pixels which will be (partly)
7962 replaced with interpolated values. The values of the next pixels
7963 immediately outside this rectangle in each direction will be used to
7964 compute the interpolated pixel values inside the rectangle.
7968 @subsection Examples
7972 Set a rectangle covering the area with top left corner coordinates 0,0
7973 and size 100x77, and a band of size 10:
7975 delogo=x=0:y=0:w=100:h=77:band=10
7982 Attempt to fix small changes in horizontal and/or vertical shift. This
7983 filter helps remove camera shake from hand-holding a camera, bumping a
7984 tripod, moving on a vehicle, etc.
7986 The filter accepts the following options:
7994 Specify a rectangular area where to limit the search for motion
7996 If desired the search for motion vectors can be limited to a
7997 rectangular area of the frame defined by its top left corner, width
7998 and height. These parameters have the same meaning as the drawbox
7999 filter which can be used to visualise the position of the bounding
8002 This is useful when simultaneous movement of subjects within the frame
8003 might be confused for camera motion by the motion vector search.
8005 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8006 then the full frame is used. This allows later options to be set
8007 without specifying the bounding box for the motion vector search.
8009 Default - search the whole frame.
8013 Specify the maximum extent of movement in x and y directions in the
8014 range 0-64 pixels. Default 16.
8017 Specify how to generate pixels to fill blanks at the edge of the
8018 frame. Available values are:
8021 Fill zeroes at blank locations
8023 Original image at blank locations
8025 Extruded edge value at blank locations
8027 Mirrored edge at blank locations
8029 Default value is @samp{mirror}.
8032 Specify the blocksize to use for motion search. Range 4-128 pixels,
8036 Specify the contrast threshold for blocks. Only blocks with more than
8037 the specified contrast (difference between darkest and lightest
8038 pixels) will be considered. Range 1-255, default 125.
8041 Specify the search strategy. Available values are:
8044 Set exhaustive search
8046 Set less exhaustive search.
8048 Default value is @samp{exhaustive}.
8051 If set then a detailed log of the motion search is written to the
8058 Remove unwanted contamination of foreground colors, caused by reflected color of
8059 greenscreen or bluescreen.
8061 This filter accepts the following options:
8065 Set what type of despill to use.
8068 Set how spillmap will be generated.
8071 Set how much to get rid of still remaining spill.
8074 Controls amount of red in spill area.
8077 Controls amount of green in spill area.
8078 Should be -1 for greenscreen.
8081 Controls amount of blue in spill area.
8082 Should be -1 for bluescreen.
8085 Controls brightness of spill area, preserving colors.
8088 Modify alpha from generated spillmap.
8093 Apply an exact inverse of the telecine operation. It requires a predefined
8094 pattern specified using the pattern option which must be the same as that passed
8095 to the telecine filter.
8097 This filter accepts the following options:
8106 The default value is @code{top}.
8110 A string of numbers representing the pulldown pattern you wish to apply.
8111 The default value is @code{23}.
8114 A number representing position of the first frame with respect to the telecine
8115 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8120 Apply dilation effect to the video.
8122 This filter replaces the pixel by the local(3x3) maximum.
8124 It accepts the following options:
8131 Limit the maximum change for each plane, default is 65535.
8132 If 0, plane will remain unchanged.
8135 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8138 Flags to local 3x3 coordinates maps like this:
8147 Displace pixels as indicated by second and third input stream.
8149 It takes three input streams and outputs one stream, the first input is the
8150 source, and second and third input are displacement maps.
8152 The second input specifies how much to displace pixels along the
8153 x-axis, while the third input specifies how much to displace pixels
8155 If one of displacement map streams terminates, last frame from that
8156 displacement map will be used.
8158 Note that once generated, displacements maps can be reused over and over again.
8160 A description of the accepted options follows.
8164 Set displace behavior for pixels that are out of range.
8166 Available values are:
8169 Missing pixels are replaced by black pixels.
8172 Adjacent pixels will spread out to replace missing pixels.
8175 Out of range pixels are wrapped so they point to pixels of other side.
8178 Out of range pixels will be replaced with mirrored pixels.
8180 Default is @samp{smear}.
8184 @subsection Examples
8188 Add ripple effect to rgb input of video size hd720:
8190 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
8194 Add wave effect to rgb input of video size hd720:
8196 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
8202 Draw a colored box on the input image.
8204 It accepts the following parameters:
8209 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8213 The expressions which specify the width and height of the box; if 0 they are interpreted as
8214 the input width and height. It defaults to 0.
8217 Specify the color of the box to write. For the general syntax of this option,
8218 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8219 value @code{invert} is used, the box edge color is the same as the
8220 video with inverted luma.
8223 The expression which sets the thickness of the box edge.
8224 A value of @code{fill} will create a filled box. Default value is @code{3}.
8226 See below for the list of accepted constants.
8229 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8230 will overwrite the video's color and alpha pixels.
8231 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8234 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8235 following constants:
8239 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8243 horizontal and vertical chroma subsample values. For example for the
8244 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8248 The input width and height.
8251 The input sample aspect ratio.
8255 The x and y offset coordinates where the box is drawn.
8259 The width and height of the drawn box.
8262 The thickness of the drawn box.
8264 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8265 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8269 @subsection Examples
8273 Draw a black box around the edge of the input image:
8279 Draw a box with color red and an opacity of 50%:
8281 drawbox=10:20:200:60:red@@0.5
8284 The previous example can be specified as:
8286 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8290 Fill the box with pink color:
8292 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8296 Draw a 2-pixel red 2.40:1 mask:
8298 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
8304 Draw a grid on the input image.
8306 It accepts the following parameters:
8311 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8315 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8316 input width and height, respectively, minus @code{thickness}, so image gets
8317 framed. Default to 0.
8320 Specify the color of the grid. For the general syntax of this option,
8321 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8322 value @code{invert} is used, the grid color is the same as the
8323 video with inverted luma.
8326 The expression which sets the thickness of the grid line. Default value is @code{1}.
8328 See below for the list of accepted constants.
8331 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8332 will overwrite the video's color and alpha pixels.
8333 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8336 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8337 following constants:
8341 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8345 horizontal and vertical chroma subsample values. For example for the
8346 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8350 The input grid cell width and height.
8353 The input sample aspect ratio.
8357 The x and y coordinates of some point of grid intersection (meant to configure offset).
8361 The width and height of the drawn cell.
8364 The thickness of the drawn cell.
8366 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8367 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8371 @subsection Examples
8375 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8377 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8381 Draw a white 3x3 grid with an opacity of 50%:
8383 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8390 Draw a text string or text from a specified file on top of a video, using the
8391 libfreetype library.
8393 To enable compilation of this filter, you need to configure FFmpeg with
8394 @code{--enable-libfreetype}.
8395 To enable default font fallback and the @var{font} option you need to
8396 configure FFmpeg with @code{--enable-libfontconfig}.
8397 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8398 @code{--enable-libfribidi}.
8402 It accepts the following parameters:
8407 Used to draw a box around text using the background color.
8408 The value must be either 1 (enable) or 0 (disable).
8409 The default value of @var{box} is 0.
8412 Set the width of the border to be drawn around the box using @var{boxcolor}.
8413 The default value of @var{boxborderw} is 0.
8416 The color to be used for drawing box around text. For the syntax of this
8417 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8419 The default value of @var{boxcolor} is "white".
8422 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8423 The default value of @var{line_spacing} is 0.
8426 Set the width of the border to be drawn around the text using @var{bordercolor}.
8427 The default value of @var{borderw} is 0.
8430 Set the color to be used for drawing border around text. For the syntax of this
8431 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8433 The default value of @var{bordercolor} is "black".
8436 Select how the @var{text} is expanded. Can be either @code{none},
8437 @code{strftime} (deprecated) or
8438 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8442 Set a start time for the count. Value is in microseconds. Only applied
8443 in the deprecated strftime expansion mode. To emulate in normal expansion
8444 mode use the @code{pts} function, supplying the start time (in seconds)
8445 as the second argument.
8448 If true, check and fix text coords to avoid clipping.
8451 The color to be used for drawing fonts. For the syntax of this option, check
8452 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8454 The default value of @var{fontcolor} is "black".
8456 @item fontcolor_expr
8457 String which is expanded the same way as @var{text} to obtain dynamic
8458 @var{fontcolor} value. By default this option has empty value and is not
8459 processed. When this option is set, it overrides @var{fontcolor} option.
8462 The font family to be used for drawing text. By default Sans.
8465 The font file to be used for drawing text. The path must be included.
8466 This parameter is mandatory if the fontconfig support is disabled.
8469 Draw the text applying alpha blending. The value can
8470 be a number between 0.0 and 1.0.
8471 The expression accepts the same variables @var{x, y} as well.
8472 The default value is 1.
8473 Please see @var{fontcolor_expr}.
8476 The font size to be used for drawing text.
8477 The default value of @var{fontsize} is 16.
8480 If set to 1, attempt to shape the text (for example, reverse the order of
8481 right-to-left text and join Arabic characters) before drawing it.
8482 Otherwise, just draw the text exactly as given.
8483 By default 1 (if supported).
8486 The flags to be used for loading the fonts.
8488 The flags map the corresponding flags supported by libfreetype, and are
8489 a combination of the following values:
8496 @item vertical_layout
8497 @item force_autohint
8500 @item ignore_global_advance_width
8502 @item ignore_transform
8508 Default value is "default".
8510 For more information consult the documentation for the FT_LOAD_*
8514 The color to be used for drawing a shadow behind the drawn text. For the
8515 syntax of this option, check the @ref{color syntax,,"Color" section in the
8516 ffmpeg-utils manual,ffmpeg-utils}.
8518 The default value of @var{shadowcolor} is "black".
8522 The x and y offsets for the text shadow position with respect to the
8523 position of the text. They can be either positive or negative
8524 values. The default value for both is "0".
8527 The starting frame number for the n/frame_num variable. The default value
8531 The size in number of spaces to use for rendering the tab.
8535 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8536 format. It can be used with or without text parameter. @var{timecode_rate}
8537 option must be specified.
8539 @item timecode_rate, rate, r
8540 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8541 integer. Minimum value is "1".
8542 Drop-frame timecode is supported for frame rates 30 & 60.
8545 If set to 1, the output of the timecode option will wrap around at 24 hours.
8546 Default is 0 (disabled).
8549 The text string to be drawn. The text must be a sequence of UTF-8
8551 This parameter is mandatory if no file is specified with the parameter
8555 A text file containing text to be drawn. The text must be a sequence
8556 of UTF-8 encoded characters.
8558 This parameter is mandatory if no text string is specified with the
8559 parameter @var{text}.
8561 If both @var{text} and @var{textfile} are specified, an error is thrown.
8564 If set to 1, the @var{textfile} will be reloaded before each frame.
8565 Be sure to update it atomically, or it may be read partially, or even fail.
8569 The expressions which specify the offsets where text will be drawn
8570 within the video frame. They are relative to the top/left border of the
8573 The default value of @var{x} and @var{y} is "0".
8575 See below for the list of accepted constants and functions.
8578 The parameters for @var{x} and @var{y} are expressions containing the
8579 following constants and functions:
8583 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8587 horizontal and vertical chroma subsample values. For example for the
8588 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8591 the height of each text line
8599 @item max_glyph_a, ascent
8600 the maximum distance from the baseline to the highest/upper grid
8601 coordinate used to place a glyph outline point, for all the rendered
8603 It is a positive value, due to the grid's orientation with the Y axis
8606 @item max_glyph_d, descent
8607 the maximum distance from the baseline to the lowest grid coordinate
8608 used to place a glyph outline point, for all the rendered glyphs.
8609 This is a negative value, due to the grid's orientation, with the Y axis
8613 maximum glyph height, that is the maximum height for all the glyphs
8614 contained in the rendered text, it is equivalent to @var{ascent} -
8618 maximum glyph width, that is the maximum width for all the glyphs
8619 contained in the rendered text
8622 the number of input frame, starting from 0
8624 @item rand(min, max)
8625 return a random number included between @var{min} and @var{max}
8628 The input sample aspect ratio.
8631 timestamp expressed in seconds, NAN if the input timestamp is unknown
8634 the height of the rendered text
8637 the width of the rendered text
8641 the x and y offset coordinates where the text is drawn.
8643 These parameters allow the @var{x} and @var{y} expressions to refer
8644 each other, so you can for example specify @code{y=x/dar}.
8647 @anchor{drawtext_expansion}
8648 @subsection Text expansion
8650 If @option{expansion} is set to @code{strftime},
8651 the filter recognizes strftime() sequences in the provided text and
8652 expands them accordingly. Check the documentation of strftime(). This
8653 feature is deprecated.
8655 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8657 If @option{expansion} is set to @code{normal} (which is the default),
8658 the following expansion mechanism is used.
8660 The backslash character @samp{\}, followed by any character, always expands to
8661 the second character.
8663 Sequences of the form @code{%@{...@}} are expanded. The text between the
8664 braces is a function name, possibly followed by arguments separated by ':'.
8665 If the arguments contain special characters or delimiters (':' or '@}'),
8666 they should be escaped.
8668 Note that they probably must also be escaped as the value for the
8669 @option{text} option in the filter argument string and as the filter
8670 argument in the filtergraph description, and possibly also for the shell,
8671 that makes up to four levels of escaping; using a text file avoids these
8674 The following functions are available:
8679 The expression evaluation result.
8681 It must take one argument specifying the expression to be evaluated,
8682 which accepts the same constants and functions as the @var{x} and
8683 @var{y} values. Note that not all constants should be used, for
8684 example the text size is not known when evaluating the expression, so
8685 the constants @var{text_w} and @var{text_h} will have an undefined
8688 @item expr_int_format, eif
8689 Evaluate the expression's value and output as formatted integer.
8691 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8692 The second argument specifies the output format. Allowed values are @samp{x},
8693 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8694 @code{printf} function.
8695 The third parameter is optional and sets the number of positions taken by the output.
8696 It can be used to add padding with zeros from the left.
8699 The time at which the filter is running, expressed in UTC.
8700 It can accept an argument: a strftime() format string.
8703 The time at which the filter is running, expressed in the local time zone.
8704 It can accept an argument: a strftime() format string.
8707 Frame metadata. Takes one or two arguments.
8709 The first argument is mandatory and specifies the metadata key.
8711 The second argument is optional and specifies a default value, used when the
8712 metadata key is not found or empty.
8715 The frame number, starting from 0.
8718 A 1 character description of the current picture type.
8721 The timestamp of the current frame.
8722 It can take up to three arguments.
8724 The first argument is the format of the timestamp; it defaults to @code{flt}
8725 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8726 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8727 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8728 @code{localtime} stands for the timestamp of the frame formatted as
8729 local time zone time.
8731 The second argument is an offset added to the timestamp.
8733 If the format is set to @code{hms}, a third argument @code{24HH} may be
8734 supplied to present the hour part of the formatted timestamp in 24h format
8737 If the format is set to @code{localtime} or @code{gmtime},
8738 a third argument may be supplied: a strftime() format string.
8739 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8742 @subsection Examples
8746 Draw "Test Text" with font FreeSerif, using the default values for the
8747 optional parameters.
8750 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8754 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8755 and y=50 (counting from the top-left corner of the screen), text is
8756 yellow with a red box around it. Both the text and the box have an
8760 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8761 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8764 Note that the double quotes are not necessary if spaces are not used
8765 within the parameter list.
8768 Show the text at the center of the video frame:
8770 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8774 Show the text at a random position, switching to a new position every 30 seconds:
8776 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)"
8780 Show a text line sliding from right to left in the last row of the video
8781 frame. The file @file{LONG_LINE} is assumed to contain a single line
8784 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8788 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8790 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8794 Draw a single green letter "g", at the center of the input video.
8795 The glyph baseline is placed at half screen height.
8797 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8801 Show text for 1 second every 3 seconds:
8803 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8807 Use fontconfig to set the font. Note that the colons need to be escaped.
8809 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8813 Print the date of a real-time encoding (see strftime(3)):
8815 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8819 Show text fading in and out (appearing/disappearing):
8822 DS=1.0 # display start
8823 DE=10.0 # display end
8824 FID=1.5 # fade in duration
8825 FOD=5 # fade out duration
8826 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 @}"
8830 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8831 and the @option{fontsize} value are included in the @option{y} offset.
8833 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8834 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8839 For more information about libfreetype, check:
8840 @url{http://www.freetype.org/}.
8842 For more information about fontconfig, check:
8843 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8845 For more information about libfribidi, check:
8846 @url{http://fribidi.org/}.
8850 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8852 The filter accepts the following options:
8857 Set low and high threshold values used by the Canny thresholding
8860 The high threshold selects the "strong" edge pixels, which are then
8861 connected through 8-connectivity with the "weak" edge pixels selected
8862 by the low threshold.
8864 @var{low} and @var{high} threshold values must be chosen in the range
8865 [0,1], and @var{low} should be lesser or equal to @var{high}.
8867 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8871 Define the drawing mode.
8875 Draw white/gray wires on black background.
8878 Mix the colors to create a paint/cartoon effect.
8881 Apply Canny edge detector on all selected planes.
8883 Default value is @var{wires}.
8886 Select planes for filtering. By default all available planes are filtered.
8889 @subsection Examples
8893 Standard edge detection with custom values for the hysteresis thresholding:
8895 edgedetect=low=0.1:high=0.4
8899 Painting effect without thresholding:
8901 edgedetect=mode=colormix:high=0
8906 Set brightness, contrast, saturation and approximate gamma adjustment.
8908 The filter accepts the following options:
8912 Set the contrast expression. The value must be a float value in range
8913 @code{-2.0} to @code{2.0}. The default value is "1".
8916 Set the brightness expression. The value must be a float value in
8917 range @code{-1.0} to @code{1.0}. The default value is "0".
8920 Set the saturation expression. The value must be a float in
8921 range @code{0.0} to @code{3.0}. The default value is "1".
8924 Set the gamma expression. The value must be a float in range
8925 @code{0.1} to @code{10.0}. The default value is "1".
8928 Set the gamma expression for red. The value must be a float in
8929 range @code{0.1} to @code{10.0}. The default value is "1".
8932 Set the gamma expression for green. The value must be a float in range
8933 @code{0.1} to @code{10.0}. The default value is "1".
8936 Set the gamma expression for blue. The value must be a float in range
8937 @code{0.1} to @code{10.0}. The default value is "1".
8940 Set the gamma weight expression. It can be used to reduce the effect
8941 of a high gamma value on bright image areas, e.g. keep them from
8942 getting overamplified and just plain white. The value must be a float
8943 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8944 gamma correction all the way down while @code{1.0} leaves it at its
8945 full strength. Default is "1".
8948 Set when the expressions for brightness, contrast, saturation and
8949 gamma expressions are evaluated.
8951 It accepts the following values:
8954 only evaluate expressions once during the filter initialization or
8955 when a command is processed
8958 evaluate expressions for each incoming frame
8961 Default value is @samp{init}.
8964 The expressions accept the following parameters:
8967 frame count of the input frame starting from 0
8970 byte position of the corresponding packet in the input file, NAN if
8974 frame rate of the input video, NAN if the input frame rate is unknown
8977 timestamp expressed in seconds, NAN if the input timestamp is unknown
8980 @subsection Commands
8981 The filter supports the following commands:
8985 Set the contrast expression.
8988 Set the brightness expression.
8991 Set the saturation expression.
8994 Set the gamma expression.
8997 Set the gamma_r expression.
9000 Set gamma_g expression.
9003 Set gamma_b expression.
9006 Set gamma_weight expression.
9008 The command accepts the same syntax of the corresponding option.
9010 If the specified expression is not valid, it is kept at its current
9017 Apply erosion effect to the video.
9019 This filter replaces the pixel by the local(3x3) minimum.
9021 It accepts the following options:
9028 Limit the maximum change for each plane, default is 65535.
9029 If 0, plane will remain unchanged.
9032 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9035 Flags to local 3x3 coordinates maps like this:
9042 @section extractplanes
9044 Extract color channel components from input video stream into
9045 separate grayscale video streams.
9047 The filter accepts the following option:
9051 Set plane(s) to extract.
9053 Available values for planes are:
9064 Choosing planes not available in the input will result in an error.
9065 That means you cannot select @code{r}, @code{g}, @code{b} planes
9066 with @code{y}, @code{u}, @code{v} planes at same time.
9069 @subsection Examples
9073 Extract luma, u and v color channel component from input video frame
9074 into 3 grayscale outputs:
9076 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
9082 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9084 For each input image, the filter will compute the optimal mapping from
9085 the input to the output given the codebook length, that is the number
9086 of distinct output colors.
9088 This filter accepts the following options.
9091 @item codebook_length, l
9092 Set codebook length. The value must be a positive integer, and
9093 represents the number of distinct output colors. Default value is 256.
9096 Set the maximum number of iterations to apply for computing the optimal
9097 mapping. The higher the value the better the result and the higher the
9098 computation time. Default value is 1.
9101 Set a random seed, must be an integer included between 0 and
9102 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9103 will try to use a good random seed on a best effort basis.
9106 Set pal8 output pixel format. This option does not work with codebook
9107 length greater than 256.
9112 Measure graylevel entropy in histogram of color channels of video frames.
9114 It accepts the following parameters:
9118 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9120 @var{diff} mode measures entropy of histogram delta values, absolute differences
9121 between neighbour histogram values.
9126 Apply a fade-in/out effect to the input video.
9128 It accepts the following parameters:
9132 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9134 Default is @code{in}.
9136 @item start_frame, s
9137 Specify the number of the frame to start applying the fade
9138 effect at. Default is 0.
9141 The number of frames that the fade effect lasts. At the end of the
9142 fade-in effect, the output video will have the same intensity as the input video.
9143 At the end of the fade-out transition, the output video will be filled with the
9144 selected @option{color}.
9148 If set to 1, fade only alpha channel, if one exists on the input.
9151 @item start_time, st
9152 Specify the timestamp (in seconds) of the frame to start to apply the fade
9153 effect. If both start_frame and start_time are specified, the fade will start at
9154 whichever comes last. Default is 0.
9157 The number of seconds for which the fade effect has to last. At the end of the
9158 fade-in effect the output video will have the same intensity as the input video,
9159 at the end of the fade-out transition the output video will be filled with the
9160 selected @option{color}.
9161 If both duration and nb_frames are specified, duration is used. Default is 0
9162 (nb_frames is used by default).
9165 Specify the color of the fade. Default is "black".
9168 @subsection Examples
9172 Fade in the first 30 frames of video:
9177 The command above is equivalent to:
9183 Fade out the last 45 frames of a 200-frame video:
9186 fade=type=out:start_frame=155:nb_frames=45
9190 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9192 fade=in:0:25, fade=out:975:25
9196 Make the first 5 frames yellow, then fade in from frame 5-24:
9198 fade=in:5:20:color=yellow
9202 Fade in alpha over first 25 frames of video:
9204 fade=in:0:25:alpha=1
9208 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9210 fade=t=in:st=5.5:d=0.5
9216 Apply arbitrary expressions to samples in frequency domain
9220 Adjust the dc value (gain) of the luma plane of the image. The filter
9221 accepts an integer value in range @code{0} to @code{1000}. The default
9222 value is set to @code{0}.
9225 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9226 filter accepts an integer value in range @code{0} to @code{1000}. The
9227 default value is set to @code{0}.
9230 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9231 filter accepts an integer value in range @code{0} to @code{1000}. The
9232 default value is set to @code{0}.
9235 Set the frequency domain weight expression for the luma plane.
9238 Set the frequency domain weight expression for the 1st chroma plane.
9241 Set the frequency domain weight expression for the 2nd chroma plane.
9244 Set when the expressions are evaluated.
9246 It accepts the following values:
9249 Only evaluate expressions once during the filter initialization.
9252 Evaluate expressions for each incoming frame.
9255 Default value is @samp{init}.
9257 The filter accepts the following variables:
9260 The coordinates of the current sample.
9264 The width and height of the image.
9267 The number of input frame, starting from 0.
9270 @subsection Examples
9276 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9282 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9288 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9294 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9300 Denoise frames using 3D FFT (frequency domain filtering).
9302 The filter accepts the following options:
9306 Set the noise sigma constant. This sets denoising strength.
9307 Default value is 1. Allowed range is from 0 to 30.
9308 Using very high sigma with low overlap may give blocking artifacts.
9311 Set amount of denoising. By default all detected noise is reduced.
9312 Default value is 1. Allowed range is from 0 to 1.
9315 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9316 Actual size of block in pixels is 2 to power of @var{block}, so by default
9317 block size in pixels is 2^4 which is 16.
9320 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9323 Set number of previous frames to use for denoising. By default is set to 0.
9326 Set number of next frames to to use for denoising. By default is set to 0.
9329 Set planes which will be filtered, by default are all available filtered
9335 Extract a single field from an interlaced image using stride
9336 arithmetic to avoid wasting CPU time. The output frames are marked as
9339 The filter accepts the following options:
9343 Specify whether to extract the top (if the value is @code{0} or
9344 @code{top}) or the bottom field (if the value is @code{1} or
9350 Create new frames by copying the top and bottom fields from surrounding frames
9351 supplied as numbers by the hint file.
9355 Set file containing hints: absolute/relative frame numbers.
9357 There must be one line for each frame in a clip. Each line must contain two
9358 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9359 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9360 is current frame number for @code{absolute} mode or out of [-1, 1] range
9361 for @code{relative} mode. First number tells from which frame to pick up top
9362 field and second number tells from which frame to pick up bottom field.
9364 If optionally followed by @code{+} output frame will be marked as interlaced,
9365 else if followed by @code{-} output frame will be marked as progressive, else
9366 it will be marked same as input frame.
9367 If line starts with @code{#} or @code{;} that line is skipped.
9370 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9373 Example of first several lines of @code{hint} file for @code{relative} mode:
9376 1,0 - # second frame, use third's frame top field and second's frame bottom field
9377 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9394 Field matching filter for inverse telecine. It is meant to reconstruct the
9395 progressive frames from a telecined stream. The filter does not drop duplicated
9396 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9397 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9399 The separation of the field matching and the decimation is notably motivated by
9400 the possibility of inserting a de-interlacing filter fallback between the two.
9401 If the source has mixed telecined and real interlaced content,
9402 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9403 But these remaining combed frames will be marked as interlaced, and thus can be
9404 de-interlaced by a later filter such as @ref{yadif} before decimation.
9406 In addition to the various configuration options, @code{fieldmatch} can take an
9407 optional second stream, activated through the @option{ppsrc} option. If
9408 enabled, the frames reconstruction will be based on the fields and frames from
9409 this second stream. This allows the first input to be pre-processed in order to
9410 help the various algorithms of the filter, while keeping the output lossless
9411 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9412 or brightness/contrast adjustments can help.
9414 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9415 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9416 which @code{fieldmatch} is based on. While the semantic and usage are very
9417 close, some behaviour and options names can differ.
9419 The @ref{decimate} filter currently only works for constant frame rate input.
9420 If your input has mixed telecined (30fps) and progressive content with a lower
9421 framerate like 24fps use the following filterchain to produce the necessary cfr
9422 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9424 The filter accepts the following options:
9428 Specify the assumed field order of the input stream. Available values are:
9432 Auto detect parity (use FFmpeg's internal parity value).
9434 Assume bottom field first.
9436 Assume top field first.
9439 Note that it is sometimes recommended not to trust the parity announced by the
9442 Default value is @var{auto}.
9445 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9446 sense that it won't risk creating jerkiness due to duplicate frames when
9447 possible, but if there are bad edits or blended fields it will end up
9448 outputting combed frames when a good match might actually exist. On the other
9449 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9450 but will almost always find a good frame if there is one. The other values are
9451 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9452 jerkiness and creating duplicate frames versus finding good matches in sections
9453 with bad edits, orphaned fields, blended fields, etc.
9455 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9457 Available values are:
9461 2-way matching (p/c)
9463 2-way matching, and trying 3rd match if still combed (p/c + n)
9465 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9467 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9468 still combed (p/c + n + u/b)
9470 3-way matching (p/c/n)
9472 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9473 detected as combed (p/c/n + u/b)
9476 The parenthesis at the end indicate the matches that would be used for that
9477 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9480 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9483 Default value is @var{pc_n}.
9486 Mark the main input stream as a pre-processed input, and enable the secondary
9487 input stream as the clean source to pick the fields from. See the filter
9488 introduction for more details. It is similar to the @option{clip2} feature from
9491 Default value is @code{0} (disabled).
9494 Set the field to match from. It is recommended to set this to the same value as
9495 @option{order} unless you experience matching failures with that setting. In
9496 certain circumstances changing the field that is used to match from can have a
9497 large impact on matching performance. Available values are:
9501 Automatic (same value as @option{order}).
9503 Match from the bottom field.
9505 Match from the top field.
9508 Default value is @var{auto}.
9511 Set whether or not chroma is included during the match comparisons. In most
9512 cases it is recommended to leave this enabled. You should set this to @code{0}
9513 only if your clip has bad chroma problems such as heavy rainbowing or other
9514 artifacts. Setting this to @code{0} could also be used to speed things up at
9515 the cost of some accuracy.
9517 Default value is @code{1}.
9521 These define an exclusion band which excludes the lines between @option{y0} and
9522 @option{y1} from being included in the field matching decision. An exclusion
9523 band can be used to ignore subtitles, a logo, or other things that may
9524 interfere with the matching. @option{y0} sets the starting scan line and
9525 @option{y1} sets the ending line; all lines in between @option{y0} and
9526 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9527 @option{y0} and @option{y1} to the same value will disable the feature.
9528 @option{y0} and @option{y1} defaults to @code{0}.
9531 Set the scene change detection threshold as a percentage of maximum change on
9532 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9533 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9534 @option{scthresh} is @code{[0.0, 100.0]}.
9536 Default value is @code{12.0}.
9539 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9540 account the combed scores of matches when deciding what match to use as the
9541 final match. Available values are:
9545 No final matching based on combed scores.
9547 Combed scores are only used when a scene change is detected.
9549 Use combed scores all the time.
9552 Default is @var{sc}.
9555 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9556 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9557 Available values are:
9561 No forced calculation.
9563 Force p/c/n calculations.
9565 Force p/c/n/u/b calculations.
9568 Default value is @var{none}.
9571 This is the area combing threshold used for combed frame detection. This
9572 essentially controls how "strong" or "visible" combing must be to be detected.
9573 Larger values mean combing must be more visible and smaller values mean combing
9574 can be less visible or strong and still be detected. Valid settings are from
9575 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9576 be detected as combed). This is basically a pixel difference value. A good
9577 range is @code{[8, 12]}.
9579 Default value is @code{9}.
9582 Sets whether or not chroma is considered in the combed frame decision. Only
9583 disable this if your source has chroma problems (rainbowing, etc.) that are
9584 causing problems for the combed frame detection with chroma enabled. Actually,
9585 using @option{chroma}=@var{0} is usually more reliable, except for the case
9586 where there is chroma only combing in the source.
9588 Default value is @code{0}.
9592 Respectively set the x-axis and y-axis size of the window used during combed
9593 frame detection. This has to do with the size of the area in which
9594 @option{combpel} pixels are required to be detected as combed for a frame to be
9595 declared combed. See the @option{combpel} parameter description for more info.
9596 Possible values are any number that is a power of 2 starting at 4 and going up
9599 Default value is @code{16}.
9602 The number of combed pixels inside any of the @option{blocky} by
9603 @option{blockx} size blocks on the frame for the frame to be detected as
9604 combed. While @option{cthresh} controls how "visible" the combing must be, this
9605 setting controls "how much" combing there must be in any localized area (a
9606 window defined by the @option{blockx} and @option{blocky} settings) on the
9607 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9608 which point no frames will ever be detected as combed). This setting is known
9609 as @option{MI} in TFM/VFM vocabulary.
9611 Default value is @code{80}.
9614 @anchor{p/c/n/u/b meaning}
9615 @subsection p/c/n/u/b meaning
9617 @subsubsection p/c/n
9619 We assume the following telecined stream:
9622 Top fields: 1 2 2 3 4
9623 Bottom fields: 1 2 3 4 4
9626 The numbers correspond to the progressive frame the fields relate to. Here, the
9627 first two frames are progressive, the 3rd and 4th are combed, and so on.
9629 When @code{fieldmatch} is configured to run a matching from bottom
9630 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9635 B 1 2 3 4 4 <-- matching reference
9644 As a result of the field matching, we can see that some frames get duplicated.
9645 To perform a complete inverse telecine, you need to rely on a decimation filter
9646 after this operation. See for instance the @ref{decimate} filter.
9648 The same operation now matching from top fields (@option{field}=@var{top})
9653 T 1 2 2 3 4 <-- matching reference
9663 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9664 basically, they refer to the frame and field of the opposite parity:
9667 @item @var{p} matches the field of the opposite parity in the previous frame
9668 @item @var{c} matches the field of the opposite parity in the current frame
9669 @item @var{n} matches the field of the opposite parity in the next frame
9674 The @var{u} and @var{b} matching are a bit special in the sense that they match
9675 from the opposite parity flag. In the following examples, we assume that we are
9676 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9677 'x' is placed above and below each matched fields.
9679 With bottom matching (@option{field}=@var{bottom}):
9684 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9685 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9693 With top matching (@option{field}=@var{top}):
9698 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9699 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9707 @subsection Examples
9709 Simple IVTC of a top field first telecined stream:
9711 fieldmatch=order=tff:combmatch=none, decimate
9714 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9716 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9721 Transform the field order of the input video.
9723 It accepts the following parameters:
9728 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9729 for bottom field first.
9732 The default value is @samp{tff}.
9734 The transformation is done by shifting the picture content up or down
9735 by one line, and filling the remaining line with appropriate picture content.
9736 This method is consistent with most broadcast field order converters.
9738 If the input video is not flagged as being interlaced, or it is already
9739 flagged as being of the required output field order, then this filter does
9740 not alter the incoming video.
9742 It is very useful when converting to or from PAL DV material,
9743 which is bottom field first.
9747 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9750 @section fifo, afifo
9752 Buffer input images and send them when they are requested.
9754 It is mainly useful when auto-inserted by the libavfilter
9757 It does not take parameters.
9759 @section fillborders
9761 Fill borders of the input video, without changing video stream dimensions.
9762 Sometimes video can have garbage at the four edges and you may not want to
9763 crop video input to keep size multiple of some number.
9765 This filter accepts the following options:
9769 Number of pixels to fill from left border.
9772 Number of pixels to fill from right border.
9775 Number of pixels to fill from top border.
9778 Number of pixels to fill from bottom border.
9783 It accepts the following values:
9786 fill pixels using outermost pixels
9789 fill pixels using mirroring
9792 fill pixels with constant value
9795 Default is @var{smear}.
9798 Set color for pixels in fixed mode. Default is @var{black}.
9803 Find a rectangular object
9805 It accepts the following options:
9809 Filepath of the object image, needs to be in gray8.
9812 Detection threshold, default is 0.5.
9815 Number of mipmaps, default is 3.
9817 @item xmin, ymin, xmax, ymax
9818 Specifies the rectangle in which to search.
9821 @subsection Examples
9825 Generate a representative palette of a given video using @command{ffmpeg}:
9827 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9833 Cover a rectangular object
9835 It accepts the following options:
9839 Filepath of the optional cover image, needs to be in yuv420.
9844 It accepts the following values:
9847 cover it by the supplied image
9849 cover it by interpolating the surrounding pixels
9852 Default value is @var{blur}.
9855 @subsection Examples
9859 Generate a representative palette of a given video using @command{ffmpeg}:
9861 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9867 Flood area with values of same pixel components with another values.
9869 It accepts the following options:
9872 Set pixel x coordinate.
9875 Set pixel y coordinate.
9878 Set source #0 component value.
9881 Set source #1 component value.
9884 Set source #2 component value.
9887 Set source #3 component value.
9890 Set destination #0 component value.
9893 Set destination #1 component value.
9896 Set destination #2 component value.
9899 Set destination #3 component value.
9905 Convert the input video to one of the specified pixel formats.
9906 Libavfilter will try to pick one that is suitable as input to
9909 It accepts the following parameters:
9913 A '|'-separated list of pixel format names, such as
9914 "pix_fmts=yuv420p|monow|rgb24".
9918 @subsection Examples
9922 Convert the input video to the @var{yuv420p} format
9924 format=pix_fmts=yuv420p
9927 Convert the input video to any of the formats in the list
9929 format=pix_fmts=yuv420p|yuv444p|yuv410p
9936 Convert the video to specified constant frame rate by duplicating or dropping
9937 frames as necessary.
9939 It accepts the following parameters:
9943 The desired output frame rate. The default is @code{25}.
9946 Assume the first PTS should be the given value, in seconds. This allows for
9947 padding/trimming at the start of stream. By default, no assumption is made
9948 about the first frame's expected PTS, so no padding or trimming is done.
9949 For example, this could be set to 0 to pad the beginning with duplicates of
9950 the first frame if a video stream starts after the audio stream or to trim any
9951 frames with a negative PTS.
9954 Timestamp (PTS) rounding method.
9956 Possible values are:
9963 round towards -infinity
9965 round towards +infinity
9969 The default is @code{near}.
9972 Action performed when reading the last frame.
9974 Possible values are:
9977 Use same timestamp rounding method as used for other frames.
9979 Pass through last frame if input duration has not been reached yet.
9981 The default is @code{round}.
9985 Alternatively, the options can be specified as a flat string:
9986 @var{fps}[:@var{start_time}[:@var{round}]].
9988 See also the @ref{setpts} filter.
9990 @subsection Examples
9994 A typical usage in order to set the fps to 25:
10000 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10002 fps=fps=film:round=near
10008 Pack two different video streams into a stereoscopic video, setting proper
10009 metadata on supported codecs. The two views should have the same size and
10010 framerate and processing will stop when the shorter video ends. Please note
10011 that you may conveniently adjust view properties with the @ref{scale} and
10014 It accepts the following parameters:
10018 The desired packing format. Supported values are:
10023 The views are next to each other (default).
10026 The views are on top of each other.
10029 The views are packed by line.
10032 The views are packed by column.
10035 The views are temporally interleaved.
10044 # Convert left and right views into a frame-sequential video
10045 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10047 # Convert views into a side-by-side video with the same output resolution as the input
10048 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
10053 Change the frame rate by interpolating new video output frames from the source
10056 This filter is not designed to function correctly with interlaced media. If
10057 you wish to change the frame rate of interlaced media then you are required
10058 to deinterlace before this filter and re-interlace after this filter.
10060 A description of the accepted options follows.
10064 Specify the output frames per second. This option can also be specified
10065 as a value alone. The default is @code{50}.
10068 Specify the start of a range where the output frame will be created as a
10069 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10070 the default is @code{15}.
10073 Specify the end of a range where the output frame will be created as a
10074 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10075 the default is @code{240}.
10078 Specify the level at which a scene change is detected as a value between
10079 0 and 100 to indicate a new scene; a low value reflects a low
10080 probability for the current frame to introduce a new scene, while a higher
10081 value means the current frame is more likely to be one.
10082 The default is @code{8.2}.
10085 Specify flags influencing the filter process.
10087 Available value for @var{flags} is:
10090 @item scene_change_detect, scd
10091 Enable scene change detection using the value of the option @var{scene}.
10092 This flag is enabled by default.
10098 Select one frame every N-th frame.
10100 This filter accepts the following option:
10103 Select frame after every @code{step} frames.
10104 Allowed values are positive integers higher than 0. Default value is @code{1}.
10107 @section freezedetect
10109 Detect frozen video.
10111 This filter logs a message and sets frame metadata when it detects that the
10112 input video has no significant change in content during a specified duration.
10113 Video freeze detection calculates the mean average absolute difference of all
10114 the components of video frames and compares it to a noise floor.
10116 The printed times and duration are expressed in seconds. The
10117 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10118 whose timestamp equals or exceeds the detection duration and it contains the
10119 timestamp of the first frame of the freeze. The
10120 @code{lavfi.freezedetect.freeze_duration} and
10121 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10124 The filter accepts the following options:
10128 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10129 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10133 Set freeze duration until notification (default is 2 seconds).
10139 Apply a frei0r effect to the input video.
10141 To enable the compilation of this filter, you need to install the frei0r
10142 header and configure FFmpeg with @code{--enable-frei0r}.
10144 It accepts the following parameters:
10149 The name of the frei0r effect to load. If the environment variable
10150 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10151 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10152 Otherwise, the standard frei0r paths are searched, in this order:
10153 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10154 @file{/usr/lib/frei0r-1/}.
10156 @item filter_params
10157 A '|'-separated list of parameters to pass to the frei0r effect.
10161 A frei0r effect parameter can be a boolean (its value is either
10162 "y" or "n"), a double, a color (specified as
10163 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10164 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10165 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10166 a position (specified as @var{X}/@var{Y}, where
10167 @var{X} and @var{Y} are floating point numbers) and/or a string.
10169 The number and types of parameters depend on the loaded effect. If an
10170 effect parameter is not specified, the default value is set.
10172 @subsection Examples
10176 Apply the distort0r effect, setting the first two double parameters:
10178 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10182 Apply the colordistance effect, taking a color as the first parameter:
10184 frei0r=colordistance:0.2/0.3/0.4
10185 frei0r=colordistance:violet
10186 frei0r=colordistance:0x112233
10190 Apply the perspective effect, specifying the top left and top right image
10193 frei0r=perspective:0.2/0.2|0.8/0.2
10197 For more information, see
10198 @url{http://frei0r.dyne.org}
10202 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10204 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10205 processing filter, one of them is performed once per block, not per pixel.
10206 This allows for much higher speed.
10208 The filter accepts the following options:
10212 Set quality. This option defines the number of levels for averaging. It accepts
10213 an integer in the range 4-5. Default value is @code{4}.
10216 Force a constant quantization parameter. It accepts an integer in range 0-63.
10217 If not set, the filter will use the QP from the video stream (if available).
10220 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10221 more details but also more artifacts, while higher values make the image smoother
10222 but also blurrier. Default value is @code{0} − PSNR optimal.
10224 @item use_bframe_qp
10225 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10226 option may cause flicker since the B-Frames have often larger QP. Default is
10227 @code{0} (not enabled).
10233 Apply Gaussian blur filter.
10235 The filter accepts the following options:
10239 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10242 Set number of steps for Gaussian approximation. Defauls is @code{1}.
10245 Set which planes to filter. By default all planes are filtered.
10248 Set vertical sigma, if negative it will be same as @code{sigma}.
10249 Default is @code{-1}.
10254 Apply generic equation to each pixel.
10256 The filter accepts the following options:
10259 @item lum_expr, lum
10260 Set the luminance expression.
10262 Set the chrominance blue expression.
10264 Set the chrominance red expression.
10265 @item alpha_expr, a
10266 Set the alpha expression.
10268 Set the red expression.
10269 @item green_expr, g
10270 Set the green expression.
10272 Set the blue expression.
10275 The colorspace is selected according to the specified options. If one
10276 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10277 options is specified, the filter will automatically select a YCbCr
10278 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10279 @option{blue_expr} options is specified, it will select an RGB
10282 If one of the chrominance expression is not defined, it falls back on the other
10283 one. If no alpha expression is specified it will evaluate to opaque value.
10284 If none of chrominance expressions are specified, they will evaluate
10285 to the luminance expression.
10287 The expressions can use the following variables and functions:
10291 The sequential number of the filtered frame, starting from @code{0}.
10295 The coordinates of the current sample.
10299 The width and height of the image.
10303 Width and height scale depending on the currently filtered plane. It is the
10304 ratio between the corresponding luma plane number of pixels and the current
10305 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10306 @code{0.5,0.5} for chroma planes.
10309 Time of the current frame, expressed in seconds.
10312 Return the value of the pixel at location (@var{x},@var{y}) of the current
10316 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10320 Return the value of the pixel at location (@var{x},@var{y}) of the
10321 blue-difference chroma plane. Return 0 if there is no such plane.
10324 Return the value of the pixel at location (@var{x},@var{y}) of the
10325 red-difference chroma plane. Return 0 if there is no such plane.
10330 Return the value of the pixel at location (@var{x},@var{y}) of the
10331 red/green/blue component. Return 0 if there is no such component.
10334 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10335 plane. Return 0 if there is no such plane.
10338 For functions, if @var{x} and @var{y} are outside the area, the value will be
10339 automatically clipped to the closer edge.
10341 @subsection Examples
10345 Flip the image horizontally:
10351 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10352 wavelength of 100 pixels:
10354 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10358 Generate a fancy enigmatic moving light:
10360 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
10364 Generate a quick emboss effect:
10366 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10370 Modify RGB components depending on pixel position:
10372 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10376 Create a radial gradient that is the same size as the input (also see
10377 the @ref{vignette} filter):
10379 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10385 Fix the banding artifacts that are sometimes introduced into nearly flat
10386 regions by truncation to 8-bit color depth.
10387 Interpolate the gradients that should go where the bands are, and
10390 It is designed for playback only. Do not use it prior to
10391 lossy compression, because compression tends to lose the dither and
10392 bring back the bands.
10394 It accepts the following parameters:
10399 The maximum amount by which the filter will change any one pixel. This is also
10400 the threshold for detecting nearly flat regions. Acceptable values range from
10401 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10405 The neighborhood to fit the gradient to. A larger radius makes for smoother
10406 gradients, but also prevents the filter from modifying the pixels near detailed
10407 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10408 values will be clipped to the valid range.
10412 Alternatively, the options can be specified as a flat string:
10413 @var{strength}[:@var{radius}]
10415 @subsection Examples
10419 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10425 Specify radius, omitting the strength (which will fall-back to the default
10433 @section graphmonitor, agraphmonitor
10434 Show various filtergraph stats.
10436 With this filter one can debug complete filtergraph.
10437 Especially issues with links filling with queued frames.
10439 The filter accepts the following options:
10443 Set video output size. Default is @var{hd720}.
10446 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10449 Set output mode, can be @var{fulll} or @var{compact}.
10450 In @var{compact} mode only filters with some queued frames have displayed stats.
10453 Set flags which enable which stats are shown in video.
10455 Available values for flags are:
10458 Display number of queued frames in each link.
10460 @item frame_count_in
10461 Display number of frames taken from filter.
10463 @item frame_count_out
10464 Display number of frames given out from filter.
10467 Display current filtered frame pts.
10470 Display current filtered frame time.
10473 Display time base for filter link.
10476 Display used format for filter link.
10479 Display video size or number of audio channels in case of audio used by filter link.
10482 Display video frame rate or sample rate in case of audio used by filter link.
10486 Set upper limit for video rate of output stream, Default value is @var{25}.
10487 This guarantee that output video frame rate will not be higher than this value.
10491 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10492 and corrects the scene colors accordingly.
10494 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10496 The filter accepts the following options:
10500 The order of differentiation to be applied on the scene. Must be chosen in the range
10501 [0,2] and default value is 1.
10504 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10505 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10506 max value instead of calculating Minkowski distance.
10509 The standard deviation of Gaussian blur to be applied on the scene. Must be
10510 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10511 can't be euqal to 0 if @var{difford} is greater than 0.
10514 @subsection Examples
10520 greyedge=difford=1:minknorm=5:sigma=2
10526 greyedge=difford=1:minknorm=0:sigma=2
10534 Apply a Hald CLUT to a video stream.
10536 First input is the video stream to process, and second one is the Hald CLUT.
10537 The Hald CLUT input can be a simple picture or a complete video stream.
10539 The filter accepts the following options:
10543 Force termination when the shortest input terminates. Default is @code{0}.
10545 Continue applying the last CLUT after the end of the stream. A value of
10546 @code{0} disable the filter after the last frame of the CLUT is reached.
10547 Default is @code{1}.
10550 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10551 filters share the same internals).
10553 More information about the Hald CLUT can be found on Eskil Steenberg's website
10554 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10556 @subsection Workflow examples
10558 @subsubsection Hald CLUT video stream
10560 Generate an identity Hald CLUT stream altered with various effects:
10562 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
10565 Note: make sure you use a lossless codec.
10567 Then use it with @code{haldclut} to apply it on some random stream:
10569 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10572 The Hald CLUT will be applied to the 10 first seconds (duration of
10573 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10574 to the remaining frames of the @code{mandelbrot} stream.
10576 @subsubsection Hald CLUT with preview
10578 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10579 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10580 biggest possible square starting at the top left of the picture. The remaining
10581 padding pixels (bottom or right) will be ignored. This area can be used to add
10582 a preview of the Hald CLUT.
10584 Typically, the following generated Hald CLUT will be supported by the
10585 @code{haldclut} filter:
10588 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10589 pad=iw+320 [padded_clut];
10590 smptebars=s=320x256, split [a][b];
10591 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10592 [main][b] overlay=W-320" -frames:v 1 clut.png
10595 It contains the original and a preview of the effect of the CLUT: SMPTE color
10596 bars are displayed on the right-top, and below the same color bars processed by
10599 Then, the effect of this Hald CLUT can be visualized with:
10601 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10606 Flip the input video horizontally.
10608 For example, to horizontally flip the input video with @command{ffmpeg}:
10610 ffmpeg -i in.avi -vf "hflip" out.avi
10614 This filter applies a global color histogram equalization on a
10617 It can be used to correct video that has a compressed range of pixel
10618 intensities. The filter redistributes the pixel intensities to
10619 equalize their distribution across the intensity range. It may be
10620 viewed as an "automatically adjusting contrast filter". This filter is
10621 useful only for correcting degraded or poorly captured source
10624 The filter accepts the following options:
10628 Determine the amount of equalization to be applied. As the strength
10629 is reduced, the distribution of pixel intensities more-and-more
10630 approaches that of the input frame. The value must be a float number
10631 in the range [0,1] and defaults to 0.200.
10634 Set the maximum intensity that can generated and scale the output
10635 values appropriately. The strength should be set as desired and then
10636 the intensity can be limited if needed to avoid washing-out. The value
10637 must be a float number in the range [0,1] and defaults to 0.210.
10640 Set the antibanding level. If enabled the filter will randomly vary
10641 the luminance of output pixels by a small amount to avoid banding of
10642 the histogram. Possible values are @code{none}, @code{weak} or
10643 @code{strong}. It defaults to @code{none}.
10648 Compute and draw a color distribution histogram for the input video.
10650 The computed histogram is a representation of the color component
10651 distribution in an image.
10653 Standard histogram displays the color components distribution in an image.
10654 Displays color graph for each color component. Shows distribution of
10655 the Y, U, V, A or R, G, B components, depending on input format, in the
10656 current frame. Below each graph a color component scale meter is shown.
10658 The filter accepts the following options:
10662 Set height of level. Default value is @code{200}.
10663 Allowed range is [50, 2048].
10666 Set height of color scale. Default value is @code{12}.
10667 Allowed range is [0, 40].
10671 It accepts the following values:
10674 Per color component graphs are placed below each other.
10677 Per color component graphs are placed side by side.
10680 Presents information identical to that in the @code{parade}, except
10681 that the graphs representing color components are superimposed directly
10684 Default is @code{stack}.
10687 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10688 Default is @code{linear}.
10691 Set what color components to display.
10692 Default is @code{7}.
10695 Set foreground opacity. Default is @code{0.7}.
10698 Set background opacity. Default is @code{0.5}.
10701 @subsection Examples
10706 Calculate and draw histogram:
10708 ffplay -i input -vf histogram
10716 This is a high precision/quality 3d denoise filter. It aims to reduce
10717 image noise, producing smooth images and making still images really
10718 still. It should enhance compressibility.
10720 It accepts the following optional parameters:
10724 A non-negative floating point number which specifies spatial luma strength.
10725 It defaults to 4.0.
10727 @item chroma_spatial
10728 A non-negative floating point number which specifies spatial chroma strength.
10729 It defaults to 3.0*@var{luma_spatial}/4.0.
10732 A floating point number which specifies luma temporal strength. It defaults to
10733 6.0*@var{luma_spatial}/4.0.
10736 A floating point number which specifies chroma temporal strength. It defaults to
10737 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10740 @anchor{hwdownload}
10741 @section hwdownload
10743 Download hardware frames to system memory.
10745 The input must be in hardware frames, and the output a non-hardware format.
10746 Not all formats will be supported on the output - it may be necessary to insert
10747 an additional @option{format} filter immediately following in the graph to get
10748 the output in a supported format.
10752 Map hardware frames to system memory or to another device.
10754 This filter has several different modes of operation; which one is used depends
10755 on the input and output formats:
10758 Hardware frame input, normal frame output
10760 Map the input frames to system memory and pass them to the output. If the
10761 original hardware frame is later required (for example, after overlaying
10762 something else on part of it), the @option{hwmap} filter can be used again
10763 in the next mode to retrieve it.
10765 Normal frame input, hardware frame output
10767 If the input is actually a software-mapped hardware frame, then unmap it -
10768 that is, return the original hardware frame.
10770 Otherwise, a device must be provided. Create new hardware surfaces on that
10771 device for the output, then map them back to the software format at the input
10772 and give those frames to the preceding filter. This will then act like the
10773 @option{hwupload} filter, but may be able to avoid an additional copy when
10774 the input is already in a compatible format.
10776 Hardware frame input and output
10778 A device must be supplied for the output, either directly or with the
10779 @option{derive_device} option. The input and output devices must be of
10780 different types and compatible - the exact meaning of this is
10781 system-dependent, but typically it means that they must refer to the same
10782 underlying hardware context (for example, refer to the same graphics card).
10784 If the input frames were originally created on the output device, then unmap
10785 to retrieve the original frames.
10787 Otherwise, map the frames to the output device - create new hardware frames
10788 on the output corresponding to the frames on the input.
10791 The following additional parameters are accepted:
10795 Set the frame mapping mode. Some combination of:
10798 The mapped frame should be readable.
10800 The mapped frame should be writeable.
10802 The mapping will always overwrite the entire frame.
10804 This may improve performance in some cases, as the original contents of the
10805 frame need not be loaded.
10807 The mapping must not involve any copying.
10809 Indirect mappings to copies of frames are created in some cases where either
10810 direct mapping is not possible or it would have unexpected properties.
10811 Setting this flag ensures that the mapping is direct and will fail if that is
10814 Defaults to @var{read+write} if not specified.
10816 @item derive_device @var{type}
10817 Rather than using the device supplied at initialisation, instead derive a new
10818 device of type @var{type} from the device the input frames exist on.
10821 In a hardware to hardware mapping, map in reverse - create frames in the sink
10822 and map them back to the source. This may be necessary in some cases where
10823 a mapping in one direction is required but only the opposite direction is
10824 supported by the devices being used.
10826 This option is dangerous - it may break the preceding filter in undefined
10827 ways if there are any additional constraints on that filter's output.
10828 Do not use it without fully understanding the implications of its use.
10834 Upload system memory frames to hardware surfaces.
10836 The device to upload to must be supplied when the filter is initialised. If
10837 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10840 @anchor{hwupload_cuda}
10841 @section hwupload_cuda
10843 Upload system memory frames to a CUDA device.
10845 It accepts the following optional parameters:
10849 The number of the CUDA device to use
10854 Apply a high-quality magnification filter designed for pixel art. This filter
10855 was originally created by Maxim Stepin.
10857 It accepts the following option:
10861 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10862 @code{hq3x} and @code{4} for @code{hq4x}.
10863 Default is @code{3}.
10867 Stack input videos horizontally.
10869 All streams must be of same pixel format and of same height.
10871 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10872 to create same output.
10874 The filter accept the following option:
10878 Set number of input streams. Default is 2.
10881 If set to 1, force the output to terminate when the shortest input
10882 terminates. Default value is 0.
10887 Modify the hue and/or the saturation of the input.
10889 It accepts the following parameters:
10893 Specify the hue angle as a number of degrees. It accepts an expression,
10894 and defaults to "0".
10897 Specify the saturation in the [-10,10] range. It accepts an expression and
10901 Specify the hue angle as a number of radians. It accepts an
10902 expression, and defaults to "0".
10905 Specify the brightness in the [-10,10] range. It accepts an expression and
10909 @option{h} and @option{H} are mutually exclusive, and can't be
10910 specified at the same time.
10912 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10913 expressions containing the following constants:
10917 frame count of the input frame starting from 0
10920 presentation timestamp of the input frame expressed in time base units
10923 frame rate of the input video, NAN if the input frame rate is unknown
10926 timestamp expressed in seconds, NAN if the input timestamp is unknown
10929 time base of the input video
10932 @subsection Examples
10936 Set the hue to 90 degrees and the saturation to 1.0:
10942 Same command but expressing the hue in radians:
10948 Rotate hue and make the saturation swing between 0
10949 and 2 over a period of 1 second:
10951 hue="H=2*PI*t: s=sin(2*PI*t)+1"
10955 Apply a 3 seconds saturation fade-in effect starting at 0:
10957 hue="s=min(t/3\,1)"
10960 The general fade-in expression can be written as:
10962 hue="s=min(0\, max((t-START)/DURATION\, 1))"
10966 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
10968 hue="s=max(0\, min(1\, (8-t)/3))"
10971 The general fade-out expression can be written as:
10973 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
10978 @subsection Commands
10980 This filter supports the following commands:
10986 Modify the hue and/or the saturation and/or brightness of the input video.
10987 The command accepts the same syntax of the corresponding option.
10989 If the specified expression is not valid, it is kept at its current
10993 @section hysteresis
10995 Grow first stream into second stream by connecting components.
10996 This makes it possible to build more robust edge masks.
10998 This filter accepts the following options:
11002 Set which planes will be processed as bitmap, unprocessed planes will be
11003 copied from first stream.
11004 By default value 0xf, all planes will be processed.
11007 Set threshold which is used in filtering. If pixel component value is higher than
11008 this value filter algorithm for connecting components is activated.
11009 By default value is 0.
11014 Detect video interlacing type.
11016 This filter tries to detect if the input frames are interlaced, progressive,
11017 top or bottom field first. It will also try to detect fields that are
11018 repeated between adjacent frames (a sign of telecine).
11020 Single frame detection considers only immediately adjacent frames when classifying each frame.
11021 Multiple frame detection incorporates the classification history of previous frames.
11023 The filter will log these metadata values:
11026 @item single.current_frame
11027 Detected type of current frame using single-frame detection. One of:
11028 ``tff'' (top field first), ``bff'' (bottom field first),
11029 ``progressive'', or ``undetermined''
11032 Cumulative number of frames detected as top field first using single-frame detection.
11035 Cumulative number of frames detected as top field first using multiple-frame detection.
11038 Cumulative number of frames detected as bottom field first using single-frame detection.
11040 @item multiple.current_frame
11041 Detected type of current frame using multiple-frame detection. One of:
11042 ``tff'' (top field first), ``bff'' (bottom field first),
11043 ``progressive'', or ``undetermined''
11046 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11048 @item single.progressive
11049 Cumulative number of frames detected as progressive using single-frame detection.
11051 @item multiple.progressive
11052 Cumulative number of frames detected as progressive using multiple-frame detection.
11054 @item single.undetermined
11055 Cumulative number of frames that could not be classified using single-frame detection.
11057 @item multiple.undetermined
11058 Cumulative number of frames that could not be classified using multiple-frame detection.
11060 @item repeated.current_frame
11061 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11063 @item repeated.neither
11064 Cumulative number of frames with no repeated field.
11067 Cumulative number of frames with the top field repeated from the previous frame's top field.
11069 @item repeated.bottom
11070 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11073 The filter accepts the following options:
11077 Set interlacing threshold.
11079 Set progressive threshold.
11081 Threshold for repeated field detection.
11083 Number of frames after which a given frame's contribution to the
11084 statistics is halved (i.e., it contributes only 0.5 to its
11085 classification). The default of 0 means that all frames seen are given
11086 full weight of 1.0 forever.
11087 @item analyze_interlaced_flag
11088 When this is not 0 then idet will use the specified number of frames to determine
11089 if the interlaced flag is accurate, it will not count undetermined frames.
11090 If the flag is found to be accurate it will be used without any further
11091 computations, if it is found to be inaccurate it will be cleared without any
11092 further computations. This allows inserting the idet filter as a low computational
11093 method to clean up the interlaced flag
11098 Deinterleave or interleave fields.
11100 This filter allows one to process interlaced images fields without
11101 deinterlacing them. Deinterleaving splits the input frame into 2
11102 fields (so called half pictures). Odd lines are moved to the top
11103 half of the output image, even lines to the bottom half.
11104 You can process (filter) them independently and then re-interleave them.
11106 The filter accepts the following options:
11110 @item chroma_mode, c
11111 @item alpha_mode, a
11112 Available values for @var{luma_mode}, @var{chroma_mode} and
11113 @var{alpha_mode} are:
11119 @item deinterleave, d
11120 Deinterleave fields, placing one above the other.
11122 @item interleave, i
11123 Interleave fields. Reverse the effect of deinterleaving.
11125 Default value is @code{none}.
11127 @item luma_swap, ls
11128 @item chroma_swap, cs
11129 @item alpha_swap, as
11130 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11135 Apply inflate effect to the video.
11137 This filter replaces the pixel by the local(3x3) average by taking into account
11138 only values higher than the pixel.
11140 It accepts the following options:
11147 Limit the maximum change for each plane, default is 65535.
11148 If 0, plane will remain unchanged.
11153 Simple interlacing filter from progressive contents. This interleaves upper (or
11154 lower) lines from odd frames with lower (or upper) lines from even frames,
11155 halving the frame rate and preserving image height.
11158 Original Original New Frame
11159 Frame 'j' Frame 'j+1' (tff)
11160 ========== =========== ==================
11161 Line 0 --------------------> Frame 'j' Line 0
11162 Line 1 Line 1 ----> Frame 'j+1' Line 1
11163 Line 2 ---------------------> Frame 'j' Line 2
11164 Line 3 Line 3 ----> Frame 'j+1' Line 3
11166 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11169 It accepts the following optional parameters:
11173 This determines whether the interlaced frame is taken from the even
11174 (tff - default) or odd (bff) lines of the progressive frame.
11177 Vertical lowpass filter to avoid twitter interlacing and
11178 reduce moire patterns.
11182 Disable vertical lowpass filter
11185 Enable linear filter (default)
11188 Enable complex filter. This will slightly less reduce twitter and moire
11189 but better retain detail and subjective sharpness impression.
11196 Deinterlace input video by applying Donald Graft's adaptive kernel
11197 deinterling. Work on interlaced parts of a video to produce
11198 progressive frames.
11200 The description of the accepted parameters follows.
11204 Set the threshold which affects the filter's tolerance when
11205 determining if a pixel line must be processed. It must be an integer
11206 in the range [0,255] and defaults to 10. A value of 0 will result in
11207 applying the process on every pixels.
11210 Paint pixels exceeding the threshold value to white if set to 1.
11214 Set the fields order. Swap fields if set to 1, leave fields alone if
11218 Enable additional sharpening if set to 1. Default is 0.
11221 Enable twoway sharpening if set to 1. Default is 0.
11224 @subsection Examples
11228 Apply default values:
11230 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11234 Enable additional sharpening:
11240 Paint processed pixels in white:
11246 @section lenscorrection
11248 Correct radial lens distortion
11250 This filter can be used to correct for radial distortion as can result from the use
11251 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11252 one can use tools available for example as part of opencv or simply trial-and-error.
11253 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11254 and extract the k1 and k2 coefficients from the resulting matrix.
11256 Note that effectively the same filter is available in the open-source tools Krita and
11257 Digikam from the KDE project.
11259 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11260 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11261 brightness distribution, so you may want to use both filters together in certain
11262 cases, though you will have to take care of ordering, i.e. whether vignetting should
11263 be applied before or after lens correction.
11265 @subsection Options
11267 The filter accepts the following options:
11271 Relative x-coordinate of the focal point of the image, and thereby the center of the
11272 distortion. This value has a range [0,1] and is expressed as fractions of the image
11273 width. Default is 0.5.
11275 Relative y-coordinate of the focal point of the image, and thereby the center of the
11276 distortion. This value has a range [0,1] and is expressed as fractions of the image
11277 height. Default is 0.5.
11279 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11280 no correction. Default is 0.
11282 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11283 0 means no correction. Default is 0.
11286 The formula that generates the correction is:
11288 @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)
11290 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11291 distances from the focal point in the source and target images, respectively.
11295 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11297 The @code{lensfun} filter requires the camera make, camera model, and lens model
11298 to apply the lens correction. The filter will load the lensfun database and
11299 query it to find the corresponding camera and lens entries in the database. As
11300 long as these entries can be found with the given options, the filter can
11301 perform corrections on frames. Note that incomplete strings will result in the
11302 filter choosing the best match with the given options, and the filter will
11303 output the chosen camera and lens models (logged with level "info"). You must
11304 provide the make, camera model, and lens model as they are required.
11306 The filter accepts the following options:
11310 The make of the camera (for example, "Canon"). This option is required.
11313 The model of the camera (for example, "Canon EOS 100D"). This option is
11317 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11318 option is required.
11321 The type of correction to apply. The following values are valid options:
11325 Enables fixing lens vignetting.
11328 Enables fixing lens geometry. This is the default.
11331 Enables fixing chromatic aberrations.
11334 Enables fixing lens vignetting and lens geometry.
11337 Enables fixing lens vignetting and chromatic aberrations.
11340 Enables fixing both lens geometry and chromatic aberrations.
11343 Enables all possible corrections.
11347 The focal length of the image/video (zoom; expected constant for video). For
11348 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11349 range should be chosen when using that lens. Default 18.
11352 The aperture of the image/video (expected constant for video). Note that
11353 aperture is only used for vignetting correction. Default 3.5.
11355 @item focus_distance
11356 The focus distance of the image/video (expected constant for video). Note that
11357 focus distance is only used for vignetting and only slightly affects the
11358 vignetting correction process. If unknown, leave it at the default value (which
11361 @item target_geometry
11362 The target geometry of the output image/video. The following values are valid
11366 @item rectilinear (default)
11369 @item equirectangular
11370 @item fisheye_orthographic
11371 @item fisheye_stereographic
11372 @item fisheye_equisolid
11373 @item fisheye_thoby
11376 Apply the reverse of image correction (instead of correcting distortion, apply
11379 @item interpolation
11380 The type of interpolation used when correcting distortion. The following values
11385 @item linear (default)
11390 @subsection Examples
11394 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11395 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11399 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
11403 Apply the same as before, but only for the first 5 seconds of video.
11406 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
11413 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11414 score between two input videos.
11416 The obtained VMAF score is printed through the logging system.
11418 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11419 After installing the library it can be enabled using:
11420 @code{./configure --enable-libvmaf --enable-version3}.
11421 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11423 The filter has following options:
11427 Set the model path which is to be used for SVM.
11428 Default value: @code{"vmaf_v0.6.1.pkl"}
11431 Set the file path to be used to store logs.
11434 Set the format of the log file (xml or json).
11436 @item enable_transform
11437 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11438 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11439 Default value: @code{false}
11442 Invokes the phone model which will generate VMAF scores higher than in the
11443 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11446 Enables computing psnr along with vmaf.
11449 Enables computing ssim along with vmaf.
11452 Enables computing ms_ssim along with vmaf.
11455 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11458 Set number of threads to be used when computing vmaf.
11461 Set interval for frame subsampling used when computing vmaf.
11463 @item enable_conf_interval
11464 Enables confidence interval.
11467 This filter also supports the @ref{framesync} options.
11469 On the below examples the input file @file{main.mpg} being processed is
11470 compared with the reference file @file{ref.mpg}.
11473 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11476 Example with options:
11478 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11483 Limits the pixel components values to the specified range [min, max].
11485 The filter accepts the following options:
11489 Lower bound. Defaults to the lowest allowed value for the input.
11492 Upper bound. Defaults to the highest allowed value for the input.
11495 Specify which planes will be processed. Defaults to all available.
11502 The filter accepts the following options:
11506 Set the number of loops. Setting this value to -1 will result in infinite loops.
11510 Set maximal size in number of frames. Default is 0.
11513 Set first frame of loop. Default is 0.
11516 @subsection Examples
11520 Loop single first frame infinitely:
11522 loop=loop=-1:size=1:start=0
11526 Loop single first frame 10 times:
11528 loop=loop=10:size=1:start=0
11532 Loop 10 first frames 5 times:
11534 loop=loop=5:size=10:start=0
11540 Apply a 1D LUT to an input video.
11542 The filter accepts the following options:
11546 Set the 1D LUT file name.
11548 Currently supported formats:
11555 Select interpolation mode.
11557 Available values are:
11561 Use values from the nearest defined point.
11563 Interpolate values using the linear interpolation.
11565 Interpolate values using the cosine interpolation.
11567 Interpolate values using the cubic interpolation.
11569 Interpolate values using the spline interpolation.
11576 Apply a 3D LUT to an input video.
11578 The filter accepts the following options:
11582 Set the 3D LUT file name.
11584 Currently supported formats:
11596 Select interpolation mode.
11598 Available values are:
11602 Use values from the nearest defined point.
11604 Interpolate values using the 8 points defining a cube.
11606 Interpolate values using a tetrahedron.
11610 This filter also supports the @ref{framesync} options.
11614 Turn certain luma values into transparency.
11616 The filter accepts the following options:
11620 Set the luma which will be used as base for transparency.
11621 Default value is @code{0}.
11624 Set the range of luma values to be keyed out.
11625 Default value is @code{0}.
11628 Set the range of softness. Default value is @code{0}.
11629 Use this to control gradual transition from zero to full transparency.
11632 @section lut, lutrgb, lutyuv
11634 Compute a look-up table for binding each pixel component input value
11635 to an output value, and apply it to the input video.
11637 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11638 to an RGB input video.
11640 These filters accept the following parameters:
11643 set first pixel component expression
11645 set second pixel component expression
11647 set third pixel component expression
11649 set fourth pixel component expression, corresponds to the alpha component
11652 set red component expression
11654 set green component expression
11656 set blue component expression
11658 alpha component expression
11661 set Y/luminance component expression
11663 set U/Cb component expression
11665 set V/Cr component expression
11668 Each of them specifies the expression to use for computing the lookup table for
11669 the corresponding pixel component values.
11671 The exact component associated to each of the @var{c*} options depends on the
11674 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11675 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11677 The expressions can contain the following constants and functions:
11682 The input width and height.
11685 The input value for the pixel component.
11688 The input value, clipped to the @var{minval}-@var{maxval} range.
11691 The maximum value for the pixel component.
11694 The minimum value for the pixel component.
11697 The negated value for the pixel component value, clipped to the
11698 @var{minval}-@var{maxval} range; it corresponds to the expression
11699 "maxval-clipval+minval".
11702 The computed value in @var{val}, clipped to the
11703 @var{minval}-@var{maxval} range.
11705 @item gammaval(gamma)
11706 The computed gamma correction value of the pixel component value,
11707 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11709 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11713 All expressions default to "val".
11715 @subsection Examples
11719 Negate input video:
11721 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11722 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11725 The above is the same as:
11727 lutrgb="r=negval:g=negval:b=negval"
11728 lutyuv="y=negval:u=negval:v=negval"
11738 Remove chroma components, turning the video into a graytone image:
11740 lutyuv="u=128:v=128"
11744 Apply a luma burning effect:
11750 Remove green and blue components:
11756 Set a constant alpha channel value on input:
11758 format=rgba,lutrgb=a="maxval-minval/2"
11762 Correct luminance gamma by a factor of 0.5:
11764 lutyuv=y=gammaval(0.5)
11768 Discard least significant bits of luma:
11770 lutyuv=y='bitand(val, 128+64+32)'
11774 Technicolor like effect:
11776 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11780 @section lut2, tlut2
11782 The @code{lut2} filter takes two input streams and outputs one
11785 The @code{tlut2} (time lut2) filter takes two consecutive frames
11786 from one single stream.
11788 This filter accepts the following parameters:
11791 set first pixel component expression
11793 set second pixel component expression
11795 set third pixel component expression
11797 set fourth pixel component expression, corresponds to the alpha component
11800 set output bit depth, only available for @code{lut2} filter. By default is 0,
11801 which means bit depth is automatically picked from first input format.
11804 Each of them specifies the expression to use for computing the lookup table for
11805 the corresponding pixel component values.
11807 The exact component associated to each of the @var{c*} options depends on the
11810 The expressions can contain the following constants:
11815 The input width and height.
11818 The first input value for the pixel component.
11821 The second input value for the pixel component.
11824 The first input video bit depth.
11827 The second input video bit depth.
11830 All expressions default to "x".
11832 @subsection Examples
11836 Highlight differences between two RGB video streams:
11838 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)'
11842 Highlight differences between two YUV video streams:
11844 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)'
11848 Show max difference between two video streams:
11850 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)))'
11854 @section maskedclamp
11856 Clamp the first input stream with the second input and third input stream.
11858 Returns the value of first stream to be between second input
11859 stream - @code{undershoot} and third input stream + @code{overshoot}.
11861 This filter accepts the following options:
11864 Default value is @code{0}.
11867 Default value is @code{0}.
11870 Set which planes will be processed as bitmap, unprocessed planes will be
11871 copied from first stream.
11872 By default value 0xf, all planes will be processed.
11875 @section maskedmerge
11877 Merge the first input stream with the second input stream using per pixel
11878 weights in the third input stream.
11880 A value of 0 in the third stream pixel component means that pixel component
11881 from first stream is returned unchanged, while maximum value (eg. 255 for
11882 8-bit videos) means that pixel component from second stream is returned
11883 unchanged. Intermediate values define the amount of merging between both
11884 input stream's pixel components.
11886 This filter accepts the following options:
11889 Set which planes will be processed as bitmap, unprocessed planes will be
11890 copied from first stream.
11891 By default value 0xf, all planes will be processed.
11896 Apply motion-compensation deinterlacing.
11898 It needs one field per frame as input and must thus be used together
11899 with yadif=1/3 or equivalent.
11901 This filter accepts the following options:
11904 Set the deinterlacing mode.
11906 It accepts one of the following values:
11911 use iterative motion estimation
11913 like @samp{slow}, but use multiple reference frames.
11915 Default value is @samp{fast}.
11918 Set the picture field parity assumed for the input video. It must be
11919 one of the following values:
11923 assume top field first
11925 assume bottom field first
11928 Default value is @samp{bff}.
11931 Set per-block quantization parameter (QP) used by the internal
11934 Higher values should result in a smoother motion vector field but less
11935 optimal individual vectors. Default value is 1.
11938 @section mergeplanes
11940 Merge color channel components from several video streams.
11942 The filter accepts up to 4 input streams, and merge selected input
11943 planes to the output video.
11945 This filter accepts the following options:
11948 Set input to output plane mapping. Default is @code{0}.
11950 The mappings is specified as a bitmap. It should be specified as a
11951 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
11952 mapping for the first plane of the output stream. 'A' sets the number of
11953 the input stream to use (from 0 to 3), and 'a' the plane number of the
11954 corresponding input to use (from 0 to 3). The rest of the mappings is
11955 similar, 'Bb' describes the mapping for the output stream second
11956 plane, 'Cc' describes the mapping for the output stream third plane and
11957 'Dd' describes the mapping for the output stream fourth plane.
11960 Set output pixel format. Default is @code{yuva444p}.
11963 @subsection Examples
11967 Merge three gray video streams of same width and height into single video stream:
11969 [a0][a1][a2]mergeplanes=0x001020:yuv444p
11973 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
11975 [a0][a1]mergeplanes=0x00010210:yuva444p
11979 Swap Y and A plane in yuva444p stream:
11981 format=yuva444p,mergeplanes=0x03010200:yuva444p
11985 Swap U and V plane in yuv420p stream:
11987 format=yuv420p,mergeplanes=0x000201:yuv420p
11991 Cast a rgb24 clip to yuv444p:
11993 format=rgb24,mergeplanes=0x000102:yuv444p
11999 Estimate and export motion vectors using block matching algorithms.
12000 Motion vectors are stored in frame side data to be used by other filters.
12002 This filter accepts the following options:
12005 Specify the motion estimation method. Accepts one of the following values:
12009 Exhaustive search algorithm.
12011 Three step search algorithm.
12013 Two dimensional logarithmic search algorithm.
12015 New three step search algorithm.
12017 Four step search algorithm.
12019 Diamond search algorithm.
12021 Hexagon-based search algorithm.
12023 Enhanced predictive zonal search algorithm.
12025 Uneven multi-hexagon search algorithm.
12027 Default value is @samp{esa}.
12030 Macroblock size. Default @code{16}.
12033 Search parameter. Default @code{7}.
12036 @section midequalizer
12038 Apply Midway Image Equalization effect using two video streams.
12040 Midway Image Equalization adjusts a pair of images to have the same
12041 histogram, while maintaining their dynamics as much as possible. It's
12042 useful for e.g. matching exposures from a pair of stereo cameras.
12044 This filter has two inputs and one output, which must be of same pixel format, but
12045 may be of different sizes. The output of filter is first input adjusted with
12046 midway histogram of both inputs.
12048 This filter accepts the following option:
12052 Set which planes to process. Default is @code{15}, which is all available planes.
12055 @section minterpolate
12057 Convert the video to specified frame rate using motion interpolation.
12059 This filter accepts the following options:
12062 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}.
12065 Motion interpolation mode. Following values are accepted:
12068 Duplicate previous or next frame for interpolating new ones.
12070 Blend source frames. Interpolated frame is mean of previous and next frames.
12072 Motion compensated interpolation. Following options are effective when this mode is selected:
12076 Motion compensation mode. Following values are accepted:
12079 Overlapped block motion compensation.
12081 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12083 Default mode is @samp{obmc}.
12086 Motion estimation mode. Following values are accepted:
12089 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12091 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12093 Default mode is @samp{bilat}.
12096 The algorithm to be used for motion estimation. Following values are accepted:
12099 Exhaustive search algorithm.
12101 Three step search algorithm.
12103 Two dimensional logarithmic search algorithm.
12105 New three step search algorithm.
12107 Four step search algorithm.
12109 Diamond search algorithm.
12111 Hexagon-based search algorithm.
12113 Enhanced predictive zonal search algorithm.
12115 Uneven multi-hexagon search algorithm.
12117 Default algorithm is @samp{epzs}.
12120 Macroblock size. Default @code{16}.
12123 Motion estimation search parameter. Default @code{32}.
12126 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).
12131 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:
12134 Disable scene change detection.
12136 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12138 Default method is @samp{fdiff}.
12140 @item scd_threshold
12141 Scene change detection threshold. Default is @code{5.0}.
12146 Mix several video input streams into one video stream.
12148 A description of the accepted options follows.
12152 The number of inputs. If unspecified, it defaults to 2.
12155 Specify weight of each input video stream as sequence.
12156 Each weight is separated by space. If number of weights
12157 is smaller than number of @var{frames} last specified
12158 weight will be used for all remaining unset weights.
12161 Specify scale, if it is set it will be multiplied with sum
12162 of each weight multiplied with pixel values to give final destination
12163 pixel value. By default @var{scale} is auto scaled to sum of weights.
12166 Specify how end of stream is determined.
12169 The duration of the longest input. (default)
12172 The duration of the shortest input.
12175 The duration of the first input.
12179 @section mpdecimate
12181 Drop frames that do not differ greatly from the previous frame in
12182 order to reduce frame rate.
12184 The main use of this filter is for very-low-bitrate encoding
12185 (e.g. streaming over dialup modem), but it could in theory be used for
12186 fixing movies that were inverse-telecined incorrectly.
12188 A description of the accepted options follows.
12192 Set the maximum number of consecutive frames which can be dropped (if
12193 positive), or the minimum interval between dropped frames (if
12194 negative). If the value is 0, the frame is dropped disregarding the
12195 number of previous sequentially dropped frames.
12197 Default value is 0.
12202 Set the dropping threshold values.
12204 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12205 represent actual pixel value differences, so a threshold of 64
12206 corresponds to 1 unit of difference for each pixel, or the same spread
12207 out differently over the block.
12209 A frame is a candidate for dropping if no 8x8 blocks differ by more
12210 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12211 meaning the whole image) differ by more than a threshold of @option{lo}.
12213 Default value for @option{hi} is 64*12, default value for @option{lo} is
12214 64*5, and default value for @option{frac} is 0.33.
12220 Negate (invert) the input video.
12222 It accepts the following option:
12227 With value 1, it negates the alpha component, if present. Default value is 0.
12233 Denoise frames using Non-Local Means algorithm.
12235 Each pixel is adjusted by looking for other pixels with similar contexts. This
12236 context similarity is defined by comparing their surrounding patches of size
12237 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12240 Note that the research area defines centers for patches, which means some
12241 patches will be made of pixels outside that research area.
12243 The filter accepts the following options.
12247 Set denoising strength.
12253 Same as @option{p} but for chroma planes.
12255 The default value is @var{0} and means automatic.
12261 Same as @option{r} but for chroma planes.
12263 The default value is @var{0} and means automatic.
12268 Deinterlace video using neural network edge directed interpolation.
12270 This filter accepts the following options:
12274 Mandatory option, without binary file filter can not work.
12275 Currently file can be found here:
12276 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12279 Set which frames to deinterlace, by default it is @code{all}.
12280 Can be @code{all} or @code{interlaced}.
12283 Set mode of operation.
12285 Can be one of the following:
12289 Use frame flags, both fields.
12291 Use frame flags, single field.
12293 Use top field only.
12295 Use bottom field only.
12297 Use both fields, top first.
12299 Use both fields, bottom first.
12303 Set which planes to process, by default filter process all frames.
12306 Set size of local neighborhood around each pixel, used by the predictor neural
12309 Can be one of the following:
12322 Set the number of neurons in predictor neural network.
12323 Can be one of the following:
12334 Controls the number of different neural network predictions that are blended
12335 together to compute the final output value. Can be @code{fast}, default or
12339 Set which set of weights to use in the predictor.
12340 Can be one of the following:
12344 weights trained to minimize absolute error
12346 weights trained to minimize squared error
12350 Controls whether or not the prescreener neural network is used to decide
12351 which pixels should be processed by the predictor neural network and which
12352 can be handled by simple cubic interpolation.
12353 The prescreener is trained to know whether cubic interpolation will be
12354 sufficient for a pixel or whether it should be predicted by the predictor nn.
12355 The computational complexity of the prescreener nn is much less than that of
12356 the predictor nn. Since most pixels can be handled by cubic interpolation,
12357 using the prescreener generally results in much faster processing.
12358 The prescreener is pretty accurate, so the difference between using it and not
12359 using it is almost always unnoticeable.
12361 Can be one of the following:
12369 Default is @code{new}.
12372 Set various debugging flags.
12377 Force libavfilter not to use any of the specified pixel formats for the
12378 input to the next filter.
12380 It accepts the following parameters:
12384 A '|'-separated list of pixel format names, such as
12385 pix_fmts=yuv420p|monow|rgb24".
12389 @subsection Examples
12393 Force libavfilter to use a format different from @var{yuv420p} for the
12394 input to the vflip filter:
12396 noformat=pix_fmts=yuv420p,vflip
12400 Convert the input video to any of the formats not contained in the list:
12402 noformat=yuv420p|yuv444p|yuv410p
12408 Add noise on video input frame.
12410 The filter accepts the following options:
12418 Set noise seed for specific pixel component or all pixel components in case
12419 of @var{all_seed}. Default value is @code{123457}.
12421 @item all_strength, alls
12422 @item c0_strength, c0s
12423 @item c1_strength, c1s
12424 @item c2_strength, c2s
12425 @item c3_strength, c3s
12426 Set noise strength for specific pixel component or all pixel components in case
12427 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12429 @item all_flags, allf
12430 @item c0_flags, c0f
12431 @item c1_flags, c1f
12432 @item c2_flags, c2f
12433 @item c3_flags, c3f
12434 Set pixel component flags or set flags for all components if @var{all_flags}.
12435 Available values for component flags are:
12438 averaged temporal noise (smoother)
12440 mix random noise with a (semi)regular pattern
12442 temporal noise (noise pattern changes between frames)
12444 uniform noise (gaussian otherwise)
12448 @subsection Examples
12450 Add temporal and uniform noise to input video:
12452 noise=alls=20:allf=t+u
12457 Normalize RGB video (aka histogram stretching, contrast stretching).
12458 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12460 For each channel of each frame, the filter computes the input range and maps
12461 it linearly to the user-specified output range. The output range defaults
12462 to the full dynamic range from pure black to pure white.
12464 Temporal smoothing can be used on the input range to reduce flickering (rapid
12465 changes in brightness) caused when small dark or bright objects enter or leave
12466 the scene. This is similar to the auto-exposure (automatic gain control) on a
12467 video camera, and, like a video camera, it may cause a period of over- or
12468 under-exposure of the video.
12470 The R,G,B channels can be normalized independently, which may cause some
12471 color shifting, or linked together as a single channel, which prevents
12472 color shifting. Linked normalization preserves hue. Independent normalization
12473 does not, so it can be used to remove some color casts. Independent and linked
12474 normalization can be combined in any ratio.
12476 The normalize filter accepts the following options:
12481 Colors which define the output range. The minimum input value is mapped to
12482 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12483 The defaults are black and white respectively. Specifying white for
12484 @var{blackpt} and black for @var{whitept} will give color-inverted,
12485 normalized video. Shades of grey can be used to reduce the dynamic range
12486 (contrast). Specifying saturated colors here can create some interesting
12490 The number of previous frames to use for temporal smoothing. The input range
12491 of each channel is smoothed using a rolling average over the current frame
12492 and the @var{smoothing} previous frames. The default is 0 (no temporal
12496 Controls the ratio of independent (color shifting) channel normalization to
12497 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12498 independent. Defaults to 1.0 (fully independent).
12501 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12502 expensive no-op. Defaults to 1.0 (full strength).
12506 @subsection Examples
12508 Stretch video contrast to use the full dynamic range, with no temporal
12509 smoothing; may flicker depending on the source content:
12511 normalize=blackpt=black:whitept=white:smoothing=0
12514 As above, but with 50 frames of temporal smoothing; flicker should be
12515 reduced, depending on the source content:
12517 normalize=blackpt=black:whitept=white:smoothing=50
12520 As above, but with hue-preserving linked channel normalization:
12522 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12525 As above, but with half strength:
12527 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12530 Map the darkest input color to red, the brightest input color to cyan:
12532 normalize=blackpt=red:whitept=cyan
12537 Pass the video source unchanged to the output.
12540 Optical Character Recognition
12542 This filter uses Tesseract for optical character recognition. To enable
12543 compilation of this filter, you need to configure FFmpeg with
12544 @code{--enable-libtesseract}.
12546 It accepts the following options:
12550 Set datapath to tesseract data. Default is to use whatever was
12551 set at installation.
12554 Set language, default is "eng".
12557 Set character whitelist.
12560 Set character blacklist.
12563 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12567 Apply a video transform using libopencv.
12569 To enable this filter, install the libopencv library and headers and
12570 configure FFmpeg with @code{--enable-libopencv}.
12572 It accepts the following parameters:
12577 The name of the libopencv filter to apply.
12579 @item filter_params
12580 The parameters to pass to the libopencv filter. If not specified, the default
12581 values are assumed.
12585 Refer to the official libopencv documentation for more precise
12587 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12589 Several libopencv filters are supported; see the following subsections.
12594 Dilate an image by using a specific structuring element.
12595 It corresponds to the libopencv function @code{cvDilate}.
12597 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12599 @var{struct_el} represents a structuring element, and has the syntax:
12600 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12602 @var{cols} and @var{rows} represent the number of columns and rows of
12603 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12604 point, and @var{shape} the shape for the structuring element. @var{shape}
12605 must be "rect", "cross", "ellipse", or "custom".
12607 If the value for @var{shape} is "custom", it must be followed by a
12608 string of the form "=@var{filename}". The file with name
12609 @var{filename} is assumed to represent a binary image, with each
12610 printable character corresponding to a bright pixel. When a custom
12611 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12612 or columns and rows of the read file are assumed instead.
12614 The default value for @var{struct_el} is "3x3+0x0/rect".
12616 @var{nb_iterations} specifies the number of times the transform is
12617 applied to the image, and defaults to 1.
12621 # Use the default values
12624 # Dilate using a structuring element with a 5x5 cross, iterating two times
12625 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12627 # Read the shape from the file diamond.shape, iterating two times.
12628 # The file diamond.shape may contain a pattern of characters like this
12634 # The specified columns and rows are ignored
12635 # but the anchor point coordinates are not
12636 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12641 Erode an image by using a specific structuring element.
12642 It corresponds to the libopencv function @code{cvErode}.
12644 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12645 with the same syntax and semantics as the @ref{dilate} filter.
12649 Smooth the input video.
12651 The filter takes the following parameters:
12652 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12654 @var{type} is the type of smooth filter to apply, and must be one of
12655 the following values: "blur", "blur_no_scale", "median", "gaussian",
12656 or "bilateral". The default value is "gaussian".
12658 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12659 depend on the smooth type. @var{param1} and
12660 @var{param2} accept integer positive values or 0. @var{param3} and
12661 @var{param4} accept floating point values.
12663 The default value for @var{param1} is 3. The default value for the
12664 other parameters is 0.
12666 These parameters correspond to the parameters assigned to the
12667 libopencv function @code{cvSmooth}.
12669 @section oscilloscope
12671 2D Video Oscilloscope.
12673 Useful to measure spatial impulse, step responses, chroma delays, etc.
12675 It accepts the following parameters:
12679 Set scope center x position.
12682 Set scope center y position.
12685 Set scope size, relative to frame diagonal.
12688 Set scope tilt/rotation.
12694 Set trace center x position.
12697 Set trace center y position.
12700 Set trace width, relative to width of frame.
12703 Set trace height, relative to height of frame.
12706 Set which components to trace. By default it traces first three components.
12709 Draw trace grid. By default is enabled.
12712 Draw some statistics. By default is enabled.
12715 Draw scope. By default is enabled.
12718 @subsection Examples
12722 Inspect full first row of video frame.
12724 oscilloscope=x=0.5:y=0:s=1
12728 Inspect full last row of video frame.
12730 oscilloscope=x=0.5:y=1:s=1
12734 Inspect full 5th line of video frame of height 1080.
12736 oscilloscope=x=0.5:y=5/1080:s=1
12740 Inspect full last column of video frame.
12742 oscilloscope=x=1:y=0.5:s=1:t=1
12750 Overlay one video on top of another.
12752 It takes two inputs and has one output. The first input is the "main"
12753 video on which the second input is overlaid.
12755 It accepts the following parameters:
12757 A description of the accepted options follows.
12762 Set the expression for the x and y coordinates of the overlaid video
12763 on the main video. Default value is "0" for both expressions. In case
12764 the expression is invalid, it is set to a huge value (meaning that the
12765 overlay will not be displayed within the output visible area).
12768 See @ref{framesync}.
12771 Set when the expressions for @option{x}, and @option{y} are evaluated.
12773 It accepts the following values:
12776 only evaluate expressions once during the filter initialization or
12777 when a command is processed
12780 evaluate expressions for each incoming frame
12783 Default value is @samp{frame}.
12786 See @ref{framesync}.
12789 Set the format for the output video.
12791 It accepts the following values:
12794 force YUV420 output
12797 force YUV422 output
12800 force YUV444 output
12803 force packed RGB output
12806 force planar RGB output
12809 automatically pick format
12812 Default value is @samp{yuv420}.
12815 See @ref{framesync}.
12818 Set format of alpha of the overlaid video, it can be @var{straight} or
12819 @var{premultiplied}. Default is @var{straight}.
12822 The @option{x}, and @option{y} expressions can contain the following
12828 The main input width and height.
12832 The overlay input width and height.
12836 The computed values for @var{x} and @var{y}. They are evaluated for
12841 horizontal and vertical chroma subsample values of the output
12842 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
12846 the number of input frame, starting from 0
12849 the position in the file of the input frame, NAN if unknown
12852 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
12856 This filter also supports the @ref{framesync} options.
12858 Note that the @var{n}, @var{pos}, @var{t} variables are available only
12859 when evaluation is done @emph{per frame}, and will evaluate to NAN
12860 when @option{eval} is set to @samp{init}.
12862 Be aware that frames are taken from each input video in timestamp
12863 order, hence, if their initial timestamps differ, it is a good idea
12864 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
12865 have them begin in the same zero timestamp, as the example for
12866 the @var{movie} filter does.
12868 You can chain together more overlays but you should test the
12869 efficiency of such approach.
12871 @subsection Commands
12873 This filter supports the following commands:
12877 Modify the x and y of the overlay input.
12878 The command accepts the same syntax of the corresponding option.
12880 If the specified expression is not valid, it is kept at its current
12884 @subsection Examples
12888 Draw the overlay at 10 pixels from the bottom right corner of the main
12891 overlay=main_w-overlay_w-10:main_h-overlay_h-10
12894 Using named options the example above becomes:
12896 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
12900 Insert a transparent PNG logo in the bottom left corner of the input,
12901 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
12903 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
12907 Insert 2 different transparent PNG logos (second logo on bottom
12908 right corner) using the @command{ffmpeg} tool:
12910 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
12914 Add a transparent color layer on top of the main video; @code{WxH}
12915 must specify the size of the main input to the overlay filter:
12917 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
12921 Play an original video and a filtered version (here with the deshake
12922 filter) side by side using the @command{ffplay} tool:
12924 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
12927 The above command is the same as:
12929 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
12933 Make a sliding overlay appearing from the left to the right top part of the
12934 screen starting since time 2:
12936 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
12940 Compose output by putting two input videos side to side:
12942 ffmpeg -i left.avi -i right.avi -filter_complex "
12943 nullsrc=size=200x100 [background];
12944 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
12945 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
12946 [background][left] overlay=shortest=1 [background+left];
12947 [background+left][right] overlay=shortest=1:x=100 [left+right]
12952 Mask 10-20 seconds of a video by applying the delogo filter to a section
12954 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
12955 -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]'
12960 Chain several overlays in cascade:
12962 nullsrc=s=200x200 [bg];
12963 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
12964 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
12965 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
12966 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
12967 [in3] null, [mid2] overlay=100:100 [out0]
12974 Apply Overcomplete Wavelet denoiser.
12976 The filter accepts the following options:
12982 Larger depth values will denoise lower frequency components more, but
12983 slow down filtering.
12985 Must be an int in the range 8-16, default is @code{8}.
12987 @item luma_strength, ls
12990 Must be a double value in the range 0-1000, default is @code{1.0}.
12992 @item chroma_strength, cs
12993 Set chroma strength.
12995 Must be a double value in the range 0-1000, default is @code{1.0}.
13001 Add paddings to the input image, and place the original input at the
13002 provided @var{x}, @var{y} coordinates.
13004 It accepts the following parameters:
13009 Specify an expression for the size of the output image with the
13010 paddings added. If the value for @var{width} or @var{height} is 0, the
13011 corresponding input size is used for the output.
13013 The @var{width} expression can reference the value set by the
13014 @var{height} expression, and vice versa.
13016 The default value of @var{width} and @var{height} is 0.
13020 Specify the offsets to place the input image at within the padded area,
13021 with respect to the top/left border of the output image.
13023 The @var{x} expression can reference the value set by the @var{y}
13024 expression, and vice versa.
13026 The default value of @var{x} and @var{y} is 0.
13028 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13029 so the input image is centered on the padded area.
13032 Specify the color of the padded area. For the syntax of this option,
13033 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13034 manual,ffmpeg-utils}.
13036 The default value of @var{color} is "black".
13039 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13041 It accepts the following values:
13045 Only evaluate expressions once during the filter initialization or when
13046 a command is processed.
13049 Evaluate expressions for each incoming frame.
13053 Default value is @samp{init}.
13056 Pad to aspect instead to a resolution.
13060 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13061 options are expressions containing the following constants:
13066 The input video width and height.
13070 These are the same as @var{in_w} and @var{in_h}.
13074 The output width and height (the size of the padded area), as
13075 specified by the @var{width} and @var{height} expressions.
13079 These are the same as @var{out_w} and @var{out_h}.
13083 The x and y offsets as specified by the @var{x} and @var{y}
13084 expressions, or NAN if not yet specified.
13087 same as @var{iw} / @var{ih}
13090 input sample aspect ratio
13093 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13097 The horizontal and vertical chroma subsample values. For example for the
13098 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13101 @subsection Examples
13105 Add paddings with the color "violet" to the input video. The output video
13106 size is 640x480, and the top-left corner of the input video is placed at
13109 pad=640:480:0:40:violet
13112 The example above is equivalent to the following command:
13114 pad=width=640:height=480:x=0:y=40:color=violet
13118 Pad the input to get an output with dimensions increased by 3/2,
13119 and put the input video at the center of the padded area:
13121 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13125 Pad the input to get a squared output with size equal to the maximum
13126 value between the input width and height, and put the input video at
13127 the center of the padded area:
13129 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13133 Pad the input to get a final w/h ratio of 16:9:
13135 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13139 In case of anamorphic video, in order to set the output display aspect
13140 correctly, it is necessary to use @var{sar} in the expression,
13141 according to the relation:
13143 (ih * X / ih) * sar = output_dar
13144 X = output_dar / sar
13147 Thus the previous example needs to be modified to:
13149 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13153 Double the output size and put the input video in the bottom-right
13154 corner of the output padded area:
13156 pad="2*iw:2*ih:ow-iw:oh-ih"
13160 @anchor{palettegen}
13161 @section palettegen
13163 Generate one palette for a whole video stream.
13165 It accepts the following options:
13169 Set the maximum number of colors to quantize in the palette.
13170 Note: the palette will still contain 256 colors; the unused palette entries
13173 @item reserve_transparent
13174 Create a palette of 255 colors maximum and reserve the last one for
13175 transparency. Reserving the transparency color is useful for GIF optimization.
13176 If not set, the maximum of colors in the palette will be 256. You probably want
13177 to disable this option for a standalone image.
13180 @item transparency_color
13181 Set the color that will be used as background for transparency.
13184 Set statistics mode.
13186 It accepts the following values:
13189 Compute full frame histograms.
13191 Compute histograms only for the part that differs from previous frame. This
13192 might be relevant to give more importance to the moving part of your input if
13193 the background is static.
13195 Compute new histogram for each frame.
13198 Default value is @var{full}.
13201 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13202 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13203 color quantization of the palette. This information is also visible at
13204 @var{info} logging level.
13206 @subsection Examples
13210 Generate a representative palette of a given video using @command{ffmpeg}:
13212 ffmpeg -i input.mkv -vf palettegen palette.png
13216 @section paletteuse
13218 Use a palette to downsample an input video stream.
13220 The filter takes two inputs: one video stream and a palette. The palette must
13221 be a 256 pixels image.
13223 It accepts the following options:
13227 Select dithering mode. Available algorithms are:
13230 Ordered 8x8 bayer dithering (deterministic)
13232 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13233 Note: this dithering is sometimes considered "wrong" and is included as a
13235 @item floyd_steinberg
13236 Floyd and Steingberg dithering (error diffusion)
13238 Frankie Sierra dithering v2 (error diffusion)
13240 Frankie Sierra dithering v2 "Lite" (error diffusion)
13243 Default is @var{sierra2_4a}.
13246 When @var{bayer} dithering is selected, this option defines the scale of the
13247 pattern (how much the crosshatch pattern is visible). A low value means more
13248 visible pattern for less banding, and higher value means less visible pattern
13249 at the cost of more banding.
13251 The option must be an integer value in the range [0,5]. Default is @var{2}.
13254 If set, define the zone to process
13258 Only the changing rectangle will be reprocessed. This is similar to GIF
13259 cropping/offsetting compression mechanism. This option can be useful for speed
13260 if only a part of the image is changing, and has use cases such as limiting the
13261 scope of the error diffusal @option{dither} to the rectangle that bounds the
13262 moving scene (it leads to more deterministic output if the scene doesn't change
13263 much, and as a result less moving noise and better GIF compression).
13266 Default is @var{none}.
13269 Take new palette for each output frame.
13271 @item alpha_threshold
13272 Sets the alpha threshold for transparency. Alpha values above this threshold
13273 will be treated as completely opaque, and values below this threshold will be
13274 treated as completely transparent.
13276 The option must be an integer value in the range [0,255]. Default is @var{128}.
13279 @subsection Examples
13283 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13284 using @command{ffmpeg}:
13286 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13290 @section perspective
13292 Correct perspective of video not recorded perpendicular to the screen.
13294 A description of the accepted parameters follows.
13305 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13306 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13307 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13308 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13309 then the corners of the source will be sent to the specified coordinates.
13311 The expressions can use the following variables:
13316 the width and height of video frame.
13320 Output frame count.
13323 @item interpolation
13324 Set interpolation for perspective correction.
13326 It accepts the following values:
13332 Default value is @samp{linear}.
13335 Set interpretation of coordinate options.
13337 It accepts the following values:
13341 Send point in the source specified by the given coordinates to
13342 the corners of the destination.
13344 @item 1, destination
13346 Send the corners of the source to the point in the destination specified
13347 by the given coordinates.
13349 Default value is @samp{source}.
13353 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13355 It accepts the following values:
13358 only evaluate expressions once during the filter initialization or
13359 when a command is processed
13362 evaluate expressions for each incoming frame
13365 Default value is @samp{init}.
13370 Delay interlaced video by one field time so that the field order changes.
13372 The intended use is to fix PAL movies that have been captured with the
13373 opposite field order to the film-to-video transfer.
13375 A description of the accepted parameters follows.
13381 It accepts the following values:
13384 Capture field order top-first, transfer bottom-first.
13385 Filter will delay the bottom field.
13388 Capture field order bottom-first, transfer top-first.
13389 Filter will delay the top field.
13392 Capture and transfer with the same field order. This mode only exists
13393 for the documentation of the other options to refer to, but if you
13394 actually select it, the filter will faithfully do nothing.
13397 Capture field order determined automatically by field flags, transfer
13399 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13400 basis using field flags. If no field information is available,
13401 then this works just like @samp{u}.
13404 Capture unknown or varying, transfer opposite.
13405 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13406 analyzing the images and selecting the alternative that produces best
13407 match between the fields.
13410 Capture top-first, transfer unknown or varying.
13411 Filter selects among @samp{t} and @samp{p} using image analysis.
13414 Capture bottom-first, transfer unknown or varying.
13415 Filter selects among @samp{b} and @samp{p} using image analysis.
13418 Capture determined by field flags, transfer unknown or varying.
13419 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13420 image analysis. If no field information is available, then this works just
13421 like @samp{U}. This is the default mode.
13424 Both capture and transfer unknown or varying.
13425 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13429 @section pixdesctest
13431 Pixel format descriptor test filter, mainly useful for internal
13432 testing. The output video should be equal to the input video.
13436 format=monow, pixdesctest
13439 can be used to test the monowhite pixel format descriptor definition.
13443 Display sample values of color channels. Mainly useful for checking color
13444 and levels. Minimum supported resolution is 640x480.
13446 The filters accept the following options:
13450 Set scope X position, relative offset on X axis.
13453 Set scope Y position, relative offset on Y axis.
13462 Set window opacity. This window also holds statistics about pixel area.
13465 Set window X position, relative offset on X axis.
13468 Set window Y position, relative offset on Y axis.
13473 Enable the specified chain of postprocessing subfilters using libpostproc. This
13474 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13475 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13476 Each subfilter and some options have a short and a long name that can be used
13477 interchangeably, i.e. dr/dering are the same.
13479 The filters accept the following options:
13483 Set postprocessing subfilters string.
13486 All subfilters share common options to determine their scope:
13490 Honor the quality commands for this subfilter.
13493 Do chrominance filtering, too (default).
13496 Do luminance filtering only (no chrominance).
13499 Do chrominance filtering only (no luminance).
13502 These options can be appended after the subfilter name, separated by a '|'.
13504 Available subfilters are:
13507 @item hb/hdeblock[|difference[|flatness]]
13508 Horizontal deblocking filter
13511 Difference factor where higher values mean more deblocking (default: @code{32}).
13513 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13516 @item vb/vdeblock[|difference[|flatness]]
13517 Vertical deblocking filter
13520 Difference factor where higher values mean more deblocking (default: @code{32}).
13522 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13525 @item ha/hadeblock[|difference[|flatness]]
13526 Accurate horizontal deblocking filter
13529 Difference factor where higher values mean more deblocking (default: @code{32}).
13531 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13534 @item va/vadeblock[|difference[|flatness]]
13535 Accurate vertical deblocking filter
13538 Difference factor where higher values mean more deblocking (default: @code{32}).
13540 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13544 The horizontal and vertical deblocking filters share the difference and
13545 flatness values so you cannot set different horizontal and vertical
13549 @item h1/x1hdeblock
13550 Experimental horizontal deblocking filter
13552 @item v1/x1vdeblock
13553 Experimental vertical deblocking filter
13558 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13561 larger -> stronger filtering
13563 larger -> stronger filtering
13565 larger -> stronger filtering
13568 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13571 Stretch luminance to @code{0-255}.
13574 @item lb/linblenddeint
13575 Linear blend deinterlacing filter that deinterlaces the given block by
13576 filtering all lines with a @code{(1 2 1)} filter.
13578 @item li/linipoldeint
13579 Linear interpolating deinterlacing filter that deinterlaces the given block by
13580 linearly interpolating every second line.
13582 @item ci/cubicipoldeint
13583 Cubic interpolating deinterlacing filter deinterlaces the given block by
13584 cubically interpolating every second line.
13586 @item md/mediandeint
13587 Median deinterlacing filter that deinterlaces the given block by applying a
13588 median filter to every second line.
13590 @item fd/ffmpegdeint
13591 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13592 second line with a @code{(-1 4 2 4 -1)} filter.
13595 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13596 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13598 @item fq/forceQuant[|quantizer]
13599 Overrides the quantizer table from the input with the constant quantizer you
13607 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13610 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13613 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13616 @subsection Examples
13620 Apply horizontal and vertical deblocking, deringing and automatic
13621 brightness/contrast:
13627 Apply default filters without brightness/contrast correction:
13633 Apply default filters and temporal denoiser:
13635 pp=default/tmpnoise|1|2|3
13639 Apply deblocking on luminance only, and switch vertical deblocking on or off
13640 automatically depending on available CPU time:
13647 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13648 similar to spp = 6 with 7 point DCT, where only the center sample is
13651 The filter accepts the following options:
13655 Force a constant quantization parameter. It accepts an integer in range
13656 0 to 63. If not set, the filter will use the QP from the video stream
13660 Set thresholding mode. Available modes are:
13664 Set hard thresholding.
13666 Set soft thresholding (better de-ringing effect, but likely blurrier).
13668 Set medium thresholding (good results, default).
13672 @section premultiply
13673 Apply alpha premultiply effect to input video stream using first plane
13674 of second stream as alpha.
13676 Both streams must have same dimensions and same pixel format.
13678 The filter accepts the following option:
13682 Set which planes will be processed, unprocessed planes will be copied.
13683 By default value 0xf, all planes will be processed.
13686 Do not require 2nd input for processing, instead use alpha plane from input stream.
13690 Apply prewitt operator to input video stream.
13692 The filter accepts the following option:
13696 Set which planes will be processed, unprocessed planes will be copied.
13697 By default value 0xf, all planes will be processed.
13700 Set value which will be multiplied with filtered result.
13703 Set value which will be added to filtered result.
13706 @anchor{program_opencl}
13707 @section program_opencl
13709 Filter video using an OpenCL program.
13714 OpenCL program source file.
13717 Kernel name in program.
13720 Number of inputs to the filter. Defaults to 1.
13723 Size of output frames. Defaults to the same as the first input.
13727 The program source file must contain a kernel function with the given name,
13728 which will be run once for each plane of the output. Each run on a plane
13729 gets enqueued as a separate 2D global NDRange with one work-item for each
13730 pixel to be generated. The global ID offset for each work-item is therefore
13731 the coordinates of a pixel in the destination image.
13733 The kernel function needs to take the following arguments:
13736 Destination image, @var{__write_only image2d_t}.
13738 This image will become the output; the kernel should write all of it.
13740 Frame index, @var{unsigned int}.
13742 This is a counter starting from zero and increasing by one for each frame.
13744 Source images, @var{__read_only image2d_t}.
13746 These are the most recent images on each input. The kernel may read from
13747 them to generate the output, but they can't be written to.
13754 Copy the input to the output (output must be the same size as the input).
13756 __kernel void copy(__write_only image2d_t destination,
13757 unsigned int index,
13758 __read_only image2d_t source)
13760 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13762 int2 location = (int2)(get_global_id(0), get_global_id(1));
13764 float4 value = read_imagef(source, sampler, location);
13766 write_imagef(destination, location, value);
13771 Apply a simple transformation, rotating the input by an amount increasing
13772 with the index counter. Pixel values are linearly interpolated by the
13773 sampler, and the output need not have the same dimensions as the input.
13775 __kernel void rotate_image(__write_only image2d_t dst,
13776 unsigned int index,
13777 __read_only image2d_t src)
13779 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13780 CLK_FILTER_LINEAR);
13782 float angle = (float)index / 100.0f;
13784 float2 dst_dim = convert_float2(get_image_dim(dst));
13785 float2 src_dim = convert_float2(get_image_dim(src));
13787 float2 dst_cen = dst_dim / 2.0f;
13788 float2 src_cen = src_dim / 2.0f;
13790 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13792 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13794 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13795 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13797 src_pos = src_pos * src_dim / dst_dim;
13799 float2 src_loc = src_pos + src_cen;
13801 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13802 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13803 write_imagef(dst, dst_loc, 0.5f);
13805 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13810 Blend two inputs together, with the amount of each input used varying
13811 with the index counter.
13813 __kernel void blend_images(__write_only image2d_t dst,
13814 unsigned int index,
13815 __read_only image2d_t src1,
13816 __read_only image2d_t src2)
13818 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13819 CLK_FILTER_LINEAR);
13821 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
13823 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13824 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
13825 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
13827 float4 val1 = read_imagef(src1, sampler, src1_loc);
13828 float4 val2 = read_imagef(src2, sampler, src2_loc);
13830 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
13836 @section pseudocolor
13838 Alter frame colors in video with pseudocolors.
13840 This filter accept the following options:
13844 set pixel first component expression
13847 set pixel second component expression
13850 set pixel third component expression
13853 set pixel fourth component expression, corresponds to the alpha component
13856 set component to use as base for altering colors
13859 Each of them specifies the expression to use for computing the lookup table for
13860 the corresponding pixel component values.
13862 The expressions can contain the following constants and functions:
13867 The input width and height.
13870 The input value for the pixel component.
13872 @item ymin, umin, vmin, amin
13873 The minimum allowed component value.
13875 @item ymax, umax, vmax, amax
13876 The maximum allowed component value.
13879 All expressions default to "val".
13881 @subsection Examples
13885 Change too high luma values to gradient:
13887 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'"
13893 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
13894 Ratio) between two input videos.
13896 This filter takes in input two input videos, the first input is
13897 considered the "main" source and is passed unchanged to the
13898 output. The second input is used as a "reference" video for computing
13901 Both video inputs must have the same resolution and pixel format for
13902 this filter to work correctly. Also it assumes that both inputs
13903 have the same number of frames, which are compared one by one.
13905 The obtained average PSNR is printed through the logging system.
13907 The filter stores the accumulated MSE (mean squared error) of each
13908 frame, and at the end of the processing it is averaged across all frames
13909 equally, and the following formula is applied to obtain the PSNR:
13912 PSNR = 10*log10(MAX^2/MSE)
13915 Where MAX is the average of the maximum values of each component of the
13918 The description of the accepted parameters follows.
13921 @item stats_file, f
13922 If specified the filter will use the named file to save the PSNR of
13923 each individual frame. When filename equals "-" the data is sent to
13926 @item stats_version
13927 Specifies which version of the stats file format to use. Details of
13928 each format are written below.
13929 Default value is 1.
13931 @item stats_add_max
13932 Determines whether the max value is output to the stats log.
13933 Default value is 0.
13934 Requires stats_version >= 2. If this is set and stats_version < 2,
13935 the filter will return an error.
13938 This filter also supports the @ref{framesync} options.
13940 The file printed if @var{stats_file} is selected, contains a sequence of
13941 key/value pairs of the form @var{key}:@var{value} for each compared
13944 If a @var{stats_version} greater than 1 is specified, a header line precedes
13945 the list of per-frame-pair stats, with key value pairs following the frame
13946 format with the following parameters:
13949 @item psnr_log_version
13950 The version of the log file format. Will match @var{stats_version}.
13953 A comma separated list of the per-frame-pair parameters included in
13957 A description of each shown per-frame-pair parameter follows:
13961 sequential number of the input frame, starting from 1
13964 Mean Square Error pixel-by-pixel average difference of the compared
13965 frames, averaged over all the image components.
13967 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
13968 Mean Square Error pixel-by-pixel average difference of the compared
13969 frames for the component specified by the suffix.
13971 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
13972 Peak Signal to Noise ratio of the compared frames for the component
13973 specified by the suffix.
13975 @item max_avg, max_y, max_u, max_v
13976 Maximum allowed value for each channel, and average over all
13982 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13983 [main][ref] psnr="stats_file=stats.log" [out]
13986 On this example the input file being processed is compared with the
13987 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
13988 is stored in @file{stats.log}.
13993 Pulldown reversal (inverse telecine) filter, capable of handling mixed
13994 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
13997 The pullup filter is designed to take advantage of future context in making
13998 its decisions. This filter is stateless in the sense that it does not lock
13999 onto a pattern to follow, but it instead looks forward to the following
14000 fields in order to identify matches and rebuild progressive frames.
14002 To produce content with an even framerate, insert the fps filter after
14003 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14004 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14006 The filter accepts the following options:
14013 These options set the amount of "junk" to ignore at the left, right, top, and
14014 bottom of the image, respectively. Left and right are in units of 8 pixels,
14015 while top and bottom are in units of 2 lines.
14016 The default is 8 pixels on each side.
14019 Set the strict breaks. Setting this option to 1 will reduce the chances of
14020 filter generating an occasional mismatched frame, but it may also cause an
14021 excessive number of frames to be dropped during high motion sequences.
14022 Conversely, setting it to -1 will make filter match fields more easily.
14023 This may help processing of video where there is slight blurring between
14024 the fields, but may also cause there to be interlaced frames in the output.
14025 Default value is @code{0}.
14028 Set the metric plane to use. It accepts the following values:
14034 Use chroma blue plane.
14037 Use chroma red plane.
14040 This option may be set to use chroma plane instead of the default luma plane
14041 for doing filter's computations. This may improve accuracy on very clean
14042 source material, but more likely will decrease accuracy, especially if there
14043 is chroma noise (rainbow effect) or any grayscale video.
14044 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14045 load and make pullup usable in realtime on slow machines.
14048 For best results (without duplicated frames in the output file) it is
14049 necessary to change the output frame rate. For example, to inverse
14050 telecine NTSC input:
14052 ffmpeg -i input -vf pullup -r 24000/1001 ...
14057 Change video quantization parameters (QP).
14059 The filter accepts the following option:
14063 Set expression for quantization parameter.
14066 The expression is evaluated through the eval API and can contain, among others,
14067 the following constants:
14071 1 if index is not 129, 0 otherwise.
14074 Sequential index starting from -129 to 128.
14077 @subsection Examples
14081 Some equation like:
14089 Flush video frames from internal cache of frames into a random order.
14090 No frame is discarded.
14091 Inspired by @ref{frei0r} nervous filter.
14095 Set size in number of frames of internal cache, in range from @code{2} to
14096 @code{512}. Default is @code{30}.
14099 Set seed for random number generator, must be an integer included between
14100 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14101 less than @code{0}, the filter will try to use a good random seed on a
14105 @section readeia608
14107 Read closed captioning (EIA-608) information from the top lines of a video frame.
14109 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14110 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14111 with EIA-608 data (starting from 0). A description of each metadata value follows:
14114 @item lavfi.readeia608.X.cc
14115 The two bytes stored as EIA-608 data (printed in hexadecimal).
14117 @item lavfi.readeia608.X.line
14118 The number of the line on which the EIA-608 data was identified and read.
14121 This filter accepts the following options:
14125 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14128 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14131 Set minimal acceptable amplitude change for sync codes detection.
14132 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14135 Set the ratio of width reserved for sync code detection.
14136 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14139 Set the max peaks height difference for sync code detection.
14140 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14143 Set max peaks period difference for sync code detection.
14144 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14147 Set the first two max start code bits differences.
14148 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14151 Set the minimum ratio of bits height compared to 3rd start code bit.
14152 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14155 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14158 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14161 Enable checking the parity bit. In the event of a parity error, the filter will output
14162 @code{0x00} for that character. Default is false.
14165 @subsection Examples
14169 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14171 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
14177 Read vertical interval timecode (VITC) information from the top lines of a
14180 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14181 timecode value, if a valid timecode has been detected. Further metadata key
14182 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14183 timecode data has been found or not.
14185 This filter accepts the following options:
14189 Set the maximum number of lines to scan for VITC data. If the value is set to
14190 @code{-1} the full video frame is scanned. Default is @code{45}.
14193 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14194 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14197 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14198 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14201 @subsection Examples
14205 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14206 draw @code{--:--:--:--} as a placeholder:
14208 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14214 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14216 Destination pixel at position (X, Y) will be picked from source (x, y) position
14217 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14218 value for pixel will be used for destination pixel.
14220 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14221 will have Xmap/Ymap video stream dimensions.
14222 Xmap and Ymap input video streams are 16bit depth, single channel.
14224 @section removegrain
14226 The removegrain filter is a spatial denoiser for progressive video.
14230 Set mode for the first plane.
14233 Set mode for the second plane.
14236 Set mode for the third plane.
14239 Set mode for the fourth plane.
14242 Range of mode is from 0 to 24. Description of each mode follows:
14246 Leave input plane unchanged. Default.
14249 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14252 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14255 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14258 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14259 This is equivalent to a median filter.
14262 Line-sensitive clipping giving the minimal change.
14265 Line-sensitive clipping, intermediate.
14268 Line-sensitive clipping, intermediate.
14271 Line-sensitive clipping, intermediate.
14274 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14277 Replaces the target pixel with the closest neighbour.
14280 [1 2 1] horizontal and vertical kernel blur.
14286 Bob mode, interpolates top field from the line where the neighbours
14287 pixels are the closest.
14290 Bob mode, interpolates bottom field from the line where the neighbours
14291 pixels are the closest.
14294 Bob mode, interpolates top field. Same as 13 but with a more complicated
14295 interpolation formula.
14298 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14299 interpolation formula.
14302 Clips the pixel with the minimum and maximum of respectively the maximum and
14303 minimum of each pair of opposite neighbour pixels.
14306 Line-sensitive clipping using opposite neighbours whose greatest distance from
14307 the current pixel is minimal.
14310 Replaces the pixel with the average of its 8 neighbours.
14313 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14316 Clips pixels using the averages of opposite neighbour.
14319 Same as mode 21 but simpler and faster.
14322 Small edge and halo removal, but reputed useless.
14328 @section removelogo
14330 Suppress a TV station logo, using an image file to determine which
14331 pixels comprise the logo. It works by filling in the pixels that
14332 comprise the logo with neighboring pixels.
14334 The filter accepts the following options:
14338 Set the filter bitmap file, which can be any image format supported by
14339 libavformat. The width and height of the image file must match those of the
14340 video stream being processed.
14343 Pixels in the provided bitmap image with a value of zero are not
14344 considered part of the logo, non-zero pixels are considered part of
14345 the logo. If you use white (255) for the logo and black (0) for the
14346 rest, you will be safe. For making the filter bitmap, it is
14347 recommended to take a screen capture of a black frame with the logo
14348 visible, and then using a threshold filter followed by the erode
14349 filter once or twice.
14351 If needed, little splotches can be fixed manually. Remember that if
14352 logo pixels are not covered, the filter quality will be much
14353 reduced. Marking too many pixels as part of the logo does not hurt as
14354 much, but it will increase the amount of blurring needed to cover over
14355 the image and will destroy more information than necessary, and extra
14356 pixels will slow things down on a large logo.
14358 @section repeatfields
14360 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14361 fields based on its value.
14365 Reverse a video clip.
14367 Warning: This filter requires memory to buffer the entire clip, so trimming
14370 @subsection Examples
14374 Take the first 5 seconds of a clip, and reverse it.
14381 Shift R/G/B/A pixels horizontally and/or vertically.
14383 The filter accepts the following options:
14386 Set amount to shift red horizontally.
14388 Set amount to shift red vertically.
14390 Set amount to shift green horizontally.
14392 Set amount to shift green vertically.
14394 Set amount to shift blue horizontally.
14396 Set amount to shift blue vertically.
14398 Set amount to shift alpha horizontally.
14400 Set amount to shift alpha vertically.
14402 Set edge mode, can be @var{smear}, default, or @var{warp}.
14406 Apply roberts cross operator to input video stream.
14408 The filter accepts the following option:
14412 Set which planes will be processed, unprocessed planes will be copied.
14413 By default value 0xf, all planes will be processed.
14416 Set value which will be multiplied with filtered result.
14419 Set value which will be added to filtered result.
14424 Rotate video by an arbitrary angle expressed in radians.
14426 The filter accepts the following options:
14428 A description of the optional parameters follows.
14431 Set an expression for the angle by which to rotate the input video
14432 clockwise, expressed as a number of radians. A negative value will
14433 result in a counter-clockwise rotation. By default it is set to "0".
14435 This expression is evaluated for each frame.
14438 Set the output width expression, default value is "iw".
14439 This expression is evaluated just once during configuration.
14442 Set the output height expression, default value is "ih".
14443 This expression is evaluated just once during configuration.
14446 Enable bilinear interpolation if set to 1, a value of 0 disables
14447 it. Default value is 1.
14450 Set the color used to fill the output area not covered by the rotated
14451 image. For the general syntax of this option, check the
14452 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14453 If the special value "none" is selected then no
14454 background is printed (useful for example if the background is never shown).
14456 Default value is "black".
14459 The expressions for the angle and the output size can contain the
14460 following constants and functions:
14464 sequential number of the input frame, starting from 0. It is always NAN
14465 before the first frame is filtered.
14468 time in seconds of the input frame, it is set to 0 when the filter is
14469 configured. It is always NAN before the first frame is filtered.
14473 horizontal and vertical chroma subsample values. For example for the
14474 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14478 the input video width and height
14482 the output width and height, that is the size of the padded area as
14483 specified by the @var{width} and @var{height} expressions
14487 the minimal width/height required for completely containing the input
14488 video rotated by @var{a} radians.
14490 These are only available when computing the @option{out_w} and
14491 @option{out_h} expressions.
14494 @subsection Examples
14498 Rotate the input by PI/6 radians clockwise:
14504 Rotate the input by PI/6 radians counter-clockwise:
14510 Rotate the input by 45 degrees clockwise:
14516 Apply a constant rotation with period T, starting from an angle of PI/3:
14518 rotate=PI/3+2*PI*t/T
14522 Make the input video rotation oscillating with a period of T
14523 seconds and an amplitude of A radians:
14525 rotate=A*sin(2*PI/T*t)
14529 Rotate the video, output size is chosen so that the whole rotating
14530 input video is always completely contained in the output:
14532 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14536 Rotate the video, reduce the output size so that no background is ever
14539 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14543 @subsection Commands
14545 The filter supports the following commands:
14549 Set the angle expression.
14550 The command accepts the same syntax of the corresponding option.
14552 If the specified expression is not valid, it is kept at its current
14558 Apply Shape Adaptive Blur.
14560 The filter accepts the following options:
14563 @item luma_radius, lr
14564 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14565 value is 1.0. A greater value will result in a more blurred image, and
14566 in slower processing.
14568 @item luma_pre_filter_radius, lpfr
14569 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14572 @item luma_strength, ls
14573 Set luma maximum difference between pixels to still be considered, must
14574 be a value in the 0.1-100.0 range, default value is 1.0.
14576 @item chroma_radius, cr
14577 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14578 greater value will result in a more blurred image, and in slower
14581 @item chroma_pre_filter_radius, cpfr
14582 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14584 @item chroma_strength, cs
14585 Set chroma maximum difference between pixels to still be considered,
14586 must be a value in the -0.9-100.0 range.
14589 Each chroma option value, if not explicitly specified, is set to the
14590 corresponding luma option value.
14595 Scale (resize) the input video, using the libswscale library.
14597 The scale filter forces the output display aspect ratio to be the same
14598 of the input, by changing the output sample aspect ratio.
14600 If the input image format is different from the format requested by
14601 the next filter, the scale filter will convert the input to the
14604 @subsection Options
14605 The filter accepts the following options, or any of the options
14606 supported by the libswscale scaler.
14608 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14609 the complete list of scaler options.
14614 Set the output video dimension expression. Default value is the input
14617 If the @var{width} or @var{w} value is 0, the input width is used for
14618 the output. If the @var{height} or @var{h} value is 0, the input height
14619 is used for the output.
14621 If one and only one of the values is -n with n >= 1, the scale filter
14622 will use a value that maintains the aspect ratio of the input image,
14623 calculated from the other specified dimension. After that it will,
14624 however, make sure that the calculated dimension is divisible by n and
14625 adjust the value if necessary.
14627 If both values are -n with n >= 1, the behavior will be identical to
14628 both values being set to 0 as previously detailed.
14630 See below for the list of accepted constants for use in the dimension
14634 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14638 Only evaluate expressions once during the filter initialization or when a command is processed.
14641 Evaluate expressions for each incoming frame.
14645 Default value is @samp{init}.
14649 Set the interlacing mode. It accepts the following values:
14653 Force interlaced aware scaling.
14656 Do not apply interlaced scaling.
14659 Select interlaced aware scaling depending on whether the source frames
14660 are flagged as interlaced or not.
14663 Default value is @samp{0}.
14666 Set libswscale scaling flags. See
14667 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14668 complete list of values. If not explicitly specified the filter applies
14672 @item param0, param1
14673 Set libswscale input parameters for scaling algorithms that need them. See
14674 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14675 complete documentation. If not explicitly specified the filter applies
14681 Set the video size. For the syntax of this option, check the
14682 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14684 @item in_color_matrix
14685 @item out_color_matrix
14686 Set in/output YCbCr color space type.
14688 This allows the autodetected value to be overridden as well as allows forcing
14689 a specific value used for the output and encoder.
14691 If not specified, the color space type depends on the pixel format.
14697 Choose automatically.
14700 Format conforming to International Telecommunication Union (ITU)
14701 Recommendation BT.709.
14704 Set color space conforming to the United States Federal Communications
14705 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14708 Set color space conforming to:
14712 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14715 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14718 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14723 Set color space conforming to SMPTE ST 240:1999.
14728 Set in/output YCbCr sample range.
14730 This allows the autodetected value to be overridden as well as allows forcing
14731 a specific value used for the output and encoder. If not specified, the
14732 range depends on the pixel format. Possible values:
14736 Choose automatically.
14739 Set full range (0-255 in case of 8-bit luma).
14741 @item mpeg/limited/tv
14742 Set "MPEG" range (16-235 in case of 8-bit luma).
14745 @item force_original_aspect_ratio
14746 Enable decreasing or increasing output video width or height if necessary to
14747 keep the original aspect ratio. Possible values:
14751 Scale the video as specified and disable this feature.
14754 The output video dimensions will automatically be decreased if needed.
14757 The output video dimensions will automatically be increased if needed.
14761 One useful instance of this option is that when you know a specific device's
14762 maximum allowed resolution, you can use this to limit the output video to
14763 that, while retaining the aspect ratio. For example, device A allows
14764 1280x720 playback, and your video is 1920x800. Using this option (set it to
14765 decrease) and specifying 1280x720 to the command line makes the output
14768 Please note that this is a different thing than specifying -1 for @option{w}
14769 or @option{h}, you still need to specify the output resolution for this option
14774 The values of the @option{w} and @option{h} options are expressions
14775 containing the following constants:
14780 The input width and height
14784 These are the same as @var{in_w} and @var{in_h}.
14788 The output (scaled) width and height
14792 These are the same as @var{out_w} and @var{out_h}
14795 The same as @var{iw} / @var{ih}
14798 input sample aspect ratio
14801 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14805 horizontal and vertical input chroma subsample values. For example for the
14806 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14810 horizontal and vertical output chroma subsample values. For example for the
14811 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14814 @subsection Examples
14818 Scale the input video to a size of 200x100
14823 This is equivalent to:
14834 Specify a size abbreviation for the output size:
14839 which can also be written as:
14845 Scale the input to 2x:
14847 scale=w=2*iw:h=2*ih
14851 The above is the same as:
14853 scale=2*in_w:2*in_h
14857 Scale the input to 2x with forced interlaced scaling:
14859 scale=2*iw:2*ih:interl=1
14863 Scale the input to half size:
14865 scale=w=iw/2:h=ih/2
14869 Increase the width, and set the height to the same size:
14875 Seek Greek harmony:
14882 Increase the height, and set the width to 3/2 of the height:
14884 scale=w=3/2*oh:h=3/5*ih
14888 Increase the size, making the size a multiple of the chroma
14891 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
14895 Increase the width to a maximum of 500 pixels,
14896 keeping the same aspect ratio as the input:
14898 scale=w='min(500\, iw*3/2):h=-1'
14902 Make pixels square by combining scale and setsar:
14904 scale='trunc(ih*dar):ih',setsar=1/1
14908 Make pixels square by combining scale and setsar,
14909 making sure the resulting resolution is even (required by some codecs):
14911 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
14915 @subsection Commands
14917 This filter supports the following commands:
14921 Set the output video dimension expression.
14922 The command accepts the same syntax of the corresponding option.
14924 If the specified expression is not valid, it is kept at its current
14930 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
14931 format conversion on CUDA video frames. Setting the output width and height
14932 works in the same way as for the @var{scale} filter.
14934 The following additional options are accepted:
14937 The pixel format of the output CUDA frames. If set to the string "same" (the
14938 default), the input format will be kept. Note that automatic format negotiation
14939 and conversion is not yet supported for hardware frames
14942 The interpolation algorithm used for resizing. One of the following:
14949 @item cubic2p_bspline
14950 2-parameter cubic (B=1, C=0)
14952 @item cubic2p_catmullrom
14953 2-parameter cubic (B=0, C=1/2)
14955 @item cubic2p_b05c03
14956 2-parameter cubic (B=1/2, C=3/10)
14968 Scale (resize) the input video, based on a reference video.
14970 See the scale filter for available options, scale2ref supports the same but
14971 uses the reference video instead of the main input as basis. scale2ref also
14972 supports the following additional constants for the @option{w} and
14973 @option{h} options:
14978 The main input video's width and height
14981 The same as @var{main_w} / @var{main_h}
14984 The main input video's sample aspect ratio
14986 @item main_dar, mdar
14987 The main input video's display aspect ratio. Calculated from
14988 @code{(main_w / main_h) * main_sar}.
14992 The main input video's horizontal and vertical chroma subsample values.
14993 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
14997 @subsection Examples
15001 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15003 'scale2ref[b][a];[a][b]overlay'
15007 @anchor{selectivecolor}
15008 @section selectivecolor
15010 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15011 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15012 by the "purity" of the color (that is, how saturated it already is).
15014 This filter is similar to the Adobe Photoshop Selective Color tool.
15016 The filter accepts the following options:
15019 @item correction_method
15020 Select color correction method.
15022 Available values are:
15025 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15028 Specified adjustments are relative to the original component value.
15030 Default is @code{absolute}.
15032 Adjustments for red pixels (pixels where the red component is the maximum)
15034 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15036 Adjustments for green pixels (pixels where the green component is the maximum)
15038 Adjustments for cyan pixels (pixels where the red component is the minimum)
15040 Adjustments for blue pixels (pixels where the blue component is the maximum)
15042 Adjustments for magenta pixels (pixels where the green component is the minimum)
15044 Adjustments for white pixels (pixels where all components are greater than 128)
15046 Adjustments for all pixels except pure black and pure white
15048 Adjustments for black pixels (pixels where all components are lesser than 128)
15050 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15053 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15054 4 space separated floating point adjustment values in the [-1,1] range,
15055 respectively to adjust the amount of cyan, magenta, yellow and black for the
15056 pixels of its range.
15058 @subsection Examples
15062 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15063 increase magenta by 27% in blue areas:
15065 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15069 Use a Photoshop selective color preset:
15071 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15075 @anchor{separatefields}
15076 @section separatefields
15078 The @code{separatefields} takes a frame-based video input and splits
15079 each frame into its components fields, producing a new half height clip
15080 with twice the frame rate and twice the frame count.
15082 This filter use field-dominance information in frame to decide which
15083 of each pair of fields to place first in the output.
15084 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15086 @section setdar, setsar
15088 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15091 This is done by changing the specified Sample (aka Pixel) Aspect
15092 Ratio, according to the following equation:
15094 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15097 Keep in mind that the @code{setdar} filter does not modify the pixel
15098 dimensions of the video frame. Also, the display aspect ratio set by
15099 this filter may be changed by later filters in the filterchain,
15100 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15103 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15104 the filter output video.
15106 Note that as a consequence of the application of this filter, the
15107 output display aspect ratio will change according to the equation
15110 Keep in mind that the sample aspect ratio set by the @code{setsar}
15111 filter may be changed by later filters in the filterchain, e.g. if
15112 another "setsar" or a "setdar" filter is applied.
15114 It accepts the following parameters:
15117 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15118 Set the aspect ratio used by the filter.
15120 The parameter can be a floating point number string, an expression, or
15121 a string of the form @var{num}:@var{den}, where @var{num} and
15122 @var{den} are the numerator and denominator of the aspect ratio. If
15123 the parameter is not specified, it is assumed the value "0".
15124 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15128 Set the maximum integer value to use for expressing numerator and
15129 denominator when reducing the expressed aspect ratio to a rational.
15130 Default value is @code{100}.
15134 The parameter @var{sar} is an expression containing
15135 the following constants:
15139 These are approximated values for the mathematical constants e
15140 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15143 The input width and height.
15146 These are the same as @var{w} / @var{h}.
15149 The input sample aspect ratio.
15152 The input display aspect ratio. It is the same as
15153 (@var{w} / @var{h}) * @var{sar}.
15156 Horizontal and vertical chroma subsample values. For example, for the
15157 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15160 @subsection Examples
15165 To change the display aspect ratio to 16:9, specify one of the following:
15172 To change the sample aspect ratio to 10:11, specify:
15178 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15179 1000 in the aspect ratio reduction, use the command:
15181 setdar=ratio=16/9:max=1000
15189 Force field for the output video frame.
15191 The @code{setfield} filter marks the interlace type field for the
15192 output frames. It does not change the input frame, but only sets the
15193 corresponding property, which affects how the frame is treated by
15194 following filters (e.g. @code{fieldorder} or @code{yadif}).
15196 The filter accepts the following options:
15201 Available values are:
15205 Keep the same field property.
15208 Mark the frame as bottom-field-first.
15211 Mark the frame as top-field-first.
15214 Mark the frame as progressive.
15221 Force frame parameter for the output video frame.
15223 The @code{setparams} filter marks interlace and color range for the
15224 output frames. It does not change the input frame, but only sets the
15225 corresponding property, which affects how the frame is treated by
15230 Available values are:
15234 Keep the same field property (default).
15237 Mark the frame as bottom-field-first.
15240 Mark the frame as top-field-first.
15243 Mark the frame as progressive.
15247 Available values are:
15251 Keep the same color range property (default).
15253 @item unspecified, unknown
15254 Mark the frame as unspecified color range.
15256 @item limited, tv, mpeg
15257 Mark the frame as limited range.
15259 @item full, pc, jpeg
15260 Mark the frame as full range.
15263 @item color_primaries
15264 Set the color primaries.
15265 Available values are:
15269 Keep the same color primaries property (default).
15286 Set the color transfert.
15287 Available values are:
15291 Keep the same color trc property (default).
15313 Set the colorspace.
15314 Available values are:
15318 Keep the same colorspace property (default).
15331 @item chroma-derived-nc
15332 @item chroma-derived-c
15339 Show a line containing various information for each input video frame.
15340 The input video is not modified.
15342 This filter supports the following options:
15346 Calculate checksums of each plane. By default enabled.
15349 The shown line contains a sequence of key/value pairs of the form
15350 @var{key}:@var{value}.
15352 The following values are shown in the output:
15356 The (sequential) number of the input frame, starting from 0.
15359 The Presentation TimeStamp of the input frame, expressed as a number of
15360 time base units. The time base unit depends on the filter input pad.
15363 The Presentation TimeStamp of the input frame, expressed as a number of
15367 The position of the frame in the input stream, or -1 if this information is
15368 unavailable and/or meaningless (for example in case of synthetic video).
15371 The pixel format name.
15374 The sample aspect ratio of the input frame, expressed in the form
15375 @var{num}/@var{den}.
15378 The size of the input frame. For the syntax of this option, check the
15379 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15382 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15383 for bottom field first).
15386 This is 1 if the frame is a key frame, 0 otherwise.
15389 The picture type of the input frame ("I" for an I-frame, "P" for a
15390 P-frame, "B" for a B-frame, or "?" for an unknown type).
15391 Also refer to the documentation of the @code{AVPictureType} enum and of
15392 the @code{av_get_picture_type_char} function defined in
15393 @file{libavutil/avutil.h}.
15396 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15398 @item plane_checksum
15399 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15400 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15403 @section showpalette
15405 Displays the 256 colors palette of each frame. This filter is only relevant for
15406 @var{pal8} pixel format frames.
15408 It accepts the following option:
15412 Set the size of the box used to represent one palette color entry. Default is
15413 @code{30} (for a @code{30x30} pixel box).
15416 @section shuffleframes
15418 Reorder and/or duplicate and/or drop video frames.
15420 It accepts the following parameters:
15424 Set the destination indexes of input frames.
15425 This is space or '|' separated list of indexes that maps input frames to output
15426 frames. Number of indexes also sets maximal value that each index may have.
15427 '-1' index have special meaning and that is to drop frame.
15430 The first frame has the index 0. The default is to keep the input unchanged.
15432 @subsection Examples
15436 Swap second and third frame of every three frames of the input:
15438 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15442 Swap 10th and 1st frame of every ten frames of the input:
15444 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15448 @section shuffleplanes
15450 Reorder and/or duplicate video planes.
15452 It accepts the following parameters:
15457 The index of the input plane to be used as the first output plane.
15460 The index of the input plane to be used as the second output plane.
15463 The index of the input plane to be used as the third output plane.
15466 The index of the input plane to be used as the fourth output plane.
15470 The first plane has the index 0. The default is to keep the input unchanged.
15472 @subsection Examples
15476 Swap the second and third planes of the input:
15478 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15482 @anchor{signalstats}
15483 @section signalstats
15484 Evaluate various visual metrics that assist in determining issues associated
15485 with the digitization of analog video media.
15487 By default the filter will log these metadata values:
15491 Display the minimal Y value contained within the input frame. Expressed in
15495 Display the Y value at the 10% percentile within the input frame. Expressed in
15499 Display the average Y value within the input frame. Expressed in range of
15503 Display the Y value at the 90% percentile within the input frame. Expressed in
15507 Display the maximum Y value contained within the input frame. Expressed in
15511 Display the minimal U value contained within the input frame. Expressed in
15515 Display the U value at the 10% percentile within the input frame. Expressed in
15519 Display the average U value within the input frame. Expressed in range of
15523 Display the U value at the 90% percentile within the input frame. Expressed in
15527 Display the maximum U value contained within the input frame. Expressed in
15531 Display the minimal V value contained within the input frame. Expressed in
15535 Display the V value at the 10% percentile within the input frame. Expressed in
15539 Display the average V value within the input frame. Expressed in range of
15543 Display the V value at the 90% percentile within the input frame. Expressed in
15547 Display the maximum V value contained within the input frame. Expressed in
15551 Display the minimal saturation value contained within the input frame.
15552 Expressed in range of [0-~181.02].
15555 Display the saturation value at the 10% percentile within the input frame.
15556 Expressed in range of [0-~181.02].
15559 Display the average saturation value within the input frame. Expressed in range
15563 Display the saturation value at the 90% percentile within the input frame.
15564 Expressed in range of [0-~181.02].
15567 Display the maximum saturation value contained within the input frame.
15568 Expressed in range of [0-~181.02].
15571 Display the median value for hue within the input frame. Expressed in range of
15575 Display the average value for hue within the input frame. Expressed in range of
15579 Display the average of sample value difference between all values of the Y
15580 plane in the current frame and corresponding values of the previous input frame.
15581 Expressed in range of [0-255].
15584 Display the average of sample value difference between all values of the U
15585 plane in the current frame and corresponding values of the previous input frame.
15586 Expressed in range of [0-255].
15589 Display the average of sample value difference between all values of the V
15590 plane in the current frame and corresponding values of the previous input frame.
15591 Expressed in range of [0-255].
15594 Display bit depth of Y plane in current frame.
15595 Expressed in range of [0-16].
15598 Display bit depth of U plane in current frame.
15599 Expressed in range of [0-16].
15602 Display bit depth of V plane in current frame.
15603 Expressed in range of [0-16].
15606 The filter accepts the following options:
15612 @option{stat} specify an additional form of image analysis.
15613 @option{out} output video with the specified type of pixel highlighted.
15615 Both options accept the following values:
15619 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15620 unlike the neighboring pixels of the same field. Examples of temporal outliers
15621 include the results of video dropouts, head clogs, or tape tracking issues.
15624 Identify @var{vertical line repetition}. Vertical line repetition includes
15625 similar rows of pixels within a frame. In born-digital video vertical line
15626 repetition is common, but this pattern is uncommon in video digitized from an
15627 analog source. When it occurs in video that results from the digitization of an
15628 analog source it can indicate concealment from a dropout compensator.
15631 Identify pixels that fall outside of legal broadcast range.
15635 Set the highlight color for the @option{out} option. The default color is
15639 @subsection Examples
15643 Output data of various video metrics:
15645 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15649 Output specific data about the minimum and maximum values of the Y plane per frame:
15651 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15655 Playback video while highlighting pixels that are outside of broadcast range in red.
15657 ffplay example.mov -vf signalstats="out=brng:color=red"
15661 Playback video with signalstats metadata drawn over the frame.
15663 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15666 The contents of signalstat_drawtext.txt used in the command are:
15669 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15670 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15671 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15672 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15680 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15681 input. In this case the matching between the inputs can be calculated additionally.
15682 The filter always passes through the first input. The signature of each stream can
15683 be written into a file.
15685 It accepts the following options:
15689 Enable or disable the matching process.
15691 Available values are:
15695 Disable the calculation of a matching (default).
15697 Calculate the matching for the whole video and output whether the whole video
15698 matches or only parts.
15700 Calculate only until a matching is found or the video ends. Should be faster in
15705 Set the number of inputs. The option value must be a non negative integer.
15706 Default value is 1.
15709 Set the path to which the output is written. If there is more than one input,
15710 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15711 integer), that will be replaced with the input number. If no filename is
15712 specified, no output will be written. This is the default.
15715 Choose the output format.
15717 Available values are:
15721 Use the specified binary representation (default).
15723 Use the specified xml representation.
15727 Set threshold to detect one word as similar. The option value must be an integer
15728 greater than zero. The default value is 9000.
15731 Set threshold to detect all words as similar. The option value must be an integer
15732 greater than zero. The default value is 60000.
15735 Set threshold to detect frames as similar. The option value must be an integer
15736 greater than zero. The default value is 116.
15739 Set the minimum length of a sequence in frames to recognize it as matching
15740 sequence. The option value must be a non negative integer value.
15741 The default value is 0.
15744 Set the minimum relation, that matching frames to all frames must have.
15745 The option value must be a double value between 0 and 1. The default value is 0.5.
15748 @subsection Examples
15752 To calculate the signature of an input video and store it in signature.bin:
15754 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15758 To detect whether two videos match and store the signatures in XML format in
15759 signature0.xml and signature1.xml:
15761 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 -
15769 Blur the input video without impacting the outlines.
15771 It accepts the following options:
15774 @item luma_radius, lr
15775 Set the luma radius. The option value must be a float number in
15776 the range [0.1,5.0] that specifies the variance of the gaussian filter
15777 used to blur the image (slower if larger). Default value is 1.0.
15779 @item luma_strength, ls
15780 Set the luma strength. The option value must be a float number
15781 in the range [-1.0,1.0] that configures the blurring. A value included
15782 in [0.0,1.0] will blur the image whereas a value included in
15783 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15785 @item luma_threshold, lt
15786 Set the luma threshold used as a coefficient to determine
15787 whether a pixel should be blurred or not. The option value must be an
15788 integer in the range [-30,30]. A value of 0 will filter all the image,
15789 a value included in [0,30] will filter flat areas and a value included
15790 in [-30,0] will filter edges. Default value is 0.
15792 @item chroma_radius, cr
15793 Set the chroma radius. The option value must be a float number in
15794 the range [0.1,5.0] that specifies the variance of the gaussian filter
15795 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15797 @item chroma_strength, cs
15798 Set the chroma strength. The option value must be a float number
15799 in the range [-1.0,1.0] that configures the blurring. A value included
15800 in [0.0,1.0] will blur the image whereas a value included in
15801 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15803 @item chroma_threshold, ct
15804 Set the chroma threshold used as a coefficient to determine
15805 whether a pixel should be blurred or not. The option value must be an
15806 integer in the range [-30,30]. A value of 0 will filter all the image,
15807 a value included in [0,30] will filter flat areas and a value included
15808 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15811 If a chroma option is not explicitly set, the corresponding luma value
15816 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
15818 This filter takes in input two input videos, the first input is
15819 considered the "main" source and is passed unchanged to the
15820 output. The second input is used as a "reference" video for computing
15823 Both video inputs must have the same resolution and pixel format for
15824 this filter to work correctly. Also it assumes that both inputs
15825 have the same number of frames, which are compared one by one.
15827 The filter stores the calculated SSIM of each frame.
15829 The description of the accepted parameters follows.
15832 @item stats_file, f
15833 If specified the filter will use the named file to save the SSIM of
15834 each individual frame. When filename equals "-" the data is sent to
15838 The file printed if @var{stats_file} is selected, contains a sequence of
15839 key/value pairs of the form @var{key}:@var{value} for each compared
15842 A description of each shown parameter follows:
15846 sequential number of the input frame, starting from 1
15848 @item Y, U, V, R, G, B
15849 SSIM of the compared frames for the component specified by the suffix.
15852 SSIM of the compared frames for the whole frame.
15855 Same as above but in dB representation.
15858 This filter also supports the @ref{framesync} options.
15862 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15863 [main][ref] ssim="stats_file=stats.log" [out]
15866 On this example the input file being processed is compared with the
15867 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
15868 is stored in @file{stats.log}.
15870 Another example with both psnr and ssim at same time:
15872 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
15877 Convert between different stereoscopic image formats.
15879 The filters accept the following options:
15883 Set stereoscopic image format of input.
15885 Available values for input image formats are:
15888 side by side parallel (left eye left, right eye right)
15891 side by side crosseye (right eye left, left eye right)
15894 side by side parallel with half width resolution
15895 (left eye left, right eye right)
15898 side by side crosseye with half width resolution
15899 (right eye left, left eye right)
15902 above-below (left eye above, right eye below)
15905 above-below (right eye above, left eye below)
15908 above-below with half height resolution
15909 (left eye above, right eye below)
15912 above-below with half height resolution
15913 (right eye above, left eye below)
15916 alternating frames (left eye first, right eye second)
15919 alternating frames (right eye first, left eye second)
15922 interleaved rows (left eye has top row, right eye starts on next row)
15925 interleaved rows (right eye has top row, left eye starts on next row)
15928 interleaved columns, left eye first
15931 interleaved columns, right eye first
15933 Default value is @samp{sbsl}.
15937 Set stereoscopic image format of output.
15941 side by side parallel (left eye left, right eye right)
15944 side by side crosseye (right eye left, left eye right)
15947 side by side parallel with half width resolution
15948 (left eye left, right eye right)
15951 side by side crosseye with half width resolution
15952 (right eye left, left eye right)
15955 above-below (left eye above, right eye below)
15958 above-below (right eye above, left eye below)
15961 above-below with half height resolution
15962 (left eye above, right eye below)
15965 above-below with half height resolution
15966 (right eye above, left eye below)
15969 alternating frames (left eye first, right eye second)
15972 alternating frames (right eye first, left eye second)
15975 interleaved rows (left eye has top row, right eye starts on next row)
15978 interleaved rows (right eye has top row, left eye starts on next row)
15981 anaglyph red/blue gray
15982 (red filter on left eye, blue filter on right eye)
15985 anaglyph red/green gray
15986 (red filter on left eye, green filter on right eye)
15989 anaglyph red/cyan gray
15990 (red filter on left eye, cyan filter on right eye)
15993 anaglyph red/cyan half colored
15994 (red filter on left eye, cyan filter on right eye)
15997 anaglyph red/cyan color
15998 (red filter on left eye, cyan filter on right eye)
16001 anaglyph red/cyan color optimized with the least squares projection of dubois
16002 (red filter on left eye, cyan filter on right eye)
16005 anaglyph green/magenta gray
16006 (green filter on left eye, magenta filter on right eye)
16009 anaglyph green/magenta half colored
16010 (green filter on left eye, magenta filter on right eye)
16013 anaglyph green/magenta colored
16014 (green filter on left eye, magenta filter on right eye)
16017 anaglyph green/magenta color optimized with the least squares projection of dubois
16018 (green filter on left eye, magenta filter on right eye)
16021 anaglyph yellow/blue gray
16022 (yellow filter on left eye, blue filter on right eye)
16025 anaglyph yellow/blue half colored
16026 (yellow filter on left eye, blue filter on right eye)
16029 anaglyph yellow/blue colored
16030 (yellow filter on left eye, blue filter on right eye)
16033 anaglyph yellow/blue color optimized with the least squares projection of dubois
16034 (yellow filter on left eye, blue filter on right eye)
16037 mono output (left eye only)
16040 mono output (right eye only)
16043 checkerboard, left eye first
16046 checkerboard, right eye first
16049 interleaved columns, left eye first
16052 interleaved columns, right eye first
16058 Default value is @samp{arcd}.
16061 @subsection Examples
16065 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16071 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16077 @section streamselect, astreamselect
16078 Select video or audio streams.
16080 The filter accepts the following options:
16084 Set number of inputs. Default is 2.
16087 Set input indexes to remap to outputs.
16090 @subsection Commands
16092 The @code{streamselect} and @code{astreamselect} filter supports the following
16097 Set input indexes to remap to outputs.
16100 @subsection Examples
16104 Select first 5 seconds 1st stream and rest of time 2nd stream:
16106 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16110 Same as above, but for audio:
16112 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16117 Apply sobel operator to input video stream.
16119 The filter accepts the following option:
16123 Set which planes will be processed, unprocessed planes will be copied.
16124 By default value 0xf, all planes will be processed.
16127 Set value which will be multiplied with filtered result.
16130 Set value which will be added to filtered result.
16136 Apply a simple postprocessing filter that compresses and decompresses the image
16137 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16138 and average the results.
16140 The filter accepts the following options:
16144 Set quality. This option defines the number of levels for averaging. It accepts
16145 an integer in the range 0-6. If set to @code{0}, the filter will have no
16146 effect. A value of @code{6} means the higher quality. For each increment of
16147 that value the speed drops by a factor of approximately 2. Default value is
16151 Force a constant quantization parameter. If not set, the filter will use the QP
16152 from the video stream (if available).
16155 Set thresholding mode. Available modes are:
16159 Set hard thresholding (default).
16161 Set soft thresholding (better de-ringing effect, but likely blurrier).
16164 @item use_bframe_qp
16165 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16166 option may cause flicker since the B-Frames have often larger QP. Default is
16167 @code{0} (not enabled).
16172 Scale the input by applying one of the super-resolution methods based on
16173 convolutional neural networks. Supported models:
16177 Super-Resolution Convolutional Neural Network model (SRCNN).
16178 See @url{https://arxiv.org/abs/1501.00092}.
16181 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16182 See @url{https://arxiv.org/abs/1609.05158}.
16185 Training scripts as well as scripts for model generation are provided in
16186 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16188 The filter accepts the following options:
16192 Specify which DNN backend to use for model loading and execution. This option accepts
16193 the following values:
16197 Native implementation of DNN loading and execution.
16200 TensorFlow backend. To enable this backend you
16201 need to install the TensorFlow for C library (see
16202 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16203 @code{--enable-libtensorflow}
16206 Default value is @samp{native}.
16209 Set path to model file specifying network architecture and its parameters.
16210 Note that different backends use different file formats. TensorFlow backend
16211 can load files for both formats, while native backend can load files for only
16215 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16216 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16217 input upscaled using bicubic upscaling with proper scale factor.
16223 Draw subtitles on top of input video using the libass library.
16225 To enable compilation of this filter you need to configure FFmpeg with
16226 @code{--enable-libass}. This filter also requires a build with libavcodec and
16227 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16228 Alpha) subtitles format.
16230 The filter accepts the following options:
16234 Set the filename of the subtitle file to read. It must be specified.
16236 @item original_size
16237 Specify the size of the original video, the video for which the ASS file
16238 was composed. For the syntax of this option, check the
16239 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16240 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16241 correctly scale the fonts if the aspect ratio has been changed.
16244 Set a directory path containing fonts that can be used by the filter.
16245 These fonts will be used in addition to whatever the font provider uses.
16248 Process alpha channel, by default alpha channel is untouched.
16251 Set subtitles input character encoding. @code{subtitles} filter only. Only
16252 useful if not UTF-8.
16254 @item stream_index, si
16255 Set subtitles stream index. @code{subtitles} filter only.
16258 Override default style or script info parameters of the subtitles. It accepts a
16259 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16262 If the first key is not specified, it is assumed that the first value
16263 specifies the @option{filename}.
16265 For example, to render the file @file{sub.srt} on top of the input
16266 video, use the command:
16271 which is equivalent to:
16273 subtitles=filename=sub.srt
16276 To render the default subtitles stream from file @file{video.mkv}, use:
16278 subtitles=video.mkv
16281 To render the second subtitles stream from that file, use:
16283 subtitles=video.mkv:si=1
16286 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16287 @code{DejaVu Serif}, use:
16289 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16292 @section super2xsai
16294 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16295 Interpolate) pixel art scaling algorithm.
16297 Useful for enlarging pixel art images without reducing sharpness.
16301 Swap two rectangular objects in video.
16303 This filter accepts the following options:
16313 Set 1st rect x coordinate.
16316 Set 1st rect y coordinate.
16319 Set 2nd rect x coordinate.
16322 Set 2nd rect y coordinate.
16324 All expressions are evaluated once for each frame.
16327 The all options are expressions containing the following constants:
16332 The input width and height.
16335 same as @var{w} / @var{h}
16338 input sample aspect ratio
16341 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16344 The number of the input frame, starting from 0.
16347 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16350 the position in the file of the input frame, NAN if unknown
16358 Apply telecine process to the video.
16360 This filter accepts the following options:
16369 The default value is @code{top}.
16373 A string of numbers representing the pulldown pattern you wish to apply.
16374 The default value is @code{23}.
16378 Some typical patterns:
16383 24p: 2332 (preferred)
16390 24p: 222222222223 ("Euro pulldown")
16397 Apply threshold effect to video stream.
16399 This filter needs four video streams to perform thresholding.
16400 First stream is stream we are filtering.
16401 Second stream is holding threshold values, third stream is holding min values,
16402 and last, fourth stream is holding max values.
16404 The filter accepts the following option:
16408 Set which planes will be processed, unprocessed planes will be copied.
16409 By default value 0xf, all planes will be processed.
16412 For example if first stream pixel's component value is less then threshold value
16413 of pixel component from 2nd threshold stream, third stream value will picked,
16414 otherwise fourth stream pixel component value will be picked.
16416 Using color source filter one can perform various types of thresholding:
16418 @subsection Examples
16422 Binary threshold, using gray color as threshold:
16424 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16428 Inverted binary threshold, using gray color as threshold:
16430 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16434 Truncate binary threshold, using gray color as threshold:
16436 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16440 Threshold to zero, using gray color as threshold:
16442 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16446 Inverted threshold to zero, using gray color as threshold:
16448 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16453 Select the most representative frame in a given sequence of consecutive frames.
16455 The filter accepts the following options:
16459 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16460 will pick one of them, and then handle the next batch of @var{n} frames until
16461 the end. Default is @code{100}.
16464 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16465 value will result in a higher memory usage, so a high value is not recommended.
16467 @subsection Examples
16471 Extract one picture each 50 frames:
16477 Complete example of a thumbnail creation with @command{ffmpeg}:
16479 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16485 Tile several successive frames together.
16487 The filter accepts the following options:
16492 Set the grid size (i.e. the number of lines and columns). For the syntax of
16493 this option, check the
16494 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16497 Set the maximum number of frames to render in the given area. It must be less
16498 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16499 the area will be used.
16502 Set the outer border margin in pixels.
16505 Set the inner border thickness (i.e. the number of pixels between frames). For
16506 more advanced padding options (such as having different values for the edges),
16507 refer to the pad video filter.
16510 Specify the color of the unused area. For the syntax of this option, check the
16511 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16512 The default value of @var{color} is "black".
16515 Set the number of frames to overlap when tiling several successive frames together.
16516 The value must be between @code{0} and @var{nb_frames - 1}.
16519 Set the number of frames to initially be empty before displaying first output frame.
16520 This controls how soon will one get first output frame.
16521 The value must be between @code{0} and @var{nb_frames - 1}.
16524 @subsection Examples
16528 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16530 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16532 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16533 duplicating each output frame to accommodate the originally detected frame
16537 Display @code{5} pictures in an area of @code{3x2} frames,
16538 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16539 mixed flat and named options:
16541 tile=3x2:nb_frames=5:padding=7:margin=2
16545 @section tinterlace
16547 Perform various types of temporal field interlacing.
16549 Frames are counted starting from 1, so the first input frame is
16552 The filter accepts the following options:
16557 Specify the mode of the interlacing. This option can also be specified
16558 as a value alone. See below for a list of values for this option.
16560 Available values are:
16564 Move odd frames into the upper field, even into the lower field,
16565 generating a double height frame at half frame rate.
16569 Frame 1 Frame 2 Frame 3 Frame 4
16571 11111 22222 33333 44444
16572 11111 22222 33333 44444
16573 11111 22222 33333 44444
16574 11111 22222 33333 44444
16588 Only output odd frames, even frames are dropped, generating a frame with
16589 unchanged height at half frame rate.
16594 Frame 1 Frame 2 Frame 3 Frame 4
16596 11111 22222 33333 44444
16597 11111 22222 33333 44444
16598 11111 22222 33333 44444
16599 11111 22222 33333 44444
16609 Only output even frames, odd frames are dropped, generating a frame with
16610 unchanged height at half frame rate.
16615 Frame 1 Frame 2 Frame 3 Frame 4
16617 11111 22222 33333 44444
16618 11111 22222 33333 44444
16619 11111 22222 33333 44444
16620 11111 22222 33333 44444
16630 Expand each frame to full height, but pad alternate lines with black,
16631 generating a frame with double height at the same input frame rate.
16636 Frame 1 Frame 2 Frame 3 Frame 4
16638 11111 22222 33333 44444
16639 11111 22222 33333 44444
16640 11111 22222 33333 44444
16641 11111 22222 33333 44444
16644 11111 ..... 33333 .....
16645 ..... 22222 ..... 44444
16646 11111 ..... 33333 .....
16647 ..... 22222 ..... 44444
16648 11111 ..... 33333 .....
16649 ..... 22222 ..... 44444
16650 11111 ..... 33333 .....
16651 ..... 22222 ..... 44444
16655 @item interleave_top, 4
16656 Interleave the upper field from odd frames with the lower field from
16657 even frames, generating a frame with unchanged height at half frame rate.
16662 Frame 1 Frame 2 Frame 3 Frame 4
16664 11111<- 22222 33333<- 44444
16665 11111 22222<- 33333 44444<-
16666 11111<- 22222 33333<- 44444
16667 11111 22222<- 33333 44444<-
16677 @item interleave_bottom, 5
16678 Interleave the lower field from odd frames with the upper field from
16679 even frames, generating a frame with unchanged height at half frame rate.
16684 Frame 1 Frame 2 Frame 3 Frame 4
16686 11111 22222<- 33333 44444<-
16687 11111<- 22222 33333<- 44444
16688 11111 22222<- 33333 44444<-
16689 11111<- 22222 33333<- 44444
16699 @item interlacex2, 6
16700 Double frame rate with unchanged height. Frames are inserted each
16701 containing the second temporal field from the previous input frame and
16702 the first temporal field from the next input frame. This mode relies on
16703 the top_field_first flag. Useful for interlaced video displays with no
16704 field synchronisation.
16709 Frame 1 Frame 2 Frame 3 Frame 4
16711 11111 22222 33333 44444
16712 11111 22222 33333 44444
16713 11111 22222 33333 44444
16714 11111 22222 33333 44444
16717 11111 22222 22222 33333 33333 44444 44444
16718 11111 11111 22222 22222 33333 33333 44444
16719 11111 22222 22222 33333 33333 44444 44444
16720 11111 11111 22222 22222 33333 33333 44444
16725 Move odd frames into the upper field, even into the lower field,
16726 generating a double height frame at same frame rate.
16731 Frame 1 Frame 2 Frame 3 Frame 4
16733 11111 22222 33333 44444
16734 11111 22222 33333 44444
16735 11111 22222 33333 44444
16736 11111 22222 33333 44444
16739 11111 33333 33333 55555
16740 22222 22222 44444 44444
16741 11111 33333 33333 55555
16742 22222 22222 44444 44444
16743 11111 33333 33333 55555
16744 22222 22222 44444 44444
16745 11111 33333 33333 55555
16746 22222 22222 44444 44444
16751 Numeric values are deprecated but are accepted for backward
16752 compatibility reasons.
16754 Default mode is @code{merge}.
16757 Specify flags influencing the filter process.
16759 Available value for @var{flags} is:
16762 @item low_pass_filter, vlfp
16763 Enable linear vertical low-pass filtering in the filter.
16764 Vertical low-pass filtering is required when creating an interlaced
16765 destination from a progressive source which contains high-frequency
16766 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16769 @item complex_filter, cvlfp
16770 Enable complex vertical low-pass filtering.
16771 This will slightly less reduce interlace 'twitter' and Moire
16772 patterning but better retain detail and subjective sharpness impression.
16776 Vertical low-pass filtering can only be enabled for @option{mode}
16777 @var{interleave_top} and @var{interleave_bottom}.
16783 Mix successive video frames.
16785 A description of the accepted options follows.
16789 The number of successive frames to mix. If unspecified, it defaults to 3.
16792 Specify weight of each input video frame.
16793 Each weight is separated by space. If number of weights is smaller than
16794 number of @var{frames} last specified weight will be used for all remaining
16798 Specify scale, if it is set it will be multiplied with sum
16799 of each weight multiplied with pixel values to give final destination
16800 pixel value. By default @var{scale} is auto scaled to sum of weights.
16803 @subsection Examples
16807 Average 7 successive frames:
16809 tmix=frames=7:weights="1 1 1 1 1 1 1"
16813 Apply simple temporal convolution:
16815 tmix=frames=3:weights="-1 3 -1"
16819 Similar as above but only showing temporal differences:
16821 tmix=frames=3:weights="-1 2 -1":scale=1
16827 Tone map colors from different dynamic ranges.
16829 This filter expects data in single precision floating point, as it needs to
16830 operate on (and can output) out-of-range values. Another filter, such as
16831 @ref{zscale}, is needed to convert the resulting frame to a usable format.
16833 The tonemapping algorithms implemented only work on linear light, so input
16834 data should be linearized beforehand (and possibly correctly tagged).
16837 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
16840 @subsection Options
16841 The filter accepts the following options.
16845 Set the tone map algorithm to use.
16847 Possible values are:
16850 Do not apply any tone map, only desaturate overbright pixels.
16853 Hard-clip any out-of-range values. Use it for perfect color accuracy for
16854 in-range values, while distorting out-of-range values.
16857 Stretch the entire reference gamut to a linear multiple of the display.
16860 Fit a logarithmic transfer between the tone curves.
16863 Preserve overall image brightness with a simple curve, using nonlinear
16864 contrast, which results in flattening details and degrading color accuracy.
16867 Preserve both dark and bright details better than @var{reinhard}, at the cost
16868 of slightly darkening everything. Use it when detail preservation is more
16869 important than color and brightness accuracy.
16872 Smoothly map out-of-range values, while retaining contrast and colors for
16873 in-range material as much as possible. Use it when color accuracy is more
16874 important than detail preservation.
16880 Tune the tone mapping algorithm.
16882 This affects the following algorithms:
16888 Specifies the scale factor to use while stretching.
16892 Specifies the exponent of the function.
16896 Specify an extra linear coefficient to multiply into the signal before clipping.
16900 Specify the local contrast coefficient at the display peak.
16901 Default to 0.5, which means that in-gamut values will be about half as bright
16908 Specify the transition point from linear to mobius transform. Every value
16909 below this point is guaranteed to be mapped 1:1. The higher the value, the
16910 more accurate the result will be, at the cost of losing bright details.
16911 Default to 0.3, which due to the steep initial slope still preserves in-range
16912 colors fairly accurately.
16916 Apply desaturation for highlights that exceed this level of brightness. The
16917 higher the parameter, the more color information will be preserved. This
16918 setting helps prevent unnaturally blown-out colors for super-highlights, by
16919 (smoothly) turning into white instead. This makes images feel more natural,
16920 at the cost of reducing information about out-of-range colors.
16922 The default of 2.0 is somewhat conservative and will mostly just apply to
16923 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
16925 This option works only if the input frame has a supported color tag.
16928 Override signal/nominal/reference peak with this value. Useful when the
16929 embedded peak information in display metadata is not reliable or when tone
16930 mapping from a lower range to a higher range.
16935 Temporarily pad video frames.
16937 The filter accepts the following options:
16941 Specify number of delay frames before input video stream.
16944 Specify number of padding frames after input video stream.
16945 Set to -1 to pad indefinitely.
16948 Set kind of frames added to beginning of stream.
16949 Can be either @var{add} or @var{clone}.
16950 With @var{add} frames of solid-color are added.
16951 With @var{clone} frames are clones of first frame.
16954 Set kind of frames added to end of stream.
16955 Can be either @var{add} or @var{clone}.
16956 With @var{add} frames of solid-color are added.
16957 With @var{clone} frames are clones of last frame.
16959 @item start_duration, stop_duration
16960 Specify the duration of the start/stop delay. See
16961 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
16962 for the accepted syntax.
16963 These options override @var{start} and @var{stop}.
16966 Specify the color of the padded area. For the syntax of this option,
16967 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
16968 manual,ffmpeg-utils}.
16970 The default value of @var{color} is "black".
16976 Transpose rows with columns in the input video and optionally flip it.
16978 It accepts the following parameters:
16983 Specify the transposition direction.
16985 Can assume the following values:
16987 @item 0, 4, cclock_flip
16988 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
16996 Rotate by 90 degrees clockwise, that is:
17004 Rotate by 90 degrees counterclockwise, that is:
17011 @item 3, 7, clock_flip
17012 Rotate by 90 degrees clockwise and vertically flip, that is:
17020 For values between 4-7, the transposition is only done if the input
17021 video geometry is portrait and not landscape. These values are
17022 deprecated, the @code{passthrough} option should be used instead.
17024 Numerical values are deprecated, and should be dropped in favor of
17025 symbolic constants.
17028 Do not apply the transposition if the input geometry matches the one
17029 specified by the specified value. It accepts the following values:
17032 Always apply transposition.
17034 Preserve portrait geometry (when @var{height} >= @var{width}).
17036 Preserve landscape geometry (when @var{width} >= @var{height}).
17039 Default value is @code{none}.
17042 For example to rotate by 90 degrees clockwise and preserve portrait
17045 transpose=dir=1:passthrough=portrait
17048 The command above can also be specified as:
17050 transpose=1:portrait
17053 @section transpose_npp
17055 Transpose rows with columns in the input video and optionally flip it.
17056 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17058 It accepts the following parameters:
17063 Specify the transposition direction.
17065 Can assume the following values:
17068 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17071 Rotate by 90 degrees clockwise.
17074 Rotate by 90 degrees counterclockwise.
17077 Rotate by 90 degrees clockwise and vertically flip.
17081 Do not apply the transposition if the input geometry matches the one
17082 specified by the specified value. It accepts the following values:
17085 Always apply transposition. (default)
17087 Preserve portrait geometry (when @var{height} >= @var{width}).
17089 Preserve landscape geometry (when @var{width} >= @var{height}).
17095 Trim the input so that the output contains one continuous subpart of the input.
17097 It accepts the following parameters:
17100 Specify the time of the start of the kept section, i.e. the frame with the
17101 timestamp @var{start} will be the first frame in the output.
17104 Specify the time of the first frame that will be dropped, i.e. the frame
17105 immediately preceding the one with the timestamp @var{end} will be the last
17106 frame in the output.
17109 This is the same as @var{start}, except this option sets the start timestamp
17110 in timebase units instead of seconds.
17113 This is the same as @var{end}, except this option sets the end timestamp
17114 in timebase units instead of seconds.
17117 The maximum duration of the output in seconds.
17120 The number of the first frame that should be passed to the output.
17123 The number of the first frame that should be dropped.
17126 @option{start}, @option{end}, and @option{duration} are expressed as time
17127 duration specifications; see
17128 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17129 for the accepted syntax.
17131 Note that the first two sets of the start/end options and the @option{duration}
17132 option look at the frame timestamp, while the _frame variants simply count the
17133 frames that pass through the filter. Also note that this filter does not modify
17134 the timestamps. If you wish for the output timestamps to start at zero, insert a
17135 setpts filter after the trim filter.
17137 If multiple start or end options are set, this filter tries to be greedy and
17138 keep all the frames that match at least one of the specified constraints. To keep
17139 only the part that matches all the constraints at once, chain multiple trim
17142 The defaults are such that all the input is kept. So it is possible to set e.g.
17143 just the end values to keep everything before the specified time.
17148 Drop everything except the second minute of input:
17150 ffmpeg -i INPUT -vf trim=60:120
17154 Keep only the first second:
17156 ffmpeg -i INPUT -vf trim=duration=1
17161 @section unpremultiply
17162 Apply alpha unpremultiply effect to input video stream using first plane
17163 of second stream as alpha.
17165 Both streams must have same dimensions and same pixel format.
17167 The filter accepts the following option:
17171 Set which planes will be processed, unprocessed planes will be copied.
17172 By default value 0xf, all planes will be processed.
17174 If the format has 1 or 2 components, then luma is bit 0.
17175 If the format has 3 or 4 components:
17176 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17177 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17178 If present, the alpha channel is always the last bit.
17181 Do not require 2nd input for processing, instead use alpha plane from input stream.
17187 Sharpen or blur the input video.
17189 It accepts the following parameters:
17192 @item luma_msize_x, lx
17193 Set the luma matrix horizontal size. It must be an odd integer between
17194 3 and 23. The default value is 5.
17196 @item luma_msize_y, ly
17197 Set the luma matrix vertical size. It must be an odd integer between 3
17198 and 23. The default value is 5.
17200 @item luma_amount, la
17201 Set the luma effect strength. It must be a floating point number, reasonable
17202 values lay between -1.5 and 1.5.
17204 Negative values will blur the input video, while positive values will
17205 sharpen it, a value of zero will disable the effect.
17207 Default value is 1.0.
17209 @item chroma_msize_x, cx
17210 Set the chroma matrix horizontal size. It must be an odd integer
17211 between 3 and 23. The default value is 5.
17213 @item chroma_msize_y, cy
17214 Set the chroma matrix vertical size. It must be an odd integer
17215 between 3 and 23. The default value is 5.
17217 @item chroma_amount, ca
17218 Set the chroma effect strength. It must be a floating point number, reasonable
17219 values lay between -1.5 and 1.5.
17221 Negative values will blur the input video, while positive values will
17222 sharpen it, a value of zero will disable the effect.
17224 Default value is 0.0.
17228 All parameters are optional and default to the equivalent of the
17229 string '5:5:1.0:5:5:0.0'.
17231 @subsection Examples
17235 Apply strong luma sharpen effect:
17237 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17241 Apply a strong blur of both luma and chroma parameters:
17243 unsharp=7:7:-2:7:7:-2
17249 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17250 the image at several (or - in the case of @option{quality} level @code{8} - all)
17251 shifts and average the results.
17253 The way this differs from the behavior of spp is that uspp actually encodes &
17254 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17255 DCT similar to MJPEG.
17257 The filter accepts the following options:
17261 Set quality. This option defines the number of levels for averaging. It accepts
17262 an integer in the range 0-8. If set to @code{0}, the filter will have no
17263 effect. A value of @code{8} means the higher quality. For each increment of
17264 that value the speed drops by a factor of approximately 2. Default value is
17268 Force a constant quantization parameter. If not set, the filter will use the QP
17269 from the video stream (if available).
17272 @section vaguedenoiser
17274 Apply a wavelet based denoiser.
17276 It transforms each frame from the video input into the wavelet domain,
17277 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17278 the obtained coefficients. It does an inverse wavelet transform after.
17279 Due to wavelet properties, it should give a nice smoothed result, and
17280 reduced noise, without blurring picture features.
17282 This filter accepts the following options:
17286 The filtering strength. The higher, the more filtered the video will be.
17287 Hard thresholding can use a higher threshold than soft thresholding
17288 before the video looks overfiltered. Default value is 2.
17291 The filtering method the filter will use.
17293 It accepts the following values:
17296 All values under the threshold will be zeroed.
17299 All values under the threshold will be zeroed. All values above will be
17300 reduced by the threshold.
17303 Scales or nullifies coefficients - intermediary between (more) soft and
17304 (less) hard thresholding.
17307 Default is garrote.
17310 Number of times, the wavelet will decompose the picture. Picture can't
17311 be decomposed beyond a particular point (typically, 8 for a 640x480
17312 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17315 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17318 A list of the planes to process. By default all planes are processed.
17321 @section vectorscope
17323 Display 2 color component values in the two dimensional graph (which is called
17326 This filter accepts the following options:
17330 Set vectorscope mode.
17332 It accepts the following values:
17335 Gray values are displayed on graph, higher brightness means more pixels have
17336 same component color value on location in graph. This is the default mode.
17339 Gray values are displayed on graph. Surrounding pixels values which are not
17340 present in video frame are drawn in gradient of 2 color components which are
17341 set by option @code{x} and @code{y}. The 3rd color component is static.
17344 Actual color components values present in video frame are displayed on graph.
17347 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17348 on graph increases value of another color component, which is luminance by
17349 default values of @code{x} and @code{y}.
17352 Actual colors present in video frame are displayed on graph. If two different
17353 colors map to same position on graph then color with higher value of component
17354 not present in graph is picked.
17357 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17358 component picked from radial gradient.
17362 Set which color component will be represented on X-axis. Default is @code{1}.
17365 Set which color component will be represented on Y-axis. Default is @code{2}.
17368 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17369 of color component which represents frequency of (X, Y) location in graph.
17374 No envelope, this is default.
17377 Instant envelope, even darkest single pixel will be clearly highlighted.
17380 Hold maximum and minimum values presented in graph over time. This way you
17381 can still spot out of range values without constantly looking at vectorscope.
17384 Peak and instant envelope combined together.
17388 Set what kind of graticule to draw.
17396 Set graticule opacity.
17399 Set graticule flags.
17403 Draw graticule for white point.
17406 Draw graticule for black point.
17409 Draw color points short names.
17413 Set background opacity.
17415 @item lthreshold, l
17416 Set low threshold for color component not represented on X or Y axis.
17417 Values lower than this value will be ignored. Default is 0.
17418 Note this value is multiplied with actual max possible value one pixel component
17419 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17422 @item hthreshold, h
17423 Set high threshold for color component not represented on X or Y axis.
17424 Values higher than this value will be ignored. Default is 1.
17425 Note this value is multiplied with actual max possible value one pixel component
17426 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17427 is 0.9 * 255 = 230.
17429 @item colorspace, c
17430 Set what kind of colorspace to use when drawing graticule.
17439 @anchor{vidstabdetect}
17440 @section vidstabdetect
17442 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17443 @ref{vidstabtransform} for pass 2.
17445 This filter generates a file with relative translation and rotation
17446 transform information about subsequent frames, which is then used by
17447 the @ref{vidstabtransform} filter.
17449 To enable compilation of this filter you need to configure FFmpeg with
17450 @code{--enable-libvidstab}.
17452 This filter accepts the following options:
17456 Set the path to the file used to write the transforms information.
17457 Default value is @file{transforms.trf}.
17460 Set how shaky the video is and how quick the camera is. It accepts an
17461 integer in the range 1-10, a value of 1 means little shakiness, a
17462 value of 10 means strong shakiness. Default value is 5.
17465 Set the accuracy of the detection process. It must be a value in the
17466 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17467 accuracy. Default value is 15.
17470 Set stepsize of the search process. The region around minimum is
17471 scanned with 1 pixel resolution. Default value is 6.
17474 Set minimum contrast. Below this value a local measurement field is
17475 discarded. Must be a floating point value in the range 0-1. Default
17479 Set reference frame number for tripod mode.
17481 If enabled, the motion of the frames is compared to a reference frame
17482 in the filtered stream, identified by the specified number. The idea
17483 is to compensate all movements in a more-or-less static scene and keep
17484 the camera view absolutely still.
17486 If set to 0, it is disabled. The frames are counted starting from 1.
17489 Show fields and transforms in the resulting frames. It accepts an
17490 integer in the range 0-2. Default value is 0, which disables any
17494 @subsection Examples
17498 Use default values:
17504 Analyze strongly shaky movie and put the results in file
17505 @file{mytransforms.trf}:
17507 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17511 Visualize the result of internal transformations in the resulting
17514 vidstabdetect=show=1
17518 Analyze a video with medium shakiness using @command{ffmpeg}:
17520 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17524 @anchor{vidstabtransform}
17525 @section vidstabtransform
17527 Video stabilization/deshaking: pass 2 of 2,
17528 see @ref{vidstabdetect} for pass 1.
17530 Read a file with transform information for each frame and
17531 apply/compensate them. Together with the @ref{vidstabdetect}
17532 filter this can be used to deshake videos. See also
17533 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17534 the @ref{unsharp} filter, see below.
17536 To enable compilation of this filter you need to configure FFmpeg with
17537 @code{--enable-libvidstab}.
17539 @subsection Options
17543 Set path to the file used to read the transforms. Default value is
17544 @file{transforms.trf}.
17547 Set the number of frames (value*2 + 1) used for lowpass filtering the
17548 camera movements. Default value is 10.
17550 For example a number of 10 means that 21 frames are used (10 in the
17551 past and 10 in the future) to smoothen the motion in the video. A
17552 larger value leads to a smoother video, but limits the acceleration of
17553 the camera (pan/tilt movements). 0 is a special case where a static
17554 camera is simulated.
17557 Set the camera path optimization algorithm.
17559 Accepted values are:
17562 gaussian kernel low-pass filter on camera motion (default)
17564 averaging on transformations
17568 Set maximal number of pixels to translate frames. Default value is -1,
17572 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17573 value is -1, meaning no limit.
17576 Specify how to deal with borders that may be visible due to movement
17579 Available values are:
17582 keep image information from previous frame (default)
17584 fill the border black
17588 Invert transforms if set to 1. Default value is 0.
17591 Consider transforms as relative to previous frame if set to 1,
17592 absolute if set to 0. Default value is 0.
17595 Set percentage to zoom. A positive value will result in a zoom-in
17596 effect, a negative value in a zoom-out effect. Default value is 0 (no
17600 Set optimal zooming to avoid borders.
17602 Accepted values are:
17607 optimal static zoom value is determined (only very strong movements
17608 will lead to visible borders) (default)
17610 optimal adaptive zoom value is determined (no borders will be
17611 visible), see @option{zoomspeed}
17614 Note that the value given at zoom is added to the one calculated here.
17617 Set percent to zoom maximally each frame (enabled when
17618 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17622 Specify type of interpolation.
17624 Available values are:
17629 linear only horizontal
17631 linear in both directions (default)
17633 cubic in both directions (slow)
17637 Enable virtual tripod mode if set to 1, which is equivalent to
17638 @code{relative=0:smoothing=0}. Default value is 0.
17640 Use also @code{tripod} option of @ref{vidstabdetect}.
17643 Increase log verbosity if set to 1. Also the detected global motions
17644 are written to the temporary file @file{global_motions.trf}. Default
17648 @subsection Examples
17652 Use @command{ffmpeg} for a typical stabilization with default values:
17654 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17657 Note the use of the @ref{unsharp} filter which is always recommended.
17660 Zoom in a bit more and load transform data from a given file:
17662 vidstabtransform=zoom=5:input="mytransforms.trf"
17666 Smoothen the video even more:
17668 vidstabtransform=smoothing=30
17674 Flip the input video vertically.
17676 For example, to vertically flip a video with @command{ffmpeg}:
17678 ffmpeg -i in.avi -vf "vflip" out.avi
17683 Detect variable frame rate video.
17685 This filter tries to detect if the input is variable or constant frame rate.
17687 At end it will output number of frames detected as having variable delta pts,
17688 and ones with constant delta pts.
17689 If there was frames with variable delta, than it will also show min and max delta
17694 Boost or alter saturation.
17696 The filter accepts the following options:
17699 Set strength of boost if positive value or strength of alter if negative value.
17700 Default is 0. Allowed range is from -2 to 2.
17703 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17706 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17709 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17712 Set the red luma coefficient.
17715 Set the green luma coefficient.
17718 Set the blue luma coefficient.
17724 Make or reverse a natural vignetting effect.
17726 The filter accepts the following options:
17730 Set lens angle expression as a number of radians.
17732 The value is clipped in the @code{[0,PI/2]} range.
17734 Default value: @code{"PI/5"}
17738 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17742 Set forward/backward mode.
17744 Available modes are:
17747 The larger the distance from the central point, the darker the image becomes.
17750 The larger the distance from the central point, the brighter the image becomes.
17751 This can be used to reverse a vignette effect, though there is no automatic
17752 detection to extract the lens @option{angle} and other settings (yet). It can
17753 also be used to create a burning effect.
17756 Default value is @samp{forward}.
17759 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17761 It accepts the following values:
17764 Evaluate expressions only once during the filter initialization.
17767 Evaluate expressions for each incoming frame. This is way slower than the
17768 @samp{init} mode since it requires all the scalers to be re-computed, but it
17769 allows advanced dynamic expressions.
17772 Default value is @samp{init}.
17775 Set dithering to reduce the circular banding effects. Default is @code{1}
17779 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17780 Setting this value to the SAR of the input will make a rectangular vignetting
17781 following the dimensions of the video.
17783 Default is @code{1/1}.
17786 @subsection Expressions
17788 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17789 following parameters.
17794 input width and height
17797 the number of input frame, starting from 0
17800 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17801 @var{TB} units, NAN if undefined
17804 frame rate of the input video, NAN if the input frame rate is unknown
17807 the PTS (Presentation TimeStamp) of the filtered video frame,
17808 expressed in seconds, NAN if undefined
17811 time base of the input video
17815 @subsection Examples
17819 Apply simple strong vignetting effect:
17825 Make a flickering vignetting:
17827 vignette='PI/4+random(1)*PI/50':eval=frame
17832 @section vmafmotion
17834 Obtain the average vmaf motion score of a video.
17835 It is one of the component filters of VMAF.
17837 The obtained average motion score is printed through the logging system.
17839 In the below example the input file @file{ref.mpg} is being processed and score
17843 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
17847 Stack input videos vertically.
17849 All streams must be of same pixel format and of same width.
17851 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
17852 to create same output.
17854 The filter accept the following option:
17858 Set number of input streams. Default is 2.
17861 If set to 1, force the output to terminate when the shortest input
17862 terminates. Default value is 0.
17867 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
17868 Deinterlacing Filter").
17870 Based on the process described by Martin Weston for BBC R&D, and
17871 implemented based on the de-interlace algorithm written by Jim
17872 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
17873 uses filter coefficients calculated by BBC R&D.
17875 There are two sets of filter coefficients, so called "simple":
17876 and "complex". Which set of filter coefficients is used can
17877 be set by passing an optional parameter:
17881 Set the interlacing filter coefficients. Accepts one of the following values:
17885 Simple filter coefficient set.
17887 More-complex filter coefficient set.
17889 Default value is @samp{complex}.
17892 Specify which frames to deinterlace. Accept one of the following values:
17896 Deinterlace all frames,
17898 Only deinterlace frames marked as interlaced.
17901 Default value is @samp{all}.
17905 Video waveform monitor.
17907 The waveform monitor plots color component intensity. By default luminance
17908 only. Each column of the waveform corresponds to a column of pixels in the
17911 It accepts the following options:
17915 Can be either @code{row}, or @code{column}. Default is @code{column}.
17916 In row mode, the graph on the left side represents color component value 0 and
17917 the right side represents value = 255. In column mode, the top side represents
17918 color component value = 0 and bottom side represents value = 255.
17921 Set intensity. Smaller values are useful to find out how many values of the same
17922 luminance are distributed across input rows/columns.
17923 Default value is @code{0.04}. Allowed range is [0, 1].
17926 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
17927 In mirrored mode, higher values will be represented on the left
17928 side for @code{row} mode and at the top for @code{column} mode. Default is
17929 @code{1} (mirrored).
17933 It accepts the following values:
17936 Presents information identical to that in the @code{parade}, except
17937 that the graphs representing color components are superimposed directly
17940 This display mode makes it easier to spot relative differences or similarities
17941 in overlapping areas of the color components that are supposed to be identical,
17942 such as neutral whites, grays, or blacks.
17945 Display separate graph for the color components side by side in
17946 @code{row} mode or one below the other in @code{column} mode.
17949 Display separate graph for the color components side by side in
17950 @code{column} mode or one below the other in @code{row} mode.
17952 Using this display mode makes it easy to spot color casts in the highlights
17953 and shadows of an image, by comparing the contours of the top and the bottom
17954 graphs of each waveform. Since whites, grays, and blacks are characterized
17955 by exactly equal amounts of red, green, and blue, neutral areas of the picture
17956 should display three waveforms of roughly equal width/height. If not, the
17957 correction is easy to perform by making level adjustments the three waveforms.
17959 Default is @code{stack}.
17961 @item components, c
17962 Set which color components to display. Default is 1, which means only luminance
17963 or red color component if input is in RGB colorspace. If is set for example to
17964 7 it will display all 3 (if) available color components.
17969 No envelope, this is default.
17972 Instant envelope, minimum and maximum values presented in graph will be easily
17973 visible even with small @code{step} value.
17976 Hold minimum and maximum values presented in graph across time. This way you
17977 can still spot out of range values without constantly looking at waveforms.
17980 Peak and instant envelope combined together.
17986 No filtering, this is default.
17989 Luma and chroma combined together.
17992 Similar as above, but shows difference between blue and red chroma.
17995 Similar as above, but use different colors.
17998 Displays only chroma.
18001 Displays actual color value on waveform.
18004 Similar as above, but with luma showing frequency of chroma values.
18008 Set which graticule to display.
18012 Do not display graticule.
18015 Display green graticule showing legal broadcast ranges.
18018 Display orange graticule showing legal broadcast ranges.
18022 Set graticule opacity.
18025 Set graticule flags.
18029 Draw numbers above lines. By default enabled.
18032 Draw dots instead of lines.
18036 Set scale used for displaying graticule.
18043 Default is digital.
18046 Set background opacity.
18049 @section weave, doubleweave
18051 The @code{weave} takes a field-based video input and join
18052 each two sequential fields into single frame, producing a new double
18053 height clip with half the frame rate and half the frame count.
18055 The @code{doubleweave} works same as @code{weave} but without
18056 halving frame rate and frame count.
18058 It accepts the following option:
18062 Set first field. Available values are:
18066 Set the frame as top-field-first.
18069 Set the frame as bottom-field-first.
18073 @subsection Examples
18077 Interlace video using @ref{select} and @ref{separatefields} filter:
18079 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18084 Apply the xBR high-quality magnification filter which is designed for pixel
18085 art. It follows a set of edge-detection rules, see
18086 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18088 It accepts the following option:
18092 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18093 @code{3xBR} and @code{4} for @code{4xBR}.
18094 Default is @code{3}.
18098 Stack video inputs into custom layout.
18100 All streams must be of same pixel format.
18102 The filter accept the following option:
18106 Set number of input streams. Default is 2.
18109 Specify layout of inputs.
18110 This option requires the desired layout configuration to be explicitly set by the user.
18111 This sets position of each video input in output. Each input
18112 is separated by '|'.
18113 The first number represents the column, and the second number represents the row.
18114 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18115 where X is video input from which to take width or height.
18116 Multiple values can be used when separated by '+'. In such
18117 case values are summed together.
18120 If set to 1, force the output to terminate when the shortest input
18121 terminates. Default value is 0.
18124 @subsection Examples
18128 Display 4 inputs into 2x2 grid,
18129 note that if inputs are of different sizes unused gaps might appear,
18130 as not all of output video is used.
18132 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18136 Display 4 inputs into 1x4 grid,
18137 note that if inputs are of different sizes unused gaps might appear,
18138 as not all of output video is used.
18140 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18144 Display 9 inputs into 3x3 grid,
18145 note that if inputs are of different sizes unused gaps might appear,
18146 as not all of output video is used.
18148 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
18155 Deinterlace the input video ("yadif" means "yet another deinterlacing
18158 It accepts the following parameters:
18164 The interlacing mode to adopt. It accepts one of the following values:
18167 @item 0, send_frame
18168 Output one frame for each frame.
18169 @item 1, send_field
18170 Output one frame for each field.
18171 @item 2, send_frame_nospatial
18172 Like @code{send_frame}, but it skips the spatial interlacing check.
18173 @item 3, send_field_nospatial
18174 Like @code{send_field}, but it skips the spatial interlacing check.
18177 The default value is @code{send_frame}.
18180 The picture field parity assumed for the input interlaced video. It accepts one
18181 of the following values:
18185 Assume the top field is first.
18187 Assume the bottom field is first.
18189 Enable automatic detection of field parity.
18192 The default value is @code{auto}.
18193 If the interlacing is unknown or the decoder does not export this information,
18194 top field first will be assumed.
18197 Specify which frames to deinterlace. Accept one of the following
18202 Deinterlace all frames.
18203 @item 1, interlaced
18204 Only deinterlace frames marked as interlaced.
18207 The default value is @code{all}.
18210 @section yadif_cuda
18212 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18213 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18216 It accepts the following parameters:
18222 The interlacing mode to adopt. It accepts one of the following values:
18225 @item 0, send_frame
18226 Output one frame for each frame.
18227 @item 1, send_field
18228 Output one frame for each field.
18229 @item 2, send_frame_nospatial
18230 Like @code{send_frame}, but it skips the spatial interlacing check.
18231 @item 3, send_field_nospatial
18232 Like @code{send_field}, but it skips the spatial interlacing check.
18235 The default value is @code{send_frame}.
18238 The picture field parity assumed for the input interlaced video. It accepts one
18239 of the following values:
18243 Assume the top field is first.
18245 Assume the bottom field is first.
18247 Enable automatic detection of field parity.
18250 The default value is @code{auto}.
18251 If the interlacing is unknown or the decoder does not export this information,
18252 top field first will be assumed.
18255 Specify which frames to deinterlace. Accept one of the following
18260 Deinterlace all frames.
18261 @item 1, interlaced
18262 Only deinterlace frames marked as interlaced.
18265 The default value is @code{all}.
18270 Apply Zoom & Pan effect.
18272 This filter accepts the following options:
18276 Set the zoom expression. Default is 1.
18280 Set the x and y expression. Default is 0.
18283 Set the duration expression in number of frames.
18284 This sets for how many number of frames effect will last for
18285 single input image.
18288 Set the output image size, default is 'hd720'.
18291 Set the output frame rate, default is '25'.
18294 Each expression can contain the following constants:
18313 Output frame count.
18317 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18318 for current input frame.
18322 'x' and 'y' of last output frame of previous input frame or 0 when there was
18323 not yet such frame (first input frame).
18326 Last calculated zoom from 'z' expression for current input frame.
18329 Last calculated zoom of last output frame of previous input frame.
18332 Number of output frames for current input frame. Calculated from 'd' expression
18333 for each input frame.
18336 number of output frames created for previous input frame
18339 Rational number: input width / input height
18342 sample aspect ratio
18345 display aspect ratio
18349 @subsection Examples
18353 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18355 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
18359 Zoom-in up to 1.5 and pan always at center of picture:
18361 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18365 Same as above but without pausing:
18367 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18373 Scale (resize) the input video, using the z.lib library:
18374 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18375 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18377 The zscale filter forces the output display aspect ratio to be the same
18378 as the input, by changing the output sample aspect ratio.
18380 If the input image format is different from the format requested by
18381 the next filter, the zscale filter will convert the input to the
18384 @subsection Options
18385 The filter accepts the following options.
18390 Set the output video dimension expression. Default value is the input
18393 If the @var{width} or @var{w} value is 0, the input width is used for
18394 the output. If the @var{height} or @var{h} value is 0, the input height
18395 is used for the output.
18397 If one and only one of the values is -n with n >= 1, the zscale filter
18398 will use a value that maintains the aspect ratio of the input image,
18399 calculated from the other specified dimension. After that it will,
18400 however, make sure that the calculated dimension is divisible by n and
18401 adjust the value if necessary.
18403 If both values are -n with n >= 1, the behavior will be identical to
18404 both values being set to 0 as previously detailed.
18406 See below for the list of accepted constants for use in the dimension
18410 Set the video size. For the syntax of this option, check the
18411 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18414 Set the dither type.
18416 Possible values are:
18421 @item error_diffusion
18427 Set the resize filter type.
18429 Possible values are:
18439 Default is bilinear.
18442 Set the color range.
18444 Possible values are:
18451 Default is same as input.
18454 Set the color primaries.
18456 Possible values are:
18466 Default is same as input.
18469 Set the transfer characteristics.
18471 Possible values are:
18485 Default is same as input.
18488 Set the colorspace matrix.
18490 Possible value are:
18501 Default is same as input.
18504 Set the input color range.
18506 Possible values are:
18513 Default is same as input.
18515 @item primariesin, pin
18516 Set the input color primaries.
18518 Possible values are:
18528 Default is same as input.
18530 @item transferin, tin
18531 Set the input transfer characteristics.
18533 Possible values are:
18544 Default is same as input.
18546 @item matrixin, min
18547 Set the input colorspace matrix.
18549 Possible value are:
18561 Set the output chroma location.
18563 Possible values are:
18574 @item chromalin, cin
18575 Set the input chroma location.
18577 Possible values are:
18589 Set the nominal peak luminance.
18592 The values of the @option{w} and @option{h} options are expressions
18593 containing the following constants:
18598 The input width and height
18602 These are the same as @var{in_w} and @var{in_h}.
18606 The output (scaled) width and height
18610 These are the same as @var{out_w} and @var{out_h}
18613 The same as @var{iw} / @var{ih}
18616 input sample aspect ratio
18619 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18623 horizontal and vertical input chroma subsample values. For example for the
18624 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18628 horizontal and vertical output chroma subsample values. For example for the
18629 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18635 @c man end VIDEO FILTERS
18637 @chapter OpenCL Video Filters
18638 @c man begin OPENCL VIDEO FILTERS
18640 Below is a description of the currently available OpenCL video filters.
18642 To enable compilation of these filters you need to configure FFmpeg with
18643 @code{--enable-opencl}.
18645 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18648 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18649 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18650 given device parameters.
18652 @item -filter_hw_device @var{name}
18653 Pass the hardware device called @var{name} to all filters in any filter graph.
18657 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18661 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18663 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18667 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.
18669 @section avgblur_opencl
18671 Apply average blur filter.
18673 The filter accepts the following options:
18677 Set horizontal radius size.
18678 Range is @code{[1, 1024]} and default value is @code{1}.
18681 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18684 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18687 @subsection Example
18691 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.
18693 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18697 @section boxblur_opencl
18699 Apply a boxblur algorithm to the input video.
18701 It accepts the following parameters:
18705 @item luma_radius, lr
18706 @item luma_power, lp
18707 @item chroma_radius, cr
18708 @item chroma_power, cp
18709 @item alpha_radius, ar
18710 @item alpha_power, ap
18714 A description of the accepted options follows.
18717 @item luma_radius, lr
18718 @item chroma_radius, cr
18719 @item alpha_radius, ar
18720 Set an expression for the box radius in pixels used for blurring the
18721 corresponding input plane.
18723 The radius value must be a non-negative number, and must not be
18724 greater than the value of the expression @code{min(w,h)/2} for the
18725 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18728 Default value for @option{luma_radius} is "2". If not specified,
18729 @option{chroma_radius} and @option{alpha_radius} default to the
18730 corresponding value set for @option{luma_radius}.
18732 The expressions can contain the following constants:
18736 The input width and height in pixels.
18740 The input chroma image width and height in pixels.
18744 The horizontal and vertical chroma subsample values. For example, for the
18745 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
18748 @item luma_power, lp
18749 @item chroma_power, cp
18750 @item alpha_power, ap
18751 Specify how many times the boxblur filter is applied to the
18752 corresponding plane.
18754 Default value for @option{luma_power} is 2. If not specified,
18755 @option{chroma_power} and @option{alpha_power} default to the
18756 corresponding value set for @option{luma_power}.
18758 A value of 0 will disable the effect.
18761 @subsection Examples
18763 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.
18767 Apply a boxblur filter with the luma, chroma, and alpha radius
18768 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.
18770 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
18771 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
18775 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.
18777 For the luma plane, a 2x2 box radius will be run once.
18779 For the chroma plane, a 4x4 box radius will be run 5 times.
18781 For the alpha plane, a 3x3 box radius will be run 7 times.
18783 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
18787 @section convolution_opencl
18789 Apply convolution of 3x3, 5x5, 7x7 matrix.
18791 The filter accepts the following options:
18798 Set matrix for each plane.
18799 Matrix is sequence of 9, 25 or 49 signed numbers.
18800 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
18806 Set multiplier for calculated value for each plane.
18807 If unset or 0, it will be sum of all matrix elements.
18808 The option value must be an float number greater or equal to @code{0.0}. Default value is @code{1.0}.
18814 Set bias for each plane. This value is added to the result of the multiplication.
18815 Useful for making the overall image brighter or darker.
18816 The option value must be an float number greater or equal to @code{0.0}. Default value is @code{0.0}.
18820 @subsection Examples
18826 -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
18832 -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
18836 Apply edge enhance:
18838 -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
18844 -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
18848 Apply laplacian edge detector which includes diagonals:
18850 -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
18856 -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
18860 @section dilation_opencl
18862 Apply dilation effect to the video.
18864 This filter replaces the pixel by the local(3x3) maximum.
18866 It accepts the following options:
18873 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18874 If @code{0}, plane will remain unchanged.
18877 Flag which specifies the pixel to refer to.
18878 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
18880 Flags to local 3x3 coordinates region centered on @code{x}:
18889 @subsection Example
18893 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.
18895 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
18899 @section erosion_opencl
18901 Apply erosion effect to the video.
18903 This filter replaces the pixel by the local(3x3) minimum.
18905 It accepts the following options:
18912 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18913 If @code{0}, plane will remain unchanged.
18916 Flag which specifies the pixel to refer to.
18917 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
18919 Flags to local 3x3 coordinates region centered on @code{x}:
18928 @subsection Example
18932 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.
18934 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
18938 @section overlay_opencl
18940 Overlay one video on top of another.
18942 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
18943 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
18945 The filter accepts the following options:
18950 Set the x coordinate of the overlaid video on the main video.
18951 Default value is @code{0}.
18954 Set the x coordinate of the overlaid video on the main video.
18955 Default value is @code{0}.
18959 @subsection Examples
18963 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
18965 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
18968 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
18970 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
18975 @section prewitt_opencl
18977 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
18979 The filter accepts the following option:
18983 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18986 Set value which will be multiplied with filtered result.
18987 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
18990 Set value which will be added to filtered result.
18991 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
18994 @subsection Example
18998 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19000 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19004 @section roberts_opencl
19005 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19007 The filter accepts the following option:
19011 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19014 Set value which will be multiplied with filtered result.
19015 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19018 Set value which will be added to filtered result.
19019 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19022 @subsection Example
19026 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19028 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19032 @section sobel_opencl
19034 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19036 The filter accepts the following option:
19040 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19043 Set value which will be multiplied with filtered result.
19044 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19047 Set value which will be added to filtered result.
19048 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19051 @subsection Example
19055 Apply sobel operator with scale set to 2 and delta set to 10
19057 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19061 @section tonemap_opencl
19063 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19065 It accepts the following parameters:
19069 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19072 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19075 Apply desaturation for highlights that exceed this level of brightness. The
19076 higher the parameter, the more color information will be preserved. This
19077 setting helps prevent unnaturally blown-out colors for super-highlights, by
19078 (smoothly) turning into white instead. This makes images feel more natural,
19079 at the cost of reducing information about out-of-range colors.
19081 The default value is 0.5, and the algorithm here is a little different from
19082 the cpu version tonemap currently. A setting of 0.0 disables this option.
19085 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19086 is used to detect whether the scene has changed or not. If the distance beween
19087 the current frame average brightness and the current running average exceeds
19088 a threshold value, we would re-calculate scene average and peak brightness.
19089 The default value is 0.2.
19092 Specify the output pixel format.
19094 Currently supported formats are:
19101 Set the output color range.
19103 Possible values are:
19109 Default is same as input.
19112 Set the output color primaries.
19114 Possible values are:
19120 Default is same as input.
19123 Set the output transfer characteristics.
19125 Possible values are:
19134 Set the output colorspace matrix.
19136 Possible value are:
19142 Default is same as input.
19146 @subsection Example
19150 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19152 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19156 @section unsharp_opencl
19158 Sharpen or blur the input video.
19160 It accepts the following parameters:
19163 @item luma_msize_x, lx
19164 Set the luma matrix horizontal size.
19165 Range is @code{[1, 23]} and default value is @code{5}.
19167 @item luma_msize_y, ly
19168 Set the luma matrix vertical size.
19169 Range is @code{[1, 23]} and default value is @code{5}.
19171 @item luma_amount, la
19172 Set the luma effect strength.
19173 Range is @code{[-10, 10]} and default value is @code{1.0}.
19175 Negative values will blur the input video, while positive values will
19176 sharpen it, a value of zero will disable the effect.
19178 @item chroma_msize_x, cx
19179 Set the chroma matrix horizontal size.
19180 Range is @code{[1, 23]} and default value is @code{5}.
19182 @item chroma_msize_y, cy
19183 Set the chroma matrix vertical size.
19184 Range is @code{[1, 23]} and default value is @code{5}.
19186 @item chroma_amount, ca
19187 Set the chroma effect strength.
19188 Range is @code{[-10, 10]} and default value is @code{0.0}.
19190 Negative values will blur the input video, while positive values will
19191 sharpen it, a value of zero will disable the effect.
19195 All parameters are optional and default to the equivalent of the
19196 string '5:5:1.0:5:5:0.0'.
19198 @subsection Examples
19202 Apply strong luma sharpen effect:
19204 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19208 Apply a strong blur of both luma and chroma parameters:
19210 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19214 @c man end OPENCL VIDEO FILTERS
19216 @chapter Video Sources
19217 @c man begin VIDEO SOURCES
19219 Below is a description of the currently available video sources.
19223 Buffer video frames, and make them available to the filter chain.
19225 This source is mainly intended for a programmatic use, in particular
19226 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19228 It accepts the following parameters:
19233 Specify the size (width and height) of the buffered video frames. For the
19234 syntax of this option, check the
19235 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19238 The input video width.
19241 The input video height.
19244 A string representing the pixel format of the buffered video frames.
19245 It may be a number corresponding to a pixel format, or a pixel format
19249 Specify the timebase assumed by the timestamps of the buffered frames.
19252 Specify the frame rate expected for the video stream.
19254 @item pixel_aspect, sar
19255 The sample (pixel) aspect ratio of the input video.
19258 Specify the optional parameters to be used for the scale filter which
19259 is automatically inserted when an input change is detected in the
19260 input size or format.
19262 @item hw_frames_ctx
19263 When using a hardware pixel format, this should be a reference to an
19264 AVHWFramesContext describing input frames.
19269 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19272 will instruct the source to accept video frames with size 320x240 and
19273 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19274 square pixels (1:1 sample aspect ratio).
19275 Since the pixel format with name "yuv410p" corresponds to the number 6
19276 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19277 this example corresponds to:
19279 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19282 Alternatively, the options can be specified as a flat string, but this
19283 syntax is deprecated:
19285 @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}]
19289 Create a pattern generated by an elementary cellular automaton.
19291 The initial state of the cellular automaton can be defined through the
19292 @option{filename} and @option{pattern} options. If such options are
19293 not specified an initial state is created randomly.
19295 At each new frame a new row in the video is filled with the result of
19296 the cellular automaton next generation. The behavior when the whole
19297 frame is filled is defined by the @option{scroll} option.
19299 This source accepts the following options:
19303 Read the initial cellular automaton state, i.e. the starting row, from
19304 the specified file.
19305 In the file, each non-whitespace character is considered an alive
19306 cell, a newline will terminate the row, and further characters in the
19307 file will be ignored.
19310 Read the initial cellular automaton state, i.e. the starting row, from
19311 the specified string.
19313 Each non-whitespace character in the string is considered an alive
19314 cell, a newline will terminate the row, and further characters in the
19315 string will be ignored.
19318 Set the video rate, that is the number of frames generated per second.
19321 @item random_fill_ratio, ratio
19322 Set the random fill ratio for the initial cellular automaton row. It
19323 is a floating point number value ranging from 0 to 1, defaults to
19326 This option is ignored when a file or a pattern is specified.
19328 @item random_seed, seed
19329 Set the seed for filling randomly the initial row, must be an integer
19330 included between 0 and UINT32_MAX. If not specified, or if explicitly
19331 set to -1, the filter will try to use a good random seed on a best
19335 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19336 Default value is 110.
19339 Set the size of the output video. For the syntax of this option, check the
19340 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19342 If @option{filename} or @option{pattern} is specified, the size is set
19343 by default to the width of the specified initial state row, and the
19344 height is set to @var{width} * PHI.
19346 If @option{size} is set, it must contain the width of the specified
19347 pattern string, and the specified pattern will be centered in the
19350 If a filename or a pattern string is not specified, the size value
19351 defaults to "320x518" (used for a randomly generated initial state).
19354 If set to 1, scroll the output upward when all the rows in the output
19355 have been already filled. If set to 0, the new generated row will be
19356 written over the top row just after the bottom row is filled.
19359 @item start_full, full
19360 If set to 1, completely fill the output with generated rows before
19361 outputting the first frame.
19362 This is the default behavior, for disabling set the value to 0.
19365 If set to 1, stitch the left and right row edges together.
19366 This is the default behavior, for disabling set the value to 0.
19369 @subsection Examples
19373 Read the initial state from @file{pattern}, and specify an output of
19376 cellauto=f=pattern:s=200x400
19380 Generate a random initial row with a width of 200 cells, with a fill
19383 cellauto=ratio=2/3:s=200x200
19387 Create a pattern generated by rule 18 starting by a single alive cell
19388 centered on an initial row with width 100:
19390 cellauto=p=@@:s=100x400:full=0:rule=18
19394 Specify a more elaborated initial pattern:
19396 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19401 @anchor{coreimagesrc}
19402 @section coreimagesrc
19403 Video source generated on GPU using Apple's CoreImage API on OSX.
19405 This video source is a specialized version of the @ref{coreimage} video filter.
19406 Use a core image generator at the beginning of the applied filterchain to
19407 generate the content.
19409 The coreimagesrc video source accepts the following options:
19411 @item list_generators
19412 List all available generators along with all their respective options as well as
19413 possible minimum and maximum values along with the default values.
19415 list_generators=true
19419 Specify the size of the sourced video. For the syntax of this option, check the
19420 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19421 The default value is @code{320x240}.
19424 Specify the frame rate of the sourced video, as the number of frames
19425 generated per second. It has to be a string in the format
19426 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19427 number or a valid video frame rate abbreviation. The default value is
19431 Set the sample aspect ratio of the sourced video.
19434 Set the duration of the sourced video. See
19435 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19436 for the accepted syntax.
19438 If not specified, or the expressed duration is negative, the video is
19439 supposed to be generated forever.
19442 Additionally, all options of the @ref{coreimage} video filter are accepted.
19443 A complete filterchain can be used for further processing of the
19444 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19445 and examples for details.
19447 @subsection Examples
19452 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19453 given as complete and escaped command-line for Apple's standard bash shell:
19455 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19457 This example is equivalent to the QRCode example of @ref{coreimage} without the
19458 need for a nullsrc video source.
19462 @section mandelbrot
19464 Generate a Mandelbrot set fractal, and progressively zoom towards the
19465 point specified with @var{start_x} and @var{start_y}.
19467 This source accepts the following options:
19472 Set the terminal pts value. Default value is 400.
19475 Set the terminal scale value.
19476 Must be a floating point value. Default value is 0.3.
19479 Set the inner coloring mode, that is the algorithm used to draw the
19480 Mandelbrot fractal internal region.
19482 It shall assume one of the following values:
19487 Show time until convergence.
19489 Set color based on point closest to the origin of the iterations.
19494 Default value is @var{mincol}.
19497 Set the bailout value. Default value is 10.0.
19500 Set the maximum of iterations performed by the rendering
19501 algorithm. Default value is 7189.
19504 Set outer coloring mode.
19505 It shall assume one of following values:
19507 @item iteration_count
19508 Set iteration cound mode.
19509 @item normalized_iteration_count
19510 set normalized iteration count mode.
19512 Default value is @var{normalized_iteration_count}.
19515 Set frame rate, expressed as number of frames per second. Default
19519 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19520 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19523 Set the initial scale value. Default value is 3.0.
19526 Set the initial x position. Must be a floating point value between
19527 -100 and 100. Default value is -0.743643887037158704752191506114774.
19530 Set the initial y position. Must be a floating point value between
19531 -100 and 100. Default value is -0.131825904205311970493132056385139.
19536 Generate various test patterns, as generated by the MPlayer test filter.
19538 The size of the generated video is fixed, and is 256x256.
19539 This source is useful in particular for testing encoding features.
19541 This source accepts the following options:
19546 Specify the frame rate of the sourced video, as the number of frames
19547 generated per second. It has to be a string in the format
19548 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19549 number or a valid video frame rate abbreviation. The default value is
19553 Set the duration of the sourced video. See
19554 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19555 for the accepted syntax.
19557 If not specified, or the expressed duration is negative, the video is
19558 supposed to be generated forever.
19562 Set the number or the name of the test to perform. Supported tests are:
19578 Default value is "all", which will cycle through the list of all tests.
19583 mptestsrc=t=dc_luma
19586 will generate a "dc_luma" test pattern.
19588 @section frei0r_src
19590 Provide a frei0r source.
19592 To enable compilation of this filter you need to install the frei0r
19593 header and configure FFmpeg with @code{--enable-frei0r}.
19595 This source accepts the following parameters:
19600 The size of the video to generate. For the syntax of this option, check the
19601 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19604 The framerate of the generated video. It may be a string of the form
19605 @var{num}/@var{den} or a frame rate abbreviation.
19608 The name to the frei0r source to load. For more information regarding frei0r and
19609 how to set the parameters, read the @ref{frei0r} section in the video filters
19612 @item filter_params
19613 A '|'-separated list of parameters to pass to the frei0r source.
19617 For example, to generate a frei0r partik0l source with size 200x200
19618 and frame rate 10 which is overlaid on the overlay filter main input:
19620 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19625 Generate a life pattern.
19627 This source is based on a generalization of John Conway's life game.
19629 The sourced input represents a life grid, each pixel represents a cell
19630 which can be in one of two possible states, alive or dead. Every cell
19631 interacts with its eight neighbours, which are the cells that are
19632 horizontally, vertically, or diagonally adjacent.
19634 At each interaction the grid evolves according to the adopted rule,
19635 which specifies the number of neighbor alive cells which will make a
19636 cell stay alive or born. The @option{rule} option allows one to specify
19639 This source accepts the following options:
19643 Set the file from which to read the initial grid state. In the file,
19644 each non-whitespace character is considered an alive cell, and newline
19645 is used to delimit the end of each row.
19647 If this option is not specified, the initial grid is generated
19651 Set the video rate, that is the number of frames generated per second.
19654 @item random_fill_ratio, ratio
19655 Set the random fill ratio for the initial random grid. It is a
19656 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19657 It is ignored when a file is specified.
19659 @item random_seed, seed
19660 Set the seed for filling the initial random grid, must be an integer
19661 included between 0 and UINT32_MAX. If not specified, or if explicitly
19662 set to -1, the filter will try to use a good random seed on a best
19668 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19669 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19670 @var{NS} specifies the number of alive neighbor cells which make a
19671 live cell stay alive, and @var{NB} the number of alive neighbor cells
19672 which make a dead cell to become alive (i.e. to "born").
19673 "s" and "b" can be used in place of "S" and "B", respectively.
19675 Alternatively a rule can be specified by an 18-bits integer. The 9
19676 high order bits are used to encode the next cell state if it is alive
19677 for each number of neighbor alive cells, the low order bits specify
19678 the rule for "borning" new cells. Higher order bits encode for an
19679 higher number of neighbor cells.
19680 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19681 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19683 Default value is "S23/B3", which is the original Conway's game of life
19684 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19685 cells, and will born a new cell if there are three alive cells around
19689 Set the size of the output video. For the syntax of this option, check the
19690 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19692 If @option{filename} is specified, the size is set by default to the
19693 same size of the input file. If @option{size} is set, it must contain
19694 the size specified in the input file, and the initial grid defined in
19695 that file is centered in the larger resulting area.
19697 If a filename is not specified, the size value defaults to "320x240"
19698 (used for a randomly generated initial grid).
19701 If set to 1, stitch the left and right grid edges together, and the
19702 top and bottom edges also. Defaults to 1.
19705 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
19706 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
19707 value from 0 to 255.
19710 Set the color of living (or new born) cells.
19713 Set the color of dead cells. If @option{mold} is set, this is the first color
19714 used to represent a dead cell.
19717 Set mold color, for definitely dead and moldy cells.
19719 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
19720 ffmpeg-utils manual,ffmpeg-utils}.
19723 @subsection Examples
19727 Read a grid from @file{pattern}, and center it on a grid of size
19730 life=f=pattern:s=300x300
19734 Generate a random grid of size 200x200, with a fill ratio of 2/3:
19736 life=ratio=2/3:s=200x200
19740 Specify a custom rule for evolving a randomly generated grid:
19746 Full example with slow death effect (mold) using @command{ffplay}:
19748 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
19755 @anchor{haldclutsrc}
19758 @anchor{pal100bars}
19759 @anchor{rgbtestsrc}
19761 @anchor{smptehdbars}
19764 @anchor{yuvtestsrc}
19765 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
19767 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
19769 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
19771 The @code{color} source provides an uniformly colored input.
19773 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
19774 @ref{haldclut} filter.
19776 The @code{nullsrc} source returns unprocessed video frames. It is
19777 mainly useful to be employed in analysis / debugging tools, or as the
19778 source for filters which ignore the input data.
19780 The @code{pal75bars} source generates a color bars pattern, based on
19781 EBU PAL recommendations with 75% color levels.
19783 The @code{pal100bars} source generates a color bars pattern, based on
19784 EBU PAL recommendations with 100% color levels.
19786 The @code{rgbtestsrc} source generates an RGB test pattern useful for
19787 detecting RGB vs BGR issues. You should see a red, green and blue
19788 stripe from top to bottom.
19790 The @code{smptebars} source generates a color bars pattern, based on
19791 the SMPTE Engineering Guideline EG 1-1990.
19793 The @code{smptehdbars} source generates a color bars pattern, based on
19794 the SMPTE RP 219-2002.
19796 The @code{testsrc} source generates a test video pattern, showing a
19797 color pattern, a scrolling gradient and a timestamp. This is mainly
19798 intended for testing purposes.
19800 The @code{testsrc2} source is similar to testsrc, but supports more
19801 pixel formats instead of just @code{rgb24}. This allows using it as an
19802 input for other tests without requiring a format conversion.
19804 The @code{yuvtestsrc} source generates an YUV test pattern. You should
19805 see a y, cb and cr stripe from top to bottom.
19807 The sources accept the following parameters:
19812 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
19813 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
19814 pixels to be used as identity matrix for 3D lookup tables. Each component is
19815 coded on a @code{1/(N*N)} scale.
19818 Specify the color of the source, only available in the @code{color}
19819 source. For the syntax of this option, check the
19820 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19823 Specify the size of the sourced video. For the syntax of this option, check the
19824 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19825 The default value is @code{320x240}.
19827 This option is not available with the @code{allrgb}, @code{allyuv}, and
19828 @code{haldclutsrc} filters.
19831 Specify the frame rate of the sourced video, as the number of frames
19832 generated per second. It has to be a string in the format
19833 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19834 number or a valid video frame rate abbreviation. The default value is
19838 Set the duration of the sourced video. See
19839 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19840 for the accepted syntax.
19842 If not specified, or the expressed duration is negative, the video is
19843 supposed to be generated forever.
19846 Set the sample aspect ratio of the sourced video.
19849 Specify the alpha (opacity) of the background, only available in the
19850 @code{testsrc2} source. The value must be between 0 (fully transparent) and
19851 255 (fully opaque, the default).
19854 Set the number of decimals to show in the timestamp, only available in the
19855 @code{testsrc} source.
19857 The displayed timestamp value will correspond to the original
19858 timestamp value multiplied by the power of 10 of the specified
19859 value. Default value is 0.
19862 @subsection Examples
19866 Generate a video with a duration of 5.3 seconds, with size
19867 176x144 and a frame rate of 10 frames per second:
19869 testsrc=duration=5.3:size=qcif:rate=10
19873 The following graph description will generate a red source
19874 with an opacity of 0.2, with size "qcif" and a frame rate of 10
19877 color=c=red@@0.2:s=qcif:r=10
19881 If the input content is to be ignored, @code{nullsrc} can be used. The
19882 following command generates noise in the luminance plane by employing
19883 the @code{geq} filter:
19885 nullsrc=s=256x256, geq=random(1)*255:128:128
19889 @subsection Commands
19891 The @code{color} source supports the following commands:
19895 Set the color of the created image. Accepts the same syntax of the
19896 corresponding @option{color} option.
19901 Generate video using an OpenCL program.
19906 OpenCL program source file.
19909 Kernel name in program.
19912 Size of frames to generate. This must be set.
19915 Pixel format to use for the generated frames. This must be set.
19918 Number of frames generated every second. Default value is '25'.
19922 For details of how the program loading works, see the @ref{program_opencl}
19929 Generate a colour ramp by setting pixel values from the position of the pixel
19930 in the output image. (Note that this will work with all pixel formats, but
19931 the generated output will not be the same.)
19933 __kernel void ramp(__write_only image2d_t dst,
19934 unsigned int index)
19936 int2 loc = (int2)(get_global_id(0), get_global_id(1));
19939 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
19941 write_imagef(dst, loc, val);
19946 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
19948 __kernel void sierpinski_carpet(__write_only image2d_t dst,
19949 unsigned int index)
19951 int2 loc = (int2)(get_global_id(0), get_global_id(1));
19953 float4 value = 0.0f;
19954 int x = loc.x + index;
19955 int y = loc.y + index;
19956 while (x > 0 || y > 0) {
19957 if (x % 3 == 1 && y % 3 == 1) {
19965 write_imagef(dst, loc, value);
19971 @c man end VIDEO SOURCES
19973 @chapter Video Sinks
19974 @c man begin VIDEO SINKS
19976 Below is a description of the currently available video sinks.
19978 @section buffersink
19980 Buffer video frames, and make them available to the end of the filter
19983 This sink is mainly intended for programmatic use, in particular
19984 through the interface defined in @file{libavfilter/buffersink.h}
19985 or the options system.
19987 It accepts a pointer to an AVBufferSinkContext structure, which
19988 defines the incoming buffers' formats, to be passed as the opaque
19989 parameter to @code{avfilter_init_filter} for initialization.
19993 Null video sink: do absolutely nothing with the input video. It is
19994 mainly useful as a template and for use in analysis / debugging
19997 @c man end VIDEO SINKS
19999 @chapter Multimedia Filters
20000 @c man begin MULTIMEDIA FILTERS
20002 Below is a description of the currently available multimedia filters.
20006 Convert input audio to a video output, displaying the audio bit scope.
20008 The filter accepts the following options:
20012 Set frame rate, expressed as number of frames per second. Default
20016 Specify the video size for the output. For the syntax of this option, check the
20017 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20018 Default value is @code{1024x256}.
20021 Specify list of colors separated by space or by '|' which will be used to
20022 draw channels. Unrecognized or missing colors will be replaced
20026 @section ahistogram
20028 Convert input audio to a video output, displaying the volume histogram.
20030 The filter accepts the following options:
20034 Specify how histogram is calculated.
20036 It accepts the following values:
20039 Use single histogram for all channels.
20041 Use separate histogram for each channel.
20043 Default is @code{single}.
20046 Set frame rate, expressed as number of frames per second. Default
20050 Specify the video size for the output. For the syntax of this option, check the
20051 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20052 Default value is @code{hd720}.
20057 It accepts the following values:
20068 reverse logarithmic
20070 Default is @code{log}.
20073 Set amplitude scale.
20075 It accepts the following values:
20082 Default is @code{log}.
20085 Set how much frames to accumulate in histogram.
20086 Defauls is 1. Setting this to -1 accumulates all frames.
20089 Set histogram ratio of window height.
20092 Set sonogram sliding.
20094 It accepts the following values:
20097 replace old rows with new ones.
20099 scroll from top to bottom.
20101 Default is @code{replace}.
20104 @section aphasemeter
20106 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20107 representing mean phase of current audio frame. A video output can also be produced and is
20108 enabled by default. The audio is passed through as first output.
20110 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20111 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20112 and @code{1} means channels are in phase.
20114 The filter accepts the following options, all related to its video output:
20118 Set the output frame rate. Default value is @code{25}.
20121 Set the video size for the output. For the syntax of this option, check the
20122 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20123 Default value is @code{800x400}.
20128 Specify the red, green, blue contrast. Default values are @code{2},
20129 @code{7} and @code{1}.
20130 Allowed range is @code{[0, 255]}.
20133 Set color which will be used for drawing median phase. If color is
20134 @code{none} which is default, no median phase value will be drawn.
20137 Enable video output. Default is enabled.
20140 @section avectorscope
20142 Convert input audio to a video output, representing the audio vector
20145 The filter is used to measure the difference between channels of stereo
20146 audio stream. A monoaural signal, consisting of identical left and right
20147 signal, results in straight vertical line. Any stereo separation is visible
20148 as a deviation from this line, creating a Lissajous figure.
20149 If the straight (or deviation from it) but horizontal line appears this
20150 indicates that the left and right channels are out of phase.
20152 The filter accepts the following options:
20156 Set the vectorscope mode.
20158 Available values are:
20161 Lissajous rotated by 45 degrees.
20164 Same as above but not rotated.
20167 Shape resembling half of circle.
20170 Default value is @samp{lissajous}.
20173 Set the video size for the output. For the syntax of this option, check the
20174 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20175 Default value is @code{400x400}.
20178 Set the output frame rate. Default value is @code{25}.
20184 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20185 @code{160}, @code{80} and @code{255}.
20186 Allowed range is @code{[0, 255]}.
20192 Specify the red, green, blue and alpha fade. Default values are @code{15},
20193 @code{10}, @code{5} and @code{5}.
20194 Allowed range is @code{[0, 255]}.
20197 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20198 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20201 Set the vectorscope drawing mode.
20203 Available values are:
20206 Draw dot for each sample.
20209 Draw line between previous and current sample.
20212 Default value is @samp{dot}.
20215 Specify amplitude scale of audio samples.
20217 Available values are:
20233 Swap left channel axis with right channel axis.
20243 Mirror only x axis.
20246 Mirror only y axis.
20254 @subsection Examples
20258 Complete example using @command{ffplay}:
20260 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20261 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20265 @section bench, abench
20267 Benchmark part of a filtergraph.
20269 The filter accepts the following options:
20273 Start or stop a timer.
20275 Available values are:
20278 Get the current time, set it as frame metadata (using the key
20279 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20282 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20283 the input frame metadata to get the time difference. Time difference, average,
20284 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20285 @code{min}) are then printed. The timestamps are expressed in seconds.
20289 @subsection Examples
20293 Benchmark @ref{selectivecolor} filter:
20295 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20301 Concatenate audio and video streams, joining them together one after the
20304 The filter works on segments of synchronized video and audio streams. All
20305 segments must have the same number of streams of each type, and that will
20306 also be the number of streams at output.
20308 The filter accepts the following options:
20313 Set the number of segments. Default is 2.
20316 Set the number of output video streams, that is also the number of video
20317 streams in each segment. Default is 1.
20320 Set the number of output audio streams, that is also the number of audio
20321 streams in each segment. Default is 0.
20324 Activate unsafe mode: do not fail if segments have a different format.
20328 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20329 @var{a} audio outputs.
20331 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20332 segment, in the same order as the outputs, then the inputs for the second
20335 Related streams do not always have exactly the same duration, for various
20336 reasons including codec frame size or sloppy authoring. For that reason,
20337 related synchronized streams (e.g. a video and its audio track) should be
20338 concatenated at once. The concat filter will use the duration of the longest
20339 stream in each segment (except the last one), and if necessary pad shorter
20340 audio streams with silence.
20342 For this filter to work correctly, all segments must start at timestamp 0.
20344 All corresponding streams must have the same parameters in all segments; the
20345 filtering system will automatically select a common pixel format for video
20346 streams, and a common sample format, sample rate and channel layout for
20347 audio streams, but other settings, such as resolution, must be converted
20348 explicitly by the user.
20350 Different frame rates are acceptable but will result in variable frame rate
20351 at output; be sure to configure the output file to handle it.
20353 @subsection Examples
20357 Concatenate an opening, an episode and an ending, all in bilingual version
20358 (video in stream 0, audio in streams 1 and 2):
20360 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20361 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20362 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20363 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20367 Concatenate two parts, handling audio and video separately, using the
20368 (a)movie sources, and adjusting the resolution:
20370 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20371 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20372 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20374 Note that a desync will happen at the stitch if the audio and video streams
20375 do not have exactly the same duration in the first file.
20379 @subsection Commands
20381 This filter supports the following commands:
20384 Close the current segment and step to the next one
20387 @section drawgraph, adrawgraph
20389 Draw a graph using input video or audio metadata.
20391 It accepts the following parameters:
20395 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20398 Set 1st foreground color expression.
20401 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20404 Set 2nd foreground color expression.
20407 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20410 Set 3rd foreground color expression.
20413 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20416 Set 4th foreground color expression.
20419 Set minimal value of metadata value.
20422 Set maximal value of metadata value.
20425 Set graph background color. Default is white.
20430 Available values for mode is:
20437 Default is @code{line}.
20442 Available values for slide is:
20445 Draw new frame when right border is reached.
20448 Replace old columns with new ones.
20451 Scroll from right to left.
20454 Scroll from left to right.
20457 Draw single picture.
20460 Default is @code{frame}.
20463 Set size of graph video. For the syntax of this option, check the
20464 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20465 The default value is @code{900x256}.
20467 The foreground color expressions can use the following variables:
20470 Minimal value of metadata value.
20473 Maximal value of metadata value.
20476 Current metadata key value.
20479 The color is defined as 0xAABBGGRR.
20482 Example using metadata from @ref{signalstats} filter:
20484 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20487 Example using metadata from @ref{ebur128} filter:
20489 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20495 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
20496 it unchanged. By default, it logs a message at a frequency of 10Hz with the
20497 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20498 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20500 The filter also has a video output (see the @var{video} option) with a real
20501 time graph to observe the loudness evolution. The graphic contains the logged
20502 message mentioned above, so it is not printed anymore when this option is set,
20503 unless the verbose logging is set. The main graphing area contains the
20504 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20505 the momentary loudness (400 milliseconds), but can optionally be configured
20506 to instead display short-term loudness (see @var{gauge}).
20508 The green area marks a +/- 1LU target range around the target loudness
20509 (-23LUFS by default, unless modified through @var{target}).
20511 More information about the Loudness Recommendation EBU R128 on
20512 @url{http://tech.ebu.ch/loudness}.
20514 The filter accepts the following options:
20519 Activate the video output. The audio stream is passed unchanged whether this
20520 option is set or no. The video stream will be the first output stream if
20521 activated. Default is @code{0}.
20524 Set the video size. This option is for video only. For the syntax of this
20526 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20527 Default and minimum resolution is @code{640x480}.
20530 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20531 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20532 other integer value between this range is allowed.
20535 Set metadata injection. If set to @code{1}, the audio input will be segmented
20536 into 100ms output frames, each of them containing various loudness information
20537 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20539 Default is @code{0}.
20542 Force the frame logging level.
20544 Available values are:
20547 information logging level
20549 verbose logging level
20552 By default, the logging level is set to @var{info}. If the @option{video} or
20553 the @option{metadata} options are set, it switches to @var{verbose}.
20558 Available modes can be cumulated (the option is a @code{flag} type). Possible
20562 Disable any peak mode (default).
20564 Enable sample-peak mode.
20566 Simple peak mode looking for the higher sample value. It logs a message
20567 for sample-peak (identified by @code{SPK}).
20569 Enable true-peak mode.
20571 If enabled, the peak lookup is done on an over-sampled version of the input
20572 stream for better peak accuracy. It logs a message for true-peak.
20573 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20574 This mode requires a build with @code{libswresample}.
20578 Treat mono input files as "dual mono". If a mono file is intended for playback
20579 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20580 If set to @code{true}, this option will compensate for this effect.
20581 Multi-channel input files are not affected by this option.
20584 Set a specific pan law to be used for the measurement of dual mono files.
20585 This parameter is optional, and has a default value of -3.01dB.
20588 Set a specific target level (in LUFS) used as relative zero in the visualization.
20589 This parameter is optional and has a default value of -23LUFS as specified
20590 by EBU R128. However, material published online may prefer a level of -16LUFS
20591 (e.g. for use with podcasts or video platforms).
20594 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20595 @code{shortterm}. By default the momentary value will be used, but in certain
20596 scenarios it may be more useful to observe the short term value instead (e.g.
20600 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20601 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20602 video output, not the summary or continuous log output.
20605 @subsection Examples
20609 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20611 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20615 Run an analysis with @command{ffmpeg}:
20617 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20621 @section interleave, ainterleave
20623 Temporally interleave frames from several inputs.
20625 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20627 These filters read frames from several inputs and send the oldest
20628 queued frame to the output.
20630 Input streams must have well defined, monotonically increasing frame
20633 In order to submit one frame to output, these filters need to enqueue
20634 at least one frame for each input, so they cannot work in case one
20635 input is not yet terminated and will not receive incoming frames.
20637 For example consider the case when one input is a @code{select} filter
20638 which always drops input frames. The @code{interleave} filter will keep
20639 reading from that input, but it will never be able to send new frames
20640 to output until the input sends an end-of-stream signal.
20642 Also, depending on inputs synchronization, the filters will drop
20643 frames in case one input receives more frames than the other ones, and
20644 the queue is already filled.
20646 These filters accept the following options:
20650 Set the number of different inputs, it is 2 by default.
20653 @subsection Examples
20657 Interleave frames belonging to different streams using @command{ffmpeg}:
20659 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20663 Add flickering blur effect:
20665 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20669 @section metadata, ametadata
20671 Manipulate frame metadata.
20673 This filter accepts the following options:
20677 Set mode of operation of the filter.
20679 Can be one of the following:
20683 If both @code{value} and @code{key} is set, select frames
20684 which have such metadata. If only @code{key} is set, select
20685 every frame that has such key in metadata.
20688 Add new metadata @code{key} and @code{value}. If key is already available
20692 Modify value of already present key.
20695 If @code{value} is set, delete only keys that have such value.
20696 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
20700 Print key and its value if metadata was found. If @code{key} is not set print all
20701 metadata values available in frame.
20705 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
20708 Set metadata value which will be used. This option is mandatory for
20709 @code{modify} and @code{add} mode.
20712 Which function to use when comparing metadata value and @code{value}.
20714 Can be one of following:
20718 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
20721 Values are interpreted as strings, returns true if metadata value starts with
20722 the @code{value} option string.
20725 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
20728 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
20731 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
20734 Values are interpreted as floats, returns true if expression from option @code{expr}
20739 Set expression which is used when @code{function} is set to @code{expr}.
20740 The expression is evaluated through the eval API and can contain the following
20745 Float representation of @code{value} from metadata key.
20748 Float representation of @code{value} as supplied by user in @code{value} option.
20752 If specified in @code{print} mode, output is written to the named file. Instead of
20753 plain filename any writable url can be specified. Filename ``-'' is a shorthand
20754 for standard output. If @code{file} option is not set, output is written to the log
20755 with AV_LOG_INFO loglevel.
20759 @subsection Examples
20763 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
20766 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
20769 Print silencedetect output to file @file{metadata.txt}.
20771 silencedetect,ametadata=mode=print:file=metadata.txt
20774 Direct all metadata to a pipe with file descriptor 4.
20776 metadata=mode=print:file='pipe\:4'
20780 @section perms, aperms
20782 Set read/write permissions for the output frames.
20784 These filters are mainly aimed at developers to test direct path in the
20785 following filter in the filtergraph.
20787 The filters accept the following options:
20791 Select the permissions mode.
20793 It accepts the following values:
20796 Do nothing. This is the default.
20798 Set all the output frames read-only.
20800 Set all the output frames directly writable.
20802 Make the frame read-only if writable, and writable if read-only.
20804 Set each output frame read-only or writable randomly.
20808 Set the seed for the @var{random} mode, must be an integer included between
20809 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
20810 @code{-1}, the filter will try to use a good random seed on a best effort
20814 Note: in case of auto-inserted filter between the permission filter and the
20815 following one, the permission might not be received as expected in that
20816 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
20817 perms/aperms filter can avoid this problem.
20819 @section realtime, arealtime
20821 Slow down filtering to match real time approximately.
20823 These filters will pause the filtering for a variable amount of time to
20824 match the output rate with the input timestamps.
20825 They are similar to the @option{re} option to @code{ffmpeg}.
20827 They accept the following options:
20831 Time limit for the pauses. Any pause longer than that will be considered
20832 a timestamp discontinuity and reset the timer. Default is 2 seconds.
20836 @section select, aselect
20838 Select frames to pass in output.
20840 This filter accepts the following options:
20845 Set expression, which is evaluated for each input frame.
20847 If the expression is evaluated to zero, the frame is discarded.
20849 If the evaluation result is negative or NaN, the frame is sent to the
20850 first output; otherwise it is sent to the output with index
20851 @code{ceil(val)-1}, assuming that the input index starts from 0.
20853 For example a value of @code{1.2} corresponds to the output with index
20854 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
20857 Set the number of outputs. The output to which to send the selected
20858 frame is based on the result of the evaluation. Default value is 1.
20861 The expression can contain the following constants:
20865 The (sequential) number of the filtered frame, starting from 0.
20868 The (sequential) number of the selected frame, starting from 0.
20870 @item prev_selected_n
20871 The sequential number of the last selected frame. It's NAN if undefined.
20874 The timebase of the input timestamps.
20877 The PTS (Presentation TimeStamp) of the filtered video frame,
20878 expressed in @var{TB} units. It's NAN if undefined.
20881 The PTS of the filtered video frame,
20882 expressed in seconds. It's NAN if undefined.
20885 The PTS of the previously filtered video frame. It's NAN if undefined.
20887 @item prev_selected_pts
20888 The PTS of the last previously filtered video frame. It's NAN if undefined.
20890 @item prev_selected_t
20891 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
20894 The PTS of the first video frame in the video. It's NAN if undefined.
20897 The time of the first video frame in the video. It's NAN if undefined.
20899 @item pict_type @emph{(video only)}
20900 The type of the filtered frame. It can assume one of the following
20912 @item interlace_type @emph{(video only)}
20913 The frame interlace type. It can assume one of the following values:
20916 The frame is progressive (not interlaced).
20918 The frame is top-field-first.
20920 The frame is bottom-field-first.
20923 @item consumed_sample_n @emph{(audio only)}
20924 the number of selected samples before the current frame
20926 @item samples_n @emph{(audio only)}
20927 the number of samples in the current frame
20929 @item sample_rate @emph{(audio only)}
20930 the input sample rate
20933 This is 1 if the filtered frame is a key-frame, 0 otherwise.
20936 the position in the file of the filtered frame, -1 if the information
20937 is not available (e.g. for synthetic video)
20939 @item scene @emph{(video only)}
20940 value between 0 and 1 to indicate a new scene; a low value reflects a low
20941 probability for the current frame to introduce a new scene, while a higher
20942 value means the current frame is more likely to be one (see the example below)
20944 @item concatdec_select
20945 The concat demuxer can select only part of a concat input file by setting an
20946 inpoint and an outpoint, but the output packets may not be entirely contained
20947 in the selected interval. By using this variable, it is possible to skip frames
20948 generated by the concat demuxer which are not exactly contained in the selected
20951 This works by comparing the frame pts against the @var{lavf.concat.start_time}
20952 and the @var{lavf.concat.duration} packet metadata values which are also
20953 present in the decoded frames.
20955 The @var{concatdec_select} variable is -1 if the frame pts is at least
20956 start_time and either the duration metadata is missing or the frame pts is less
20957 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
20960 That basically means that an input frame is selected if its pts is within the
20961 interval set by the concat demuxer.
20965 The default value of the select expression is "1".
20967 @subsection Examples
20971 Select all frames in input:
20976 The example above is the same as:
20988 Select only I-frames:
20990 select='eq(pict_type\,I)'
20994 Select one frame every 100:
20996 select='not(mod(n\,100))'
21000 Select only frames contained in the 10-20 time interval:
21002 select=between(t\,10\,20)
21006 Select only I-frames contained in the 10-20 time interval:
21008 select=between(t\,10\,20)*eq(pict_type\,I)
21012 Select frames with a minimum distance of 10 seconds:
21014 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21018 Use aselect to select only audio frames with samples number > 100:
21020 aselect='gt(samples_n\,100)'
21024 Create a mosaic of the first scenes:
21026 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21029 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21033 Send even and odd frames to separate outputs, and compose them:
21035 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21039 Select useful frames from an ffconcat file which is using inpoints and
21040 outpoints but where the source files are not intra frame only.
21042 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21046 @section sendcmd, asendcmd
21048 Send commands to filters in the filtergraph.
21050 These filters read commands to be sent to other filters in the
21053 @code{sendcmd} must be inserted between two video filters,
21054 @code{asendcmd} must be inserted between two audio filters, but apart
21055 from that they act the same way.
21057 The specification of commands can be provided in the filter arguments
21058 with the @var{commands} option, or in a file specified by the
21059 @var{filename} option.
21061 These filters accept the following options:
21064 Set the commands to be read and sent to the other filters.
21066 Set the filename of the commands to be read and sent to the other
21070 @subsection Commands syntax
21072 A commands description consists of a sequence of interval
21073 specifications, comprising a list of commands to be executed when a
21074 particular event related to that interval occurs. The occurring event
21075 is typically the current frame time entering or leaving a given time
21078 An interval is specified by the following syntax:
21080 @var{START}[-@var{END}] @var{COMMANDS};
21083 The time interval is specified by the @var{START} and @var{END} times.
21084 @var{END} is optional and defaults to the maximum time.
21086 The current frame time is considered within the specified interval if
21087 it is included in the interval [@var{START}, @var{END}), that is when
21088 the time is greater or equal to @var{START} and is lesser than
21091 @var{COMMANDS} consists of a sequence of one or more command
21092 specifications, separated by ",", relating to that interval. The
21093 syntax of a command specification is given by:
21095 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21098 @var{FLAGS} is optional and specifies the type of events relating to
21099 the time interval which enable sending the specified command, and must
21100 be a non-null sequence of identifier flags separated by "+" or "|" and
21101 enclosed between "[" and "]".
21103 The following flags are recognized:
21106 The command is sent when the current frame timestamp enters the
21107 specified interval. In other words, the command is sent when the
21108 previous frame timestamp was not in the given interval, and the
21112 The command is sent when the current frame timestamp leaves the
21113 specified interval. In other words, the command is sent when the
21114 previous frame timestamp was in the given interval, and the
21118 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21121 @var{TARGET} specifies the target of the command, usually the name of
21122 the filter class or a specific filter instance name.
21124 @var{COMMAND} specifies the name of the command for the target filter.
21126 @var{ARG} is optional and specifies the optional list of argument for
21127 the given @var{COMMAND}.
21129 Between one interval specification and another, whitespaces, or
21130 sequences of characters starting with @code{#} until the end of line,
21131 are ignored and can be used to annotate comments.
21133 A simplified BNF description of the commands specification syntax
21136 @var{COMMAND_FLAG} ::= "enter" | "leave"
21137 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21138 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21139 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21140 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21141 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21144 @subsection Examples
21148 Specify audio tempo change at second 4:
21150 asendcmd=c='4.0 atempo tempo 1.5',atempo
21154 Target a specific filter instance:
21156 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21160 Specify a list of drawtext and hue commands in a file.
21162 # show text in the interval 5-10
21163 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21164 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21166 # desaturate the image in the interval 15-20
21167 15.0-20.0 [enter] hue s 0,
21168 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21170 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21172 # apply an exponential saturation fade-out effect, starting from time 25
21173 25 [enter] hue s exp(25-t)
21176 A filtergraph allowing to read and process the above command list
21177 stored in a file @file{test.cmd}, can be specified with:
21179 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21184 @section setpts, asetpts
21186 Change the PTS (presentation timestamp) of the input frames.
21188 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21190 This filter accepts the following options:
21195 The expression which is evaluated for each frame to construct its timestamp.
21199 The expression is evaluated through the eval API and can contain the following
21203 @item FRAME_RATE, FR
21204 frame rate, only defined for constant frame-rate video
21207 The presentation timestamp in input
21210 The count of the input frame for video or the number of consumed samples,
21211 not including the current frame for audio, starting from 0.
21213 @item NB_CONSUMED_SAMPLES
21214 The number of consumed samples, not including the current frame (only
21217 @item NB_SAMPLES, S
21218 The number of samples in the current frame (only audio)
21220 @item SAMPLE_RATE, SR
21221 The audio sample rate.
21224 The PTS of the first frame.
21227 the time in seconds of the first frame
21230 State whether the current frame is interlaced.
21233 the time in seconds of the current frame
21236 original position in the file of the frame, or undefined if undefined
21237 for the current frame
21240 The previous input PTS.
21243 previous input time in seconds
21246 The previous output PTS.
21249 previous output time in seconds
21252 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21256 The wallclock (RTC) time at the start of the movie in microseconds.
21259 The timebase of the input timestamps.
21263 @subsection Examples
21267 Start counting PTS from zero
21269 setpts=PTS-STARTPTS
21273 Apply fast motion effect:
21279 Apply slow motion effect:
21285 Set fixed rate of 25 frames per second:
21291 Set fixed rate 25 fps with some jitter:
21293 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21297 Apply an offset of 10 seconds to the input PTS:
21303 Generate timestamps from a "live source" and rebase onto the current timebase:
21305 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21309 Generate timestamps by counting samples:
21318 Force color range for the output video frame.
21320 The @code{setrange} filter marks the color range property for the
21321 output frames. It does not change the input frame, but only sets the
21322 corresponding property, which affects how the frame is treated by
21325 The filter accepts the following options:
21330 Available values are:
21334 Keep the same color range property.
21336 @item unspecified, unknown
21337 Set the color range as unspecified.
21339 @item limited, tv, mpeg
21340 Set the color range as limited.
21342 @item full, pc, jpeg
21343 Set the color range as full.
21347 @section settb, asettb
21349 Set the timebase to use for the output frames timestamps.
21350 It is mainly useful for testing timebase configuration.
21352 It accepts the following parameters:
21357 The expression which is evaluated into the output timebase.
21361 The value for @option{tb} is an arithmetic expression representing a
21362 rational. The expression can contain the constants "AVTB" (the default
21363 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21364 audio only). Default value is "intb".
21366 @subsection Examples
21370 Set the timebase to 1/25:
21376 Set the timebase to 1/10:
21382 Set the timebase to 1001/1000:
21388 Set the timebase to 2*intb:
21394 Set the default timebase value:
21401 Convert input audio to a video output representing frequency spectrum
21402 logarithmically using Brown-Puckette constant Q transform algorithm with
21403 direct frequency domain coefficient calculation (but the transform itself
21404 is not really constant Q, instead the Q factor is actually variable/clamped),
21405 with musical tone scale, from E0 to D#10.
21407 The filter accepts the following options:
21411 Specify the video size for the output. It must be even. For the syntax of this option,
21412 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21413 Default value is @code{1920x1080}.
21416 Set the output frame rate. Default value is @code{25}.
21419 Set the bargraph height. It must be even. Default value is @code{-1} which
21420 computes the bargraph height automatically.
21423 Set the axis height. It must be even. Default value is @code{-1} which computes
21424 the axis height automatically.
21427 Set the sonogram height. It must be even. Default value is @code{-1} which
21428 computes the sonogram height automatically.
21431 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21432 instead. Default value is @code{1}.
21434 @item sono_v, volume
21435 Specify the sonogram volume expression. It can contain variables:
21438 the @var{bar_v} evaluated expression
21439 @item frequency, freq, f
21440 the frequency where it is evaluated
21441 @item timeclamp, tc
21442 the value of @var{timeclamp} option
21446 @item a_weighting(f)
21447 A-weighting of equal loudness
21448 @item b_weighting(f)
21449 B-weighting of equal loudness
21450 @item c_weighting(f)
21451 C-weighting of equal loudness.
21453 Default value is @code{16}.
21455 @item bar_v, volume2
21456 Specify the bargraph volume expression. It can contain variables:
21459 the @var{sono_v} evaluated expression
21460 @item frequency, freq, f
21461 the frequency where it is evaluated
21462 @item timeclamp, tc
21463 the value of @var{timeclamp} option
21467 @item a_weighting(f)
21468 A-weighting of equal loudness
21469 @item b_weighting(f)
21470 B-weighting of equal loudness
21471 @item c_weighting(f)
21472 C-weighting of equal loudness.
21474 Default value is @code{sono_v}.
21476 @item sono_g, gamma
21477 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21478 higher gamma makes the spectrum having more range. Default value is @code{3}.
21479 Acceptable range is @code{[1, 7]}.
21481 @item bar_g, gamma2
21482 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21486 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21487 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21489 @item timeclamp, tc
21490 Specify the transform timeclamp. At low frequency, there is trade-off between
21491 accuracy in time domain and frequency domain. If timeclamp is lower,
21492 event in time domain is represented more accurately (such as fast bass drum),
21493 otherwise event in frequency domain is represented more accurately
21494 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21497 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21498 limits future samples by applying asymmetric windowing in time domain, useful
21499 when low latency is required. Accepted range is @code{[0, 1]}.
21502 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21503 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21506 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21507 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21510 This option is deprecated and ignored.
21513 Specify the transform length in time domain. Use this option to control accuracy
21514 trade-off between time domain and frequency domain at every frequency sample.
21515 It can contain variables:
21517 @item frequency, freq, f
21518 the frequency where it is evaluated
21519 @item timeclamp, tc
21520 the value of @var{timeclamp} option.
21522 Default value is @code{384*tc/(384+tc*f)}.
21525 Specify the transform count for every video frame. Default value is @code{6}.
21526 Acceptable range is @code{[1, 30]}.
21529 Specify the transform count for every single pixel. Default value is @code{0},
21530 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21533 Specify font file for use with freetype to draw the axis. If not specified,
21534 use embedded font. Note that drawing with font file or embedded font is not
21535 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21539 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21540 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21543 Specify font color expression. This is arithmetic expression that should return
21544 integer value 0xRRGGBB. It can contain variables:
21546 @item frequency, freq, f
21547 the frequency where it is evaluated
21548 @item timeclamp, tc
21549 the value of @var{timeclamp} option
21554 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21555 @item r(x), g(x), b(x)
21556 red, green, and blue value of intensity x.
21558 Default value is @code{st(0, (midi(f)-59.5)/12);
21559 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21560 r(1-ld(1)) + b(ld(1))}.
21563 Specify image file to draw the axis. This option override @var{fontfile} and
21564 @var{fontcolor} option.
21567 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21568 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21569 Default value is @code{1}.
21572 Set colorspace. The accepted values are:
21575 Unspecified (default)
21584 BT.470BG or BT.601-6 625
21587 SMPTE-170M or BT.601-6 525
21593 BT.2020 with non-constant luminance
21598 Set spectrogram color scheme. This is list of floating point values with format
21599 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21600 The default is @code{1|0.5|0|0|0.5|1}.
21604 @subsection Examples
21608 Playing audio while showing the spectrum:
21610 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21614 Same as above, but with frame rate 30 fps:
21616 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21620 Playing at 1280x720:
21622 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21626 Disable sonogram display:
21632 A1 and its harmonics: A1, A2, (near)E3, A3:
21634 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),
21635 asplit[a][out1]; [a] showcqt [out0]'
21639 Same as above, but with more accuracy in frequency domain:
21641 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),
21642 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21648 bar_v=10:sono_v=bar_v*a_weighting(f)
21652 Custom gamma, now spectrum is linear to the amplitude.
21658 Custom tlength equation:
21660 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)))'
21664 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21666 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21670 Custom font using fontconfig:
21672 font='Courier New,Monospace,mono|bold'
21676 Custom frequency range with custom axis using image file:
21678 axisfile=myaxis.png:basefreq=40:endfreq=10000
21684 Convert input audio to video output representing the audio power spectrum.
21685 Audio amplitude is on Y-axis while frequency is on X-axis.
21687 The filter accepts the following options:
21691 Specify size of video. For the syntax of this option, check the
21692 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21693 Default is @code{1024x512}.
21697 This set how each frequency bin will be represented.
21699 It accepts the following values:
21705 Default is @code{bar}.
21708 Set amplitude scale.
21710 It accepts the following values:
21724 Default is @code{log}.
21727 Set frequency scale.
21729 It accepts the following values:
21738 Reverse logarithmic scale.
21740 Default is @code{lin}.
21745 It accepts the following values:
21761 Default is @code{w2048}
21764 Set windowing function.
21766 It accepts the following values:
21789 Default is @code{hanning}.
21792 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
21793 which means optimal overlap for selected window function will be picked.
21796 Set time averaging. Setting this to 0 will display current maximal peaks.
21797 Default is @code{1}, which means time averaging is disabled.
21800 Specify list of colors separated by space or by '|' which will be used to
21801 draw channel frequencies. Unrecognized or missing colors will be replaced
21805 Set channel display mode.
21807 It accepts the following values:
21812 Default is @code{combined}.
21815 Set minimum amplitude used in @code{log} amplitude scaler.
21819 @anchor{showspectrum}
21820 @section showspectrum
21822 Convert input audio to a video output, representing the audio frequency
21825 The filter accepts the following options:
21829 Specify the video size for the output. For the syntax of this option, check the
21830 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21831 Default value is @code{640x512}.
21834 Specify how the spectrum should slide along the window.
21836 It accepts the following values:
21839 the samples start again on the left when they reach the right
21841 the samples scroll from right to left
21843 frames are only produced when the samples reach the right
21845 the samples scroll from left to right
21848 Default value is @code{replace}.
21851 Specify display mode.
21853 It accepts the following values:
21856 all channels are displayed in the same row
21858 all channels are displayed in separate rows
21861 Default value is @samp{combined}.
21864 Specify display color mode.
21866 It accepts the following values:
21869 each channel is displayed in a separate color
21871 each channel is displayed using the same color scheme
21873 each channel is displayed using the rainbow color scheme
21875 each channel is displayed using the moreland color scheme
21877 each channel is displayed using the nebulae color scheme
21879 each channel is displayed using the fire color scheme
21881 each channel is displayed using the fiery color scheme
21883 each channel is displayed using the fruit color scheme
21885 each channel is displayed using the cool color scheme
21887 each channel is displayed using the magma color scheme
21889 each channel is displayed using the green color scheme
21891 each channel is displayed using the viridis color scheme
21893 each channel is displayed using the plasma color scheme
21895 each channel is displayed using the cividis color scheme
21897 each channel is displayed using the terrain color scheme
21900 Default value is @samp{channel}.
21903 Specify scale used for calculating intensity color values.
21905 It accepts the following values:
21910 square root, default
21921 Default value is @samp{sqrt}.
21924 Set saturation modifier for displayed colors. Negative values provide
21925 alternative color scheme. @code{0} is no saturation at all.
21926 Saturation must be in [-10.0, 10.0] range.
21927 Default value is @code{1}.
21930 Set window function.
21932 It accepts the following values:
21957 Default value is @code{hann}.
21960 Set orientation of time vs frequency axis. Can be @code{vertical} or
21961 @code{horizontal}. Default is @code{vertical}.
21964 Set ratio of overlap window. Default value is @code{0}.
21965 When value is @code{1} overlap is set to recommended size for specific
21966 window function currently used.
21969 Set scale gain for calculating intensity color values.
21970 Default value is @code{1}.
21973 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
21976 Set color rotation, must be in [-1.0, 1.0] range.
21977 Default value is @code{0}.
21980 Set start frequency from which to display spectrogram. Default is @code{0}.
21983 Set stop frequency to which to display spectrogram. Default is @code{0}.
21986 Set upper frame rate limit. Default is @code{auto}, unlimited.
21989 Draw time and frequency axes and legends. Default is disabled.
21992 The usage is very similar to the showwaves filter; see the examples in that
21995 @subsection Examples
21999 Large window with logarithmic color scaling:
22001 showspectrum=s=1280x480:scale=log
22005 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22007 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22008 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22012 @section showspectrumpic
22014 Convert input audio to a single video frame, representing the audio frequency
22017 The filter accepts the following options:
22021 Specify the video size for the output. For the syntax of this option, check the
22022 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22023 Default value is @code{4096x2048}.
22026 Specify display mode.
22028 It accepts the following values:
22031 all channels are displayed in the same row
22033 all channels are displayed in separate rows
22035 Default value is @samp{combined}.
22038 Specify display color mode.
22040 It accepts the following values:
22043 each channel is displayed in a separate color
22045 each channel is displayed using the same color scheme
22047 each channel is displayed using the rainbow color scheme
22049 each channel is displayed using the moreland color scheme
22051 each channel is displayed using the nebulae color scheme
22053 each channel is displayed using the fire color scheme
22055 each channel is displayed using the fiery color scheme
22057 each channel is displayed using the fruit color scheme
22059 each channel is displayed using the cool color scheme
22061 each channel is displayed using the magma color scheme
22063 each channel is displayed using the green color scheme
22065 each channel is displayed using the viridis color scheme
22067 each channel is displayed using the plasma color scheme
22069 each channel is displayed using the cividis color scheme
22071 each channel is displayed using the terrain color scheme
22073 Default value is @samp{intensity}.
22076 Specify scale used for calculating intensity color values.
22078 It accepts the following values:
22083 square root, default
22093 Default value is @samp{log}.
22096 Set saturation modifier for displayed colors. Negative values provide
22097 alternative color scheme. @code{0} is no saturation at all.
22098 Saturation must be in [-10.0, 10.0] range.
22099 Default value is @code{1}.
22102 Set window function.
22104 It accepts the following values:
22128 Default value is @code{hann}.
22131 Set orientation of time vs frequency axis. Can be @code{vertical} or
22132 @code{horizontal}. Default is @code{vertical}.
22135 Set scale gain for calculating intensity color values.
22136 Default value is @code{1}.
22139 Draw time and frequency axes and legends. Default is enabled.
22142 Set color rotation, must be in [-1.0, 1.0] range.
22143 Default value is @code{0}.
22146 Set start frequency from which to display spectrogram. Default is @code{0}.
22149 Set stop frequency to which to display spectrogram. Default is @code{0}.
22152 @subsection Examples
22156 Extract an audio spectrogram of a whole audio track
22157 in a 1024x1024 picture using @command{ffmpeg}:
22159 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22163 @section showvolume
22165 Convert input audio volume to a video output.
22167 The filter accepts the following options:
22174 Set border width, allowed range is [0, 5]. Default is 1.
22177 Set channel width, allowed range is [80, 8192]. Default is 400.
22180 Set channel height, allowed range is [1, 900]. Default is 20.
22183 Set fade, allowed range is [0, 1]. Default is 0.95.
22186 Set volume color expression.
22188 The expression can use the following variables:
22192 Current max volume of channel in dB.
22198 Current channel number, starting from 0.
22202 If set, displays channel names. Default is enabled.
22205 If set, displays volume values. Default is enabled.
22208 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22209 default is @code{h}.
22212 Set step size, allowed range is [0, 5]. Default is 0, which means
22216 Set background opacity, allowed range is [0, 1]. Default is 0.
22219 Set metering mode, can be peak: @code{p} or rms: @code{r},
22220 default is @code{p}.
22223 Set display scale, can be linear: @code{lin} or log: @code{log},
22224 default is @code{lin}.
22228 If set to > 0., display a line for the max level
22229 in the previous seconds.
22230 default is disabled: @code{0.}
22233 The color of the max line. Use when @code{dm} option is set to > 0.
22234 default is: @code{orange}
22239 Convert input audio to a video output, representing the samples waves.
22241 The filter accepts the following options:
22245 Specify the video size for the output. For the syntax of this option, check the
22246 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22247 Default value is @code{600x240}.
22252 Available values are:
22255 Draw a point for each sample.
22258 Draw a vertical line for each sample.
22261 Draw a point for each sample and a line between them.
22264 Draw a centered vertical line for each sample.
22267 Default value is @code{point}.
22270 Set the number of samples which are printed on the same column. A
22271 larger value will decrease the frame rate. Must be a positive
22272 integer. This option can be set only if the value for @var{rate}
22273 is not explicitly specified.
22276 Set the (approximate) output frame rate. This is done by setting the
22277 option @var{n}. Default value is "25".
22279 @item split_channels
22280 Set if channels should be drawn separately or overlap. Default value is 0.
22283 Set colors separated by '|' which are going to be used for drawing of each channel.
22286 Set amplitude scale.
22288 Available values are:
22306 Set the draw mode. This is mostly useful to set for high @var{n}.
22308 Available values are:
22311 Scale pixel values for each drawn sample.
22314 Draw every sample directly.
22317 Default value is @code{scale}.
22320 @subsection Examples
22324 Output the input file audio and the corresponding video representation
22327 amovie=a.mp3,asplit[out0],showwaves[out1]
22331 Create a synthetic signal and show it with showwaves, forcing a
22332 frame rate of 30 frames per second:
22334 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22338 @section showwavespic
22340 Convert input audio to a single video frame, representing the samples waves.
22342 The filter accepts the following options:
22346 Specify the video size for the output. For the syntax of this option, check the
22347 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22348 Default value is @code{600x240}.
22350 @item split_channels
22351 Set if channels should be drawn separately or overlap. Default value is 0.
22354 Set colors separated by '|' which are going to be used for drawing of each channel.
22357 Set amplitude scale.
22359 Available values are:
22377 @subsection Examples
22381 Extract a channel split representation of the wave form of a whole audio track
22382 in a 1024x800 picture using @command{ffmpeg}:
22384 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22388 @section sidedata, asidedata
22390 Delete frame side data, or select frames based on it.
22392 This filter accepts the following options:
22396 Set mode of operation of the filter.
22398 Can be one of the following:
22402 Select every frame with side data of @code{type}.
22405 Delete side data of @code{type}. If @code{type} is not set, delete all side
22411 Set side data type used with all modes. Must be set for @code{select} mode. For
22412 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22413 in @file{libavutil/frame.h}. For example, to choose
22414 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22418 @section spectrumsynth
22420 Sythesize audio from 2 input video spectrums, first input stream represents
22421 magnitude across time and second represents phase across time.
22422 The filter will transform from frequency domain as displayed in videos back
22423 to time domain as presented in audio output.
22425 This filter is primarily created for reversing processed @ref{showspectrum}
22426 filter outputs, but can synthesize sound from other spectrograms too.
22427 But in such case results are going to be poor if the phase data is not
22428 available, because in such cases phase data need to be recreated, usually
22429 its just recreated from random noise.
22430 For best results use gray only output (@code{channel} color mode in
22431 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22432 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22433 @code{data} option. Inputs videos should generally use @code{fullframe}
22434 slide mode as that saves resources needed for decoding video.
22436 The filter accepts the following options:
22440 Specify sample rate of output audio, the sample rate of audio from which
22441 spectrum was generated may differ.
22444 Set number of channels represented in input video spectrums.
22447 Set scale which was used when generating magnitude input spectrum.
22448 Can be @code{lin} or @code{log}. Default is @code{log}.
22451 Set slide which was used when generating inputs spectrums.
22452 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22453 Default is @code{fullframe}.
22456 Set window function used for resynthesis.
22459 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22460 which means optimal overlap for selected window function will be picked.
22463 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22464 Default is @code{vertical}.
22467 @subsection Examples
22471 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22472 then resynthesize videos back to audio with spectrumsynth:
22474 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
22475 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
22476 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22480 @section split, asplit
22482 Split input into several identical outputs.
22484 @code{asplit} works with audio input, @code{split} with video.
22486 The filter accepts a single parameter which specifies the number of outputs. If
22487 unspecified, it defaults to 2.
22489 @subsection Examples
22493 Create two separate outputs from the same input:
22495 [in] split [out0][out1]
22499 To create 3 or more outputs, you need to specify the number of
22502 [in] asplit=3 [out0][out1][out2]
22506 Create two separate outputs from the same input, one cropped and
22509 [in] split [splitout1][splitout2];
22510 [splitout1] crop=100:100:0:0 [cropout];
22511 [splitout2] pad=200:200:100:100 [padout];
22515 Create 5 copies of the input audio with @command{ffmpeg}:
22517 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22523 Receive commands sent through a libzmq client, and forward them to
22524 filters in the filtergraph.
22526 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22527 must be inserted between two video filters, @code{azmq} between two
22528 audio filters. Both are capable to send messages to any filter type.
22530 To enable these filters you need to install the libzmq library and
22531 headers and configure FFmpeg with @code{--enable-libzmq}.
22533 For more information about libzmq see:
22534 @url{http://www.zeromq.org/}
22536 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22537 receives messages sent through a network interface defined by the
22538 @option{bind_address} (or the abbreviation "@option{b}") option.
22539 Default value of this option is @file{tcp://localhost:5555}. You may
22540 want to alter this value to your needs, but do not forget to escape any
22541 ':' signs (see @ref{filtergraph escaping}).
22543 The received message must be in the form:
22545 @var{TARGET} @var{COMMAND} [@var{ARG}]
22548 @var{TARGET} specifies the target of the command, usually the name of
22549 the filter class or a specific filter instance name. The default
22550 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22551 but you can override this by using the @samp{filter_name@@id} syntax
22552 (see @ref{Filtergraph syntax}).
22554 @var{COMMAND} specifies the name of the command for the target filter.
22556 @var{ARG} is optional and specifies the optional argument list for the
22557 given @var{COMMAND}.
22559 Upon reception, the message is processed and the corresponding command
22560 is injected into the filtergraph. Depending on the result, the filter
22561 will send a reply to the client, adopting the format:
22563 @var{ERROR_CODE} @var{ERROR_REASON}
22567 @var{MESSAGE} is optional.
22569 @subsection Examples
22571 Look at @file{tools/zmqsend} for an example of a zmq client which can
22572 be used to send commands processed by these filters.
22574 Consider the following filtergraph generated by @command{ffplay}.
22575 In this example the last overlay filter has an instance name. All other
22576 filters will have default instance names.
22579 ffplay -dumpgraph 1 -f lavfi "
22580 color=s=100x100:c=red [l];
22581 color=s=100x100:c=blue [r];
22582 nullsrc=s=200x100, zmq [bg];
22583 [bg][l] overlay [bg+l];
22584 [bg+l][r] overlay@@my=x=100 "
22587 To change the color of the left side of the video, the following
22588 command can be used:
22590 echo Parsed_color_0 c yellow | tools/zmqsend
22593 To change the right side:
22595 echo Parsed_color_1 c pink | tools/zmqsend
22598 To change the position of the right side:
22600 echo overlay@@my x 150 | tools/zmqsend
22604 @c man end MULTIMEDIA FILTERS
22606 @chapter Multimedia Sources
22607 @c man begin MULTIMEDIA SOURCES
22609 Below is a description of the currently available multimedia sources.
22613 This is the same as @ref{movie} source, except it selects an audio
22619 Read audio and/or video stream(s) from a movie container.
22621 It accepts the following parameters:
22625 The name of the resource to read (not necessarily a file; it can also be a
22626 device or a stream accessed through some protocol).
22628 @item format_name, f
22629 Specifies the format assumed for the movie to read, and can be either
22630 the name of a container or an input device. If not specified, the
22631 format is guessed from @var{movie_name} or by probing.
22633 @item seek_point, sp
22634 Specifies the seek point in seconds. The frames will be output
22635 starting from this seek point. The parameter is evaluated with
22636 @code{av_strtod}, so the numerical value may be suffixed by an IS
22637 postfix. The default value is "0".
22640 Specifies the streams to read. Several streams can be specified,
22641 separated by "+". The source will then have as many outputs, in the
22642 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22643 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22644 respectively the default (best suited) video and audio stream. Default
22645 is "dv", or "da" if the filter is called as "amovie".
22647 @item stream_index, si
22648 Specifies the index of the video stream to read. If the value is -1,
22649 the most suitable video stream will be automatically selected. The default
22650 value is "-1". Deprecated. If the filter is called "amovie", it will select
22651 audio instead of video.
22654 Specifies how many times to read the stream in sequence.
22655 If the value is 0, the stream will be looped infinitely.
22656 Default value is "1".
22658 Note that when the movie is looped the source timestamps are not
22659 changed, so it will generate non monotonically increasing timestamps.
22661 @item discontinuity
22662 Specifies the time difference between frames above which the point is
22663 considered a timestamp discontinuity which is removed by adjusting the later
22667 It allows overlaying a second video on top of the main input of
22668 a filtergraph, as shown in this graph:
22670 input -----------> deltapts0 --> overlay --> output
22673 movie --> scale--> deltapts1 -------+
22675 @subsection Examples
22679 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
22680 on top of the input labelled "in":
22682 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
22683 [in] setpts=PTS-STARTPTS [main];
22684 [main][over] overlay=16:16 [out]
22688 Read from a video4linux2 device, and overlay it on top of the input
22691 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
22692 [in] setpts=PTS-STARTPTS [main];
22693 [main][over] overlay=16:16 [out]
22697 Read the first video stream and the audio stream with id 0x81 from
22698 dvd.vob; the video is connected to the pad named "video" and the audio is
22699 connected to the pad named "audio":
22701 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
22705 @subsection Commands
22707 Both movie and amovie support the following commands:
22710 Perform seek using "av_seek_frame".
22711 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
22714 @var{stream_index}: If stream_index is -1, a default
22715 stream is selected, and @var{timestamp} is automatically converted
22716 from AV_TIME_BASE units to the stream specific time_base.
22718 @var{timestamp}: Timestamp in AVStream.time_base units
22719 or, if no stream is specified, in AV_TIME_BASE units.
22721 @var{flags}: Flags which select direction and seeking mode.
22725 Get movie duration in AV_TIME_BASE units.
22729 @c man end MULTIMEDIA SOURCES