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 compression/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 detected as noise are spaced less than this value then any
609 sample between 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
970 Fade in first 15 seconds of audio:
976 Fade out last 25 seconds of a 900 seconds audio:
978 afade=t=out:st=875:d=25
983 Denoise audio samples with FFT.
985 A description of the accepted parameters follows.
989 Set the noise reduction in dB, allowed range is 0.01 to 97.
990 Default value is 12 dB.
993 Set the noise floor in dB, allowed range is -80 to -20.
994 Default value is -50 dB.
999 It accepts the following values:
1008 Select shellac noise.
1011 Select custom noise, defined in @code{bn} option.
1013 Default value is white noise.
1017 Set custom band noise for every one of 15 bands.
1018 Bands are separated by ' ' or '|'.
1021 Set the residual floor in dB, allowed range is -80 to -20.
1022 Default value is -38 dB.
1025 Enable noise tracking. By default is disabled.
1026 With this enabled, noise floor is automatically adjusted.
1029 Enable residual tracking. By default is disabled.
1032 Set the output mode.
1034 It accepts the following values:
1037 Pass input unchanged.
1040 Pass noise filtered out.
1045 Default value is @var{o}.
1049 @subsection Commands
1051 This filter supports the following commands:
1053 @item sample_noise, sn
1054 Start or stop measuring noise profile.
1055 Syntax for the command is : "start" or "stop" string.
1056 After measuring noise profile is stopped it will be
1057 automatically applied in filtering.
1059 @item noise_reduction, nr
1060 Change noise reduction. Argument is single float number.
1061 Syntax for the command is : "@var{noise_reduction}"
1063 @item noise_floor, nf
1064 Change noise floor. Argument is single float number.
1065 Syntax for the command is : "@var{noise_floor}"
1067 @item output_mode, om
1068 Change output mode operation.
1069 Syntax for the command is : "i", "o" or "n" string.
1073 Apply arbitrary expressions to samples in frequency domain.
1077 Set frequency domain real expression for each separate channel separated
1078 by '|'. Default is "re".
1079 If the number of input channels is greater than the number of
1080 expressions, the last specified expression is used for the remaining
1084 Set frequency domain imaginary expression for each separate channel
1085 separated by '|'. Default is "im".
1087 Each expression in @var{real} and @var{imag} can contain the following
1088 constants and functions:
1095 current frequency bin number
1098 number of available bins
1101 channel number of the current expression
1110 current real part of frequency bin of current channel
1113 current imaginary part of frequency bin of current channel
1116 Return the value of real part of frequency bin at location (@var{bin},@var{channel})
1119 Return the value of imaginary part of frequency bin at location (@var{bin},@var{channel})
1125 It accepts the following values:
1141 Default is @code{w4096}
1144 Set window function. Default is @code{hann}.
1147 Set window overlap. If set to 1, the recommended overlap for selected
1148 window function will be picked. Default is @code{0.75}.
1151 @subsection Examples
1155 Leave almost only low frequencies in audio:
1157 afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
1164 Apply an arbitrary Frequency Impulse Response filter.
1166 This filter is designed for applying long FIR filters,
1167 up to 60 seconds long.
1169 It can be used as component for digital crossover filters,
1170 room equalization, cross talk cancellation, wavefield synthesis,
1171 auralization, ambiophonics, ambisonics and spatialization.
1173 This filter uses second stream as FIR coefficients.
1174 If second stream holds single channel, it will be used
1175 for all input channels in first stream, otherwise
1176 number of channels in second stream must be same as
1177 number of channels in first stream.
1179 It accepts the following parameters:
1183 Set dry gain. This sets input gain.
1186 Set wet gain. This sets final output gain.
1189 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
1192 Enable applying gain measured from power of IR.
1194 Set which approach to use for auto gain measurement.
1198 Do not apply any gain.
1201 select peak gain, very conservative approach. This is default value.
1204 select DC gain, limited application.
1207 select gain to noise approach, this is most popular one.
1211 Set gain to be applied to IR coefficients before filtering.
1212 Allowed range is 0 to 1. This gain is applied after any gain applied with @var{gtype} option.
1215 Set format of IR stream. Can be @code{mono} or @code{input}.
1216 Default is @code{input}.
1219 Set max allowed Impulse Response filter duration in seconds. Default is 30 seconds.
1220 Allowed range is 0.1 to 60 seconds.
1223 Show IR frequency response, magnitude(magenta), phase(green) and group delay(yellow) in additional video stream.
1224 By default it is disabled.
1227 Set for which IR channel to display frequency response. By default is first channel
1228 displayed. This option is used only when @var{response} is enabled.
1231 Set video stream size. This option is used only when @var{response} is enabled.
1234 Set video stream frame rate. This option is used only when @var{response} is enabled.
1237 Set minimal partition size used for convolution. Default is @var{8192}.
1238 Allowed range is from @var{8} to @var{32768}.
1239 Lower values decreases latency at cost of higher CPU usage.
1242 Set maximal partition size used for convolution. Default is @var{8192}.
1243 Allowed range is from @var{8} to @var{32768}.
1244 Lower values may increase CPU usage.
1247 @subsection Examples
1251 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
1253 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
1260 Set output format constraints for the input audio. The framework will
1261 negotiate the most appropriate format to minimize conversions.
1263 It accepts the following parameters:
1267 A '|'-separated list of requested sample formats.
1270 A '|'-separated list of requested sample rates.
1272 @item channel_layouts
1273 A '|'-separated list of requested channel layouts.
1275 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1276 for the required syntax.
1279 If a parameter is omitted, all values are allowed.
1281 Force the output to either unsigned 8-bit or signed 16-bit stereo
1283 aformat=sample_fmts=u8|s16:channel_layouts=stereo
1288 A gate is mainly used to reduce lower parts of a signal. This kind of signal
1289 processing reduces disturbing noise between useful signals.
1291 Gating is done by detecting the volume below a chosen level @var{threshold}
1292 and dividing it by the factor set with @var{ratio}. The bottom of the noise
1293 floor is set via @var{range}. Because an exact manipulation of the signal
1294 would cause distortion of the waveform the reduction can be levelled over
1295 time. This is done by setting @var{attack} and @var{release}.
1297 @var{attack} determines how long the signal has to fall below the threshold
1298 before any reduction will occur and @var{release} sets the time the signal
1299 has to rise above the threshold to reduce the reduction again.
1300 Shorter signals than the chosen attack time will be left untouched.
1304 Set input level before filtering.
1305 Default is 1. Allowed range is from 0.015625 to 64.
1308 Set the level of gain reduction when the signal is below the threshold.
1309 Default is 0.06125. Allowed range is from 0 to 1.
1312 If a signal rises above this level the gain reduction is released.
1313 Default is 0.125. Allowed range is from 0 to 1.
1316 Set a ratio by which the signal is reduced.
1317 Default is 2. Allowed range is from 1 to 9000.
1320 Amount of milliseconds the signal has to rise above the threshold before gain
1322 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
1325 Amount of milliseconds the signal has to fall below the threshold before the
1326 reduction is increased again. Default is 250 milliseconds.
1327 Allowed range is from 0.01 to 9000.
1330 Set amount of amplification of signal after processing.
1331 Default is 1. Allowed range is from 1 to 64.
1334 Curve the sharp knee around the threshold to enter gain reduction more softly.
1335 Default is 2.828427125. Allowed range is from 1 to 8.
1338 Choose if exact signal should be taken for detection or an RMS like one.
1339 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1342 Choose if the average level between all channels or the louder channel affects
1344 Default is @code{average}. Can be @code{average} or @code{maximum}.
1349 Apply an arbitrary Infinite Impulse Response filter.
1351 It accepts the following parameters:
1355 Set numerator/zeros coefficients.
1358 Set denominator/poles coefficients.
1370 Set coefficients format.
1376 Z-plane zeros/poles, cartesian (default)
1378 Z-plane zeros/poles, polar radians
1380 Z-plane zeros/poles, polar degrees
1384 Set kind of processing.
1385 Can be @code{d} - direct or @code{s} - serial cascading. Default is @code{s}.
1388 Set filtering precision.
1392 double-precision floating-point (default)
1394 single-precision floating-point
1402 Show IR frequency response, magnitude and phase in additional video stream.
1403 By default it is disabled.
1406 Set for which IR channel to display frequency response. By default is first channel
1407 displayed. This option is used only when @var{response} is enabled.
1410 Set video stream size. This option is used only when @var{response} is enabled.
1413 Coefficients in @code{tf} format are separated by spaces and are in ascending
1416 Coefficients in @code{zp} format are separated by spaces and order of coefficients
1417 doesn't matter. Coefficients in @code{zp} format are complex numbers with @var{i}
1420 Different coefficients and gains can be provided for every channel, in such case
1421 use '|' to separate coefficients or gains. Last provided coefficients will be
1422 used for all remaining channels.
1424 @subsection Examples
1428 Apply 2 pole elliptic notch at around 5000Hz for 48000 Hz sample rate:
1430 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
1434 Same as above but in @code{zp} format:
1436 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
1442 The limiter prevents an input signal from rising over a desired threshold.
1443 This limiter uses lookahead technology to prevent your signal from distorting.
1444 It means that there is a small delay after the signal is processed. Keep in mind
1445 that the delay it produces is the attack time you set.
1447 The filter accepts the following options:
1451 Set input gain. Default is 1.
1454 Set output gain. Default is 1.
1457 Don't let signals above this level pass the limiter. Default is 1.
1460 The limiter will reach its attenuation level in this amount of time in
1461 milliseconds. Default is 5 milliseconds.
1464 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1465 Default is 50 milliseconds.
1468 When gain reduction is always needed ASC takes care of releasing to an
1469 average reduction level rather than reaching a reduction of 0 in the release
1473 Select how much the release time is affected by ASC, 0 means nearly no changes
1474 in release time while 1 produces higher release times.
1477 Auto level output signal. Default is enabled.
1478 This normalizes audio back to 0dB if enabled.
1481 Depending on picked setting it is recommended to upsample input 2x or 4x times
1482 with @ref{aresample} before applying this filter.
1486 Apply a two-pole all-pass filter with central frequency (in Hz)
1487 @var{frequency}, and filter-width @var{width}.
1488 An all-pass filter changes the audio's frequency to phase relationship
1489 without changing its frequency to amplitude relationship.
1491 The filter accepts the following options:
1495 Set frequency in Hz.
1498 Set method to specify band-width of filter.
1513 Specify the band-width of a filter in width_type units.
1516 Specify which channels to filter, by default all available are filtered.
1519 @subsection Commands
1521 This filter supports the following commands:
1524 Change allpass frequency.
1525 Syntax for the command is : "@var{frequency}"
1528 Change allpass width_type.
1529 Syntax for the command is : "@var{width_type}"
1532 Change allpass width.
1533 Syntax for the command is : "@var{width}"
1540 The filter accepts the following options:
1544 Set the number of loops. Setting this value to -1 will result in infinite loops.
1548 Set maximal number of samples. Default is 0.
1551 Set first sample of loop. Default is 0.
1557 Merge two or more audio streams into a single multi-channel stream.
1559 The filter accepts the following options:
1564 Set the number of inputs. Default is 2.
1568 If the channel layouts of the inputs are disjoint, and therefore compatible,
1569 the channel layout of the output will be set accordingly and the channels
1570 will be reordered as necessary. If the channel layouts of the inputs are not
1571 disjoint, the output will have all the channels of the first input then all
1572 the channels of the second input, in that order, and the channel layout of
1573 the output will be the default value corresponding to the total number of
1576 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1577 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1578 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1579 first input, b1 is the first channel of the second input).
1581 On the other hand, if both input are in stereo, the output channels will be
1582 in the default order: a1, a2, b1, b2, and the channel layout will be
1583 arbitrarily set to 4.0, which may or may not be the expected value.
1585 All inputs must have the same sample rate, and format.
1587 If inputs do not have the same duration, the output will stop with the
1590 @subsection Examples
1594 Merge two mono files into a stereo stream:
1596 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1600 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1602 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
1608 Mixes multiple audio inputs into a single output.
1610 Note that this filter only supports float samples (the @var{amerge}
1611 and @var{pan} audio filters support many formats). If the @var{amix}
1612 input has integer samples then @ref{aresample} will be automatically
1613 inserted to perform the conversion to float samples.
1617 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1619 will mix 3 input audio streams to a single output with the same duration as the
1620 first input and a dropout transition time of 3 seconds.
1622 It accepts the following parameters:
1626 The number of inputs. If unspecified, it defaults to 2.
1629 How to determine the end-of-stream.
1633 The duration of the longest input. (default)
1636 The duration of the shortest input.
1639 The duration of the first input.
1643 @item dropout_transition
1644 The transition time, in seconds, for volume renormalization when an input
1645 stream ends. The default value is 2 seconds.
1648 Specify weight of each input audio stream as sequence.
1649 Each weight is separated by space. By default all inputs have same weight.
1654 Multiply first audio stream with second audio stream and store result
1655 in output audio stream. Multiplication is done by multiplying each
1656 sample from first stream with sample at same position from second stream.
1658 With this element-wise multiplication one can create amplitude fades and
1659 amplitude modulations.
1661 @section anequalizer
1663 High-order parametric multiband equalizer for each channel.
1665 It accepts the following parameters:
1669 This option string is in format:
1670 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1671 Each equalizer band is separated by '|'.
1675 Set channel number to which equalization will be applied.
1676 If input doesn't have that channel the entry is ignored.
1679 Set central frequency for band.
1680 If input doesn't have that frequency the entry is ignored.
1683 Set band width in hertz.
1686 Set band gain in dB.
1689 Set filter type for band, optional, can be:
1693 Butterworth, this is default.
1704 With this option activated frequency response of anequalizer is displayed
1708 Set video stream size. Only useful if curves option is activated.
1711 Set max gain that will be displayed. Only useful if curves option is activated.
1712 Setting this to a reasonable value makes it possible to display gain which is derived from
1713 neighbour bands which are too close to each other and thus produce higher gain
1714 when both are activated.
1717 Set frequency scale used to draw frequency response in video output.
1718 Can be linear or logarithmic. Default is logarithmic.
1721 Set color for each channel curve which is going to be displayed in video stream.
1722 This is list of color names separated by space or by '|'.
1723 Unrecognised or missing colors will be replaced by white color.
1726 @subsection Examples
1730 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1731 for first 2 channels using Chebyshev type 1 filter:
1733 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1737 @subsection Commands
1739 This filter supports the following commands:
1742 Alter existing filter parameters.
1743 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1745 @var{fN} is existing filter number, starting from 0, if no such filter is available
1747 @var{freq} set new frequency parameter.
1748 @var{width} set new width parameter in herz.
1749 @var{gain} set new gain parameter in dB.
1751 Full filter invocation with asendcmd may look like this:
1752 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1757 Reduce broadband noise in audio samples using Non-Local Means algorithm.
1759 Each sample is adjusted by looking for other samples with similar contexts. This
1760 context similarity is defined by comparing their surrounding patches of size
1761 @option{p}. Patches are searched in an area of @option{r} around the sample.
1763 The filter accepts the following options.
1767 Set denoising strength. Allowed range is from 0.00001 to 10. Default value is 0.00001.
1770 Set patch radius duration. Allowed range is from 1 to 100 milliseconds.
1771 Default value is 2 milliseconds.
1774 Set research radius duration. Allowed range is from 2 to 300 milliseconds.
1775 Default value is 6 milliseconds.
1778 Set the output mode.
1780 It accepts the following values:
1783 Pass input unchanged.
1786 Pass noise filtered out.
1791 Default value is @var{o}.
1797 Pass the audio source unchanged to the output.
1801 Pad the end of an audio stream with silence.
1803 This can be used together with @command{ffmpeg} @option{-shortest} to
1804 extend audio streams to the same length as the video stream.
1806 A description of the accepted options follows.
1810 Set silence packet size. Default value is 4096.
1813 Set the number of samples of silence to add to the end. After the
1814 value is reached, the stream is terminated. This option is mutually
1815 exclusive with @option{whole_len}.
1818 Set the minimum total number of samples in the output audio stream. If
1819 the value is longer than the input audio length, silence is added to
1820 the end, until the value is reached. This option is mutually exclusive
1821 with @option{pad_len}.
1824 Specify the duration of samples of silence to add. See
1825 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1826 for the accepted syntax. Used only if set to non-zero value.
1829 Specify the minimum total duration in the output audio stream. See
1830 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
1831 for the accepted syntax. Used only if set to non-zero value. If the value is longer than
1832 the input audio length, silence is added to the end, until the value is reached.
1833 This option is mutually exclusive with @option{pad_dur}
1836 If neither the @option{pad_len} nor the @option{whole_len} nor @option{pad_dur}
1837 nor @option{whole_dur} option is set, the filter will add silence to the end of
1838 the input stream indefinitely.
1840 @subsection Examples
1844 Add 1024 samples of silence to the end of the input:
1850 Make sure the audio output will contain at least 10000 samples, pad
1851 the input with silence if required:
1853 apad=whole_len=10000
1857 Use @command{ffmpeg} to pad the audio input with silence, so that the
1858 video stream will always result the shortest and will be converted
1859 until the end in the output file when using the @option{shortest}
1862 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1867 Add a phasing effect to the input audio.
1869 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1870 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1872 A description of the accepted parameters follows.
1876 Set input gain. Default is 0.4.
1879 Set output gain. Default is 0.74
1882 Set delay in milliseconds. Default is 3.0.
1885 Set decay. Default is 0.4.
1888 Set modulation speed in Hz. Default is 0.5.
1891 Set modulation type. Default is triangular.
1893 It accepts the following values:
1902 Audio pulsator is something between an autopanner and a tremolo.
1903 But it can produce funny stereo effects as well. Pulsator changes the volume
1904 of the left and right channel based on a LFO (low frequency oscillator) with
1905 different waveforms and shifted phases.
1906 This filter have the ability to define an offset between left and right
1907 channel. An offset of 0 means that both LFO shapes match each other.
1908 The left and right channel are altered equally - a conventional tremolo.
1909 An offset of 50% means that the shape of the right channel is exactly shifted
1910 in phase (or moved backwards about half of the frequency) - pulsator acts as
1911 an autopanner. At 1 both curves match again. Every setting in between moves the
1912 phase shift gapless between all stages and produces some "bypassing" sounds with
1913 sine and triangle waveforms. The more you set the offset near 1 (starting from
1914 the 0.5) the faster the signal passes from the left to the right speaker.
1916 The filter accepts the following options:
1920 Set input gain. By default it is 1. Range is [0.015625 - 64].
1923 Set output gain. By default it is 1. Range is [0.015625 - 64].
1926 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1927 sawup or sawdown. Default is sine.
1930 Set modulation. Define how much of original signal is affected by the LFO.
1933 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1936 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1939 Set pulse width. Default is 1. Allowed range is [0 - 2].
1942 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1945 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1949 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1953 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1954 if timing is set to hz.
1960 Resample the input audio to the specified parameters, using the
1961 libswresample library. If none are specified then the filter will
1962 automatically convert between its input and output.
1964 This filter is also able to stretch/squeeze the audio data to make it match
1965 the timestamps or to inject silence / cut out audio to make it match the
1966 timestamps, do a combination of both or do neither.
1968 The filter accepts the syntax
1969 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1970 expresses a sample rate and @var{resampler_options} is a list of
1971 @var{key}=@var{value} pairs, separated by ":". See the
1972 @ref{Resampler Options,,"Resampler Options" section in the
1973 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1974 for the complete list of supported options.
1976 @subsection Examples
1980 Resample the input audio to 44100Hz:
1986 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1987 samples per second compensation:
1989 aresample=async=1000
1995 Reverse an audio clip.
1997 Warning: This filter requires memory to buffer the entire clip, so trimming
2000 @subsection Examples
2004 Take the first 5 seconds of a clip, and reverse it.
2006 atrim=end=5,areverse
2010 @section asetnsamples
2012 Set the number of samples per each output audio frame.
2014 The last output packet may contain a different number of samples, as
2015 the filter will flush all the remaining samples when the input audio
2018 The filter accepts the following options:
2022 @item nb_out_samples, n
2023 Set the number of frames per each output audio frame. The number is
2024 intended as the number of samples @emph{per each channel}.
2025 Default value is 1024.
2028 If set to 1, the filter will pad the last audio frame with zeroes, so
2029 that the last frame will contain the same number of samples as the
2030 previous ones. Default value is 1.
2033 For example, to set the number of per-frame samples to 1234 and
2034 disable padding for the last frame, use:
2036 asetnsamples=n=1234:p=0
2041 Set the sample rate without altering the PCM data.
2042 This will result in a change of speed and pitch.
2044 The filter accepts the following options:
2047 @item sample_rate, r
2048 Set the output sample rate. Default is 44100 Hz.
2053 Show a line containing various information for each input audio frame.
2054 The input audio is not modified.
2056 The shown line contains a sequence of key/value pairs of the form
2057 @var{key}:@var{value}.
2059 The following values are shown in the output:
2063 The (sequential) number of the input frame, starting from 0.
2066 The presentation timestamp of the input frame, in time base units; the time base
2067 depends on the filter input pad, and is usually 1/@var{sample_rate}.
2070 The presentation timestamp of the input frame in seconds.
2073 position of the frame in the input stream, -1 if this information in
2074 unavailable and/or meaningless (for example in case of synthetic audio)
2083 The sample rate for the audio frame.
2086 The number of samples (per channel) in the frame.
2089 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
2090 audio, the data is treated as if all the planes were concatenated.
2092 @item plane_checksums
2093 A list of Adler-32 checksums for each data plane.
2099 Display time domain statistical information about the audio channels.
2100 Statistics are calculated and displayed for each audio channel and,
2101 where applicable, an overall figure is also given.
2103 It accepts the following option:
2106 Short window length in seconds, used for peak and trough RMS measurement.
2107 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.01 - 10]}.
2111 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
2112 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
2115 Available keys for each channel are:
2151 For example full key look like this @code{lavfi.astats.1.DC_offset} or
2152 this @code{lavfi.astats.Overall.Peak_count}.
2154 For description what each key means read below.
2157 Set number of frame after which stats are going to be recalculated.
2158 Default is disabled.
2161 A description of each shown parameter follows:
2165 Mean amplitude displacement from zero.
2168 Minimal sample level.
2171 Maximal sample level.
2173 @item Min difference
2174 Minimal difference between two consecutive samples.
2176 @item Max difference
2177 Maximal difference between two consecutive samples.
2179 @item Mean difference
2180 Mean difference between two consecutive samples.
2181 The average of each difference between two consecutive samples.
2183 @item RMS difference
2184 Root Mean Square difference between two consecutive samples.
2188 Standard peak and RMS level measured in dBFS.
2192 Peak and trough values for RMS level measured over a short window.
2195 Standard ratio of peak to RMS level (note: not in dB).
2198 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
2199 (i.e. either @var{Min level} or @var{Max level}).
2202 Number of occasions (not the number of samples) that the signal attained either
2203 @var{Min level} or @var{Max level}.
2206 Overall bit depth of audio. Number of bits used for each sample.
2209 Measured dynamic range of audio in dB.
2211 @item Zero crossings
2212 Number of points where the waveform crosses the zero level axis.
2214 @item Zero crossings rate
2215 Rate of Zero crossings and number of audio samples.
2222 The filter accepts exactly one parameter, the audio tempo. If not
2223 specified then the filter will assume nominal 1.0 tempo. Tempo must
2224 be in the [0.5, 100.0] range.
2226 Note that tempo greater than 2 will skip some samples rather than
2227 blend them in. If for any reason this is a concern it is always
2228 possible to daisy-chain several instances of atempo to achieve the
2229 desired product tempo.
2231 @subsection Examples
2235 Slow down audio to 80% tempo:
2241 To speed up audio to 300% tempo:
2247 To speed up audio to 300% tempo by daisy-chaining two atempo instances:
2249 atempo=sqrt(3),atempo=sqrt(3)
2255 Trim the input so that the output contains one continuous subpart of the input.
2257 It accepts the following parameters:
2260 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
2261 sample with the timestamp @var{start} will be the first sample in the output.
2264 Specify time of the first audio sample that will be dropped, i.e. the
2265 audio sample immediately preceding the one with the timestamp @var{end} will be
2266 the last sample in the output.
2269 Same as @var{start}, except this option sets the start timestamp in samples
2273 Same as @var{end}, except this option sets the end timestamp in samples instead
2277 The maximum duration of the output in seconds.
2280 The number of the first sample that should be output.
2283 The number of the first sample that should be dropped.
2286 @option{start}, @option{end}, and @option{duration} are expressed as time
2287 duration specifications; see
2288 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
2290 Note that the first two sets of the start/end options and the @option{duration}
2291 option look at the frame timestamp, while the _sample options simply count the
2292 samples that pass through the filter. So start/end_pts and start/end_sample will
2293 give different results when the timestamps are wrong, inexact or do not start at
2294 zero. Also note that this filter does not modify the timestamps. If you wish
2295 to have the output timestamps start at zero, insert the asetpts filter after the
2298 If multiple start or end options are set, this filter tries to be greedy and
2299 keep all samples that match at least one of the specified constraints. To keep
2300 only the part that matches all the constraints at once, chain multiple atrim
2303 The defaults are such that all the input is kept. So it is possible to set e.g.
2304 just the end values to keep everything before the specified time.
2309 Drop everything except the second minute of input:
2311 ffmpeg -i INPUT -af atrim=60:120
2315 Keep only the first 1000 samples:
2317 ffmpeg -i INPUT -af atrim=end_sample=1000
2324 Apply a two-pole Butterworth band-pass filter with central
2325 frequency @var{frequency}, and (3dB-point) band-width width.
2326 The @var{csg} option selects a constant skirt gain (peak gain = Q)
2327 instead of the default: constant 0dB peak gain.
2328 The filter roll off at 6dB per octave (20dB per decade).
2330 The filter accepts the following options:
2334 Set the filter's central frequency. Default is @code{3000}.
2337 Constant skirt gain if set to 1. Defaults to 0.
2340 Set method to specify band-width of filter.
2355 Specify the band-width of a filter in width_type units.
2358 Specify which channels to filter, by default all available are filtered.
2361 @subsection Commands
2363 This filter supports the following commands:
2366 Change bandpass frequency.
2367 Syntax for the command is : "@var{frequency}"
2370 Change bandpass width_type.
2371 Syntax for the command is : "@var{width_type}"
2374 Change bandpass width.
2375 Syntax for the command is : "@var{width}"
2380 Apply a two-pole Butterworth band-reject filter with central
2381 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
2382 The filter roll off at 6dB per octave (20dB per decade).
2384 The filter accepts the following options:
2388 Set the filter's central frequency. Default is @code{3000}.
2391 Set method to specify band-width of filter.
2406 Specify the band-width of a filter in width_type units.
2409 Specify which channels to filter, by default all available are filtered.
2412 @subsection Commands
2414 This filter supports the following commands:
2417 Change bandreject frequency.
2418 Syntax for the command is : "@var{frequency}"
2421 Change bandreject width_type.
2422 Syntax for the command is : "@var{width_type}"
2425 Change bandreject width.
2426 Syntax for the command is : "@var{width}"
2429 @section bass, lowshelf
2431 Boost or cut the bass (lower) frequencies of the audio using a two-pole
2432 shelving filter with a response similar to that of a standard
2433 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
2435 The filter accepts the following options:
2439 Give the gain at 0 Hz. Its useful range is about -20
2440 (for a large cut) to +20 (for a large boost).
2441 Beware of clipping when using a positive gain.
2444 Set the filter's central frequency and so can be used
2445 to extend or reduce the frequency range to be boosted or cut.
2446 The default value is @code{100} Hz.
2449 Set method to specify band-width of filter.
2464 Determine how steep is the filter's shelf transition.
2467 Specify which channels to filter, by default all available are filtered.
2470 @subsection Commands
2472 This filter supports the following commands:
2475 Change bass frequency.
2476 Syntax for the command is : "@var{frequency}"
2479 Change bass width_type.
2480 Syntax for the command is : "@var{width_type}"
2484 Syntax for the command is : "@var{width}"
2488 Syntax for the command is : "@var{gain}"
2493 Apply a biquad IIR filter with the given coefficients.
2494 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
2495 are the numerator and denominator coefficients respectively.
2496 and @var{channels}, @var{c} specify which channels to filter, by default all
2497 available are filtered.
2499 @subsection Commands
2501 This filter supports the following commands:
2509 Change biquad parameter.
2510 Syntax for the command is : "@var{value}"
2514 Bauer stereo to binaural transformation, which improves headphone listening of
2515 stereo audio records.
2517 To enable compilation of this filter you need to configure FFmpeg with
2518 @code{--enable-libbs2b}.
2520 It accepts the following parameters:
2524 Pre-defined crossfeed level.
2528 Default level (fcut=700, feed=50).
2531 Chu Moy circuit (fcut=700, feed=60).
2534 Jan Meier circuit (fcut=650, feed=95).
2539 Cut frequency (in Hz).
2548 Remap input channels to new locations.
2550 It accepts the following parameters:
2553 Map channels from input to output. The argument is a '|'-separated list of
2554 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
2555 @var{in_channel} form. @var{in_channel} can be either the name of the input
2556 channel (e.g. FL for front left) or its index in the input channel layout.
2557 @var{out_channel} is the name of the output channel or its index in the output
2558 channel layout. If @var{out_channel} is not given then it is implicitly an
2559 index, starting with zero and increasing by one for each mapping.
2561 @item channel_layout
2562 The channel layout of the output stream.
2565 If no mapping is present, the filter will implicitly map input channels to
2566 output channels, preserving indices.
2568 @subsection Examples
2572 For example, assuming a 5.1+downmix input MOV file,
2574 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
2576 will create an output WAV file tagged as stereo from the downmix channels of
2580 To fix a 5.1 WAV improperly encoded in AAC's native channel order
2582 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
2586 @section channelsplit
2588 Split each channel from an input audio stream into a separate output stream.
2590 It accepts the following parameters:
2592 @item channel_layout
2593 The channel layout of the input stream. The default is "stereo".
2595 A channel layout describing the channels to be extracted as separate output streams
2596 or "all" to extract each input channel as a separate stream. The default is "all".
2598 Choosing channels not present in channel layout in the input will result in an error.
2601 @subsection Examples
2605 For example, assuming a stereo input MP3 file,
2607 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
2609 will create an output Matroska file with two audio streams, one containing only
2610 the left channel and the other the right channel.
2613 Split a 5.1 WAV file into per-channel files:
2615 ffmpeg -i in.wav -filter_complex
2616 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
2617 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
2618 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2623 Extract only LFE from a 5.1 WAV file:
2625 ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
2626 -map '[LFE]' lfe.wav
2631 Add a chorus effect to the audio.
2633 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2635 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2636 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2637 The modulation depth defines the range the modulated delay is played before or after
2638 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2639 sound tuned around the original one, like in a chorus where some vocals are slightly
2642 It accepts the following parameters:
2645 Set input gain. Default is 0.4.
2648 Set output gain. Default is 0.4.
2651 Set delays. A typical delay is around 40ms to 60ms.
2663 @subsection Examples
2669 chorus=0.7:0.9:55:0.4:0.25:2
2675 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2679 Fuller sounding chorus with three delays:
2681 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
2686 Compress or expand the audio's dynamic range.
2688 It accepts the following parameters:
2694 A list of times in seconds for each channel over which the instantaneous level
2695 of the input signal is averaged to determine its volume. @var{attacks} refers to
2696 increase of volume and @var{decays} refers to decrease of volume. For most
2697 situations, the attack time (response to the audio getting louder) should be
2698 shorter than the decay time, because the human ear is more sensitive to sudden
2699 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2700 a typical value for decay is 0.8 seconds.
2701 If specified number of attacks & decays is lower than number of channels, the last
2702 set attack/decay will be used for all remaining channels.
2705 A list of points for the transfer function, specified in dB relative to the
2706 maximum possible signal amplitude. Each key points list must be defined using
2707 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2708 @code{x0/y0 x1/y1 x2/y2 ....}
2710 The input values must be in strictly increasing order but the transfer function
2711 does not have to be monotonically rising. The point @code{0/0} is assumed but
2712 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2713 function are @code{-70/-70|-60/-20|1/0}.
2716 Set the curve radius in dB for all joints. It defaults to 0.01.
2719 Set the additional gain in dB to be applied at all points on the transfer
2720 function. This allows for easy adjustment of the overall gain.
2724 Set an initial volume, in dB, to be assumed for each channel when filtering
2725 starts. This permits the user to supply a nominal level initially, so that, for
2726 example, a very large gain is not applied to initial signal levels before the
2727 companding has begun to operate. A typical value for audio which is initially
2728 quiet is -90 dB. It defaults to 0.
2731 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2732 delayed before being fed to the volume adjuster. Specifying a delay
2733 approximately equal to the attack/decay times allows the filter to effectively
2734 operate in predictive rather than reactive mode. It defaults to 0.
2738 @subsection Examples
2742 Make music with both quiet and loud passages suitable for listening to in a
2745 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2748 Another example for audio with whisper and explosion parts:
2750 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2754 A noise gate for when the noise is at a lower level than the signal:
2756 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2760 Here is another noise gate, this time for when the noise is at a higher level
2761 than the signal (making it, in some ways, similar to squelch):
2763 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2767 2:1 compression starting at -6dB:
2769 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2773 2:1 compression starting at -9dB:
2775 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2779 2:1 compression starting at -12dB:
2781 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2785 2:1 compression starting at -18dB:
2787 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2791 3:1 compression starting at -15dB:
2793 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2799 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2805 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
2809 Hard limiter at -6dB:
2811 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2815 Hard limiter at -12dB:
2817 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2821 Hard noise gate at -35 dB:
2823 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2829 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2833 @section compensationdelay
2835 Compensation Delay Line is a metric based delay to compensate differing
2836 positions of microphones or speakers.
2838 For example, you have recorded guitar with two microphones placed in
2839 different location. Because the front of sound wave has fixed speed in
2840 normal conditions, the phasing of microphones can vary and depends on
2841 their location and interposition. The best sound mix can be achieved when
2842 these microphones are in phase (synchronized). Note that distance of
2843 ~30 cm between microphones makes one microphone to capture signal in
2844 antiphase to another microphone. That makes the final mix sounding moody.
2845 This filter helps to solve phasing problems by adding different delays
2846 to each microphone track and make them synchronized.
2848 The best result can be reached when you take one track as base and
2849 synchronize other tracks one by one with it.
2850 Remember that synchronization/delay tolerance depends on sample rate, too.
2851 Higher sample rates will give more tolerance.
2853 It accepts the following parameters:
2857 Set millimeters distance. This is compensation distance for fine tuning.
2861 Set cm distance. This is compensation distance for tightening distance setup.
2865 Set meters distance. This is compensation distance for hard distance setup.
2869 Set dry amount. Amount of unprocessed (dry) signal.
2873 Set wet amount. Amount of processed (wet) signal.
2877 Set temperature degree in Celsius. This is the temperature of the environment.
2882 Apply headphone crossfeed filter.
2884 Crossfeed is the process of blending the left and right channels of stereo
2886 It is mainly used to reduce extreme stereo separation of low frequencies.
2888 The intent is to produce more speaker like sound to the listener.
2890 The filter accepts the following options:
2894 Set strength of crossfeed. Default is 0.2. Allowed range is from 0 to 1.
2895 This sets gain of low shelf filter for side part of stereo image.
2896 Default is -6dB. Max allowed is -30db when strength is set to 1.
2899 Set soundstage wideness. Default is 0.5. Allowed range is from 0 to 1.
2900 This sets cut off frequency of low shelf filter. Default is cut off near
2901 1550 Hz. With range set to 1 cut off frequency is set to 2100 Hz.
2904 Set input gain. Default is 0.9.
2907 Set output gain. Default is 1.
2910 @section crystalizer
2911 Simple algorithm to expand audio dynamic range.
2913 The filter accepts the following options:
2917 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2918 (unchanged sound) to 10.0 (maximum effect).
2921 Enable clipping. By default is enabled.
2925 Apply a DC shift to the audio.
2927 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2928 in the recording chain) from the audio. The effect of a DC offset is reduced
2929 headroom and hence volume. The @ref{astats} filter can be used to determine if
2930 a signal has a DC offset.
2934 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2938 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2939 used to prevent clipping.
2943 Measure audio dynamic range.
2945 DR values of 14 and higher is found in very dynamic material. DR of 8 to 13
2946 is found in transition material. And anything less that 8 have very poor dynamics
2947 and is very compressed.
2949 The filter accepts the following options:
2953 Set window length in seconds used to split audio into segments of equal length.
2954 Default is 3 seconds.
2958 Dynamic Audio Normalizer.
2960 This filter applies a certain amount of gain to the input audio in order
2961 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2962 contrast to more "simple" normalization algorithms, the Dynamic Audio
2963 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2964 This allows for applying extra gain to the "quiet" sections of the audio
2965 while avoiding distortions or clipping the "loud" sections. In other words:
2966 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2967 sections, in the sense that the volume of each section is brought to the
2968 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2969 this goal *without* applying "dynamic range compressing". It will retain 100%
2970 of the dynamic range *within* each section of the audio file.
2974 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2975 Default is 500 milliseconds.
2976 The Dynamic Audio Normalizer processes the input audio in small chunks,
2977 referred to as frames. This is required, because a peak magnitude has no
2978 meaning for just a single sample value. Instead, we need to determine the
2979 peak magnitude for a contiguous sequence of sample values. While a "standard"
2980 normalizer would simply use the peak magnitude of the complete file, the
2981 Dynamic Audio Normalizer determines the peak magnitude individually for each
2982 frame. The length of a frame is specified in milliseconds. By default, the
2983 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2984 been found to give good results with most files.
2985 Note that the exact frame length, in number of samples, will be determined
2986 automatically, based on the sampling rate of the individual input audio file.
2989 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2990 number. Default is 31.
2991 Probably the most important parameter of the Dynamic Audio Normalizer is the
2992 @code{window size} of the Gaussian smoothing filter. The filter's window size
2993 is specified in frames, centered around the current frame. For the sake of
2994 simplicity, this must be an odd number. Consequently, the default value of 31
2995 takes into account the current frame, as well as the 15 preceding frames and
2996 the 15 subsequent frames. Using a larger window results in a stronger
2997 smoothing effect and thus in less gain variation, i.e. slower gain
2998 adaptation. Conversely, using a smaller window results in a weaker smoothing
2999 effect and thus in more gain variation, i.e. faster gain adaptation.
3000 In other words, the more you increase this value, the more the Dynamic Audio
3001 Normalizer will behave like a "traditional" normalization filter. On the
3002 contrary, the more you decrease this value, the more the Dynamic Audio
3003 Normalizer will behave like a dynamic range compressor.
3006 Set the target peak value. This specifies the highest permissible magnitude
3007 level for the normalized audio input. This filter will try to approach the
3008 target peak magnitude as closely as possible, but at the same time it also
3009 makes sure that the normalized signal will never exceed the peak magnitude.
3010 A frame's maximum local gain factor is imposed directly by the target peak
3011 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
3012 It is not recommended to go above this value.
3015 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
3016 The Dynamic Audio Normalizer determines the maximum possible (local) gain
3017 factor for each input frame, i.e. the maximum gain factor that does not
3018 result in clipping or distortion. The maximum gain factor is determined by
3019 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
3020 additionally bounds the frame's maximum gain factor by a predetermined
3021 (global) maximum gain factor. This is done in order to avoid excessive gain
3022 factors in "silent" or almost silent frames. By default, the maximum gain
3023 factor is 10.0, For most inputs the default value should be sufficient and
3024 it usually is not recommended to increase this value. Though, for input
3025 with an extremely low overall volume level, it may be necessary to allow even
3026 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
3027 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
3028 Instead, a "sigmoid" threshold function will be applied. This way, the
3029 gain factors will smoothly approach the threshold value, but never exceed that
3033 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
3034 By default, the Dynamic Audio Normalizer performs "peak" normalization.
3035 This means that the maximum local gain factor for each frame is defined
3036 (only) by the frame's highest magnitude sample. This way, the samples can
3037 be amplified as much as possible without exceeding the maximum signal
3038 level, i.e. without clipping. Optionally, however, the Dynamic Audio
3039 Normalizer can also take into account the frame's root mean square,
3040 abbreviated RMS. In electrical engineering, the RMS is commonly used to
3041 determine the power of a time-varying signal. It is therefore considered
3042 that the RMS is a better approximation of the "perceived loudness" than
3043 just looking at the signal's peak magnitude. Consequently, by adjusting all
3044 frames to a constant RMS value, a uniform "perceived loudness" can be
3045 established. If a target RMS value has been specified, a frame's local gain
3046 factor is defined as the factor that would result in exactly that RMS value.
3047 Note, however, that the maximum local gain factor is still restricted by the
3048 frame's highest magnitude sample, in order to prevent clipping.
3051 Enable channels coupling. By default is enabled.
3052 By default, the Dynamic Audio Normalizer will amplify all channels by the same
3053 amount. This means the same gain factor will be applied to all channels, i.e.
3054 the maximum possible gain factor is determined by the "loudest" channel.
3055 However, in some recordings, it may happen that the volume of the different
3056 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
3057 In this case, this option can be used to disable the channel coupling. This way,
3058 the gain factor will be determined independently for each channel, depending
3059 only on the individual channel's highest magnitude sample. This allows for
3060 harmonizing the volume of the different channels.
3063 Enable DC bias correction. By default is disabled.
3064 An audio signal (in the time domain) is a sequence of sample values.
3065 In the Dynamic Audio Normalizer these sample values are represented in the
3066 -1.0 to 1.0 range, regardless of the original input format. Normally, the
3067 audio signal, or "waveform", should be centered around the zero point.
3068 That means if we calculate the mean value of all samples in a file, or in a
3069 single frame, then the result should be 0.0 or at least very close to that
3070 value. If, however, there is a significant deviation of the mean value from
3071 0.0, in either positive or negative direction, this is referred to as a
3072 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
3073 Audio Normalizer provides optional DC bias correction.
3074 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
3075 the mean value, or "DC correction" offset, of each input frame and subtract
3076 that value from all of the frame's sample values which ensures those samples
3077 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
3078 boundaries, the DC correction offset values will be interpolated smoothly
3079 between neighbouring frames.
3082 Enable alternative boundary mode. By default is disabled.
3083 The Dynamic Audio Normalizer takes into account a certain neighbourhood
3084 around each frame. This includes the preceding frames as well as the
3085 subsequent frames. However, for the "boundary" frames, located at the very
3086 beginning and at the very end of the audio file, not all neighbouring
3087 frames are available. In particular, for the first few frames in the audio
3088 file, the preceding frames are not known. And, similarly, for the last few
3089 frames in the audio file, the subsequent frames are not known. Thus, the
3090 question arises which gain factors should be assumed for the missing frames
3091 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
3092 to deal with this situation. The default boundary mode assumes a gain factor
3093 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
3094 "fade out" at the beginning and at the end of the input, respectively.
3097 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
3098 By default, the Dynamic Audio Normalizer does not apply "traditional"
3099 compression. This means that signal peaks will not be pruned and thus the
3100 full dynamic range will be retained within each local neighbourhood. However,
3101 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
3102 normalization algorithm with a more "traditional" compression.
3103 For this purpose, the Dynamic Audio Normalizer provides an optional compression
3104 (thresholding) function. If (and only if) the compression feature is enabled,
3105 all input frames will be processed by a soft knee thresholding function prior
3106 to the actual normalization process. Put simply, the thresholding function is
3107 going to prune all samples whose magnitude exceeds a certain threshold value.
3108 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
3109 value. Instead, the threshold value will be adjusted for each individual
3111 In general, smaller parameters result in stronger compression, and vice versa.
3112 Values below 3.0 are not recommended, because audible distortion may appear.
3117 Make audio easier to listen to on headphones.
3119 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
3120 so that when listened to on headphones the stereo image is moved from
3121 inside your head (standard for headphones) to outside and in front of
3122 the listener (standard for speakers).
3128 Apply a two-pole peaking equalisation (EQ) filter. With this
3129 filter, the signal-level at and around a selected frequency can
3130 be increased or decreased, whilst (unlike bandpass and bandreject
3131 filters) that at all other frequencies is unchanged.
3133 In order to produce complex equalisation curves, this filter can
3134 be given several times, each with a different central frequency.
3136 The filter accepts the following options:
3140 Set the filter's central frequency in Hz.
3143 Set method to specify band-width of filter.
3158 Specify the band-width of a filter in width_type units.
3161 Set the required gain or attenuation in dB.
3162 Beware of clipping when using a positive gain.
3165 Specify which channels to filter, by default all available are filtered.
3168 @subsection Examples
3171 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
3173 equalizer=f=1000:t=h:width=200:g=-10
3177 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
3179 equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
3183 @subsection Commands
3185 This filter supports the following commands:
3188 Change equalizer frequency.
3189 Syntax for the command is : "@var{frequency}"
3192 Change equalizer width_type.
3193 Syntax for the command is : "@var{width_type}"
3196 Change equalizer width.
3197 Syntax for the command is : "@var{width}"
3200 Change equalizer gain.
3201 Syntax for the command is : "@var{gain}"
3204 @section extrastereo
3206 Linearly increases the difference between left and right channels which
3207 adds some sort of "live" effect to playback.
3209 The filter accepts the following options:
3213 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
3214 (average of both channels), with 1.0 sound will be unchanged, with
3215 -1.0 left and right channels will be swapped.
3218 Enable clipping. By default is enabled.
3221 @section firequalizer
3222 Apply FIR Equalization using arbitrary frequency response.
3224 The filter accepts the following option:
3228 Set gain curve equation (in dB). The expression can contain variables:
3231 the evaluated frequency
3235 channel number, set to 0 when multichannels evaluation is disabled
3237 channel id, see libavutil/channel_layout.h, set to the first channel id when
3238 multichannels evaluation is disabled
3242 channel_layout, see libavutil/channel_layout.h
3247 @item gain_interpolate(f)
3248 interpolate gain on frequency f based on gain_entry
3249 @item cubic_interpolate(f)
3250 same as gain_interpolate, but smoother
3252 This option is also available as command. Default is @code{gain_interpolate(f)}.
3255 Set gain entry for gain_interpolate function. The expression can
3259 store gain entry at frequency f with value g
3261 This option is also available as command.
3264 Set filter delay in seconds. Higher value means more accurate.
3265 Default is @code{0.01}.
3268 Set filter accuracy in Hz. Lower value means more accurate.
3269 Default is @code{5}.
3272 Set window function. Acceptable values are:
3275 rectangular window, useful when gain curve is already smooth
3277 hann window (default)
3283 3-terms continuous 1st derivative nuttall window
3285 minimum 3-terms discontinuous nuttall window
3287 4-terms continuous 1st derivative nuttall window
3289 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
3291 blackman-harris window
3297 If enabled, use fixed number of audio samples. This improves speed when
3298 filtering with large delay. Default is disabled.
3301 Enable multichannels evaluation on gain. Default is disabled.
3304 Enable zero phase mode by subtracting timestamp to compensate delay.
3305 Default is disabled.
3308 Set scale used by gain. Acceptable values are:
3311 linear frequency, linear gain
3313 linear frequency, logarithmic (in dB) gain (default)
3315 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
3317 logarithmic frequency, logarithmic gain
3321 Set file for dumping, suitable for gnuplot.
3324 Set scale for dumpfile. Acceptable values are same with scale option.
3328 Enable 2-channel convolution using complex FFT. This improves speed significantly.
3329 Default is disabled.
3332 Enable minimum phase impulse response. Default is disabled.
3335 @subsection Examples
3340 firequalizer=gain='if(lt(f,1000), 0, -INF)'
3343 lowpass at 1000 Hz with gain_entry:
3345 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
3348 custom equalization:
3350 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
3353 higher delay with zero phase to compensate delay:
3355 firequalizer=delay=0.1:fixed=on:zero_phase=on
3358 lowpass on left channel, highpass on right channel:
3360 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
3361 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
3366 Apply a flanging effect to the audio.
3368 The filter accepts the following options:
3372 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
3375 Set added sweep delay in milliseconds. Range from 0 to 10. Default value is 2.
3378 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
3382 Set percentage of delayed signal mixed with original. Range from 0 to 100.
3383 Default value is 71.
3386 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
3389 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
3390 Default value is @var{sinusoidal}.
3393 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
3394 Default value is 25.
3397 Set delay-line interpolation, @var{linear} or @var{quadratic}.
3398 Default is @var{linear}.
3402 Apply Haas effect to audio.
3404 Note that this makes most sense to apply on mono signals.
3405 With this filter applied to mono signals it give some directionality and
3406 stretches its stereo image.
3408 The filter accepts the following options:
3412 Set input level. By default is @var{1}, or 0dB
3415 Set output level. By default is @var{1}, or 0dB.
3418 Set gain applied to side part of signal. By default is @var{1}.
3421 Set kind of middle source. Can be one of the following:
3431 Pick middle part signal of stereo image.
3434 Pick side part signal of stereo image.
3438 Change middle phase. By default is disabled.
3441 Set left channel delay. By default is @var{2.05} milliseconds.
3444 Set left channel balance. By default is @var{-1}.
3447 Set left channel gain. By default is @var{1}.
3450 Change left phase. By default is disabled.
3453 Set right channel delay. By defaults is @var{2.12} milliseconds.
3456 Set right channel balance. By default is @var{1}.
3459 Set right channel gain. By default is @var{1}.
3462 Change right phase. By default is enabled.
3467 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
3468 embedded HDCD codes is expanded into a 20-bit PCM stream.
3470 The filter supports the Peak Extend and Low-level Gain Adjustment features
3471 of HDCD, and detects the Transient Filter flag.
3474 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
3477 When using the filter with wav, note the default encoding for wav is 16-bit,
3478 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
3479 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
3481 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
3482 ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
3485 The filter accepts the following options:
3488 @item disable_autoconvert
3489 Disable any automatic format conversion or resampling in the filter graph.
3491 @item process_stereo
3492 Process the stereo channels together. If target_gain does not match between
3493 channels, consider it invalid and use the last valid target_gain.
3496 Set the code detect timer period in ms.
3499 Always extend peaks above -3dBFS even if PE isn't signaled.
3502 Replace audio with a solid tone and adjust the amplitude to signal some
3503 specific aspect of the decoding process. The output file can be loaded in
3504 an audio editor alongside the original to aid analysis.
3506 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
3513 Gain adjustment level at each sample
3515 Samples where peak extend occurs
3517 Samples where the code detect timer is active
3519 Samples where the target gain does not match between channels
3525 Apply head-related transfer functions (HRTFs) to create virtual
3526 loudspeakers around the user for binaural listening via headphones.
3527 The HRIRs are provided via additional streams, for each channel
3528 one stereo input stream is needed.
3530 The filter accepts the following options:
3534 Set mapping of input streams for convolution.
3535 The argument is a '|'-separated list of channel names in order as they
3536 are given as additional stream inputs for filter.
3537 This also specify number of input streams. Number of input streams
3538 must be not less than number of channels in first stream plus one.
3541 Set gain applied to audio. Value is in dB. Default is 0.
3544 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3545 processing audio in time domain which is slow.
3546 @var{freq} is processing audio in frequency domain which is fast.
3547 Default is @var{freq}.
3550 Set custom gain for LFE channels. Value is in dB. Default is 0.
3553 Set size of frame in number of samples which will be processed at once.
3554 Default value is @var{1024}. Allowed range is from 1024 to 96000.
3557 Set format of hrir stream.
3558 Default value is @var{stereo}. Alternative value is @var{multich}.
3559 If value is set to @var{stereo}, number of additional streams should
3560 be greater or equal to number of input channels in first input stream.
3561 Also each additional stream should have stereo number of channels.
3562 If value is set to @var{multich}, number of additional streams should
3563 be exactly one. Also number of input channels of additional stream
3564 should be equal or greater than twice number of channels of first input
3568 @subsection Examples
3572 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3573 each amovie filter use stereo file with IR coefficients as input.
3574 The files give coefficients for each position of virtual loudspeaker:
3577 -filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
3582 Full example using wav files as coefficients with amovie filters for 7.1 downmix,
3583 but now in @var{multich} @var{hrir} format.
3585 ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
3592 Apply a high-pass filter with 3dB point frequency.
3593 The filter can be either single-pole, or double-pole (the default).
3594 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3596 The filter accepts the following options:
3600 Set frequency in Hz. Default is 3000.
3603 Set number of poles. Default is 2.
3606 Set method to specify band-width of filter.
3621 Specify the band-width of a filter in width_type units.
3622 Applies only to double-pole filter.
3623 The default is 0.707q and gives a Butterworth response.
3626 Specify which channels to filter, by default all available are filtered.
3629 @subsection Commands
3631 This filter supports the following commands:
3634 Change highpass frequency.
3635 Syntax for the command is : "@var{frequency}"
3638 Change highpass width_type.
3639 Syntax for the command is : "@var{width_type}"
3642 Change highpass width.
3643 Syntax for the command is : "@var{width}"
3648 Join multiple input streams into one multi-channel stream.
3650 It accepts the following parameters:
3654 The number of input streams. It defaults to 2.
3656 @item channel_layout
3657 The desired output channel layout. It defaults to stereo.
3660 Map channels from inputs to output. The argument is a '|'-separated list of
3661 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
3662 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
3663 can be either the name of the input channel (e.g. FL for front left) or its
3664 index in the specified input stream. @var{out_channel} is the name of the output
3668 The filter will attempt to guess the mappings when they are not specified
3669 explicitly. It does so by first trying to find an unused matching input channel
3670 and if that fails it picks the first unused input channel.
3672 Join 3 inputs (with properly set channel layouts):
3674 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
3677 Build a 5.1 output from 6 single-channel streams:
3679 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
3680 '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'
3686 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
3688 To enable compilation of this filter you need to configure FFmpeg with
3689 @code{--enable-ladspa}.
3693 Specifies the name of LADSPA plugin library to load. If the environment
3694 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
3695 each one of the directories specified by the colon separated list in
3696 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
3697 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
3698 @file{/usr/lib/ladspa/}.
3701 Specifies the plugin within the library. Some libraries contain only
3702 one plugin, but others contain many of them. If this is not set filter
3703 will list all available plugins within the specified library.
3706 Set the '|' separated list of controls which are zero or more floating point
3707 values that determine the behavior of the loaded plugin (for example delay,
3709 Controls need to be defined using the following syntax:
3710 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
3711 @var{valuei} is the value set on the @var{i}-th control.
3712 Alternatively they can be also defined using the following syntax:
3713 @var{value0}|@var{value1}|@var{value2}|..., where
3714 @var{valuei} is the value set on the @var{i}-th control.
3715 If @option{controls} is set to @code{help}, all available controls and
3716 their valid ranges are printed.
3718 @item sample_rate, s
3719 Specify the sample rate, default to 44100. Only used if plugin have
3723 Set the number of samples per channel per each output frame, default
3724 is 1024. Only used if plugin have zero inputs.
3727 Set the minimum duration of the sourced audio. See
3728 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3729 for the accepted syntax.
3730 Note that the resulting duration may be greater than the specified duration,
3731 as the generated audio is always cut at the end of a complete frame.
3732 If not specified, or the expressed duration is negative, the audio is
3733 supposed to be generated forever.
3734 Only used if plugin have zero inputs.
3738 @subsection Examples
3742 List all available plugins within amp (LADSPA example plugin) library:
3748 List all available controls and their valid ranges for @code{vcf_notch}
3749 plugin from @code{VCF} library:
3751 ladspa=f=vcf:p=vcf_notch:c=help
3755 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
3758 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
3762 Add reverberation to the audio using TAP-plugins
3763 (Tom's Audio Processing plugins):
3765 ladspa=file=tap_reverb:tap_reverb
3769 Generate white noise, with 0.2 amplitude:
3771 ladspa=file=cmt:noise_source_white:c=c0=.2
3775 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
3776 @code{C* Audio Plugin Suite} (CAPS) library:
3778 ladspa=file=caps:Click:c=c1=20'
3782 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
3784 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
3788 Increase volume by 20dB using fast lookahead limiter from Steve Harris
3789 @code{SWH Plugins} collection:
3791 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
3795 Attenuate low frequencies using Multiband EQ from Steve Harris
3796 @code{SWH Plugins} collection:
3798 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
3802 Reduce stereo image using @code{Narrower} from the @code{C* Audio Plugin Suite}
3805 ladspa=caps:Narrower
3809 Another white noise, now using @code{C* Audio Plugin Suite} (CAPS) library:
3811 ladspa=caps:White:.2
3815 Some fractal noise, using @code{C* Audio Plugin Suite} (CAPS) library:
3817 ladspa=caps:Fractal:c=c1=1
3821 Dynamic volume normalization using @code{VLevel} plugin:
3823 ladspa=vlevel-ladspa:vlevel_mono
3827 @subsection Commands
3829 This filter supports the following commands:
3832 Modify the @var{N}-th control value.
3834 If the specified value is not valid, it is ignored and prior one is kept.
3839 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
3840 Support for both single pass (livestreams, files) and double pass (files) modes.
3841 This algorithm can target IL, LRA, and maximum true peak. To accurately detect true peaks,
3842 the audio stream will be upsampled to 192 kHz unless the normalization mode is linear.
3843 Use the @code{-ar} option or @code{aresample} filter to explicitly set an output sample rate.
3845 The filter accepts the following options:
3849 Set integrated loudness target.
3850 Range is -70.0 - -5.0. Default value is -24.0.
3853 Set loudness range target.
3854 Range is 1.0 - 20.0. Default value is 7.0.
3857 Set maximum true peak.
3858 Range is -9.0 - +0.0. Default value is -2.0.
3860 @item measured_I, measured_i
3861 Measured IL of input file.
3862 Range is -99.0 - +0.0.
3864 @item measured_LRA, measured_lra
3865 Measured LRA of input file.
3866 Range is 0.0 - 99.0.
3868 @item measured_TP, measured_tp
3869 Measured true peak of input file.
3870 Range is -99.0 - +99.0.
3872 @item measured_thresh
3873 Measured threshold of input file.
3874 Range is -99.0 - +0.0.
3877 Set offset gain. Gain is applied before the true-peak limiter.
3878 Range is -99.0 - +99.0. Default is +0.0.
3881 Normalize linearly if possible.
3882 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3883 to be specified in order to use this mode.
3884 Options are true or false. Default is true.
3887 Treat mono input files as "dual-mono". If a mono file is intended for playback
3888 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3889 If set to @code{true}, this option will compensate for this effect.
3890 Multi-channel input files are not affected by this option.
3891 Options are true or false. Default is false.
3894 Set print format for stats. Options are summary, json, or none.
3895 Default value is none.
3900 Apply a low-pass filter with 3dB point frequency.
3901 The filter can be either single-pole or double-pole (the default).
3902 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3904 The filter accepts the following options:
3908 Set frequency in Hz. Default is 500.
3911 Set number of poles. Default is 2.
3914 Set method to specify band-width of filter.
3929 Specify the band-width of a filter in width_type units.
3930 Applies only to double-pole filter.
3931 The default is 0.707q and gives a Butterworth response.
3934 Specify which channels to filter, by default all available are filtered.
3937 @subsection Examples
3940 Lowpass only LFE channel, it LFE is not present it does nothing:
3946 @subsection Commands
3948 This filter supports the following commands:
3951 Change lowpass frequency.
3952 Syntax for the command is : "@var{frequency}"
3955 Change lowpass width_type.
3956 Syntax for the command is : "@var{width_type}"
3959 Change lowpass width.
3960 Syntax for the command is : "@var{width}"
3965 Load a LV2 (LADSPA Version 2) plugin.
3967 To enable compilation of this filter you need to configure FFmpeg with
3968 @code{--enable-lv2}.
3972 Specifies the plugin URI. You may need to escape ':'.
3975 Set the '|' separated list of controls which are zero or more floating point
3976 values that determine the behavior of the loaded plugin (for example delay,
3978 If @option{controls} is set to @code{help}, all available controls and
3979 their valid ranges are printed.
3981 @item sample_rate, s
3982 Specify the sample rate, default to 44100. Only used if plugin have
3986 Set the number of samples per channel per each output frame, default
3987 is 1024. Only used if plugin have zero inputs.
3990 Set the minimum duration of the sourced audio. See
3991 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
3992 for the accepted syntax.
3993 Note that the resulting duration may be greater than the specified duration,
3994 as the generated audio is always cut at the end of a complete frame.
3995 If not specified, or the expressed duration is negative, the audio is
3996 supposed to be generated forever.
3997 Only used if plugin have zero inputs.
4000 @subsection Examples
4004 Apply bass enhancer plugin from Calf:
4006 lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
4010 Apply vinyl plugin from Calf:
4012 lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
4016 Apply bit crusher plugin from ArtyFX:
4018 lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
4023 Multiband Compress or expand the audio's dynamic range.
4025 The input audio is divided into bands using 4th order Linkwitz-Riley IIRs.
4026 This is akin to the crossover of a loudspeaker, and results in flat frequency
4027 response when absent compander action.
4029 It accepts the following parameters:
4033 This option syntax is:
4034 attack,decay,[attack,decay..] soft-knee points crossover_frequency [delay [initial_volume [gain]]] | attack,decay ...
4035 For explanation of each item refer to compand filter documentation.
4041 Mix channels with specific gain levels. The filter accepts the output
4042 channel layout followed by a set of channels definitions.
4044 This filter is also designed to efficiently remap the channels of an audio
4047 The filter accepts parameters of the form:
4048 "@var{l}|@var{outdef}|@var{outdef}|..."
4052 output channel layout or number of channels
4055 output channel specification, of the form:
4056 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
4059 output channel to define, either a channel name (FL, FR, etc.) or a channel
4060 number (c0, c1, etc.)
4063 multiplicative coefficient for the channel, 1 leaving the volume unchanged
4066 input channel to use, see out_name for details; it is not possible to mix
4067 named and numbered input channels
4070 If the `=' in a channel specification is replaced by `<', then the gains for
4071 that specification will be renormalized so that the total is 1, thus
4072 avoiding clipping noise.
4074 @subsection Mixing examples
4076 For example, if you want to down-mix from stereo to mono, but with a bigger
4077 factor for the left channel:
4079 pan=1c|c0=0.9*c0+0.1*c1
4082 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
4083 7-channels surround:
4085 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
4088 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
4089 that should be preferred (see "-ac" option) unless you have very specific
4092 @subsection Remapping examples
4094 The channel remapping will be effective if, and only if:
4097 @item gain coefficients are zeroes or ones,
4098 @item only one input per channel output,
4101 If all these conditions are satisfied, the filter will notify the user ("Pure
4102 channel mapping detected"), and use an optimized and lossless method to do the
4105 For example, if you have a 5.1 source and want a stereo audio stream by
4106 dropping the extra channels:
4108 pan="stereo| c0=FL | c1=FR"
4111 Given the same source, you can also switch front left and front right channels
4112 and keep the input channel layout:
4114 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
4117 If the input is a stereo audio stream, you can mute the front left channel (and
4118 still keep the stereo channel layout) with:
4123 Still with a stereo audio stream input, you can copy the right channel in both
4124 front left and right:
4126 pan="stereo| c0=FR | c1=FR"
4131 ReplayGain scanner filter. This filter takes an audio stream as an input and
4132 outputs it unchanged.
4133 At end of filtering it displays @code{track_gain} and @code{track_peak}.
4137 Convert the audio sample format, sample rate and channel layout. It is
4138 not meant to be used directly.
4141 Apply time-stretching and pitch-shifting with librubberband.
4143 To enable compilation of this filter, you need to configure FFmpeg with
4144 @code{--enable-librubberband}.
4146 The filter accepts the following options:
4150 Set tempo scale factor.
4153 Set pitch scale factor.
4156 Set transients detector.
4157 Possible values are:
4166 Possible values are:
4175 Possible values are:
4182 Set processing window size.
4183 Possible values are:
4192 Possible values are:
4199 Enable formant preservation when shift pitching.
4200 Possible values are:
4208 Possible values are:
4217 Possible values are:
4224 @section sidechaincompress
4226 This filter acts like normal compressor but has the ability to compress
4227 detected signal using second input signal.
4228 It needs two input streams and returns one output stream.
4229 First input stream will be processed depending on second stream signal.
4230 The filtered signal then can be filtered with other filters in later stages of
4231 processing. See @ref{pan} and @ref{amerge} filter.
4233 The filter accepts the following options:
4237 Set input gain. Default is 1. Range is between 0.015625 and 64.
4240 If a signal of second stream raises above this level it will affect the gain
4241 reduction of first stream.
4242 By default is 0.125. Range is between 0.00097563 and 1.
4245 Set a ratio about which the signal is reduced. 1:2 means that if the level
4246 raised 4dB above the threshold, it will be only 2dB above after the reduction.
4247 Default is 2. Range is between 1 and 20.
4250 Amount of milliseconds the signal has to rise above the threshold before gain
4251 reduction starts. Default is 20. Range is between 0.01 and 2000.
4254 Amount of milliseconds the signal has to fall below the threshold before
4255 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
4258 Set the amount by how much signal will be amplified after processing.
4259 Default is 1. Range is from 1 to 64.
4262 Curve the sharp knee around the threshold to enter gain reduction more softly.
4263 Default is 2.82843. Range is between 1 and 8.
4266 Choose if the @code{average} level between all channels of side-chain stream
4267 or the louder(@code{maximum}) channel of side-chain stream affects the
4268 reduction. Default is @code{average}.
4271 Should the exact signal be taken in case of @code{peak} or an RMS one in case
4272 of @code{rms}. Default is @code{rms} which is mainly smoother.
4275 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
4278 How much to use compressed signal in output. Default is 1.
4279 Range is between 0 and 1.
4282 @subsection Examples
4286 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
4287 depending on the signal of 2nd input and later compressed signal to be
4288 merged with 2nd input:
4290 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
4294 @section sidechaingate
4296 A sidechain gate acts like a normal (wideband) gate but has the ability to
4297 filter the detected signal before sending it to the gain reduction stage.
4298 Normally a gate uses the full range signal to detect a level above the
4300 For example: If you cut all lower frequencies from your sidechain signal
4301 the gate will decrease the volume of your track only if not enough highs
4302 appear. With this technique you are able to reduce the resonation of a
4303 natural drum or remove "rumbling" of muted strokes from a heavily distorted
4305 It needs two input streams and returns one output stream.
4306 First input stream will be processed depending on second stream signal.
4308 The filter accepts the following options:
4312 Set input level before filtering.
4313 Default is 1. Allowed range is from 0.015625 to 64.
4316 Set the level of gain reduction when the signal is below the threshold.
4317 Default is 0.06125. Allowed range is from 0 to 1.
4320 If a signal rises above this level the gain reduction is released.
4321 Default is 0.125. Allowed range is from 0 to 1.
4324 Set a ratio about which the signal is reduced.
4325 Default is 2. Allowed range is from 1 to 9000.
4328 Amount of milliseconds the signal has to rise above the threshold before gain
4330 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
4333 Amount of milliseconds the signal has to fall below the threshold before the
4334 reduction is increased again. Default is 250 milliseconds.
4335 Allowed range is from 0.01 to 9000.
4338 Set amount of amplification of signal after processing.
4339 Default is 1. Allowed range is from 1 to 64.
4342 Curve the sharp knee around the threshold to enter gain reduction more softly.
4343 Default is 2.828427125. Allowed range is from 1 to 8.
4346 Choose if exact signal should be taken for detection or an RMS like one.
4347 Default is rms. Can be peak or rms.
4350 Choose if the average level between all channels or the louder channel affects
4352 Default is average. Can be average or maximum.
4355 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
4358 @section silencedetect
4360 Detect silence in an audio stream.
4362 This filter logs a message when it detects that the input audio volume is less
4363 or equal to a noise tolerance value for a duration greater or equal to the
4364 minimum detected noise duration.
4366 The printed times and duration are expressed in seconds.
4368 The filter accepts the following options:
4372 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
4373 specified value) or amplitude ratio. Default is -60dB, or 0.001.
4376 Set silence duration until notification (default is 2 seconds).
4379 Process each channel separately, instead of combined. By default is disabled.
4382 @subsection Examples
4386 Detect 5 seconds of silence with -50dB noise tolerance:
4388 silencedetect=n=-50dB:d=5
4392 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
4393 tolerance in @file{silence.mp3}:
4395 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
4399 @section silenceremove
4401 Remove silence from the beginning, middle or end of the audio.
4403 The filter accepts the following options:
4407 This value is used to indicate if audio should be trimmed at beginning of
4408 the audio. A value of zero indicates no silence should be trimmed from the
4409 beginning. When specifying a non-zero value, it trims audio up until it
4410 finds non-silence. Normally, when trimming silence from beginning of audio
4411 the @var{start_periods} will be @code{1} but it can be increased to higher
4412 values to trim all audio up to specific count of non-silence periods.
4413 Default value is @code{0}.
4415 @item start_duration
4416 Specify the amount of time that non-silence must be detected before it stops
4417 trimming audio. By increasing the duration, bursts of noises can be treated
4418 as silence and trimmed off. Default value is @code{0}.
4420 @item start_threshold
4421 This indicates what sample value should be treated as silence. For digital
4422 audio, a value of @code{0} may be fine but for audio recorded from analog,
4423 you may wish to increase the value to account for background noise.
4424 Can be specified in dB (in case "dB" is appended to the specified value)
4425 or amplitude ratio. Default value is @code{0}.
4428 Specify max duration of silence at beginning that will be kept after
4429 trimming. Default is 0, which is equal to trimming all samples detected
4433 Specify mode of detection of silence end in start of multi-channel audio.
4434 Can be @var{any} or @var{all}. Default is @var{any}.
4435 With @var{any}, any sample that is detected as non-silence will cause
4436 stopped trimming of silence.
4437 With @var{all}, only if all channels are detected as non-silence will cause
4438 stopped trimming of silence.
4441 Set the count for trimming silence from the end of audio.
4442 To remove silence from the middle of a file, specify a @var{stop_periods}
4443 that is negative. This value is then treated as a positive value and is
4444 used to indicate the effect should restart processing as specified by
4445 @var{start_periods}, making it suitable for removing periods of silence
4446 in the middle of the audio.
4447 Default value is @code{0}.
4450 Specify a duration of silence that must exist before audio is not copied any
4451 more. By specifying a higher duration, silence that is wanted can be left in
4453 Default value is @code{0}.
4455 @item stop_threshold
4456 This is the same as @option{start_threshold} but for trimming silence from
4458 Can be specified in dB (in case "dB" is appended to the specified value)
4459 or amplitude ratio. Default value is @code{0}.
4462 Specify max duration of silence at end that will be kept after
4463 trimming. Default is 0, which is equal to trimming all samples detected
4467 Specify mode of detection of silence start in end of multi-channel audio.
4468 Can be @var{any} or @var{all}. Default is @var{any}.
4469 With @var{any}, any sample that is detected as non-silence will cause
4470 stopped trimming of silence.
4471 With @var{all}, only if all channels are detected as non-silence will cause
4472 stopped trimming of silence.
4475 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
4476 and works better with digital silence which is exactly 0.
4477 Default value is @code{rms}.
4480 Set duration in number of seconds used to calculate size of window in number
4481 of samples for detecting silence.
4482 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
4485 @subsection Examples
4489 The following example shows how this filter can be used to start a recording
4490 that does not contain the delay at the start which usually occurs between
4491 pressing the record button and the start of the performance:
4493 silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
4497 Trim all silence encountered from beginning to end where there is more than 1
4498 second of silence in audio:
4500 silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
4506 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
4507 loudspeakers around the user for binaural listening via headphones (audio
4508 formats up to 9 channels supported).
4509 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
4510 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
4511 Austrian Academy of Sciences.
4513 To enable compilation of this filter you need to configure FFmpeg with
4514 @code{--enable-libmysofa}.
4516 The filter accepts the following options:
4520 Set the SOFA file used for rendering.
4523 Set gain applied to audio. Value is in dB. Default is 0.
4526 Set rotation of virtual loudspeakers in deg. Default is 0.
4529 Set elevation of virtual speakers in deg. Default is 0.
4532 Set distance in meters between loudspeakers and the listener with near-field
4533 HRTFs. Default is 1.
4536 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
4537 processing audio in time domain which is slow.
4538 @var{freq} is processing audio in frequency domain which is fast.
4539 Default is @var{freq}.
4542 Set custom positions of virtual loudspeakers. Syntax for this option is:
4543 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
4544 Each virtual loudspeaker is described with short channel name following with
4545 azimuth and elevation in degrees.
4546 Each virtual loudspeaker description is separated by '|'.
4547 For example to override front left and front right channel positions use:
4548 'speakers=FL 45 15|FR 345 15'.
4549 Descriptions with unrecognised channel names are ignored.
4552 Set custom gain for LFE channels. Value is in dB. Default is 0.
4555 Set custom frame size in number of samples. Default is 1024.
4556 Allowed range is from 1024 to 96000. Only used if option @samp{type}
4557 is set to @var{freq}.
4560 Should all IRs be normalized upon importing SOFA file.
4561 By default is enabled.
4564 Should nearest IRs be interpolated with neighbor IRs if exact position
4565 does not match. By default is disabled.
4568 Minphase all IRs upon loading of SOFA file. By default is disabled.
4571 Set neighbor search angle step. Only used if option @var{interpolate} is enabled.
4574 Set neighbor search radius step. Only used if option @var{interpolate} is enabled.
4577 @subsection Examples
4581 Using ClubFritz6 sofa file:
4583 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
4587 Using ClubFritz12 sofa file and bigger radius with small rotation:
4589 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
4593 Similar as above but with custom speaker positions for front left, front right, back left and back right
4594 and also with custom gain:
4596 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
4600 @section stereotools
4602 This filter has some handy utilities to manage stereo signals, for converting
4603 M/S stereo recordings to L/R signal while having control over the parameters
4604 or spreading the stereo image of master track.
4606 The filter accepts the following options:
4610 Set input level before filtering for both channels. Defaults is 1.
4611 Allowed range is from 0.015625 to 64.
4614 Set output level after filtering for both channels. Defaults is 1.
4615 Allowed range is from 0.015625 to 64.
4618 Set input balance between both channels. Default is 0.
4619 Allowed range is from -1 to 1.
4622 Set output balance between both channels. Default is 0.
4623 Allowed range is from -1 to 1.
4626 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
4627 clipping. Disabled by default.
4630 Mute the left channel. Disabled by default.
4633 Mute the right channel. Disabled by default.
4636 Change the phase of the left channel. Disabled by default.
4639 Change the phase of the right channel. Disabled by default.
4642 Set stereo mode. Available values are:
4646 Left/Right to Left/Right, this is default.
4649 Left/Right to Mid/Side.
4652 Mid/Side to Left/Right.
4655 Left/Right to Left/Left.
4658 Left/Right to Right/Right.
4661 Left/Right to Left + Right.
4664 Left/Right to Right/Left.
4667 Mid/Side to Left/Left.
4670 Mid/Side to Right/Right.
4674 Set level of side signal. Default is 1.
4675 Allowed range is from 0.015625 to 64.
4678 Set balance of side signal. Default is 0.
4679 Allowed range is from -1 to 1.
4682 Set level of the middle signal. Default is 1.
4683 Allowed range is from 0.015625 to 64.
4686 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
4689 Set stereo base between mono and inversed channels. Default is 0.
4690 Allowed range is from -1 to 1.
4693 Set delay in milliseconds how much to delay left from right channel and
4694 vice versa. Default is 0. Allowed range is from -20 to 20.
4697 Set S/C level. Default is 1. Allowed range is from 1 to 100.
4700 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
4702 @item bmode_in, bmode_out
4703 Set balance mode for balance_in/balance_out option.
4705 Can be one of the following:
4709 Classic balance mode. Attenuate one channel at time.
4710 Gain is raised up to 1.
4713 Similar as classic mode above but gain is raised up to 2.
4716 Equal power distribution, from -6dB to +6dB range.
4720 @subsection Examples
4724 Apply karaoke like effect:
4726 stereotools=mlev=0.015625
4730 Convert M/S signal to L/R:
4732 "stereotools=mode=ms>lr"
4736 @section stereowiden
4738 This filter enhance the stereo effect by suppressing signal common to both
4739 channels and by delaying the signal of left into right and vice versa,
4740 thereby widening the stereo effect.
4742 The filter accepts the following options:
4746 Time in milliseconds of the delay of left signal into right and vice versa.
4747 Default is 20 milliseconds.
4750 Amount of gain in delayed signal into right and vice versa. Gives a delay
4751 effect of left signal in right output and vice versa which gives widening
4752 effect. Default is 0.3.
4755 Cross feed of left into right with inverted phase. This helps in suppressing
4756 the mono. If the value is 1 it will cancel all the signal common to both
4757 channels. Default is 0.3.
4760 Set level of input signal of original channel. Default is 0.8.
4763 @section superequalizer
4764 Apply 18 band equalizer.
4766 The filter accepts the following options:
4773 Set 131Hz band gain.
4775 Set 185Hz band gain.
4777 Set 262Hz band gain.
4779 Set 370Hz band gain.
4781 Set 523Hz band gain.
4783 Set 740Hz band gain.
4785 Set 1047Hz band gain.
4787 Set 1480Hz band gain.
4789 Set 2093Hz band gain.
4791 Set 2960Hz band gain.
4793 Set 4186Hz band gain.
4795 Set 5920Hz band gain.
4797 Set 8372Hz band gain.
4799 Set 11840Hz band gain.
4801 Set 16744Hz band gain.
4803 Set 20000Hz band gain.
4807 Apply audio surround upmix filter.
4809 This filter allows to produce multichannel output from audio stream.
4811 The filter accepts the following options:
4815 Set output channel layout. By default, this is @var{5.1}.
4817 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4818 for the required syntax.
4821 Set input channel layout. By default, this is @var{stereo}.
4823 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4824 for the required syntax.
4827 Set input volume level. By default, this is @var{1}.
4830 Set output volume level. By default, this is @var{1}.
4833 Enable LFE channel output if output channel layout has it. By default, this is enabled.
4836 Set LFE low cut off frequency. By default, this is @var{128} Hz.
4839 Set LFE high cut off frequency. By default, this is @var{256} Hz.
4842 Set front center input volume. By default, this is @var{1}.
4845 Set front center output volume. By default, this is @var{1}.
4848 Set LFE input volume. By default, this is @var{1}.
4851 Set LFE output volume. By default, this is @var{1}.
4854 @section treble, highshelf
4856 Boost or cut treble (upper) frequencies of the audio using a two-pole
4857 shelving filter with a response similar to that of a standard
4858 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
4860 The filter accepts the following options:
4864 Give the gain at whichever is the lower of ~22 kHz and the
4865 Nyquist frequency. Its useful range is about -20 (for a large cut)
4866 to +20 (for a large boost). Beware of clipping when using a positive gain.
4869 Set the filter's central frequency and so can be used
4870 to extend or reduce the frequency range to be boosted or cut.
4871 The default value is @code{3000} Hz.
4874 Set method to specify band-width of filter.
4889 Determine how steep is the filter's shelf transition.
4892 Specify which channels to filter, by default all available are filtered.
4895 @subsection Commands
4897 This filter supports the following commands:
4900 Change treble frequency.
4901 Syntax for the command is : "@var{frequency}"
4904 Change treble width_type.
4905 Syntax for the command is : "@var{width_type}"
4908 Change treble width.
4909 Syntax for the command is : "@var{width}"
4913 Syntax for the command is : "@var{gain}"
4918 Sinusoidal amplitude modulation.
4920 The filter accepts the following options:
4924 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
4925 (20 Hz or lower) will result in a tremolo effect.
4926 This filter may also be used as a ring modulator by specifying
4927 a modulation frequency higher than 20 Hz.
4928 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4931 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4932 Default value is 0.5.
4937 Sinusoidal phase modulation.
4939 The filter accepts the following options:
4943 Modulation frequency in Hertz.
4944 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
4947 Depth of modulation as a percentage. Range is 0.0 - 1.0.
4948 Default value is 0.5.
4953 Adjust the input audio volume.
4955 It accepts the following parameters:
4959 Set audio volume expression.
4961 Output values are clipped to the maximum value.
4963 The output audio volume is given by the relation:
4965 @var{output_volume} = @var{volume} * @var{input_volume}
4968 The default value for @var{volume} is "1.0".
4971 This parameter represents the mathematical precision.
4973 It determines which input sample formats will be allowed, which affects the
4974 precision of the volume scaling.
4978 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
4980 32-bit floating-point; this limits input sample format to FLT. (default)
4982 64-bit floating-point; this limits input sample format to DBL.
4986 Choose the behaviour on encountering ReplayGain side data in input frames.
4990 Remove ReplayGain side data, ignoring its contents (the default).
4993 Ignore ReplayGain side data, but leave it in the frame.
4996 Prefer the track gain, if present.
4999 Prefer the album gain, if present.
5002 @item replaygain_preamp
5003 Pre-amplification gain in dB to apply to the selected replaygain gain.
5005 Default value for @var{replaygain_preamp} is 0.0.
5008 Set when the volume expression is evaluated.
5010 It accepts the following values:
5013 only evaluate expression once during the filter initialization, or
5014 when the @samp{volume} command is sent
5017 evaluate expression for each incoming frame
5020 Default value is @samp{once}.
5023 The volume expression can contain the following parameters.
5027 frame number (starting at zero)
5030 @item nb_consumed_samples
5031 number of samples consumed by the filter
5033 number of samples in the current frame
5035 original frame position in the file
5041 PTS at start of stream
5043 time at start of stream
5049 last set volume value
5052 Note that when @option{eval} is set to @samp{once} only the
5053 @var{sample_rate} and @var{tb} variables are available, all other
5054 variables will evaluate to NAN.
5056 @subsection Commands
5058 This filter supports the following commands:
5061 Modify the volume expression.
5062 The command accepts the same syntax of the corresponding option.
5064 If the specified expression is not valid, it is kept at its current
5066 @item replaygain_noclip
5067 Prevent clipping by limiting the gain applied.
5069 Default value for @var{replaygain_noclip} is 1.
5073 @subsection Examples
5077 Halve the input audio volume:
5081 volume=volume=-6.0206dB
5084 In all the above example the named key for @option{volume} can be
5085 omitted, for example like in:
5091 Increase input audio power by 6 decibels using fixed-point precision:
5093 volume=volume=6dB:precision=fixed
5097 Fade volume after time 10 with an annihilation period of 5 seconds:
5099 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
5103 @section volumedetect
5105 Detect the volume of the input video.
5107 The filter has no parameters. The input is not modified. Statistics about
5108 the volume will be printed in the log when the input stream end is reached.
5110 In particular it will show the mean volume (root mean square), maximum
5111 volume (on a per-sample basis), and the beginning of a histogram of the
5112 registered volume values (from the maximum value to a cumulated 1/1000 of
5115 All volumes are in decibels relative to the maximum PCM value.
5117 @subsection Examples
5119 Here is an excerpt of the output:
5121 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
5122 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
5123 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
5124 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
5125 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
5126 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
5127 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
5128 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
5129 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
5135 The mean square energy is approximately -27 dB, or 10^-2.7.
5137 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
5139 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
5142 In other words, raising the volume by +4 dB does not cause any clipping,
5143 raising it by +5 dB causes clipping for 6 samples, etc.
5145 @c man end AUDIO FILTERS
5147 @chapter Audio Sources
5148 @c man begin AUDIO SOURCES
5150 Below is a description of the currently available audio sources.
5154 Buffer audio frames, and make them available to the filter chain.
5156 This source is mainly intended for a programmatic use, in particular
5157 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
5159 It accepts the following parameters:
5163 The timebase which will be used for timestamps of submitted frames. It must be
5164 either a floating-point number or in @var{numerator}/@var{denominator} form.
5167 The sample rate of the incoming audio buffers.
5170 The sample format of the incoming audio buffers.
5171 Either a sample format name or its corresponding integer representation from
5172 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
5174 @item channel_layout
5175 The channel layout of the incoming audio buffers.
5176 Either a channel layout name from channel_layout_map in
5177 @file{libavutil/channel_layout.c} or its corresponding integer representation
5178 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
5181 The number of channels of the incoming audio buffers.
5182 If both @var{channels} and @var{channel_layout} are specified, then they
5187 @subsection Examples
5190 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
5193 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
5194 Since the sample format with name "s16p" corresponds to the number
5195 6 and the "stereo" channel layout corresponds to the value 0x3, this is
5198 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
5203 Generate an audio signal specified by an expression.
5205 This source accepts in input one or more expressions (one for each
5206 channel), which are evaluated and used to generate a corresponding
5209 This source accepts the following options:
5213 Set the '|'-separated expressions list for each separate channel. In case the
5214 @option{channel_layout} option is not specified, the selected channel layout
5215 depends on the number of provided expressions. Otherwise the last
5216 specified expression is applied to the remaining output channels.
5218 @item channel_layout, c
5219 Set the channel layout. The number of channels in the specified layout
5220 must be equal to the number of specified expressions.
5223 Set the minimum duration of the sourced audio. See
5224 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
5225 for the accepted syntax.
5226 Note that the resulting duration may be greater than the specified
5227 duration, as the generated audio is always cut at the end of a
5230 If not specified, or the expressed duration is negative, the audio is
5231 supposed to be generated forever.
5234 Set the number of samples per channel per each output frame,
5237 @item sample_rate, s
5238 Specify the sample rate, default to 44100.
5241 Each expression in @var{exprs} can contain the following constants:
5245 number of the evaluated sample, starting from 0
5248 time of the evaluated sample expressed in seconds, starting from 0
5255 @subsection Examples
5265 Generate a sin signal with frequency of 440 Hz, set sample rate to
5268 aevalsrc="sin(440*2*PI*t):s=8000"
5272 Generate a two channels signal, specify the channel layout (Front
5273 Center + Back Center) explicitly:
5275 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
5279 Generate white noise:
5281 aevalsrc="-2+random(0)"
5285 Generate an amplitude modulated signal:
5287 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
5291 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
5293 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
5300 The null audio source, return unprocessed audio frames. It is mainly useful
5301 as a template and to be employed in analysis / debugging tools, or as
5302 the source for filters which ignore the input data (for example the sox
5305 This source accepts the following options:
5309 @item channel_layout, cl
5311 Specifies the channel layout, and can be either an integer or a string
5312 representing a channel layout. The default value of @var{channel_layout}
5315 Check the channel_layout_map definition in
5316 @file{libavutil/channel_layout.c} for the mapping between strings and
5317 channel layout values.
5319 @item sample_rate, r
5320 Specifies the sample rate, and defaults to 44100.
5323 Set the number of samples per requested frames.
5327 @subsection Examples
5331 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
5333 anullsrc=r=48000:cl=4
5337 Do the same operation with a more obvious syntax:
5339 anullsrc=r=48000:cl=mono
5343 All the parameters need to be explicitly defined.
5347 Synthesize a voice utterance using the libflite library.
5349 To enable compilation of this filter you need to configure FFmpeg with
5350 @code{--enable-libflite}.
5352 Note that versions of the flite library prior to 2.0 are not thread-safe.
5354 The filter accepts the following options:
5359 If set to 1, list the names of the available voices and exit
5360 immediately. Default value is 0.
5363 Set the maximum number of samples per frame. Default value is 512.
5366 Set the filename containing the text to speak.
5369 Set the text to speak.
5372 Set the voice to use for the speech synthesis. Default value is
5373 @code{kal}. See also the @var{list_voices} option.
5376 @subsection Examples
5380 Read from file @file{speech.txt}, and synthesize the text using the
5381 standard flite voice:
5383 flite=textfile=speech.txt
5387 Read the specified text selecting the @code{slt} voice:
5389 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5393 Input text to ffmpeg:
5395 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
5399 Make @file{ffplay} speak the specified text, using @code{flite} and
5400 the @code{lavfi} device:
5402 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
5406 For more information about libflite, check:
5407 @url{http://www.festvox.org/flite/}
5411 Generate a noise audio signal.
5413 The filter accepts the following options:
5416 @item sample_rate, r
5417 Specify the sample rate. Default value is 48000 Hz.
5420 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
5424 Specify the duration of the generated audio stream. Not specifying this option
5425 results in noise with an infinite length.
5427 @item color, colour, c
5428 Specify the color of noise. Available noise colors are white, pink, brown,
5429 blue and violet. Default color is white.
5432 Specify a value used to seed the PRNG.
5435 Set the number of samples per each output frame, default is 1024.
5438 @subsection Examples
5443 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
5445 anoisesrc=d=60:c=pink:r=44100:a=0.5
5451 Generate odd-tap Hilbert transform FIR coefficients.
5453 The resulting stream can be used with @ref{afir} filter for phase-shifting
5454 the signal by 90 degrees.
5456 This is used in many matrix coding schemes and for analytic signal generation.
5457 The process is often written as a multiplication by i (or j), the imaginary unit.
5459 The filter accepts the following options:
5463 @item sample_rate, s
5464 Set sample rate, default is 44100.
5467 Set length of FIR filter, default is 22051.
5470 Set number of samples per each frame.
5473 Set window function to be used when generating FIR coefficients.
5478 Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
5480 The resulting stream can be used with @ref{afir} filter for filtering the audio signal.
5482 The filter accepts the following options:
5485 @item sample_rate, r
5486 Set sample rate, default is 44100.
5489 Set number of samples per each frame. Default is 1024.
5492 Set high-pass frequency. Default is 0.
5495 Set low-pass frequency. Default is 0.
5496 If high-pass frequency is lower than low-pass frequency and low-pass frequency
5497 is higher than 0 then filter will create band-pass filter coefficients,
5498 otherwise band-reject filter coefficients.
5501 Set filter phase response. Default is 50. Allowed range is from 0 to 100.
5504 Set Kaiser window beta.
5507 Set stop-band attenuation. Default is 120dB, allowed range is from 40 to 180 dB.
5510 Enable rounding, by default is disabled.
5513 Set number of taps for high-pass filter.
5516 Set number of taps for low-pass filter.
5521 Generate an audio signal made of a sine wave with amplitude 1/8.
5523 The audio signal is bit-exact.
5525 The filter accepts the following options:
5530 Set the carrier frequency. Default is 440 Hz.
5532 @item beep_factor, b
5533 Enable a periodic beep every second with frequency @var{beep_factor} times
5534 the carrier frequency. Default is 0, meaning the beep is disabled.
5536 @item sample_rate, r
5537 Specify the sample rate, default is 44100.
5540 Specify the duration of the generated audio stream.
5542 @item samples_per_frame
5543 Set the number of samples per output frame.
5545 The expression can contain the following constants:
5549 The (sequential) number of the output audio frame, starting from 0.
5552 The PTS (Presentation TimeStamp) of the output audio frame,
5553 expressed in @var{TB} units.
5556 The PTS of the output audio frame, expressed in seconds.
5559 The timebase of the output audio frames.
5562 Default is @code{1024}.
5565 @subsection Examples
5570 Generate a simple 440 Hz sine wave:
5576 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
5580 sine=frequency=220:beep_factor=4:duration=5
5584 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
5587 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
5591 @c man end AUDIO SOURCES
5593 @chapter Audio Sinks
5594 @c man begin AUDIO SINKS
5596 Below is a description of the currently available audio sinks.
5598 @section abuffersink
5600 Buffer audio frames, and make them available to the end of filter chain.
5602 This sink is mainly intended for programmatic use, in particular
5603 through the interface defined in @file{libavfilter/buffersink.h}
5604 or the options system.
5606 It accepts a pointer to an AVABufferSinkContext structure, which
5607 defines the incoming buffers' formats, to be passed as the opaque
5608 parameter to @code{avfilter_init_filter} for initialization.
5611 Null audio sink; do absolutely nothing with the input audio. It is
5612 mainly useful as a template and for use in analysis / debugging
5615 @c man end AUDIO SINKS
5617 @chapter Video Filters
5618 @c man begin VIDEO FILTERS
5620 When you configure your FFmpeg build, you can disable any of the
5621 existing filters using @code{--disable-filters}.
5622 The configure output will show the video filters included in your
5625 Below is a description of the currently available video filters.
5627 @section alphaextract
5629 Extract the alpha component from the input as a grayscale video. This
5630 is especially useful with the @var{alphamerge} filter.
5634 Add or replace the alpha component of the primary input with the
5635 grayscale value of a second input. This is intended for use with
5636 @var{alphaextract} to allow the transmission or storage of frame
5637 sequences that have alpha in a format that doesn't support an alpha
5640 For example, to reconstruct full frames from a normal YUV-encoded video
5641 and a separate video created with @var{alphaextract}, you might use:
5643 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
5646 Since this filter is designed for reconstruction, it operates on frame
5647 sequences without considering timestamps, and terminates when either
5648 input reaches end of stream. This will cause problems if your encoding
5649 pipeline drops frames. If you're trying to apply an image as an
5650 overlay to a video stream, consider the @var{overlay} filter instead.
5654 Amplify differences between current pixel and pixels of adjacent frames in
5655 same pixel location.
5657 This filter accepts the following options:
5661 Set frame radius. Default is 2. Allowed range is from 1 to 63.
5662 For example radius of 3 will instruct filter to calculate average of 7 frames.
5665 Set factor to amplify difference. Default is 2. Allowed range is from 0 to 65535.
5668 Set threshold for difference amplification. Any difference greater or equal to
5669 this value will not alter source pixel. Default is 10.
5670 Allowed range is from 0 to 65535.
5673 Set tolerance for difference amplification. Any difference lower to
5674 this value will not alter source pixel. Default is 0.
5675 Allowed range is from 0 to 65535.
5678 Set lower limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5679 This option controls maximum possible value that will decrease source pixel value.
5682 Set high limit for changing source pixel. Default is 65535. Allowed range is from 0 to 65535.
5683 This option controls maximum possible value that will increase source pixel value.
5686 Set which planes to filter. Default is all. Allowed range is from 0 to 15.
5691 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
5692 and libavformat to work. On the other hand, it is limited to ASS (Advanced
5693 Substation Alpha) subtitles files.
5695 This filter accepts the following option in addition to the common options from
5696 the @ref{subtitles} filter:
5700 Set the shaping engine
5702 Available values are:
5705 The default libass shaping engine, which is the best available.
5707 Fast, font-agnostic shaper that can do only substitutions
5709 Slower shaper using OpenType for substitutions and positioning
5712 The default is @code{auto}.
5716 Apply an Adaptive Temporal Averaging Denoiser to the video input.
5718 The filter accepts the following options:
5722 Set threshold A for 1st plane. Default is 0.02.
5723 Valid range is 0 to 0.3.
5726 Set threshold B for 1st plane. Default is 0.04.
5727 Valid range is 0 to 5.
5730 Set threshold A for 2nd plane. Default is 0.02.
5731 Valid range is 0 to 0.3.
5734 Set threshold B for 2nd plane. Default is 0.04.
5735 Valid range is 0 to 5.
5738 Set threshold A for 3rd plane. Default is 0.02.
5739 Valid range is 0 to 0.3.
5742 Set threshold B for 3rd plane. Default is 0.04.
5743 Valid range is 0 to 5.
5745 Threshold A is designed to react on abrupt changes in the input signal and
5746 threshold B is designed to react on continuous changes in the input signal.
5749 Set number of frames filter will use for averaging. Default is 9. Must be odd
5750 number in range [5, 129].
5753 Set what planes of frame filter will use for averaging. Default is all.
5758 Apply average blur filter.
5760 The filter accepts the following options:
5764 Set horizontal radius size.
5767 Set which planes to filter. By default all planes are filtered.
5770 Set vertical radius size, if zero it will be same as @code{sizeX}.
5771 Default is @code{0}.
5776 Compute the bounding box for the non-black pixels in the input frame
5779 This filter computes the bounding box containing all the pixels with a
5780 luminance value greater than the minimum allowed value.
5781 The parameters describing the bounding box are printed on the filter
5784 The filter accepts the following option:
5788 Set the minimal luminance value. Default is @code{16}.
5791 @section bitplanenoise
5793 Show and measure bit plane noise.
5795 The filter accepts the following options:
5799 Set which plane to analyze. Default is @code{1}.
5802 Filter out noisy pixels from @code{bitplane} set above.
5803 Default is disabled.
5806 @section blackdetect
5808 Detect video intervals that are (almost) completely black. Can be
5809 useful to detect chapter transitions, commercials, or invalid
5810 recordings. Output lines contains the time for the start, end and
5811 duration of the detected black interval expressed in seconds.
5813 In order to display the output lines, you need to set the loglevel at
5814 least to the AV_LOG_INFO value.
5816 The filter accepts the following options:
5819 @item black_min_duration, d
5820 Set the minimum detected black duration expressed in seconds. It must
5821 be a non-negative floating point number.
5823 Default value is 2.0.
5825 @item picture_black_ratio_th, pic_th
5826 Set the threshold for considering a picture "black".
5827 Express the minimum value for the ratio:
5829 @var{nb_black_pixels} / @var{nb_pixels}
5832 for which a picture is considered black.
5833 Default value is 0.98.
5835 @item pixel_black_th, pix_th
5836 Set the threshold for considering a pixel "black".
5838 The threshold expresses the maximum pixel luminance value for which a
5839 pixel is considered "black". The provided value is scaled according to
5840 the following equation:
5842 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
5845 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
5846 the input video format, the range is [0-255] for YUV full-range
5847 formats and [16-235] for YUV non full-range formats.
5849 Default value is 0.10.
5852 The following example sets the maximum pixel threshold to the minimum
5853 value, and detects only black intervals of 2 or more seconds:
5855 blackdetect=d=2:pix_th=0.00
5860 Detect frames that are (almost) completely black. Can be useful to
5861 detect chapter transitions or commercials. Output lines consist of
5862 the frame number of the detected frame, the percentage of blackness,
5863 the position in the file if known or -1 and the timestamp in seconds.
5865 In order to display the output lines, you need to set the loglevel at
5866 least to the AV_LOG_INFO value.
5868 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
5869 The value represents the percentage of pixels in the picture that
5870 are below the threshold value.
5872 It accepts the following parameters:
5877 The percentage of the pixels that have to be below the threshold; it defaults to
5880 @item threshold, thresh
5881 The threshold below which a pixel value is considered black; it defaults to
5886 @section blend, tblend
5888 Blend two video frames into each other.
5890 The @code{blend} filter takes two input streams and outputs one
5891 stream, the first input is the "top" layer and second input is
5892 "bottom" layer. By default, the output terminates when the longest input terminates.
5894 The @code{tblend} (time blend) filter takes two consecutive frames
5895 from one single stream, and outputs the result obtained by blending
5896 the new frame on top of the old frame.
5898 A description of the accepted options follows.
5906 Set blend mode for specific pixel component or all pixel components in case
5907 of @var{all_mode}. Default value is @code{normal}.
5909 Available values for component modes are:
5951 Set blend opacity for specific pixel component or all pixel components in case
5952 of @var{all_opacity}. Only used in combination with pixel component blend modes.
5959 Set blend expression for specific pixel component or all pixel components in case
5960 of @var{all_expr}. Note that related mode options will be ignored if those are set.
5962 The expressions can use the following variables:
5966 The sequential number of the filtered frame, starting from @code{0}.
5970 the coordinates of the current sample
5974 the width and height of currently filtered plane
5978 Width and height scale for the plane being filtered. It is the
5979 ratio between the dimensions of the current plane to the luma plane,
5980 e.g. for a @code{yuv420p} frame, the values are @code{1,1} for
5981 the luma plane and @code{0.5,0.5} for the chroma planes.
5984 Time of the current frame, expressed in seconds.
5987 Value of pixel component at current location for first video frame (top layer).
5990 Value of pixel component at current location for second video frame (bottom layer).
5994 The @code{blend} filter also supports the @ref{framesync} options.
5996 @subsection Examples
6000 Apply transition from bottom layer to top layer in first 10 seconds:
6002 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
6006 Apply linear horizontal transition from top layer to bottom layer:
6008 blend=all_expr='A*(X/W)+B*(1-X/W)'
6012 Apply 1x1 checkerboard effect:
6014 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
6018 Apply uncover left effect:
6020 blend=all_expr='if(gte(N*SW+X,W),A,B)'
6024 Apply uncover down effect:
6026 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
6030 Apply uncover up-left effect:
6032 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
6036 Split diagonally video and shows top and bottom layer on each side:
6038 blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
6042 Display differences between the current and the previous frame:
6044 tblend=all_mode=grainextract
6050 Denoise frames using Block-Matching 3D algorithm.
6052 The filter accepts the following options.
6056 Set denoising strength. Default value is 1.
6057 Allowed range is from 0 to 999.9.
6058 The denoising algorithm is very sensitive to sigma, so adjust it
6059 according to the source.
6062 Set local patch size. This sets dimensions in 2D.
6065 Set sliding step for processing blocks. Default value is 4.
6066 Allowed range is from 1 to 64.
6067 Smaller values allows processing more reference blocks and is slower.
6070 Set maximal number of similar blocks for 3rd dimension. Default value is 1.
6071 When set to 1, no block matching is done. Larger values allows more blocks
6073 Allowed range is from 1 to 256.
6076 Set radius for search block matching. Default is 9.
6077 Allowed range is from 1 to INT32_MAX.
6080 Set step between two search locations for block matching. Default is 1.
6081 Allowed range is from 1 to 64. Smaller is slower.
6084 Set threshold of mean square error for block matching. Valid range is 0 to
6088 Set thresholding parameter for hard thresholding in 3D transformed domain.
6089 Larger values results in stronger hard-thresholding filtering in frequency
6093 Set filtering estimation mode. Can be @code{basic} or @code{final}.
6094 Default is @code{basic}.
6097 If enabled, filter will use 2nd stream for block matching.
6098 Default is disabled for @code{basic} value of @var{estim} option,
6099 and always enabled if value of @var{estim} is @code{final}.
6102 Set planes to filter. Default is all available except alpha.
6105 @subsection Examples
6109 Basic filtering with bm3d:
6111 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
6115 Same as above, but filtering only luma:
6117 bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
6121 Same as above, but with both estimation modes:
6123 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
6127 Same as above, but prefilter with @ref{nlmeans} filter instead:
6129 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
6135 Apply a boxblur algorithm to the input video.
6137 It accepts the following parameters:
6141 @item luma_radius, lr
6142 @item luma_power, lp
6143 @item chroma_radius, cr
6144 @item chroma_power, cp
6145 @item alpha_radius, ar
6146 @item alpha_power, ap
6150 A description of the accepted options follows.
6153 @item luma_radius, lr
6154 @item chroma_radius, cr
6155 @item alpha_radius, ar
6156 Set an expression for the box radius in pixels used for blurring the
6157 corresponding input plane.
6159 The radius value must be a non-negative number, and must not be
6160 greater than the value of the expression @code{min(w,h)/2} for the
6161 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
6164 Default value for @option{luma_radius} is "2". If not specified,
6165 @option{chroma_radius} and @option{alpha_radius} default to the
6166 corresponding value set for @option{luma_radius}.
6168 The expressions can contain the following constants:
6172 The input width and height in pixels.
6176 The input chroma image width and height in pixels.
6180 The horizontal and vertical chroma subsample values. For example, for the
6181 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
6184 @item luma_power, lp
6185 @item chroma_power, cp
6186 @item alpha_power, ap
6187 Specify how many times the boxblur filter is applied to the
6188 corresponding plane.
6190 Default value for @option{luma_power} is 2. If not specified,
6191 @option{chroma_power} and @option{alpha_power} default to the
6192 corresponding value set for @option{luma_power}.
6194 A value of 0 will disable the effect.
6197 @subsection Examples
6201 Apply a boxblur filter with the luma, chroma, and alpha radii
6204 boxblur=luma_radius=2:luma_power=1
6209 Set the luma radius to 2, and alpha and chroma radius to 0:
6211 boxblur=2:1:cr=0:ar=0
6215 Set the luma and chroma radii to a fraction of the video dimension:
6217 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
6223 Deinterlace the input video ("bwdif" stands for "Bob Weaver
6224 Deinterlacing Filter").
6226 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
6227 interpolation algorithms.
6228 It accepts the following parameters:
6232 The interlacing mode to adopt. It accepts one of the following values:
6236 Output one frame for each frame.
6238 Output one frame for each field.
6241 The default value is @code{send_field}.
6244 The picture field parity assumed for the input interlaced video. It accepts one
6245 of the following values:
6249 Assume the top field is first.
6251 Assume the bottom field is first.
6253 Enable automatic detection of field parity.
6256 The default value is @code{auto}.
6257 If the interlacing is unknown or the decoder does not export this information,
6258 top field first will be assumed.
6261 Specify which frames to deinterlace. Accept one of the following
6266 Deinterlace all frames.
6268 Only deinterlace frames marked as interlaced.
6271 The default value is @code{all}.
6275 Remove all color information for all colors except for certain one.
6277 The filter accepts the following options:
6281 The color which will not be replaced with neutral chroma.
6284 Similarity percentage with the above color.
6285 0.01 matches only the exact key color, while 1.0 matches everything.
6288 Signals that the color passed is already in YUV instead of RGB.
6290 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6291 This can be used to pass exact YUV values as hexadecimal numbers.
6295 YUV colorspace color/chroma keying.
6297 The filter accepts the following options:
6301 The color which will be replaced with transparency.
6304 Similarity percentage with the key color.
6306 0.01 matches only the exact key color, while 1.0 matches everything.
6311 0.0 makes pixels either fully transparent, or not transparent at all.
6313 Higher values result in semi-transparent pixels, with a higher transparency
6314 the more similar the pixels color is to the key color.
6317 Signals that the color passed is already in YUV instead of RGB.
6319 Literal colors like "green" or "red" don't make sense with this enabled anymore.
6320 This can be used to pass exact YUV values as hexadecimal numbers.
6323 @subsection Examples
6327 Make every green pixel in the input image transparent:
6329 ffmpeg -i input.png -vf chromakey=green out.png
6333 Overlay a greenscreen-video on top of a static black background.
6335 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
6339 @section chromashift
6340 Shift chroma pixels horizontally and/or vertically.
6342 The filter accepts the following options:
6345 Set amount to shift chroma-blue horizontally.
6347 Set amount to shift chroma-blue vertically.
6349 Set amount to shift chroma-red horizontally.
6351 Set amount to shift chroma-red vertically.
6353 Set edge mode, can be @var{smear}, default, or @var{warp}.
6358 Display CIE color diagram with pixels overlaid onto it.
6360 The filter accepts the following options:
6375 @item uhdtv, rec2020
6388 Set what gamuts to draw.
6390 See @code{system} option for available values.
6393 Set ciescope size, by default set to 512.
6396 Set intensity used to map input pixel values to CIE diagram.
6399 Set contrast used to draw tongue colors that are out of active color system gamut.
6402 Correct gamma displayed on scope, by default enabled.
6405 Show white point on CIE diagram, by default disabled.
6408 Set input gamma. Used only with XYZ input color space.
6413 Visualize information exported by some codecs.
6415 Some codecs can export information through frames using side-data or other
6416 means. For example, some MPEG based codecs export motion vectors through the
6417 @var{export_mvs} flag in the codec @option{flags2} option.
6419 The filter accepts the following option:
6423 Set motion vectors to visualize.
6425 Available flags for @var{mv} are:
6429 forward predicted MVs of P-frames
6431 forward predicted MVs of B-frames
6433 backward predicted MVs of B-frames
6437 Display quantization parameters using the chroma planes.
6440 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
6442 Available flags for @var{mv_type} are:
6446 forward predicted MVs
6448 backward predicted MVs
6451 @item frame_type, ft
6452 Set frame type to visualize motion vectors of.
6454 Available flags for @var{frame_type} are:
6458 intra-coded frames (I-frames)
6460 predicted frames (P-frames)
6462 bi-directionally predicted frames (B-frames)
6466 @subsection Examples
6470 Visualize forward predicted MVs of all frames using @command{ffplay}:
6472 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
6476 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
6478 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
6482 @section colorbalance
6483 Modify intensity of primary colors (red, green and blue) of input frames.
6485 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
6486 regions for the red-cyan, green-magenta or blue-yellow balance.
6488 A positive adjustment value shifts the balance towards the primary color, a negative
6489 value towards the complementary color.
6491 The filter accepts the following options:
6497 Adjust red, green and blue shadows (darkest pixels).
6502 Adjust red, green and blue midtones (medium pixels).
6507 Adjust red, green and blue highlights (brightest pixels).
6509 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6512 @subsection Examples
6516 Add red color cast to shadows:
6523 RGB colorspace color keying.
6525 The filter accepts the following options:
6529 The color which will be replaced with transparency.
6532 Similarity percentage with the key color.
6534 0.01 matches only the exact key color, while 1.0 matches everything.
6539 0.0 makes pixels either fully transparent, or not transparent at all.
6541 Higher values result in semi-transparent pixels, with a higher transparency
6542 the more similar the pixels color is to the key color.
6545 @subsection Examples
6549 Make every green pixel in the input image transparent:
6551 ffmpeg -i input.png -vf colorkey=green out.png
6555 Overlay a greenscreen-video on top of a static background image.
6557 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
6561 @section colorlevels
6563 Adjust video input frames using levels.
6565 The filter accepts the following options:
6572 Adjust red, green, blue and alpha input black point.
6573 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
6579 Adjust red, green, blue and alpha input white point.
6580 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
6582 Input levels are used to lighten highlights (bright tones), darken shadows
6583 (dark tones), change the balance of bright and dark tones.
6589 Adjust red, green, blue and alpha output black point.
6590 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
6596 Adjust red, green, blue and alpha output white point.
6597 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
6599 Output levels allows manual selection of a constrained output level range.
6602 @subsection Examples
6606 Make video output darker:
6608 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
6614 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
6618 Make video output lighter:
6620 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
6624 Increase brightness:
6626 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
6630 @section colorchannelmixer
6632 Adjust video input frames by re-mixing color channels.
6634 This filter modifies a color channel by adding the values associated to
6635 the other channels of the same pixels. For example if the value to
6636 modify is red, the output value will be:
6638 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
6641 The filter accepts the following options:
6648 Adjust contribution of input red, green, blue and alpha channels for output red channel.
6649 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
6655 Adjust contribution of input red, green, blue and alpha channels for output green channel.
6656 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
6662 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
6663 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
6669 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
6670 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
6672 Allowed ranges for options are @code{[-2.0, 2.0]}.
6675 @subsection Examples
6679 Convert source to grayscale:
6681 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
6684 Simulate sepia tones:
6686 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
6690 @section colormatrix
6692 Convert color matrix.
6694 The filter accepts the following options:
6699 Specify the source and destination color matrix. Both values must be
6702 The accepted values are:
6730 For example to convert from BT.601 to SMPTE-240M, use the command:
6732 colormatrix=bt601:smpte240m
6737 Convert colorspace, transfer characteristics or color primaries.
6738 Input video needs to have an even size.
6740 The filter accepts the following options:
6745 Specify all color properties at once.
6747 The accepted values are:
6777 Specify output colorspace.
6779 The accepted values are:
6788 BT.470BG or BT.601-6 625
6791 SMPTE-170M or BT.601-6 525
6800 BT.2020 with non-constant luminance
6806 Specify output transfer characteristics.
6808 The accepted values are:
6820 Constant gamma of 2.2
6823 Constant gamma of 2.8
6826 SMPTE-170M, BT.601-6 625 or BT.601-6 525
6844 BT.2020 for 10-bits content
6847 BT.2020 for 12-bits content
6853 Specify output color primaries.
6855 The accepted values are:
6864 BT.470BG or BT.601-6 625
6867 SMPTE-170M or BT.601-6 525
6891 Specify output color range.
6893 The accepted values are:
6896 TV (restricted) range
6899 MPEG (restricted) range
6910 Specify output color format.
6912 The accepted values are:
6915 YUV 4:2:0 planar 8-bits
6918 YUV 4:2:0 planar 10-bits
6921 YUV 4:2:0 planar 12-bits
6924 YUV 4:2:2 planar 8-bits
6927 YUV 4:2:2 planar 10-bits
6930 YUV 4:2:2 planar 12-bits
6933 YUV 4:4:4 planar 8-bits
6936 YUV 4:4:4 planar 10-bits
6939 YUV 4:4:4 planar 12-bits
6944 Do a fast conversion, which skips gamma/primary correction. This will take
6945 significantly less CPU, but will be mathematically incorrect. To get output
6946 compatible with that produced by the colormatrix filter, use fast=1.
6949 Specify dithering mode.
6951 The accepted values are:
6957 Floyd-Steinberg dithering
6961 Whitepoint adaptation mode.
6963 The accepted values are:
6966 Bradford whitepoint adaptation
6969 von Kries whitepoint adaptation
6972 identity whitepoint adaptation (i.e. no whitepoint adaptation)
6976 Override all input properties at once. Same accepted values as @ref{all}.
6979 Override input colorspace. Same accepted values as @ref{space}.
6982 Override input color primaries. Same accepted values as @ref{primaries}.
6985 Override input transfer characteristics. Same accepted values as @ref{trc}.
6988 Override input color range. Same accepted values as @ref{range}.
6992 The filter converts the transfer characteristics, color space and color
6993 primaries to the specified user values. The output value, if not specified,
6994 is set to a default value based on the "all" property. If that property is
6995 also not specified, the filter will log an error. The output color range and
6996 format default to the same value as the input color range and format. The
6997 input transfer characteristics, color space, color primaries and color range
6998 should be set on the input data. If any of these are missing, the filter will
6999 log an error and no conversion will take place.
7001 For example to convert the input to SMPTE-240M, use the command:
7003 colorspace=smpte240m
7006 @section convolution
7008 Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
7010 The filter accepts the following options:
7017 Set matrix for each plane.
7018 Matrix is sequence of 9, 25 or 49 signed integers in @var{square} mode,
7019 and from 1 to 49 odd number of signed integers in @var{row} mode.
7025 Set multiplier for calculated value for each plane.
7026 If unset or 0, it will be sum of all matrix elements.
7032 Set bias for each plane. This value is added to the result of the multiplication.
7033 Useful for making the overall image brighter or darker. Default is 0.0.
7039 Set matrix mode for each plane. Can be @var{square}, @var{row} or @var{column}.
7040 Default is @var{square}.
7043 @subsection Examples
7049 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"
7055 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"
7061 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"
7067 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"
7071 Apply laplacian edge detector which includes diagonals:
7073 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"
7079 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"
7085 Apply 2D convolution of video stream in frequency domain using second stream
7088 The filter accepts the following options:
7092 Set which planes to process.
7095 Set which impulse video frames will be processed, can be @var{first}
7096 or @var{all}. Default is @var{all}.
7099 The @code{convolve} filter also supports the @ref{framesync} options.
7103 Copy the input video source unchanged to the output. This is mainly useful for
7108 Video filtering on GPU using Apple's CoreImage API on OSX.
7110 Hardware acceleration is based on an OpenGL context. Usually, this means it is
7111 processed by video hardware. However, software-based OpenGL implementations
7112 exist which means there is no guarantee for hardware processing. It depends on
7115 There are many filters and image generators provided by Apple that come with a
7116 large variety of options. The filter has to be referenced by its name along
7119 The coreimage filter accepts the following options:
7122 List all available filters and generators along with all their respective
7123 options as well as possible minimum and maximum values along with the default
7130 Specify all filters by their respective name and options.
7131 Use @var{list_filters} to determine all valid filter names and options.
7132 Numerical options are specified by a float value and are automatically clamped
7133 to their respective value range. Vector and color options have to be specified
7134 by a list of space separated float values. Character escaping has to be done.
7135 A special option name @code{default} is available to use default options for a
7138 It is required to specify either @code{default} or at least one of the filter options.
7139 All omitted options are used with their default values.
7140 The syntax of the filter string is as follows:
7142 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
7146 Specify a rectangle where the output of the filter chain is copied into the
7147 input image. It is given by a list of space separated float values:
7149 output_rect=x\ y\ width\ height
7151 If not given, the output rectangle equals the dimensions of the input image.
7152 The output rectangle is automatically cropped at the borders of the input
7153 image. Negative values are valid for each component.
7155 output_rect=25\ 25\ 100\ 100
7159 Several filters can be chained for successive processing without GPU-HOST
7160 transfers allowing for fast processing of complex filter chains.
7161 Currently, only filters with zero (generators) or exactly one (filters) input
7162 image and one output image are supported. Also, transition filters are not yet
7165 Some filters generate output images with additional padding depending on the
7166 respective filter kernel. The padding is automatically removed to ensure the
7167 filter output has the same size as the input image.
7169 For image generators, the size of the output image is determined by the
7170 previous output image of the filter chain or the input image of the whole
7171 filterchain, respectively. The generators do not use the pixel information of
7172 this image to generate their output. However, the generated output is
7173 blended onto this image, resulting in partial or complete coverage of the
7176 The @ref{coreimagesrc} video source can be used for generating input images
7177 which are directly fed into the filter chain. By using it, providing input
7178 images by another video source or an input video is not required.
7180 @subsection Examples
7185 List all filters available:
7187 coreimage=list_filters=true
7191 Use the CIBoxBlur filter with default options to blur an image:
7193 coreimage=filter=CIBoxBlur@@default
7197 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
7198 its center at 100x100 and a radius of 50 pixels:
7200 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
7204 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
7205 given as complete and escaped command-line for Apple's standard bash shell:
7207 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
7213 Crop the input video to given dimensions.
7215 It accepts the following parameters:
7219 The width of the output video. It defaults to @code{iw}.
7220 This expression is evaluated only once during the filter
7221 configuration, or when the @samp{w} or @samp{out_w} command is sent.
7224 The height of the output video. It defaults to @code{ih}.
7225 This expression is evaluated only once during the filter
7226 configuration, or when the @samp{h} or @samp{out_h} command is sent.
7229 The horizontal position, in the input video, of the left edge of the output
7230 video. It defaults to @code{(in_w-out_w)/2}.
7231 This expression is evaluated per-frame.
7234 The vertical position, in the input video, of the top edge of the output video.
7235 It defaults to @code{(in_h-out_h)/2}.
7236 This expression is evaluated per-frame.
7239 If set to 1 will force the output display aspect ratio
7240 to be the same of the input, by changing the output sample aspect
7241 ratio. It defaults to 0.
7244 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
7245 width/height/x/y as specified and will not be rounded to nearest smaller value.
7249 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
7250 expressions containing the following constants:
7255 The computed values for @var{x} and @var{y}. They are evaluated for
7260 The input width and height.
7264 These are the same as @var{in_w} and @var{in_h}.
7268 The output (cropped) width and height.
7272 These are the same as @var{out_w} and @var{out_h}.
7275 same as @var{iw} / @var{ih}
7278 input sample aspect ratio
7281 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
7285 horizontal and vertical chroma subsample values. For example for the
7286 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7289 The number of the input frame, starting from 0.
7292 the position in the file of the input frame, NAN if unknown
7295 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
7299 The expression for @var{out_w} may depend on the value of @var{out_h},
7300 and the expression for @var{out_h} may depend on @var{out_w}, but they
7301 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
7302 evaluated after @var{out_w} and @var{out_h}.
7304 The @var{x} and @var{y} parameters specify the expressions for the
7305 position of the top-left corner of the output (non-cropped) area. They
7306 are evaluated for each frame. If the evaluated value is not valid, it
7307 is approximated to the nearest valid value.
7309 The expression for @var{x} may depend on @var{y}, and the expression
7310 for @var{y} may depend on @var{x}.
7312 @subsection Examples
7316 Crop area with size 100x100 at position (12,34).
7321 Using named options, the example above becomes:
7323 crop=w=100:h=100:x=12:y=34
7327 Crop the central input area with size 100x100:
7333 Crop the central input area with size 2/3 of the input video:
7335 crop=2/3*in_w:2/3*in_h
7339 Crop the input video central square:
7346 Delimit the rectangle with the top-left corner placed at position
7347 100:100 and the right-bottom corner corresponding to the right-bottom
7348 corner of the input image.
7350 crop=in_w-100:in_h-100:100:100
7354 Crop 10 pixels from the left and right borders, and 20 pixels from
7355 the top and bottom borders
7357 crop=in_w-2*10:in_h-2*20
7361 Keep only the bottom right quarter of the input image:
7363 crop=in_w/2:in_h/2:in_w/2:in_h/2
7367 Crop height for getting Greek harmony:
7369 crop=in_w:1/PHI*in_w
7373 Apply trembling effect:
7375 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)
7379 Apply erratic camera effect depending on timestamp:
7381 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)"
7385 Set x depending on the value of y:
7387 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
7391 @subsection Commands
7393 This filter supports the following commands:
7399 Set width/height of the output video and the horizontal/vertical position
7401 The command accepts the same syntax of the corresponding option.
7403 If the specified expression is not valid, it is kept at its current
7409 Auto-detect the crop size.
7411 It calculates the necessary cropping parameters and prints the
7412 recommended parameters via the logging system. The detected dimensions
7413 correspond to the non-black area of the input video.
7415 It accepts the following parameters:
7420 Set higher black value threshold, which can be optionally specified
7421 from nothing (0) to everything (255 for 8-bit based formats). An intensity
7422 value greater to the set value is considered non-black. It defaults to 24.
7423 You can also specify a value between 0.0 and 1.0 which will be scaled depending
7424 on the bitdepth of the pixel format.
7427 The value which the width/height should be divisible by. It defaults to
7428 16. The offset is automatically adjusted to center the video. Use 2 to
7429 get only even dimensions (needed for 4:2:2 video). 16 is best when
7430 encoding to most video codecs.
7432 @item reset_count, reset
7433 Set the counter that determines after how many frames cropdetect will
7434 reset the previously detected largest video area and start over to
7435 detect the current optimal crop area. Default value is 0.
7437 This can be useful when channel logos distort the video area. 0
7438 indicates 'never reset', and returns the largest area encountered during
7445 Delay video filtering until a given wallclock timestamp. The filter first
7446 passes on @option{preroll} amount of frames, then it buffers at most
7447 @option{buffer} amount of frames and waits for the cue. After reaching the cue
7448 it forwards the buffered frames and also any subsequent frames coming in its
7451 The filter can be used synchronize the output of multiple ffmpeg processes for
7452 realtime output devices like decklink. By putting the delay in the filtering
7453 chain and pre-buffering frames the process can pass on data to output almost
7454 immediately after the target wallclock timestamp is reached.
7456 Perfect frame accuracy cannot be guaranteed, but the result is good enough for
7462 The cue timestamp expressed in a UNIX timestamp in microseconds. Default is 0.
7465 The duration of content to pass on as preroll expressed in seconds. Default is 0.
7468 The maximum duration of content to buffer before waiting for the cue expressed
7469 in seconds. Default is 0.
7476 Apply color adjustments using curves.
7478 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
7479 component (red, green and blue) has its values defined by @var{N} key points
7480 tied from each other using a smooth curve. The x-axis represents the pixel
7481 values from the input frame, and the y-axis the new pixel values to be set for
7484 By default, a component curve is defined by the two points @var{(0;0)} and
7485 @var{(1;1)}. This creates a straight line where each original pixel value is
7486 "adjusted" to its own value, which means no change to the image.
7488 The filter allows you to redefine these two points and add some more. A new
7489 curve (using a natural cubic spline interpolation) will be define to pass
7490 smoothly through all these new coordinates. The new defined points needs to be
7491 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
7492 be in the @var{[0;1]} interval. If the computed curves happened to go outside
7493 the vector spaces, the values will be clipped accordingly.
7495 The filter accepts the following options:
7499 Select one of the available color presets. This option can be used in addition
7500 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
7501 options takes priority on the preset values.
7502 Available presets are:
7505 @item color_negative
7508 @item increase_contrast
7510 @item linear_contrast
7511 @item medium_contrast
7513 @item strong_contrast
7516 Default is @code{none}.
7518 Set the master key points. These points will define a second pass mapping. It
7519 is sometimes called a "luminance" or "value" mapping. It can be used with
7520 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
7521 post-processing LUT.
7523 Set the key points for the red component.
7525 Set the key points for the green component.
7527 Set the key points for the blue component.
7529 Set the key points for all components (not including master).
7530 Can be used in addition to the other key points component
7531 options. In this case, the unset component(s) will fallback on this
7532 @option{all} setting.
7534 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
7536 Save Gnuplot script of the curves in specified file.
7539 To avoid some filtergraph syntax conflicts, each key points list need to be
7540 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
7542 @subsection Examples
7546 Increase slightly the middle level of blue:
7548 curves=blue='0/0 0.5/0.58 1/1'
7554 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'
7556 Here we obtain the following coordinates for each components:
7559 @code{(0;0.11) (0.42;0.51) (1;0.95)}
7561 @code{(0;0) (0.50;0.48) (1;1)}
7563 @code{(0;0.22) (0.49;0.44) (1;0.80)}
7567 The previous example can also be achieved with the associated built-in preset:
7569 curves=preset=vintage
7579 Use a Photoshop preset and redefine the points of the green component:
7581 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
7585 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
7586 and @command{gnuplot}:
7588 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
7589 gnuplot -p /tmp/curves.plt
7595 Video data analysis filter.
7597 This filter shows hexadecimal pixel values of part of video.
7599 The filter accepts the following options:
7603 Set output video size.
7606 Set x offset from where to pick pixels.
7609 Set y offset from where to pick pixels.
7612 Set scope mode, can be one of the following:
7615 Draw hexadecimal pixel values with white color on black background.
7618 Draw hexadecimal pixel values with input video pixel color on black
7622 Draw hexadecimal pixel values on color background picked from input video,
7623 the text color is picked in such way so its always visible.
7627 Draw rows and columns numbers on left and top of video.
7630 Set background opacity.
7635 Denoise frames using 2D DCT (frequency domain filtering).
7637 This filter is not designed for real time.
7639 The filter accepts the following options:
7643 Set the noise sigma constant.
7645 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
7646 coefficient (absolute value) below this threshold with be dropped.
7648 If you need a more advanced filtering, see @option{expr}.
7650 Default is @code{0}.
7653 Set number overlapping pixels for each block. Since the filter can be slow, you
7654 may want to reduce this value, at the cost of a less effective filter and the
7655 risk of various artefacts.
7657 If the overlapping value doesn't permit processing the whole input width or
7658 height, a warning will be displayed and according borders won't be denoised.
7660 Default value is @var{blocksize}-1, which is the best possible setting.
7663 Set the coefficient factor expression.
7665 For each coefficient of a DCT block, this expression will be evaluated as a
7666 multiplier value for the coefficient.
7668 If this is option is set, the @option{sigma} option will be ignored.
7670 The absolute value of the coefficient can be accessed through the @var{c}
7674 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
7675 @var{blocksize}, which is the width and height of the processed blocks.
7677 The default value is @var{3} (8x8) and can be raised to @var{4} for a
7678 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
7679 on the speed processing. Also, a larger block size does not necessarily means a
7683 @subsection Examples
7685 Apply a denoise with a @option{sigma} of @code{4.5}:
7690 The same operation can be achieved using the expression system:
7692 dctdnoiz=e='gte(c, 4.5*3)'
7695 Violent denoise using a block size of @code{16x16}:
7702 Remove banding artifacts from input video.
7703 It works by replacing banded pixels with average value of referenced pixels.
7705 The filter accepts the following options:
7712 Set banding detection threshold for each plane. Default is 0.02.
7713 Valid range is 0.00003 to 0.5.
7714 If difference between current pixel and reference pixel is less than threshold,
7715 it will be considered as banded.
7718 Banding detection range in pixels. Default is 16. If positive, random number
7719 in range 0 to set value will be used. If negative, exact absolute value
7721 The range defines square of four pixels around current pixel.
7724 Set direction in radians from which four pixel will be compared. If positive,
7725 random direction from 0 to set direction will be picked. If negative, exact of
7726 absolute value will be picked. For example direction 0, -PI or -2*PI radians
7727 will pick only pixels on same row and -PI/2 will pick only pixels on same
7731 If enabled, current pixel is compared with average value of all four
7732 surrounding pixels. The default is enabled. If disabled current pixel is
7733 compared with all four surrounding pixels. The pixel is considered banded
7734 if only all four differences with surrounding pixels are less than threshold.
7737 If enabled, current pixel is changed if and only if all pixel components are banded,
7738 e.g. banding detection threshold is triggered for all color components.
7739 The default is disabled.
7744 Remove blocking artifacts from input video.
7746 The filter accepts the following options:
7750 Set filter type, can be @var{weak} or @var{strong}. Default is @var{strong}.
7751 This controls what kind of deblocking is applied.
7754 Set size of block, allowed range is from 4 to 512. Default is @var{8}.
7760 Set blocking detection thresholds. Allowed range is 0 to 1.
7761 Defaults are: @var{0.098} for @var{alpha} and @var{0.05} for the rest.
7762 Using higher threshold gives more deblocking strength.
7763 Setting @var{alpha} controls threshold detection at exact edge of block.
7764 Remaining options controls threshold detection near the edge. Each one for
7765 below/above or left/right. Setting any of those to @var{0} disables
7769 Set planes to filter. Default is to filter all available planes.
7772 @subsection Examples
7776 Deblock using weak filter and block size of 4 pixels.
7778 deblock=filter=weak:block=4
7782 Deblock using strong filter, block size of 4 pixels and custom thresholds for
7783 deblocking more edges.
7785 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
7789 Similar as above, but filter only first plane.
7791 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
7795 Similar as above, but filter only second and third plane.
7797 deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
7804 Drop duplicated frames at regular intervals.
7806 The filter accepts the following options:
7810 Set the number of frames from which one will be dropped. Setting this to
7811 @var{N} means one frame in every batch of @var{N} frames will be dropped.
7812 Default is @code{5}.
7815 Set the threshold for duplicate detection. If the difference metric for a frame
7816 is less than or equal to this value, then it is declared as duplicate. Default
7820 Set scene change threshold. Default is @code{15}.
7824 Set the size of the x and y-axis blocks used during metric calculations.
7825 Larger blocks give better noise suppression, but also give worse detection of
7826 small movements. Must be a power of two. Default is @code{32}.
7829 Mark main input as a pre-processed input and activate clean source input
7830 stream. This allows the input to be pre-processed with various filters to help
7831 the metrics calculation while keeping the frame selection lossless. When set to
7832 @code{1}, the first stream is for the pre-processed input, and the second
7833 stream is the clean source from where the kept frames are chosen. Default is
7837 Set whether or not chroma is considered in the metric calculations. Default is
7843 Apply 2D deconvolution of video stream in frequency domain using second stream
7846 The filter accepts the following options:
7850 Set which planes to process.
7853 Set which impulse video frames will be processed, can be @var{first}
7854 or @var{all}. Default is @var{all}.
7857 Set noise when doing divisions. Default is @var{0.0000001}. Useful when width
7858 and height are not same and not power of 2 or if stream prior to convolving
7862 The @code{deconvolve} filter also supports the @ref{framesync} options.
7866 Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
7868 It accepts the following options:
7872 Set mode of operation. Can be combination of @var{dotcrawl} for cross-luminance reduction and/or
7873 @var{rainbows} for cross-color reduction.
7876 Set spatial luma threshold. Lower values increases reduction of cross-luminance.
7879 Set tolerance for temporal luma. Higher values increases reduction of cross-luminance.
7882 Set tolerance for chroma temporal variation. Higher values increases reduction of cross-color.
7885 Set temporal chroma threshold. Lower values increases reduction of cross-color.
7890 Apply deflate effect to the video.
7892 This filter replaces the pixel by the local(3x3) average by taking into account
7893 only values lower than the pixel.
7895 It accepts the following options:
7902 Limit the maximum change for each plane, default is 65535.
7903 If 0, plane will remain unchanged.
7908 Remove temporal frame luminance variations.
7910 It accepts the following options:
7914 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
7917 Set averaging mode to smooth temporal luminance variations.
7919 Available values are:
7944 Do not actually modify frame. Useful when one only wants metadata.
7949 Remove judder produced by partially interlaced telecined content.
7951 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
7952 source was partially telecined content then the output of @code{pullup,dejudder}
7953 will have a variable frame rate. May change the recorded frame rate of the
7954 container. Aside from that change, this filter will not affect constant frame
7957 The option available in this filter is:
7961 Specify the length of the window over which the judder repeats.
7963 Accepts any integer greater than 1. Useful values are:
7967 If the original was telecined from 24 to 30 fps (Film to NTSC).
7970 If the original was telecined from 25 to 30 fps (PAL to NTSC).
7973 If a mixture of the two.
7976 The default is @samp{4}.
7981 Suppress a TV station logo by a simple interpolation of the surrounding
7982 pixels. Just set a rectangle covering the logo and watch it disappear
7983 (and sometimes something even uglier appear - your mileage may vary).
7985 It accepts the following parameters:
7990 Specify the top left corner coordinates of the logo. They must be
7995 Specify the width and height of the logo to clear. They must be
7999 Specify the thickness of the fuzzy edge of the rectangle (added to
8000 @var{w} and @var{h}). The default value is 1. This option is
8001 deprecated, setting higher values should no longer be necessary and
8005 When set to 1, a green rectangle is drawn on the screen to simplify
8006 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
8007 The default value is 0.
8009 The rectangle is drawn on the outermost pixels which will be (partly)
8010 replaced with interpolated values. The values of the next pixels
8011 immediately outside this rectangle in each direction will be used to
8012 compute the interpolated pixel values inside the rectangle.
8016 @subsection Examples
8020 Set a rectangle covering the area with top left corner coordinates 0,0
8021 and size 100x77, and a band of size 10:
8023 delogo=x=0:y=0:w=100:h=77:band=10
8030 Attempt to fix small changes in horizontal and/or vertical shift. This
8031 filter helps remove camera shake from hand-holding a camera, bumping a
8032 tripod, moving on a vehicle, etc.
8034 The filter accepts the following options:
8042 Specify a rectangular area where to limit the search for motion
8044 If desired the search for motion vectors can be limited to a
8045 rectangular area of the frame defined by its top left corner, width
8046 and height. These parameters have the same meaning as the drawbox
8047 filter which can be used to visualise the position of the bounding
8050 This is useful when simultaneous movement of subjects within the frame
8051 might be confused for camera motion by the motion vector search.
8053 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
8054 then the full frame is used. This allows later options to be set
8055 without specifying the bounding box for the motion vector search.
8057 Default - search the whole frame.
8061 Specify the maximum extent of movement in x and y directions in the
8062 range 0-64 pixels. Default 16.
8065 Specify how to generate pixels to fill blanks at the edge of the
8066 frame. Available values are:
8069 Fill zeroes at blank locations
8071 Original image at blank locations
8073 Extruded edge value at blank locations
8075 Mirrored edge at blank locations
8077 Default value is @samp{mirror}.
8080 Specify the blocksize to use for motion search. Range 4-128 pixels,
8084 Specify the contrast threshold for blocks. Only blocks with more than
8085 the specified contrast (difference between darkest and lightest
8086 pixels) will be considered. Range 1-255, default 125.
8089 Specify the search strategy. Available values are:
8092 Set exhaustive search
8094 Set less exhaustive search.
8096 Default value is @samp{exhaustive}.
8099 If set then a detailed log of the motion search is written to the
8106 Remove unwanted contamination of foreground colors, caused by reflected color of
8107 greenscreen or bluescreen.
8109 This filter accepts the following options:
8113 Set what type of despill to use.
8116 Set how spillmap will be generated.
8119 Set how much to get rid of still remaining spill.
8122 Controls amount of red in spill area.
8125 Controls amount of green in spill area.
8126 Should be -1 for greenscreen.
8129 Controls amount of blue in spill area.
8130 Should be -1 for bluescreen.
8133 Controls brightness of spill area, preserving colors.
8136 Modify alpha from generated spillmap.
8141 Apply an exact inverse of the telecine operation. It requires a predefined
8142 pattern specified using the pattern option which must be the same as that passed
8143 to the telecine filter.
8145 This filter accepts the following options:
8154 The default value is @code{top}.
8158 A string of numbers representing the pulldown pattern you wish to apply.
8159 The default value is @code{23}.
8162 A number representing position of the first frame with respect to the telecine
8163 pattern. This is to be used if the stream is cut. The default value is @code{0}.
8168 Apply dilation effect to the video.
8170 This filter replaces the pixel by the local(3x3) maximum.
8172 It accepts the following options:
8179 Limit the maximum change for each plane, default is 65535.
8180 If 0, plane will remain unchanged.
8183 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
8186 Flags to local 3x3 coordinates maps like this:
8195 Displace pixels as indicated by second and third input stream.
8197 It takes three input streams and outputs one stream, the first input is the
8198 source, and second and third input are displacement maps.
8200 The second input specifies how much to displace pixels along the
8201 x-axis, while the third input specifies how much to displace pixels
8203 If one of displacement map streams terminates, last frame from that
8204 displacement map will be used.
8206 Note that once generated, displacements maps can be reused over and over again.
8208 A description of the accepted options follows.
8212 Set displace behavior for pixels that are out of range.
8214 Available values are:
8217 Missing pixels are replaced by black pixels.
8220 Adjacent pixels will spread out to replace missing pixels.
8223 Out of range pixels are wrapped so they point to pixels of other side.
8226 Out of range pixels will be replaced with mirrored pixels.
8228 Default is @samp{smear}.
8232 @subsection Examples
8236 Add ripple effect to rgb input of video size hd720:
8238 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
8242 Add wave effect to rgb input of video size hd720:
8244 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
8250 Draw a colored box on the input image.
8252 It accepts the following parameters:
8257 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
8261 The expressions which specify the width and height of the box; if 0 they are interpreted as
8262 the input width and height. It defaults to 0.
8265 Specify the color of the box to write. For the general syntax of this option,
8266 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8267 value @code{invert} is used, the box edge color is the same as the
8268 video with inverted luma.
8271 The expression which sets the thickness of the box edge.
8272 A value of @code{fill} will create a filled box. Default value is @code{3}.
8274 See below for the list of accepted constants.
8277 Applicable if the input has alpha. With value @code{1}, the pixels of the painted box
8278 will overwrite the video's color and alpha pixels.
8279 Default is @code{0}, which composites the box onto the input, leaving the video's alpha intact.
8282 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8283 following constants:
8287 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8291 horizontal and vertical chroma subsample values. For example for the
8292 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8296 The input width and height.
8299 The input sample aspect ratio.
8303 The x and y offset coordinates where the box is drawn.
8307 The width and height of the drawn box.
8310 The thickness of the drawn box.
8312 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8313 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8317 @subsection Examples
8321 Draw a black box around the edge of the input image:
8327 Draw a box with color red and an opacity of 50%:
8329 drawbox=10:20:200:60:red@@0.5
8332 The previous example can be specified as:
8334 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
8338 Fill the box with pink color:
8340 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=fill
8344 Draw a 2-pixel red 2.40:1 mask:
8346 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
8352 Draw a grid on the input image.
8354 It accepts the following parameters:
8359 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
8363 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
8364 input width and height, respectively, minus @code{thickness}, so image gets
8365 framed. Default to 0.
8368 Specify the color of the grid. For the general syntax of this option,
8369 check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}. If the special
8370 value @code{invert} is used, the grid color is the same as the
8371 video with inverted luma.
8374 The expression which sets the thickness of the grid line. Default value is @code{1}.
8376 See below for the list of accepted constants.
8379 Applicable if the input has alpha. With @code{1} the pixels of the painted grid
8380 will overwrite the video's color and alpha pixels.
8381 Default is @code{0}, which composites the grid onto the input, leaving the video's alpha intact.
8384 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
8385 following constants:
8389 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
8393 horizontal and vertical chroma subsample values. For example for the
8394 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8398 The input grid cell width and height.
8401 The input sample aspect ratio.
8405 The x and y coordinates of some point of grid intersection (meant to configure offset).
8409 The width and height of the drawn cell.
8412 The thickness of the drawn cell.
8414 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
8415 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
8419 @subsection Examples
8423 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
8425 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
8429 Draw a white 3x3 grid with an opacity of 50%:
8431 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
8438 Draw a text string or text from a specified file on top of a video, using the
8439 libfreetype library.
8441 To enable compilation of this filter, you need to configure FFmpeg with
8442 @code{--enable-libfreetype}.
8443 To enable default font fallback and the @var{font} option you need to
8444 configure FFmpeg with @code{--enable-libfontconfig}.
8445 To enable the @var{text_shaping} option, you need to configure FFmpeg with
8446 @code{--enable-libfribidi}.
8450 It accepts the following parameters:
8455 Used to draw a box around text using the background color.
8456 The value must be either 1 (enable) or 0 (disable).
8457 The default value of @var{box} is 0.
8460 Set the width of the border to be drawn around the box using @var{boxcolor}.
8461 The default value of @var{boxborderw} is 0.
8464 The color to be used for drawing box around text. For the syntax of this
8465 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8467 The default value of @var{boxcolor} is "white".
8470 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
8471 The default value of @var{line_spacing} is 0.
8474 Set the width of the border to be drawn around the text using @var{bordercolor}.
8475 The default value of @var{borderw} is 0.
8478 Set the color to be used for drawing border around text. For the syntax of this
8479 option, check the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8481 The default value of @var{bordercolor} is "black".
8484 Select how the @var{text} is expanded. Can be either @code{none},
8485 @code{strftime} (deprecated) or
8486 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
8490 Set a start time for the count. Value is in microseconds. Only applied
8491 in the deprecated strftime expansion mode. To emulate in normal expansion
8492 mode use the @code{pts} function, supplying the start time (in seconds)
8493 as the second argument.
8496 If true, check and fix text coords to avoid clipping.
8499 The color to be used for drawing fonts. For the syntax of this option, check
8500 the @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
8502 The default value of @var{fontcolor} is "black".
8504 @item fontcolor_expr
8505 String which is expanded the same way as @var{text} to obtain dynamic
8506 @var{fontcolor} value. By default this option has empty value and is not
8507 processed. When this option is set, it overrides @var{fontcolor} option.
8510 The font family to be used for drawing text. By default Sans.
8513 The font file to be used for drawing text. The path must be included.
8514 This parameter is mandatory if the fontconfig support is disabled.
8517 Draw the text applying alpha blending. The value can
8518 be a number between 0.0 and 1.0.
8519 The expression accepts the same variables @var{x, y} as well.
8520 The default value is 1.
8521 Please see @var{fontcolor_expr}.
8524 The font size to be used for drawing text.
8525 The default value of @var{fontsize} is 16.
8528 If set to 1, attempt to shape the text (for example, reverse the order of
8529 right-to-left text and join Arabic characters) before drawing it.
8530 Otherwise, just draw the text exactly as given.
8531 By default 1 (if supported).
8534 The flags to be used for loading the fonts.
8536 The flags map the corresponding flags supported by libfreetype, and are
8537 a combination of the following values:
8544 @item vertical_layout
8545 @item force_autohint
8548 @item ignore_global_advance_width
8550 @item ignore_transform
8556 Default value is "default".
8558 For more information consult the documentation for the FT_LOAD_*
8562 The color to be used for drawing a shadow behind the drawn text. For the
8563 syntax of this option, check the @ref{color syntax,,"Color" section in the
8564 ffmpeg-utils manual,ffmpeg-utils}.
8566 The default value of @var{shadowcolor} is "black".
8570 The x and y offsets for the text shadow position with respect to the
8571 position of the text. They can be either positive or negative
8572 values. The default value for both is "0".
8575 The starting frame number for the n/frame_num variable. The default value
8579 The size in number of spaces to use for rendering the tab.
8583 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
8584 format. It can be used with or without text parameter. @var{timecode_rate}
8585 option must be specified.
8587 @item timecode_rate, rate, r
8588 Set the timecode frame rate (timecode only). Value will be rounded to nearest
8589 integer. Minimum value is "1".
8590 Drop-frame timecode is supported for frame rates 30 & 60.
8593 If set to 1, the output of the timecode option will wrap around at 24 hours.
8594 Default is 0 (disabled).
8597 The text string to be drawn. The text must be a sequence of UTF-8
8599 This parameter is mandatory if no file is specified with the parameter
8603 A text file containing text to be drawn. The text must be a sequence
8604 of UTF-8 encoded characters.
8606 This parameter is mandatory if no text string is specified with the
8607 parameter @var{text}.
8609 If both @var{text} and @var{textfile} are specified, an error is thrown.
8612 If set to 1, the @var{textfile} will be reloaded before each frame.
8613 Be sure to update it atomically, or it may be read partially, or even fail.
8617 The expressions which specify the offsets where text will be drawn
8618 within the video frame. They are relative to the top/left border of the
8621 The default value of @var{x} and @var{y} is "0".
8623 See below for the list of accepted constants and functions.
8626 The parameters for @var{x} and @var{y} are expressions containing the
8627 following constants and functions:
8631 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
8635 horizontal and vertical chroma subsample values. For example for the
8636 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
8639 the height of each text line
8647 @item max_glyph_a, ascent
8648 the maximum distance from the baseline to the highest/upper grid
8649 coordinate used to place a glyph outline point, for all the rendered
8651 It is a positive value, due to the grid's orientation with the Y axis
8654 @item max_glyph_d, descent
8655 the maximum distance from the baseline to the lowest grid coordinate
8656 used to place a glyph outline point, for all the rendered glyphs.
8657 This is a negative value, due to the grid's orientation, with the Y axis
8661 maximum glyph height, that is the maximum height for all the glyphs
8662 contained in the rendered text, it is equivalent to @var{ascent} -
8666 maximum glyph width, that is the maximum width for all the glyphs
8667 contained in the rendered text
8670 the number of input frame, starting from 0
8672 @item rand(min, max)
8673 return a random number included between @var{min} and @var{max}
8676 The input sample aspect ratio.
8679 timestamp expressed in seconds, NAN if the input timestamp is unknown
8682 the height of the rendered text
8685 the width of the rendered text
8689 the x and y offset coordinates where the text is drawn.
8691 These parameters allow the @var{x} and @var{y} expressions to refer
8692 each other, so you can for example specify @code{y=x/dar}.
8695 @anchor{drawtext_expansion}
8696 @subsection Text expansion
8698 If @option{expansion} is set to @code{strftime},
8699 the filter recognizes strftime() sequences in the provided text and
8700 expands them accordingly. Check the documentation of strftime(). This
8701 feature is deprecated.
8703 If @option{expansion} is set to @code{none}, the text is printed verbatim.
8705 If @option{expansion} is set to @code{normal} (which is the default),
8706 the following expansion mechanism is used.
8708 The backslash character @samp{\}, followed by any character, always expands to
8709 the second character.
8711 Sequences of the form @code{%@{...@}} are expanded. The text between the
8712 braces is a function name, possibly followed by arguments separated by ':'.
8713 If the arguments contain special characters or delimiters (':' or '@}'),
8714 they should be escaped.
8716 Note that they probably must also be escaped as the value for the
8717 @option{text} option in the filter argument string and as the filter
8718 argument in the filtergraph description, and possibly also for the shell,
8719 that makes up to four levels of escaping; using a text file avoids these
8722 The following functions are available:
8727 The expression evaluation result.
8729 It must take one argument specifying the expression to be evaluated,
8730 which accepts the same constants and functions as the @var{x} and
8731 @var{y} values. Note that not all constants should be used, for
8732 example the text size is not known when evaluating the expression, so
8733 the constants @var{text_w} and @var{text_h} will have an undefined
8736 @item expr_int_format, eif
8737 Evaluate the expression's value and output as formatted integer.
8739 The first argument is the expression to be evaluated, just as for the @var{expr} function.
8740 The second argument specifies the output format. Allowed values are @samp{x},
8741 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
8742 @code{printf} function.
8743 The third parameter is optional and sets the number of positions taken by the output.
8744 It can be used to add padding with zeros from the left.
8747 The time at which the filter is running, expressed in UTC.
8748 It can accept an argument: a strftime() format string.
8751 The time at which the filter is running, expressed in the local time zone.
8752 It can accept an argument: a strftime() format string.
8755 Frame metadata. Takes one or two arguments.
8757 The first argument is mandatory and specifies the metadata key.
8759 The second argument is optional and specifies a default value, used when the
8760 metadata key is not found or empty.
8763 The frame number, starting from 0.
8766 A 1 character description of the current picture type.
8769 The timestamp of the current frame.
8770 It can take up to three arguments.
8772 The first argument is the format of the timestamp; it defaults to @code{flt}
8773 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
8774 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
8775 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
8776 @code{localtime} stands for the timestamp of the frame formatted as
8777 local time zone time.
8779 The second argument is an offset added to the timestamp.
8781 If the format is set to @code{hms}, a third argument @code{24HH} may be
8782 supplied to present the hour part of the formatted timestamp in 24h format
8785 If the format is set to @code{localtime} or @code{gmtime},
8786 a third argument may be supplied: a strftime() format string.
8787 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
8790 @subsection Examples
8794 Draw "Test Text" with font FreeSerif, using the default values for the
8795 optional parameters.
8798 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
8802 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
8803 and y=50 (counting from the top-left corner of the screen), text is
8804 yellow with a red box around it. Both the text and the box have an
8808 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
8809 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
8812 Note that the double quotes are not necessary if spaces are not used
8813 within the parameter list.
8816 Show the text at the center of the video frame:
8818 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
8822 Show the text at a random position, switching to a new position every 30 seconds:
8824 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)"
8828 Show a text line sliding from right to left in the last row of the video
8829 frame. The file @file{LONG_LINE} is assumed to contain a single line
8832 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
8836 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
8838 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
8842 Draw a single green letter "g", at the center of the input video.
8843 The glyph baseline is placed at half screen height.
8845 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
8849 Show text for 1 second every 3 seconds:
8851 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
8855 Use fontconfig to set the font. Note that the colons need to be escaped.
8857 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
8861 Print the date of a real-time encoding (see strftime(3)):
8863 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
8867 Show text fading in and out (appearing/disappearing):
8870 DS=1.0 # display start
8871 DE=10.0 # display end
8872 FID=1.5 # fade in duration
8873 FOD=5 # fade out duration
8874 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 @}"
8878 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
8879 and the @option{fontsize} value are included in the @option{y} offset.
8881 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
8882 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
8887 For more information about libfreetype, check:
8888 @url{http://www.freetype.org/}.
8890 For more information about fontconfig, check:
8891 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
8893 For more information about libfribidi, check:
8894 @url{http://fribidi.org/}.
8898 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
8900 The filter accepts the following options:
8905 Set low and high threshold values used by the Canny thresholding
8908 The high threshold selects the "strong" edge pixels, which are then
8909 connected through 8-connectivity with the "weak" edge pixels selected
8910 by the low threshold.
8912 @var{low} and @var{high} threshold values must be chosen in the range
8913 [0,1], and @var{low} should be lesser or equal to @var{high}.
8915 Default value for @var{low} is @code{20/255}, and default value for @var{high}
8919 Define the drawing mode.
8923 Draw white/gray wires on black background.
8926 Mix the colors to create a paint/cartoon effect.
8929 Apply Canny edge detector on all selected planes.
8931 Default value is @var{wires}.
8934 Select planes for filtering. By default all available planes are filtered.
8937 @subsection Examples
8941 Standard edge detection with custom values for the hysteresis thresholding:
8943 edgedetect=low=0.1:high=0.4
8947 Painting effect without thresholding:
8949 edgedetect=mode=colormix:high=0
8954 Set brightness, contrast, saturation and approximate gamma adjustment.
8956 The filter accepts the following options:
8960 Set the contrast expression. The value must be a float value in range
8961 @code{-2.0} to @code{2.0}. The default value is "1".
8964 Set the brightness expression. The value must be a float value in
8965 range @code{-1.0} to @code{1.0}. The default value is "0".
8968 Set the saturation expression. The value must be a float in
8969 range @code{0.0} to @code{3.0}. The default value is "1".
8972 Set the gamma expression. The value must be a float in range
8973 @code{0.1} to @code{10.0}. The default value is "1".
8976 Set the gamma expression for red. The value must be a float in
8977 range @code{0.1} to @code{10.0}. The default value is "1".
8980 Set the gamma expression for green. The value must be a float in range
8981 @code{0.1} to @code{10.0}. The default value is "1".
8984 Set the gamma expression for blue. The value must be a float in range
8985 @code{0.1} to @code{10.0}. The default value is "1".
8988 Set the gamma weight expression. It can be used to reduce the effect
8989 of a high gamma value on bright image areas, e.g. keep them from
8990 getting overamplified and just plain white. The value must be a float
8991 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
8992 gamma correction all the way down while @code{1.0} leaves it at its
8993 full strength. Default is "1".
8996 Set when the expressions for brightness, contrast, saturation and
8997 gamma expressions are evaluated.
8999 It accepts the following values:
9002 only evaluate expressions once during the filter initialization or
9003 when a command is processed
9006 evaluate expressions for each incoming frame
9009 Default value is @samp{init}.
9012 The expressions accept the following parameters:
9015 frame count of the input frame starting from 0
9018 byte position of the corresponding packet in the input file, NAN if
9022 frame rate of the input video, NAN if the input frame rate is unknown
9025 timestamp expressed in seconds, NAN if the input timestamp is unknown
9028 @subsection Commands
9029 The filter supports the following commands:
9033 Set the contrast expression.
9036 Set the brightness expression.
9039 Set the saturation expression.
9042 Set the gamma expression.
9045 Set the gamma_r expression.
9048 Set gamma_g expression.
9051 Set gamma_b expression.
9054 Set gamma_weight expression.
9056 The command accepts the same syntax of the corresponding option.
9058 If the specified expression is not valid, it is kept at its current
9065 Apply erosion effect to the video.
9067 This filter replaces the pixel by the local(3x3) minimum.
9069 It accepts the following options:
9076 Limit the maximum change for each plane, default is 65535.
9077 If 0, plane will remain unchanged.
9080 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
9083 Flags to local 3x3 coordinates maps like this:
9090 @section extractplanes
9092 Extract color channel components from input video stream into
9093 separate grayscale video streams.
9095 The filter accepts the following option:
9099 Set plane(s) to extract.
9101 Available values for planes are:
9112 Choosing planes not available in the input will result in an error.
9113 That means you cannot select @code{r}, @code{g}, @code{b} planes
9114 with @code{y}, @code{u}, @code{v} planes at same time.
9117 @subsection Examples
9121 Extract luma, u and v color channel component from input video frame
9122 into 3 grayscale outputs:
9124 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
9130 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
9132 For each input image, the filter will compute the optimal mapping from
9133 the input to the output given the codebook length, that is the number
9134 of distinct output colors.
9136 This filter accepts the following options.
9139 @item codebook_length, l
9140 Set codebook length. The value must be a positive integer, and
9141 represents the number of distinct output colors. Default value is 256.
9144 Set the maximum number of iterations to apply for computing the optimal
9145 mapping. The higher the value the better the result and the higher the
9146 computation time. Default value is 1.
9149 Set a random seed, must be an integer included between 0 and
9150 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
9151 will try to use a good random seed on a best effort basis.
9154 Set pal8 output pixel format. This option does not work with codebook
9155 length greater than 256.
9160 Measure graylevel entropy in histogram of color channels of video frames.
9162 It accepts the following parameters:
9166 Can be either @var{normal} or @var{diff}. Default is @var{normal}.
9168 @var{diff} mode measures entropy of histogram delta values, absolute differences
9169 between neighbour histogram values.
9174 Apply a fade-in/out effect to the input video.
9176 It accepts the following parameters:
9180 The effect type can be either "in" for a fade-in, or "out" for a fade-out
9182 Default is @code{in}.
9184 @item start_frame, s
9185 Specify the number of the frame to start applying the fade
9186 effect at. Default is 0.
9189 The number of frames that the fade effect lasts. At the end of the
9190 fade-in effect, the output video will have the same intensity as the input video.
9191 At the end of the fade-out transition, the output video will be filled with the
9192 selected @option{color}.
9196 If set to 1, fade only alpha channel, if one exists on the input.
9199 @item start_time, st
9200 Specify the timestamp (in seconds) of the frame to start to apply the fade
9201 effect. If both start_frame and start_time are specified, the fade will start at
9202 whichever comes last. Default is 0.
9205 The number of seconds for which the fade effect has to last. At the end of the
9206 fade-in effect the output video will have the same intensity as the input video,
9207 at the end of the fade-out transition the output video will be filled with the
9208 selected @option{color}.
9209 If both duration and nb_frames are specified, duration is used. Default is 0
9210 (nb_frames is used by default).
9213 Specify the color of the fade. Default is "black".
9216 @subsection Examples
9220 Fade in the first 30 frames of video:
9225 The command above is equivalent to:
9231 Fade out the last 45 frames of a 200-frame video:
9234 fade=type=out:start_frame=155:nb_frames=45
9238 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
9240 fade=in:0:25, fade=out:975:25
9244 Make the first 5 frames yellow, then fade in from frame 5-24:
9246 fade=in:5:20:color=yellow
9250 Fade in alpha over first 25 frames of video:
9252 fade=in:0:25:alpha=1
9256 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
9258 fade=t=in:st=5.5:d=0.5
9264 Apply arbitrary expressions to samples in frequency domain
9268 Adjust the dc value (gain) of the luma plane of the image. The filter
9269 accepts an integer value in range @code{0} to @code{1000}. The default
9270 value is set to @code{0}.
9273 Adjust the dc value (gain) of the 1st chroma plane of the image. The
9274 filter accepts an integer value in range @code{0} to @code{1000}. The
9275 default value is set to @code{0}.
9278 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
9279 filter accepts an integer value in range @code{0} to @code{1000}. The
9280 default value is set to @code{0}.
9283 Set the frequency domain weight expression for the luma plane.
9286 Set the frequency domain weight expression for the 1st chroma plane.
9289 Set the frequency domain weight expression for the 2nd chroma plane.
9292 Set when the expressions are evaluated.
9294 It accepts the following values:
9297 Only evaluate expressions once during the filter initialization.
9300 Evaluate expressions for each incoming frame.
9303 Default value is @samp{init}.
9305 The filter accepts the following variables:
9308 The coordinates of the current sample.
9312 The width and height of the image.
9315 The number of input frame, starting from 0.
9318 @subsection Examples
9324 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
9330 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
9336 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
9342 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
9348 Denoise frames using 3D FFT (frequency domain filtering).
9350 The filter accepts the following options:
9354 Set the noise sigma constant. This sets denoising strength.
9355 Default value is 1. Allowed range is from 0 to 30.
9356 Using very high sigma with low overlap may give blocking artifacts.
9359 Set amount of denoising. By default all detected noise is reduced.
9360 Default value is 1. Allowed range is from 0 to 1.
9363 Set size of block, Default is 4, can be 3, 4, 5 or 6.
9364 Actual size of block in pixels is 2 to power of @var{block}, so by default
9365 block size in pixels is 2^4 which is 16.
9368 Set block overlap. Default is 0.5. Allowed range is from 0.2 to 0.8.
9371 Set number of previous frames to use for denoising. By default is set to 0.
9374 Set number of next frames to to use for denoising. By default is set to 0.
9377 Set planes which will be filtered, by default are all available filtered
9383 Extract a single field from an interlaced image using stride
9384 arithmetic to avoid wasting CPU time. The output frames are marked as
9387 The filter accepts the following options:
9391 Specify whether to extract the top (if the value is @code{0} or
9392 @code{top}) or the bottom field (if the value is @code{1} or
9398 Create new frames by copying the top and bottom fields from surrounding frames
9399 supplied as numbers by the hint file.
9403 Set file containing hints: absolute/relative frame numbers.
9405 There must be one line for each frame in a clip. Each line must contain two
9406 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
9407 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
9408 is current frame number for @code{absolute} mode or out of [-1, 1] range
9409 for @code{relative} mode. First number tells from which frame to pick up top
9410 field and second number tells from which frame to pick up bottom field.
9412 If optionally followed by @code{+} output frame will be marked as interlaced,
9413 else if followed by @code{-} output frame will be marked as progressive, else
9414 it will be marked same as input frame.
9415 If line starts with @code{#} or @code{;} that line is skipped.
9418 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
9421 Example of first several lines of @code{hint} file for @code{relative} mode:
9424 1,0 - # second frame, use third's frame top field and second's frame bottom field
9425 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
9442 Field matching filter for inverse telecine. It is meant to reconstruct the
9443 progressive frames from a telecined stream. The filter does not drop duplicated
9444 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
9445 followed by a decimation filter such as @ref{decimate} in the filtergraph.
9447 The separation of the field matching and the decimation is notably motivated by
9448 the possibility of inserting a de-interlacing filter fallback between the two.
9449 If the source has mixed telecined and real interlaced content,
9450 @code{fieldmatch} will not be able to match fields for the interlaced parts.
9451 But these remaining combed frames will be marked as interlaced, and thus can be
9452 de-interlaced by a later filter such as @ref{yadif} before decimation.
9454 In addition to the various configuration options, @code{fieldmatch} can take an
9455 optional second stream, activated through the @option{ppsrc} option. If
9456 enabled, the frames reconstruction will be based on the fields and frames from
9457 this second stream. This allows the first input to be pre-processed in order to
9458 help the various algorithms of the filter, while keeping the output lossless
9459 (assuming the fields are matched properly). Typically, a field-aware denoiser,
9460 or brightness/contrast adjustments can help.
9462 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
9463 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
9464 which @code{fieldmatch} is based on. While the semantic and usage are very
9465 close, some behaviour and options names can differ.
9467 The @ref{decimate} filter currently only works for constant frame rate input.
9468 If your input has mixed telecined (30fps) and progressive content with a lower
9469 framerate like 24fps use the following filterchain to produce the necessary cfr
9470 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
9472 The filter accepts the following options:
9476 Specify the assumed field order of the input stream. Available values are:
9480 Auto detect parity (use FFmpeg's internal parity value).
9482 Assume bottom field first.
9484 Assume top field first.
9487 Note that it is sometimes recommended not to trust the parity announced by the
9490 Default value is @var{auto}.
9493 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
9494 sense that it won't risk creating jerkiness due to duplicate frames when
9495 possible, but if there are bad edits or blended fields it will end up
9496 outputting combed frames when a good match might actually exist. On the other
9497 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
9498 but will almost always find a good frame if there is one. The other values are
9499 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
9500 jerkiness and creating duplicate frames versus finding good matches in sections
9501 with bad edits, orphaned fields, blended fields, etc.
9503 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
9505 Available values are:
9509 2-way matching (p/c)
9511 2-way matching, and trying 3rd match if still combed (p/c + n)
9513 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
9515 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
9516 still combed (p/c + n + u/b)
9518 3-way matching (p/c/n)
9520 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
9521 detected as combed (p/c/n + u/b)
9524 The parenthesis at the end indicate the matches that would be used for that
9525 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
9528 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
9531 Default value is @var{pc_n}.
9534 Mark the main input stream as a pre-processed input, and enable the secondary
9535 input stream as the clean source to pick the fields from. See the filter
9536 introduction for more details. It is similar to the @option{clip2} feature from
9539 Default value is @code{0} (disabled).
9542 Set the field to match from. It is recommended to set this to the same value as
9543 @option{order} unless you experience matching failures with that setting. In
9544 certain circumstances changing the field that is used to match from can have a
9545 large impact on matching performance. Available values are:
9549 Automatic (same value as @option{order}).
9551 Match from the bottom field.
9553 Match from the top field.
9556 Default value is @var{auto}.
9559 Set whether or not chroma is included during the match comparisons. In most
9560 cases it is recommended to leave this enabled. You should set this to @code{0}
9561 only if your clip has bad chroma problems such as heavy rainbowing or other
9562 artifacts. Setting this to @code{0} could also be used to speed things up at
9563 the cost of some accuracy.
9565 Default value is @code{1}.
9569 These define an exclusion band which excludes the lines between @option{y0} and
9570 @option{y1} from being included in the field matching decision. An exclusion
9571 band can be used to ignore subtitles, a logo, or other things that may
9572 interfere with the matching. @option{y0} sets the starting scan line and
9573 @option{y1} sets the ending line; all lines in between @option{y0} and
9574 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
9575 @option{y0} and @option{y1} to the same value will disable the feature.
9576 @option{y0} and @option{y1} defaults to @code{0}.
9579 Set the scene change detection threshold as a percentage of maximum change on
9580 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
9581 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
9582 @option{scthresh} is @code{[0.0, 100.0]}.
9584 Default value is @code{12.0}.
9587 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
9588 account the combed scores of matches when deciding what match to use as the
9589 final match. Available values are:
9593 No final matching based on combed scores.
9595 Combed scores are only used when a scene change is detected.
9597 Use combed scores all the time.
9600 Default is @var{sc}.
9603 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
9604 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
9605 Available values are:
9609 No forced calculation.
9611 Force p/c/n calculations.
9613 Force p/c/n/u/b calculations.
9616 Default value is @var{none}.
9619 This is the area combing threshold used for combed frame detection. This
9620 essentially controls how "strong" or "visible" combing must be to be detected.
9621 Larger values mean combing must be more visible and smaller values mean combing
9622 can be less visible or strong and still be detected. Valid settings are from
9623 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
9624 be detected as combed). This is basically a pixel difference value. A good
9625 range is @code{[8, 12]}.
9627 Default value is @code{9}.
9630 Sets whether or not chroma is considered in the combed frame decision. Only
9631 disable this if your source has chroma problems (rainbowing, etc.) that are
9632 causing problems for the combed frame detection with chroma enabled. Actually,
9633 using @option{chroma}=@var{0} is usually more reliable, except for the case
9634 where there is chroma only combing in the source.
9636 Default value is @code{0}.
9640 Respectively set the x-axis and y-axis size of the window used during combed
9641 frame detection. This has to do with the size of the area in which
9642 @option{combpel} pixels are required to be detected as combed for a frame to be
9643 declared combed. See the @option{combpel} parameter description for more info.
9644 Possible values are any number that is a power of 2 starting at 4 and going up
9647 Default value is @code{16}.
9650 The number of combed pixels inside any of the @option{blocky} by
9651 @option{blockx} size blocks on the frame for the frame to be detected as
9652 combed. While @option{cthresh} controls how "visible" the combing must be, this
9653 setting controls "how much" combing there must be in any localized area (a
9654 window defined by the @option{blockx} and @option{blocky} settings) on the
9655 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
9656 which point no frames will ever be detected as combed). This setting is known
9657 as @option{MI} in TFM/VFM vocabulary.
9659 Default value is @code{80}.
9662 @anchor{p/c/n/u/b meaning}
9663 @subsection p/c/n/u/b meaning
9665 @subsubsection p/c/n
9667 We assume the following telecined stream:
9670 Top fields: 1 2 2 3 4
9671 Bottom fields: 1 2 3 4 4
9674 The numbers correspond to the progressive frame the fields relate to. Here, the
9675 first two frames are progressive, the 3rd and 4th are combed, and so on.
9677 When @code{fieldmatch} is configured to run a matching from bottom
9678 (@option{field}=@var{bottom}) this is how this input stream get transformed:
9683 B 1 2 3 4 4 <-- matching reference
9692 As a result of the field matching, we can see that some frames get duplicated.
9693 To perform a complete inverse telecine, you need to rely on a decimation filter
9694 after this operation. See for instance the @ref{decimate} filter.
9696 The same operation now matching from top fields (@option{field}=@var{top})
9701 T 1 2 2 3 4 <-- matching reference
9711 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
9712 basically, they refer to the frame and field of the opposite parity:
9715 @item @var{p} matches the field of the opposite parity in the previous frame
9716 @item @var{c} matches the field of the opposite parity in the current frame
9717 @item @var{n} matches the field of the opposite parity in the next frame
9722 The @var{u} and @var{b} matching are a bit special in the sense that they match
9723 from the opposite parity flag. In the following examples, we assume that we are
9724 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
9725 'x' is placed above and below each matched fields.
9727 With bottom matching (@option{field}=@var{bottom}):
9732 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9733 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9741 With top matching (@option{field}=@var{top}):
9746 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
9747 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
9755 @subsection Examples
9757 Simple IVTC of a top field first telecined stream:
9759 fieldmatch=order=tff:combmatch=none, decimate
9762 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
9764 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
9769 Transform the field order of the input video.
9771 It accepts the following parameters:
9776 The output field order. Valid values are @var{tff} for top field first or @var{bff}
9777 for bottom field first.
9780 The default value is @samp{tff}.
9782 The transformation is done by shifting the picture content up or down
9783 by one line, and filling the remaining line with appropriate picture content.
9784 This method is consistent with most broadcast field order converters.
9786 If the input video is not flagged as being interlaced, or it is already
9787 flagged as being of the required output field order, then this filter does
9788 not alter the incoming video.
9790 It is very useful when converting to or from PAL DV material,
9791 which is bottom field first.
9795 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
9798 @section fifo, afifo
9800 Buffer input images and send them when they are requested.
9802 It is mainly useful when auto-inserted by the libavfilter
9805 It does not take parameters.
9807 @section fillborders
9809 Fill borders of the input video, without changing video stream dimensions.
9810 Sometimes video can have garbage at the four edges and you may not want to
9811 crop video input to keep size multiple of some number.
9813 This filter accepts the following options:
9817 Number of pixels to fill from left border.
9820 Number of pixels to fill from right border.
9823 Number of pixels to fill from top border.
9826 Number of pixels to fill from bottom border.
9831 It accepts the following values:
9834 fill pixels using outermost pixels
9837 fill pixels using mirroring
9840 fill pixels with constant value
9843 Default is @var{smear}.
9846 Set color for pixels in fixed mode. Default is @var{black}.
9851 Find a rectangular object
9853 It accepts the following options:
9857 Filepath of the object image, needs to be in gray8.
9860 Detection threshold, default is 0.5.
9863 Number of mipmaps, default is 3.
9865 @item xmin, ymin, xmax, ymax
9866 Specifies the rectangle in which to search.
9869 @subsection Examples
9873 Generate a representative palette of a given video using @command{ffmpeg}:
9875 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9881 Cover a rectangular object
9883 It accepts the following options:
9887 Filepath of the optional cover image, needs to be in yuv420.
9892 It accepts the following values:
9895 cover it by the supplied image
9897 cover it by interpolating the surrounding pixels
9900 Default value is @var{blur}.
9903 @subsection Examples
9907 Generate a representative palette of a given video using @command{ffmpeg}:
9909 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
9915 Flood area with values of same pixel components with another values.
9917 It accepts the following options:
9920 Set pixel x coordinate.
9923 Set pixel y coordinate.
9926 Set source #0 component value.
9929 Set source #1 component value.
9932 Set source #2 component value.
9935 Set source #3 component value.
9938 Set destination #0 component value.
9941 Set destination #1 component value.
9944 Set destination #2 component value.
9947 Set destination #3 component value.
9953 Convert the input video to one of the specified pixel formats.
9954 Libavfilter will try to pick one that is suitable as input to
9957 It accepts the following parameters:
9961 A '|'-separated list of pixel format names, such as
9962 "pix_fmts=yuv420p|monow|rgb24".
9966 @subsection Examples
9970 Convert the input video to the @var{yuv420p} format
9972 format=pix_fmts=yuv420p
9975 Convert the input video to any of the formats in the list
9977 format=pix_fmts=yuv420p|yuv444p|yuv410p
9984 Convert the video to specified constant frame rate by duplicating or dropping
9985 frames as necessary.
9987 It accepts the following parameters:
9991 The desired output frame rate. The default is @code{25}.
9994 Assume the first PTS should be the given value, in seconds. This allows for
9995 padding/trimming at the start of stream. By default, no assumption is made
9996 about the first frame's expected PTS, so no padding or trimming is done.
9997 For example, this could be set to 0 to pad the beginning with duplicates of
9998 the first frame if a video stream starts after the audio stream or to trim any
9999 frames with a negative PTS.
10002 Timestamp (PTS) rounding method.
10004 Possible values are:
10011 round towards -infinity
10013 round towards +infinity
10017 The default is @code{near}.
10020 Action performed when reading the last frame.
10022 Possible values are:
10025 Use same timestamp rounding method as used for other frames.
10027 Pass through last frame if input duration has not been reached yet.
10029 The default is @code{round}.
10033 Alternatively, the options can be specified as a flat string:
10034 @var{fps}[:@var{start_time}[:@var{round}]].
10036 See also the @ref{setpts} filter.
10038 @subsection Examples
10042 A typical usage in order to set the fps to 25:
10048 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
10050 fps=fps=film:round=near
10056 Pack two different video streams into a stereoscopic video, setting proper
10057 metadata on supported codecs. The two views should have the same size and
10058 framerate and processing will stop when the shorter video ends. Please note
10059 that you may conveniently adjust view properties with the @ref{scale} and
10062 It accepts the following parameters:
10066 The desired packing format. Supported values are:
10071 The views are next to each other (default).
10074 The views are on top of each other.
10077 The views are packed by line.
10080 The views are packed by column.
10083 The views are temporally interleaved.
10092 # Convert left and right views into a frame-sequential video
10093 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
10095 # Convert views into a side-by-side video with the same output resolution as the input
10096 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
10101 Change the frame rate by interpolating new video output frames from the source
10104 This filter is not designed to function correctly with interlaced media. If
10105 you wish to change the frame rate of interlaced media then you are required
10106 to deinterlace before this filter and re-interlace after this filter.
10108 A description of the accepted options follows.
10112 Specify the output frames per second. This option can also be specified
10113 as a value alone. The default is @code{50}.
10116 Specify the start of a range where the output frame will be created as a
10117 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10118 the default is @code{15}.
10121 Specify the end of a range where the output frame will be created as a
10122 linear interpolation of two frames. The range is [@code{0}-@code{255}],
10123 the default is @code{240}.
10126 Specify the level at which a scene change is detected as a value between
10127 0 and 100 to indicate a new scene; a low value reflects a low
10128 probability for the current frame to introduce a new scene, while a higher
10129 value means the current frame is more likely to be one.
10130 The default is @code{8.2}.
10133 Specify flags influencing the filter process.
10135 Available value for @var{flags} is:
10138 @item scene_change_detect, scd
10139 Enable scene change detection using the value of the option @var{scene}.
10140 This flag is enabled by default.
10146 Select one frame every N-th frame.
10148 This filter accepts the following option:
10151 Select frame after every @code{step} frames.
10152 Allowed values are positive integers higher than 0. Default value is @code{1}.
10155 @section freezedetect
10157 Detect frozen video.
10159 This filter logs a message and sets frame metadata when it detects that the
10160 input video has no significant change in content during a specified duration.
10161 Video freeze detection calculates the mean average absolute difference of all
10162 the components of video frames and compares it to a noise floor.
10164 The printed times and duration are expressed in seconds. The
10165 @code{lavfi.freezedetect.freeze_start} metadata key is set on the first frame
10166 whose timestamp equals or exceeds the detection duration and it contains the
10167 timestamp of the first frame of the freeze. The
10168 @code{lavfi.freezedetect.freeze_duration} and
10169 @code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
10172 The filter accepts the following options:
10176 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
10177 specified value) or as a difference ratio between 0 and 1. Default is -60dB, or
10181 Set freeze duration until notification (default is 2 seconds).
10187 Apply a frei0r effect to the input video.
10189 To enable the compilation of this filter, you need to install the frei0r
10190 header and configure FFmpeg with @code{--enable-frei0r}.
10192 It accepts the following parameters:
10197 The name of the frei0r effect to load. If the environment variable
10198 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
10199 directories specified by the colon-separated list in @env{FREI0R_PATH}.
10200 Otherwise, the standard frei0r paths are searched, in this order:
10201 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
10202 @file{/usr/lib/frei0r-1/}.
10204 @item filter_params
10205 A '|'-separated list of parameters to pass to the frei0r effect.
10209 A frei0r effect parameter can be a boolean (its value is either
10210 "y" or "n"), a double, a color (specified as
10211 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
10212 numbers between 0.0 and 1.0, inclusive) or a color description as specified in the
10213 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils},
10214 a position (specified as @var{X}/@var{Y}, where
10215 @var{X} and @var{Y} are floating point numbers) and/or a string.
10217 The number and types of parameters depend on the loaded effect. If an
10218 effect parameter is not specified, the default value is set.
10220 @subsection Examples
10224 Apply the distort0r effect, setting the first two double parameters:
10226 frei0r=filter_name=distort0r:filter_params=0.5|0.01
10230 Apply the colordistance effect, taking a color as the first parameter:
10232 frei0r=colordistance:0.2/0.3/0.4
10233 frei0r=colordistance:violet
10234 frei0r=colordistance:0x112233
10238 Apply the perspective effect, specifying the top left and top right image
10241 frei0r=perspective:0.2/0.2|0.8/0.2
10245 For more information, see
10246 @url{http://frei0r.dyne.org}
10250 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
10252 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
10253 processing filter, one of them is performed once per block, not per pixel.
10254 This allows for much higher speed.
10256 The filter accepts the following options:
10260 Set quality. This option defines the number of levels for averaging. It accepts
10261 an integer in the range 4-5. Default value is @code{4}.
10264 Force a constant quantization parameter. It accepts an integer in range 0-63.
10265 If not set, the filter will use the QP from the video stream (if available).
10268 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
10269 more details but also more artifacts, while higher values make the image smoother
10270 but also blurrier. Default value is @code{0} − PSNR optimal.
10272 @item use_bframe_qp
10273 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
10274 option may cause flicker since the B-Frames have often larger QP. Default is
10275 @code{0} (not enabled).
10281 Apply Gaussian blur filter.
10283 The filter accepts the following options:
10287 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
10290 Set number of steps for Gaussian approximation. Default is @code{1}.
10293 Set which planes to filter. By default all planes are filtered.
10296 Set vertical sigma, if negative it will be same as @code{sigma}.
10297 Default is @code{-1}.
10302 Apply generic equation to each pixel.
10304 The filter accepts the following options:
10307 @item lum_expr, lum
10308 Set the luminance expression.
10310 Set the chrominance blue expression.
10312 Set the chrominance red expression.
10313 @item alpha_expr, a
10314 Set the alpha expression.
10316 Set the red expression.
10317 @item green_expr, g
10318 Set the green expression.
10320 Set the blue expression.
10323 The colorspace is selected according to the specified options. If one
10324 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
10325 options is specified, the filter will automatically select a YCbCr
10326 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
10327 @option{blue_expr} options is specified, it will select an RGB
10330 If one of the chrominance expression is not defined, it falls back on the other
10331 one. If no alpha expression is specified it will evaluate to opaque value.
10332 If none of chrominance expressions are specified, they will evaluate
10333 to the luminance expression.
10335 The expressions can use the following variables and functions:
10339 The sequential number of the filtered frame, starting from @code{0}.
10343 The coordinates of the current sample.
10347 The width and height of the image.
10351 Width and height scale depending on the currently filtered plane. It is the
10352 ratio between the corresponding luma plane number of pixels and the current
10353 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
10354 @code{0.5,0.5} for chroma planes.
10357 Time of the current frame, expressed in seconds.
10360 Return the value of the pixel at location (@var{x},@var{y}) of the current
10364 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
10368 Return the value of the pixel at location (@var{x},@var{y}) of the
10369 blue-difference chroma plane. Return 0 if there is no such plane.
10372 Return the value of the pixel at location (@var{x},@var{y}) of the
10373 red-difference chroma plane. Return 0 if there is no such plane.
10378 Return the value of the pixel at location (@var{x},@var{y}) of the
10379 red/green/blue component. Return 0 if there is no such component.
10382 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
10383 plane. Return 0 if there is no such plane.
10386 For functions, if @var{x} and @var{y} are outside the area, the value will be
10387 automatically clipped to the closer edge.
10389 @subsection Examples
10393 Flip the image horizontally:
10399 Generate a bidimensional sine wave, with angle @code{PI/3} and a
10400 wavelength of 100 pixels:
10402 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
10406 Generate a fancy enigmatic moving light:
10408 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
10412 Generate a quick emboss effect:
10414 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
10418 Modify RGB components depending on pixel position:
10420 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
10424 Create a radial gradient that is the same size as the input (also see
10425 the @ref{vignette} filter):
10427 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
10433 Fix the banding artifacts that are sometimes introduced into nearly flat
10434 regions by truncation to 8-bit color depth.
10435 Interpolate the gradients that should go where the bands are, and
10438 It is designed for playback only. Do not use it prior to
10439 lossy compression, because compression tends to lose the dither and
10440 bring back the bands.
10442 It accepts the following parameters:
10447 The maximum amount by which the filter will change any one pixel. This is also
10448 the threshold for detecting nearly flat regions. Acceptable values range from
10449 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
10453 The neighborhood to fit the gradient to. A larger radius makes for smoother
10454 gradients, but also prevents the filter from modifying the pixels near detailed
10455 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
10456 values will be clipped to the valid range.
10460 Alternatively, the options can be specified as a flat string:
10461 @var{strength}[:@var{radius}]
10463 @subsection Examples
10467 Apply the filter with a @code{3.5} strength and radius of @code{8}:
10473 Specify radius, omitting the strength (which will fall-back to the default
10481 @section graphmonitor, agraphmonitor
10482 Show various filtergraph stats.
10484 With this filter one can debug complete filtergraph.
10485 Especially issues with links filling with queued frames.
10487 The filter accepts the following options:
10491 Set video output size. Default is @var{hd720}.
10494 Set video opacity. Default is @var{0.9}. Allowed range is from @var{0} to @var{1}.
10497 Set output mode, can be @var{fulll} or @var{compact}.
10498 In @var{compact} mode only filters with some queued frames have displayed stats.
10501 Set flags which enable which stats are shown in video.
10503 Available values for flags are:
10506 Display number of queued frames in each link.
10508 @item frame_count_in
10509 Display number of frames taken from filter.
10511 @item frame_count_out
10512 Display number of frames given out from filter.
10515 Display current filtered frame pts.
10518 Display current filtered frame time.
10521 Display time base for filter link.
10524 Display used format for filter link.
10527 Display video size or number of audio channels in case of audio used by filter link.
10530 Display video frame rate or sample rate in case of audio used by filter link.
10534 Set upper limit for video rate of output stream, Default value is @var{25}.
10535 This guarantee that output video frame rate will not be higher than this value.
10539 A color constancy variation filter which estimates scene illumination via grey edge algorithm
10540 and corrects the scene colors accordingly.
10542 See: @url{https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf}
10544 The filter accepts the following options:
10548 The order of differentiation to be applied on the scene. Must be chosen in the range
10549 [0,2] and default value is 1.
10552 The Minkowski parameter to be used for calculating the Minkowski distance. Must
10553 be chosen in the range [0,20] and default value is 1. Set to 0 for getting
10554 max value instead of calculating Minkowski distance.
10557 The standard deviation of Gaussian blur to be applied on the scene. Must be
10558 chosen in the range [0,1024.0] and default value = 1. floor( @var{sigma} * break_off_sigma(3) )
10559 can't be equal to 0 if @var{difford} is greater than 0.
10562 @subsection Examples
10568 greyedge=difford=1:minknorm=5:sigma=2
10574 greyedge=difford=1:minknorm=0:sigma=2
10582 Apply a Hald CLUT to a video stream.
10584 First input is the video stream to process, and second one is the Hald CLUT.
10585 The Hald CLUT input can be a simple picture or a complete video stream.
10587 The filter accepts the following options:
10591 Force termination when the shortest input terminates. Default is @code{0}.
10593 Continue applying the last CLUT after the end of the stream. A value of
10594 @code{0} disable the filter after the last frame of the CLUT is reached.
10595 Default is @code{1}.
10598 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
10599 filters share the same internals).
10601 More information about the Hald CLUT can be found on Eskil Steenberg's website
10602 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
10604 @subsection Workflow examples
10606 @subsubsection Hald CLUT video stream
10608 Generate an identity Hald CLUT stream altered with various effects:
10610 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
10613 Note: make sure you use a lossless codec.
10615 Then use it with @code{haldclut} to apply it on some random stream:
10617 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
10620 The Hald CLUT will be applied to the 10 first seconds (duration of
10621 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
10622 to the remaining frames of the @code{mandelbrot} stream.
10624 @subsubsection Hald CLUT with preview
10626 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
10627 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
10628 biggest possible square starting at the top left of the picture. The remaining
10629 padding pixels (bottom or right) will be ignored. This area can be used to add
10630 a preview of the Hald CLUT.
10632 Typically, the following generated Hald CLUT will be supported by the
10633 @code{haldclut} filter:
10636 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
10637 pad=iw+320 [padded_clut];
10638 smptebars=s=320x256, split [a][b];
10639 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
10640 [main][b] overlay=W-320" -frames:v 1 clut.png
10643 It contains the original and a preview of the effect of the CLUT: SMPTE color
10644 bars are displayed on the right-top, and below the same color bars processed by
10647 Then, the effect of this Hald CLUT can be visualized with:
10649 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
10654 Flip the input video horizontally.
10656 For example, to horizontally flip the input video with @command{ffmpeg}:
10658 ffmpeg -i in.avi -vf "hflip" out.avi
10662 This filter applies a global color histogram equalization on a
10665 It can be used to correct video that has a compressed range of pixel
10666 intensities. The filter redistributes the pixel intensities to
10667 equalize their distribution across the intensity range. It may be
10668 viewed as an "automatically adjusting contrast filter". This filter is
10669 useful only for correcting degraded or poorly captured source
10672 The filter accepts the following options:
10676 Determine the amount of equalization to be applied. As the strength
10677 is reduced, the distribution of pixel intensities more-and-more
10678 approaches that of the input frame. The value must be a float number
10679 in the range [0,1] and defaults to 0.200.
10682 Set the maximum intensity that can generated and scale the output
10683 values appropriately. The strength should be set as desired and then
10684 the intensity can be limited if needed to avoid washing-out. The value
10685 must be a float number in the range [0,1] and defaults to 0.210.
10688 Set the antibanding level. If enabled the filter will randomly vary
10689 the luminance of output pixels by a small amount to avoid banding of
10690 the histogram. Possible values are @code{none}, @code{weak} or
10691 @code{strong}. It defaults to @code{none}.
10696 Compute and draw a color distribution histogram for the input video.
10698 The computed histogram is a representation of the color component
10699 distribution in an image.
10701 Standard histogram displays the color components distribution in an image.
10702 Displays color graph for each color component. Shows distribution of
10703 the Y, U, V, A or R, G, B components, depending on input format, in the
10704 current frame. Below each graph a color component scale meter is shown.
10706 The filter accepts the following options:
10710 Set height of level. Default value is @code{200}.
10711 Allowed range is [50, 2048].
10714 Set height of color scale. Default value is @code{12}.
10715 Allowed range is [0, 40].
10719 It accepts the following values:
10722 Per color component graphs are placed below each other.
10725 Per color component graphs are placed side by side.
10728 Presents information identical to that in the @code{parade}, except
10729 that the graphs representing color components are superimposed directly
10732 Default is @code{stack}.
10735 Set mode. Can be either @code{linear}, or @code{logarithmic}.
10736 Default is @code{linear}.
10739 Set what color components to display.
10740 Default is @code{7}.
10743 Set foreground opacity. Default is @code{0.7}.
10746 Set background opacity. Default is @code{0.5}.
10749 @subsection Examples
10754 Calculate and draw histogram:
10756 ffplay -i input -vf histogram
10764 This is a high precision/quality 3d denoise filter. It aims to reduce
10765 image noise, producing smooth images and making still images really
10766 still. It should enhance compressibility.
10768 It accepts the following optional parameters:
10772 A non-negative floating point number which specifies spatial luma strength.
10773 It defaults to 4.0.
10775 @item chroma_spatial
10776 A non-negative floating point number which specifies spatial chroma strength.
10777 It defaults to 3.0*@var{luma_spatial}/4.0.
10780 A floating point number which specifies luma temporal strength. It defaults to
10781 6.0*@var{luma_spatial}/4.0.
10784 A floating point number which specifies chroma temporal strength. It defaults to
10785 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
10788 @anchor{hwdownload}
10789 @section hwdownload
10791 Download hardware frames to system memory.
10793 The input must be in hardware frames, and the output a non-hardware format.
10794 Not all formats will be supported on the output - it may be necessary to insert
10795 an additional @option{format} filter immediately following in the graph to get
10796 the output in a supported format.
10800 Map hardware frames to system memory or to another device.
10802 This filter has several different modes of operation; which one is used depends
10803 on the input and output formats:
10806 Hardware frame input, normal frame output
10808 Map the input frames to system memory and pass them to the output. If the
10809 original hardware frame is later required (for example, after overlaying
10810 something else on part of it), the @option{hwmap} filter can be used again
10811 in the next mode to retrieve it.
10813 Normal frame input, hardware frame output
10815 If the input is actually a software-mapped hardware frame, then unmap it -
10816 that is, return the original hardware frame.
10818 Otherwise, a device must be provided. Create new hardware surfaces on that
10819 device for the output, then map them back to the software format at the input
10820 and give those frames to the preceding filter. This will then act like the
10821 @option{hwupload} filter, but may be able to avoid an additional copy when
10822 the input is already in a compatible format.
10824 Hardware frame input and output
10826 A device must be supplied for the output, either directly or with the
10827 @option{derive_device} option. The input and output devices must be of
10828 different types and compatible - the exact meaning of this is
10829 system-dependent, but typically it means that they must refer to the same
10830 underlying hardware context (for example, refer to the same graphics card).
10832 If the input frames were originally created on the output device, then unmap
10833 to retrieve the original frames.
10835 Otherwise, map the frames to the output device - create new hardware frames
10836 on the output corresponding to the frames on the input.
10839 The following additional parameters are accepted:
10843 Set the frame mapping mode. Some combination of:
10846 The mapped frame should be readable.
10848 The mapped frame should be writeable.
10850 The mapping will always overwrite the entire frame.
10852 This may improve performance in some cases, as the original contents of the
10853 frame need not be loaded.
10855 The mapping must not involve any copying.
10857 Indirect mappings to copies of frames are created in some cases where either
10858 direct mapping is not possible or it would have unexpected properties.
10859 Setting this flag ensures that the mapping is direct and will fail if that is
10862 Defaults to @var{read+write} if not specified.
10864 @item derive_device @var{type}
10865 Rather than using the device supplied at initialisation, instead derive a new
10866 device of type @var{type} from the device the input frames exist on.
10869 In a hardware to hardware mapping, map in reverse - create frames in the sink
10870 and map them back to the source. This may be necessary in some cases where
10871 a mapping in one direction is required but only the opposite direction is
10872 supported by the devices being used.
10874 This option is dangerous - it may break the preceding filter in undefined
10875 ways if there are any additional constraints on that filter's output.
10876 Do not use it without fully understanding the implications of its use.
10882 Upload system memory frames to hardware surfaces.
10884 The device to upload to must be supplied when the filter is initialised. If
10885 using ffmpeg, select the appropriate device with the @option{-filter_hw_device}
10888 @anchor{hwupload_cuda}
10889 @section hwupload_cuda
10891 Upload system memory frames to a CUDA device.
10893 It accepts the following optional parameters:
10897 The number of the CUDA device to use
10902 Apply a high-quality magnification filter designed for pixel art. This filter
10903 was originally created by Maxim Stepin.
10905 It accepts the following option:
10909 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
10910 @code{hq3x} and @code{4} for @code{hq4x}.
10911 Default is @code{3}.
10915 Stack input videos horizontally.
10917 All streams must be of same pixel format and of same height.
10919 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
10920 to create same output.
10922 The filter accept the following option:
10926 Set number of input streams. Default is 2.
10929 If set to 1, force the output to terminate when the shortest input
10930 terminates. Default value is 0.
10935 Modify the hue and/or the saturation of the input.
10937 It accepts the following parameters:
10941 Specify the hue angle as a number of degrees. It accepts an expression,
10942 and defaults to "0".
10945 Specify the saturation in the [-10,10] range. It accepts an expression and
10949 Specify the hue angle as a number of radians. It accepts an
10950 expression, and defaults to "0".
10953 Specify the brightness in the [-10,10] range. It accepts an expression and
10957 @option{h} and @option{H} are mutually exclusive, and can't be
10958 specified at the same time.
10960 The @option{b}, @option{h}, @option{H} and @option{s} option values are
10961 expressions containing the following constants:
10965 frame count of the input frame starting from 0
10968 presentation timestamp of the input frame expressed in time base units
10971 frame rate of the input video, NAN if the input frame rate is unknown
10974 timestamp expressed in seconds, NAN if the input timestamp is unknown
10977 time base of the input video
10980 @subsection Examples
10984 Set the hue to 90 degrees and the saturation to 1.0:
10990 Same command but expressing the hue in radians:
10996 Rotate hue and make the saturation swing between 0
10997 and 2 over a period of 1 second:
10999 hue="H=2*PI*t: s=sin(2*PI*t)+1"
11003 Apply a 3 seconds saturation fade-in effect starting at 0:
11005 hue="s=min(t/3\,1)"
11008 The general fade-in expression can be written as:
11010 hue="s=min(0\, max((t-START)/DURATION\, 1))"
11014 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
11016 hue="s=max(0\, min(1\, (8-t)/3))"
11019 The general fade-out expression can be written as:
11021 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
11026 @subsection Commands
11028 This filter supports the following commands:
11034 Modify the hue and/or the saturation and/or brightness of the input video.
11035 The command accepts the same syntax of the corresponding option.
11037 If the specified expression is not valid, it is kept at its current
11041 @section hysteresis
11043 Grow first stream into second stream by connecting components.
11044 This makes it possible to build more robust edge masks.
11046 This filter accepts the following options:
11050 Set which planes will be processed as bitmap, unprocessed planes will be
11051 copied from first stream.
11052 By default value 0xf, all planes will be processed.
11055 Set threshold which is used in filtering. If pixel component value is higher than
11056 this value filter algorithm for connecting components is activated.
11057 By default value is 0.
11062 Detect video interlacing type.
11064 This filter tries to detect if the input frames are interlaced, progressive,
11065 top or bottom field first. It will also try to detect fields that are
11066 repeated between adjacent frames (a sign of telecine).
11068 Single frame detection considers only immediately adjacent frames when classifying each frame.
11069 Multiple frame detection incorporates the classification history of previous frames.
11071 The filter will log these metadata values:
11074 @item single.current_frame
11075 Detected type of current frame using single-frame detection. One of:
11076 ``tff'' (top field first), ``bff'' (bottom field first),
11077 ``progressive'', or ``undetermined''
11080 Cumulative number of frames detected as top field first using single-frame detection.
11083 Cumulative number of frames detected as top field first using multiple-frame detection.
11086 Cumulative number of frames detected as bottom field first using single-frame detection.
11088 @item multiple.current_frame
11089 Detected type of current frame using multiple-frame detection. One of:
11090 ``tff'' (top field first), ``bff'' (bottom field first),
11091 ``progressive'', or ``undetermined''
11094 Cumulative number of frames detected as bottom field first using multiple-frame detection.
11096 @item single.progressive
11097 Cumulative number of frames detected as progressive using single-frame detection.
11099 @item multiple.progressive
11100 Cumulative number of frames detected as progressive using multiple-frame detection.
11102 @item single.undetermined
11103 Cumulative number of frames that could not be classified using single-frame detection.
11105 @item multiple.undetermined
11106 Cumulative number of frames that could not be classified using multiple-frame detection.
11108 @item repeated.current_frame
11109 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
11111 @item repeated.neither
11112 Cumulative number of frames with no repeated field.
11115 Cumulative number of frames with the top field repeated from the previous frame's top field.
11117 @item repeated.bottom
11118 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
11121 The filter accepts the following options:
11125 Set interlacing threshold.
11127 Set progressive threshold.
11129 Threshold for repeated field detection.
11131 Number of frames after which a given frame's contribution to the
11132 statistics is halved (i.e., it contributes only 0.5 to its
11133 classification). The default of 0 means that all frames seen are given
11134 full weight of 1.0 forever.
11135 @item analyze_interlaced_flag
11136 When this is not 0 then idet will use the specified number of frames to determine
11137 if the interlaced flag is accurate, it will not count undetermined frames.
11138 If the flag is found to be accurate it will be used without any further
11139 computations, if it is found to be inaccurate it will be cleared without any
11140 further computations. This allows inserting the idet filter as a low computational
11141 method to clean up the interlaced flag
11146 Deinterleave or interleave fields.
11148 This filter allows one to process interlaced images fields without
11149 deinterlacing them. Deinterleaving splits the input frame into 2
11150 fields (so called half pictures). Odd lines are moved to the top
11151 half of the output image, even lines to the bottom half.
11152 You can process (filter) them independently and then re-interleave them.
11154 The filter accepts the following options:
11158 @item chroma_mode, c
11159 @item alpha_mode, a
11160 Available values for @var{luma_mode}, @var{chroma_mode} and
11161 @var{alpha_mode} are:
11167 @item deinterleave, d
11168 Deinterleave fields, placing one above the other.
11170 @item interleave, i
11171 Interleave fields. Reverse the effect of deinterleaving.
11173 Default value is @code{none}.
11175 @item luma_swap, ls
11176 @item chroma_swap, cs
11177 @item alpha_swap, as
11178 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
11183 Apply inflate effect to the video.
11185 This filter replaces the pixel by the local(3x3) average by taking into account
11186 only values higher than the pixel.
11188 It accepts the following options:
11195 Limit the maximum change for each plane, default is 65535.
11196 If 0, plane will remain unchanged.
11201 Simple interlacing filter from progressive contents. This interleaves upper (or
11202 lower) lines from odd frames with lower (or upper) lines from even frames,
11203 halving the frame rate and preserving image height.
11206 Original Original New Frame
11207 Frame 'j' Frame 'j+1' (tff)
11208 ========== =========== ==================
11209 Line 0 --------------------> Frame 'j' Line 0
11210 Line 1 Line 1 ----> Frame 'j+1' Line 1
11211 Line 2 ---------------------> Frame 'j' Line 2
11212 Line 3 Line 3 ----> Frame 'j+1' Line 3
11214 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
11217 It accepts the following optional parameters:
11221 This determines whether the interlaced frame is taken from the even
11222 (tff - default) or odd (bff) lines of the progressive frame.
11225 Vertical lowpass filter to avoid twitter interlacing and
11226 reduce moire patterns.
11230 Disable vertical lowpass filter
11233 Enable linear filter (default)
11236 Enable complex filter. This will slightly less reduce twitter and moire
11237 but better retain detail and subjective sharpness impression.
11244 Deinterlace input video by applying Donald Graft's adaptive kernel
11245 deinterling. Work on interlaced parts of a video to produce
11246 progressive frames.
11248 The description of the accepted parameters follows.
11252 Set the threshold which affects the filter's tolerance when
11253 determining if a pixel line must be processed. It must be an integer
11254 in the range [0,255] and defaults to 10. A value of 0 will result in
11255 applying the process on every pixels.
11258 Paint pixels exceeding the threshold value to white if set to 1.
11262 Set the fields order. Swap fields if set to 1, leave fields alone if
11266 Enable additional sharpening if set to 1. Default is 0.
11269 Enable twoway sharpening if set to 1. Default is 0.
11272 @subsection Examples
11276 Apply default values:
11278 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
11282 Enable additional sharpening:
11288 Paint processed pixels in white:
11294 @section lenscorrection
11296 Correct radial lens distortion
11298 This filter can be used to correct for radial distortion as can result from the use
11299 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
11300 one can use tools available for example as part of opencv or simply trial-and-error.
11301 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
11302 and extract the k1 and k2 coefficients from the resulting matrix.
11304 Note that effectively the same filter is available in the open-source tools Krita and
11305 Digikam from the KDE project.
11307 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
11308 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
11309 brightness distribution, so you may want to use both filters together in certain
11310 cases, though you will have to take care of ordering, i.e. whether vignetting should
11311 be applied before or after lens correction.
11313 @subsection Options
11315 The filter accepts the following options:
11319 Relative x-coordinate of the focal point of the image, and thereby the center of the
11320 distortion. This value has a range [0,1] and is expressed as fractions of the image
11321 width. Default is 0.5.
11323 Relative y-coordinate of the focal point of the image, and thereby the center of the
11324 distortion. This value has a range [0,1] and is expressed as fractions of the image
11325 height. Default is 0.5.
11327 Coefficient of the quadratic correction term. This value has a range [-1,1]. 0 means
11328 no correction. Default is 0.
11330 Coefficient of the double quadratic correction term. This value has a range [-1,1].
11331 0 means no correction. Default is 0.
11334 The formula that generates the correction is:
11336 @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)
11338 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
11339 distances from the focal point in the source and target images, respectively.
11343 Apply lens correction via the lensfun library (@url{http://lensfun.sourceforge.net/}).
11345 The @code{lensfun} filter requires the camera make, camera model, and lens model
11346 to apply the lens correction. The filter will load the lensfun database and
11347 query it to find the corresponding camera and lens entries in the database. As
11348 long as these entries can be found with the given options, the filter can
11349 perform corrections on frames. Note that incomplete strings will result in the
11350 filter choosing the best match with the given options, and the filter will
11351 output the chosen camera and lens models (logged with level "info"). You must
11352 provide the make, camera model, and lens model as they are required.
11354 The filter accepts the following options:
11358 The make of the camera (for example, "Canon"). This option is required.
11361 The model of the camera (for example, "Canon EOS 100D"). This option is
11365 The model of the lens (for example, "Canon EF-S 18-55mm f/3.5-5.6 IS STM"). This
11366 option is required.
11369 The type of correction to apply. The following values are valid options:
11373 Enables fixing lens vignetting.
11376 Enables fixing lens geometry. This is the default.
11379 Enables fixing chromatic aberrations.
11382 Enables fixing lens vignetting and lens geometry.
11385 Enables fixing lens vignetting and chromatic aberrations.
11388 Enables fixing both lens geometry and chromatic aberrations.
11391 Enables all possible corrections.
11395 The focal length of the image/video (zoom; expected constant for video). For
11396 example, a 18--55mm lens has focal length range of [18--55], so a value in that
11397 range should be chosen when using that lens. Default 18.
11400 The aperture of the image/video (expected constant for video). Note that
11401 aperture is only used for vignetting correction. Default 3.5.
11403 @item focus_distance
11404 The focus distance of the image/video (expected constant for video). Note that
11405 focus distance is only used for vignetting and only slightly affects the
11406 vignetting correction process. If unknown, leave it at the default value (which
11409 @item target_geometry
11410 The target geometry of the output image/video. The following values are valid
11414 @item rectilinear (default)
11417 @item equirectangular
11418 @item fisheye_orthographic
11419 @item fisheye_stereographic
11420 @item fisheye_equisolid
11421 @item fisheye_thoby
11424 Apply the reverse of image correction (instead of correcting distortion, apply
11427 @item interpolation
11428 The type of interpolation used when correcting distortion. The following values
11433 @item linear (default)
11438 @subsection Examples
11442 Apply lens correction with make "Canon", camera model "Canon EOS 100D", and lens
11443 model "Canon EF-S 18-55mm f/3.5-5.6 IS STM" with focal length of "18" and
11447 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
11451 Apply the same as before, but only for the first 5 seconds of video.
11454 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
11461 Obtain the VMAF (Video Multi-Method Assessment Fusion)
11462 score between two input videos.
11464 The obtained VMAF score is printed through the logging system.
11466 It requires Netflix's vmaf library (libvmaf) as a pre-requisite.
11467 After installing the library it can be enabled using:
11468 @code{./configure --enable-libvmaf --enable-version3}.
11469 If no model path is specified it uses the default model: @code{vmaf_v0.6.1.pkl}.
11471 The filter has following options:
11475 Set the model path which is to be used for SVM.
11476 Default value: @code{"vmaf_v0.6.1.pkl"}
11479 Set the file path to be used to store logs.
11482 Set the format of the log file (xml or json).
11484 @item enable_transform
11485 This option can enable/disable the @code{score_transform} applied to the final predicted VMAF score,
11486 if you have specified score_transform option in the input parameter file passed to @code{run_vmaf_training.py}
11487 Default value: @code{false}
11490 Invokes the phone model which will generate VMAF scores higher than in the
11491 regular model, which is more suitable for laptop, TV, etc. viewing conditions.
11494 Enables computing psnr along with vmaf.
11497 Enables computing ssim along with vmaf.
11500 Enables computing ms_ssim along with vmaf.
11503 Set the pool method (mean, min or harmonic mean) to be used for computing vmaf.
11506 Set number of threads to be used when computing vmaf.
11509 Set interval for frame subsampling used when computing vmaf.
11511 @item enable_conf_interval
11512 Enables confidence interval.
11515 This filter also supports the @ref{framesync} options.
11517 On the below examples the input file @file{main.mpg} being processed is
11518 compared with the reference file @file{ref.mpg}.
11521 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf -f null -
11524 Example with options:
11526 ffmpeg -i main.mpg -i ref.mpg -lavfi libvmaf="psnr=1:log_fmt=json" -f null -
11531 Limits the pixel components values to the specified range [min, max].
11533 The filter accepts the following options:
11537 Lower bound. Defaults to the lowest allowed value for the input.
11540 Upper bound. Defaults to the highest allowed value for the input.
11543 Specify which planes will be processed. Defaults to all available.
11550 The filter accepts the following options:
11554 Set the number of loops. Setting this value to -1 will result in infinite loops.
11558 Set maximal size in number of frames. Default is 0.
11561 Set first frame of loop. Default is 0.
11564 @subsection Examples
11568 Loop single first frame infinitely:
11570 loop=loop=-1:size=1:start=0
11574 Loop single first frame 10 times:
11576 loop=loop=10:size=1:start=0
11580 Loop 10 first frames 5 times:
11582 loop=loop=5:size=10:start=0
11588 Apply a 1D LUT to an input video.
11590 The filter accepts the following options:
11594 Set the 1D LUT file name.
11596 Currently supported formats:
11603 Select interpolation mode.
11605 Available values are:
11609 Use values from the nearest defined point.
11611 Interpolate values using the linear interpolation.
11613 Interpolate values using the cosine interpolation.
11615 Interpolate values using the cubic interpolation.
11617 Interpolate values using the spline interpolation.
11624 Apply a 3D LUT to an input video.
11626 The filter accepts the following options:
11630 Set the 3D LUT file name.
11632 Currently supported formats:
11644 Select interpolation mode.
11646 Available values are:
11650 Use values from the nearest defined point.
11652 Interpolate values using the 8 points defining a cube.
11654 Interpolate values using a tetrahedron.
11658 This filter also supports the @ref{framesync} options.
11662 Turn certain luma values into transparency.
11664 The filter accepts the following options:
11668 Set the luma which will be used as base for transparency.
11669 Default value is @code{0}.
11672 Set the range of luma values to be keyed out.
11673 Default value is @code{0}.
11676 Set the range of softness. Default value is @code{0}.
11677 Use this to control gradual transition from zero to full transparency.
11680 @section lut, lutrgb, lutyuv
11682 Compute a look-up table for binding each pixel component input value
11683 to an output value, and apply it to the input video.
11685 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
11686 to an RGB input video.
11688 These filters accept the following parameters:
11691 set first pixel component expression
11693 set second pixel component expression
11695 set third pixel component expression
11697 set fourth pixel component expression, corresponds to the alpha component
11700 set red component expression
11702 set green component expression
11704 set blue component expression
11706 alpha component expression
11709 set Y/luminance component expression
11711 set U/Cb component expression
11713 set V/Cr component expression
11716 Each of them specifies the expression to use for computing the lookup table for
11717 the corresponding pixel component values.
11719 The exact component associated to each of the @var{c*} options depends on the
11722 The @var{lut} filter requires either YUV or RGB pixel formats in input,
11723 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
11725 The expressions can contain the following constants and functions:
11730 The input width and height.
11733 The input value for the pixel component.
11736 The input value, clipped to the @var{minval}-@var{maxval} range.
11739 The maximum value for the pixel component.
11742 The minimum value for the pixel component.
11745 The negated value for the pixel component value, clipped to the
11746 @var{minval}-@var{maxval} range; it corresponds to the expression
11747 "maxval-clipval+minval".
11750 The computed value in @var{val}, clipped to the
11751 @var{minval}-@var{maxval} range.
11753 @item gammaval(gamma)
11754 The computed gamma correction value of the pixel component value,
11755 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
11757 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
11761 All expressions default to "val".
11763 @subsection Examples
11767 Negate input video:
11769 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
11770 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
11773 The above is the same as:
11775 lutrgb="r=negval:g=negval:b=negval"
11776 lutyuv="y=negval:u=negval:v=negval"
11786 Remove chroma components, turning the video into a graytone image:
11788 lutyuv="u=128:v=128"
11792 Apply a luma burning effect:
11798 Remove green and blue components:
11804 Set a constant alpha channel value on input:
11806 format=rgba,lutrgb=a="maxval-minval/2"
11810 Correct luminance gamma by a factor of 0.5:
11812 lutyuv=y=gammaval(0.5)
11816 Discard least significant bits of luma:
11818 lutyuv=y='bitand(val, 128+64+32)'
11822 Technicolor like effect:
11824 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
11828 @section lut2, tlut2
11830 The @code{lut2} filter takes two input streams and outputs one
11833 The @code{tlut2} (time lut2) filter takes two consecutive frames
11834 from one single stream.
11836 This filter accepts the following parameters:
11839 set first pixel component expression
11841 set second pixel component expression
11843 set third pixel component expression
11845 set fourth pixel component expression, corresponds to the alpha component
11848 set output bit depth, only available for @code{lut2} filter. By default is 0,
11849 which means bit depth is automatically picked from first input format.
11852 Each of them specifies the expression to use for computing the lookup table for
11853 the corresponding pixel component values.
11855 The exact component associated to each of the @var{c*} options depends on the
11858 The expressions can contain the following constants:
11863 The input width and height.
11866 The first input value for the pixel component.
11869 The second input value for the pixel component.
11872 The first input video bit depth.
11875 The second input video bit depth.
11878 All expressions default to "x".
11880 @subsection Examples
11884 Highlight differences between two RGB video streams:
11886 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)'
11890 Highlight differences between two YUV video streams:
11892 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)'
11896 Show max difference between two video streams:
11898 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)))'
11902 @section maskedclamp
11904 Clamp the first input stream with the second input and third input stream.
11906 Returns the value of first stream to be between second input
11907 stream - @code{undershoot} and third input stream + @code{overshoot}.
11909 This filter accepts the following options:
11912 Default value is @code{0}.
11915 Default value is @code{0}.
11918 Set which planes will be processed as bitmap, unprocessed planes will be
11919 copied from first stream.
11920 By default value 0xf, all planes will be processed.
11923 @section maskedmerge
11925 Merge the first input stream with the second input stream using per pixel
11926 weights in the third input stream.
11928 A value of 0 in the third stream pixel component means that pixel component
11929 from first stream is returned unchanged, while maximum value (eg. 255 for
11930 8-bit videos) means that pixel component from second stream is returned
11931 unchanged. Intermediate values define the amount of merging between both
11932 input stream's pixel components.
11934 This filter accepts the following options:
11937 Set which planes will be processed as bitmap, unprocessed planes will be
11938 copied from first stream.
11939 By default value 0xf, all planes will be processed.
11943 Create mask from input video.
11945 For example it is useful to create motion masks after @code{tblend} filter.
11947 This filter accepts the following options:
11951 Set low threshold. Any pixel component lower or exact than this value will be set to 0.
11954 Set high threshold. Any pixel component higher than this value will be set to max value
11955 allowed for current pixel format.
11958 Set planes to filter, by default all available planes are filtered.
11961 Fill all frame pixels with this value.
11964 Set max average pixel value for frame. If sum of all pixel components is higher that this
11965 average, output frame will be completely filled with value set by @var{fill} option.
11966 Typically useful for scene changes when used in combination with @code{tblend} filter.
11971 Apply motion-compensation deinterlacing.
11973 It needs one field per frame as input and must thus be used together
11974 with yadif=1/3 or equivalent.
11976 This filter accepts the following options:
11979 Set the deinterlacing mode.
11981 It accepts one of the following values:
11986 use iterative motion estimation
11988 like @samp{slow}, but use multiple reference frames.
11990 Default value is @samp{fast}.
11993 Set the picture field parity assumed for the input video. It must be
11994 one of the following values:
11998 assume top field first
12000 assume bottom field first
12003 Default value is @samp{bff}.
12006 Set per-block quantization parameter (QP) used by the internal
12009 Higher values should result in a smoother motion vector field but less
12010 optimal individual vectors. Default value is 1.
12013 @section mergeplanes
12015 Merge color channel components from several video streams.
12017 The filter accepts up to 4 input streams, and merge selected input
12018 planes to the output video.
12020 This filter accepts the following options:
12023 Set input to output plane mapping. Default is @code{0}.
12025 The mappings is specified as a bitmap. It should be specified as a
12026 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
12027 mapping for the first plane of the output stream. 'A' sets the number of
12028 the input stream to use (from 0 to 3), and 'a' the plane number of the
12029 corresponding input to use (from 0 to 3). The rest of the mappings is
12030 similar, 'Bb' describes the mapping for the output stream second
12031 plane, 'Cc' describes the mapping for the output stream third plane and
12032 'Dd' describes the mapping for the output stream fourth plane.
12035 Set output pixel format. Default is @code{yuva444p}.
12038 @subsection Examples
12042 Merge three gray video streams of same width and height into single video stream:
12044 [a0][a1][a2]mergeplanes=0x001020:yuv444p
12048 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
12050 [a0][a1]mergeplanes=0x00010210:yuva444p
12054 Swap Y and A plane in yuva444p stream:
12056 format=yuva444p,mergeplanes=0x03010200:yuva444p
12060 Swap U and V plane in yuv420p stream:
12062 format=yuv420p,mergeplanes=0x000201:yuv420p
12066 Cast a rgb24 clip to yuv444p:
12068 format=rgb24,mergeplanes=0x000102:yuv444p
12074 Estimate and export motion vectors using block matching algorithms.
12075 Motion vectors are stored in frame side data to be used by other filters.
12077 This filter accepts the following options:
12080 Specify the motion estimation method. Accepts one of the following values:
12084 Exhaustive search algorithm.
12086 Three step search algorithm.
12088 Two dimensional logarithmic search algorithm.
12090 New three step search algorithm.
12092 Four step search algorithm.
12094 Diamond search algorithm.
12096 Hexagon-based search algorithm.
12098 Enhanced predictive zonal search algorithm.
12100 Uneven multi-hexagon search algorithm.
12102 Default value is @samp{esa}.
12105 Macroblock size. Default @code{16}.
12108 Search parameter. Default @code{7}.
12111 @section midequalizer
12113 Apply Midway Image Equalization effect using two video streams.
12115 Midway Image Equalization adjusts a pair of images to have the same
12116 histogram, while maintaining their dynamics as much as possible. It's
12117 useful for e.g. matching exposures from a pair of stereo cameras.
12119 This filter has two inputs and one output, which must be of same pixel format, but
12120 may be of different sizes. The output of filter is first input adjusted with
12121 midway histogram of both inputs.
12123 This filter accepts the following option:
12127 Set which planes to process. Default is @code{15}, which is all available planes.
12130 @section minterpolate
12132 Convert the video to specified frame rate using motion interpolation.
12134 This filter accepts the following options:
12137 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}.
12140 Motion interpolation mode. Following values are accepted:
12143 Duplicate previous or next frame for interpolating new ones.
12145 Blend source frames. Interpolated frame is mean of previous and next frames.
12147 Motion compensated interpolation. Following options are effective when this mode is selected:
12151 Motion compensation mode. Following values are accepted:
12154 Overlapped block motion compensation.
12156 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
12158 Default mode is @samp{obmc}.
12161 Motion estimation mode. Following values are accepted:
12164 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
12166 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
12168 Default mode is @samp{bilat}.
12171 The algorithm to be used for motion estimation. Following values are accepted:
12174 Exhaustive search algorithm.
12176 Three step search algorithm.
12178 Two dimensional logarithmic search algorithm.
12180 New three step search algorithm.
12182 Four step search algorithm.
12184 Diamond search algorithm.
12186 Hexagon-based search algorithm.
12188 Enhanced predictive zonal search algorithm.
12190 Uneven multi-hexagon search algorithm.
12192 Default algorithm is @samp{epzs}.
12195 Macroblock size. Default @code{16}.
12198 Motion estimation search parameter. Default @code{32}.
12201 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).
12206 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:
12209 Disable scene change detection.
12211 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
12213 Default method is @samp{fdiff}.
12215 @item scd_threshold
12216 Scene change detection threshold. Default is @code{5.0}.
12221 Mix several video input streams into one video stream.
12223 A description of the accepted options follows.
12227 The number of inputs. If unspecified, it defaults to 2.
12230 Specify weight of each input video stream as sequence.
12231 Each weight is separated by space. If number of weights
12232 is smaller than number of @var{frames} last specified
12233 weight will be used for all remaining unset weights.
12236 Specify scale, if it is set it will be multiplied with sum
12237 of each weight multiplied with pixel values to give final destination
12238 pixel value. By default @var{scale} is auto scaled to sum of weights.
12241 Specify how end of stream is determined.
12244 The duration of the longest input. (default)
12247 The duration of the shortest input.
12250 The duration of the first input.
12254 @section mpdecimate
12256 Drop frames that do not differ greatly from the previous frame in
12257 order to reduce frame rate.
12259 The main use of this filter is for very-low-bitrate encoding
12260 (e.g. streaming over dialup modem), but it could in theory be used for
12261 fixing movies that were inverse-telecined incorrectly.
12263 A description of the accepted options follows.
12267 Set the maximum number of consecutive frames which can be dropped (if
12268 positive), or the minimum interval between dropped frames (if
12269 negative). If the value is 0, the frame is dropped disregarding the
12270 number of previous sequentially dropped frames.
12272 Default value is 0.
12277 Set the dropping threshold values.
12279 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
12280 represent actual pixel value differences, so a threshold of 64
12281 corresponds to 1 unit of difference for each pixel, or the same spread
12282 out differently over the block.
12284 A frame is a candidate for dropping if no 8x8 blocks differ by more
12285 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
12286 meaning the whole image) differ by more than a threshold of @option{lo}.
12288 Default value for @option{hi} is 64*12, default value for @option{lo} is
12289 64*5, and default value for @option{frac} is 0.33.
12295 Negate (invert) the input video.
12297 It accepts the following option:
12302 With value 1, it negates the alpha component, if present. Default value is 0.
12308 Denoise frames using Non-Local Means algorithm.
12310 Each pixel is adjusted by looking for other pixels with similar contexts. This
12311 context similarity is defined by comparing their surrounding patches of size
12312 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
12315 Note that the research area defines centers for patches, which means some
12316 patches will be made of pixels outside that research area.
12318 The filter accepts the following options.
12322 Set denoising strength. Default is 1.0. Must be in range [1.0, 30.0].
12325 Set patch size. Default is 7. Must be odd number in range [0, 99].
12328 Same as @option{p} but for chroma planes.
12330 The default value is @var{0} and means automatic.
12333 Set research size. Default is 15. Must be odd number in range [0, 99].
12336 Same as @option{r} but for chroma planes.
12338 The default value is @var{0} and means automatic.
12343 Deinterlace video using neural network edge directed interpolation.
12345 This filter accepts the following options:
12349 Mandatory option, without binary file filter can not work.
12350 Currently file can be found here:
12351 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
12354 Set which frames to deinterlace, by default it is @code{all}.
12355 Can be @code{all} or @code{interlaced}.
12358 Set mode of operation.
12360 Can be one of the following:
12364 Use frame flags, both fields.
12366 Use frame flags, single field.
12368 Use top field only.
12370 Use bottom field only.
12372 Use both fields, top first.
12374 Use both fields, bottom first.
12378 Set which planes to process, by default filter process all frames.
12381 Set size of local neighborhood around each pixel, used by the predictor neural
12384 Can be one of the following:
12397 Set the number of neurons in predictor neural network.
12398 Can be one of the following:
12409 Controls the number of different neural network predictions that are blended
12410 together to compute the final output value. Can be @code{fast}, default or
12414 Set which set of weights to use in the predictor.
12415 Can be one of the following:
12419 weights trained to minimize absolute error
12421 weights trained to minimize squared error
12425 Controls whether or not the prescreener neural network is used to decide
12426 which pixels should be processed by the predictor neural network and which
12427 can be handled by simple cubic interpolation.
12428 The prescreener is trained to know whether cubic interpolation will be
12429 sufficient for a pixel or whether it should be predicted by the predictor nn.
12430 The computational complexity of the prescreener nn is much less than that of
12431 the predictor nn. Since most pixels can be handled by cubic interpolation,
12432 using the prescreener generally results in much faster processing.
12433 The prescreener is pretty accurate, so the difference between using it and not
12434 using it is almost always unnoticeable.
12436 Can be one of the following:
12444 Default is @code{new}.
12447 Set various debugging flags.
12452 Force libavfilter not to use any of the specified pixel formats for the
12453 input to the next filter.
12455 It accepts the following parameters:
12459 A '|'-separated list of pixel format names, such as
12460 pix_fmts=yuv420p|monow|rgb24".
12464 @subsection Examples
12468 Force libavfilter to use a format different from @var{yuv420p} for the
12469 input to the vflip filter:
12471 noformat=pix_fmts=yuv420p,vflip
12475 Convert the input video to any of the formats not contained in the list:
12477 noformat=yuv420p|yuv444p|yuv410p
12483 Add noise on video input frame.
12485 The filter accepts the following options:
12493 Set noise seed for specific pixel component or all pixel components in case
12494 of @var{all_seed}. Default value is @code{123457}.
12496 @item all_strength, alls
12497 @item c0_strength, c0s
12498 @item c1_strength, c1s
12499 @item c2_strength, c2s
12500 @item c3_strength, c3s
12501 Set noise strength for specific pixel component or all pixel components in case
12502 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
12504 @item all_flags, allf
12505 @item c0_flags, c0f
12506 @item c1_flags, c1f
12507 @item c2_flags, c2f
12508 @item c3_flags, c3f
12509 Set pixel component flags or set flags for all components if @var{all_flags}.
12510 Available values for component flags are:
12513 averaged temporal noise (smoother)
12515 mix random noise with a (semi)regular pattern
12517 temporal noise (noise pattern changes between frames)
12519 uniform noise (gaussian otherwise)
12523 @subsection Examples
12525 Add temporal and uniform noise to input video:
12527 noise=alls=20:allf=t+u
12532 Normalize RGB video (aka histogram stretching, contrast stretching).
12533 See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
12535 For each channel of each frame, the filter computes the input range and maps
12536 it linearly to the user-specified output range. The output range defaults
12537 to the full dynamic range from pure black to pure white.
12539 Temporal smoothing can be used on the input range to reduce flickering (rapid
12540 changes in brightness) caused when small dark or bright objects enter or leave
12541 the scene. This is similar to the auto-exposure (automatic gain control) on a
12542 video camera, and, like a video camera, it may cause a period of over- or
12543 under-exposure of the video.
12545 The R,G,B channels can be normalized independently, which may cause some
12546 color shifting, or linked together as a single channel, which prevents
12547 color shifting. Linked normalization preserves hue. Independent normalization
12548 does not, so it can be used to remove some color casts. Independent and linked
12549 normalization can be combined in any ratio.
12551 The normalize filter accepts the following options:
12556 Colors which define the output range. The minimum input value is mapped to
12557 the @var{blackpt}. The maximum input value is mapped to the @var{whitept}.
12558 The defaults are black and white respectively. Specifying white for
12559 @var{blackpt} and black for @var{whitept} will give color-inverted,
12560 normalized video. Shades of grey can be used to reduce the dynamic range
12561 (contrast). Specifying saturated colors here can create some interesting
12565 The number of previous frames to use for temporal smoothing. The input range
12566 of each channel is smoothed using a rolling average over the current frame
12567 and the @var{smoothing} previous frames. The default is 0 (no temporal
12571 Controls the ratio of independent (color shifting) channel normalization to
12572 linked (color preserving) normalization. 0.0 is fully linked, 1.0 is fully
12573 independent. Defaults to 1.0 (fully independent).
12576 Overall strength of the filter. 1.0 is full strength. 0.0 is a rather
12577 expensive no-op. Defaults to 1.0 (full strength).
12581 @subsection Examples
12583 Stretch video contrast to use the full dynamic range, with no temporal
12584 smoothing; may flicker depending on the source content:
12586 normalize=blackpt=black:whitept=white:smoothing=0
12589 As above, but with 50 frames of temporal smoothing; flicker should be
12590 reduced, depending on the source content:
12592 normalize=blackpt=black:whitept=white:smoothing=50
12595 As above, but with hue-preserving linked channel normalization:
12597 normalize=blackpt=black:whitept=white:smoothing=50:independence=0
12600 As above, but with half strength:
12602 normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
12605 Map the darkest input color to red, the brightest input color to cyan:
12607 normalize=blackpt=red:whitept=cyan
12612 Pass the video source unchanged to the output.
12615 Optical Character Recognition
12617 This filter uses Tesseract for optical character recognition. To enable
12618 compilation of this filter, you need to configure FFmpeg with
12619 @code{--enable-libtesseract}.
12621 It accepts the following options:
12625 Set datapath to tesseract data. Default is to use whatever was
12626 set at installation.
12629 Set language, default is "eng".
12632 Set character whitelist.
12635 Set character blacklist.
12638 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
12642 Apply a video transform using libopencv.
12644 To enable this filter, install the libopencv library and headers and
12645 configure FFmpeg with @code{--enable-libopencv}.
12647 It accepts the following parameters:
12652 The name of the libopencv filter to apply.
12654 @item filter_params
12655 The parameters to pass to the libopencv filter. If not specified, the default
12656 values are assumed.
12660 Refer to the official libopencv documentation for more precise
12662 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
12664 Several libopencv filters are supported; see the following subsections.
12669 Dilate an image by using a specific structuring element.
12670 It corresponds to the libopencv function @code{cvDilate}.
12672 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
12674 @var{struct_el} represents a structuring element, and has the syntax:
12675 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
12677 @var{cols} and @var{rows} represent the number of columns and rows of
12678 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
12679 point, and @var{shape} the shape for the structuring element. @var{shape}
12680 must be "rect", "cross", "ellipse", or "custom".
12682 If the value for @var{shape} is "custom", it must be followed by a
12683 string of the form "=@var{filename}". The file with name
12684 @var{filename} is assumed to represent a binary image, with each
12685 printable character corresponding to a bright pixel. When a custom
12686 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
12687 or columns and rows of the read file are assumed instead.
12689 The default value for @var{struct_el} is "3x3+0x0/rect".
12691 @var{nb_iterations} specifies the number of times the transform is
12692 applied to the image, and defaults to 1.
12696 # Use the default values
12699 # Dilate using a structuring element with a 5x5 cross, iterating two times
12700 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
12702 # Read the shape from the file diamond.shape, iterating two times.
12703 # The file diamond.shape may contain a pattern of characters like this
12709 # The specified columns and rows are ignored
12710 # but the anchor point coordinates are not
12711 ocv=dilate:0x0+2x2/custom=diamond.shape|2
12716 Erode an image by using a specific structuring element.
12717 It corresponds to the libopencv function @code{cvErode}.
12719 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
12720 with the same syntax and semantics as the @ref{dilate} filter.
12724 Smooth the input video.
12726 The filter takes the following parameters:
12727 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
12729 @var{type} is the type of smooth filter to apply, and must be one of
12730 the following values: "blur", "blur_no_scale", "median", "gaussian",
12731 or "bilateral". The default value is "gaussian".
12733 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
12734 depend on the smooth type. @var{param1} and
12735 @var{param2} accept integer positive values or 0. @var{param3} and
12736 @var{param4} accept floating point values.
12738 The default value for @var{param1} is 3. The default value for the
12739 other parameters is 0.
12741 These parameters correspond to the parameters assigned to the
12742 libopencv function @code{cvSmooth}.
12744 @section oscilloscope
12746 2D Video Oscilloscope.
12748 Useful to measure spatial impulse, step responses, chroma delays, etc.
12750 It accepts the following parameters:
12754 Set scope center x position.
12757 Set scope center y position.
12760 Set scope size, relative to frame diagonal.
12763 Set scope tilt/rotation.
12769 Set trace center x position.
12772 Set trace center y position.
12775 Set trace width, relative to width of frame.
12778 Set trace height, relative to height of frame.
12781 Set which components to trace. By default it traces first three components.
12784 Draw trace grid. By default is enabled.
12787 Draw some statistics. By default is enabled.
12790 Draw scope. By default is enabled.
12793 @subsection Examples
12797 Inspect full first row of video frame.
12799 oscilloscope=x=0.5:y=0:s=1
12803 Inspect full last row of video frame.
12805 oscilloscope=x=0.5:y=1:s=1
12809 Inspect full 5th line of video frame of height 1080.
12811 oscilloscope=x=0.5:y=5/1080:s=1
12815 Inspect full last column of video frame.
12817 oscilloscope=x=1:y=0.5:s=1:t=1
12825 Overlay one video on top of another.
12827 It takes two inputs and has one output. The first input is the "main"
12828 video on which the second input is overlaid.
12830 It accepts the following parameters:
12832 A description of the accepted options follows.
12837 Set the expression for the x and y coordinates of the overlaid video
12838 on the main video. Default value is "0" for both expressions. In case
12839 the expression is invalid, it is set to a huge value (meaning that the
12840 overlay will not be displayed within the output visible area).
12843 See @ref{framesync}.
12846 Set when the expressions for @option{x}, and @option{y} are evaluated.
12848 It accepts the following values:
12851 only evaluate expressions once during the filter initialization or
12852 when a command is processed
12855 evaluate expressions for each incoming frame
12858 Default value is @samp{frame}.
12861 See @ref{framesync}.
12864 Set the format for the output video.
12866 It accepts the following values:
12869 force YUV420 output
12872 force YUV422 output
12875 force YUV444 output
12878 force packed RGB output
12881 force planar RGB output
12884 automatically pick format
12887 Default value is @samp{yuv420}.
12890 See @ref{framesync}.
12893 Set format of alpha of the overlaid video, it can be @var{straight} or
12894 @var{premultiplied}. Default is @var{straight}.
12897 The @option{x}, and @option{y} expressions can contain the following
12903 The main input width and height.
12907 The overlay input width and height.
12911 The computed values for @var{x} and @var{y}. They are evaluated for
12916 horizontal and vertical chroma subsample values of the output
12917 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
12921 the number of input frame, starting from 0
12924 the position in the file of the input frame, NAN if unknown
12927 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
12931 This filter also supports the @ref{framesync} options.
12933 Note that the @var{n}, @var{pos}, @var{t} variables are available only
12934 when evaluation is done @emph{per frame}, and will evaluate to NAN
12935 when @option{eval} is set to @samp{init}.
12937 Be aware that frames are taken from each input video in timestamp
12938 order, hence, if their initial timestamps differ, it is a good idea
12939 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
12940 have them begin in the same zero timestamp, as the example for
12941 the @var{movie} filter does.
12943 You can chain together more overlays but you should test the
12944 efficiency of such approach.
12946 @subsection Commands
12948 This filter supports the following commands:
12952 Modify the x and y of the overlay input.
12953 The command accepts the same syntax of the corresponding option.
12955 If the specified expression is not valid, it is kept at its current
12959 @subsection Examples
12963 Draw the overlay at 10 pixels from the bottom right corner of the main
12966 overlay=main_w-overlay_w-10:main_h-overlay_h-10
12969 Using named options the example above becomes:
12971 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
12975 Insert a transparent PNG logo in the bottom left corner of the input,
12976 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
12978 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
12982 Insert 2 different transparent PNG logos (second logo on bottom
12983 right corner) using the @command{ffmpeg} tool:
12985 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
12989 Add a transparent color layer on top of the main video; @code{WxH}
12990 must specify the size of the main input to the overlay filter:
12992 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
12996 Play an original video and a filtered version (here with the deshake
12997 filter) side by side using the @command{ffplay} tool:
12999 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
13002 The above command is the same as:
13004 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
13008 Make a sliding overlay appearing from the left to the right top part of the
13009 screen starting since time 2:
13011 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
13015 Compose output by putting two input videos side to side:
13017 ffmpeg -i left.avi -i right.avi -filter_complex "
13018 nullsrc=size=200x100 [background];
13019 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
13020 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
13021 [background][left] overlay=shortest=1 [background+left];
13022 [background+left][right] overlay=shortest=1:x=100 [left+right]
13027 Mask 10-20 seconds of a video by applying the delogo filter to a section
13029 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
13030 -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]'
13035 Chain several overlays in cascade:
13037 nullsrc=s=200x200 [bg];
13038 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
13039 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
13040 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
13041 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
13042 [in3] null, [mid2] overlay=100:100 [out0]
13049 Apply Overcomplete Wavelet denoiser.
13051 The filter accepts the following options:
13057 Larger depth values will denoise lower frequency components more, but
13058 slow down filtering.
13060 Must be an int in the range 8-16, default is @code{8}.
13062 @item luma_strength, ls
13065 Must be a double value in the range 0-1000, default is @code{1.0}.
13067 @item chroma_strength, cs
13068 Set chroma strength.
13070 Must be a double value in the range 0-1000, default is @code{1.0}.
13076 Add paddings to the input image, and place the original input at the
13077 provided @var{x}, @var{y} coordinates.
13079 It accepts the following parameters:
13084 Specify an expression for the size of the output image with the
13085 paddings added. If the value for @var{width} or @var{height} is 0, the
13086 corresponding input size is used for the output.
13088 The @var{width} expression can reference the value set by the
13089 @var{height} expression, and vice versa.
13091 The default value of @var{width} and @var{height} is 0.
13095 Specify the offsets to place the input image at within the padded area,
13096 with respect to the top/left border of the output image.
13098 The @var{x} expression can reference the value set by the @var{y}
13099 expression, and vice versa.
13101 The default value of @var{x} and @var{y} is 0.
13103 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
13104 so the input image is centered on the padded area.
13107 Specify the color of the padded area. For the syntax of this option,
13108 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
13109 manual,ffmpeg-utils}.
13111 The default value of @var{color} is "black".
13114 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
13116 It accepts the following values:
13120 Only evaluate expressions once during the filter initialization or when
13121 a command is processed.
13124 Evaluate expressions for each incoming frame.
13128 Default value is @samp{init}.
13131 Pad to aspect instead to a resolution.
13135 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
13136 options are expressions containing the following constants:
13141 The input video width and height.
13145 These are the same as @var{in_w} and @var{in_h}.
13149 The output width and height (the size of the padded area), as
13150 specified by the @var{width} and @var{height} expressions.
13154 These are the same as @var{out_w} and @var{out_h}.
13158 The x and y offsets as specified by the @var{x} and @var{y}
13159 expressions, or NAN if not yet specified.
13162 same as @var{iw} / @var{ih}
13165 input sample aspect ratio
13168 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
13172 The horizontal and vertical chroma subsample values. For example for the
13173 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
13176 @subsection Examples
13180 Add paddings with the color "violet" to the input video. The output video
13181 size is 640x480, and the top-left corner of the input video is placed at
13184 pad=640:480:0:40:violet
13187 The example above is equivalent to the following command:
13189 pad=width=640:height=480:x=0:y=40:color=violet
13193 Pad the input to get an output with dimensions increased by 3/2,
13194 and put the input video at the center of the padded area:
13196 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
13200 Pad the input to get a squared output with size equal to the maximum
13201 value between the input width and height, and put the input video at
13202 the center of the padded area:
13204 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
13208 Pad the input to get a final w/h ratio of 16:9:
13210 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
13214 In case of anamorphic video, in order to set the output display aspect
13215 correctly, it is necessary to use @var{sar} in the expression,
13216 according to the relation:
13218 (ih * X / ih) * sar = output_dar
13219 X = output_dar / sar
13222 Thus the previous example needs to be modified to:
13224 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
13228 Double the output size and put the input video in the bottom-right
13229 corner of the output padded area:
13231 pad="2*iw:2*ih:ow-iw:oh-ih"
13235 @anchor{palettegen}
13236 @section palettegen
13238 Generate one palette for a whole video stream.
13240 It accepts the following options:
13244 Set the maximum number of colors to quantize in the palette.
13245 Note: the palette will still contain 256 colors; the unused palette entries
13248 @item reserve_transparent
13249 Create a palette of 255 colors maximum and reserve the last one for
13250 transparency. Reserving the transparency color is useful for GIF optimization.
13251 If not set, the maximum of colors in the palette will be 256. You probably want
13252 to disable this option for a standalone image.
13255 @item transparency_color
13256 Set the color that will be used as background for transparency.
13259 Set statistics mode.
13261 It accepts the following values:
13264 Compute full frame histograms.
13266 Compute histograms only for the part that differs from previous frame. This
13267 might be relevant to give more importance to the moving part of your input if
13268 the background is static.
13270 Compute new histogram for each frame.
13273 Default value is @var{full}.
13276 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
13277 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
13278 color quantization of the palette. This information is also visible at
13279 @var{info} logging level.
13281 @subsection Examples
13285 Generate a representative palette of a given video using @command{ffmpeg}:
13287 ffmpeg -i input.mkv -vf palettegen palette.png
13291 @section paletteuse
13293 Use a palette to downsample an input video stream.
13295 The filter takes two inputs: one video stream and a palette. The palette must
13296 be a 256 pixels image.
13298 It accepts the following options:
13302 Select dithering mode. Available algorithms are:
13305 Ordered 8x8 bayer dithering (deterministic)
13307 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
13308 Note: this dithering is sometimes considered "wrong" and is included as a
13310 @item floyd_steinberg
13311 Floyd and Steingberg dithering (error diffusion)
13313 Frankie Sierra dithering v2 (error diffusion)
13315 Frankie Sierra dithering v2 "Lite" (error diffusion)
13318 Default is @var{sierra2_4a}.
13321 When @var{bayer} dithering is selected, this option defines the scale of the
13322 pattern (how much the crosshatch pattern is visible). A low value means more
13323 visible pattern for less banding, and higher value means less visible pattern
13324 at the cost of more banding.
13326 The option must be an integer value in the range [0,5]. Default is @var{2}.
13329 If set, define the zone to process
13333 Only the changing rectangle will be reprocessed. This is similar to GIF
13334 cropping/offsetting compression mechanism. This option can be useful for speed
13335 if only a part of the image is changing, and has use cases such as limiting the
13336 scope of the error diffusal @option{dither} to the rectangle that bounds the
13337 moving scene (it leads to more deterministic output if the scene doesn't change
13338 much, and as a result less moving noise and better GIF compression).
13341 Default is @var{none}.
13344 Take new palette for each output frame.
13346 @item alpha_threshold
13347 Sets the alpha threshold for transparency. Alpha values above this threshold
13348 will be treated as completely opaque, and values below this threshold will be
13349 treated as completely transparent.
13351 The option must be an integer value in the range [0,255]. Default is @var{128}.
13354 @subsection Examples
13358 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
13359 using @command{ffmpeg}:
13361 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
13365 @section perspective
13367 Correct perspective of video not recorded perpendicular to the screen.
13369 A description of the accepted parameters follows.
13380 Set coordinates expression for top left, top right, bottom left and bottom right corners.
13381 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
13382 If the @code{sense} option is set to @code{source}, then the specified points will be sent
13383 to the corners of the destination. If the @code{sense} option is set to @code{destination},
13384 then the corners of the source will be sent to the specified coordinates.
13386 The expressions can use the following variables:
13391 the width and height of video frame.
13395 Output frame count.
13398 @item interpolation
13399 Set interpolation for perspective correction.
13401 It accepts the following values:
13407 Default value is @samp{linear}.
13410 Set interpretation of coordinate options.
13412 It accepts the following values:
13416 Send point in the source specified by the given coordinates to
13417 the corners of the destination.
13419 @item 1, destination
13421 Send the corners of the source to the point in the destination specified
13422 by the given coordinates.
13424 Default value is @samp{source}.
13428 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
13430 It accepts the following values:
13433 only evaluate expressions once during the filter initialization or
13434 when a command is processed
13437 evaluate expressions for each incoming frame
13440 Default value is @samp{init}.
13445 Delay interlaced video by one field time so that the field order changes.
13447 The intended use is to fix PAL movies that have been captured with the
13448 opposite field order to the film-to-video transfer.
13450 A description of the accepted parameters follows.
13456 It accepts the following values:
13459 Capture field order top-first, transfer bottom-first.
13460 Filter will delay the bottom field.
13463 Capture field order bottom-first, transfer top-first.
13464 Filter will delay the top field.
13467 Capture and transfer with the same field order. This mode only exists
13468 for the documentation of the other options to refer to, but if you
13469 actually select it, the filter will faithfully do nothing.
13472 Capture field order determined automatically by field flags, transfer
13474 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
13475 basis using field flags. If no field information is available,
13476 then this works just like @samp{u}.
13479 Capture unknown or varying, transfer opposite.
13480 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
13481 analyzing the images and selecting the alternative that produces best
13482 match between the fields.
13485 Capture top-first, transfer unknown or varying.
13486 Filter selects among @samp{t} and @samp{p} using image analysis.
13489 Capture bottom-first, transfer unknown or varying.
13490 Filter selects among @samp{b} and @samp{p} using image analysis.
13493 Capture determined by field flags, transfer unknown or varying.
13494 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
13495 image analysis. If no field information is available, then this works just
13496 like @samp{U}. This is the default mode.
13499 Both capture and transfer unknown or varying.
13500 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
13504 @section pixdesctest
13506 Pixel format descriptor test filter, mainly useful for internal
13507 testing. The output video should be equal to the input video.
13511 format=monow, pixdesctest
13514 can be used to test the monowhite pixel format descriptor definition.
13518 Display sample values of color channels. Mainly useful for checking color
13519 and levels. Minimum supported resolution is 640x480.
13521 The filters accept the following options:
13525 Set scope X position, relative offset on X axis.
13528 Set scope Y position, relative offset on Y axis.
13537 Set window opacity. This window also holds statistics about pixel area.
13540 Set window X position, relative offset on X axis.
13543 Set window Y position, relative offset on Y axis.
13548 Enable the specified chain of postprocessing subfilters using libpostproc. This
13549 library should be automatically selected with a GPL build (@code{--enable-gpl}).
13550 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
13551 Each subfilter and some options have a short and a long name that can be used
13552 interchangeably, i.e. dr/dering are the same.
13554 The filters accept the following options:
13558 Set postprocessing subfilters string.
13561 All subfilters share common options to determine their scope:
13565 Honor the quality commands for this subfilter.
13568 Do chrominance filtering, too (default).
13571 Do luminance filtering only (no chrominance).
13574 Do chrominance filtering only (no luminance).
13577 These options can be appended after the subfilter name, separated by a '|'.
13579 Available subfilters are:
13582 @item hb/hdeblock[|difference[|flatness]]
13583 Horizontal deblocking filter
13586 Difference factor where higher values mean more deblocking (default: @code{32}).
13588 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13591 @item vb/vdeblock[|difference[|flatness]]
13592 Vertical deblocking filter
13595 Difference factor where higher values mean more deblocking (default: @code{32}).
13597 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13600 @item ha/hadeblock[|difference[|flatness]]
13601 Accurate horizontal deblocking filter
13604 Difference factor where higher values mean more deblocking (default: @code{32}).
13606 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13609 @item va/vadeblock[|difference[|flatness]]
13610 Accurate vertical deblocking filter
13613 Difference factor where higher values mean more deblocking (default: @code{32}).
13615 Flatness threshold where lower values mean more deblocking (default: @code{39}).
13619 The horizontal and vertical deblocking filters share the difference and
13620 flatness values so you cannot set different horizontal and vertical
13624 @item h1/x1hdeblock
13625 Experimental horizontal deblocking filter
13627 @item v1/x1vdeblock
13628 Experimental vertical deblocking filter
13633 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
13636 larger -> stronger filtering
13638 larger -> stronger filtering
13640 larger -> stronger filtering
13643 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
13646 Stretch luminance to @code{0-255}.
13649 @item lb/linblenddeint
13650 Linear blend deinterlacing filter that deinterlaces the given block by
13651 filtering all lines with a @code{(1 2 1)} filter.
13653 @item li/linipoldeint
13654 Linear interpolating deinterlacing filter that deinterlaces the given block by
13655 linearly interpolating every second line.
13657 @item ci/cubicipoldeint
13658 Cubic interpolating deinterlacing filter deinterlaces the given block by
13659 cubically interpolating every second line.
13661 @item md/mediandeint
13662 Median deinterlacing filter that deinterlaces the given block by applying a
13663 median filter to every second line.
13665 @item fd/ffmpegdeint
13666 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
13667 second line with a @code{(-1 4 2 4 -1)} filter.
13670 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
13671 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
13673 @item fq/forceQuant[|quantizer]
13674 Overrides the quantizer table from the input with the constant quantizer you
13682 Default pp filter combination (@code{hb|a,vb|a,dr|a})
13685 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
13688 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
13691 @subsection Examples
13695 Apply horizontal and vertical deblocking, deringing and automatic
13696 brightness/contrast:
13702 Apply default filters without brightness/contrast correction:
13708 Apply default filters and temporal denoiser:
13710 pp=default/tmpnoise|1|2|3
13714 Apply deblocking on luminance only, and switch vertical deblocking on or off
13715 automatically depending on available CPU time:
13722 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
13723 similar to spp = 6 with 7 point DCT, where only the center sample is
13726 The filter accepts the following options:
13730 Force a constant quantization parameter. It accepts an integer in range
13731 0 to 63. If not set, the filter will use the QP from the video stream
13735 Set thresholding mode. Available modes are:
13739 Set hard thresholding.
13741 Set soft thresholding (better de-ringing effect, but likely blurrier).
13743 Set medium thresholding (good results, default).
13747 @section premultiply
13748 Apply alpha premultiply effect to input video stream using first plane
13749 of second stream as alpha.
13751 Both streams must have same dimensions and same pixel format.
13753 The filter accepts the following option:
13757 Set which planes will be processed, unprocessed planes will be copied.
13758 By default value 0xf, all planes will be processed.
13761 Do not require 2nd input for processing, instead use alpha plane from input stream.
13765 Apply prewitt operator to input video stream.
13767 The filter accepts the following option:
13771 Set which planes will be processed, unprocessed planes will be copied.
13772 By default value 0xf, all planes will be processed.
13775 Set value which will be multiplied with filtered result.
13778 Set value which will be added to filtered result.
13781 @anchor{program_opencl}
13782 @section program_opencl
13784 Filter video using an OpenCL program.
13789 OpenCL program source file.
13792 Kernel name in program.
13795 Number of inputs to the filter. Defaults to 1.
13798 Size of output frames. Defaults to the same as the first input.
13802 The program source file must contain a kernel function with the given name,
13803 which will be run once for each plane of the output. Each run on a plane
13804 gets enqueued as a separate 2D global NDRange with one work-item for each
13805 pixel to be generated. The global ID offset for each work-item is therefore
13806 the coordinates of a pixel in the destination image.
13808 The kernel function needs to take the following arguments:
13811 Destination image, @var{__write_only image2d_t}.
13813 This image will become the output; the kernel should write all of it.
13815 Frame index, @var{unsigned int}.
13817 This is a counter starting from zero and increasing by one for each frame.
13819 Source images, @var{__read_only image2d_t}.
13821 These are the most recent images on each input. The kernel may read from
13822 them to generate the output, but they can't be written to.
13829 Copy the input to the output (output must be the same size as the input).
13831 __kernel void copy(__write_only image2d_t destination,
13832 unsigned int index,
13833 __read_only image2d_t source)
13835 const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
13837 int2 location = (int2)(get_global_id(0), get_global_id(1));
13839 float4 value = read_imagef(source, sampler, location);
13841 write_imagef(destination, location, value);
13846 Apply a simple transformation, rotating the input by an amount increasing
13847 with the index counter. Pixel values are linearly interpolated by the
13848 sampler, and the output need not have the same dimensions as the input.
13850 __kernel void rotate_image(__write_only image2d_t dst,
13851 unsigned int index,
13852 __read_only image2d_t src)
13854 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13855 CLK_FILTER_LINEAR);
13857 float angle = (float)index / 100.0f;
13859 float2 dst_dim = convert_float2(get_image_dim(dst));
13860 float2 src_dim = convert_float2(get_image_dim(src));
13862 float2 dst_cen = dst_dim / 2.0f;
13863 float2 src_cen = src_dim / 2.0f;
13865 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13867 float2 dst_pos = convert_float2(dst_loc) - dst_cen;
13869 cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
13870 sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
13872 src_pos = src_pos * src_dim / dst_dim;
13874 float2 src_loc = src_pos + src_cen;
13876 if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
13877 src_loc.x > src_dim.x || src_loc.y > src_dim.y)
13878 write_imagef(dst, dst_loc, 0.5f);
13880 write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
13885 Blend two inputs together, with the amount of each input used varying
13886 with the index counter.
13888 __kernel void blend_images(__write_only image2d_t dst,
13889 unsigned int index,
13890 __read_only image2d_t src1,
13891 __read_only image2d_t src2)
13893 const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
13894 CLK_FILTER_LINEAR);
13896 float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
13898 int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
13899 int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
13900 int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
13902 float4 val1 = read_imagef(src1, sampler, src1_loc);
13903 float4 val2 = read_imagef(src2, sampler, src2_loc);
13905 write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
13911 @section pseudocolor
13913 Alter frame colors in video with pseudocolors.
13915 This filter accept the following options:
13919 set pixel first component expression
13922 set pixel second component expression
13925 set pixel third component expression
13928 set pixel fourth component expression, corresponds to the alpha component
13931 set component to use as base for altering colors
13934 Each of them specifies the expression to use for computing the lookup table for
13935 the corresponding pixel component values.
13937 The expressions can contain the following constants and functions:
13942 The input width and height.
13945 The input value for the pixel component.
13947 @item ymin, umin, vmin, amin
13948 The minimum allowed component value.
13950 @item ymax, umax, vmax, amax
13951 The maximum allowed component value.
13954 All expressions default to "val".
13956 @subsection Examples
13960 Change too high luma values to gradient:
13962 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'"
13968 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
13969 Ratio) between two input videos.
13971 This filter takes in input two input videos, the first input is
13972 considered the "main" source and is passed unchanged to the
13973 output. The second input is used as a "reference" video for computing
13976 Both video inputs must have the same resolution and pixel format for
13977 this filter to work correctly. Also it assumes that both inputs
13978 have the same number of frames, which are compared one by one.
13980 The obtained average PSNR is printed through the logging system.
13982 The filter stores the accumulated MSE (mean squared error) of each
13983 frame, and at the end of the processing it is averaged across all frames
13984 equally, and the following formula is applied to obtain the PSNR:
13987 PSNR = 10*log10(MAX^2/MSE)
13990 Where MAX is the average of the maximum values of each component of the
13993 The description of the accepted parameters follows.
13996 @item stats_file, f
13997 If specified the filter will use the named file to save the PSNR of
13998 each individual frame. When filename equals "-" the data is sent to
14001 @item stats_version
14002 Specifies which version of the stats file format to use. Details of
14003 each format are written below.
14004 Default value is 1.
14006 @item stats_add_max
14007 Determines whether the max value is output to the stats log.
14008 Default value is 0.
14009 Requires stats_version >= 2. If this is set and stats_version < 2,
14010 the filter will return an error.
14013 This filter also supports the @ref{framesync} options.
14015 The file printed if @var{stats_file} is selected, contains a sequence of
14016 key/value pairs of the form @var{key}:@var{value} for each compared
14019 If a @var{stats_version} greater than 1 is specified, a header line precedes
14020 the list of per-frame-pair stats, with key value pairs following the frame
14021 format with the following parameters:
14024 @item psnr_log_version
14025 The version of the log file format. Will match @var{stats_version}.
14028 A comma separated list of the per-frame-pair parameters included in
14032 A description of each shown per-frame-pair parameter follows:
14036 sequential number of the input frame, starting from 1
14039 Mean Square Error pixel-by-pixel average difference of the compared
14040 frames, averaged over all the image components.
14042 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_b, mse_a
14043 Mean Square Error pixel-by-pixel average difference of the compared
14044 frames for the component specified by the suffix.
14046 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
14047 Peak Signal to Noise ratio of the compared frames for the component
14048 specified by the suffix.
14050 @item max_avg, max_y, max_u, max_v
14051 Maximum allowed value for each channel, and average over all
14057 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
14058 [main][ref] psnr="stats_file=stats.log" [out]
14061 On this example the input file being processed is compared with the
14062 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
14063 is stored in @file{stats.log}.
14068 Pulldown reversal (inverse telecine) filter, capable of handling mixed
14069 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
14072 The pullup filter is designed to take advantage of future context in making
14073 its decisions. This filter is stateless in the sense that it does not lock
14074 onto a pattern to follow, but it instead looks forward to the following
14075 fields in order to identify matches and rebuild progressive frames.
14077 To produce content with an even framerate, insert the fps filter after
14078 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
14079 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
14081 The filter accepts the following options:
14088 These options set the amount of "junk" to ignore at the left, right, top, and
14089 bottom of the image, respectively. Left and right are in units of 8 pixels,
14090 while top and bottom are in units of 2 lines.
14091 The default is 8 pixels on each side.
14094 Set the strict breaks. Setting this option to 1 will reduce the chances of
14095 filter generating an occasional mismatched frame, but it may also cause an
14096 excessive number of frames to be dropped during high motion sequences.
14097 Conversely, setting it to -1 will make filter match fields more easily.
14098 This may help processing of video where there is slight blurring between
14099 the fields, but may also cause there to be interlaced frames in the output.
14100 Default value is @code{0}.
14103 Set the metric plane to use. It accepts the following values:
14109 Use chroma blue plane.
14112 Use chroma red plane.
14115 This option may be set to use chroma plane instead of the default luma plane
14116 for doing filter's computations. This may improve accuracy on very clean
14117 source material, but more likely will decrease accuracy, especially if there
14118 is chroma noise (rainbow effect) or any grayscale video.
14119 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
14120 load and make pullup usable in realtime on slow machines.
14123 For best results (without duplicated frames in the output file) it is
14124 necessary to change the output frame rate. For example, to inverse
14125 telecine NTSC input:
14127 ffmpeg -i input -vf pullup -r 24000/1001 ...
14132 Change video quantization parameters (QP).
14134 The filter accepts the following option:
14138 Set expression for quantization parameter.
14141 The expression is evaluated through the eval API and can contain, among others,
14142 the following constants:
14146 1 if index is not 129, 0 otherwise.
14149 Sequential index starting from -129 to 128.
14152 @subsection Examples
14156 Some equation like:
14164 Flush video frames from internal cache of frames into a random order.
14165 No frame is discarded.
14166 Inspired by @ref{frei0r} nervous filter.
14170 Set size in number of frames of internal cache, in range from @code{2} to
14171 @code{512}. Default is @code{30}.
14174 Set seed for random number generator, must be an integer included between
14175 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
14176 less than @code{0}, the filter will try to use a good random seed on a
14180 @section readeia608
14182 Read closed captioning (EIA-608) information from the top lines of a video frame.
14184 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
14185 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
14186 with EIA-608 data (starting from 0). A description of each metadata value follows:
14189 @item lavfi.readeia608.X.cc
14190 The two bytes stored as EIA-608 data (printed in hexadecimal).
14192 @item lavfi.readeia608.X.line
14193 The number of the line on which the EIA-608 data was identified and read.
14196 This filter accepts the following options:
14200 Set the line to start scanning for EIA-608 data. Default is @code{0}.
14203 Set the line to end scanning for EIA-608 data. Default is @code{29}.
14206 Set minimal acceptable amplitude change for sync codes detection.
14207 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
14210 Set the ratio of width reserved for sync code detection.
14211 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
14214 Set the max peaks height difference for sync code detection.
14215 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14218 Set max peaks period difference for sync code detection.
14219 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
14222 Set the first two max start code bits differences.
14223 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
14226 Set the minimum ratio of bits height compared to 3rd start code bit.
14227 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
14230 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
14233 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
14236 Enable checking the parity bit. In the event of a parity error, the filter will output
14237 @code{0x00} for that character. Default is false.
14240 @subsection Examples
14244 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
14246 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
14252 Read vertical interval timecode (VITC) information from the top lines of a
14255 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
14256 timecode value, if a valid timecode has been detected. Further metadata key
14257 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
14258 timecode data has been found or not.
14260 This filter accepts the following options:
14264 Set the maximum number of lines to scan for VITC data. If the value is set to
14265 @code{-1} the full video frame is scanned. Default is @code{45}.
14268 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
14269 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
14272 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
14273 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
14276 @subsection Examples
14280 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
14281 draw @code{--:--:--:--} as a placeholder:
14283 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
14289 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
14291 Destination pixel at position (X, Y) will be picked from source (x, y) position
14292 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
14293 value for pixel will be used for destination pixel.
14295 Xmap and Ymap input video streams must be of same dimensions. Output video stream
14296 will have Xmap/Ymap video stream dimensions.
14297 Xmap and Ymap input video streams are 16bit depth, single channel.
14299 @section removegrain
14301 The removegrain filter is a spatial denoiser for progressive video.
14305 Set mode for the first plane.
14308 Set mode for the second plane.
14311 Set mode for the third plane.
14314 Set mode for the fourth plane.
14317 Range of mode is from 0 to 24. Description of each mode follows:
14321 Leave input plane unchanged. Default.
14324 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
14327 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
14330 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
14333 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
14334 This is equivalent to a median filter.
14337 Line-sensitive clipping giving the minimal change.
14340 Line-sensitive clipping, intermediate.
14343 Line-sensitive clipping, intermediate.
14346 Line-sensitive clipping, intermediate.
14349 Line-sensitive clipping on a line where the neighbours pixels are the closest.
14352 Replaces the target pixel with the closest neighbour.
14355 [1 2 1] horizontal and vertical kernel blur.
14361 Bob mode, interpolates top field from the line where the neighbours
14362 pixels are the closest.
14365 Bob mode, interpolates bottom field from the line where the neighbours
14366 pixels are the closest.
14369 Bob mode, interpolates top field. Same as 13 but with a more complicated
14370 interpolation formula.
14373 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
14374 interpolation formula.
14377 Clips the pixel with the minimum and maximum of respectively the maximum and
14378 minimum of each pair of opposite neighbour pixels.
14381 Line-sensitive clipping using opposite neighbours whose greatest distance from
14382 the current pixel is minimal.
14385 Replaces the pixel with the average of its 8 neighbours.
14388 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
14391 Clips pixels using the averages of opposite neighbour.
14394 Same as mode 21 but simpler and faster.
14397 Small edge and halo removal, but reputed useless.
14403 @section removelogo
14405 Suppress a TV station logo, using an image file to determine which
14406 pixels comprise the logo. It works by filling in the pixels that
14407 comprise the logo with neighboring pixels.
14409 The filter accepts the following options:
14413 Set the filter bitmap file, which can be any image format supported by
14414 libavformat. The width and height of the image file must match those of the
14415 video stream being processed.
14418 Pixels in the provided bitmap image with a value of zero are not
14419 considered part of the logo, non-zero pixels are considered part of
14420 the logo. If you use white (255) for the logo and black (0) for the
14421 rest, you will be safe. For making the filter bitmap, it is
14422 recommended to take a screen capture of a black frame with the logo
14423 visible, and then using a threshold filter followed by the erode
14424 filter once or twice.
14426 If needed, little splotches can be fixed manually. Remember that if
14427 logo pixels are not covered, the filter quality will be much
14428 reduced. Marking too many pixels as part of the logo does not hurt as
14429 much, but it will increase the amount of blurring needed to cover over
14430 the image and will destroy more information than necessary, and extra
14431 pixels will slow things down on a large logo.
14433 @section repeatfields
14435 This filter uses the repeat_field flag from the Video ES headers and hard repeats
14436 fields based on its value.
14440 Reverse a video clip.
14442 Warning: This filter requires memory to buffer the entire clip, so trimming
14445 @subsection Examples
14449 Take the first 5 seconds of a clip, and reverse it.
14456 Shift R/G/B/A pixels horizontally and/or vertically.
14458 The filter accepts the following options:
14461 Set amount to shift red horizontally.
14463 Set amount to shift red vertically.
14465 Set amount to shift green horizontally.
14467 Set amount to shift green vertically.
14469 Set amount to shift blue horizontally.
14471 Set amount to shift blue vertically.
14473 Set amount to shift alpha horizontally.
14475 Set amount to shift alpha vertically.
14477 Set edge mode, can be @var{smear}, default, or @var{warp}.
14481 Apply roberts cross operator to input video stream.
14483 The filter accepts the following option:
14487 Set which planes will be processed, unprocessed planes will be copied.
14488 By default value 0xf, all planes will be processed.
14491 Set value which will be multiplied with filtered result.
14494 Set value which will be added to filtered result.
14499 Rotate video by an arbitrary angle expressed in radians.
14501 The filter accepts the following options:
14503 A description of the optional parameters follows.
14506 Set an expression for the angle by which to rotate the input video
14507 clockwise, expressed as a number of radians. A negative value will
14508 result in a counter-clockwise rotation. By default it is set to "0".
14510 This expression is evaluated for each frame.
14513 Set the output width expression, default value is "iw".
14514 This expression is evaluated just once during configuration.
14517 Set the output height expression, default value is "ih".
14518 This expression is evaluated just once during configuration.
14521 Enable bilinear interpolation if set to 1, a value of 0 disables
14522 it. Default value is 1.
14525 Set the color used to fill the output area not covered by the rotated
14526 image. For the general syntax of this option, check the
14527 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
14528 If the special value "none" is selected then no
14529 background is printed (useful for example if the background is never shown).
14531 Default value is "black".
14534 The expressions for the angle and the output size can contain the
14535 following constants and functions:
14539 sequential number of the input frame, starting from 0. It is always NAN
14540 before the first frame is filtered.
14543 time in seconds of the input frame, it is set to 0 when the filter is
14544 configured. It is always NAN before the first frame is filtered.
14548 horizontal and vertical chroma subsample values. For example for the
14549 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14553 the input video width and height
14557 the output width and height, that is the size of the padded area as
14558 specified by the @var{width} and @var{height} expressions
14562 the minimal width/height required for completely containing the input
14563 video rotated by @var{a} radians.
14565 These are only available when computing the @option{out_w} and
14566 @option{out_h} expressions.
14569 @subsection Examples
14573 Rotate the input by PI/6 radians clockwise:
14579 Rotate the input by PI/6 radians counter-clockwise:
14585 Rotate the input by 45 degrees clockwise:
14591 Apply a constant rotation with period T, starting from an angle of PI/3:
14593 rotate=PI/3+2*PI*t/T
14597 Make the input video rotation oscillating with a period of T
14598 seconds and an amplitude of A radians:
14600 rotate=A*sin(2*PI/T*t)
14604 Rotate the video, output size is chosen so that the whole rotating
14605 input video is always completely contained in the output:
14607 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
14611 Rotate the video, reduce the output size so that no background is ever
14614 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
14618 @subsection Commands
14620 The filter supports the following commands:
14624 Set the angle expression.
14625 The command accepts the same syntax of the corresponding option.
14627 If the specified expression is not valid, it is kept at its current
14633 Apply Shape Adaptive Blur.
14635 The filter accepts the following options:
14638 @item luma_radius, lr
14639 Set luma blur filter strength, must be a value in range 0.1-4.0, default
14640 value is 1.0. A greater value will result in a more blurred image, and
14641 in slower processing.
14643 @item luma_pre_filter_radius, lpfr
14644 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
14647 @item luma_strength, ls
14648 Set luma maximum difference between pixels to still be considered, must
14649 be a value in the 0.1-100.0 range, default value is 1.0.
14651 @item chroma_radius, cr
14652 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
14653 greater value will result in a more blurred image, and in slower
14656 @item chroma_pre_filter_radius, cpfr
14657 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
14659 @item chroma_strength, cs
14660 Set chroma maximum difference between pixels to still be considered,
14661 must be a value in the -0.9-100.0 range.
14664 Each chroma option value, if not explicitly specified, is set to the
14665 corresponding luma option value.
14670 Scale (resize) the input video, using the libswscale library.
14672 The scale filter forces the output display aspect ratio to be the same
14673 of the input, by changing the output sample aspect ratio.
14675 If the input image format is different from the format requested by
14676 the next filter, the scale filter will convert the input to the
14679 @subsection Options
14680 The filter accepts the following options, or any of the options
14681 supported by the libswscale scaler.
14683 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
14684 the complete list of scaler options.
14689 Set the output video dimension expression. Default value is the input
14692 If the @var{width} or @var{w} value is 0, the input width is used for
14693 the output. If the @var{height} or @var{h} value is 0, the input height
14694 is used for the output.
14696 If one and only one of the values is -n with n >= 1, the scale filter
14697 will use a value that maintains the aspect ratio of the input image,
14698 calculated from the other specified dimension. After that it will,
14699 however, make sure that the calculated dimension is divisible by n and
14700 adjust the value if necessary.
14702 If both values are -n with n >= 1, the behavior will be identical to
14703 both values being set to 0 as previously detailed.
14705 See below for the list of accepted constants for use in the dimension
14709 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
14713 Only evaluate expressions once during the filter initialization or when a command is processed.
14716 Evaluate expressions for each incoming frame.
14720 Default value is @samp{init}.
14724 Set the interlacing mode. It accepts the following values:
14728 Force interlaced aware scaling.
14731 Do not apply interlaced scaling.
14734 Select interlaced aware scaling depending on whether the source frames
14735 are flagged as interlaced or not.
14738 Default value is @samp{0}.
14741 Set libswscale scaling flags. See
14742 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14743 complete list of values. If not explicitly specified the filter applies
14747 @item param0, param1
14748 Set libswscale input parameters for scaling algorithms that need them. See
14749 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
14750 complete documentation. If not explicitly specified the filter applies
14756 Set the video size. For the syntax of this option, check the
14757 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
14759 @item in_color_matrix
14760 @item out_color_matrix
14761 Set in/output YCbCr color space type.
14763 This allows the autodetected value to be overridden as well as allows forcing
14764 a specific value used for the output and encoder.
14766 If not specified, the color space type depends on the pixel format.
14772 Choose automatically.
14775 Format conforming to International Telecommunication Union (ITU)
14776 Recommendation BT.709.
14779 Set color space conforming to the United States Federal Communications
14780 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
14783 Set color space conforming to:
14787 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
14790 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
14793 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
14798 Set color space conforming to SMPTE ST 240:1999.
14803 Set in/output YCbCr sample range.
14805 This allows the autodetected value to be overridden as well as allows forcing
14806 a specific value used for the output and encoder. If not specified, the
14807 range depends on the pixel format. Possible values:
14811 Choose automatically.
14814 Set full range (0-255 in case of 8-bit luma).
14816 @item mpeg/limited/tv
14817 Set "MPEG" range (16-235 in case of 8-bit luma).
14820 @item force_original_aspect_ratio
14821 Enable decreasing or increasing output video width or height if necessary to
14822 keep the original aspect ratio. Possible values:
14826 Scale the video as specified and disable this feature.
14829 The output video dimensions will automatically be decreased if needed.
14832 The output video dimensions will automatically be increased if needed.
14836 One useful instance of this option is that when you know a specific device's
14837 maximum allowed resolution, you can use this to limit the output video to
14838 that, while retaining the aspect ratio. For example, device A allows
14839 1280x720 playback, and your video is 1920x800. Using this option (set it to
14840 decrease) and specifying 1280x720 to the command line makes the output
14843 Please note that this is a different thing than specifying -1 for @option{w}
14844 or @option{h}, you still need to specify the output resolution for this option
14849 The values of the @option{w} and @option{h} options are expressions
14850 containing the following constants:
14855 The input width and height
14859 These are the same as @var{in_w} and @var{in_h}.
14863 The output (scaled) width and height
14867 These are the same as @var{out_w} and @var{out_h}
14870 The same as @var{iw} / @var{ih}
14873 input sample aspect ratio
14876 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
14880 horizontal and vertical input chroma subsample values. For example for the
14881 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14885 horizontal and vertical output chroma subsample values. For example for the
14886 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
14889 @subsection Examples
14893 Scale the input video to a size of 200x100
14898 This is equivalent to:
14909 Specify a size abbreviation for the output size:
14914 which can also be written as:
14920 Scale the input to 2x:
14922 scale=w=2*iw:h=2*ih
14926 The above is the same as:
14928 scale=2*in_w:2*in_h
14932 Scale the input to 2x with forced interlaced scaling:
14934 scale=2*iw:2*ih:interl=1
14938 Scale the input to half size:
14940 scale=w=iw/2:h=ih/2
14944 Increase the width, and set the height to the same size:
14950 Seek Greek harmony:
14957 Increase the height, and set the width to 3/2 of the height:
14959 scale=w=3/2*oh:h=3/5*ih
14963 Increase the size, making the size a multiple of the chroma
14966 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
14970 Increase the width to a maximum of 500 pixels,
14971 keeping the same aspect ratio as the input:
14973 scale=w='min(500\, iw*3/2):h=-1'
14977 Make pixels square by combining scale and setsar:
14979 scale='trunc(ih*dar):ih',setsar=1/1
14983 Make pixels square by combining scale and setsar,
14984 making sure the resulting resolution is even (required by some codecs):
14986 scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
14990 @subsection Commands
14992 This filter supports the following commands:
14996 Set the output video dimension expression.
14997 The command accepts the same syntax of the corresponding option.
14999 If the specified expression is not valid, it is kept at its current
15005 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
15006 format conversion on CUDA video frames. Setting the output width and height
15007 works in the same way as for the @var{scale} filter.
15009 The following additional options are accepted:
15012 The pixel format of the output CUDA frames. If set to the string "same" (the
15013 default), the input format will be kept. Note that automatic format negotiation
15014 and conversion is not yet supported for hardware frames
15017 The interpolation algorithm used for resizing. One of the following:
15024 @item cubic2p_bspline
15025 2-parameter cubic (B=1, C=0)
15027 @item cubic2p_catmullrom
15028 2-parameter cubic (B=0, C=1/2)
15030 @item cubic2p_b05c03
15031 2-parameter cubic (B=1/2, C=3/10)
15043 Scale (resize) the input video, based on a reference video.
15045 See the scale filter for available options, scale2ref supports the same but
15046 uses the reference video instead of the main input as basis. scale2ref also
15047 supports the following additional constants for the @option{w} and
15048 @option{h} options:
15053 The main input video's width and height
15056 The same as @var{main_w} / @var{main_h}
15059 The main input video's sample aspect ratio
15061 @item main_dar, mdar
15062 The main input video's display aspect ratio. Calculated from
15063 @code{(main_w / main_h) * main_sar}.
15067 The main input video's horizontal and vertical chroma subsample values.
15068 For example for the pixel format "yuv422p" @var{hsub} is 2 and @var{vsub}
15072 @subsection Examples
15076 Scale a subtitle stream (b) to match the main video (a) in size before overlaying
15078 'scale2ref[b][a];[a][b]overlay'
15082 @anchor{selectivecolor}
15083 @section selectivecolor
15085 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
15086 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
15087 by the "purity" of the color (that is, how saturated it already is).
15089 This filter is similar to the Adobe Photoshop Selective Color tool.
15091 The filter accepts the following options:
15094 @item correction_method
15095 Select color correction method.
15097 Available values are:
15100 Specified adjustments are applied "as-is" (added/subtracted to original pixel
15103 Specified adjustments are relative to the original component value.
15105 Default is @code{absolute}.
15107 Adjustments for red pixels (pixels where the red component is the maximum)
15109 Adjustments for yellow pixels (pixels where the blue component is the minimum)
15111 Adjustments for green pixels (pixels where the green component is the maximum)
15113 Adjustments for cyan pixels (pixels where the red component is the minimum)
15115 Adjustments for blue pixels (pixels where the blue component is the maximum)
15117 Adjustments for magenta pixels (pixels where the green component is the minimum)
15119 Adjustments for white pixels (pixels where all components are greater than 128)
15121 Adjustments for all pixels except pure black and pure white
15123 Adjustments for black pixels (pixels where all components are lesser than 128)
15125 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
15128 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
15129 4 space separated floating point adjustment values in the [-1,1] range,
15130 respectively to adjust the amount of cyan, magenta, yellow and black for the
15131 pixels of its range.
15133 @subsection Examples
15137 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
15138 increase magenta by 27% in blue areas:
15140 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
15144 Use a Photoshop selective color preset:
15146 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
15150 @anchor{separatefields}
15151 @section separatefields
15153 The @code{separatefields} takes a frame-based video input and splits
15154 each frame into its components fields, producing a new half height clip
15155 with twice the frame rate and twice the frame count.
15157 This filter use field-dominance information in frame to decide which
15158 of each pair of fields to place first in the output.
15159 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
15161 @section setdar, setsar
15163 The @code{setdar} filter sets the Display Aspect Ratio for the filter
15166 This is done by changing the specified Sample (aka Pixel) Aspect
15167 Ratio, according to the following equation:
15169 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
15172 Keep in mind that the @code{setdar} filter does not modify the pixel
15173 dimensions of the video frame. Also, the display aspect ratio set by
15174 this filter may be changed by later filters in the filterchain,
15175 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
15178 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
15179 the filter output video.
15181 Note that as a consequence of the application of this filter, the
15182 output display aspect ratio will change according to the equation
15185 Keep in mind that the sample aspect ratio set by the @code{setsar}
15186 filter may be changed by later filters in the filterchain, e.g. if
15187 another "setsar" or a "setdar" filter is applied.
15189 It accepts the following parameters:
15192 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
15193 Set the aspect ratio used by the filter.
15195 The parameter can be a floating point number string, an expression, or
15196 a string of the form @var{num}:@var{den}, where @var{num} and
15197 @var{den} are the numerator and denominator of the aspect ratio. If
15198 the parameter is not specified, it is assumed the value "0".
15199 In case the form "@var{num}:@var{den}" is used, the @code{:} character
15203 Set the maximum integer value to use for expressing numerator and
15204 denominator when reducing the expressed aspect ratio to a rational.
15205 Default value is @code{100}.
15209 The parameter @var{sar} is an expression containing
15210 the following constants:
15214 These are approximated values for the mathematical constants e
15215 (Euler's number), pi (Greek pi), and phi (the golden ratio).
15218 The input width and height.
15221 These are the same as @var{w} / @var{h}.
15224 The input sample aspect ratio.
15227 The input display aspect ratio. It is the same as
15228 (@var{w} / @var{h}) * @var{sar}.
15231 Horizontal and vertical chroma subsample values. For example, for the
15232 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15235 @subsection Examples
15240 To change the display aspect ratio to 16:9, specify one of the following:
15247 To change the sample aspect ratio to 10:11, specify:
15253 To set a display aspect ratio of 16:9, and specify a maximum integer value of
15254 1000 in the aspect ratio reduction, use the command:
15256 setdar=ratio=16/9:max=1000
15264 Force field for the output video frame.
15266 The @code{setfield} filter marks the interlace type field for the
15267 output frames. It does not change the input frame, but only sets the
15268 corresponding property, which affects how the frame is treated by
15269 following filters (e.g. @code{fieldorder} or @code{yadif}).
15271 The filter accepts the following options:
15276 Available values are:
15280 Keep the same field property.
15283 Mark the frame as bottom-field-first.
15286 Mark the frame as top-field-first.
15289 Mark the frame as progressive.
15296 Force frame parameter for the output video frame.
15298 The @code{setparams} filter marks interlace and color range for the
15299 output frames. It does not change the input frame, but only sets the
15300 corresponding property, which affects how the frame is treated by
15305 Available values are:
15309 Keep the same field property (default).
15312 Mark the frame as bottom-field-first.
15315 Mark the frame as top-field-first.
15318 Mark the frame as progressive.
15322 Available values are:
15326 Keep the same color range property (default).
15328 @item unspecified, unknown
15329 Mark the frame as unspecified color range.
15331 @item limited, tv, mpeg
15332 Mark the frame as limited range.
15334 @item full, pc, jpeg
15335 Mark the frame as full range.
15338 @item color_primaries
15339 Set the color primaries.
15340 Available values are:
15344 Keep the same color primaries property (default).
15361 Set the color transfer.
15362 Available values are:
15366 Keep the same color trc property (default).
15388 Set the colorspace.
15389 Available values are:
15393 Keep the same colorspace property (default).
15406 @item chroma-derived-nc
15407 @item chroma-derived-c
15414 Show a line containing various information for each input video frame.
15415 The input video is not modified.
15417 This filter supports the following options:
15421 Calculate checksums of each plane. By default enabled.
15424 The shown line contains a sequence of key/value pairs of the form
15425 @var{key}:@var{value}.
15427 The following values are shown in the output:
15431 The (sequential) number of the input frame, starting from 0.
15434 The Presentation TimeStamp of the input frame, expressed as a number of
15435 time base units. The time base unit depends on the filter input pad.
15438 The Presentation TimeStamp of the input frame, expressed as a number of
15442 The position of the frame in the input stream, or -1 if this information is
15443 unavailable and/or meaningless (for example in case of synthetic video).
15446 The pixel format name.
15449 The sample aspect ratio of the input frame, expressed in the form
15450 @var{num}/@var{den}.
15453 The size of the input frame. For the syntax of this option, check the
15454 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15457 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
15458 for bottom field first).
15461 This is 1 if the frame is a key frame, 0 otherwise.
15464 The picture type of the input frame ("I" for an I-frame, "P" for a
15465 P-frame, "B" for a B-frame, or "?" for an unknown type).
15466 Also refer to the documentation of the @code{AVPictureType} enum and of
15467 the @code{av_get_picture_type_char} function defined in
15468 @file{libavutil/avutil.h}.
15471 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
15473 @item plane_checksum
15474 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
15475 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
15478 @section showpalette
15480 Displays the 256 colors palette of each frame. This filter is only relevant for
15481 @var{pal8} pixel format frames.
15483 It accepts the following option:
15487 Set the size of the box used to represent one palette color entry. Default is
15488 @code{30} (for a @code{30x30} pixel box).
15491 @section shuffleframes
15493 Reorder and/or duplicate and/or drop video frames.
15495 It accepts the following parameters:
15499 Set the destination indexes of input frames.
15500 This is space or '|' separated list of indexes that maps input frames to output
15501 frames. Number of indexes also sets maximal value that each index may have.
15502 '-1' index have special meaning and that is to drop frame.
15505 The first frame has the index 0. The default is to keep the input unchanged.
15507 @subsection Examples
15511 Swap second and third frame of every three frames of the input:
15513 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
15517 Swap 10th and 1st frame of every ten frames of the input:
15519 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
15523 @section shuffleplanes
15525 Reorder and/or duplicate video planes.
15527 It accepts the following parameters:
15532 The index of the input plane to be used as the first output plane.
15535 The index of the input plane to be used as the second output plane.
15538 The index of the input plane to be used as the third output plane.
15541 The index of the input plane to be used as the fourth output plane.
15545 The first plane has the index 0. The default is to keep the input unchanged.
15547 @subsection Examples
15551 Swap the second and third planes of the input:
15553 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
15557 @anchor{signalstats}
15558 @section signalstats
15559 Evaluate various visual metrics that assist in determining issues associated
15560 with the digitization of analog video media.
15562 By default the filter will log these metadata values:
15566 Display the minimal Y value contained within the input frame. Expressed in
15570 Display the Y value at the 10% percentile within the input frame. Expressed in
15574 Display the average Y value within the input frame. Expressed in range of
15578 Display the Y value at the 90% percentile within the input frame. Expressed in
15582 Display the maximum Y value contained within the input frame. Expressed in
15586 Display the minimal U value contained within the input frame. Expressed in
15590 Display the U value at the 10% percentile within the input frame. Expressed in
15594 Display the average U value within the input frame. Expressed in range of
15598 Display the U value at the 90% percentile within the input frame. Expressed in
15602 Display the maximum U value contained within the input frame. Expressed in
15606 Display the minimal V value contained within the input frame. Expressed in
15610 Display the V value at the 10% percentile within the input frame. Expressed in
15614 Display the average V value within the input frame. Expressed in range of
15618 Display the V value at the 90% percentile within the input frame. Expressed in
15622 Display the maximum V value contained within the input frame. Expressed in
15626 Display the minimal saturation value contained within the input frame.
15627 Expressed in range of [0-~181.02].
15630 Display the saturation value at the 10% percentile within the input frame.
15631 Expressed in range of [0-~181.02].
15634 Display the average saturation value within the input frame. Expressed in range
15638 Display the saturation value at the 90% percentile within the input frame.
15639 Expressed in range of [0-~181.02].
15642 Display the maximum saturation value contained within the input frame.
15643 Expressed in range of [0-~181.02].
15646 Display the median value for hue within the input frame. Expressed in range of
15650 Display the average value for hue within the input frame. Expressed in range of
15654 Display the average of sample value difference between all values of the Y
15655 plane in the current frame and corresponding values of the previous input frame.
15656 Expressed in range of [0-255].
15659 Display the average of sample value difference between all values of the U
15660 plane in the current frame and corresponding values of the previous input frame.
15661 Expressed in range of [0-255].
15664 Display the average of sample value difference between all values of the V
15665 plane in the current frame and corresponding values of the previous input frame.
15666 Expressed in range of [0-255].
15669 Display bit depth of Y plane in current frame.
15670 Expressed in range of [0-16].
15673 Display bit depth of U plane in current frame.
15674 Expressed in range of [0-16].
15677 Display bit depth of V plane in current frame.
15678 Expressed in range of [0-16].
15681 The filter accepts the following options:
15687 @option{stat} specify an additional form of image analysis.
15688 @option{out} output video with the specified type of pixel highlighted.
15690 Both options accept the following values:
15694 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
15695 unlike the neighboring pixels of the same field. Examples of temporal outliers
15696 include the results of video dropouts, head clogs, or tape tracking issues.
15699 Identify @var{vertical line repetition}. Vertical line repetition includes
15700 similar rows of pixels within a frame. In born-digital video vertical line
15701 repetition is common, but this pattern is uncommon in video digitized from an
15702 analog source. When it occurs in video that results from the digitization of an
15703 analog source it can indicate concealment from a dropout compensator.
15706 Identify pixels that fall outside of legal broadcast range.
15710 Set the highlight color for the @option{out} option. The default color is
15714 @subsection Examples
15718 Output data of various video metrics:
15720 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
15724 Output specific data about the minimum and maximum values of the Y plane per frame:
15726 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
15730 Playback video while highlighting pixels that are outside of broadcast range in red.
15732 ffplay example.mov -vf signalstats="out=brng:color=red"
15736 Playback video with signalstats metadata drawn over the frame.
15738 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
15741 The contents of signalstat_drawtext.txt used in the command are:
15744 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
15745 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
15746 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
15747 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
15755 Calculates the MPEG-7 Video Signature. The filter can handle more than one
15756 input. In this case the matching between the inputs can be calculated additionally.
15757 The filter always passes through the first input. The signature of each stream can
15758 be written into a file.
15760 It accepts the following options:
15764 Enable or disable the matching process.
15766 Available values are:
15770 Disable the calculation of a matching (default).
15772 Calculate the matching for the whole video and output whether the whole video
15773 matches or only parts.
15775 Calculate only until a matching is found or the video ends. Should be faster in
15780 Set the number of inputs. The option value must be a non negative integer.
15781 Default value is 1.
15784 Set the path to which the output is written. If there is more than one input,
15785 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
15786 integer), that will be replaced with the input number. If no filename is
15787 specified, no output will be written. This is the default.
15790 Choose the output format.
15792 Available values are:
15796 Use the specified binary representation (default).
15798 Use the specified xml representation.
15802 Set threshold to detect one word as similar. The option value must be an integer
15803 greater than zero. The default value is 9000.
15806 Set threshold to detect all words as similar. The option value must be an integer
15807 greater than zero. The default value is 60000.
15810 Set threshold to detect frames as similar. The option value must be an integer
15811 greater than zero. The default value is 116.
15814 Set the minimum length of a sequence in frames to recognize it as matching
15815 sequence. The option value must be a non negative integer value.
15816 The default value is 0.
15819 Set the minimum relation, that matching frames to all frames must have.
15820 The option value must be a double value between 0 and 1. The default value is 0.5.
15823 @subsection Examples
15827 To calculate the signature of an input video and store it in signature.bin:
15829 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
15833 To detect whether two videos match and store the signatures in XML format in
15834 signature0.xml and signature1.xml:
15836 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 -
15844 Blur the input video without impacting the outlines.
15846 It accepts the following options:
15849 @item luma_radius, lr
15850 Set the luma radius. The option value must be a float number in
15851 the range [0.1,5.0] that specifies the variance of the gaussian filter
15852 used to blur the image (slower if larger). Default value is 1.0.
15854 @item luma_strength, ls
15855 Set the luma strength. The option value must be a float number
15856 in the range [-1.0,1.0] that configures the blurring. A value included
15857 in [0.0,1.0] will blur the image whereas a value included in
15858 [-1.0,0.0] will sharpen the image. Default value is 1.0.
15860 @item luma_threshold, lt
15861 Set the luma threshold used as a coefficient to determine
15862 whether a pixel should be blurred or not. The option value must be an
15863 integer in the range [-30,30]. A value of 0 will filter all the image,
15864 a value included in [0,30] will filter flat areas and a value included
15865 in [-30,0] will filter edges. Default value is 0.
15867 @item chroma_radius, cr
15868 Set the chroma radius. The option value must be a float number in
15869 the range [0.1,5.0] that specifies the variance of the gaussian filter
15870 used to blur the image (slower if larger). Default value is @option{luma_radius}.
15872 @item chroma_strength, cs
15873 Set the chroma strength. The option value must be a float number
15874 in the range [-1.0,1.0] that configures the blurring. A value included
15875 in [0.0,1.0] will blur the image whereas a value included in
15876 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
15878 @item chroma_threshold, ct
15879 Set the chroma threshold used as a coefficient to determine
15880 whether a pixel should be blurred or not. The option value must be an
15881 integer in the range [-30,30]. A value of 0 will filter all the image,
15882 a value included in [0,30] will filter flat areas and a value included
15883 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
15886 If a chroma option is not explicitly set, the corresponding luma value
15891 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
15893 This filter takes in input two input videos, the first input is
15894 considered the "main" source and is passed unchanged to the
15895 output. The second input is used as a "reference" video for computing
15898 Both video inputs must have the same resolution and pixel format for
15899 this filter to work correctly. Also it assumes that both inputs
15900 have the same number of frames, which are compared one by one.
15902 The filter stores the calculated SSIM of each frame.
15904 The description of the accepted parameters follows.
15907 @item stats_file, f
15908 If specified the filter will use the named file to save the SSIM of
15909 each individual frame. When filename equals "-" the data is sent to
15913 The file printed if @var{stats_file} is selected, contains a sequence of
15914 key/value pairs of the form @var{key}:@var{value} for each compared
15917 A description of each shown parameter follows:
15921 sequential number of the input frame, starting from 1
15923 @item Y, U, V, R, G, B
15924 SSIM of the compared frames for the component specified by the suffix.
15927 SSIM of the compared frames for the whole frame.
15930 Same as above but in dB representation.
15933 This filter also supports the @ref{framesync} options.
15937 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
15938 [main][ref] ssim="stats_file=stats.log" [out]
15941 On this example the input file being processed is compared with the
15942 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
15943 is stored in @file{stats.log}.
15945 Another example with both psnr and ssim at same time:
15947 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
15952 Convert between different stereoscopic image formats.
15954 The filters accept the following options:
15958 Set stereoscopic image format of input.
15960 Available values for input image formats are:
15963 side by side parallel (left eye left, right eye right)
15966 side by side crosseye (right eye left, left eye right)
15969 side by side parallel with half width resolution
15970 (left eye left, right eye right)
15973 side by side crosseye with half width resolution
15974 (right eye left, left eye right)
15977 above-below (left eye above, right eye below)
15980 above-below (right eye above, left eye below)
15983 above-below with half height resolution
15984 (left eye above, right eye below)
15987 above-below with half height resolution
15988 (right eye above, left eye below)
15991 alternating frames (left eye first, right eye second)
15994 alternating frames (right eye first, left eye second)
15997 interleaved rows (left eye has top row, right eye starts on next row)
16000 interleaved rows (right eye has top row, left eye starts on next row)
16003 interleaved columns, left eye first
16006 interleaved columns, right eye first
16008 Default value is @samp{sbsl}.
16012 Set stereoscopic image format of output.
16016 side by side parallel (left eye left, right eye right)
16019 side by side crosseye (right eye left, left eye right)
16022 side by side parallel with half width resolution
16023 (left eye left, right eye right)
16026 side by side crosseye with half width resolution
16027 (right eye left, left eye right)
16030 above-below (left eye above, right eye below)
16033 above-below (right eye above, left eye below)
16036 above-below with half height resolution
16037 (left eye above, right eye below)
16040 above-below with half height resolution
16041 (right eye above, left eye below)
16044 alternating frames (left eye first, right eye second)
16047 alternating frames (right eye first, left eye second)
16050 interleaved rows (left eye has top row, right eye starts on next row)
16053 interleaved rows (right eye has top row, left eye starts on next row)
16056 anaglyph red/blue gray
16057 (red filter on left eye, blue filter on right eye)
16060 anaglyph red/green gray
16061 (red filter on left eye, green filter on right eye)
16064 anaglyph red/cyan gray
16065 (red filter on left eye, cyan filter on right eye)
16068 anaglyph red/cyan half colored
16069 (red filter on left eye, cyan filter on right eye)
16072 anaglyph red/cyan color
16073 (red filter on left eye, cyan filter on right eye)
16076 anaglyph red/cyan color optimized with the least squares projection of dubois
16077 (red filter on left eye, cyan filter on right eye)
16080 anaglyph green/magenta gray
16081 (green filter on left eye, magenta filter on right eye)
16084 anaglyph green/magenta half colored
16085 (green filter on left eye, magenta filter on right eye)
16088 anaglyph green/magenta colored
16089 (green filter on left eye, magenta filter on right eye)
16092 anaglyph green/magenta color optimized with the least squares projection of dubois
16093 (green filter on left eye, magenta filter on right eye)
16096 anaglyph yellow/blue gray
16097 (yellow filter on left eye, blue filter on right eye)
16100 anaglyph yellow/blue half colored
16101 (yellow filter on left eye, blue filter on right eye)
16104 anaglyph yellow/blue colored
16105 (yellow filter on left eye, blue filter on right eye)
16108 anaglyph yellow/blue color optimized with the least squares projection of dubois
16109 (yellow filter on left eye, blue filter on right eye)
16112 mono output (left eye only)
16115 mono output (right eye only)
16118 checkerboard, left eye first
16121 checkerboard, right eye first
16124 interleaved columns, left eye first
16127 interleaved columns, right eye first
16133 Default value is @samp{arcd}.
16136 @subsection Examples
16140 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
16146 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
16152 @section streamselect, astreamselect
16153 Select video or audio streams.
16155 The filter accepts the following options:
16159 Set number of inputs. Default is 2.
16162 Set input indexes to remap to outputs.
16165 @subsection Commands
16167 The @code{streamselect} and @code{astreamselect} filter supports the following
16172 Set input indexes to remap to outputs.
16175 @subsection Examples
16179 Select first 5 seconds 1st stream and rest of time 2nd stream:
16181 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
16185 Same as above, but for audio:
16187 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
16192 Apply sobel operator to input video stream.
16194 The filter accepts the following option:
16198 Set which planes will be processed, unprocessed planes will be copied.
16199 By default value 0xf, all planes will be processed.
16202 Set value which will be multiplied with filtered result.
16205 Set value which will be added to filtered result.
16211 Apply a simple postprocessing filter that compresses and decompresses the image
16212 at several (or - in the case of @option{quality} level @code{6} - all) shifts
16213 and average the results.
16215 The filter accepts the following options:
16219 Set quality. This option defines the number of levels for averaging. It accepts
16220 an integer in the range 0-6. If set to @code{0}, the filter will have no
16221 effect. A value of @code{6} means the higher quality. For each increment of
16222 that value the speed drops by a factor of approximately 2. Default value is
16226 Force a constant quantization parameter. If not set, the filter will use the QP
16227 from the video stream (if available).
16230 Set thresholding mode. Available modes are:
16234 Set hard thresholding (default).
16236 Set soft thresholding (better de-ringing effect, but likely blurrier).
16239 @item use_bframe_qp
16240 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
16241 option may cause flicker since the B-Frames have often larger QP. Default is
16242 @code{0} (not enabled).
16247 Scale the input by applying one of the super-resolution methods based on
16248 convolutional neural networks. Supported models:
16252 Super-Resolution Convolutional Neural Network model (SRCNN).
16253 See @url{https://arxiv.org/abs/1501.00092}.
16256 Efficient Sub-Pixel Convolutional Neural Network model (ESPCN).
16257 See @url{https://arxiv.org/abs/1609.05158}.
16260 Training scripts as well as scripts for model generation are provided in
16261 the repository at @url{https://github.com/HighVoltageRocknRoll/sr.git}.
16263 The filter accepts the following options:
16267 Specify which DNN backend to use for model loading and execution. This option accepts
16268 the following values:
16272 Native implementation of DNN loading and execution.
16275 TensorFlow backend. To enable this backend you
16276 need to install the TensorFlow for C library (see
16277 @url{https://www.tensorflow.org/install/install_c}) and configure FFmpeg with
16278 @code{--enable-libtensorflow}
16281 Default value is @samp{native}.
16284 Set path to model file specifying network architecture and its parameters.
16285 Note that different backends use different file formats. TensorFlow backend
16286 can load files for both formats, while native backend can load files for only
16290 Set scale factor for SRCNN model. Allowed values are @code{2}, @code{3} and @code{4}.
16291 Default value is @code{2}. Scale factor is necessary for SRCNN model, because it accepts
16292 input upscaled using bicubic upscaling with proper scale factor.
16298 Draw subtitles on top of input video using the libass library.
16300 To enable compilation of this filter you need to configure FFmpeg with
16301 @code{--enable-libass}. This filter also requires a build with libavcodec and
16302 libavformat to convert the passed subtitles file to ASS (Advanced Substation
16303 Alpha) subtitles format.
16305 The filter accepts the following options:
16309 Set the filename of the subtitle file to read. It must be specified.
16311 @item original_size
16312 Specify the size of the original video, the video for which the ASS file
16313 was composed. For the syntax of this option, check the
16314 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16315 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
16316 correctly scale the fonts if the aspect ratio has been changed.
16319 Set a directory path containing fonts that can be used by the filter.
16320 These fonts will be used in addition to whatever the font provider uses.
16323 Process alpha channel, by default alpha channel is untouched.
16326 Set subtitles input character encoding. @code{subtitles} filter only. Only
16327 useful if not UTF-8.
16329 @item stream_index, si
16330 Set subtitles stream index. @code{subtitles} filter only.
16333 Override default style or script info parameters of the subtitles. It accepts a
16334 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
16337 If the first key is not specified, it is assumed that the first value
16338 specifies the @option{filename}.
16340 For example, to render the file @file{sub.srt} on top of the input
16341 video, use the command:
16346 which is equivalent to:
16348 subtitles=filename=sub.srt
16351 To render the default subtitles stream from file @file{video.mkv}, use:
16353 subtitles=video.mkv
16356 To render the second subtitles stream from that file, use:
16358 subtitles=video.mkv:si=1
16361 To make the subtitles stream from @file{sub.srt} appear in 80% transparent blue
16362 @code{DejaVu Serif}, use:
16364 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HCCFF0000'
16367 @section super2xsai
16369 Scale the input by 2x and smooth using the Super2xSaI (Scale and
16370 Interpolate) pixel art scaling algorithm.
16372 Useful for enlarging pixel art images without reducing sharpness.
16376 Swap two rectangular objects in video.
16378 This filter accepts the following options:
16388 Set 1st rect x coordinate.
16391 Set 1st rect y coordinate.
16394 Set 2nd rect x coordinate.
16397 Set 2nd rect y coordinate.
16399 All expressions are evaluated once for each frame.
16402 The all options are expressions containing the following constants:
16407 The input width and height.
16410 same as @var{w} / @var{h}
16413 input sample aspect ratio
16416 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
16419 The number of the input frame, starting from 0.
16422 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
16425 the position in the file of the input frame, NAN if unknown
16433 Apply telecine process to the video.
16435 This filter accepts the following options:
16444 The default value is @code{top}.
16448 A string of numbers representing the pulldown pattern you wish to apply.
16449 The default value is @code{23}.
16453 Some typical patterns:
16458 24p: 2332 (preferred)
16465 24p: 222222222223 ("Euro pulldown")
16472 Apply threshold effect to video stream.
16474 This filter needs four video streams to perform thresholding.
16475 First stream is stream we are filtering.
16476 Second stream is holding threshold values, third stream is holding min values,
16477 and last, fourth stream is holding max values.
16479 The filter accepts the following option:
16483 Set which planes will be processed, unprocessed planes will be copied.
16484 By default value 0xf, all planes will be processed.
16487 For example if first stream pixel's component value is less then threshold value
16488 of pixel component from 2nd threshold stream, third stream value will picked,
16489 otherwise fourth stream pixel component value will be picked.
16491 Using color source filter one can perform various types of thresholding:
16493 @subsection Examples
16497 Binary threshold, using gray color as threshold:
16499 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
16503 Inverted binary threshold, using gray color as threshold:
16505 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
16509 Truncate binary threshold, using gray color as threshold:
16511 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
16515 Threshold to zero, using gray color as threshold:
16517 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
16521 Inverted threshold to zero, using gray color as threshold:
16523 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
16528 Select the most representative frame in a given sequence of consecutive frames.
16530 The filter accepts the following options:
16534 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
16535 will pick one of them, and then handle the next batch of @var{n} frames until
16536 the end. Default is @code{100}.
16539 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
16540 value will result in a higher memory usage, so a high value is not recommended.
16542 @subsection Examples
16546 Extract one picture each 50 frames:
16552 Complete example of a thumbnail creation with @command{ffmpeg}:
16554 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
16560 Tile several successive frames together.
16562 The filter accepts the following options:
16567 Set the grid size (i.e. the number of lines and columns). For the syntax of
16568 this option, check the
16569 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16572 Set the maximum number of frames to render in the given area. It must be less
16573 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
16574 the area will be used.
16577 Set the outer border margin in pixels.
16580 Set the inner border thickness (i.e. the number of pixels between frames). For
16581 more advanced padding options (such as having different values for the edges),
16582 refer to the pad video filter.
16585 Specify the color of the unused area. For the syntax of this option, check the
16586 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
16587 The default value of @var{color} is "black".
16590 Set the number of frames to overlap when tiling several successive frames together.
16591 The value must be between @code{0} and @var{nb_frames - 1}.
16594 Set the number of frames to initially be empty before displaying first output frame.
16595 This controls how soon will one get first output frame.
16596 The value must be between @code{0} and @var{nb_frames - 1}.
16599 @subsection Examples
16603 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
16605 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
16607 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
16608 duplicating each output frame to accommodate the originally detected frame
16612 Display @code{5} pictures in an area of @code{3x2} frames,
16613 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
16614 mixed flat and named options:
16616 tile=3x2:nb_frames=5:padding=7:margin=2
16620 @section tinterlace
16622 Perform various types of temporal field interlacing.
16624 Frames are counted starting from 1, so the first input frame is
16627 The filter accepts the following options:
16632 Specify the mode of the interlacing. This option can also be specified
16633 as a value alone. See below for a list of values for this option.
16635 Available values are:
16639 Move odd frames into the upper field, even into the lower field,
16640 generating a double height frame at half frame rate.
16644 Frame 1 Frame 2 Frame 3 Frame 4
16646 11111 22222 33333 44444
16647 11111 22222 33333 44444
16648 11111 22222 33333 44444
16649 11111 22222 33333 44444
16663 Only output odd frames, even frames are dropped, generating a frame with
16664 unchanged height at half frame rate.
16669 Frame 1 Frame 2 Frame 3 Frame 4
16671 11111 22222 33333 44444
16672 11111 22222 33333 44444
16673 11111 22222 33333 44444
16674 11111 22222 33333 44444
16684 Only output even frames, odd frames are dropped, generating a frame with
16685 unchanged height at half frame rate.
16690 Frame 1 Frame 2 Frame 3 Frame 4
16692 11111 22222 33333 44444
16693 11111 22222 33333 44444
16694 11111 22222 33333 44444
16695 11111 22222 33333 44444
16705 Expand each frame to full height, but pad alternate lines with black,
16706 generating a frame with double height at the same input frame rate.
16711 Frame 1 Frame 2 Frame 3 Frame 4
16713 11111 22222 33333 44444
16714 11111 22222 33333 44444
16715 11111 22222 33333 44444
16716 11111 22222 33333 44444
16719 11111 ..... 33333 .....
16720 ..... 22222 ..... 44444
16721 11111 ..... 33333 .....
16722 ..... 22222 ..... 44444
16723 11111 ..... 33333 .....
16724 ..... 22222 ..... 44444
16725 11111 ..... 33333 .....
16726 ..... 22222 ..... 44444
16730 @item interleave_top, 4
16731 Interleave the upper field from odd frames with the lower field from
16732 even frames, generating a frame with unchanged height at half frame rate.
16737 Frame 1 Frame 2 Frame 3 Frame 4
16739 11111<- 22222 33333<- 44444
16740 11111 22222<- 33333 44444<-
16741 11111<- 22222 33333<- 44444
16742 11111 22222<- 33333 44444<-
16752 @item interleave_bottom, 5
16753 Interleave the lower field from odd frames with the upper field from
16754 even frames, generating a frame with unchanged height at half frame rate.
16759 Frame 1 Frame 2 Frame 3 Frame 4
16761 11111 22222<- 33333 44444<-
16762 11111<- 22222 33333<- 44444
16763 11111 22222<- 33333 44444<-
16764 11111<- 22222 33333<- 44444
16774 @item interlacex2, 6
16775 Double frame rate with unchanged height. Frames are inserted each
16776 containing the second temporal field from the previous input frame and
16777 the first temporal field from the next input frame. This mode relies on
16778 the top_field_first flag. Useful for interlaced video displays with no
16779 field synchronisation.
16784 Frame 1 Frame 2 Frame 3 Frame 4
16786 11111 22222 33333 44444
16787 11111 22222 33333 44444
16788 11111 22222 33333 44444
16789 11111 22222 33333 44444
16792 11111 22222 22222 33333 33333 44444 44444
16793 11111 11111 22222 22222 33333 33333 44444
16794 11111 22222 22222 33333 33333 44444 44444
16795 11111 11111 22222 22222 33333 33333 44444
16800 Move odd frames into the upper field, even into the lower field,
16801 generating a double height frame at same frame rate.
16806 Frame 1 Frame 2 Frame 3 Frame 4
16808 11111 22222 33333 44444
16809 11111 22222 33333 44444
16810 11111 22222 33333 44444
16811 11111 22222 33333 44444
16814 11111 33333 33333 55555
16815 22222 22222 44444 44444
16816 11111 33333 33333 55555
16817 22222 22222 44444 44444
16818 11111 33333 33333 55555
16819 22222 22222 44444 44444
16820 11111 33333 33333 55555
16821 22222 22222 44444 44444
16826 Numeric values are deprecated but are accepted for backward
16827 compatibility reasons.
16829 Default mode is @code{merge}.
16832 Specify flags influencing the filter process.
16834 Available value for @var{flags} is:
16837 @item low_pass_filter, vlfp
16838 Enable linear vertical low-pass filtering in the filter.
16839 Vertical low-pass filtering is required when creating an interlaced
16840 destination from a progressive source which contains high-frequency
16841 vertical detail. Filtering will reduce interlace 'twitter' and Moire
16844 @item complex_filter, cvlfp
16845 Enable complex vertical low-pass filtering.
16846 This will slightly less reduce interlace 'twitter' and Moire
16847 patterning but better retain detail and subjective sharpness impression.
16851 Vertical low-pass filtering can only be enabled for @option{mode}
16852 @var{interleave_top} and @var{interleave_bottom}.
16858 Mix successive video frames.
16860 A description of the accepted options follows.
16864 The number of successive frames to mix. If unspecified, it defaults to 3.
16867 Specify weight of each input video frame.
16868 Each weight is separated by space. If number of weights is smaller than
16869 number of @var{frames} last specified weight will be used for all remaining
16873 Specify scale, if it is set it will be multiplied with sum
16874 of each weight multiplied with pixel values to give final destination
16875 pixel value. By default @var{scale} is auto scaled to sum of weights.
16878 @subsection Examples
16882 Average 7 successive frames:
16884 tmix=frames=7:weights="1 1 1 1 1 1 1"
16888 Apply simple temporal convolution:
16890 tmix=frames=3:weights="-1 3 -1"
16894 Similar as above but only showing temporal differences:
16896 tmix=frames=3:weights="-1 2 -1":scale=1
16902 Tone map colors from different dynamic ranges.
16904 This filter expects data in single precision floating point, as it needs to
16905 operate on (and can output) out-of-range values. Another filter, such as
16906 @ref{zscale}, is needed to convert the resulting frame to a usable format.
16908 The tonemapping algorithms implemented only work on linear light, so input
16909 data should be linearized beforehand (and possibly correctly tagged).
16912 ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
16915 @subsection Options
16916 The filter accepts the following options.
16920 Set the tone map algorithm to use.
16922 Possible values are:
16925 Do not apply any tone map, only desaturate overbright pixels.
16928 Hard-clip any out-of-range values. Use it for perfect color accuracy for
16929 in-range values, while distorting out-of-range values.
16932 Stretch the entire reference gamut to a linear multiple of the display.
16935 Fit a logarithmic transfer between the tone curves.
16938 Preserve overall image brightness with a simple curve, using nonlinear
16939 contrast, which results in flattening details and degrading color accuracy.
16942 Preserve both dark and bright details better than @var{reinhard}, at the cost
16943 of slightly darkening everything. Use it when detail preservation is more
16944 important than color and brightness accuracy.
16947 Smoothly map out-of-range values, while retaining contrast and colors for
16948 in-range material as much as possible. Use it when color accuracy is more
16949 important than detail preservation.
16955 Tune the tone mapping algorithm.
16957 This affects the following algorithms:
16963 Specifies the scale factor to use while stretching.
16967 Specifies the exponent of the function.
16971 Specify an extra linear coefficient to multiply into the signal before clipping.
16975 Specify the local contrast coefficient at the display peak.
16976 Default to 0.5, which means that in-gamut values will be about half as bright
16983 Specify the transition point from linear to mobius transform. Every value
16984 below this point is guaranteed to be mapped 1:1. The higher the value, the
16985 more accurate the result will be, at the cost of losing bright details.
16986 Default to 0.3, which due to the steep initial slope still preserves in-range
16987 colors fairly accurately.
16991 Apply desaturation for highlights that exceed this level of brightness. The
16992 higher the parameter, the more color information will be preserved. This
16993 setting helps prevent unnaturally blown-out colors for super-highlights, by
16994 (smoothly) turning into white instead. This makes images feel more natural,
16995 at the cost of reducing information about out-of-range colors.
16997 The default of 2.0 is somewhat conservative and will mostly just apply to
16998 skies or directly sunlit surfaces. A setting of 0.0 disables this option.
17000 This option works only if the input frame has a supported color tag.
17003 Override signal/nominal/reference peak with this value. Useful when the
17004 embedded peak information in display metadata is not reliable or when tone
17005 mapping from a lower range to a higher range.
17010 Temporarily pad video frames.
17012 The filter accepts the following options:
17016 Specify number of delay frames before input video stream.
17019 Specify number of padding frames after input video stream.
17020 Set to -1 to pad indefinitely.
17023 Set kind of frames added to beginning of stream.
17024 Can be either @var{add} or @var{clone}.
17025 With @var{add} frames of solid-color are added.
17026 With @var{clone} frames are clones of first frame.
17029 Set kind of frames added to end of stream.
17030 Can be either @var{add} or @var{clone}.
17031 With @var{add} frames of solid-color are added.
17032 With @var{clone} frames are clones of last frame.
17034 @item start_duration, stop_duration
17035 Specify the duration of the start/stop delay. See
17036 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17037 for the accepted syntax.
17038 These options override @var{start} and @var{stop}.
17041 Specify the color of the padded area. For the syntax of this option,
17042 check the @ref{color syntax,,"Color" section in the ffmpeg-utils
17043 manual,ffmpeg-utils}.
17045 The default value of @var{color} is "black".
17051 Transpose rows with columns in the input video and optionally flip it.
17053 It accepts the following parameters:
17058 Specify the transposition direction.
17060 Can assume the following values:
17062 @item 0, 4, cclock_flip
17063 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
17071 Rotate by 90 degrees clockwise, that is:
17079 Rotate by 90 degrees counterclockwise, that is:
17086 @item 3, 7, clock_flip
17087 Rotate by 90 degrees clockwise and vertically flip, that is:
17095 For values between 4-7, the transposition is only done if the input
17096 video geometry is portrait and not landscape. These values are
17097 deprecated, the @code{passthrough} option should be used instead.
17099 Numerical values are deprecated, and should be dropped in favor of
17100 symbolic constants.
17103 Do not apply the transposition if the input geometry matches the one
17104 specified by the specified value. It accepts the following values:
17107 Always apply transposition.
17109 Preserve portrait geometry (when @var{height} >= @var{width}).
17111 Preserve landscape geometry (when @var{width} >= @var{height}).
17114 Default value is @code{none}.
17117 For example to rotate by 90 degrees clockwise and preserve portrait
17120 transpose=dir=1:passthrough=portrait
17123 The command above can also be specified as:
17125 transpose=1:portrait
17128 @section transpose_npp
17130 Transpose rows with columns in the input video and optionally flip it.
17131 For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
17133 It accepts the following parameters:
17138 Specify the transposition direction.
17140 Can assume the following values:
17143 Rotate by 90 degrees counterclockwise and vertically flip. (default)
17146 Rotate by 90 degrees clockwise.
17149 Rotate by 90 degrees counterclockwise.
17152 Rotate by 90 degrees clockwise and vertically flip.
17156 Do not apply the transposition if the input geometry matches the one
17157 specified by the specified value. It accepts the following values:
17160 Always apply transposition. (default)
17162 Preserve portrait geometry (when @var{height} >= @var{width}).
17164 Preserve landscape geometry (when @var{width} >= @var{height}).
17170 Trim the input so that the output contains one continuous subpart of the input.
17172 It accepts the following parameters:
17175 Specify the time of the start of the kept section, i.e. the frame with the
17176 timestamp @var{start} will be the first frame in the output.
17179 Specify the time of the first frame that will be dropped, i.e. the frame
17180 immediately preceding the one with the timestamp @var{end} will be the last
17181 frame in the output.
17184 This is the same as @var{start}, except this option sets the start timestamp
17185 in timebase units instead of seconds.
17188 This is the same as @var{end}, except this option sets the end timestamp
17189 in timebase units instead of seconds.
17192 The maximum duration of the output in seconds.
17195 The number of the first frame that should be passed to the output.
17198 The number of the first frame that should be dropped.
17201 @option{start}, @option{end}, and @option{duration} are expressed as time
17202 duration specifications; see
17203 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
17204 for the accepted syntax.
17206 Note that the first two sets of the start/end options and the @option{duration}
17207 option look at the frame timestamp, while the _frame variants simply count the
17208 frames that pass through the filter. Also note that this filter does not modify
17209 the timestamps. If you wish for the output timestamps to start at zero, insert a
17210 setpts filter after the trim filter.
17212 If multiple start or end options are set, this filter tries to be greedy and
17213 keep all the frames that match at least one of the specified constraints. To keep
17214 only the part that matches all the constraints at once, chain multiple trim
17217 The defaults are such that all the input is kept. So it is possible to set e.g.
17218 just the end values to keep everything before the specified time.
17223 Drop everything except the second minute of input:
17225 ffmpeg -i INPUT -vf trim=60:120
17229 Keep only the first second:
17231 ffmpeg -i INPUT -vf trim=duration=1
17236 @section unpremultiply
17237 Apply alpha unpremultiply effect to input video stream using first plane
17238 of second stream as alpha.
17240 Both streams must have same dimensions and same pixel format.
17242 The filter accepts the following option:
17246 Set which planes will be processed, unprocessed planes will be copied.
17247 By default value 0xf, all planes will be processed.
17249 If the format has 1 or 2 components, then luma is bit 0.
17250 If the format has 3 or 4 components:
17251 for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red;
17252 for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V.
17253 If present, the alpha channel is always the last bit.
17256 Do not require 2nd input for processing, instead use alpha plane from input stream.
17262 Sharpen or blur the input video.
17264 It accepts the following parameters:
17267 @item luma_msize_x, lx
17268 Set the luma matrix horizontal size. It must be an odd integer between
17269 3 and 23. The default value is 5.
17271 @item luma_msize_y, ly
17272 Set the luma matrix vertical size. It must be an odd integer between 3
17273 and 23. The default value is 5.
17275 @item luma_amount, la
17276 Set the luma effect strength. It must be a floating point number, reasonable
17277 values lay between -1.5 and 1.5.
17279 Negative values will blur the input video, while positive values will
17280 sharpen it, a value of zero will disable the effect.
17282 Default value is 1.0.
17284 @item chroma_msize_x, cx
17285 Set the chroma matrix horizontal size. It must be an odd integer
17286 between 3 and 23. The default value is 5.
17288 @item chroma_msize_y, cy
17289 Set the chroma matrix vertical size. It must be an odd integer
17290 between 3 and 23. The default value is 5.
17292 @item chroma_amount, ca
17293 Set the chroma effect strength. It must be a floating point number, reasonable
17294 values lay between -1.5 and 1.5.
17296 Negative values will blur the input video, while positive values will
17297 sharpen it, a value of zero will disable the effect.
17299 Default value is 0.0.
17303 All parameters are optional and default to the equivalent of the
17304 string '5:5:1.0:5:5:0.0'.
17306 @subsection Examples
17310 Apply strong luma sharpen effect:
17312 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
17316 Apply a strong blur of both luma and chroma parameters:
17318 unsharp=7:7:-2:7:7:-2
17324 Apply ultra slow/simple postprocessing filter that compresses and decompresses
17325 the image at several (or - in the case of @option{quality} level @code{8} - all)
17326 shifts and average the results.
17328 The way this differs from the behavior of spp is that uspp actually encodes &
17329 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
17330 DCT similar to MJPEG.
17332 The filter accepts the following options:
17336 Set quality. This option defines the number of levels for averaging. It accepts
17337 an integer in the range 0-8. If set to @code{0}, the filter will have no
17338 effect. A value of @code{8} means the higher quality. For each increment of
17339 that value the speed drops by a factor of approximately 2. Default value is
17343 Force a constant quantization parameter. If not set, the filter will use the QP
17344 from the video stream (if available).
17347 @section vaguedenoiser
17349 Apply a wavelet based denoiser.
17351 It transforms each frame from the video input into the wavelet domain,
17352 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
17353 the obtained coefficients. It does an inverse wavelet transform after.
17354 Due to wavelet properties, it should give a nice smoothed result, and
17355 reduced noise, without blurring picture features.
17357 This filter accepts the following options:
17361 The filtering strength. The higher, the more filtered the video will be.
17362 Hard thresholding can use a higher threshold than soft thresholding
17363 before the video looks overfiltered. Default value is 2.
17366 The filtering method the filter will use.
17368 It accepts the following values:
17371 All values under the threshold will be zeroed.
17374 All values under the threshold will be zeroed. All values above will be
17375 reduced by the threshold.
17378 Scales or nullifies coefficients - intermediary between (more) soft and
17379 (less) hard thresholding.
17382 Default is garrote.
17385 Number of times, the wavelet will decompose the picture. Picture can't
17386 be decomposed beyond a particular point (typically, 8 for a 640x480
17387 frame - as 2^9 = 512 > 480). Valid values are integers between 1 and 32. Default value is 6.
17390 Partial of full denoising (limited coefficients shrinking), from 0 to 100. Default value is 85.
17393 A list of the planes to process. By default all planes are processed.
17396 @section vectorscope
17398 Display 2 color component values in the two dimensional graph (which is called
17401 This filter accepts the following options:
17405 Set vectorscope mode.
17407 It accepts the following values:
17410 Gray values are displayed on graph, higher brightness means more pixels have
17411 same component color value on location in graph. This is the default mode.
17414 Gray values are displayed on graph. Surrounding pixels values which are not
17415 present in video frame are drawn in gradient of 2 color components which are
17416 set by option @code{x} and @code{y}. The 3rd color component is static.
17419 Actual color components values present in video frame are displayed on graph.
17422 Similar as color2 but higher frequency of same values @code{x} and @code{y}
17423 on graph increases value of another color component, which is luminance by
17424 default values of @code{x} and @code{y}.
17427 Actual colors present in video frame are displayed on graph. If two different
17428 colors map to same position on graph then color with higher value of component
17429 not present in graph is picked.
17432 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
17433 component picked from radial gradient.
17437 Set which color component will be represented on X-axis. Default is @code{1}.
17440 Set which color component will be represented on Y-axis. Default is @code{2}.
17443 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
17444 of color component which represents frequency of (X, Y) location in graph.
17449 No envelope, this is default.
17452 Instant envelope, even darkest single pixel will be clearly highlighted.
17455 Hold maximum and minimum values presented in graph over time. This way you
17456 can still spot out of range values without constantly looking at vectorscope.
17459 Peak and instant envelope combined together.
17463 Set what kind of graticule to draw.
17471 Set graticule opacity.
17474 Set graticule flags.
17478 Draw graticule for white point.
17481 Draw graticule for black point.
17484 Draw color points short names.
17488 Set background opacity.
17490 @item lthreshold, l
17491 Set low threshold for color component not represented on X or Y axis.
17492 Values lower than this value will be ignored. Default is 0.
17493 Note this value is multiplied with actual max possible value one pixel component
17494 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
17497 @item hthreshold, h
17498 Set high threshold for color component not represented on X or Y axis.
17499 Values higher than this value will be ignored. Default is 1.
17500 Note this value is multiplied with actual max possible value one pixel component
17501 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
17502 is 0.9 * 255 = 230.
17504 @item colorspace, c
17505 Set what kind of colorspace to use when drawing graticule.
17514 @anchor{vidstabdetect}
17515 @section vidstabdetect
17517 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
17518 @ref{vidstabtransform} for pass 2.
17520 This filter generates a file with relative translation and rotation
17521 transform information about subsequent frames, which is then used by
17522 the @ref{vidstabtransform} filter.
17524 To enable compilation of this filter you need to configure FFmpeg with
17525 @code{--enable-libvidstab}.
17527 This filter accepts the following options:
17531 Set the path to the file used to write the transforms information.
17532 Default value is @file{transforms.trf}.
17535 Set how shaky the video is and how quick the camera is. It accepts an
17536 integer in the range 1-10, a value of 1 means little shakiness, a
17537 value of 10 means strong shakiness. Default value is 5.
17540 Set the accuracy of the detection process. It must be a value in the
17541 range 1-15. A value of 1 means low accuracy, a value of 15 means high
17542 accuracy. Default value is 15.
17545 Set stepsize of the search process. The region around minimum is
17546 scanned with 1 pixel resolution. Default value is 6.
17549 Set minimum contrast. Below this value a local measurement field is
17550 discarded. Must be a floating point value in the range 0-1. Default
17554 Set reference frame number for tripod mode.
17556 If enabled, the motion of the frames is compared to a reference frame
17557 in the filtered stream, identified by the specified number. The idea
17558 is to compensate all movements in a more-or-less static scene and keep
17559 the camera view absolutely still.
17561 If set to 0, it is disabled. The frames are counted starting from 1.
17564 Show fields and transforms in the resulting frames. It accepts an
17565 integer in the range 0-2. Default value is 0, which disables any
17569 @subsection Examples
17573 Use default values:
17579 Analyze strongly shaky movie and put the results in file
17580 @file{mytransforms.trf}:
17582 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
17586 Visualize the result of internal transformations in the resulting
17589 vidstabdetect=show=1
17593 Analyze a video with medium shakiness using @command{ffmpeg}:
17595 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
17599 @anchor{vidstabtransform}
17600 @section vidstabtransform
17602 Video stabilization/deshaking: pass 2 of 2,
17603 see @ref{vidstabdetect} for pass 1.
17605 Read a file with transform information for each frame and
17606 apply/compensate them. Together with the @ref{vidstabdetect}
17607 filter this can be used to deshake videos. See also
17608 @url{http://public.hronopik.de/vid.stab}. It is important to also use
17609 the @ref{unsharp} filter, see below.
17611 To enable compilation of this filter you need to configure FFmpeg with
17612 @code{--enable-libvidstab}.
17614 @subsection Options
17618 Set path to the file used to read the transforms. Default value is
17619 @file{transforms.trf}.
17622 Set the number of frames (value*2 + 1) used for lowpass filtering the
17623 camera movements. Default value is 10.
17625 For example a number of 10 means that 21 frames are used (10 in the
17626 past and 10 in the future) to smoothen the motion in the video. A
17627 larger value leads to a smoother video, but limits the acceleration of
17628 the camera (pan/tilt movements). 0 is a special case where a static
17629 camera is simulated.
17632 Set the camera path optimization algorithm.
17634 Accepted values are:
17637 gaussian kernel low-pass filter on camera motion (default)
17639 averaging on transformations
17643 Set maximal number of pixels to translate frames. Default value is -1,
17647 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
17648 value is -1, meaning no limit.
17651 Specify how to deal with borders that may be visible due to movement
17654 Available values are:
17657 keep image information from previous frame (default)
17659 fill the border black
17663 Invert transforms if set to 1. Default value is 0.
17666 Consider transforms as relative to previous frame if set to 1,
17667 absolute if set to 0. Default value is 0.
17670 Set percentage to zoom. A positive value will result in a zoom-in
17671 effect, a negative value in a zoom-out effect. Default value is 0 (no
17675 Set optimal zooming to avoid borders.
17677 Accepted values are:
17682 optimal static zoom value is determined (only very strong movements
17683 will lead to visible borders) (default)
17685 optimal adaptive zoom value is determined (no borders will be
17686 visible), see @option{zoomspeed}
17689 Note that the value given at zoom is added to the one calculated here.
17692 Set percent to zoom maximally each frame (enabled when
17693 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
17697 Specify type of interpolation.
17699 Available values are:
17704 linear only horizontal
17706 linear in both directions (default)
17708 cubic in both directions (slow)
17712 Enable virtual tripod mode if set to 1, which is equivalent to
17713 @code{relative=0:smoothing=0}. Default value is 0.
17715 Use also @code{tripod} option of @ref{vidstabdetect}.
17718 Increase log verbosity if set to 1. Also the detected global motions
17719 are written to the temporary file @file{global_motions.trf}. Default
17723 @subsection Examples
17727 Use @command{ffmpeg} for a typical stabilization with default values:
17729 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
17732 Note the use of the @ref{unsharp} filter which is always recommended.
17735 Zoom in a bit more and load transform data from a given file:
17737 vidstabtransform=zoom=5:input="mytransforms.trf"
17741 Smoothen the video even more:
17743 vidstabtransform=smoothing=30
17749 Flip the input video vertically.
17751 For example, to vertically flip a video with @command{ffmpeg}:
17753 ffmpeg -i in.avi -vf "vflip" out.avi
17758 Detect variable frame rate video.
17760 This filter tries to detect if the input is variable or constant frame rate.
17762 At end it will output number of frames detected as having variable delta pts,
17763 and ones with constant delta pts.
17764 If there was frames with variable delta, than it will also show min and max delta
17769 Boost or alter saturation.
17771 The filter accepts the following options:
17774 Set strength of boost if positive value or strength of alter if negative value.
17775 Default is 0. Allowed range is from -2 to 2.
17778 Set the red balance. Default is 1. Allowed range is from -10 to 10.
17781 Set the green balance. Default is 1. Allowed range is from -10 to 10.
17784 Set the blue balance. Default is 1. Allowed range is from -10 to 10.
17787 Set the red luma coefficient.
17790 Set the green luma coefficient.
17793 Set the blue luma coefficient.
17799 Make or reverse a natural vignetting effect.
17801 The filter accepts the following options:
17805 Set lens angle expression as a number of radians.
17807 The value is clipped in the @code{[0,PI/2]} range.
17809 Default value: @code{"PI/5"}
17813 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
17817 Set forward/backward mode.
17819 Available modes are:
17822 The larger the distance from the central point, the darker the image becomes.
17825 The larger the distance from the central point, the brighter the image becomes.
17826 This can be used to reverse a vignette effect, though there is no automatic
17827 detection to extract the lens @option{angle} and other settings (yet). It can
17828 also be used to create a burning effect.
17831 Default value is @samp{forward}.
17834 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
17836 It accepts the following values:
17839 Evaluate expressions only once during the filter initialization.
17842 Evaluate expressions for each incoming frame. This is way slower than the
17843 @samp{init} mode since it requires all the scalers to be re-computed, but it
17844 allows advanced dynamic expressions.
17847 Default value is @samp{init}.
17850 Set dithering to reduce the circular banding effects. Default is @code{1}
17854 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
17855 Setting this value to the SAR of the input will make a rectangular vignetting
17856 following the dimensions of the video.
17858 Default is @code{1/1}.
17861 @subsection Expressions
17863 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
17864 following parameters.
17869 input width and height
17872 the number of input frame, starting from 0
17875 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
17876 @var{TB} units, NAN if undefined
17879 frame rate of the input video, NAN if the input frame rate is unknown
17882 the PTS (Presentation TimeStamp) of the filtered video frame,
17883 expressed in seconds, NAN if undefined
17886 time base of the input video
17890 @subsection Examples
17894 Apply simple strong vignetting effect:
17900 Make a flickering vignetting:
17902 vignette='PI/4+random(1)*PI/50':eval=frame
17907 @section vmafmotion
17909 Obtain the average vmaf motion score of a video.
17910 It is one of the component filters of VMAF.
17912 The obtained average motion score is printed through the logging system.
17914 In the below example the input file @file{ref.mpg} is being processed and score
17918 ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
17922 Stack input videos vertically.
17924 All streams must be of same pixel format and of same width.
17926 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
17927 to create same output.
17929 The filter accept the following option:
17933 Set number of input streams. Default is 2.
17936 If set to 1, force the output to terminate when the shortest input
17937 terminates. Default value is 0.
17942 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
17943 Deinterlacing Filter").
17945 Based on the process described by Martin Weston for BBC R&D, and
17946 implemented based on the de-interlace algorithm written by Jim
17947 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
17948 uses filter coefficients calculated by BBC R&D.
17950 There are two sets of filter coefficients, so called "simple":
17951 and "complex". Which set of filter coefficients is used can
17952 be set by passing an optional parameter:
17956 Set the interlacing filter coefficients. Accepts one of the following values:
17960 Simple filter coefficient set.
17962 More-complex filter coefficient set.
17964 Default value is @samp{complex}.
17967 Specify which frames to deinterlace. Accept one of the following values:
17971 Deinterlace all frames,
17973 Only deinterlace frames marked as interlaced.
17976 Default value is @samp{all}.
17980 Video waveform monitor.
17982 The waveform monitor plots color component intensity. By default luminance
17983 only. Each column of the waveform corresponds to a column of pixels in the
17986 It accepts the following options:
17990 Can be either @code{row}, or @code{column}. Default is @code{column}.
17991 In row mode, the graph on the left side represents color component value 0 and
17992 the right side represents value = 255. In column mode, the top side represents
17993 color component value = 0 and bottom side represents value = 255.
17996 Set intensity. Smaller values are useful to find out how many values of the same
17997 luminance are distributed across input rows/columns.
17998 Default value is @code{0.04}. Allowed range is [0, 1].
18001 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
18002 In mirrored mode, higher values will be represented on the left
18003 side for @code{row} mode and at the top for @code{column} mode. Default is
18004 @code{1} (mirrored).
18008 It accepts the following values:
18011 Presents information identical to that in the @code{parade}, except
18012 that the graphs representing color components are superimposed directly
18015 This display mode makes it easier to spot relative differences or similarities
18016 in overlapping areas of the color components that are supposed to be identical,
18017 such as neutral whites, grays, or blacks.
18020 Display separate graph for the color components side by side in
18021 @code{row} mode or one below the other in @code{column} mode.
18024 Display separate graph for the color components side by side in
18025 @code{column} mode or one below the other in @code{row} mode.
18027 Using this display mode makes it easy to spot color casts in the highlights
18028 and shadows of an image, by comparing the contours of the top and the bottom
18029 graphs of each waveform. Since whites, grays, and blacks are characterized
18030 by exactly equal amounts of red, green, and blue, neutral areas of the picture
18031 should display three waveforms of roughly equal width/height. If not, the
18032 correction is easy to perform by making level adjustments the three waveforms.
18034 Default is @code{stack}.
18036 @item components, c
18037 Set which color components to display. Default is 1, which means only luminance
18038 or red color component if input is in RGB colorspace. If is set for example to
18039 7 it will display all 3 (if) available color components.
18044 No envelope, this is default.
18047 Instant envelope, minimum and maximum values presented in graph will be easily
18048 visible even with small @code{step} value.
18051 Hold minimum and maximum values presented in graph across time. This way you
18052 can still spot out of range values without constantly looking at waveforms.
18055 Peak and instant envelope combined together.
18061 No filtering, this is default.
18064 Luma and chroma combined together.
18067 Similar as above, but shows difference between blue and red chroma.
18070 Similar as above, but use different colors.
18073 Displays only chroma.
18076 Displays actual color value on waveform.
18079 Similar as above, but with luma showing frequency of chroma values.
18083 Set which graticule to display.
18087 Do not display graticule.
18090 Display green graticule showing legal broadcast ranges.
18093 Display orange graticule showing legal broadcast ranges.
18097 Set graticule opacity.
18100 Set graticule flags.
18104 Draw numbers above lines. By default enabled.
18107 Draw dots instead of lines.
18111 Set scale used for displaying graticule.
18118 Default is digital.
18121 Set background opacity.
18124 @section weave, doubleweave
18126 The @code{weave} takes a field-based video input and join
18127 each two sequential fields into single frame, producing a new double
18128 height clip with half the frame rate and half the frame count.
18130 The @code{doubleweave} works same as @code{weave} but without
18131 halving frame rate and frame count.
18133 It accepts the following option:
18137 Set first field. Available values are:
18141 Set the frame as top-field-first.
18144 Set the frame as bottom-field-first.
18148 @subsection Examples
18152 Interlace video using @ref{select} and @ref{separatefields} filter:
18154 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
18159 Apply the xBR high-quality magnification filter which is designed for pixel
18160 art. It follows a set of edge-detection rules, see
18161 @url{https://forums.libretro.com/t/xbr-algorithm-tutorial/123}.
18163 It accepts the following option:
18167 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
18168 @code{3xBR} and @code{4} for @code{4xBR}.
18169 Default is @code{3}.
18173 Stack video inputs into custom layout.
18175 All streams must be of same pixel format.
18177 The filter accept the following option:
18181 Set number of input streams. Default is 2.
18184 Specify layout of inputs.
18185 This option requires the desired layout configuration to be explicitly set by the user.
18186 This sets position of each video input in output. Each input
18187 is separated by '|'.
18188 The first number represents the column, and the second number represents the row.
18189 Numbers start at 0 and are separated by '_'. Optionally one can use wX and hX,
18190 where X is video input from which to take width or height.
18191 Multiple values can be used when separated by '+'. In such
18192 case values are summed together.
18195 If set to 1, force the output to terminate when the shortest input
18196 terminates. Default value is 0.
18199 @subsection Examples
18203 Display 4 inputs into 2x2 grid,
18204 note that if inputs are of different sizes unused gaps might appear,
18205 as not all of output video is used.
18207 xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
18211 Display 4 inputs into 1x4 grid,
18212 note that if inputs are of different sizes unused gaps might appear,
18213 as not all of output video is used.
18215 xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
18219 Display 9 inputs into 3x3 grid,
18220 note that if inputs are of different sizes unused gaps might appear,
18221 as not all of output video is used.
18223 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
18230 Deinterlace the input video ("yadif" means "yet another deinterlacing
18233 It accepts the following parameters:
18239 The interlacing mode to adopt. It accepts one of the following values:
18242 @item 0, send_frame
18243 Output one frame for each frame.
18244 @item 1, send_field
18245 Output one frame for each field.
18246 @item 2, send_frame_nospatial
18247 Like @code{send_frame}, but it skips the spatial interlacing check.
18248 @item 3, send_field_nospatial
18249 Like @code{send_field}, but it skips the spatial interlacing check.
18252 The default value is @code{send_frame}.
18255 The picture field parity assumed for the input interlaced video. It accepts one
18256 of the following values:
18260 Assume the top field is first.
18262 Assume the bottom field is first.
18264 Enable automatic detection of field parity.
18267 The default value is @code{auto}.
18268 If the interlacing is unknown or the decoder does not export this information,
18269 top field first will be assumed.
18272 Specify which frames to deinterlace. Accept one of the following
18277 Deinterlace all frames.
18278 @item 1, interlaced
18279 Only deinterlace frames marked as interlaced.
18282 The default value is @code{all}.
18285 @section yadif_cuda
18287 Deinterlace the input video using the @ref{yadif} algorithm, but implemented
18288 in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
18291 It accepts the following parameters:
18297 The interlacing mode to adopt. It accepts one of the following values:
18300 @item 0, send_frame
18301 Output one frame for each frame.
18302 @item 1, send_field
18303 Output one frame for each field.
18304 @item 2, send_frame_nospatial
18305 Like @code{send_frame}, but it skips the spatial interlacing check.
18306 @item 3, send_field_nospatial
18307 Like @code{send_field}, but it skips the spatial interlacing check.
18310 The default value is @code{send_frame}.
18313 The picture field parity assumed for the input interlaced video. It accepts one
18314 of the following values:
18318 Assume the top field is first.
18320 Assume the bottom field is first.
18322 Enable automatic detection of field parity.
18325 The default value is @code{auto}.
18326 If the interlacing is unknown or the decoder does not export this information,
18327 top field first will be assumed.
18330 Specify which frames to deinterlace. Accept one of the following
18335 Deinterlace all frames.
18336 @item 1, interlaced
18337 Only deinterlace frames marked as interlaced.
18340 The default value is @code{all}.
18345 Apply Zoom & Pan effect.
18347 This filter accepts the following options:
18351 Set the zoom expression. Default is 1.
18355 Set the x and y expression. Default is 0.
18358 Set the duration expression in number of frames.
18359 This sets for how many number of frames effect will last for
18360 single input image.
18363 Set the output image size, default is 'hd720'.
18366 Set the output frame rate, default is '25'.
18369 Each expression can contain the following constants:
18388 Output frame count.
18392 Last calculated 'x' and 'y' position from 'x' and 'y' expression
18393 for current input frame.
18397 'x' and 'y' of last output frame of previous input frame or 0 when there was
18398 not yet such frame (first input frame).
18401 Last calculated zoom from 'z' expression for current input frame.
18404 Last calculated zoom of last output frame of previous input frame.
18407 Number of output frames for current input frame. Calculated from 'd' expression
18408 for each input frame.
18411 number of output frames created for previous input frame
18414 Rational number: input width / input height
18417 sample aspect ratio
18420 display aspect ratio
18424 @subsection Examples
18428 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
18430 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
18434 Zoom-in up to 1.5 and pan always at center of picture:
18436 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18440 Same as above but without pausing:
18442 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
18448 Scale (resize) the input video, using the z.lib library:
18449 @url{https://github.com/sekrit-twc/zimg}. To enable compilation of this
18450 filter, you need to configure FFmpeg with @code{--enable-libzimg}.
18452 The zscale filter forces the output display aspect ratio to be the same
18453 as the input, by changing the output sample aspect ratio.
18455 If the input image format is different from the format requested by
18456 the next filter, the zscale filter will convert the input to the
18459 @subsection Options
18460 The filter accepts the following options.
18465 Set the output video dimension expression. Default value is the input
18468 If the @var{width} or @var{w} value is 0, the input width is used for
18469 the output. If the @var{height} or @var{h} value is 0, the input height
18470 is used for the output.
18472 If one and only one of the values is -n with n >= 1, the zscale filter
18473 will use a value that maintains the aspect ratio of the input image,
18474 calculated from the other specified dimension. After that it will,
18475 however, make sure that the calculated dimension is divisible by n and
18476 adjust the value if necessary.
18478 If both values are -n with n >= 1, the behavior will be identical to
18479 both values being set to 0 as previously detailed.
18481 See below for the list of accepted constants for use in the dimension
18485 Set the video size. For the syntax of this option, check the
18486 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18489 Set the dither type.
18491 Possible values are:
18496 @item error_diffusion
18502 Set the resize filter type.
18504 Possible values are:
18514 Default is bilinear.
18517 Set the color range.
18519 Possible values are:
18526 Default is same as input.
18529 Set the color primaries.
18531 Possible values are:
18541 Default is same as input.
18544 Set the transfer characteristics.
18546 Possible values are:
18560 Default is same as input.
18563 Set the colorspace matrix.
18565 Possible value are:
18576 Default is same as input.
18579 Set the input color range.
18581 Possible values are:
18588 Default is same as input.
18590 @item primariesin, pin
18591 Set the input color primaries.
18593 Possible values are:
18603 Default is same as input.
18605 @item transferin, tin
18606 Set the input transfer characteristics.
18608 Possible values are:
18619 Default is same as input.
18621 @item matrixin, min
18622 Set the input colorspace matrix.
18624 Possible value are:
18636 Set the output chroma location.
18638 Possible values are:
18649 @item chromalin, cin
18650 Set the input chroma location.
18652 Possible values are:
18664 Set the nominal peak luminance.
18667 The values of the @option{w} and @option{h} options are expressions
18668 containing the following constants:
18673 The input width and height
18677 These are the same as @var{in_w} and @var{in_h}.
18681 The output (scaled) width and height
18685 These are the same as @var{out_w} and @var{out_h}
18688 The same as @var{iw} / @var{ih}
18691 input sample aspect ratio
18694 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
18698 horizontal and vertical input chroma subsample values. For example for the
18699 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18703 horizontal and vertical output chroma subsample values. For example for the
18704 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
18710 @c man end VIDEO FILTERS
18712 @chapter OpenCL Video Filters
18713 @c man begin OPENCL VIDEO FILTERS
18715 Below is a description of the currently available OpenCL video filters.
18717 To enable compilation of these filters you need to configure FFmpeg with
18718 @code{--enable-opencl}.
18720 Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
18723 @item -init_hw_device opencl[=@var{name}][:@var{device}[,@var{key=value}...]]
18724 Initialise a new hardware device of type @var{opencl} called @var{name}, using the
18725 given device parameters.
18727 @item -filter_hw_device @var{name}
18728 Pass the hardware device called @var{name} to all filters in any filter graph.
18732 For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
18736 Example of choosing the first device on the second platform and running avgblur_opencl filter with default parameters on it.
18738 -init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
18742 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.
18744 @section avgblur_opencl
18746 Apply average blur filter.
18748 The filter accepts the following options:
18752 Set horizontal radius size.
18753 Range is @code{[1, 1024]} and default value is @code{1}.
18756 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
18759 Set vertical radius size. Range is @code{[1, 1024]} and default value is @code{0}. If zero, @code{sizeX} value will be used.
18762 @subsection Example
18766 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.
18768 -i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
18772 @section boxblur_opencl
18774 Apply a boxblur algorithm to the input video.
18776 It accepts the following parameters:
18780 @item luma_radius, lr
18781 @item luma_power, lp
18782 @item chroma_radius, cr
18783 @item chroma_power, cp
18784 @item alpha_radius, ar
18785 @item alpha_power, ap
18789 A description of the accepted options follows.
18792 @item luma_radius, lr
18793 @item chroma_radius, cr
18794 @item alpha_radius, ar
18795 Set an expression for the box radius in pixels used for blurring the
18796 corresponding input plane.
18798 The radius value must be a non-negative number, and must not be
18799 greater than the value of the expression @code{min(w,h)/2} for the
18800 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
18803 Default value for @option{luma_radius} is "2". If not specified,
18804 @option{chroma_radius} and @option{alpha_radius} default to the
18805 corresponding value set for @option{luma_radius}.
18807 The expressions can contain the following constants:
18811 The input width and height in pixels.
18815 The input chroma image width and height in pixels.
18819 The horizontal and vertical chroma subsample values. For example, for the
18820 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
18823 @item luma_power, lp
18824 @item chroma_power, cp
18825 @item alpha_power, ap
18826 Specify how many times the boxblur filter is applied to the
18827 corresponding plane.
18829 Default value for @option{luma_power} is 2. If not specified,
18830 @option{chroma_power} and @option{alpha_power} default to the
18831 corresponding value set for @option{luma_power}.
18833 A value of 0 will disable the effect.
18836 @subsection Examples
18838 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.
18842 Apply a boxblur filter with the luma, chroma, and alpha radius
18843 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.
18845 -i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
18846 -i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
18850 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.
18852 For the luma plane, a 2x2 box radius will be run once.
18854 For the chroma plane, a 4x4 box radius will be run 5 times.
18856 For the alpha plane, a 3x3 box radius will be run 7 times.
18858 -i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
18862 @section convolution_opencl
18864 Apply convolution of 3x3, 5x5, 7x7 matrix.
18866 The filter accepts the following options:
18873 Set matrix for each plane.
18874 Matrix is sequence of 9, 25 or 49 signed numbers.
18875 Default value for each plane is @code{0 0 0 0 1 0 0 0 0}.
18881 Set multiplier for calculated value for each plane.
18882 If unset or 0, it will be sum of all matrix elements.
18883 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{1.0}.
18889 Set bias for each plane. This value is added to the result of the multiplication.
18890 Useful for making the overall image brighter or darker.
18891 The option value must be a float number greater or equal to @code{0.0}. Default value is @code{0.0}.
18895 @subsection Examples
18901 -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
18907 -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
18911 Apply edge enhance:
18913 -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
18919 -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
18923 Apply laplacian edge detector which includes diagonals:
18925 -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
18931 -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
18935 @section dilation_opencl
18937 Apply dilation effect to the video.
18939 This filter replaces the pixel by the local(3x3) maximum.
18941 It accepts the following options:
18948 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18949 If @code{0}, plane will remain unchanged.
18952 Flag which specifies the pixel to refer to.
18953 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
18955 Flags to local 3x3 coordinates region centered on @code{x}:
18964 @subsection Example
18968 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.
18970 -i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
18974 @section erosion_opencl
18976 Apply erosion effect to the video.
18978 This filter replaces the pixel by the local(3x3) minimum.
18980 It accepts the following options:
18987 Limit the maximum change for each plane. Range is @code{[0, 65535]} and default value is @code{65535}.
18988 If @code{0}, plane will remain unchanged.
18991 Flag which specifies the pixel to refer to.
18992 Range is @code{[0, 255]} and default value is @code{255}, i.e. all eight pixels are used.
18994 Flags to local 3x3 coordinates region centered on @code{x}:
19003 @subsection Example
19007 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.
19009 -i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
19013 @section overlay_opencl
19015 Overlay one video on top of another.
19017 It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
19018 This filter requires same memory layout for all the inputs. So, format conversion may be needed.
19020 The filter accepts the following options:
19025 Set the x coordinate of the overlaid video on the main video.
19026 Default value is @code{0}.
19029 Set the x coordinate of the overlaid video on the main video.
19030 Default value is @code{0}.
19034 @subsection Examples
19038 Overlay an image LOGO at the top-left corner of the INPUT video. Both inputs are yuv420p format.
19040 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19043 The inputs have same memory layout for color channels , the overlay has additional alpha plane, like INPUT is yuv420p, and the LOGO is yuva420p.
19045 -i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
19050 @section prewitt_opencl
19052 Apply the Prewitt operator (@url{https://en.wikipedia.org/wiki/Prewitt_operator}) to input video stream.
19054 The filter accepts the following option:
19058 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19061 Set value which will be multiplied with filtered result.
19062 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19065 Set value which will be added to filtered result.
19066 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19069 @subsection Example
19073 Apply the Prewitt operator with scale set to 2 and delta set to 10.
19075 -i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
19079 @section roberts_opencl
19080 Apply the Roberts cross operator (@url{https://en.wikipedia.org/wiki/Roberts_cross}) to input video stream.
19082 The filter accepts the following option:
19086 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19089 Set value which will be multiplied with filtered result.
19090 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19093 Set value which will be added to filtered result.
19094 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19097 @subsection Example
19101 Apply the Roberts cross operator with scale set to 2 and delta set to 10
19103 -i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
19107 @section sobel_opencl
19109 Apply the Sobel operator (@url{https://en.wikipedia.org/wiki/Sobel_operator}) to input video stream.
19111 The filter accepts the following option:
19115 Set which planes to filter. Default value is @code{0xf}, by which all planes are processed.
19118 Set value which will be multiplied with filtered result.
19119 Range is @code{[0.0, 65535]} and default value is @code{1.0}.
19122 Set value which will be added to filtered result.
19123 Range is @code{[-65535, 65535]} and default value is @code{0.0}.
19126 @subsection Example
19130 Apply sobel operator with scale set to 2 and delta set to 10
19132 -i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
19136 @section tonemap_opencl
19138 Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
19140 It accepts the following parameters:
19144 Specify the tone-mapping operator to be used. Same as tonemap option in @ref{tonemap}.
19147 Tune the tone mapping algorithm. same as param option in @ref{tonemap}.
19150 Apply desaturation for highlights that exceed this level of brightness. The
19151 higher the parameter, the more color information will be preserved. This
19152 setting helps prevent unnaturally blown-out colors for super-highlights, by
19153 (smoothly) turning into white instead. This makes images feel more natural,
19154 at the cost of reducing information about out-of-range colors.
19156 The default value is 0.5, and the algorithm here is a little different from
19157 the cpu version tonemap currently. A setting of 0.0 disables this option.
19160 The tonemapping algorithm parameters is fine-tuned per each scene. And a threshold
19161 is used to detect whether the scene has changed or not. If the distance between
19162 the current frame average brightness and the current running average exceeds
19163 a threshold value, we would re-calculate scene average and peak brightness.
19164 The default value is 0.2.
19167 Specify the output pixel format.
19169 Currently supported formats are:
19176 Set the output color range.
19178 Possible values are:
19184 Default is same as input.
19187 Set the output color primaries.
19189 Possible values are:
19195 Default is same as input.
19198 Set the output transfer characteristics.
19200 Possible values are:
19209 Set the output colorspace matrix.
19211 Possible value are:
19217 Default is same as input.
19221 @subsection Example
19225 Convert HDR(PQ/HLG) video to bt2020-transfer-characteristic p010 format using linear operator.
19227 -i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
19231 @section unsharp_opencl
19233 Sharpen or blur the input video.
19235 It accepts the following parameters:
19238 @item luma_msize_x, lx
19239 Set the luma matrix horizontal size.
19240 Range is @code{[1, 23]} and default value is @code{5}.
19242 @item luma_msize_y, ly
19243 Set the luma matrix vertical size.
19244 Range is @code{[1, 23]} and default value is @code{5}.
19246 @item luma_amount, la
19247 Set the luma effect strength.
19248 Range is @code{[-10, 10]} and default value is @code{1.0}.
19250 Negative values will blur the input video, while positive values will
19251 sharpen it, a value of zero will disable the effect.
19253 @item chroma_msize_x, cx
19254 Set the chroma matrix horizontal size.
19255 Range is @code{[1, 23]} and default value is @code{5}.
19257 @item chroma_msize_y, cy
19258 Set the chroma matrix vertical size.
19259 Range is @code{[1, 23]} and default value is @code{5}.
19261 @item chroma_amount, ca
19262 Set the chroma effect strength.
19263 Range is @code{[-10, 10]} and default value is @code{0.0}.
19265 Negative values will blur the input video, while positive values will
19266 sharpen it, a value of zero will disable the effect.
19270 All parameters are optional and default to the equivalent of the
19271 string '5:5:1.0:5:5:0.0'.
19273 @subsection Examples
19277 Apply strong luma sharpen effect:
19279 -i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
19283 Apply a strong blur of both luma and chroma parameters:
19285 -i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
19289 @c man end OPENCL VIDEO FILTERS
19291 @chapter Video Sources
19292 @c man begin VIDEO SOURCES
19294 Below is a description of the currently available video sources.
19298 Buffer video frames, and make them available to the filter chain.
19300 This source is mainly intended for a programmatic use, in particular
19301 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
19303 It accepts the following parameters:
19308 Specify the size (width and height) of the buffered video frames. For the
19309 syntax of this option, check the
19310 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19313 The input video width.
19316 The input video height.
19319 A string representing the pixel format of the buffered video frames.
19320 It may be a number corresponding to a pixel format, or a pixel format
19324 Specify the timebase assumed by the timestamps of the buffered frames.
19327 Specify the frame rate expected for the video stream.
19329 @item pixel_aspect, sar
19330 The sample (pixel) aspect ratio of the input video.
19333 Specify the optional parameters to be used for the scale filter which
19334 is automatically inserted when an input change is detected in the
19335 input size or format.
19337 @item hw_frames_ctx
19338 When using a hardware pixel format, this should be a reference to an
19339 AVHWFramesContext describing input frames.
19344 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
19347 will instruct the source to accept video frames with size 320x240 and
19348 with format "yuv410p", assuming 1/24 as the timestamps timebase and
19349 square pixels (1:1 sample aspect ratio).
19350 Since the pixel format with name "yuv410p" corresponds to the number 6
19351 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
19352 this example corresponds to:
19354 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
19357 Alternatively, the options can be specified as a flat string, but this
19358 syntax is deprecated:
19360 @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}]
19364 Create a pattern generated by an elementary cellular automaton.
19366 The initial state of the cellular automaton can be defined through the
19367 @option{filename} and @option{pattern} options. If such options are
19368 not specified an initial state is created randomly.
19370 At each new frame a new row in the video is filled with the result of
19371 the cellular automaton next generation. The behavior when the whole
19372 frame is filled is defined by the @option{scroll} option.
19374 This source accepts the following options:
19378 Read the initial cellular automaton state, i.e. the starting row, from
19379 the specified file.
19380 In the file, each non-whitespace character is considered an alive
19381 cell, a newline will terminate the row, and further characters in the
19382 file will be ignored.
19385 Read the initial cellular automaton state, i.e. the starting row, from
19386 the specified string.
19388 Each non-whitespace character in the string is considered an alive
19389 cell, a newline will terminate the row, and further characters in the
19390 string will be ignored.
19393 Set the video rate, that is the number of frames generated per second.
19396 @item random_fill_ratio, ratio
19397 Set the random fill ratio for the initial cellular automaton row. It
19398 is a floating point number value ranging from 0 to 1, defaults to
19401 This option is ignored when a file or a pattern is specified.
19403 @item random_seed, seed
19404 Set the seed for filling randomly the initial row, must be an integer
19405 included between 0 and UINT32_MAX. If not specified, or if explicitly
19406 set to -1, the filter will try to use a good random seed on a best
19410 Set the cellular automaton rule, it is a number ranging from 0 to 255.
19411 Default value is 110.
19414 Set the size of the output video. For the syntax of this option, check the
19415 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19417 If @option{filename} or @option{pattern} is specified, the size is set
19418 by default to the width of the specified initial state row, and the
19419 height is set to @var{width} * PHI.
19421 If @option{size} is set, it must contain the width of the specified
19422 pattern string, and the specified pattern will be centered in the
19425 If a filename or a pattern string is not specified, the size value
19426 defaults to "320x518" (used for a randomly generated initial state).
19429 If set to 1, scroll the output upward when all the rows in the output
19430 have been already filled. If set to 0, the new generated row will be
19431 written over the top row just after the bottom row is filled.
19434 @item start_full, full
19435 If set to 1, completely fill the output with generated rows before
19436 outputting the first frame.
19437 This is the default behavior, for disabling set the value to 0.
19440 If set to 1, stitch the left and right row edges together.
19441 This is the default behavior, for disabling set the value to 0.
19444 @subsection Examples
19448 Read the initial state from @file{pattern}, and specify an output of
19451 cellauto=f=pattern:s=200x400
19455 Generate a random initial row with a width of 200 cells, with a fill
19458 cellauto=ratio=2/3:s=200x200
19462 Create a pattern generated by rule 18 starting by a single alive cell
19463 centered on an initial row with width 100:
19465 cellauto=p=@@:s=100x400:full=0:rule=18
19469 Specify a more elaborated initial pattern:
19471 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
19476 @anchor{coreimagesrc}
19477 @section coreimagesrc
19478 Video source generated on GPU using Apple's CoreImage API on OSX.
19480 This video source is a specialized version of the @ref{coreimage} video filter.
19481 Use a core image generator at the beginning of the applied filterchain to
19482 generate the content.
19484 The coreimagesrc video source accepts the following options:
19486 @item list_generators
19487 List all available generators along with all their respective options as well as
19488 possible minimum and maximum values along with the default values.
19490 list_generators=true
19494 Specify the size of the sourced video. For the syntax of this option, check the
19495 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19496 The default value is @code{320x240}.
19499 Specify the frame rate of the sourced video, as the number of frames
19500 generated per second. It has to be a string in the format
19501 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19502 number or a valid video frame rate abbreviation. The default value is
19506 Set the sample aspect ratio of the sourced video.
19509 Set the duration of the sourced video. See
19510 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19511 for the accepted syntax.
19513 If not specified, or the expressed duration is negative, the video is
19514 supposed to be generated forever.
19517 Additionally, all options of the @ref{coreimage} video filter are accepted.
19518 A complete filterchain can be used for further processing of the
19519 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
19520 and examples for details.
19522 @subsection Examples
19527 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
19528 given as complete and escaped command-line for Apple's standard bash shell:
19530 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
19532 This example is equivalent to the QRCode example of @ref{coreimage} without the
19533 need for a nullsrc video source.
19537 @section mandelbrot
19539 Generate a Mandelbrot set fractal, and progressively zoom towards the
19540 point specified with @var{start_x} and @var{start_y}.
19542 This source accepts the following options:
19547 Set the terminal pts value. Default value is 400.
19550 Set the terminal scale value.
19551 Must be a floating point value. Default value is 0.3.
19554 Set the inner coloring mode, that is the algorithm used to draw the
19555 Mandelbrot fractal internal region.
19557 It shall assume one of the following values:
19562 Show time until convergence.
19564 Set color based on point closest to the origin of the iterations.
19569 Default value is @var{mincol}.
19572 Set the bailout value. Default value is 10.0.
19575 Set the maximum of iterations performed by the rendering
19576 algorithm. Default value is 7189.
19579 Set outer coloring mode.
19580 It shall assume one of following values:
19582 @item iteration_count
19583 Set iteration count mode.
19584 @item normalized_iteration_count
19585 set normalized iteration count mode.
19587 Default value is @var{normalized_iteration_count}.
19590 Set frame rate, expressed as number of frames per second. Default
19594 Set frame size. For the syntax of this option, check the @ref{video size syntax,,"Video
19595 size" section in the ffmpeg-utils manual,ffmpeg-utils}. Default value is "640x480".
19598 Set the initial scale value. Default value is 3.0.
19601 Set the initial x position. Must be a floating point value between
19602 -100 and 100. Default value is -0.743643887037158704752191506114774.
19605 Set the initial y position. Must be a floating point value between
19606 -100 and 100. Default value is -0.131825904205311970493132056385139.
19611 Generate various test patterns, as generated by the MPlayer test filter.
19613 The size of the generated video is fixed, and is 256x256.
19614 This source is useful in particular for testing encoding features.
19616 This source accepts the following options:
19621 Specify the frame rate of the sourced video, as the number of frames
19622 generated per second. It has to be a string in the format
19623 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19624 number or a valid video frame rate abbreviation. The default value is
19628 Set the duration of the sourced video. See
19629 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19630 for the accepted syntax.
19632 If not specified, or the expressed duration is negative, the video is
19633 supposed to be generated forever.
19637 Set the number or the name of the test to perform. Supported tests are:
19653 Default value is "all", which will cycle through the list of all tests.
19658 mptestsrc=t=dc_luma
19661 will generate a "dc_luma" test pattern.
19663 @section frei0r_src
19665 Provide a frei0r source.
19667 To enable compilation of this filter you need to install the frei0r
19668 header and configure FFmpeg with @code{--enable-frei0r}.
19670 This source accepts the following parameters:
19675 The size of the video to generate. For the syntax of this option, check the
19676 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19679 The framerate of the generated video. It may be a string of the form
19680 @var{num}/@var{den} or a frame rate abbreviation.
19683 The name to the frei0r source to load. For more information regarding frei0r and
19684 how to set the parameters, read the @ref{frei0r} section in the video filters
19687 @item filter_params
19688 A '|'-separated list of parameters to pass to the frei0r source.
19692 For example, to generate a frei0r partik0l source with size 200x200
19693 and frame rate 10 which is overlaid on the overlay filter main input:
19695 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
19700 Generate a life pattern.
19702 This source is based on a generalization of John Conway's life game.
19704 The sourced input represents a life grid, each pixel represents a cell
19705 which can be in one of two possible states, alive or dead. Every cell
19706 interacts with its eight neighbours, which are the cells that are
19707 horizontally, vertically, or diagonally adjacent.
19709 At each interaction the grid evolves according to the adopted rule,
19710 which specifies the number of neighbor alive cells which will make a
19711 cell stay alive or born. The @option{rule} option allows one to specify
19714 This source accepts the following options:
19718 Set the file from which to read the initial grid state. In the file,
19719 each non-whitespace character is considered an alive cell, and newline
19720 is used to delimit the end of each row.
19722 If this option is not specified, the initial grid is generated
19726 Set the video rate, that is the number of frames generated per second.
19729 @item random_fill_ratio, ratio
19730 Set the random fill ratio for the initial random grid. It is a
19731 floating point number value ranging from 0 to 1, defaults to 1/PHI.
19732 It is ignored when a file is specified.
19734 @item random_seed, seed
19735 Set the seed for filling the initial random grid, must be an integer
19736 included between 0 and UINT32_MAX. If not specified, or if explicitly
19737 set to -1, the filter will try to use a good random seed on a best
19743 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
19744 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
19745 @var{NS} specifies the number of alive neighbor cells which make a
19746 live cell stay alive, and @var{NB} the number of alive neighbor cells
19747 which make a dead cell to become alive (i.e. to "born").
19748 "s" and "b" can be used in place of "S" and "B", respectively.
19750 Alternatively a rule can be specified by an 18-bits integer. The 9
19751 high order bits are used to encode the next cell state if it is alive
19752 for each number of neighbor alive cells, the low order bits specify
19753 the rule for "borning" new cells. Higher order bits encode for an
19754 higher number of neighbor cells.
19755 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
19756 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
19758 Default value is "S23/B3", which is the original Conway's game of life
19759 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
19760 cells, and will born a new cell if there are three alive cells around
19764 Set the size of the output video. For the syntax of this option, check the
19765 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19767 If @option{filename} is specified, the size is set by default to the
19768 same size of the input file. If @option{size} is set, it must contain
19769 the size specified in the input file, and the initial grid defined in
19770 that file is centered in the larger resulting area.
19772 If a filename is not specified, the size value defaults to "320x240"
19773 (used for a randomly generated initial grid).
19776 If set to 1, stitch the left and right grid edges together, and the
19777 top and bottom edges also. Defaults to 1.
19780 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
19781 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
19782 value from 0 to 255.
19785 Set the color of living (or new born) cells.
19788 Set the color of dead cells. If @option{mold} is set, this is the first color
19789 used to represent a dead cell.
19792 Set mold color, for definitely dead and moldy cells.
19794 For the syntax of these 3 color options, check the @ref{color syntax,,"Color" section in the
19795 ffmpeg-utils manual,ffmpeg-utils}.
19798 @subsection Examples
19802 Read a grid from @file{pattern}, and center it on a grid of size
19805 life=f=pattern:s=300x300
19809 Generate a random grid of size 200x200, with a fill ratio of 2/3:
19811 life=ratio=2/3:s=200x200
19815 Specify a custom rule for evolving a randomly generated grid:
19821 Full example with slow death effect (mold) using @command{ffplay}:
19823 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
19830 @anchor{haldclutsrc}
19833 @anchor{pal100bars}
19834 @anchor{rgbtestsrc}
19836 @anchor{smptehdbars}
19839 @anchor{yuvtestsrc}
19840 @section allrgb, allyuv, color, haldclutsrc, nullsrc, pal75bars, pal100bars, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
19842 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
19844 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
19846 The @code{color} source provides an uniformly colored input.
19848 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
19849 @ref{haldclut} filter.
19851 The @code{nullsrc} source returns unprocessed video frames. It is
19852 mainly useful to be employed in analysis / debugging tools, or as the
19853 source for filters which ignore the input data.
19855 The @code{pal75bars} source generates a color bars pattern, based on
19856 EBU PAL recommendations with 75% color levels.
19858 The @code{pal100bars} source generates a color bars pattern, based on
19859 EBU PAL recommendations with 100% color levels.
19861 The @code{rgbtestsrc} source generates an RGB test pattern useful for
19862 detecting RGB vs BGR issues. You should see a red, green and blue
19863 stripe from top to bottom.
19865 The @code{smptebars} source generates a color bars pattern, based on
19866 the SMPTE Engineering Guideline EG 1-1990.
19868 The @code{smptehdbars} source generates a color bars pattern, based on
19869 the SMPTE RP 219-2002.
19871 The @code{testsrc} source generates a test video pattern, showing a
19872 color pattern, a scrolling gradient and a timestamp. This is mainly
19873 intended for testing purposes.
19875 The @code{testsrc2} source is similar to testsrc, but supports more
19876 pixel formats instead of just @code{rgb24}. This allows using it as an
19877 input for other tests without requiring a format conversion.
19879 The @code{yuvtestsrc} source generates an YUV test pattern. You should
19880 see a y, cb and cr stripe from top to bottom.
19882 The sources accept the following parameters:
19887 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
19888 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
19889 pixels to be used as identity matrix for 3D lookup tables. Each component is
19890 coded on a @code{1/(N*N)} scale.
19893 Specify the color of the source, only available in the @code{color}
19894 source. For the syntax of this option, check the
19895 @ref{color syntax,,"Color" section in the ffmpeg-utils manual,ffmpeg-utils}.
19898 Specify the size of the sourced video. For the syntax of this option, check the
19899 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
19900 The default value is @code{320x240}.
19902 This option is not available with the @code{allrgb}, @code{allyuv}, and
19903 @code{haldclutsrc} filters.
19906 Specify the frame rate of the sourced video, as the number of frames
19907 generated per second. It has to be a string in the format
19908 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
19909 number or a valid video frame rate abbreviation. The default value is
19913 Set the duration of the sourced video. See
19914 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
19915 for the accepted syntax.
19917 If not specified, or the expressed duration is negative, the video is
19918 supposed to be generated forever.
19921 Set the sample aspect ratio of the sourced video.
19924 Specify the alpha (opacity) of the background, only available in the
19925 @code{testsrc2} source. The value must be between 0 (fully transparent) and
19926 255 (fully opaque, the default).
19929 Set the number of decimals to show in the timestamp, only available in the
19930 @code{testsrc} source.
19932 The displayed timestamp value will correspond to the original
19933 timestamp value multiplied by the power of 10 of the specified
19934 value. Default value is 0.
19937 @subsection Examples
19941 Generate a video with a duration of 5.3 seconds, with size
19942 176x144 and a frame rate of 10 frames per second:
19944 testsrc=duration=5.3:size=qcif:rate=10
19948 The following graph description will generate a red source
19949 with an opacity of 0.2, with size "qcif" and a frame rate of 10
19952 color=c=red@@0.2:s=qcif:r=10
19956 If the input content is to be ignored, @code{nullsrc} can be used. The
19957 following command generates noise in the luminance plane by employing
19958 the @code{geq} filter:
19960 nullsrc=s=256x256, geq=random(1)*255:128:128
19964 @subsection Commands
19966 The @code{color} source supports the following commands:
19970 Set the color of the created image. Accepts the same syntax of the
19971 corresponding @option{color} option.
19976 Generate video using an OpenCL program.
19981 OpenCL program source file.
19984 Kernel name in program.
19987 Size of frames to generate. This must be set.
19990 Pixel format to use for the generated frames. This must be set.
19993 Number of frames generated every second. Default value is '25'.
19997 For details of how the program loading works, see the @ref{program_opencl}
20004 Generate a colour ramp by setting pixel values from the position of the pixel
20005 in the output image. (Note that this will work with all pixel formats, but
20006 the generated output will not be the same.)
20008 __kernel void ramp(__write_only image2d_t dst,
20009 unsigned int index)
20011 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20014 val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
20016 write_imagef(dst, loc, val);
20021 Generate a Sierpinski carpet pattern, panning by a single pixel each frame.
20023 __kernel void sierpinski_carpet(__write_only image2d_t dst,
20024 unsigned int index)
20026 int2 loc = (int2)(get_global_id(0), get_global_id(1));
20028 float4 value = 0.0f;
20029 int x = loc.x + index;
20030 int y = loc.y + index;
20031 while (x > 0 || y > 0) {
20032 if (x % 3 == 1 && y % 3 == 1) {
20040 write_imagef(dst, loc, value);
20046 @c man end VIDEO SOURCES
20048 @chapter Video Sinks
20049 @c man begin VIDEO SINKS
20051 Below is a description of the currently available video sinks.
20053 @section buffersink
20055 Buffer video frames, and make them available to the end of the filter
20058 This sink is mainly intended for programmatic use, in particular
20059 through the interface defined in @file{libavfilter/buffersink.h}
20060 or the options system.
20062 It accepts a pointer to an AVBufferSinkContext structure, which
20063 defines the incoming buffers' formats, to be passed as the opaque
20064 parameter to @code{avfilter_init_filter} for initialization.
20068 Null video sink: do absolutely nothing with the input video. It is
20069 mainly useful as a template and for use in analysis / debugging
20072 @c man end VIDEO SINKS
20074 @chapter Multimedia Filters
20075 @c man begin MULTIMEDIA FILTERS
20077 Below is a description of the currently available multimedia filters.
20081 Convert input audio to a video output, displaying the audio bit scope.
20083 The filter accepts the following options:
20087 Set frame rate, expressed as number of frames per second. Default
20091 Specify the video size for the output. For the syntax of this option, check the
20092 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20093 Default value is @code{1024x256}.
20096 Specify list of colors separated by space or by '|' which will be used to
20097 draw channels. Unrecognized or missing colors will be replaced
20101 @section ahistogram
20103 Convert input audio to a video output, displaying the volume histogram.
20105 The filter accepts the following options:
20109 Specify how histogram is calculated.
20111 It accepts the following values:
20114 Use single histogram for all channels.
20116 Use separate histogram for each channel.
20118 Default is @code{single}.
20121 Set frame rate, expressed as number of frames per second. Default
20125 Specify the video size for the output. For the syntax of this option, check the
20126 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20127 Default value is @code{hd720}.
20132 It accepts the following values:
20143 reverse logarithmic
20145 Default is @code{log}.
20148 Set amplitude scale.
20150 It accepts the following values:
20157 Default is @code{log}.
20160 Set how much frames to accumulate in histogram.
20161 Default is 1. Setting this to -1 accumulates all frames.
20164 Set histogram ratio of window height.
20167 Set sonogram sliding.
20169 It accepts the following values:
20172 replace old rows with new ones.
20174 scroll from top to bottom.
20176 Default is @code{replace}.
20179 @section aphasemeter
20181 Measures phase of input audio, which is exported as metadata @code{lavfi.aphasemeter.phase},
20182 representing mean phase of current audio frame. A video output can also be produced and is
20183 enabled by default. The audio is passed through as first output.
20185 Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in
20186 range @code{[-1, 1]} where @code{-1} means left and right channels are completely out of phase
20187 and @code{1} means channels are in phase.
20189 The filter accepts the following options, all related to its video output:
20193 Set the output frame rate. Default value is @code{25}.
20196 Set the video size for the output. For the syntax of this option, check the
20197 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20198 Default value is @code{800x400}.
20203 Specify the red, green, blue contrast. Default values are @code{2},
20204 @code{7} and @code{1}.
20205 Allowed range is @code{[0, 255]}.
20208 Set color which will be used for drawing median phase. If color is
20209 @code{none} which is default, no median phase value will be drawn.
20212 Enable video output. Default is enabled.
20215 @section avectorscope
20217 Convert input audio to a video output, representing the audio vector
20220 The filter is used to measure the difference between channels of stereo
20221 audio stream. A monoaural signal, consisting of identical left and right
20222 signal, results in straight vertical line. Any stereo separation is visible
20223 as a deviation from this line, creating a Lissajous figure.
20224 If the straight (or deviation from it) but horizontal line appears this
20225 indicates that the left and right channels are out of phase.
20227 The filter accepts the following options:
20231 Set the vectorscope mode.
20233 Available values are:
20236 Lissajous rotated by 45 degrees.
20239 Same as above but not rotated.
20242 Shape resembling half of circle.
20245 Default value is @samp{lissajous}.
20248 Set the video size for the output. For the syntax of this option, check the
20249 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20250 Default value is @code{400x400}.
20253 Set the output frame rate. Default value is @code{25}.
20259 Specify the red, green, blue and alpha contrast. Default values are @code{40},
20260 @code{160}, @code{80} and @code{255}.
20261 Allowed range is @code{[0, 255]}.
20267 Specify the red, green, blue and alpha fade. Default values are @code{15},
20268 @code{10}, @code{5} and @code{5}.
20269 Allowed range is @code{[0, 255]}.
20272 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[0, 10]}.
20273 Values lower than @var{1} will auto adjust zoom factor to maximal possible value.
20276 Set the vectorscope drawing mode.
20278 Available values are:
20281 Draw dot for each sample.
20284 Draw line between previous and current sample.
20287 Default value is @samp{dot}.
20290 Specify amplitude scale of audio samples.
20292 Available values are:
20308 Swap left channel axis with right channel axis.
20318 Mirror only x axis.
20321 Mirror only y axis.
20329 @subsection Examples
20333 Complete example using @command{ffplay}:
20335 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
20336 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
20340 @section bench, abench
20342 Benchmark part of a filtergraph.
20344 The filter accepts the following options:
20348 Start or stop a timer.
20350 Available values are:
20353 Get the current time, set it as frame metadata (using the key
20354 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
20357 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
20358 the input frame metadata to get the time difference. Time difference, average,
20359 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
20360 @code{min}) are then printed. The timestamps are expressed in seconds.
20364 @subsection Examples
20368 Benchmark @ref{selectivecolor} filter:
20370 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
20376 Concatenate audio and video streams, joining them together one after the
20379 The filter works on segments of synchronized video and audio streams. All
20380 segments must have the same number of streams of each type, and that will
20381 also be the number of streams at output.
20383 The filter accepts the following options:
20388 Set the number of segments. Default is 2.
20391 Set the number of output video streams, that is also the number of video
20392 streams in each segment. Default is 1.
20395 Set the number of output audio streams, that is also the number of audio
20396 streams in each segment. Default is 0.
20399 Activate unsafe mode: do not fail if segments have a different format.
20403 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
20404 @var{a} audio outputs.
20406 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
20407 segment, in the same order as the outputs, then the inputs for the second
20410 Related streams do not always have exactly the same duration, for various
20411 reasons including codec frame size or sloppy authoring. For that reason,
20412 related synchronized streams (e.g. a video and its audio track) should be
20413 concatenated at once. The concat filter will use the duration of the longest
20414 stream in each segment (except the last one), and if necessary pad shorter
20415 audio streams with silence.
20417 For this filter to work correctly, all segments must start at timestamp 0.
20419 All corresponding streams must have the same parameters in all segments; the
20420 filtering system will automatically select a common pixel format for video
20421 streams, and a common sample format, sample rate and channel layout for
20422 audio streams, but other settings, such as resolution, must be converted
20423 explicitly by the user.
20425 Different frame rates are acceptable but will result in variable frame rate
20426 at output; be sure to configure the output file to handle it.
20428 @subsection Examples
20432 Concatenate an opening, an episode and an ending, all in bilingual version
20433 (video in stream 0, audio in streams 1 and 2):
20435 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
20436 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
20437 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
20438 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
20442 Concatenate two parts, handling audio and video separately, using the
20443 (a)movie sources, and adjusting the resolution:
20445 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
20446 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
20447 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
20449 Note that a desync will happen at the stitch if the audio and video streams
20450 do not have exactly the same duration in the first file.
20454 @subsection Commands
20456 This filter supports the following commands:
20459 Close the current segment and step to the next one
20462 @section drawgraph, adrawgraph
20464 Draw a graph using input video or audio metadata.
20466 It accepts the following parameters:
20470 Set 1st frame metadata key from which metadata values will be used to draw a graph.
20473 Set 1st foreground color expression.
20476 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
20479 Set 2nd foreground color expression.
20482 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
20485 Set 3rd foreground color expression.
20488 Set 4th frame metadata key from which metadata values will be used to draw a graph.
20491 Set 4th foreground color expression.
20494 Set minimal value of metadata value.
20497 Set maximal value of metadata value.
20500 Set graph background color. Default is white.
20505 Available values for mode is:
20512 Default is @code{line}.
20517 Available values for slide is:
20520 Draw new frame when right border is reached.
20523 Replace old columns with new ones.
20526 Scroll from right to left.
20529 Scroll from left to right.
20532 Draw single picture.
20535 Default is @code{frame}.
20538 Set size of graph video. For the syntax of this option, check the
20539 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20540 The default value is @code{900x256}.
20542 The foreground color expressions can use the following variables:
20545 Minimal value of metadata value.
20548 Maximal value of metadata value.
20551 Current metadata key value.
20554 The color is defined as 0xAABBGGRR.
20557 Example using metadata from @ref{signalstats} filter:
20559 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
20562 Example using metadata from @ref{ebur128} filter:
20564 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
20570 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
20571 it unchanged. By default, it logs a message at a frequency of 10Hz with the
20572 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
20573 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
20575 The filter also has a video output (see the @var{video} option) with a real
20576 time graph to observe the loudness evolution. The graphic contains the logged
20577 message mentioned above, so it is not printed anymore when this option is set,
20578 unless the verbose logging is set. The main graphing area contains the
20579 short-term loudness (3 seconds of analysis), and the gauge on the right is for
20580 the momentary loudness (400 milliseconds), but can optionally be configured
20581 to instead display short-term loudness (see @var{gauge}).
20583 The green area marks a +/- 1LU target range around the target loudness
20584 (-23LUFS by default, unless modified through @var{target}).
20586 More information about the Loudness Recommendation EBU R128 on
20587 @url{http://tech.ebu.ch/loudness}.
20589 The filter accepts the following options:
20594 Activate the video output. The audio stream is passed unchanged whether this
20595 option is set or no. The video stream will be the first output stream if
20596 activated. Default is @code{0}.
20599 Set the video size. This option is for video only. For the syntax of this
20601 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
20602 Default and minimum resolution is @code{640x480}.
20605 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
20606 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
20607 other integer value between this range is allowed.
20610 Set metadata injection. If set to @code{1}, the audio input will be segmented
20611 into 100ms output frames, each of them containing various loudness information
20612 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
20614 Default is @code{0}.
20617 Force the frame logging level.
20619 Available values are:
20622 information logging level
20624 verbose logging level
20627 By default, the logging level is set to @var{info}. If the @option{video} or
20628 the @option{metadata} options are set, it switches to @var{verbose}.
20633 Available modes can be cumulated (the option is a @code{flag} type). Possible
20637 Disable any peak mode (default).
20639 Enable sample-peak mode.
20641 Simple peak mode looking for the higher sample value. It logs a message
20642 for sample-peak (identified by @code{SPK}).
20644 Enable true-peak mode.
20646 If enabled, the peak lookup is done on an over-sampled version of the input
20647 stream for better peak accuracy. It logs a message for true-peak.
20648 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
20649 This mode requires a build with @code{libswresample}.
20653 Treat mono input files as "dual mono". If a mono file is intended for playback
20654 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
20655 If set to @code{true}, this option will compensate for this effect.
20656 Multi-channel input files are not affected by this option.
20659 Set a specific pan law to be used for the measurement of dual mono files.
20660 This parameter is optional, and has a default value of -3.01dB.
20663 Set a specific target level (in LUFS) used as relative zero in the visualization.
20664 This parameter is optional and has a default value of -23LUFS as specified
20665 by EBU R128. However, material published online may prefer a level of -16LUFS
20666 (e.g. for use with podcasts or video platforms).
20669 Set the value displayed by the gauge. Valid values are @code{momentary} and s
20670 @code{shortterm}. By default the momentary value will be used, but in certain
20671 scenarios it may be more useful to observe the short term value instead (e.g.
20675 Sets the display scale for the loudness. Valid parameters are @code{absolute}
20676 (in LUFS) or @code{relative} (LU) relative to the target. This only affects the
20677 video output, not the summary or continuous log output.
20680 @subsection Examples
20684 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
20686 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
20690 Run an analysis with @command{ffmpeg}:
20692 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
20696 @section interleave, ainterleave
20698 Temporally interleave frames from several inputs.
20700 @code{interleave} works with video inputs, @code{ainterleave} with audio.
20702 These filters read frames from several inputs and send the oldest
20703 queued frame to the output.
20705 Input streams must have well defined, monotonically increasing frame
20708 In order to submit one frame to output, these filters need to enqueue
20709 at least one frame for each input, so they cannot work in case one
20710 input is not yet terminated and will not receive incoming frames.
20712 For example consider the case when one input is a @code{select} filter
20713 which always drops input frames. The @code{interleave} filter will keep
20714 reading from that input, but it will never be able to send new frames
20715 to output until the input sends an end-of-stream signal.
20717 Also, depending on inputs synchronization, the filters will drop
20718 frames in case one input receives more frames than the other ones, and
20719 the queue is already filled.
20721 These filters accept the following options:
20725 Set the number of different inputs, it is 2 by default.
20728 @subsection Examples
20732 Interleave frames belonging to different streams using @command{ffmpeg}:
20734 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
20738 Add flickering blur effect:
20740 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
20744 @section metadata, ametadata
20746 Manipulate frame metadata.
20748 This filter accepts the following options:
20752 Set mode of operation of the filter.
20754 Can be one of the following:
20758 If both @code{value} and @code{key} is set, select frames
20759 which have such metadata. If only @code{key} is set, select
20760 every frame that has such key in metadata.
20763 Add new metadata @code{key} and @code{value}. If key is already available
20767 Modify value of already present key.
20770 If @code{value} is set, delete only keys that have such value.
20771 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
20775 Print key and its value if metadata was found. If @code{key} is not set print all
20776 metadata values available in frame.
20780 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
20783 Set metadata value which will be used. This option is mandatory for
20784 @code{modify} and @code{add} mode.
20787 Which function to use when comparing metadata value and @code{value}.
20789 Can be one of following:
20793 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
20796 Values are interpreted as strings, returns true if metadata value starts with
20797 the @code{value} option string.
20800 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
20803 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
20806 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
20809 Values are interpreted as floats, returns true if expression from option @code{expr}
20814 Set expression which is used when @code{function} is set to @code{expr}.
20815 The expression is evaluated through the eval API and can contain the following
20820 Float representation of @code{value} from metadata key.
20823 Float representation of @code{value} as supplied by user in @code{value} option.
20827 If specified in @code{print} mode, output is written to the named file. Instead of
20828 plain filename any writable url can be specified. Filename ``-'' is a shorthand
20829 for standard output. If @code{file} option is not set, output is written to the log
20830 with AV_LOG_INFO loglevel.
20834 @subsection Examples
20838 Print all metadata values for frames with key @code{lavfi.signalstats.YDIF} with values
20841 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
20844 Print silencedetect output to file @file{metadata.txt}.
20846 silencedetect,ametadata=mode=print:file=metadata.txt
20849 Direct all metadata to a pipe with file descriptor 4.
20851 metadata=mode=print:file='pipe\:4'
20855 @section perms, aperms
20857 Set read/write permissions for the output frames.
20859 These filters are mainly aimed at developers to test direct path in the
20860 following filter in the filtergraph.
20862 The filters accept the following options:
20866 Select the permissions mode.
20868 It accepts the following values:
20871 Do nothing. This is the default.
20873 Set all the output frames read-only.
20875 Set all the output frames directly writable.
20877 Make the frame read-only if writable, and writable if read-only.
20879 Set each output frame read-only or writable randomly.
20883 Set the seed for the @var{random} mode, must be an integer included between
20884 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
20885 @code{-1}, the filter will try to use a good random seed on a best effort
20889 Note: in case of auto-inserted filter between the permission filter and the
20890 following one, the permission might not be received as expected in that
20891 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
20892 perms/aperms filter can avoid this problem.
20894 @section realtime, arealtime
20896 Slow down filtering to match real time approximately.
20898 These filters will pause the filtering for a variable amount of time to
20899 match the output rate with the input timestamps.
20900 They are similar to the @option{re} option to @code{ffmpeg}.
20902 They accept the following options:
20906 Time limit for the pauses. Any pause longer than that will be considered
20907 a timestamp discontinuity and reset the timer. Default is 2 seconds.
20911 @section select, aselect
20913 Select frames to pass in output.
20915 This filter accepts the following options:
20920 Set expression, which is evaluated for each input frame.
20922 If the expression is evaluated to zero, the frame is discarded.
20924 If the evaluation result is negative or NaN, the frame is sent to the
20925 first output; otherwise it is sent to the output with index
20926 @code{ceil(val)-1}, assuming that the input index starts from 0.
20928 For example a value of @code{1.2} corresponds to the output with index
20929 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
20932 Set the number of outputs. The output to which to send the selected
20933 frame is based on the result of the evaluation. Default value is 1.
20936 The expression can contain the following constants:
20940 The (sequential) number of the filtered frame, starting from 0.
20943 The (sequential) number of the selected frame, starting from 0.
20945 @item prev_selected_n
20946 The sequential number of the last selected frame. It's NAN if undefined.
20949 The timebase of the input timestamps.
20952 The PTS (Presentation TimeStamp) of the filtered video frame,
20953 expressed in @var{TB} units. It's NAN if undefined.
20956 The PTS of the filtered video frame,
20957 expressed in seconds. It's NAN if undefined.
20960 The PTS of the previously filtered video frame. It's NAN if undefined.
20962 @item prev_selected_pts
20963 The PTS of the last previously filtered video frame. It's NAN if undefined.
20965 @item prev_selected_t
20966 The PTS of the last previously selected video frame, expressed in seconds. It's NAN if undefined.
20969 The PTS of the first video frame in the video. It's NAN if undefined.
20972 The time of the first video frame in the video. It's NAN if undefined.
20974 @item pict_type @emph{(video only)}
20975 The type of the filtered frame. It can assume one of the following
20987 @item interlace_type @emph{(video only)}
20988 The frame interlace type. It can assume one of the following values:
20991 The frame is progressive (not interlaced).
20993 The frame is top-field-first.
20995 The frame is bottom-field-first.
20998 @item consumed_sample_n @emph{(audio only)}
20999 the number of selected samples before the current frame
21001 @item samples_n @emph{(audio only)}
21002 the number of samples in the current frame
21004 @item sample_rate @emph{(audio only)}
21005 the input sample rate
21008 This is 1 if the filtered frame is a key-frame, 0 otherwise.
21011 the position in the file of the filtered frame, -1 if the information
21012 is not available (e.g. for synthetic video)
21014 @item scene @emph{(video only)}
21015 value between 0 and 1 to indicate a new scene; a low value reflects a low
21016 probability for the current frame to introduce a new scene, while a higher
21017 value means the current frame is more likely to be one (see the example below)
21019 @item concatdec_select
21020 The concat demuxer can select only part of a concat input file by setting an
21021 inpoint and an outpoint, but the output packets may not be entirely contained
21022 in the selected interval. By using this variable, it is possible to skip frames
21023 generated by the concat demuxer which are not exactly contained in the selected
21026 This works by comparing the frame pts against the @var{lavf.concat.start_time}
21027 and the @var{lavf.concat.duration} packet metadata values which are also
21028 present in the decoded frames.
21030 The @var{concatdec_select} variable is -1 if the frame pts is at least
21031 start_time and either the duration metadata is missing or the frame pts is less
21032 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
21035 That basically means that an input frame is selected if its pts is within the
21036 interval set by the concat demuxer.
21040 The default value of the select expression is "1".
21042 @subsection Examples
21046 Select all frames in input:
21051 The example above is the same as:
21063 Select only I-frames:
21065 select='eq(pict_type\,I)'
21069 Select one frame every 100:
21071 select='not(mod(n\,100))'
21075 Select only frames contained in the 10-20 time interval:
21077 select=between(t\,10\,20)
21081 Select only I-frames contained in the 10-20 time interval:
21083 select=between(t\,10\,20)*eq(pict_type\,I)
21087 Select frames with a minimum distance of 10 seconds:
21089 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
21093 Use aselect to select only audio frames with samples number > 100:
21095 aselect='gt(samples_n\,100)'
21099 Create a mosaic of the first scenes:
21101 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
21104 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
21108 Send even and odd frames to separate outputs, and compose them:
21110 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
21114 Select useful frames from an ffconcat file which is using inpoints and
21115 outpoints but where the source files are not intra frame only.
21117 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
21121 @section sendcmd, asendcmd
21123 Send commands to filters in the filtergraph.
21125 These filters read commands to be sent to other filters in the
21128 @code{sendcmd} must be inserted between two video filters,
21129 @code{asendcmd} must be inserted between two audio filters, but apart
21130 from that they act the same way.
21132 The specification of commands can be provided in the filter arguments
21133 with the @var{commands} option, or in a file specified by the
21134 @var{filename} option.
21136 These filters accept the following options:
21139 Set the commands to be read and sent to the other filters.
21141 Set the filename of the commands to be read and sent to the other
21145 @subsection Commands syntax
21147 A commands description consists of a sequence of interval
21148 specifications, comprising a list of commands to be executed when a
21149 particular event related to that interval occurs. The occurring event
21150 is typically the current frame time entering or leaving a given time
21153 An interval is specified by the following syntax:
21155 @var{START}[-@var{END}] @var{COMMANDS};
21158 The time interval is specified by the @var{START} and @var{END} times.
21159 @var{END} is optional and defaults to the maximum time.
21161 The current frame time is considered within the specified interval if
21162 it is included in the interval [@var{START}, @var{END}), that is when
21163 the time is greater or equal to @var{START} and is lesser than
21166 @var{COMMANDS} consists of a sequence of one or more command
21167 specifications, separated by ",", relating to that interval. The
21168 syntax of a command specification is given by:
21170 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
21173 @var{FLAGS} is optional and specifies the type of events relating to
21174 the time interval which enable sending the specified command, and must
21175 be a non-null sequence of identifier flags separated by "+" or "|" and
21176 enclosed between "[" and "]".
21178 The following flags are recognized:
21181 The command is sent when the current frame timestamp enters the
21182 specified interval. In other words, the command is sent when the
21183 previous frame timestamp was not in the given interval, and the
21187 The command is sent when the current frame timestamp leaves the
21188 specified interval. In other words, the command is sent when the
21189 previous frame timestamp was in the given interval, and the
21193 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
21196 @var{TARGET} specifies the target of the command, usually the name of
21197 the filter class or a specific filter instance name.
21199 @var{COMMAND} specifies the name of the command for the target filter.
21201 @var{ARG} is optional and specifies the optional list of argument for
21202 the given @var{COMMAND}.
21204 Between one interval specification and another, whitespaces, or
21205 sequences of characters starting with @code{#} until the end of line,
21206 are ignored and can be used to annotate comments.
21208 A simplified BNF description of the commands specification syntax
21211 @var{COMMAND_FLAG} ::= "enter" | "leave"
21212 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
21213 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
21214 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
21215 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
21216 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
21219 @subsection Examples
21223 Specify audio tempo change at second 4:
21225 asendcmd=c='4.0 atempo tempo 1.5',atempo
21229 Target a specific filter instance:
21231 asendcmd=c='4.0 atempo@@my tempo 1.5',atempo@@my
21235 Specify a list of drawtext and hue commands in a file.
21237 # show text in the interval 5-10
21238 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
21239 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
21241 # desaturate the image in the interval 15-20
21242 15.0-20.0 [enter] hue s 0,
21243 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
21245 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
21247 # apply an exponential saturation fade-out effect, starting from time 25
21248 25 [enter] hue s exp(25-t)
21251 A filtergraph allowing to read and process the above command list
21252 stored in a file @file{test.cmd}, can be specified with:
21254 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
21259 @section setpts, asetpts
21261 Change the PTS (presentation timestamp) of the input frames.
21263 @code{setpts} works on video frames, @code{asetpts} on audio frames.
21265 This filter accepts the following options:
21270 The expression which is evaluated for each frame to construct its timestamp.
21274 The expression is evaluated through the eval API and can contain the following
21278 @item FRAME_RATE, FR
21279 frame rate, only defined for constant frame-rate video
21282 The presentation timestamp in input
21285 The count of the input frame for video or the number of consumed samples,
21286 not including the current frame for audio, starting from 0.
21288 @item NB_CONSUMED_SAMPLES
21289 The number of consumed samples, not including the current frame (only
21292 @item NB_SAMPLES, S
21293 The number of samples in the current frame (only audio)
21295 @item SAMPLE_RATE, SR
21296 The audio sample rate.
21299 The PTS of the first frame.
21302 the time in seconds of the first frame
21305 State whether the current frame is interlaced.
21308 the time in seconds of the current frame
21311 original position in the file of the frame, or undefined if undefined
21312 for the current frame
21315 The previous input PTS.
21318 previous input time in seconds
21321 The previous output PTS.
21324 previous output time in seconds
21327 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
21331 The wallclock (RTC) time at the start of the movie in microseconds.
21334 The timebase of the input timestamps.
21338 @subsection Examples
21342 Start counting PTS from zero
21344 setpts=PTS-STARTPTS
21348 Apply fast motion effect:
21354 Apply slow motion effect:
21360 Set fixed rate of 25 frames per second:
21366 Set fixed rate 25 fps with some jitter:
21368 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
21372 Apply an offset of 10 seconds to the input PTS:
21378 Generate timestamps from a "live source" and rebase onto the current timebase:
21380 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
21384 Generate timestamps by counting samples:
21393 Force color range for the output video frame.
21395 The @code{setrange} filter marks the color range property for the
21396 output frames. It does not change the input frame, but only sets the
21397 corresponding property, which affects how the frame is treated by
21400 The filter accepts the following options:
21405 Available values are:
21409 Keep the same color range property.
21411 @item unspecified, unknown
21412 Set the color range as unspecified.
21414 @item limited, tv, mpeg
21415 Set the color range as limited.
21417 @item full, pc, jpeg
21418 Set the color range as full.
21422 @section settb, asettb
21424 Set the timebase to use for the output frames timestamps.
21425 It is mainly useful for testing timebase configuration.
21427 It accepts the following parameters:
21432 The expression which is evaluated into the output timebase.
21436 The value for @option{tb} is an arithmetic expression representing a
21437 rational. The expression can contain the constants "AVTB" (the default
21438 timebase), "intb" (the input timebase) and "sr" (the sample rate,
21439 audio only). Default value is "intb".
21441 @subsection Examples
21445 Set the timebase to 1/25:
21451 Set the timebase to 1/10:
21457 Set the timebase to 1001/1000:
21463 Set the timebase to 2*intb:
21469 Set the default timebase value:
21476 Convert input audio to a video output representing frequency spectrum
21477 logarithmically using Brown-Puckette constant Q transform algorithm with
21478 direct frequency domain coefficient calculation (but the transform itself
21479 is not really constant Q, instead the Q factor is actually variable/clamped),
21480 with musical tone scale, from E0 to D#10.
21482 The filter accepts the following options:
21486 Specify the video size for the output. It must be even. For the syntax of this option,
21487 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21488 Default value is @code{1920x1080}.
21491 Set the output frame rate. Default value is @code{25}.
21494 Set the bargraph height. It must be even. Default value is @code{-1} which
21495 computes the bargraph height automatically.
21498 Set the axis height. It must be even. Default value is @code{-1} which computes
21499 the axis height automatically.
21502 Set the sonogram height. It must be even. Default value is @code{-1} which
21503 computes the sonogram height automatically.
21506 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
21507 instead. Default value is @code{1}.
21509 @item sono_v, volume
21510 Specify the sonogram volume expression. It can contain variables:
21513 the @var{bar_v} evaluated expression
21514 @item frequency, freq, f
21515 the frequency where it is evaluated
21516 @item timeclamp, tc
21517 the value of @var{timeclamp} option
21521 @item a_weighting(f)
21522 A-weighting of equal loudness
21523 @item b_weighting(f)
21524 B-weighting of equal loudness
21525 @item c_weighting(f)
21526 C-weighting of equal loudness.
21528 Default value is @code{16}.
21530 @item bar_v, volume2
21531 Specify the bargraph volume expression. It can contain variables:
21534 the @var{sono_v} evaluated expression
21535 @item frequency, freq, f
21536 the frequency where it is evaluated
21537 @item timeclamp, tc
21538 the value of @var{timeclamp} option
21542 @item a_weighting(f)
21543 A-weighting of equal loudness
21544 @item b_weighting(f)
21545 B-weighting of equal loudness
21546 @item c_weighting(f)
21547 C-weighting of equal loudness.
21549 Default value is @code{sono_v}.
21551 @item sono_g, gamma
21552 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
21553 higher gamma makes the spectrum having more range. Default value is @code{3}.
21554 Acceptable range is @code{[1, 7]}.
21556 @item bar_g, gamma2
21557 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
21561 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
21562 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
21564 @item timeclamp, tc
21565 Specify the transform timeclamp. At low frequency, there is trade-off between
21566 accuracy in time domain and frequency domain. If timeclamp is lower,
21567 event in time domain is represented more accurately (such as fast bass drum),
21568 otherwise event in frequency domain is represented more accurately
21569 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
21572 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
21573 limits future samples by applying asymmetric windowing in time domain, useful
21574 when low latency is required. Accepted range is @code{[0, 1]}.
21577 Specify the transform base frequency. Default value is @code{20.01523126408007475},
21578 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
21581 Specify the transform end frequency. Default value is @code{20495.59681441799654},
21582 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
21585 This option is deprecated and ignored.
21588 Specify the transform length in time domain. Use this option to control accuracy
21589 trade-off between time domain and frequency domain at every frequency sample.
21590 It can contain variables:
21592 @item frequency, freq, f
21593 the frequency where it is evaluated
21594 @item timeclamp, tc
21595 the value of @var{timeclamp} option.
21597 Default value is @code{384*tc/(384+tc*f)}.
21600 Specify the transform count for every video frame. Default value is @code{6}.
21601 Acceptable range is @code{[1, 30]}.
21604 Specify the transform count for every single pixel. Default value is @code{0},
21605 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
21608 Specify font file for use with freetype to draw the axis. If not specified,
21609 use embedded font. Note that drawing with font file or embedded font is not
21610 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
21614 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
21615 The : in the pattern may be replaced by | to avoid unnecessary escaping.
21618 Specify font color expression. This is arithmetic expression that should return
21619 integer value 0xRRGGBB. It can contain variables:
21621 @item frequency, freq, f
21622 the frequency where it is evaluated
21623 @item timeclamp, tc
21624 the value of @var{timeclamp} option
21629 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
21630 @item r(x), g(x), b(x)
21631 red, green, and blue value of intensity x.
21633 Default value is @code{st(0, (midi(f)-59.5)/12);
21634 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
21635 r(1-ld(1)) + b(ld(1))}.
21638 Specify image file to draw the axis. This option override @var{fontfile} and
21639 @var{fontcolor} option.
21642 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
21643 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
21644 Default value is @code{1}.
21647 Set colorspace. The accepted values are:
21650 Unspecified (default)
21659 BT.470BG or BT.601-6 625
21662 SMPTE-170M or BT.601-6 525
21668 BT.2020 with non-constant luminance
21673 Set spectrogram color scheme. This is list of floating point values with format
21674 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
21675 The default is @code{1|0.5|0|0|0.5|1}.
21679 @subsection Examples
21683 Playing audio while showing the spectrum:
21685 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
21689 Same as above, but with frame rate 30 fps:
21691 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
21695 Playing at 1280x720:
21697 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
21701 Disable sonogram display:
21707 A1 and its harmonics: A1, A2, (near)E3, A3:
21709 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),
21710 asplit[a][out1]; [a] showcqt [out0]'
21714 Same as above, but with more accuracy in frequency domain:
21716 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),
21717 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
21723 bar_v=10:sono_v=bar_v*a_weighting(f)
21727 Custom gamma, now spectrum is linear to the amplitude.
21733 Custom tlength equation:
21735 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)))'
21739 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
21741 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
21745 Custom font using fontconfig:
21747 font='Courier New,Monospace,mono|bold'
21751 Custom frequency range with custom axis using image file:
21753 axisfile=myaxis.png:basefreq=40:endfreq=10000
21759 Convert input audio to video output representing the audio power spectrum.
21760 Audio amplitude is on Y-axis while frequency is on X-axis.
21762 The filter accepts the following options:
21766 Specify size of video. For the syntax of this option, check the
21767 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21768 Default is @code{1024x512}.
21772 This set how each frequency bin will be represented.
21774 It accepts the following values:
21780 Default is @code{bar}.
21783 Set amplitude scale.
21785 It accepts the following values:
21799 Default is @code{log}.
21802 Set frequency scale.
21804 It accepts the following values:
21813 Reverse logarithmic scale.
21815 Default is @code{lin}.
21820 It accepts the following values:
21836 Default is @code{w2048}
21839 Set windowing function.
21841 It accepts the following values:
21864 Default is @code{hanning}.
21867 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
21868 which means optimal overlap for selected window function will be picked.
21871 Set time averaging. Setting this to 0 will display current maximal peaks.
21872 Default is @code{1}, which means time averaging is disabled.
21875 Specify list of colors separated by space or by '|' which will be used to
21876 draw channel frequencies. Unrecognized or missing colors will be replaced
21880 Set channel display mode.
21882 It accepts the following values:
21887 Default is @code{combined}.
21890 Set minimum amplitude used in @code{log} amplitude scaler.
21894 @anchor{showspectrum}
21895 @section showspectrum
21897 Convert input audio to a video output, representing the audio frequency
21900 The filter accepts the following options:
21904 Specify the video size for the output. For the syntax of this option, check the
21905 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
21906 Default value is @code{640x512}.
21909 Specify how the spectrum should slide along the window.
21911 It accepts the following values:
21914 the samples start again on the left when they reach the right
21916 the samples scroll from right to left
21918 frames are only produced when the samples reach the right
21920 the samples scroll from left to right
21923 Default value is @code{replace}.
21926 Specify display mode.
21928 It accepts the following values:
21931 all channels are displayed in the same row
21933 all channels are displayed in separate rows
21936 Default value is @samp{combined}.
21939 Specify display color mode.
21941 It accepts the following values:
21944 each channel is displayed in a separate color
21946 each channel is displayed using the same color scheme
21948 each channel is displayed using the rainbow color scheme
21950 each channel is displayed using the moreland color scheme
21952 each channel is displayed using the nebulae color scheme
21954 each channel is displayed using the fire color scheme
21956 each channel is displayed using the fiery color scheme
21958 each channel is displayed using the fruit color scheme
21960 each channel is displayed using the cool color scheme
21962 each channel is displayed using the magma color scheme
21964 each channel is displayed using the green color scheme
21966 each channel is displayed using the viridis color scheme
21968 each channel is displayed using the plasma color scheme
21970 each channel is displayed using the cividis color scheme
21972 each channel is displayed using the terrain color scheme
21975 Default value is @samp{channel}.
21978 Specify scale used for calculating intensity color values.
21980 It accepts the following values:
21985 square root, default
21996 Default value is @samp{sqrt}.
21999 Set saturation modifier for displayed colors. Negative values provide
22000 alternative color scheme. @code{0} is no saturation at all.
22001 Saturation must be in [-10.0, 10.0] range.
22002 Default value is @code{1}.
22005 Set window function.
22007 It accepts the following values:
22032 Default value is @code{hann}.
22035 Set orientation of time vs frequency axis. Can be @code{vertical} or
22036 @code{horizontal}. Default is @code{vertical}.
22039 Set ratio of overlap window. Default value is @code{0}.
22040 When value is @code{1} overlap is set to recommended size for specific
22041 window function currently used.
22044 Set scale gain for calculating intensity color values.
22045 Default value is @code{1}.
22048 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
22051 Set color rotation, must be in [-1.0, 1.0] range.
22052 Default value is @code{0}.
22055 Set start frequency from which to display spectrogram. Default is @code{0}.
22058 Set stop frequency to which to display spectrogram. Default is @code{0}.
22061 Set upper frame rate limit. Default is @code{auto}, unlimited.
22064 Draw time and frequency axes and legends. Default is disabled.
22067 The usage is very similar to the showwaves filter; see the examples in that
22070 @subsection Examples
22074 Large window with logarithmic color scaling:
22076 showspectrum=s=1280x480:scale=log
22080 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
22082 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
22083 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
22087 @section showspectrumpic
22089 Convert input audio to a single video frame, representing the audio frequency
22092 The filter accepts the following options:
22096 Specify the video size for the output. For the syntax of this option, check the
22097 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22098 Default value is @code{4096x2048}.
22101 Specify display mode.
22103 It accepts the following values:
22106 all channels are displayed in the same row
22108 all channels are displayed in separate rows
22110 Default value is @samp{combined}.
22113 Specify display color mode.
22115 It accepts the following values:
22118 each channel is displayed in a separate color
22120 each channel is displayed using the same color scheme
22122 each channel is displayed using the rainbow color scheme
22124 each channel is displayed using the moreland color scheme
22126 each channel is displayed using the nebulae color scheme
22128 each channel is displayed using the fire color scheme
22130 each channel is displayed using the fiery color scheme
22132 each channel is displayed using the fruit color scheme
22134 each channel is displayed using the cool color scheme
22136 each channel is displayed using the magma color scheme
22138 each channel is displayed using the green color scheme
22140 each channel is displayed using the viridis color scheme
22142 each channel is displayed using the plasma color scheme
22144 each channel is displayed using the cividis color scheme
22146 each channel is displayed using the terrain color scheme
22148 Default value is @samp{intensity}.
22151 Specify scale used for calculating intensity color values.
22153 It accepts the following values:
22158 square root, default
22168 Default value is @samp{log}.
22171 Set saturation modifier for displayed colors. Negative values provide
22172 alternative color scheme. @code{0} is no saturation at all.
22173 Saturation must be in [-10.0, 10.0] range.
22174 Default value is @code{1}.
22177 Set window function.
22179 It accepts the following values:
22203 Default value is @code{hann}.
22206 Set orientation of time vs frequency axis. Can be @code{vertical} or
22207 @code{horizontal}. Default is @code{vertical}.
22210 Set scale gain for calculating intensity color values.
22211 Default value is @code{1}.
22214 Draw time and frequency axes and legends. Default is enabled.
22217 Set color rotation, must be in [-1.0, 1.0] range.
22218 Default value is @code{0}.
22221 Set start frequency from which to display spectrogram. Default is @code{0}.
22224 Set stop frequency to which to display spectrogram. Default is @code{0}.
22227 @subsection Examples
22231 Extract an audio spectrogram of a whole audio track
22232 in a 1024x1024 picture using @command{ffmpeg}:
22234 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
22238 @section showvolume
22240 Convert input audio volume to a video output.
22242 The filter accepts the following options:
22249 Set border width, allowed range is [0, 5]. Default is 1.
22252 Set channel width, allowed range is [80, 8192]. Default is 400.
22255 Set channel height, allowed range is [1, 900]. Default is 20.
22258 Set fade, allowed range is [0, 1]. Default is 0.95.
22261 Set volume color expression.
22263 The expression can use the following variables:
22267 Current max volume of channel in dB.
22273 Current channel number, starting from 0.
22277 If set, displays channel names. Default is enabled.
22280 If set, displays volume values. Default is enabled.
22283 Set orientation, can be horizontal: @code{h} or vertical: @code{v},
22284 default is @code{h}.
22287 Set step size, allowed range is [0, 5]. Default is 0, which means
22291 Set background opacity, allowed range is [0, 1]. Default is 0.
22294 Set metering mode, can be peak: @code{p} or rms: @code{r},
22295 default is @code{p}.
22298 Set display scale, can be linear: @code{lin} or log: @code{log},
22299 default is @code{lin}.
22303 If set to > 0., display a line for the max level
22304 in the previous seconds.
22305 default is disabled: @code{0.}
22308 The color of the max line. Use when @code{dm} option is set to > 0.
22309 default is: @code{orange}
22314 Convert input audio to a video output, representing the samples waves.
22316 The filter accepts the following options:
22320 Specify the video size for the output. For the syntax of this option, check the
22321 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22322 Default value is @code{600x240}.
22327 Available values are:
22330 Draw a point for each sample.
22333 Draw a vertical line for each sample.
22336 Draw a point for each sample and a line between them.
22339 Draw a centered vertical line for each sample.
22342 Default value is @code{point}.
22345 Set the number of samples which are printed on the same column. A
22346 larger value will decrease the frame rate. Must be a positive
22347 integer. This option can be set only if the value for @var{rate}
22348 is not explicitly specified.
22351 Set the (approximate) output frame rate. This is done by setting the
22352 option @var{n}. Default value is "25".
22354 @item split_channels
22355 Set if channels should be drawn separately or overlap. Default value is 0.
22358 Set colors separated by '|' which are going to be used for drawing of each channel.
22361 Set amplitude scale.
22363 Available values are:
22381 Set the draw mode. This is mostly useful to set for high @var{n}.
22383 Available values are:
22386 Scale pixel values for each drawn sample.
22389 Draw every sample directly.
22392 Default value is @code{scale}.
22395 @subsection Examples
22399 Output the input file audio and the corresponding video representation
22402 amovie=a.mp3,asplit[out0],showwaves[out1]
22406 Create a synthetic signal and show it with showwaves, forcing a
22407 frame rate of 30 frames per second:
22409 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
22413 @section showwavespic
22415 Convert input audio to a single video frame, representing the samples waves.
22417 The filter accepts the following options:
22421 Specify the video size for the output. For the syntax of this option, check the
22422 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
22423 Default value is @code{600x240}.
22425 @item split_channels
22426 Set if channels should be drawn separately or overlap. Default value is 0.
22429 Set colors separated by '|' which are going to be used for drawing of each channel.
22432 Set amplitude scale.
22434 Available values are:
22452 @subsection Examples
22456 Extract a channel split representation of the wave form of a whole audio track
22457 in a 1024x800 picture using @command{ffmpeg}:
22459 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
22463 @section sidedata, asidedata
22465 Delete frame side data, or select frames based on it.
22467 This filter accepts the following options:
22471 Set mode of operation of the filter.
22473 Can be one of the following:
22477 Select every frame with side data of @code{type}.
22480 Delete side data of @code{type}. If @code{type} is not set, delete all side
22486 Set side data type used with all modes. Must be set for @code{select} mode. For
22487 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
22488 in @file{libavutil/frame.h}. For example, to choose
22489 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
22493 @section spectrumsynth
22495 Sythesize audio from 2 input video spectrums, first input stream represents
22496 magnitude across time and second represents phase across time.
22497 The filter will transform from frequency domain as displayed in videos back
22498 to time domain as presented in audio output.
22500 This filter is primarily created for reversing processed @ref{showspectrum}
22501 filter outputs, but can synthesize sound from other spectrograms too.
22502 But in such case results are going to be poor if the phase data is not
22503 available, because in such cases phase data need to be recreated, usually
22504 it's just recreated from random noise.
22505 For best results use gray only output (@code{channel} color mode in
22506 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
22507 @code{lin} scale for phase video. To produce phase, for 2nd video, use
22508 @code{data} option. Inputs videos should generally use @code{fullframe}
22509 slide mode as that saves resources needed for decoding video.
22511 The filter accepts the following options:
22515 Specify sample rate of output audio, the sample rate of audio from which
22516 spectrum was generated may differ.
22519 Set number of channels represented in input video spectrums.
22522 Set scale which was used when generating magnitude input spectrum.
22523 Can be @code{lin} or @code{log}. Default is @code{log}.
22526 Set slide which was used when generating inputs spectrums.
22527 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
22528 Default is @code{fullframe}.
22531 Set window function used for resynthesis.
22534 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
22535 which means optimal overlap for selected window function will be picked.
22538 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
22539 Default is @code{vertical}.
22542 @subsection Examples
22546 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
22547 then resynthesize videos back to audio with spectrumsynth:
22549 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
22550 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
22551 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
22555 @section split, asplit
22557 Split input into several identical outputs.
22559 @code{asplit} works with audio input, @code{split} with video.
22561 The filter accepts a single parameter which specifies the number of outputs. If
22562 unspecified, it defaults to 2.
22564 @subsection Examples
22568 Create two separate outputs from the same input:
22570 [in] split [out0][out1]
22574 To create 3 or more outputs, you need to specify the number of
22577 [in] asplit=3 [out0][out1][out2]
22581 Create two separate outputs from the same input, one cropped and
22584 [in] split [splitout1][splitout2];
22585 [splitout1] crop=100:100:0:0 [cropout];
22586 [splitout2] pad=200:200:100:100 [padout];
22590 Create 5 copies of the input audio with @command{ffmpeg}:
22592 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
22598 Receive commands sent through a libzmq client, and forward them to
22599 filters in the filtergraph.
22601 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
22602 must be inserted between two video filters, @code{azmq} between two
22603 audio filters. Both are capable to send messages to any filter type.
22605 To enable these filters you need to install the libzmq library and
22606 headers and configure FFmpeg with @code{--enable-libzmq}.
22608 For more information about libzmq see:
22609 @url{http://www.zeromq.org/}
22611 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
22612 receives messages sent through a network interface defined by the
22613 @option{bind_address} (or the abbreviation "@option{b}") option.
22614 Default value of this option is @file{tcp://localhost:5555}. You may
22615 want to alter this value to your needs, but do not forget to escape any
22616 ':' signs (see @ref{filtergraph escaping}).
22618 The received message must be in the form:
22620 @var{TARGET} @var{COMMAND} [@var{ARG}]
22623 @var{TARGET} specifies the target of the command, usually the name of
22624 the filter class or a specific filter instance name. The default
22625 filter instance name uses the pattern @samp{Parsed_<filter_name>_<index>},
22626 but you can override this by using the @samp{filter_name@@id} syntax
22627 (see @ref{Filtergraph syntax}).
22629 @var{COMMAND} specifies the name of the command for the target filter.
22631 @var{ARG} is optional and specifies the optional argument list for the
22632 given @var{COMMAND}.
22634 Upon reception, the message is processed and the corresponding command
22635 is injected into the filtergraph. Depending on the result, the filter
22636 will send a reply to the client, adopting the format:
22638 @var{ERROR_CODE} @var{ERROR_REASON}
22642 @var{MESSAGE} is optional.
22644 @subsection Examples
22646 Look at @file{tools/zmqsend} for an example of a zmq client which can
22647 be used to send commands processed by these filters.
22649 Consider the following filtergraph generated by @command{ffplay}.
22650 In this example the last overlay filter has an instance name. All other
22651 filters will have default instance names.
22654 ffplay -dumpgraph 1 -f lavfi "
22655 color=s=100x100:c=red [l];
22656 color=s=100x100:c=blue [r];
22657 nullsrc=s=200x100, zmq [bg];
22658 [bg][l] overlay [bg+l];
22659 [bg+l][r] overlay@@my=x=100 "
22662 To change the color of the left side of the video, the following
22663 command can be used:
22665 echo Parsed_color_0 c yellow | tools/zmqsend
22668 To change the right side:
22670 echo Parsed_color_1 c pink | tools/zmqsend
22673 To change the position of the right side:
22675 echo overlay@@my x 150 | tools/zmqsend
22679 @c man end MULTIMEDIA FILTERS
22681 @chapter Multimedia Sources
22682 @c man begin MULTIMEDIA SOURCES
22684 Below is a description of the currently available multimedia sources.
22688 This is the same as @ref{movie} source, except it selects an audio
22694 Read audio and/or video stream(s) from a movie container.
22696 It accepts the following parameters:
22700 The name of the resource to read (not necessarily a file; it can also be a
22701 device or a stream accessed through some protocol).
22703 @item format_name, f
22704 Specifies the format assumed for the movie to read, and can be either
22705 the name of a container or an input device. If not specified, the
22706 format is guessed from @var{movie_name} or by probing.
22708 @item seek_point, sp
22709 Specifies the seek point in seconds. The frames will be output
22710 starting from this seek point. The parameter is evaluated with
22711 @code{av_strtod}, so the numerical value may be suffixed by an IS
22712 postfix. The default value is "0".
22715 Specifies the streams to read. Several streams can be specified,
22716 separated by "+". The source will then have as many outputs, in the
22717 same order. The syntax is explained in the @ref{Stream specifiers,,"Stream specifiers"
22718 section in the ffmpeg manual,ffmpeg}. Two special names, "dv" and "da" specify
22719 respectively the default (best suited) video and audio stream. Default
22720 is "dv", or "da" if the filter is called as "amovie".
22722 @item stream_index, si
22723 Specifies the index of the video stream to read. If the value is -1,
22724 the most suitable video stream will be automatically selected. The default
22725 value is "-1". Deprecated. If the filter is called "amovie", it will select
22726 audio instead of video.
22729 Specifies how many times to read the stream in sequence.
22730 If the value is 0, the stream will be looped infinitely.
22731 Default value is "1".
22733 Note that when the movie is looped the source timestamps are not
22734 changed, so it will generate non monotonically increasing timestamps.
22736 @item discontinuity
22737 Specifies the time difference between frames above which the point is
22738 considered a timestamp discontinuity which is removed by adjusting the later
22742 It allows overlaying a second video on top of the main input of
22743 a filtergraph, as shown in this graph:
22745 input -----------> deltapts0 --> overlay --> output
22748 movie --> scale--> deltapts1 -------+
22750 @subsection Examples
22754 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
22755 on top of the input labelled "in":
22757 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
22758 [in] setpts=PTS-STARTPTS [main];
22759 [main][over] overlay=16:16 [out]
22763 Read from a video4linux2 device, and overlay it on top of the input
22766 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
22767 [in] setpts=PTS-STARTPTS [main];
22768 [main][over] overlay=16:16 [out]
22772 Read the first video stream and the audio stream with id 0x81 from
22773 dvd.vob; the video is connected to the pad named "video" and the audio is
22774 connected to the pad named "audio":
22776 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
22780 @subsection Commands
22782 Both movie and amovie support the following commands:
22785 Perform seek using "av_seek_frame".
22786 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
22789 @var{stream_index}: If stream_index is -1, a default
22790 stream is selected, and @var{timestamp} is automatically converted
22791 from AV_TIME_BASE units to the stream specific time_base.
22793 @var{timestamp}: Timestamp in AVStream.time_base units
22794 or, if no stream is specified, in AV_TIME_BASE units.
22796 @var{flags}: Flags which select direction and seeking mode.
22800 Get movie duration in AV_TIME_BASE units.
22804 @c man end MULTIMEDIA SOURCES